(Debating Ethics) Benatar, David - Wasserman, David-Debating Procreation - Is It Wrong To Reproduce - Oxford University Press (2015)

DEBATING PROCREATION
Debating Ethics
General Editor
Christopher Heath Wellman
Washington University of St. Louis
Debating Ethics is a series of volumes in which leading
scholars defend opposing views on timely ethical questions
and core theoretical issues in contemporary moral,
political, and legal philosophy.
Debating the Ethics of Immigration
Is There a Right to Exclude?
Christopher Heath Wellman and Philip Cole
Debating Brain Drain
May Governments Restrict Emigration?
Gillian Brock and Michael Blake
Debating Procreation
Is It Wrong to Reproduce?
David Benatar and David Wasserman
Debating
Procreation
Is It Wrong to Reproduce?
D AV I D B E N ATA R
DAV I D WA S S E R M A N
1
Oxford University Press is a department of the University of Oxford.
It furthers the Universitys objective of excellence in research, scholarship,
and education by publishing worldwide.
Oxford New York
Auckland Cape Town Dar es Salaam Hong Kong Karachi
Kuala Lumpur Madrid Melbourne Mexico City Nairobi
New Delhi Shanghai Taipei Toronto
With offices in
Argentina Austria Brazil Chile Czech Republic France Greece
Guatemala Hungary Italy Japan Poland Portugal Singapore
South Korea Switzerland Thailand Turkey Ukraine Vietnam
Oxford is a registered trade mark of Oxford University Press
in the UK and certain other countries.
Published in the United States of America by
Oxford University Press
198 Madison Avenue, New York, NY 10016
Oxford University Press 2015

All rights reserved. No part of this publication may be reproduced,
stored in a retrieval system, or transmitted, in any form or by any means,
without the prior permission in writing of Oxford University Press,
or as expressly permitted by law, by license, or under terms agreed with the
appropriate reproduction rights organization. Inquiries concerning
reproduction outside the scope of the above should be sent to the Rights
Department, Oxford University Press, at the address above.
You must not circulate this work in any other form
and you must impose this same condition on any acquirer.
Library of Congress Cataloging-in-Publication Data
Debating procreation : is it wrong to reproduce? / David Benatar
and David Wasserman.
p. cm. (Debating ethics)
Includes bibliographical references.
ISBN 9780199333554 (pbk. : alk. paper) ISBN 9780199333547
(hardcover : alk. paper) 1. Human reproductionMoral and ethical aspects.
2. LifeMoral and ethical aspects. I. Benatar, David, author.
II. Wasserman, David, author.
BJ1335.D42 2015
176dc23
2014042303
135798642
Printed in the United States of America on acid-free paper
CONTENTS
Introduction
By David Benatar and David Wasserman
PART I ANTI-NATALISM
By David Benatar
1.
2.
3.
4.
5.
Introducing Anti-Natalism
The Asymmetry Argument
The Quality-of-Life Argument
The Misanthropic Argument
Contra Procreation
11
18
40
78
122
PART II PRO-NATALISM
By David Wasserman
6.
7.
Better to Have Lived and Lost?

Against Anti-Natalism
135
148
vi |
C ontents
8. The Good of the Future Child and the

Parent-Child Relationship as Goals
of Procreation
9.
Impersonal Constraints on Procreation
10. Alternatives to Impersonal Approaches:
Birthrights and Role-Based Duties
182
209
Index
265
228
Introduction
DAV ID BE NATA R A ND DAV ID WA S SER M A N
READERS OF A BOOK DEBATING the ethics of procreation
would be excused for thinking that the book might be premised on the assumption that procreation is sometimes
permissible and would then discuss the conditions under
which it is and is not morally acceptable. In fact, we shall
be debating the more basic questionwhether procreation
is ever morally justifiable.
On the face of it, there are only two responses to this question: no and yes. Those, in broad terms, are the views that
we, respectively, defend. David Benatar argues that procreation is never morally permissible, while David Wasserman
argues that procreation is sometimes morally permissible
and that there can be positive value in creating children.
However, as one should expect, there are numerous
views on the ethics of procreation. For example, the view
that procreation is sometimes permissible is actually a
2 |
D E B A T I N G P R O C R E A T I O N
range of views along the spectrum covered by the vagueness of the word sometimes. To say that it is sometimes
morally permissible to have children leaves open the question of how often it is permissible. Some views are more
permissive; others less so. David Wasserman will examine
some of these views and defend one account of when it is
permissible to create children.
The respective positions we take are not symmetrical. David Wasserman does not offer a categorical defense
of procreation that mirrors David Benatars categorical
attack. Although David Wassermans account is more permissive than most mainstream views, he rejects the view
that procreation can never be faulted on the grounds of
harming or wronging the children brought into existence.
Like David Benatar, he holds that procreation can harm
and wrong the child created, as well as wronging other
individuals.
There is another asymmetry between our two positions. The overwhelming majority of people think that procreation is generally morally acceptable and many of those
are outraged at any suggestion to the contrary. Thus David
Wasserman defends a general position that enjoys widespread support while David Benatar attacks a very fecund
holy cow. The latter author, therefore, has the harder task
of defending a heresy. Shifting metaphors, we are two
Davids, but only one of us is attacking a Goliath. Thus, at
least in terms of persuading people, David Benatar bears
the burden of proof, although he argues that, given the
harms of procreating, it is the defender of procreation who
bears the moral burden of proof.
Although the specific view that David Wasserman defends is not heretical, it is, at the very least, unorthodox.
I ntroduction
| 3
Unlike the majority of pro-natalists, he rejects the idea

that that it is even pro tanto wrong to have a less happy
child when one could have a happier one insteador (what
is not the same thing) a disabled one when one could have
a non-disabled one instead. While agreeing that it is problematic to select a less happy or a disabled child, he argues
that it is also problematic for prospective parents to select
against such a childoptions that prospective parents
increasingly have. He also argues that the intentions or
motivation with which they have a child can affect the permissibility of their actions, an equally unorthodox view.
And he argues, here against David Benatar and the mainstream, that there is a perfectly intelligible sense in which
prospective parents can create a child for reasons that concern the good of that child; indeed, that they should only
procreate with such an intention.
Both of us have written previously about procreation
and the reader may want to know how what we say here differs from what we have said before. In Part I, the axiological argument will be familiar to those who have read David
Benatars Better Never to Have Been. However, whereas the
argument was presented in fuller form there, only an overview of it is presented here.1 This is in order to cater to the
broader readership for which the current book is intended.
It is nonetheless the most technical of the arguments in
Part I. Readers daunted by more technical arguments can
skip to the subsequent arguments in Part I.
The core of the quality-of-life argument is also drawn
from Better Never to Have Been, but it does not cover exactly the same ground. Some details of the earlier argument have been omitted, but in other ways the argument
has been expanded here. For example, the risk argument,
4 |
which is a version of the quality-of-life argument, is discussed here in more detail than it was before.
The misanthropic argument did not feature at all in
Better Never to Have Been, where, because the focus was
on the interests of the person brought into existence, all
the arguments were philanthropic. The misanthropic argument does not obviously show that it is better never to
have been,2 but it does support the anti-natalist conclusion
that it is better not to procreate.
The concluding chapter of Part I surveys a number of
considerations that might be advanced in favor of procreation, and argues that these are not sufficient to outweigh
the anti-natalist arguments advanced in the previous
chapters.
In summary, then, the arguments in Part I constitute a clear and accessible statement of the arguments for
anti-natalism.
In Part II, David Wasserman offers his qualified defense of procreation. He begins with a critique of several
anti-natalist arguments. He focuses on David Benatar,
but also considers two other writers: Seana Shiffrin, who
argues that procreation is morally problematic in imposing
unconsented harm; and Matti Hayry, who maintains that
every child faces the realistic possibility, however slight,
of an awful lifean unacceptable risk that makes procreation wrongful. In each case, David Wasserman rejects the
categorical claim, arguing that bearing children can be, and
often is, permissible, despite the certainty of significant
harm and the possibility of an awful life. He does not address David Benatars axiological argument in detail, which
claims that procreation always harms the person created.
That more technical argument has been subject to a great
I ntroduction
| 5
deal of technical criticism. David Wasserman contends that

procreation need not wrong the child, even if it harms him
in the sense that David Benatar claims. His principal targets are David Benatars quality-of-life and misanthropic
arguments. He maintains that the former rests on unduly
pessimistic assessments and inappropriately perfectionist
standards, while the latter makes moral demands on prospective parents that they can rightly reject.
He then argues that prospective parents should have
good reasons for creating a child, given the hardships and
risks of even the most secure and comfortable life. He contends that they can and should create children in part for
reasons that concern the good of those children. Those reasons concern the good of a particular kind of lifea life
with them and their family, which they want to share with
a now-unknown being. It is these reasons that can and
often do motivate procreation, not the barely intelligible
desire to rescue a possible child from the limbo of nonexistence. He argues that these reasons do not treat existence
as a good that is bestowed on a child, and that they can
motivate both prospective biological and adoptive parents.
He maintains that these reasons have important similarities to those with which people seek intimate relationships
with other now-unknown adults, reasons that concern the
good of both parties to the desired relationship.
David Wasserman then moves from justifying the decision to have a child to justifying standards for permissible procreation. He reviews two general approaches to
setting standards. The first holds that procreative decision making is a domain governed by impersonal considerations that dictate the choice of a child free of serious
suffering or limitation, the best-off child, the child who
6 |
will make the world best off, or some combination of these

desiderata. The second approach holds that procreative
decision making is governed by the rights of children or
the duties of prospective parents: the rights of children to
have been created with the expectation that they would
lead lives satisfying a minimal standard of well-being; the
duties of prospective parents to have children with certain
expectations, attitudes, or commitments towards them.
David Wasserman rejects the first approach as uncertain in scope, unjustified in imposing special impersonal
duties on prospective parents, and counterintuitive in entailing strong moral reasons to procreate, an implication
at odds with the deeply held conviction that procreation is
morally optional. He defends the second approach in general terms, arguing that standards for procreation should
be based on the rights of future children and the duties
of prospective parents. He outlines several proposed standards for permissible procreation, arguing for the most
permissive one, which does not require the prospective
parents to choose the child that is expected to have the
best or a better life, and does not require them to avoid
having a child that is expected to have a merely adequate
one. That standard, however, is demanding as well as permissive. It rests on a view of the moral posture prospective
parents should adopt towards their future children, one
that discourages any kind of selectivity.
David Wasserman concludes the chapter by acknowledging that his defense of procreation has not only been
qualified, but piecemeal, and he defends its piecemeal
character.
While we disagree about whether procreation is ever
morally permissible, we are in agreement that not all
I ntroduction
| 7
procreation is morally acceptable. We reject the view, usually implicit, that might be expressed in the slogan Have
genitals, may procreate.3 The decision to bring a child into
existence is a matter of much greater moral import than
its frequency would suggest. Potential parents should be
thinking very carefully about whether to have a child. The
conflicting views presented in this book should be grist for
that reproductive decision mill.4
NOTE S
1. There are also some differences in the way it is presented.
For example, in the previous statement of the argument
David Benatar focused on pains and pleasures. Although he
was explicit that these were exemplars of harms and benefits more generally, many readers insisted on interpreting
the argument as a hedonistic one. In a bid to forestall this
misinterpretation, the axiological asymmetry is now presented in terms of harms and benefits rather than in terms
of hedonic exemplars of harms and benefits.
2. We say obviously because it might be argued that having
a defective nature and being the cause of harm actually
harms oneself.
3. This is actually an understatement. Many pro-natalists
think that even those without (functional) genitals may
procreateindeed, that they are entitled to assistance in
doing so.
4. We are grateful to Rivka Weinberg for her helpful comments
and to our editors at Oxford University Press.
PART I
ANTI-NATALISM
DAV I D B E NATA R
Introducing Anti-Natalism
IT IS ALWAYS DIFFICULT TO convince people that a wide-
spread practice in which they participate is morally wrong.

This is because people have difficulty believing that they
and so many others could be acting immorally. The task is
made still more difficult when the practice is one that is fed
by powerful biological drives with deep evolutionary roots.
It is thus unsurprising how challenging it is to try to convince people that procreation is wrong. The overwhelming
majority of humans produce children sometime during the
course of their lives,1 and the desire to procreate is among
the most powerful.
Although most people want to and do procreate, they
do not always want to have children as early and often as
they do have them. Billions of people have been brought
into existence unintentionally.2 They were mistakes. Very
often, therefore, people procreate without thinking. Their
children are brought into existence not as a result of a decision to procreate, but instead as a result of a decisionor,
12 |
more commonly, a mere impulseto have sex.3 A small

proportion of those children who are unintended are also
actively unwanted and are either abandoned or given over
to institutional care or adoptive parents.4 There are millions of such children.5
Many of those procreators who do think before reproducing are willing to acknowledge that unintentional procreation is undesirable and that there should be less of it.
Most people are even willing to say that there should be
less procreation of the intentional kind too. They recognize
that many people prefer to have more children than they
should have. I concur with these views but I shall argue
that they do not go far enough. More specifically, I shall
argue that having even one child, whether intentionally or
not, is having too many. In other words, while many arguments against procreation conclude that only some procreation is wrong, I shall argue for the conclusion that all
procreation is wrong.6
Procreation may seem like an innocuous activity, but it
is in fact deeply harmful. In creating a child one creates the
basic condition for all the terrible things that will or could
befall it. One creates the vulnerability to all (other) harms,
from the mildest to the most unspeakable horrors. While
not everybody suffers the worst fates, nobody escapes serious harm entirely, as I shall show later. The surest way
to prevent the awful things that will happen to ones child
is not to have that child. Yet prospective parents blithely
create new beings that only the naively optimistic could
think would escape serious harm.
Viewed in these terms, it is truly astounding not only
that procreation is not criticized, but that it is widely
lauded. Parents are congratulated on the birth of a child.
I ntroducing A nti - N atalism
| 1 3
The parents themselves often seem hugely impressed with

themselves for having produced offspring, even though
this achievement is something that they have in common
with billions of humans and other animals with functional
genitalia and reproductive systems.
These attitudes are part of a deep and widespread pronatalisman encouragement or at least endorsement of
procreation. Those opposed to procreationanti-natalists
have to confront this powerful force, which has biological,
cultural, social, religious, and legal manifestations. Anti-
natalism, like pro-natalism, can vary in its scope. Thus somebody could be an anti-natalist in the sense of advocating a
reduction but not complete cessation of procreation.7 This,
however, is not the sense in which I shall be using the term.
Instead I shall use it to refer to the more extreme position
that opposes all procreation. Anti-natalism, in this sense, is
the position I shall defend. Not every argument I shall advance will, by itself, yield this radical conclusion. However,
other arguments will lead to this conclusion, and thus the set
of arguments together also does so.
Anti-natalism (in this sense) implies that it would be
better if there were no more humans. The further implication of this is that it would be better if humans became
extinct, at least if extinction were brought about by not
creating new members of the species. Many people have
difficulty accepting this implication. They find the prospect
of a world without humans to be tragic. This view arises
from a misguided sentimentality about the human species.
The demise of humanity would have some serious costs for
the final people, but, as I shall argue later, the state of there
being no more humans is not something we should resist
or mourn. The world will someday be devoid of humans.
14 |
This outcome is certain. The uncertainty concerns when

this will happen. We do not know humanitys expiry date,
but the earlier it is, the more suffering will be avoided.8
This does not imply either suicide or speciecide, both of
which involve taking lives. Taking lives has important costs
that are not incurred when people desist from procreation.
Killing (other) people who do not wish to die violates their
rights and thus incurs a serious moral cost. Killing people
for whom death is not (yet) in their interests harms them,
usually very severely. By contrast, failing to bring somebody into existence neither violates the rights of nor harms
the merely possible people who never become actual.
Anti-natalism, again like pro-natalism, can be applied
to different spheres. Especially important is the distinction between anti-natalism as a view about the ethics of
procreation and anti-natalism as a view about the ethics
of regulating procreation. I shall argue that procreation is
morally wrong. This is distinct from and does not imply the
claim that we may prevent humanity from procreating. The
absence of a moral right to procreate does not imply that
there should be no legal right to procreate.9
There are lots of good reasons why an anti-natalist
view about the ethics of procreation does not commit
one to thinking that we may prevent people from reproducing. One such reason is that it is very likely to do
more harm than good. Efforts to prevent people from
reproducing would be met with stern opposition from
those wanting to reproduce. Severe invasions of privacy
and the use of force would quickly become necessary
to enforce a prohibition or prevention of procreation,
and these would likely be resisted, very probably even
violently. Thus there would be serious costs without the
| 1 5
likelihood of success. Anti-natalists are interested in reducing harm, not increasing it.
The sad truth is that the human species is not voluntarily going to cease reproducing, and any attempt by a
minority to prevent the rest from procreating is unlikely
to work. That does not mean that individual humans will
not desist from procreation. Some of them will desist as a
result of considering arguments for anti-natalism. Every
decision not to procreate is a decision to spare a potential
person from serious harm and is thus to be welcomed.
In the chapter that follows this introduction, I shall
advance what I shall call the axiological asymmetry argument for the conclusion that coming into existence is
always a harm. This argument creates a presumption against
procreation, but it does not by itself generate anti-natalism.
Harming somebody requires justification and if the harm of
coming into existence were a relatively minor one then the
justificatory burden might not be difficult to meet.
The quality-of-life argument in the following chapter argues that coming into existence is a serious harm. It
is so serious a harm that it is unlikely that inflicting this
harm could be justified. That chapter includes a fallback
argumentthat even if the quality of life is not always
terrible, the risks of terrible things happening are so high
that it is morally wrong to impose those risks on children
by creating them.
Both the axiological asymmetry argument and the
quality-of-life argument are what I call philanthropic arguments. They arise from concern for the beings that will
be brought into existence. Having presented those arguments, I shall turn to the misanthropic argument for
anti-natalism. This argument arises from concern for the
16 |
victims of those humans that would be brought into existence. Humanity is highly destructive and in almost all
cases, as I shall show, creating a new member of the species
contributes to that destruction.
In the final chapter of my section of this book, I consider various arguments in favor of procreation. These are
arguments that claim we have a reason or even a duty to
create children. Some of these arguments are stronger
than others, but even the most compelling of them is insufficient to rescue procreation from the challenges of
anti-natalism.
NOTE S
1. The global median of childlessness among women is 4.5%.
(World Fertility Report 2009, New York: United Nations,
2011, 59.) Some of these women would like to have had children but were or became infertile. Others are intentionally
childless.
2. In recent years around a quarter of births were unintended:
Of the estimated 208 million pregnancies that occurred
worldwide in 2008, 102 million resulted in intended births
(49%), 41 million ended in induced abortions (20%), 33
million in unintended births (16%), and about 31 million
in miscarriages (15%)some from unintended and some
from intended pregnancies . . . S. Singh, et al. Abortion
Worldwide: A Decade of Uneven Progress (New York: Guttmacher Institute, 2009), 39.
3. At least by one of the parties because some pregnancies
result from rape.
4. According to a recent report the median number of children
in the care of institutions and family foster care is 492.9 per
100,000 children (under the age of 18). Only 8.3% of these
have been adopted. See Child Adoption: Trends and Policies
(New York: United Nations, 2009), 122.
| 1 7
5. It is unlikely that this exhausts the number of unwanted

children. Many parents may care for children that they do
not want.
6. Or almost all procreation. In Better Never to Have Been
(Oxford: Oxford University Press, 2006), 182193, I considered, although did not specifically endorse, the possibility
that procreation might be permissible in some limited circumstances as part of a program of phasing out humanity.
7. Pro- and anti-natalism are positions on a spectrum, the
midpoint of which is what we might call neutro-natalism,
a position of indifference towards procreation.
8. It is true that the sooner humanity ends the less human
pleasure there will be. However, for reasons I shall explain,
we should be more interested in there being less suffering.
9. See Benatar, Better Never to Have Been, 102113.
The Asymmetry Argument
MANY PEOPLE THINK THAT COMING into existence is a
harm only if the person brought into existence has a life

in which there is more bad than good. In the next chapter I shall argue that, contrary to what most people believe, the bad (almost) always outweighs the good. In the
current chapter, however, I shall argue that even if there
were more good than bad, the presence of any bad would
be sufficient for coming into existence to be a harm. Because every life includes some bad, coming into existence
is always a harm.
THE NON-IDENTITY PROBLE M

Before I advance the arguments for this conclusion, a prior
issue needs to be examined. I need to respond to the notorious problem that is variously known as the non-identity
problem1 or the paradox of future individuals.2
T he A symmetry A rgument
| 1 9
The problem arises in those circumstances in which a

procreative choice seems contrary to the interests of the
future person even though that persons life would still be
worth living.3 Consider, for example, a case in which one
is faced with the choice of either conceiving a child that
would suffer from a severe genetic abnormality (although
not so severe an abnormality as to make its life not worth
living) or conceiving a different child that has no such abnormality. Most people think that creating the child with
the genetic defect would be wrongand that it would be
wrong because of the harm it does to the child.
However, it has been argued that there is a logical obstacle to making this claim. If the childs life is worth living
then it is, by definition, not worse than nonexistenceand
if one is not made worse off one cannot be harmed. Accordingly, we cannot say that the child would be harmed by
being brought into existence.
This is a problem in need of resolution, and various solutions have been proposed. One possibility is to deny that
the assumed conception of harm is the only possible one.
One could acknowledge that in non-procreative contexts
to harm is typically to make worse off, but that in procreative contexts a different conception is required. According
to this view, a person is harmed if (i) he suffers a condition
that is bad for him, and (ii) the alternative would not have
been bad. Troubling non-identity cases, it might be said,
meet both of these criteria: the person brought into existence suffers a condition that is bad, and the alternative
never existingwould not have been bad.
In response to this, it might be objected that this solution would work only if the persons life is not worth
living. On this view, it is not bad for somebody to come
20 |
into existence unless his life is not worth living. We should

reject this view. A life can be very bad without being so bad
that it is not worth living. A life need only be very bad to
meet condition (i). It does not need to be so bad that the life
is not worth living.
Another response to the non-identity problem is that
it arises, or at least gains strength, because of a failure to
recognize a crucial ambiguity in the phrase a life worth
living. This phrase can mean either a life worth starting
or a life worth continuing. Glossing over that ambiguity
would not be problematic if the standards for determining
when a life is worth starting and the standards for determining when a life is worth continuing were the same. In
fact, however, there is good reason to think that different
standards should apply. More specifically, the quality of
a life must be better to warrant the judgment that a life
is worth starting than it must be in order to warrant the
judgment that it is worth continuing. There is good reason
for this. When there is an existing person the bad things in
life need to be sufficiently bad to defeat that persons interest in continuing to exist. By contrast, when one is considering whether to bring somebody into existence, that
possible person has no interest in coming into existence
and thus there is no interest that needs to be defeated.
We can now see how ignoring that distinction leads to
the non-identity problem. One premise of the argument
that generates the problem is that if a life is worth living
then it is, by definition, not worse than existence. This
claim is false, however, if it means that if a life is worth
continuing then it is, by definition, not worse than never
coming into existence. It is perfectly intelligible to say
that although coming into existence was a harm because
| 2 1
the life was not good enough to be worth starting, it is

nonetheless not so bad that it is not worth continuing.
Now, it might be suggested that there are non-identity
cases in which the defect is sufficiently minor that the
life was worth starting and that it must be said of such
lives that they are not worse than never existing. However, any defect that was sufficiently minor to pass the
life worth starting test would also be minor enough for
us to be able to embrace the conclusion that no harm was
done to the child who was brought into existence with
that defect.
Of course, I think that there are no such casesbecause
I think that no lives are worth starting. Those who think
that at least some lives are worth starting need to decide
where they set the bar and then accept the implications.
THE AXIOLOG ICAL ASYM METRY

The conclusion that coming into existence is always a
harmastounding to many peoplefollows from an axiological asymmetry between harms and benefits.
Consider two scenarios, one in which a person, X,
exists, and one in which X never exists. To determine which
of these two scenariosexisting or never existingis
better for X, we must compare those two scenarios. However, there are two ways of comparing them and only one of
these is the relevant way. The wrong way to compare them
is by asking which of the scenarios is impersonally better.
While the impersonal comparison does compare the two
scenarios and does ask which is better, it does not ask
which one is better for X. An impersonal evaluation makes
22 |
no reference to the interests of the person who either is

or is not brought into existence. This is clearly the wrong
kind of comparison to make if we are seeking to determine
whether it is coming into existence or never coming into
existence that is best for X.
Instead we need to make the comparison with reference to the interests of X, because we want to know whether
it is better for X to come into existence or never to come
into existence. Some people have difficulty making sense
of such a comparison. This is because X exists in only one
of the two scenarios and thus, it is said, has interests only
in that scenario. We cannot compare those interests with
Xs interests in the alternative scenario because X does not
exist in that alternative scenario.
This concern seems to take the task of comparison too
literally. It is obviously the case in the scenario in which X
does not exist that there is no person and thus no persons
interests. However, that does not prevent us comparing
that possible world with another possible world in which
X does exist, and it does not stop us comparing the value
of those two worlds with reference to the interests of the
person who exists in one but only one of them.4
Harm befalls only those who come into existence. That
is the obvious disadvantage of coming into existence. Optimists will be quick to note that it is equally true that benefits also accrue only to those who come into existence. They
argue, therefore, that coming into existence, like never
coming into existence, has both advantages and disadvantages. However, although it is good for those who exist to
enjoy benefits, those benefits are not a net advantage over
never existing. This is because of a crucial difference between harms and benefits.5
| 2 3
Whereas:
1. The presence of harm is bad; and
2. the presence of benefit is good,
an asymmetrical evaluation applies to the absence of harm
and benefit:
3. The absence of harm is good, even if that good is
not enjoyed by anyone;
but
4. the absence of benefit is not bad unless there is
somebody for whom this absence is a deprivation.
This asymmetry can be represented in the following
diagram:

Scenario A
(X exists)
Scenario B
(X never exists)
(1)
(3)
Presence of Harm
Absence of Harm

(Bad)
(Good)
(2)
(4)
Presence of Benefit
Absence of Benefit
(Good)
figure 2.1 The Axiological Asymmetry
(Not Bad)
24 |
Some people think that the way to determine whether

coming into existence is a harm is by comparing (1) and
(2). On this view, coming into existence is a harm only if (1)
outweighs (2). However, this is not a comparison between
coming into existence and never coming into existence and
thus I cannot see how the difference between (1) and (2)
can determine whether Scenario A is better or worse than
Scenario B.
To decide the relative advantages and disadvantages of
coming into existence and of never coming into existence,
we need to compare (1) with (3), and (2) with (4). When we
make the first comparison we find that never existing is
preferable to coming into existence. The absence of harm
in Scenario B is an advantage over the presence of harm in
Scenario A. However, when we compare (2) with (4) we see
that the presence of benefit in Scenario A, although good
for X, is not an advantage over the absence of benefit in
Scenario B. In other words, Scenario B has an advantage
over Scenario A, but Scenario A has no advantage over Scenario B. We see then that the axiological asymmetry leads
to the conclusion that coming into existence is always a
net harm.6
The axiological asymmetry is widely acceptedthat is,
until people see where it leads. Then many (but not all) of
them seek to deny it. Because people are not inclined to
dispute (1) and (2), those who wish to deny the asymmetry
are determined to find some way of rejecting either one or
both of (3) and (4).
It is difficult to prove definitively that we must accept
the axiological asymmetry. However, there is a constellation of interconnecting reasons why we should accept it.
One does not have to think that any of these reasons by
| 2 5
itself provides insurmountable evidence for the axiological

asymmetry in order to think that collectively they provide
good reason to accept it.
T H E E X P L A N AT O R Y VA LU E
OF THE AXIOLOG ICAL ASYM METRY
First, the axiological asymmetry is the best explanation
for a number of other asymmetries:
1. The asymmetry of procreational (reasons and)
duties: While we have a duty to avoid bringing into existence people who would lead miserable lives, we have no
duty to bring into existence those who would lead happy
lives.
Not everybody accepts this asymmetry. Some religious
people believe that we have a duty to be fruitful and multiply,7 and some positive utilitarians believe that we have
a duty to create new people if that will increase positive
utility.8 However, the vast majority of philosophers who
have thought about the ethics of creating new people have
accepted the asymmetry of procreational duties.9
The axiological asymmetry explains this deontic asymmetry. We have a duty to avoid creating miserable lives (partly)
because the presence of that misery would be bad, but we have
no duty to create (purportedly) happy lives because although
that happiness would be good, its absence is not bad.
2. The prospective beneficence asymmetry: It is
strange to cite as a reason for having a child the fact that
26 |
the child will thereby be benefited, whereas it is not similarly strange to cite as a reason for not having a child that
that child will suffer. In other words, it is odd to have a
child for its own sake, but it is not in the least odd to desist
from having a child out of concern for the interests of the
child that would exist if one procreated.
This claim is not a logical or metaphysical one. It is not the
claim that it is logically or metaphysically incoherent to
take the same view about the benefit and the harm of potential people. Instead the claim is that it is axiologically
odd, and this is explained by the axiological asymmetry:
the absent benefits of possible people who never become
actual people are not bad, whereas the absent harms of
such people are good.
3. The retrospective beneficence asymmetry: One can
regret having brought a suffering child into existence, and
one can regret it for the sake of that child. However, when
one fails to bring a happy child into existence, one cannot
regret that for the sake of the child one did not bring into
existence.
The axiological asymmetry explains this too, because the
presence of harm in the suffering child is bad, whereas the
absence of benefit in the happy child is not bad.
4. The asymmetry of distant suffering and absent
happy people: We are rightly sad for suffering people in
distant places, but we are not similarly sad for the absence
of what would have been happy people on uninhabited islands or areas of earth or on other planets. When we think
| 2 7
about war-torn and poverty-stricken parts of the planet

we regret this suffering, and we do so for the poor beings
living in these conditions. We do not spend any timeand
rightly soworrying about the happy people who could
have existed on Mars, for example.
The axiological asymmetry explains this because the
absent benefits of the nonexistent are not bad, and are
thus not a cause for regret. By contrast, the present harms
are bad, and thus are a cause for regret.
Various attempts can be made to undermine the explanatory value of the axiological asymmetry. This is done
either (a) by suggesting that there are alternative explanations for each of the four other asymmetries just listed;
or (b) by denying these other asymmetries; or (c) by some
combination of (a) and (b).
For example, it might be argued that (i) the asymmetry
of procreational duties can be explained by appealing to
the view that while we have negative duties to avoid harm
we have no positive duties to benefit. According to this argument, the asymmetry of procreational duties is merely an
instance of a more general deontic asymmetry.
While this argument is open to those who deny that
we have any duties to benefit, it is not an option for those
who think that we do have such duties but who deny that
a duty to procreate is among those duties. Perhaps such
people could argue that the reason why a duty to procreate is not among our duties to benefit is because procreation involves considerable sacrificesfor the gestating
woman and for the parents who then rear the child
and we cannot be duty bound to make sacrifices of that
magnitude.
28 |
However, if it is the sacrifice involved in procreation

that stands in the way of there being a duty to procreate
then it must be conceded that the benefit of future possible people provides us with a reason, albeit a defeasible
one, to bring those people into existence. Yet that is at odds
with an asymmetry of procreational reasons: that while we
have a moral reason, grounded in the interests of potential
people, to avoid bringing into existence people who would
lead miserable lives, we have no such reason to bring into
existence those who would lead happy lives.
Perhaps this asymmetry is not quite as widely accepted
as the asymmetry of procreational duties. It is nonetheless
widely accepted. Those who wish to give it up are committed to thinking that potential happy people provide us
with a reason to create those people. Then they must either
concede that the reason for having a happy child and the
reason for not having a suffering child have asymmetrical strength, or they must claim that these reasons are
equally strong. If they think that the reason to avoid a suffering child is stronger than a reason to create a happy
child, they need to explain that asymmetry.
By contrast, if they think that the reason to create a
happy child is as strong as the reason to avoid creating a
suffering child, then they need to recognize that any defeating reasons need to be equally strong. Thus, it will simply
not be sufficient for prospective parents of a happy child
to defend their non-procreation merely by saying that they
do not wish to incur the costs of having that child. This is
because prospective parents of a suffering child could not
justify their procreation merely by saying that foregoing
procreation would be too great a sacrifice for them.
Those who wish to deny axiological asymmetry must
be held to a full accounting of their alternative and to
| 2 9
accepting the implications of it, because it is far too easy to

say that one rejects axiological asymmetry when one is not
held to such account.
Another attempt to circumvent the axiological asymmetry is to suggest that both (iii) the retrospective beneficence asymmetry and (iv) the asymmetry of distant
suffering and absent happy people can be explained by
whether or not there is a subject of harm or benefit.10 The
suggestion is that regret makes sense only if there is a subject of harm. In (iii) there is a suffering child for whom to
have regrets, whereas there is no happy child for whom to
have regrets, and in (iv) there are distant suffering people
whose suffering we can regret but there are no happy
people whose absent benefit we can regret.
There are a number of problems with this explanation
for (iii) and (iv). First, its explanatory capacity is limited. It
does not also explain either (i) the asymmetry of procreational duties or (ii) the prospective beneficence asymmetry. This is because there will be a subject of both harm and
benefit if we do create the suffering child and the happy
child, and there will not be a subject of harm or benefit
if we create neither of them. As a result, those seeking to
explain (iii) and (iv) without reference to the axiological
asymmetry must either lose the unificatory explanatory
value of the axiological asymmetry or they must deny (i)
and (ii). Either of those is a cost.
PROBLEMS WITH SYMMETRY

A second reason to accept the axiological asymmetry is
that there are problems with abandoning it in favor of
symmetry. Symmetry might be sought in at least two ways.
30 |
One way is to suggest that the absence of benefit in never

existing people is bad. However, if that were the case then
we would have to reject not only asymmetries (i) and (ii)
but also (iii) and (iv). We would need to regret the absence
of happy people and we would need to regret this based on
the interests of those people.
The other way to aim at symmetry is to judge (3) the
absence of harm in Scenario B to be not bad and (4) the
absence of benefit to be not good. These evaluations are
ambiguous. On one interpretation not bad means good
and not good means bad. However, if we adopt this interpretation then this means of attaining symmetry really
collapses into the previous one. To differentiate the current attempt at symmetry we need to interpret it differently. More specifically, we need to understand it to mean
that the absence of harm is not bad, but not good either
and the absence of benefit is not good, but not bad either.
However, the first of these claims is too weak. Avoiding the
harms of existence is not merely not bad. It is good. And
the claim that the absence of benefit in (4) is not good, but
not bad either is really just a fuller description of the claim
that it is not bad. Thus, even if we did opt for the fuller description there would be no symmetry. If (3) is good and
(4) is not good, but not bad either there is no symmetry.
SYMMETRY AND RELIEF BENEFITS

A third consideration against axiological symmetry is that
such symmetry would present problems for relief benefits. Such benefits are meliorations of harms. They include
quenching of thirst, relief from pain, cure from disease,
| 3 1
assuagement of guilt, and emergence from ignorance. The

defender of symmetry would have to say that the absence
of such benefits in those who never exist is bad. However,
it is extraordinarily difficult to make sense of that claim.
How could the absence of relief from an absent harm be
bad? How can it be bad that there is no relief if there is
nothing from which relief is needed?
Perhaps the defender of symmetry will respond by
wanting to distinguish between relief benefits and intrinsic benefits, the latter being those things that are good in
themselves. The argument would presumably be that the
axiological asymmetry does apply to absent relief benefits
but not to absent intrinsic benefits.
The problem with this response is that there is no
sharp line to be drawn between relief benefits and intrinsic
benefits. Many (if not all) plausible candidates for intrinsic
benefits may well be relief benefits. For example, having
rewarding work, interesting pastimes, and satisfying personal relationships, may seem to be intrinsic benefits.
However, they are also ways of preventing such harms as
dissatisfaction, boredom, loneliness, sadness, and stress. A
life devoid of these goods would be a boring life and thus
their presence is a way of driving out (some) harm.
In response it might be argued that these goods are not
relief benefits unless they temporally follow the respective harms and cause those harms to wane. However, this
is either too literal or too narrow an interpretation of the
concept of relief benefit. For a benefit to count as a relief
benefit (at least in the broader sense of that term), it is sufficient that it block the emergence of harm. In other words,
relief benefits should be understood to include those benefits that amount to either the alleviation or the prevention
32 |
of harms. For example, continuous hydration can prevent

thirst, but that does not mean that thirst-preventing hydration is not a relief benefit in the appropriately broader
sense. Hydration is a benefit only because its absence is a
harm. The absence of such a benefit is not bad unless the
absence amounts to a harm.
Some go so far as to say that all benefits are relief benefits in this sensethat all benefits are successes in the
struggle to keep harm at bay. There is much to be said for
this Schopenhauerian view,11 a contemporary version of
which is known as anti-frustrationism.12 According to this
view, advanced by Christoph Fehige, a satisfied preference
and no preference are equally good. Only an unsatisfied
preference is bad. In support of this view, Professor Fehige
asks us to consider a case in which we paint the tree nearest the Sydney Opera house red and give Kate a pill that
makes her wish that the tree nearest the Sydney Opera
house were red.13 He claims, entirely plausibly, that Kate
is no better off than if we had neither created nor fulfilled
the desire.
Professor Fehige speaks in terms of preferences and
desires, but a similar point can be made about other benefits, such as the fulfillment of needs. Compare, for example, a bird that needs a nest and a fish that neither has nor
needs a nest. It seems highly implausible to suggest that
it is better to have a fulfilled need for a nest than to have
no such need. By the same token, humans need oxygen
and it is obviously good when that need is fulfilled, but the
fulfilled need is no better than a possible world in which
humans had no need for and no access to oxygen. Similarly,
an existing humans fulfilled need for oxygen is good, but it
is no better than the absence of a never-existing humans
| 3 3
need for oxygen. Fulfillment of a need is good only if the

need exists.
It should be clear that anti-frustrationism and other
views that see all benefits as relief benefits strongly support my axiological asymmetry against attempts to symmetrize. If benefits are the relief from or absence of harms,
then Scenario A has no advantages over Scenario B in
Figure 2.1. This is why views such as anti-frustrationism
embodying an axiological asymmetry similar, if not identical, to mine14 also lead to the conclusion that coming
into existence is always harm.
For this reason, the very people most disposed to resist
my axiological asymmetry may also want to resist arguments that all benefits are relief benefits. This of course
does not release them from the burden to provide compelling counterarguments. However, it is worth noting that
even if one does not go so far as to say that all benefits
are relief benefits, the extent to which relief benefits and
intrinsic benefits are bound up with one another makes it
difficult to disentangle them. Those who wish to attach a
different evaluation to the absent relief benefits and the
absent intrinsic benefits of absent people face a problem
that those who accept the axiological asymmetry do not
have. They need to explain to us which benefitsthe purportedly intrinsic onesare not ones that consist in relieving or preventing harm. Alternatively, if they think that
some benefits are both relief and intrinsic benefits, they
are going to need to differentiate these and then apply a
more nuanced evaluative schema rather than merely denying axiological asymmetry.
Moreover, even if they can do this, the process will
mean that there will be many fewer intrinsic benefits and
34 |
thus the cost-benefit calculation is much less likely to come

out in favor of existence even if one does reject the axiological asymmetry for intrinsic benefits.
S O LV I N G P O P U L AT I O N E T H I C S
PROBLEMS
Another consideration in favor of accepting the axiological
asymmetry is that doing so avoids various problems that
have plagued moral theory about population, and more
specifically about the ethics of creating new people.15 It
must be said that it solves these problems in an unusual
way. Nevertheless, the fact that it does solve those problems must surely be a consideration in its favor, even
though it is secondary to the previous ones. In other words,
this is less an argument for accepting the axiological asymmetry and more a supporting consideration.
The foundational problem of population ethics is the
non-identity problem, which was outlined at the beginning
of this chapter. This is the problem of explaining why it is
wrong to create a suffering person when the alternative
would have been not to create that person at all. Theories
that attempt to explain the wrong with reference to the
interests of the person who is createdso called person-
affecting theoriesconfront the problem of explaining
how that person is harmed. I suggested some ways in which
this problem might be overcome.
Depending on how one interprets them, those ways
do not necessarily presuppose the axiological asymmetry.
However, unless one accepts that asymmetry and reaches
the conclusion that coming into existence is always a harm,
| 3 5
there will be the residual problem of explaining how bad

a life needs to be in order for the creation of such a life to
constitute a harm. In other words, the axiological asymmetry leads to a very clear answer to this question.
Not all responses to the non-identity problem are
person-affecting. That is to say, not all responses a ttempt
to explain the wrong in terms of harm to the person
brought into existence. Impersonal views can circumvent
the non-identity problem by saying that an outcome in
which a suffering person is created is worse (impersonally)
than an outcome in which either no such person is created
or a person with a better quality of life is created.
Although this circumvents the non-identity problem,
it does not solve it. It does not explain how the person
brought into existence is harmed. Perhaps circumvention
of the problem will be thought to be sufficient. However,
the difficulty with adopting an impersonal view in response to the non-identity problem is that this gives rise
to a number of other problems. Which problems arise depends on whether one adopts a total or an average version of the impersonal view.
According to the total impersonal view, we ought to aim
at the outcome in which there would be the greatest amount
of whatever makes life worth living.16 According to this view,
a smaller population with a higher quality of life is worse
than a larger population with a worse quality of life, as long
as there are enough extra people in the latter to outweigh the
lower quality of life. This implies that an immensely populous world in which everybody has a life just barely worth
living (World Z) would be preferable to a world in which there
were only a few people with a very high quality of life (World
A). This is known as the Repugnant Conclusion.17
36 |
The impersonal average view, according to which the

best outcome is the one in which peoples lives go on average best,18 can avoid this problem. This is because the
average quality of life in World Z is much lower than it is
in World A. However, the impersonal average view faces
another problemthe Mere Addition problem.19 To understand this problem, consider a scenario in which if you
procreate your child will have a good quality of life. According to the impersonal average view, whether this additional
child should be created depends on whether, for example,
the ancient Egyptians had an even higher quality of life. If
they did, then your procreation would lower the average
quality of life and would thus be wrong. Yet, as Derek Parfit
claims, research in Egyptology cannot be relevant to our
decision whether to have children20 and thus the impersonal average view is problematic.
The axiological asymmetry and its implication that
coming into existence is always a harm avoid all of these
problems. I have already explained how the non-identity
problem is avoided. It obviates the need to resort to an impersonal view and thus prevents the problems that arise
from adopting such a view. However, even if one does embrace an impersonal view, my view might prompt one to restrict the scope of that view to existing people by giving up
the assumption that it is good to create additional lives that
are worth continuing. This would then explain why World
Z is worse than World A and would avoid the Repugnant
Conclusion. It would also say why it would be bad to add
more people and thus avoid the problem of Mere Addition.
It is true that an impersonal view that was restricted
to existing people would tell us nothing about the ethics
of creating new people. However, axiological asymmetry
| 3 7
would contribute importantly to answering such questions. It would tell us that creating new people is always
harmful to them.
THE AXIOLOG ICAL ASYM METRY

INCONTEXT
The axiological asymmetry strikes me not merely as true,
but also as clearly true. I suspect that more people would
concede this if it did not lead to the conclusion that coming
into existence is always a harma conclusion that many
people find unbearable. However, because this conclusion
is taken to be so unacceptable, people scramble desperately
to undermine the asymmetry.
The axiological argument is among the most technical of the arguments that support anti-natalism, and the
responses to it are similarly technical. This is not necessarily a bad thing. However, one consequence is that the moral
callousness of rejecting asymmetry is camouflaged by the
technicality.
The moral insensitivity becomes more apparent when
people attempt to reject the (generally less technical)
anti-natalist arguments I shall advance in the next two
chapters. In the first of these chapters, I shall highlight
a number of empirical asymmetries between harms and
benefits. For example, I shall show that pains tend to last
longer than pleasures, that the worst pains are worse than
the best pleasures are good, and that positive states are
generally less stable than negative ones.
These and other empirical asymmetries, being empirical, are easier to demonstrate. They also provide further
38 |
support for axiological asymmetry. Given the empirical

asymmetries, we should fully expect that asymmetrical
values and duties attach to harms and benefits. It is axiological and deontic symmetry, not asymmetry, that would
be surprising. Thus, it seems that the empirical asymmetries place the burden of proof on those who assert axiological and deontic symmetry.
NOTE S
1. Derek Parfit, Reasons and Persons (Oxford: Clarendon,
1984), 359.
2. Gregory Kavka, The Paradox of Future Individuals, Philosophy and Public Affairs 11, no. 2 (1982): 93112.
3. On some views, the problem arises even in cases when the
life would not be worth living.
4. For more on this see Benatar, Better Never to Have Been,
4 and 31, and David Benatar, Still Better Never to Have
Been: A Reply to (More of) My Critics, Journal of Ethics 17,
no. 12 (2013): 121151.
5. In Better Never to Have Been I spoke of pains and pleasures
as exemplars of harms and benefits, thus implying that the
axiological asymmetry applied to all harms and benefits,
not only pains and pleasures. That did not stop innumerable commentators from claiming that mine is a hedonistic
argument. (Many went so far as to say that it was a hedonistic utilitarian argument.) In a further bid to avoid this
misinterpretation, I am here describing the asymmetry in
terms of harms and benefits rather than in terms of pain
and pleasure. See also David Benatar, Every Conceivable
Harm: A Further Defence of Anti-Natalism, South African
Journal of Philosophy 31, no. 1 (2012): 131132.
6. This is not (yet) the anti-natalist conclusion. Anti-natalism
does not follow immediately from the claim that coming
into existence is always a harm. If the harm were a minor
one, then procreation might nonetheless be permissible.
| 3 9
7. Genesis 1:28. Note, however, that not all religious people

believe that there is a duty to procreate. Catholic priests and
nuns, for example, do not believe that they have any such
duty.
8. Not all positive utilitarians are committed to such a duty.
Those who distinguish between making people happy and
making happy people and who think that our positive
utilitarian duties are restricted to the former need not embrace a duty to procreate.
9. Notice that the religious and positive utilitarian arguments
for a duty to procreate do not ground this duty on the interests of the future possible people.
10. David DeGrazia makes this suggestion (Is It Wrong To
Impose the Harms of Human Life?: A Reply to Benatar,
Theoretical Medicine and Bioethics 31.4 (2010): 317331. For a
more detailed response than I offer here, see Benatar, Still
Better Never to Have Been, Journal of Ethics, 2013.
11. Arthur Schopenhauer, On the Sufferings of the World, in
Complete Essays of Schopenhauer, trans. T. Bailey Saunders,
vol. 4, Studies in Pessimism (New York: Wiley, 1942), 5.
12. Paradoxically, it is also known as frustrationism, which
sounds like its opposite.
13. Christoph Fehige, A Pareto Principle for Possible People, in
Preferences, eds. Christoph Fehige and Ulla Wessels (Berlin:
Walter de Guyter, 1998), 513514.
14. For more on Christoph Fehiges view and its relationship to
my axiological asymmetry, see Better Never to Have Been,
5457.
15. I shall provide a brief account here of how it solves these
problems. For a much fuller account please see Better Never
to Have Been, 168178.
16. Parfit, Reasons and Persons, 387.
17. Ibid., 388.
18. Ibid., 386.
19. Ibid., 420.
20. Ibid., 420.
The Quality-of-Life Argument
THE ASYMMETRY ARGUMENT IS sufficient to reach the con-
clusion that coming into existence is always a harm. However, it is not sufficient to show that bringing somebody
into existence is always wrong. If the harm of coming into
existence were a minor one, one might think that procreation could be justified by the benefits it brings to others. In
fact, as I shall argue in this chapter, the harms of existence
are considerable. Thus the quality-of-life argument is one
argument that can supplement the asymmetry argument
to yield the anti-natalist conclusion. The quality-of-life argument can also be viewed as an independent argument
that is sufficient to generate anti-natalism, even by those
who reject the asymmetry argument.
The quality of peoples lives obviously varies immensely.
However, thinking that some lives are worse or better than
others is merely a comparative claim. It tells us nothing about
whether the worse lives are bad enough to count as bad lives
or whether the better lives are good enough to count as good
T he Q uality - of - L ife A rgument
| 4 1
lives. The common view, however, is that the quality of some

lives qualifies as bad and the quality of others qualifies as
good. In contrast to this view, I believe that while some lives
are better than others none are (noncomparatively) good.
The obvious objection to this view is that billions of
people judge the quality of their own lives to be good. How
can it possibly be argued that they are mistaken and that
the quality of their lives is, in fact, bad?
The response to this objection consists of two main
steps. The first is to demonstrate that people are very unreliable judges of the quality of their own lives. The second
step is to show that when we correct for the biases that
explain the unreliability of these assessments and we look
at human lives more accurately we find that the quality (of
even the best lives) is actually very poor.
WHY PEOPLE S JUDG MENTS ABOUT

T H E Q UA L I T Y O F T H E I R L I V E S A R E
UNRELIABLE
Peoples self-assessments of well-being are unreliable
indicators of a persons quality of life because these self-
assessments are influenced by three psychological phenomena, the existence of which has been well demonstrated.
The first of these is an optimism bias, sometimes known
as Pollyannaism. For example, when asked to rate how happy
they are, peoples responses are disproportionately toward
the happier end of the spectrum. Only a small minority of
people rate themselves as not too happy.1 When people are
asked to rate their well-being relative to others, the typical response is that they are doing better than the most
42 |
commonly experienced level, suggesting, in the words of

two authors, an interesting bias in perception.2 It is unsurprising that peoples reports of their overall well-being
are unduly optimistic because the building blocks of that
judgment are similarly prone to an optimism bias. People
have been shown to recall positive experiences more than
negative experiences.3 They are similarly optimistic in their
projections of what will happen to them in the future.4
Judgments about the overall quality of ones life that are inadequately informed by the bad things that have happened
and will happen to one are not reliable judgments.
There is ample evidence of an optimism bias among
humans. This is not to say that the extent of the bias is invariable. The inhabitants of some countries report greater
subjective well-being than those of other countries even
when the objective conditions are similar.5 This has been
attributed, in part, to cultural variation.6 However, optimism bias is found everywhere, even though the extent of
the bias varies.7
A second psychological phenomenon that provides
us with grounds for distrusting self-assessments of well-
being is what is known as accommodation, adaptation, or
habituation. If ones self-assessments were reliable they
would track improvements and deteriorations in ones
objective conditions. That is to say, if ones condition improved or deteriorated one would perceive ones condition
to have improved or deteriorated to that degree. Ones
self-assessment would then remain fixed until there was a
further improvement or deterioration, in response to which
ones self-assessment would respond commensurately.
However, that is not what happens. Our subjective assessments do respond to shifts in our objective conditions
| 4 3
but the altered self-assessment is not stable. As we adjust

to our new condition we cease to rate our condition as we
did when it first improved or deteriorated. For example, if
one suddenly loses the use of ones legs, ones subjective assessment will drop precipitously. In time, however, ones
subjective assessment of ones lifes quality will improve as
one adjusts to the paralysis. Ones objective condition will
not have improvedone is still paralyzedbut one will
judge ones life to be going less badly than one did immediately after one became paralyzed.8
There is some disagreement about the extent to which
we adapt. Some have suggested that it is completethat we
return to a baseline or set-point level of subjective well-
being. Others deny that the evidence shows this, at least not
in every domain of our lives.9 However, there is no dispute
that there is some adaptation and that it is sometimes significant. This is all that is required to lend support to the
claim that our subjective assessments are unreliable.
The third feature of human psychology that compromises the reliability of subjective assessments of well-being
is what we might call comparison. Subjective assessments
of well-being implicitly involve comparison with the well-
being of others.10 Our judgments about the quality of our
own lives are influenced by the (perceived) quality of the
lives of others. One consequence of this is that bad features
of all human lives are substantially overlooked in judging
the quality of ones life. Because these features of ones
life make one no worse off than others, one omits them in
reaching a judgment about the quality of ones life.
Whereas Pollyannaism biases judgments only in the
optimistic direction, adaptation and comparison are more
complicated. One adapts not only to deteriorations but
44 |
also to improvements in ones objective condition. Similarly one can compare oneself not only to those worse off
than oneself but also to those better off than oneself. It
would be a mistake, however, to think that the net effect
is to cancel any bias. This is because both adaptation and
comparison work against the backdrop of the optimism
bias. They may moderate the optimism bias but they do not
cancel it. Moreover, there is an optimism bias in the manifestation of these other traits. For example, we are more
likely to compare ourselves with those who are worse off
than with those who are better off.11 For these reasons, the
net effect of the three traits is for us to overestimate the
actual quality of our lives.
It is unreasonable to ignore the vast body of evidence
for these psychological characteristics of humans. To
insist, in the face of the evidence, that subjective appraisals of quality of life are reliable is a kind of denialism. This
is not to say that every human overestimates the quality of
his or her life. The evidence shows that the phenomenon is
widespreadnot that it is universal. There are some people
who have an accurate assessment, but these are the minority and very likely include those who do not take issue with
my grim view about the quality of human life.
Nor does acknowledging the unreliability of subjective assessments imply that subjective assessments are irrelevant. Thinking that ones life is better than it actually
is can make it better that it would otherwise be. In other
words, there can be a feedback loop whereby a positive
subjective assessment actually improves ones objective
well-being. However, there is a difference between a subjective assessment of ones well-being influencing the objective level and a subjective assessment determining the
| 4 5
objective level. Even if ones overly optimistic subjective

assessment makes ones life better than it would otherwise
be, it does not follow that ones life is actually going as well
as one thinks it is.
I have shown so far that there is excellent reason to
distrust cheery subjective assessments about the quality of
human life. However, to show that the quality of peoples
lives is worse than they think it is, is not to show that the
quality of their lives is very bad. That conclusion requires
further argument, which I shall now provide.
T H E P O O R QUA L I T Y O F H U M A N L I F E
Most people recognize that human lives can be of an appallingly low qualityso low that it would have been better if
those lives had never been created. The tendency, however,
is to think that this is true of other peoples lives, not ones
own.12 However, if we look dispassionately at human life,
and control for our biases, we find that human life is permeated by bad.
Even in good health, much of every day is spent in
discomfort. Within hours we become thirsty and hungry.
Many millions of people are chronically hungry. When we
can access food and beverages and thus succeed in warding
off hunger and thirst for a while we then come to feel the
discomfort of distended bladders and bowels. Sometimes
relief can be obtained relatively easily, but on other occasions the opportunity for (dignified13) relief is not as forthcoming as we would like. We also spend much of our time
in thermal discomfortfeeling either too hot or too cold.
Unless one naps at the first sign of weariness, one spends
46 |
quite a bit of the day tired. Indeed, many people wake up

tired and spend the day in that state.
With the exception of chronic hunger among the worlds
poor, these discomforts all tend to be dismissed as minor
matters. While they are minor relative to the other bad
things that befall people, they are not inconsequential. A
blessed species that never experienced these discomforts
would rightly note that if we take discomfort to be bad
then we should take the daily discomforts that humans
experience more seriously than we do.
Other negative states are experienced regularly even if
not daily or by everybody. Itches and allergies are common.
Minor illnesses like colds are suffered by almost everybody.
For some people this happens multiple times a year. For
others it occurs annually or every few years. Many women
of reproductive years suffer regular menstrual pains and
menopausal women suffer hot flashes.14 Conditions such
as nausea, hypoglycemia, seizures, and chronic pain are
widespread.
Nor are the negative features of life restricted to unpleasant physical sensations. For example, we frequently
encounter frustrations and irritations. We have to wait in
traffic or in queues. We encounter inefficiency, stupidity,
evil, Byzantine bureaucracies, and other obstacles that
can take thousands of hours to overcomeif they can
be overcome at all. Many important aspirations are unfulfilled. Millions of people seek a job but remain unemployed. Of those who have jobs, many are dissatisfied with
them, or even loathe them. Even those who enjoy their
work may have professional aspirations that remain unfulfilled. Most people yearn for close and rewarding personal relationships, not least with a lifelong partner or
| 4 7
spouse. For some this desire is never fulfilled. For others

it temporarily is, but then they find that the relationship
is trying and stultifying, or their partner betrays them or
becomes exploitative or abusive. Most people are unhappy
in some or other way with their appearancethey are too
fat, or they are too short, or their ears are too big. People
want to be, look, and feel younger and yet they age relentlessly. They have high hopes for their childrenbut these
are often thwarted, when, for example, the children prove
a disappointment in some way or other. When those close
to us suffer, we suffer at the sight of it. When they die we
are bereft.
Millions are ravaged by natural disastersfloods,
earthquakes, tornadoes and hurricanes. They lose their
property and often life or limb. There are also innumerable
harms that people suffer at the hands of other humans, including being humiliated, shamed, denigrated, maligned,
beaten, assaulted, raped, tortured, and murdered.
The select few for whom things go relatively well while
they are in their prime forget about what terrible fates could
still befall them. Many fortunes and loves have been lost,
many reputations ruined, and many dreams destroyed.
Sometimes lives start terribly. However, the worst parts of
a life are often not at the beginning but rather toward the
end. Consider the millions of people who gradually succumb
to cancer, AIDS, or neurodegenerative diseases. Millions of
people struggle to breathe or to walk. Many languish for
years following a stroke, unable to speak, walk, feed, or
clean themselves. In these and other conditions, people are
reduced to states of indignity and dependence that most of
them deeply resent. There is little or no hope of recovery.
Very few lives, if any, contain no serious suffering.
48 |
Some deaths are better than others, but there is always

a serious cost. The obvious cost in protracted deaths or in
deaths that follow protracted illness is the sheer amount of
suffering, including the mental suffering that arises from
knowing that one is dying. Deaths that come swiftly are
often premature or, if they occur after a (relatively) long
life, typically (even if not always) follow a period of decrepitude. Whatever the manner and timing of our deaths, the
fact is that we do all die and that is a serious harm. (Where
death is a release, the tragedy is that life became so bad
that death, ordinarily harmful, is either not harmful or is
the lesser of two evils.)
Optimists will very likely suggest that this is a onesided picturethat lives typically contain not only bad
but also good. However, while it is true that lives are not
usually unadulteratedly bad, there is much more bad than
good even for the luckiest humans. Things are worse still
for unluckier people, many of whom have almost nothing
going in their favor.
Our lives contain so much more bad than good in part
because of a series of empirical differences between the bad
things and the good things. These differences show that in
addition to the axiological asymmetry between good and
bad, which I discussed in the previous chapter, there are
also empirical asymmetries.
For example, the most intense pleasures are short-lived
whereas the worst pains can be much more enduring. Orgasms, for example, pass quickly. Gastronomic pleasures
last a bit longer but even if the pleasure of good food is protracted it lasts no more than a few hours whereas severe
pains can endure for days, months, and years. Indeed, pleasures in generalnot only the most sublime of themtend
| 4 9
to be shorter lived than pains. Chronic pain is rampant, but

there is no such thing as chronic pleasure. There are people
who have an enduring sense of contentment or satisfaction, but that is not the same as chronic pleasure. Moreover, discontent and dissatisfaction can be as enduring as
contentment and satisfaction, which means that the positive states are not advantaged in this realm. Indeed the
positive states are less stable because it is much easier for
things to go wrong than to go right (as I shall explain more
fully below).
The worst pains are also worse than the best pleasures
are good. Those who deny this should consider whether
they would accept an hour of the most delightful pleasures
in exchange for an hour of the worst tortures. Arthur Schopenhauer makes a similar point when he asks us to compare the respective feelings of two animals, one of which
is engaged in eating the other.15 The animal being eaten
suffers and loses vastly more than the animal that is eating
gains from this one meal.
Consider too the temporal dimensions of injury or illness and recovery. One can be injured in seconds: one is
hit by a bullet or projectile, or is knocked over or falls,
or suffers a stroke or heart attack. In these and other
ways one can instantly lose ones sight or hearing or the
use of a limb or years of learning. The path to recovery is
slow. In many cases full recovery is never attained. Injury
comes in an instant and the resultant suffering can last
a lifetime. Even lesser injuries and illnesses are typically
incurred much more quickly than one recovers from
them. For example, the common cold strikes quickly and
is defeated much more slowly by ones immune system.
The symptoms manifest with increasing intensity within
50 |
hours but they take at least days, if not weeks, to disappear entirely.
There are, of course, conditions in which one declines
gradually rather than suddenly, but the great majority of
theseincluding age-related physical decline, dementia,
neuromuscular degenerative diseases, and the deterioration from advancing cancersare conditions from which
there is no recovery. Where there are treatments, some are
merely palliative. When treatments are potentially curative, the decline is the default and one has to battle against
that presumption, sometimes successfully but other times
not. Moreover, billions of people simply have no access to
either curative or palliative treatments.
Things are also stacked against us in the fulfillment of
our desires and the satisfaction of our preferences.16 Many
of our desires are never fulfilled. There are thus more unfulfilled than fulfilled desires. Even when desires are fulfilled,
they are not fulfilled immediately. Thus there is a period
during which those desires remain unfulfilled. Sometimes
that is a relatively short period (such as between thirst and,
in ordinary circumstances, its quenching) but in the case
of more ambitious desires, they can take months, years, or
decades to fulfill. Some desires that are fulfilled prove less
satisfying than we had imagined. One wants a specific job
or to marry a particular person, but upon attaining ones
goal one learns that the job is less interesting or the spouse
is more irritating than one thought.
Even when fulfilled desires are everything that they were
expected to be, the satisfaction is typically transitory, as the
fulfilled desires yield to new desires. Sometimes the new
desires are more of the same. For example, one eats to satiety but then hunger gradually sets in again and one desires
| 5 1
more food. The treadmill of desires works in another way

too. When one can regularly satisfy ones lower level desires,
a new and more demanding level of desire emerges. Thus
those who cannot provide for their own basic needs spend
their time striving to fulfill these. Those who can satisfy the
recurring basic needs develop what Abraham Maslow calls
a higher discontent17 that they seek to satisfy. When that
level of desires can be satisfied, the aspirations shift to a yet
higher level.
Life is thus a constant state of striving. There are sometimes reprieves, but the striving ends only with the end of
life. Moreover, as should be obvious, the striving is to ward
off bad things and to attain good things. Indeed some of the
good things amount merely to the (temporary) relief from
the bad things. For example, one satisfies ones hunger or
quenches ones thirst. Notice too that while the bad things
come without any effort, one has to strive to ward them off
and to attain the good things. Ignorance, for example, is
effortless but knowledge requires (usually hard) work.
Even the extent to which our desires and goals are fulfilled creates a misleadingly optimistic impression of how
well our lives are going. This is because there is actually a
form of self-censorship in the formulation of our desires
and goals. While many of them are never fulfilled, there
are many more potential desires and goals that we do not
even formulate because we know that they are unattainable. For example, we know that we cannot live for a few
hundred years and that we cannot gain expertise in all the
subjects in which we are interested. Thus we set goals that
are less unrealistic (even if many of them are nonetheless
somewhat optimistic). Thus one hopes to live a life that is,
by human standards, a long life and gain expertise in some
52 |
perhaps very focused area. What this means is that even

if all our desires and goals were fulfilled our lives are not
going as well as they would be going if the formulation of
our desires had not been artificially restricted.
Further insight into the poor quality of human life can
be gained from considering various traits that are often
thought to be components of a good life and by noting
what limited quantities of these characterize even the best
human lives. For example, knowledge and understanding are widely thought to be goods and people are often
in awe of how much knowledge and understanding (some)
humans have. The sad truth, however, is that, on the spectrum from no knowledge and no understanding to omniscience, even the cleverest, best educated humans are much
closer to the unfortunate end of the spectrum. There are
billions more things we do not know or understand than
we do know and understand. If knowledge really is a good
thing and we have so little of it, our lives are not going very
well in this regard.
Similarly, we consider longevity to be a good thing (at
least if the life is above a minimum quality threshold). Yet
even the longest human lives are fleeting. If we think that
longevity is a good thing then a life of a thousand years
(in full vigor) would be much better than a life of eighty
or ninety years (especially where the last few decades are
years of decline and decrepitude). Ninety is much closer to
one than it is to a thousand. It is even more distant from
two or three or more thousand. If, all things being equal,
longer lives are better than shorter ones, human lives do
not fare well at all.18
It is not surprising that we fail to notice this heavy preponderance of bad in human life. The facts I have described
| 5 3
are deep and intractable features of human (and other) life.

Most humans have accommodated to the human condition
and thus fail to notice just how bad it is. Their expectations
and evaluations are rooted in this unfortunate baseline.
Longevity, for example, is judged relative to the longest
actual human lifespans and not relative to an ideal standard. The same is true of knowledge, understanding, moral
goodness, and aesthetic appreciation. Similarly, we expect
recovery to take longer than injury and thus we judge the
quality of human life off that baseline even though it is an
appalling fact of life that the dice are stacked against us in
this and other ways.
The psychological trait of comparison is obviously also
operative. Because the negative features I have described
are common to all lives, they play very little role in how
people assess the quality of their lives. It is true for
everybody that the worst pains are worse than the best
pleasures are good, and that pleasures can and often do last
much longer than pleasures. Everybody must work hard to
ward off unpleasantness and to seek the good things. Thus
when people judge the quality of their own lives and do so
by comparing them to the lives of others, they tend to overlook these and other such features.
All this occurs against the backdrop of an optimism
bias, under which people are already inclined to focus on
the good more than the bad. The fact that we fail to notice
how bad human life is does not detract from the arguments
I have given that there is much more bad than good. Human
life would be vastly better if pain were fleeting and pleasure protracted, if the pleasures were much better than the
pains are bad, if it were really difficult to be injured or get
sick, if recovery were swift when injury or illness did befall
54 |
us, and if our desires were fulfilled instantly and if they did
not give way to new desires. Human life would also be immensely better if we lived for many thousands of years in
good health and if we were much wiser, cleverer, and morally better than we are.
SECULAR OPTIMISTIC THEODICIE S

Human optimism is resilient. It does not wilt in the face of
evidence. No matter how much evidence one provides for
psychological traits such as optimism bias, and no matter
how much evidence there is that the quality of human life
is very bad, most humans will adhere to their optimistic
view. Sometimes this optimism manifests, at least in part,
as religious faith,19 with people declaring the goodness of
God and his creation. Religious optimism of this kind is
often challenged by the argument from evil, which suggests that the existence of an omniscient, omnipotent, omnibenevolent God is incompatible with the vast amount of
evil that exists in the world. Theodicy is the (optimistic)
practice of trying to reconcile Gods existence with that
evil. However, many atheists, while critical of theodicy, are
themselves engaged in a kind of secular theodicyan attempt to reconcile their optimistic views with the unfortunate facts about the human condition.
There are many secular theodicies. One of the most
commonly expressed is that the bad things in life are necessary. For example, it is suggested that without pain we
would incur more injuries. Indeed those people with congenital insensitivity to pain harm themselves unwittingly
by, for example, grasping and continuing to hold objects
| 5 5
that are dangerously hot, or by unrestricted use of limbs in

which a bone has been broken. In the absence of pain they
are simply not alerted to the danger.
It is also suggested that the bad things in life are necessary in order to appreciate the good things, or at least to
appreciate them fully. On this view, we can only enjoy pleasures (as much as we do) because we also experience pain.
Similarly, our achievements are more satisfying if we have
to work hard to attain them, and fulfilled desires mean
more to us because we know that desires are not always
fulfilled.
There are many problems with this sort of argument.
First, these sorts of claims are not always true. Lots of pain
serves no useful purpose. There is no value in labor pains
or in pain resulting from terminal diseases, for example.
While the pain associated with kidney stones might now
lead somebody to seek medical help, for most of human history such pain served no purpose, as there was absolutely
nothing anybody could do about kidney stones.20 Moreover there are at least some pleasures we can enjoy without
having to experience pain. Pleasant tastes, for example, do
not require any experience of pain or unpleasantness. Similarly many achievements can be satisfying even if they involve less or no striving. There may be a special satisfaction
in the ease of attainment. There may be some individual
variation. Perhaps some people are more capable of enjoying pleasure without having to experience pain and more
capable of taking satisfaction in achievements that come
with ease.
Second, insofar as the good things in life do require a
contrast in order to be fully appreciated, it is not clear that
this appreciation requires as much bad as there is. We do
56 |
not, for example, require millions of people suffering from

chronic pain, infectious diseases, advancing paralysis, and
tumors in order to appreciate the good things in life. We
could enjoy our achievements without having to work quite
so hard to attain them.
Finally, and perhaps most importantly, to the extent
that the bad things in life really are necessary, our lives are
worse than they would be if the bad things were not necessary. There are both real and conceivable beings in which
nociceptors lead to behavior aversive to noxious stimuli,
but without being mediated by pain. This is true of plants
and simple animal organisms, and it is also true of the
reflex arc in higher animals, such as humans.21 We can also
imagine beings much more rational than humans, in which
nociceptors and aversive behavior were mediated by a rational faculty rather than a capacity to feel pain. In such
beings a noxious stimulus would be received but not felt
(or at least not in the way pain is), and the rational faculty
would, as reliably as pain, induce the being to withdraw. It
would be much better to be that sort of being than to be
our sort of being. It would similarly be better to be the sort
of being that can appreciate the good things in life without
having to experience bad things or without having to work
really hard to attain the good things. Lives in which there
is no gain without the pain22 are much worse than lives
in which there could be the same gain without the pain.
A second theodicy picks up here. It insists that the
perfectionist standards I am using to judge the quality of
human life are too demanding and not appropriate. One
version of this critique says that we must adopt a human
perspective, not the so-called perspective of the universe
in determining what is good for humans.23 Now, of course
| 5 7
there is a sense in which it is true that we need to take

account of what sort of beings humans are in order to determine what is good for them. For example, given that we
are terrestrial animals, submerging a human under water
(without breathing equipment) is going to be bad for that
human even though it would not be bad for a fish. Yet we
can say that it would certainly be better for humans if
they could not drownthat is, if they had the capacity to
breathe not only in air but also in water.
Here another version of the second theodicy is often
invoked. It claims that there are constraints on how good
a human life can be while still being a human life. A being
that could breathe not only in air but also under water
would not be a human. A life without pain would not be
a human life. Nor should we judge the extent of human
knowledge, understanding, and goodness by the standards
of omniscience and omnibenevolence, because the latter
standards are not human standards. An omniscient, omnibenevolent being would not be a human. It would be God.
This version of the argument is also unconvincing. The
problem is that it fetishizes human life. Some emotional
distance might be required to realize this and thus consider an imaginary species rather than humans. Members
of this fictional species, which we might call Homo infortunatus, have an even more wretched quality of life than
most humans have, but their lives are not devoid of all
pleasure and other goods. Now imagine that a pessimistic philosopher among them observes how appalling their
lives are. He points to how much better things could be.
For example, instead of living only thirty years, they might
live to eighty or ninety. Instead of being in an almost constant state of hunger, they might get hungry only between
58 |
three regular meals a day. Instead of being sick every week,

they might suffer illness only annually or even less often.
In response to such observations, the optimistic members
of the speciesa vast majoritywould object that if their
lives were better in those ways they would no longer be infortunati. That observation, even if true, would not detract
from the claim that the quality of life of the infortunati is
wretched. There is, after all, a difference between asking
how good the quality of life of a particular species is and
asking whether a much better life is compatible with being
a member of that species. Perhaps we would not be human
if the quality of our lives were much better than it is. It does
not follow that the quality of human life is good.
To prefer a human life to a better life suggests a distracting sentimentality about humanity. It is to think that
it is more important to be human than to have a better quality of life. Yet the typical reasons provided for the value of
being human rather than some other species seem to imply
that it would be better to be better than to be human, even
if that implication is not typically noticed. For example,
most humans think that it is better to have the higher cognitive capacity of Homo sapiens than the lesser capacity of
Homo erectus. It seems that the logic underlying this judgment is that greater cognitive capacity is better than lesser
cognitive capacity. But this logic supports a further judgment that it would be better to have the still greater cognitive capacity of a superhuman species.
One way to ward off this implication would be to claim
that there is a Goldilocks level of cognitive capacity. On
this view, it is bad to have too little but also bad to have too
much. (Perhaps too much cognitive capacity either gives
one insights that are conducive to unhappiness or can lead
| 5 9
to unacceptable levels of destructiveness.) The problem

with the Goldilocks argument is that if there is some optimum level of cognitive sophistication it is both too convenient and implausible to think that the level is that of
Homo sapiens.
It is difficult to prove this to those who take it as an
article of faith that humans have the optimum level of this
trait. However, consider that humans greater cognitive
capacity has led them to be much more destructive than
their fellow hominids and other primates. Yet humans lack
the still further sophistication that would check that destructiveness by, for example, enabling them to think and
act more rationally. Perhaps it will be argued in response
that although humans would become less destructive if
they were cognitively more sophisticated, they would acquire, with that greater cognitive capacity, unbearable insight into the human predicament, thereby making them
more unhappy. But humans do suffer a great deal from
such angst, which suggests that they may already have too
much cognitive capacity for their own happiness.
I have referred to cognitive capacity as a trait. It is,
however, a constellation of traits. As implausible as it is to
claim that humans possess the optimum degree of cognitive capacity overall, it is still less plausible to make this
claim with respect to some of the component capacities.
Think of computational ability, for example. It would be
better if ordinary humans had greater computational ability than they currently have, at least if this did not involve
a reduction in any other capacities.
It is even harder to argue that humans occupy a Goldilocks position on the spectra of other attributes. For example, it would be exceedingly difficult to defend the view
60 |
that humans have an optimum degree of moral goodness,

as this would imply that it would be worse if they were
morally better. If this is not absurd it is at least highly
implausible.
Not all optimists fetishize humanity. Among the advocates of human enhancement are those who envisage and
welcome the prospect of a post-human futurea future
in which humans have been so enhanced (physically, mentally, and morally) that they are no longer recognizably
human. These advocates of transhumanism think it is
much more important to improve the quality of life than
for the enhanced future beings to be human.
While there are many who object to the wisdom and
morality of seeking such enhancements, I am not among
those categorically opposed to technological enhancements. If the choice is between a lower quality of life and a
higher quality of life, the latter is preferable even if the enhanced beings with the better-quality lives can no longer
be categorized as humans. To be sure, any enhancements
will need to be subject to the usual moral constraints. For
example, enhancements that carried significant risks of
causing serious harms might fall afoul of such constraints.
And attention would need to be paid to fair access to enhancement technologies. None of this, however, rules out
the transhumanist project.
However, while transhumanists are not fixated on
whether a life is human, they are nonetheless engaged in another kind of secular optimistic theodicy. They believe that
the enhancements that will become possible will improve
the quality of life sufficiently that life will be not merely
better but good. We might say they have faith in the salvific
or redemptive powers of enhancement. Humans may not
| 6 1
have fallen but they are nonetheless low. The good news,
though, on this view, is that things can get much better in a
future messianic era of enhancement.24
When this view is criticized as being too optimistic
the criticism is usually that the hoped-for enhancements
are unlikely to be achievable (or achievable within the projected time-frame).25 The suggestion is that advocates of
enhancement have an exaggerated view about what kinds
of enhancement will be possible. According to this criticism it is navely optimistic to think, for example, that
major life span extension is possible or that human cognitive capacities could be radically enhanced.
Even if we assume, however, that transhumanism is
not overly optimistic in this regard, it is unduly optimistic in another way. It assumes that the quality of life after
the anticipated enhancements would be good (enough).
This assumption is problematic. While the quality of life
would be better, it is not clear that it would be good enough
to count as good.26 For example, it would be better to live
for much longer in good health, and it would be better if
we knew much more than we do, but even lives enhanced
in these and other ways would be far from the ideal. We
would still die and we would still have vastly more ignorance than knowledge.
The relative force of the two charges of optimism is interactive. The more ambitious the claims about what improvements can be made, the more susceptible these claims are to
the first kind of objectionnamely, that the projections are
overly optimistic. On the other hand, the more modest the
claims about what can be achieved, the more susceptible the
view is to the charge that enhancement is merely a mollification of lifes harshness and not the promise of Eden.
62 |
THE RISK OF SERIOUS HARM

I have argued that the quality of all human lives is very
bad. If every time we create a new human life we are creating a being with a very poor quality of life, there is a strong
presumptive case against creating any such lives. Perhaps
there could be rare exceptional situations in which this
presumption could be defeated and one would be justified
in inflicting that kind of harm. However, if there really
were any such cases they would be very rare.27
Not all quality-of-life arguments for anti-natalism
must claim that the quality of every life is very bad. Some
such arguments are based on a more limited claimthat
bringing people into existence puts them at risk of serious harm. Terrible things can befall people. Any child you
bring into existence could be assaulted, raped, tortured, or
murdered. It could be sent to war. It could be kidnapped,
abducted, imprisoned, or executed. It could, because of
a spinal injury, a stroke, or a degenerative neurological
condition, become paralyzed. It could suffer bad burns or
some other mutilation or disfigurement. It could succumb
to a virus or a malignancy or any of thousands of other
conditions.
Perhaps merely naming these conditions does not
convey the horror of them. Think, then, about the kinds of
suffering they precipitate.28 Rape,29 for example, can instill
terror in the victim before and while she or he is violated.
Physical injury, including bruising and laceration, is not an
uncommon consequence of the assault. There can be lifelong psychological repercussions, including rage, shame,
feelings of worthlessness, and difficulties with intimacy.
A pregnancy can result if the victim is a fertile female.
| 6 3
Even where abortions are freely available there can be psychic trauma in terminating the pregnancy. Carrying the
fetus to term can be psychologically even more distressing. Rape victims can also contract sexually transmitted
diseases from their assailants. These in turn have not only
many harmful physical effects, but can cause great mental
trauma as well.
Burn victims suffer excruciating pain, not only in the
moment but also for years thereafter. The wound itself is
obviously painful, but the treatment intensifies and protracts the pain. One such victim describes his daily bath
in a disinfectant, which would sting intact skin but causes
unspeakable pain where there is little or no skin. The bandages stick to the flesh and removing them, which can take
an hour or more if the burns are extensive, causes indescribable pain.30 Repeated surgery can be required, but
even with the best treatment the victim is left with lifelong
disfigurement and the social and psychological difficulties
associated with this.
Consider next those who are quadriplegic or, worse
still, suffering from Locked-in Syndrome. This is sheer
mental torture. One eloquent ALS sufferer describes this
disease as progressive imprisonment without parole31
on account of the advancing and irreversible paralysis.
Dictating an essay at the point he had become quadriplegic, and before losing the ability to speak, he describes his
torments, which are most acute at night. When he is put
to bed he has to have his limbs placed in exactly the position he wants them for the night. He says that if he allows
a stray limb to be misplaced or fails to insist on having
[his] midriff carefully aligned with legs and head he will
suffer the agonies of the damned later in the night.32 He
64 |
invites us to consider how often we shift and move during

the course of a night, and he tells that enforced stillness
for hours on end is not only physically uncomfortable but
psychologically close to intolerable.33 He lies on his back in
a semi-upright position, attached to a breathing device and
left alone with his thoughts. Unable to move, any itch must
go unscratched. His condition, he says, is one of humiliating helplessness.34
Cancers reputation as a dreaded disease is well deserved. There is much suffering in dying from this disease,
but at least as much in the treatments that are usually necessary to cure the patient of the malignancy. In the worst
scenarios the patient suffers both from the treatment and
from its failure.
Where symptoms have not precipitated the diagnosis,
the first blow is the diagnosis itself. Arthur Frank says that
on receiving the news that he had a malignancy, he felt as
though his body had become a quicksand in which he was
sinking.35 But that is only the beginning. For example, radiation treatment for esophageal cancel left Christopher
Hitchens desperately attempting to avoid the inevitable
need to swallow. Every time he did swallow, a hellish tide
of pain would flow up [his] throat, culminating in what felt
like a mule kick in the small of [his] back.36 Ruth Rakoff,
after receiving chemotherapy for breast cancer described
her insides as raw.37 Treatment can result in nausea, vomiting, both constipation and diarrhea, and gum and dental
soreness. Food tastes bad and appetite is lost. Unsurprisingly, all this results in weight loss and fatigue. Neuropathy is another common side effect, as is hair loss. Many
of the same symptoms can be experienced even in the absence of treatment or once treatment has been abandoned.
| 6 5
Moreover, tumors pressing on brains and bowels and

bones can cause excruciating pain. Where the pain can be
controlled this is sometimes at the expense of consciousness or at least lucidity.
Pain accompanies many conditions, but we should remember that much of it is not attendant upon visible conditions. It is often hidden from those not experiencing it.
One sufferer from chronic pain describes it as debilitating and observes that it can take over ones life, sap ones
energy, and negate or neutralize joy and well-being.38
Not all suffering is physical, although psychological ailments can certainly have bodily sequelae. William Styron,
describing his depression, says that ultimately, the body
is affected and feels sapped, drained.39 He speaks of his
slowed-down responses, near paralysis, psychic energy
throttled back close to zero.40 Sleep is disrupted, with the
sufferer staring up into yawning darkness, wondering
and writhing at the devastation41 of his mind. The sufferer from depression, we are told, is like a walking casualty of war.42
To bring a new person into existence is to create a
being that is vulnerable to these and thousands of other
kinds of appalling suffering. To procreate is thus to engage
in a kind of Russian roulette, but one in which the gun
is aimed not at oneself but instead at ones offspring. You
trigger a new life and thereby subject that new life to the
risk of unspeakable suffering. Even those who are not antinatalists should be alarmed at how little thought seems to
be given to this by the overwhelming majority of procreators. However, the risks of these harms should not merely
give potential procreators pause. It should stop them from
breeding.
66 |
Against this conclusion, a number of defenses of procreation might be offered. However, the most foundational
of these argues that we need to distinguish between the severity of the harm and the magnitude of the riskthat is,
between how bad a harm is and how likely it is to eventuate. This argument can grant that the sorts of harms I have
mentioned are unspeakable and yet claim that because the
chances of them befalling any child are (often) sufficiently
small, potential procreators are justified in reproducing.
Procreation, on this view, is simply not as risky as Russian
roulette (even if the worst harms that can result from each
are equally bad).
Relatively wealthy people living in comfortable conditions in developed liberal democracies are inclined to say
that their children are very likely to lead good lives. Their
children, they think, are unlikely to succumb to poverty,
oppression, or political violence. They will have good nutrition and access to education and good healthcare. These are
comforting thoughts but they are far too credulous.
First, they ignore the fact that knowledge of the optimism bias provides us with excellent reason to distrust intuitions about the level of risk. To see why this is so, it may
help to look at humans who, by the standards of privileged
countries, have led or are leading very poor-quality lives.
These include people today who are living in extreme poverty or under repressive regimes or in politically unstable or
failed states. It also includes humans through most of history, when, for example, infant mortality was high, life expectancy was short, repression and cruelty were even more
rampant than they are today, and there were no anesthetics.
Consider the last of those for a moment. Before the discovery of (effective) anesthetics,43 creating a new person
| 6 7
was to create a being vulnerable to the horrors of surgical

procedures without the benefit of anesthesia.44 The alternative to the excruciating pain and torment of such operations45 was, for example, to endure the chronic pain from
the un-excised tumor or to die from a gangrenous limb
that was not amputated. It is salutary to consider that in
a time of these horrors humans still saw fit to create new
sentient beings vulnerable to such unspeakable suffering.
At least in cases where their procreation was the result of a
decision, they seem to have found the risks acceptable. This
should lead to grave doubts about the human capacity to
think carefully about these matters. If people are inclined
to think that the risks of harm to their prospective children are acceptable almost irrespective of how bad those
childrens lives are likely to be, we have very good reason
for distrusting peoples judgments in this regard.
Second, let us assume, for the sake of argument, that it
is true that the chances of something very bad happening
to the average personor even the average person in more
privileged societiesare small. There may still be good
reason to be risk averse in this context. The basic axiological asymmetry discussed in chapter 2 is one reason. When
considering the interests of the prospective child, there is
nothing to be lost by desisting from bringing it into existence. There is however a potentially very serious cost if the
created person suffers in one of the ways I have mentioned.
A second reason to be risk averse arises from the empirical
asymmetries between harms and benefits that I enumerated earlier in this chapter. For example, the worst things
in life are worse than the best things are good and the bad
things tend to last longer than the good things. There is
therefore more reason to avoid the bad things even if we
68 |
grant that this comes at the cost of not attaining the good
things.
As it happens, however, the chance of very bad things
happening to any person one creates is actually far from
remote. Indeed, the chance is actually alarmingly high. Just
how high it is does depend, of course, on where one sets the
threshold for what counts as very bad. However, one does
not have to set that bar impossibly low to reach a pessimistic conclusion. Cancer, for example, is a plausible candidate
for something that is very bad. In the United States, it
has been estimated that one in two men and one in three
women will develop cancer, and one in four men and one in
five women will die from it.46 It has recently been suggested
that estimates of lifetime risk of developing cancer may be
exaggerated by the fact that some people develop cancer
more than once. However, even if we opt for the more
conservative estimate of lifetime risk of first primary, we
find that forty percent of men and thirty-seven percent of
women in the United Kingdom will develop cancer.47 Those
who do not get cancer are still at risk of hundreds of other
possible causes of suffering. Thus (even) in privileged countries such as the United States and the United Kingdom
any prospective parents who propose to bring a child into
existence are subjecting that child to an extraordinarily
high risk of suffering some very bad condition.
It is, of course, more commonly older people who get
cancer.48 However, although it is, all things being equal,
worse to die when one is younger than when one is older,
the physical and psychological symptoms of life with cancer
and dying from cancer are no less appalling at older ages.
Furthermore, the Grim Reaper wields a two-edged
sword, cutting down both young and old. The swiftest,
| 6 9
most painless deaths, ending lives with the least ill health,
are typically those of people cut down before old age, often
in their prime. They are sudden deaths by injury, stroke, or
heart attack, for example. Those who live into old age are
more likely to suffer from the decline and decrepitude that
accompanies advancing age. Thus even when the elderly
do die suddenly, it is typically after suffering from some
(other) age-related cause. There are a relatively charmed few
who live in excellent health and full vigor and then die suddenly in advanced age. However, there are very few such
people. The result is that most of those who avoid the suffering that tends to precede or accompany death at older
ages are among those who die young, which is another kind
of very bad thing. Many have the worst of bothpainful
deaths at a young age.
Thus whether one is speaking about potential procreators in deprived circumstances, whose offspring are likely
to die young, or potential procreators in more privileged
circumstances, whose offspring are likely to die at older
ages, the chances of terrible things befalling the offspring
are very high.
Perhaps some especially brazen procreators in (relatively) privileged circumstances will acknowledge that
their children have a high risk of terrible suffering but say
that if that suffering comes late enough in life and is preceded by enough good then the suffering will be worth it.
This sort of defense of procreation seems callous. It fails to
appreciate just how bad it is to suffer from a condition like
cancer. It also ignores a very important feature of procreation, namely that although it is the parents who decide
to procreate, it is their children who pay the price of being
brought into existence. It is one thing to assume high risks
70 |
of terrible harm for oneself. It is quite another thing to

assume those risks for others, even if those others are ones
children.49
It is, of course, an immutable fact of procreation that
one cannot obtain advance consent from the being brought
into existence. Thus some parents might seek to defend
their procreation by observing that they could not have
sought their childrens consent before they brought them
into existence. Such a defense is the wrong response to
the impossibility of consent. The appropriate response is
instead to desist from imposing the risks and harms by desisting from procreation.
Seana Shiffrin does not go so far as to say that the risks
and harms in life make procreation wrong, although it is
not clear why she does not go that far given that her argument seems to yield that conclusion.50 Instead she adopts
what she calls an equivocal view about routine procreation.51 This is the view that procreation is intrinsically
and not just epistemically a morally hard case. This, she
says, is because procreation involves imposing serious
harms and risks on someone who is not in danger of suffering greater harm if one does not act.52
Her argument does not assume that coming into existence has no benefits. She is willing to assume, at least
for the sake of argument, that it can benefit the person
brought into existence. However, she points to a (deontic)
asymmetry between harms and benefits: It is permissible
and perhaps even obligatory (in the absence of a persons
wishes to the contrary) to inflict a lesser harm on somebody in order to save that person from a greater harm, but
it is not permissible to inflict a harm on somebody in order
to bestow a greater (pure) benefit on that person.
| 7 1
This, she says, is why it would be permissible to break

somebodys arm if that were necessary to save his life, but it
would not be permissible to break somebodys arm in order
to bestow some benefit such as supernormal memory, a
useful store of encyclopedic knowledge, 20 IQ points worth
of extra intellectual ability, or the ability to consume immoderate amounts of alcohol or fat without side effects.53
Creating somebody always causes that person harm
the harms of existenceand we cannot justify the infliction of those harms because of the purportedly greater
benefits of existence.
Professor Shiffrin denies that we can presume hypothetical consent in order to justify procreation. She provides four reasons for this: (a) the person is not harmed if
we do not bring him or her into existence; (b) the harms of
existence may be severe; (c) the harms of existence cannot
be escaped without considerable cost; and (d) the hypothetical consent is not based on the individuals values or
attitudes toward risk.54
These are all important considerations, but in the context of the quality-of-life argument it is worth highlighting
the third. Some crass optimists attempt to justify procreation in the face of the risks by pointing to the option of
suicide. Their argument is that if the person brought into
existence finds the quality of his or her life to be unacceptably low there is a way of opting out of existence.
These really are shallow words that belie the depth of
suffering that precedes and precipitates suicide. The life
drive is immensely powerful even when people are enduring unspeakable suffering. This is one reason why suicide is
by no means an easy option. Another is that suicide causes
great distress to the family and friends of the person who
72 |
takes his or her own life. Thus any (rational) person considering suicide has to consider that the contemplated action
would inflict great tragedy on others. One cannot be glib
about that.
The serious costs of suicide stand in stark contrast to
never coming into existence. Whereas taking ones own
life has costs to oneself, never existing has no costs to the
person who is never brought into existence. Childlessness
can be a source of pain, but, unlike suicide, it is not a pain
typically inflicted on others. Quite the opposite: avoiding
the distress of childlessness is usually achieved by inflicting great harm on othersthe children brought into existence. Moreover, it is much worse to lose a loved one to
suicide than for that loved one never to have come into
existence.55
Ceasing to exist and never coming into existence are,
unsurprisingly, asymmetrical. Existence is a terrible business but never existing is immeasurably better than ceasing to exist.
NOTE S
1. These findings are mentioned by David G. Myers and Ed
Diener, The Pursuit of Happiness, Scientific American 274,
no. 5 (May 1996): 7072. See also Angus Campbell, Philip
E. Converse, and Willard L. Rodgers, The Quality of American
Life: Perceptions, Evaluations, and Satisfactions (New York:
Russell Sage Foundation, 1976), 25.
2. Frank M. Andrews and Stephen B Withey, Social Indicators of Well-Being: Americans Perceptions of Life Quality (NY:
Plenum, 1976), 334.
3. See Shelley E. Taylor, Positive Illusions: Creative Self-Deception
and the Healthy Mind (New York: Basic, 1989); Margaret W.
| 7 3
Matlin and David J. Stang, The Pollyanna Principle: Selectivity

in Language, Memory, and Thought (Cambridge, MA: Schenkman, 1978).
4. This evidence is reviewed by Shelley Taylor and Jonathon
Brown, Illusion and Well-Being: A Social Psychological Perspective on Mental Health, Psychological Bulletin 103, no. 2
(1998): 193210.
5. Ronald Inglehart, Culture Shift in Advanced Industrial Society
(Princeton NJ: Princeton University Press, 1990), 241246.
6. Ibid., 242.
7. Ibid., 246.
8. Perhaps it will be argued that adapting to ones paralysis
does constitute an improvement in ones objective condition. Some might respond to this objection by saying that
it ignores the distinction between the objective condition
paralysis in this caseand how one subjectively reacts to
the objective condition. However, one can reject the objection even if one concedes that a feedback loop is possible,
such that ones subjective assessment can, to some extent,
actually affect ones objective condition. More specifically,
one can concede that the feedback loop leads to some improvement in ones objective condition, but as long as one
remains paralyzed, ones objective condition is considerably
worse than ones subjective assessment may recognize.
9. Richard A. Easterlin, Explaining Happiness, Proceedings of
the National Academy of Sciences 16, 100, no. 19 (September
2003): 1117611183.
10. See, for example, Joanne V. Wood, What Is Social Comparison and How Should We Study It? Personality and Social Psychology Bulletin 22, no. 5 (1996): 520537.
11. For more on this see Jonathon D. Brown and Keith A.
Dutton, Truth and Consequences: The Costs and Benefits
of Accurate Self-Knowledge, Personality and Social Psychology Bulletin 21, no. 12 (1995): 1292.
12. It is true, of course, that some people suffering intolerable
conditions come to believe that it would be better if their
lives were to end, but even many of them cling to the belief
that they were not harmed by being brought into existence.
The rationale here is typically that it was worth coming into
74 |
existence for that period of their life that preceded the onset
of whatever condition made their continued life intolerable.
13. One can soil oneself anywhere and thus this qualification is
necessary.
14. For a description of what this can feel like, see Patricia A
Marshall, Resilience and the Art of Living in Remission,
in Malignant: Medical Ethicists Confront Cancer, ed. Rebecca
Dresser (New York: Oxford University Press, 2012), 94.
15. Arthur Schopenhauer, On the Sufferings of the World, in
Complete Essays of Schopenhauer, trans. by T. Bailey Saunders, vol. 5: Studies in Pessimism (New York: Wiley, 1942), 2.
16. Henceforth I shall not distinguish between desires and
preferences as similar observations apply to both.
17. Abraham Maslow, Motivation and Personality, 2nd Edition
(New York: Harper & Row, 1970), xv.
18. Similar points can be made about moral goodness, aesthetic
experience, and other capacities and traits.
19. It should be noted that although religious faith can be optimistic, it is not always so. There are pessimistic religious
views too.
20. Perhaps it will be suggested that these pains are byproducts of the instrumental value of pain in other contexts. If
that is true then our lives would be better if pain were present only when it had instrumental valuethat is, if there
were no spillover into cases where pain has no instrumental
value.
21. In the case of the reflex arc, pain accompanies reflexive
aversive behavior, but the pain plays no mediating role.
22. This is, of course, a variant of the stoical motto No pain,
no gain. Insofar as this motto is true, it is an unfortunate
truth. (I am reminded here of the alternative motto for
those less sanguine about pain: No pain ... no pain.)
23. Among those who have offered this version of the argument
are: Thaddeus Metz, Are Lives Worth Creating? Philosophical Papers 40, no. 2 (July 2011): 252253; and David DeGrazia, Is It Wrong To Impose the Harms of Human Life? A
Reply to Benatar, Theoretical Medicine and Bioethics 31, no. 4
(2010): 317331.
| 7 5
24. It should be obvious that advocates of enhancement do not

use the religious language I have included in scare quotes. I
am using that language to highlight the parallels.
25. The qualification is important for the following reason: If
human life is not worth creating in the absence of the enhancements but the enhancements were of a sufficient magnitude to make life worth starting, it would be difficult to
justify procreation if it would take a very long time for the
enhancements to be brought about. The longer it takes to
bring about the necessary enhancements, the longer people
are creating lives that are not worth starting.
26. Given the psychological phenomenon of comparison described above, such a life is likely to look much better to us
than it actually is.
27. In Better Never to Have Been, 182193, I considered whether
limited procreation might temporarily be permissible as
part of a process of phasing human extinction in order to
reduce the number of people who would suffer the costs of
being the final generation. I suggested that on some views
this might be permissible but that on other views it would
not be.
28. In his response to what I say here, David Wasserman contends that not everybody views the fates I shall describe
as terrible. Perhaps there are some people who take their
lives to have an acceptable quality despite the conditions I
shall describe. It seems obvious to me that such people are
deeply in the grip of coping mechanisms. Of course I do not
begrudge them any comfort they can scrounge from their
situation. It is quite another matter to think that their conditions are not appalling. After you have read David Wassermans response, return to the descriptions I provide
here and ask yourself whether you can, with a straight face,
say that prospective parents are morally entitled to create
beings who stand a significant risk of suffering at least one
of these fates.
29. One rape victim observes that imagining what it is like to
be a rape victim is no simple matter, since much of what a
victim goes through is unimaginable. See Susan J. Brison,
76 |
Aftermath: Violence and the Remaking of a Self (Princeton, NJ:

Princeton University Press, 2002), 5.
30. Dan Ariely, Predictably Irrational: The Hidden Forces That
Shape Our Decisions, Revised and Expanded Edition (New
York: HarperCollins, 2009), xxiiixxiv.
31. Tony Judt, The Memory Chalet (New York: Penguin, 2010),
15.
32. Ibid., 17.
33. Ibid.
34. Ibid., 20.
35. Arthur Frank, At the Will of the Body: Reflections on Illness
(Boston: Houghton Mifflin, 1991), 27.
36. Christopher Hitchens, Mortality (New York: Twelve, 2012),
67.
37. Ruth Rakoff, When My World Was Very Small (Toronto:
Random House Canada, 2010), 99.
38. Philip A. Pizzo, Lessons in Pain ReliefA Personal Postgraduate Experience, New England Journal of Medicine 369,
no. 12, (September 2013): 1093.
39. William Styron, Darkness Visible: A Memoir of Madness (New
York: Random House, 1990), 47.
40. Ibid.
41. Ibid., 49.
42. Ibid., 62.
43. The first use of ether as an anesthetic was in 1842. See
Harvey Graham, Surgeons All (London: Rich & Cowan,
1939), 324. Prior to this date various attempts at dulling
the senses were used but none successfully prevented the
patient from feeling pain.
44. The same applies to dental procedures, although it has been
claimed that dental pain was most widespread from the late
seventeenth century until the discovery of anesthetics. See
N.W. Kerr, Dental Pain and Suffering Prior To the Advent
of Modern Dentistry, British Dental Journal 184, no. 8
(April 1998): 397399.)
45. For a first-person account of enduring such an experience
see Frances Burneys account of her mastectomy, performed
while she was fully conscious. (Part of her account is quoted
in Roy Porter, The Greatest Benefit to Mankind: A Medical
| 7 7
History of Humanity from Antiquity to the Present (London:

HarperCollins, 1997), 365.)
46. American Cancer Society. Lifetime Risk of Developing
or Dying from Cancer. http://www.cancer.org/cancer/
cancerbasics/lifetime-probability-of-developing-or-dyingfrom-cancer (Accessed October 2, 2013).
47. Cancer Risk UK, Lifetime Risk of Cancer, http://www.
cancerresearchuk.org/cancer-info/cancerstats/incidence/
risk/statistics-on-the-risk-of-developing-cancer (Accessed
October 6, 2013).
48. Older is a relative term. There are many children, young
adults, and middle-aged ones who suffer from cancer, but
septuagenarians, for example, are more likely to get cancer
than children, young adults, and the middle aged.
49. Nor does it become acceptable to have assumed those risks
for ones children if those offspring, as a result of the coping
mechanisms I have discussed, also come to think that the
suffering is worth it. Prospective parents do not know if
their children will reach this conclusion and it is indecent
to assume that they will. In any event, the subjective assessments that the suffering is worth it are unreliable for all
the reasons I have mentioned.
50. I discuss her argument at greater length in Better Never to
Have Been, 4954.
51. Seana Valentine Shiffrin, Wrongful Life, Procreative Responsibility and the Significance of Harm, Legal Theory 5,
no. 2 (1999): 136.
52. Ibid.
53. Ibid., 127.
54. Ibid., 131133.
55. If, for some people, it is not worse, it still should be worse.
The Misanthropic Argument
SOME ARGUMENTS FOR THE CONCLUSION that it is
(always) wrong to bring somebody into existence are philanthropic arguments. They are rooted in a concern for the
welfare of those who would be brought into existence. According to these arguments, coming into existence is such
a serious harm or carries such a severe risk of serious harm
to those people brought into existence that we should
desist from creating them.
However, philanthropy is not the only route to anti-
natalism. There are also anti-natalist arguments that we
can characterize as misanthropic. These arguments focus
on the terrible evil that humans wreak, and on various
negative characteristics of our species. This chapter will
be devoted to advancing what I take to be the strongest of
these argumentsa moral argument.1
Misanthropic anti-natal arguments are likely to be met
with an even more hostile reaction than are philanthropic
anti-natal arguments. It is not hard to see why this is the
T he M isanthropic A rgument
| 7 9
case. First, we tend to dislike those who hate us. The misanthrope is, purportedly, one who hates humans and thus
it is unsurprising that the misanthrope is disliked. Second,
people do not like to hear bad things about themselves and
the misanthrope has lots of bad things to say.
A few comments may be offered to mitigate this instinctive response. First, the arguments that will be advanced here are misanthropic only in the sense that they
point to unpleasant facts about humans. Accepting these
arguments does not commit one to hating humans. Indeed,
I shall argue that the misanthropic arguments are not incompatible with the philanthropic ones. Thus the characterization of the arguments as misanthropic should not be
taken too literally or overinterpreted.
Second, the unpleasant claims that the misanthropic
arguments make about humanity may well be true. To
refuse to believe them merely because they are unpleasant
would provide the misanthrope with further grounds for
complaint. Failing to acknowledge ones flaws is itself another flaw.
The strongest misanthropic argument for anti-natalism
is, I said, a moral one. It can be presented in various ways,
but here is one:
1. We have a (presumptive) duty to desist from
bringing into existence new members of species
that cause (and will likely continue to cause) vast
amounts of pain, suffering, and death.
2. Humans cause vast amounts of pain, suffering,
and death.
3. Therefore, we have a (presumptive) duty to desist
from bringing new humans into existence.2
80 |
I shall delay discussion of the first premise until later, and

shall begin now by demonstrating the truth of the second
premise. I do so in two stages. First I highlight the dark side
of human nature; then I show just how harmful humans
are. The two are connected because the dark side of human
nature partially explains why humans are so harmful.
More specifically, it explains how, in certain situations, the
dark side manifests, with destructive results. I shall provide more detail than some might think necessary. I do
so because some people are inclined to underestimate the
extent of human destructiveness and I need to forestall
that sort of glib response.
H U M A N N AT U R E T H E D A R K S I D E
Our species is prone to a flattering view of itself. Humans
have regarded themselves as the pinnacle of creation,
formed by and in the image of an omnibenevolent, omniscient, and omnipotent God, and inhabiting a planet at the
center of the universea planet around which all others
revolve.3 Science has done much to debunk some of these
ideas. We now know that our planet is not at the center
of the universe: the earth revolves around the sun rather
than vice versa. And we knowor at least some of us do
that we are johnny-come-lately products of a long, blind
evolutionary process.
However the inclination toward self-adulation is remarkably resilient and it simply manifests differently in
the scientific paradigm than it does in the religious one.
Thus in our taxonomy of species we designate ourselves as
Homo sapiensthe thinking human. There is, of course,
| 8 1
some truth in this designation. As a species we do think

more than other animals, and we have greater technical capacity than they do. The human elite has had some remarkable achievements. However, we would do well to note that
these achievements are thought to be remarkable only because (a) they are not within the reach of most humans;
(b) the cleverest humans typically produce them only by
pushing the limits of their capacity; and (c) there are no
more cognitively capable species on the planet to put our
achievements into a humbling perspective.4 There is thus
something unfair about judging our entire species by the
achievements of its elite. Even the cognitive capacities of
the elite are massively deficient in countless ways.
We fancy ourselves as rational beings, but there is ample
evidence that we regularly fall far short of thinking and
acting rationally.5 For example, we have instincts to make
intuitive judgments that, on reflection, we can see to be
mistaken. But we are also often too lazy to do the necessary
reflection. Our decisions can be influenced by the framing
effectthat is, our decisions are likely to differ depending on whether the same information is presented one way
or another. In states of sexual arousal we make decisions
that we know to be irrational when we are not aroused. We
are willing to pay more to retain something than we are to
obtain something of the same value even when there is no
reason to be attached to the object we already ownthe socalled endowment effect. Humans also have a tendency to
be overly optimistic and we have considerable capacity for
self-deception. And these are but a few of hundreds of possible examples that could be provided.
For all the thinking that we do we are actually an
amazingly stupid species. There is much evidence of this
82 |
stupidity. It is to be found in those who start smoking cigarettes (despite all that is known about their dangers and
their addictive content), and in the excessive consumption
of alcohol and especially in those who drive while under
its influence. It is to be found in the achievements of the
advertising industry, which bear ample testament to the
credulity of humanity. It is also to be found in the successes
of political sloganeering, demagoguery, and spin, to which
billions of people fall prey. The seriousness with which so
many people take matters of utter inconsequencesuch as
sport and the vicissitudes of particular sport teamsand
the popular adulation of shallow, dysfunctional sports,
music, and film stars, are also items of evidence.6 Further
signs are to be found in the fads and fashions and delusional obsessions that run rampant.7
Our cognitive and other deficiencies are troubling in
their own right, but some of these deficiencies, predictably, also incline us to various moral failings. These failings explain or partially explain some of the terrible things
humans do and which form the basis for the strongest versions of the misanthropic argument against creating new
people.
Consider, for example, the human tendency toward
conformity. In one influential study8 demonstrating this
phenomenon, groups of subjects were shown a linethe
standardand then asked which of three other lines was
the same length as the standard. The nonequivalent lines
were of sufficiently different length that the correct answer
was clear. In each group of subjects all but one were confederates of the experimenter and they were instructed,
in some conditions of the experiment, to give the wrong
answer. A significant number of the individuals who were
| 8 3
the actual experimental subjects yielded to the majoritys

answer. Subsequent analysis has confirmed these findings
but also shown that the extent of conformity is influenced
by cultural variables.9
The studies have also shown that the extent of conformity is influenced by the degree to which the stimulus
was ambiguous. The less clear the correct answer is the
more likely people are to conform to the majority view. We
should thus expect that when we shift from simple factual
matters such as the length of lines to more complicated
matters, including evaluative ones, people will be even
more likely to conform. That is to say, when everybody is
admiring the legendary emperors new clothes, it is less
likely that the lone individual will announce that the emperor is not wearing any clothes. We know how dangerous
conformity can be in certain circumstances. One context
in which it has manifested is that of witch hunts. Judging
by the actual incidence of witches (where witch is understood the way witch hunters understand it), a witch hunt
should be about as successful as a unicorn hunt. Yet tens of
thousands of purported witches were found and killed
between 1450 and 1700.10 There have been sporadic witch
hunts since then, including in our own times.11
Humans also have a propensity to obey authority and
will often do so even when they are asked to do terrible
things.12 Most people have difficulty believing that they
would be among those who would obey orders to commit
atrocities. While it is true that there are some people with
the strength to resist authority where it is appropriate to
do so, it is not the case that everybody who thinks that they
fall into that category is as exceptional as they think they
are. Indeed in the famous psychological experiments that
84 |
demonstrated the tendency to obey authority, some of the

subjects who had thought well of themselves were shocked
to find they had followed orders. One of them began to
call himself Eichmann,13 a reference to Adolf Eichmann,
whose defense at his trial in Jerusalem was that he was
obeying orders.14
An even more graphic experimental example of how
ordinary people can quickly descend into barbarism is
that of the Stanford prison experiments.15 In this experiment twenty-four healthy student volunteers were randomly assigned to the roles of either guards or prisoners
in a faux prison located in the basement of the Stanford
psychology building. Both groups adapted very rapidly
to their respective roles, with the authoritarian guards
humiliating and psychologically torturing the prisoners. The treatment of the prisoners got so bad that the
experiment had to be terminated prematurely, after a
mere six days.
I have pointed to evidence that humans are neither as
clever nor as good as they often think they are. None of this
is to deny that there are also positive features to human
nature. For example, we can (even if we do not always) employ
reason to a greater degree than other animals. And we can
feel empathy and act on it (as some animals also do). I have
focused on the negative, not to deny the existence of the positive, but instead to highlight what is ignored in the general
self-conception of our species.16 Moreover, the dark side is
arguably the more primitive. To avoid its manifestation, considerable effort has to be expended in educating and training
people and in constructing and maintaining circumstances
and institutions that inhibit the dangerous lapses to which
people are prone.
| 8 5
HOMO PERNICIOSUS
Humans may exceed other animals in their sapient capacities, but we also surpass other species in our destructiveness. Many animals cause harm, but we are the most lethal
species ever to have inhabited our planet. It is revealing
that we do not refer to this superlative property in identifying ourselves. There is ample evidence that we are Homo
perniciosusthe dangerous, destructive human.17
In what follows, I shall first demonstrate how much
harm humans do. I shall consider three categories of such
harm: harm to other humans, harm to animals, and harm
to both humans and animals via harm to the environment.
While it is obviously impossible to provide a full catalogue
of human destructiveness, I do plan to survey a wide range
of types and provide some examples.
Inhumanity to Humans
Humans have harmed other humans for as long as there have
been humans. The earliest destruction was on a relatively
small scale, not least because there were so few humans at
the beginning of the speciess history. The harms inflicted
were, most likely, assault and murder committed by individuals or small groups, with the victims being either individuals or other small groups. In other words, the totality of
the destruction would have been very similar to that seen in
some species of nonhuman primates today.
Although humans continue to inflict such harms, when
we think today of the destruction that humans wreak, we
are more likely to think first of much larger-scale destruction. Humans have killed many millions of other humans
86 |
in war and in other mass atrocities, such as slavery, purges,

and genocides.
The number has been increasing, partly because of the
burgeoning number of humans that there are to kill, and
partly because humans destructive capacity has increased
so significantly. That said, it is alarming just how lethal
primitively armed humans can be.
Many hundreds of millions have been murdered in
mass killings. In the twentieth century, the genocides
include those against the Herero in German South West
Africa; the Armenians in Turkey; the Jews, Roma, and
Sinti in Germany and Nazi-occupied Europe; the Tutsis
in Rwanda; and the Bosnian Muslims in the former Yugoslavia. Other twentieth century mass killings were those
perpetrated by Mao Zedong, Joseph Stalin, and Pol Pot
and his Khmer Rouge. But these mass killings were by no
means the first. Genghis Khan, for example, was responsible for killing 11.1% of all human inhabitants of earth
during his reign in the thirteenth century.18
The gargantuan numbers should not obscure the gruesome details of the how these deaths are inflicted and the
sorts of suffering the victims endure on their way to death.
Humans kill other humans by hacking, knifing, hanging,
bludgeoning, decapitating, shooting, starving, freezing,
suffocating, drowning, crushing, gassing, poisoning, and
bombing them. Sometimes the victims are killed one at a
time and sometimes they are killed en masse in a single
action. Although the killing is sometimes at a distance
where the suffering can be obscured from the killer, at
other times it is up closethe killer, covered in the blood
and splattered brains of his victims, continues on his destructive path through further victims.
| 8 7
Mass killings are obviously not the only form of destruction wrought by humans. There are smaller-scale killings and there are various barbarities other than killing.
Humans rape, assault, flog, maim, brand, kidnap, enslave,
torture, and torment other humans. Brutal punishments
are inflicted on people, sometimes for real crimes but sometimes merely because of their religious or political views,
their race or ethnicity, or their sexual orientation or practices. There are so-called honor killings and mutilations
for perceived or suspected violations of rigid codes. And
humans have performed human sacrifices to their deities.
It is hard to fathom the depth and variety of the barbarism. Consider, for example, the case of Ren de Permentier, a Belgian officer in the Congo in the 1890s:
He had all the bushes and trees cut down around his house
... so that from his porch he could use passersby for target
practice. If he found a leaf in a courtyard that women prisoners had swept, he ordered a dozen of them beheaded. If he
found a path in the forest not well-maintained, he ordered a
child killed in the nearest village.19
Or consider what was done to Ahmad Qabazard, a

nineteen-year old Kuwati detained by the Iraqis. His parents were advised that he would soon be released. When
they heard a car approaching, they went to the door:
When Ahmad was taken out of the car, they saw that his
ears, nose and genitalia had been cut off. He was coming
out of the car with his eyes in his hands. Then the Iraqis
shot him, once in the stomach and once in the head, and
told his mother to be sure not to move the body for three
days. 20
88 |
Militiamen in Congo cut the flesh from living victims and

force them to eat it, a practice known, macabrely, as autocannibalism.21 Other practices include cutting a fetus out
of a womans uterus and then making her friends eat it, and
inserting the end of an AK-47 rifle into a womans vagina
and then pulling the trigger.22 Annually, fighters from the
Lords Resistance Army would club hundreds of people to
death as they raided villages and kidnapped children.23
Kidnapping of children in such contexts is the first
stage in making them child soldiers. Sometimes they are
forced to kill members of their own family24 or others, typically in gruesome ways. In one case a boy was told to pound
to death the baby of a woman he knew.25 If the inductees
refuse orders they are beaten savagely or even killed. Indoctrination is another component of their training. It
is estimated that there are currently about three hundred
thousand child soldiers in conflicts in Asia, Africa, the
Americas, and elsewhere.26
In other situations those who are kidnapped are sold
into slavery. They are torn away from their families and
sometimes shipped great distances, often in fetid, crowded
conditions, in which many die. They are subjected to savage
beatings, rape, and other indignities. Nor does the slaves
commercial value mean that slaves were not killed. In one
horrendous case, 133 live slaves were thrown overboard
on the orders of the captain, who had insured them for
30 each.27
Some people think that slavery is no longer practiced.
However, it persists even in some jurisdictions where it is
illegal.28 In some places, young girls are still sold into sexual
slavery. One young Cambodian girl, Long Pross, who was
kidnapped and forced into prostitution, has related how
| 8 9
she was beaten and subjected to electric shocks. Two crude

abortions were performed on her. When the second left her
in great pain she pleaded to be able to rest. In response her
owner gouged out her right eye.29
Now it may well be suggested that terrible though all
these actions are, it is a minority of humans who actually behave in these ways. In response to this comforting
thought, a few other less comforting ones need to be considered. First, some of the serious harms humans inflict on
other humans are not as aberrational as one might think.
For example, there was a time when slave owning was
widespread. Slave merchants might have been a small minority, but slave owners were far more common. Rape remains widespread today. It is probably a minority of people
who are rapists, but it is not a negligible minority.
Second, even where people are not themselves perpetrators they often facilitate the atrocities committed
by others. For example, they might support the infliction of torture and cruel punishments, or policies that
discriminate against people on the basis of their race,
religion, sex, or sexual orientation, and vote for governments that implement such practices and policies. Sometimes large numbers of people endorse a worldview in
which honor killings thrive, or in which terrorists are
hailed as heroes. Sometimes the facilitation of evil results not from an endorsement of the evil but from stupidity, gullibility, dogma, or some other failing. Consider,
for example, so-called useful idiots, those well-meaning
people who support a cause without realizing how evil it
really is. Well-intentioned people in the West who sympathized with the Soviets are common examples. Many
of them would have been horrified by the brutality of
90 |
and the repression within the Soviet Union, but their navet blinded them to the realities of the Soviet regime.
Perhaps the most tragic situations are those in which
well-meaning people inadvertently cause even more suffering. For example, there is some evidence that western
media attention to amputations in Sierra Leone actually
encouraged further amputations by those seeking the
media attention. 30
Third, we should remember how easily ordinary people
can slip into contemptible behavior. One such scenario
can be found in crowds of people clambering in shops for
sale items or products in limited supply. In late 1998, the
Furby (a stuffed-toy pony) was the seasons hottest toy in
the United States, and customers were jostling to buy the
limited stock. One woman in the crowd was pushed into a
door, where her arm was badly bruised.31 In another shop
a thirteen-year-old girl reported that when she picked up a
Furby, a woman took her hand and chomped on it in order
to force her to let go of it.32 The problem of shopper crowd
violence is a recurring one.33 One year a Walmart employee
in Valley Stream, New York was trampled to death by shoppers who stormed into the store looking for bargains.34
Nor is this the worst kind of crowd behavior. Lynch
mobs, whose collective intentions are to kill, are notorious
examples. The members of such mobs have often been respectable members of society. In 1672, the De Witt brothers were lynched in The Hague. The mobs intention had
been to hang them but they were so viciously attacked
that they died before reaching the scaffold. The bodies were
then hung up by the feet, stripped bare, and literally torn
to pieces.35 The philosopher Baruch Spinoza was stunned
by these acts of barbarity, perpetrated not by some roving
| 9 1
band of thieves, but by a crowd of citizens that included

respectable middle-class burghers.36
Although this particular example and some of the
other evidence I have provided of atrocities is historical, it
certainly cannot be claimed that the worst human destructiveness is restricted to the past. I have provided plenty of
evidence of ongoing harm that humans do.37 Moreover, the
historical evidence is often pertinent to the present and
future. What people have done in the past provides good
evidence of the kinds of things people can do under certain circumstances. Sometimes relevantly similar circumstances reemerge. One of the reasons why the Holocaust
is so shocking is that it was conceived and implemented
by what had been thought to be so civilized a society. It is
unduly optimistic to think that civilization cannot backslide into barbarism. We saw earlier, when I described the
dark side of human nature, some of the features of the
human character that makes this possible. It is thus altogether too convenient to assume that there are only a few
evil people who do things of which the rest of humanity
is incapable. Sometimes, for example, it is only moral luck
that prevents somebody from becoming a genocidaire.
Fourth, human destruction comes in degrees and not
all of it involves the worst atrocities. There are many more
minor and sometimes quotidian harms that humans inflict on others. They lie, steal, cheat, speak hurtfully, break
confidences and promises, violate privacy, and act ungratefully, inconsiderately, duplicitously, impatiently, and unfaithfully. As a result peoples property is lost or damaged,
their feelings hurt, their confidence shattered, their trust
destroyed, and their psyches scarred. These are not murder,
mutilation, torture, and rape, but they are nonetheless
92 |
deep, often life-altering hurts. In the more extreme cases

the victims take their own lives as a result of the hurt, but
it does not have to reach that level for it to be worthy of our
moral revulsion.
Although humans do have a sense of justice, and human
societies often respond to injustice in order to punish it,
rectify it, and prevent future instances of it, injustice all too
frequently prevails. For example, most of the perpetrators
of human historys worst atrocities lived out their natural
lives without penalty. Forty-nine percent continued ruling
until their deaths by natural causes and a further 11% had
peaceful retirements. For an additional 8% the only penalty
was exile.38 Consider too, the number of unreported rapes,
unsolved murders, and other crimes for which nobody is
ever convicted. Whistleblowers and others who refuse to
countenance bad behavior by powerful people often pay a
high price.39 Evildoers often act with impunity.40
Nor should we lose sight of the myriad lesser (but
nonetheless significant) injustices of human life. One such
perpetrator was anatomist Henry Gray, who systematically
downplayed the role of his collaborator and illustrator,
Henry Carter, in the production of what, tellingly, became
known as Grays Anatomy.41 Another was Selman Waksman, who successfully connived to rob his student, Albert
Schatz, of credit for the latters discovery of streptomycin.
As a result Dr Waksman but not Dr Schatz won the Nobel
Prize for the discovery.42 Despite many attempts to rectify
the injustice, Dr Schatz went to his grave without achieving the recognition he was due.
Bad guys regularly finish first. They lack the scruples that provide an inner restraint, and the external restraints are either absent or inadequate.
| 9 3
Brutality to Brutes
Humans inflict untold suffering and death on many billions of animals every year, and the overwhelming majority of humans are heavily complicit.
Over 63 billion sheep, pigs, cattle, horses, goats, camels,
buffalo, rabbits, chickens, ducks, geese, turkeys and other
such animals are slaughtered every year for human consumption.43 In addition approximately 103.6 billion aquatic
animals are killed for human consumption and non-food
uses.44
Nor is the sum of these figuresover 166 billion
animalsthe total number of animals killed annually
in the industries that provide humans with animal flesh.

Excluded are hundreds of millions of male chicks that are
culled by the poultry industry because they will be unable
to produce eggs. There do not seem to be any estimates of
the annual number of such kills globally. However, there
are figures for some specific countries and regions, including the United States (260 million45) and the European
Union (330 million46).
Nor do the official slaughter figures include the dogs
and cats that are eaten in Asia. Reliable figures are even
harder to obtain here, but one calculation puts the annual
number at between thirteen and sixteen million dogs
and about four million cats.47 Similarly excluded are bycatchanimals such as turtles, dolphins, sharks, and sea
birds that are caught up in nets even when they are not
the intended catch. There are no reliable figures for the
numbers of animals killed in this category, but a subset of
bycatch, those discarded overboard, amounts to about a
further five billion marine animals.48
94 |
The deaths of the overwhelming majority of these animals are painful and stressful. Humans kill the millions of
male chicks in a variety of ways. In the United States most
are killed by being sucked at high speed to a kill plate,
which is sometimes electrified.49 Elsewhere they are killed
by suffocating or crushing, or, in the United Kingdom, by
gas or instantaneous maceration.50 Broiler chickens and
spent layer hens are suspended upside down on conveyer
belts and have their throats slit. Pigs and other animals
are beaten and shocked to coax them to move along in the
slaughterhouses, where their throats are cut or stabbed,
sometimes after stunning but sometimes not.
Marine animals do not fare any better. They typically
suffocate to death once out of the water, but there is suffering even on the way to the surface. Fish that are rapidly
hauled by trawlers from great depths suffer barometric
trauma. Gas bubbles form inside the body causing extreme
pain. Their swim bladders also become hugely inflated.
Sometimes the pressure is so great their stomach and intestines are pushed out of their mouth and anus. Eyes also
become distorted and bulge out.51 Fish caught on a smaller
scale, with line and bait, suffer the trauma of the hook as
they fight for their lives. Some humans would like to believe
that fish do not feel pain, but this comforting fiction, once
held about mammalian animals, withers in the light of the
evidence.52 The deaths of dolphins, which are highly intelligent mammals, may be even worse. When they are not
bycatch but instead the intended prey of fishermen, they
are driven into bays where they are butchered. Whales, also
mammals, are hunted at sea, where they are harpooned.
Animal suffering at the hands of humans is not restricted to the time that humans kill animals. Chickens,
| 9 5
for example, are typically reared in the extremely confined

spaces of the battery cages. They cannot spread their wings
or move about. They cannot engage in any of the activities,
such as dust bathing, which they would instinctually perform. They stand, with discomfort, on a sloped wire floor.
Because such conditions disturb the birds and cause them
to peck at one another, chicks destined for this life of suffering are de-beaked with a hot blade. When the egg yield
from a battery of hens declines, the hens are shoved into
crates and transported to slaughter.
Veal calves and farrowing sows are confined to such
small spaces that they can barely move for the duration
of their lives. Cows are fed Bovine Growth Hormone to increase milk production, but this often causes mastitis
painful inflammation of the cows udders. Humans mutilate
various animals, including pigs and cattle, by docking their
tails, castrating, dehorning, and branding them, all without anesthetic. Animals are often transported immense
distances by truck and ship in cramped and foul conditions
to be slaughtered at their destinations.
Producing food for humans is by no means the only
context in which animals are maltreated. It is hard to know
how many millions of animals are affected by scientific experiments53 each year, but a conservative calculation suggests that it is at least 115.3 million.54 Moreover, despite a
commitment to the three Rs of animal use in science
replacement, reduction, and refinementat least some
countries are actually increasing the number of animals
used each year.55
Many horrific experiments have been performed. It is
hard to summarize the full range of torturous treatments
to which animals have been subjected, but some examples
96 |
illustrate the sorts of cruelties humans have inflicted on

animals. There was a time when animals would be dissected
while fully conscious.56 As recently as the 1960s conscious
dogs were subjected to microwave blasts, resulting in the
swelling of their tongues, the crisping of their skin and,
if the temperatures were high enough, in death.57 In that
decade and the following one, monkeys were exposed, by
the US military, to massive doses of radiation, resulting in
the monkeys going into convulsions, stumbling, falling,
vomiting, twisting in an apparent endless and futile search
for a comfortable position.58
Psychological trauma has also been inflicted. In one
(in)famous set of experiments, infant monkeys were separated from their mothers, causing severe distress to both
mother and infant. The infants were then deprived of any
live contact. Their mothers were replaced with mannequins
that blasted the infants with air, or rattled them until their
teeth chattered, or catapulted them across their cages, or
stabbed them with spikes.59 Females reared in these ways
were then forcibly impregnated. Given their own upbringing they were unable, unsurprisingly, to care for the resultant offspring, and instead assaulted, maimed, and even
killed their infants.60
By current standards, many such experiments would
not receive the approval of animal research ethics committees. However, the current standards still allow humans to
inflict significant harms, including death, on animals. For
example, toxicity tests (for both medicines and cosmetics)
are performed where the intended or expected outcome is
death, typically preceded by the suffering that accompanies
the path to death by poisoning. Other animals are genetically
engineered to experience motor neurone degeneration,61 or,
| 9 7
like the oncomouse,62 to develop cancer. Humans also perform surgery on animals to produce experimental models of
painful conditions, such as sciatica,63 and they cause strokelike symptoms in a variety of animals, including rats, rabbits, cats, dogs and monkeys.64 They subject animals to
substances such as ethanol65 and methamphetamine66 and
to the effects these substances have on them. Those performing such experiments receive acclaim from the majority of their fellow humans.
Even more damning of our species than cases where
cruelty is inflicted as a result of indifference are cases
where the cruelty is brought about for human entertainment. Consider the baiting of bulls, bears, badgers, and
other animals. The baited animal is tethered to a pole and
then attacked by dogs for the pleasure of human spectators. Cockfighting, dogfighting, and bullfighting continue
even today.
Other sports also inflict suffering and death on animals even where this is not the goal. Horses are whipped
on the racetrack to entice them to run faster. They are injected with performance-enhancing drugs, often illegally.
They regularly break bones while racing and are then euthanized.67 Horses that are too old or weak to run are sent
for slaughter. Other animals that suffer for human entertainment are those confined to zoos or made to perform in
circuses.
Even those animals with whom humans have the closest emotional bondsdomestic companion animals such as
dogs and catsare not immune to ill-treatment on a colossal scale. Some humans treat these animals with immense
cruelty. They confine them in small spaces, beat them, and
fail to exercise or feed them adequately. The permutations
98 |
of cruelty are endless. For example, Henry Morton Stanley,

the famous nineteenth century explorer, cut off his dogs
tail, cooked it, and fed it to the dog.68 Terrible cruelty persists in our own times. In August 2006 a woman in England
attempted to drown a puppy in boiling water. The puppy
survived that attempt and was then left to die, which took
possibly as long as a week.69 In other recent cases a man
killed a dog by baking it in an oven,70 and another decapitated a cat with a machete.71 There are thousands of other
such cases.
Millions of dogs and cats are abandoned each year. In
the shelters to which they are sent, the overwhelming majority are killed because homes cannot be found for them.72
It is astounding that, in the context of so many unwanted
domestic animals, humans actively breed more such animals, which only exacerbates the problem. Sometimes
these breeding activities are informal and small scale. A
much greater problem, however, are the so-called puppy
mills (or kitty mills), which produce large numbers of
animals, who are often kept in poor conditions and given
inadequate attention. The aim is to maximize profits for
the breeders, and scant if any attention is giving to animal
welfare.
The human penchant for purebreds also leads to
animal suffering. Many such animals suffer from congenital problems that impair their ability to breathe, or that
render their spines vulnerable to injury, or their hips to
dysplasia.73 Other bizarre human aesthetic preferences
lead dogs to have their tails docked or their ears cropped,
often without anesthesia. Animals are also declawed
and debarked for the convenience of the humans whose
homes they share.
| 9 9
The above categories of human-inflicted harm on animals do not include the miscellaneous other ways in which
our species spreads misery. For example, in Asia, bears are
milked for their bile, a substance still used in traditional
medicine, even though no medicinal value has ever been
demonstrated. To facilitate the harvesting of their bile, the
bears are confined for the duration of their lives to crush
cages in which they cannot stand up or move around. In
these conditions their muscles atrophy and they go mad.
The catheters cause pain and the wounds can become infected, often leading to death.
Even more widespread than the abuse of bears is the
fur industry. Mink, foxes, rabbits, dogs, cats, and others
are its victims. Many of these animals are reared on fur
farms, in intensive conditions that cause significant suffering. They are then killed so that humans can wear their
furs. Many humans seem to think that fashion is a good
reason to make an animal suffer and die.
Toxic to the Environment
Some of the harm that humans cause to other humans
and to animals is mediated by the destructive effect that
humans have on the environment. For much of human history, the damage was local. Groups of humans fouled their
immediate environment. In recent centuries the human
impact has increased exponentially and the threat is now
to the global environment. The increased threat is a product of two interacting factorsthe exponential growth of
the human population combined with significant increases
in negative effects per capita. The latter is the result of industrialization and increased consumption.
100 |
The consequences include unprecedented levels of pollution. Filth is spewed in massive quantities into the air,
the rivers, lakes, and seas, with obvious effects on those
humans and animals who breath the air; live in or near the
rivers, lakes and seas; or get their water from those sources.
The carbon dioxide emissions are having a greenhouse
effect, leading to global warming. As a result, the icecaps
are melting, water levels are rising, and climate patterns
are changing. The melting icecaps are depriving some animals of their natural habitat. The rising sea levels endanger
coastal communities and threaten to engulf small low-lying
island states, such as Nauru, Tuvalu, and the Maldives.
Such an outcome would be an obvious harm to its citizens
and other inhabitants. The depletion of the ozone is exposing earths inhabitants to greater levels of ultraviolet light.
Humans are encroaching on the wild, leading to animal
(and plant) extinctions. The destruction of the rainforests exacerbates the global warming problem by removing
the trees that would help counter the increasing levels of
carbon dioxide.
There are some people, of course, who deny that
humans are having at least some of these large-scale negative effects on the environment. However, this is not the
placeand I am not the personto argue against the climate change denialists. Those who do deny that humans
are having a deleterious effect on the environment may
simply exclude the relevant harms. Humans are so destructive even without these harms that the second premise can
easily survive their exclusion. By contrast, those who do
recognize that humans are damaging the environment can
simply add this to the previous list.74
| 1 0 1
T H E N O R M AT I V E P R E M I S E
We have seen that humans cause colossal amounts of suffering and death. Having demonstrated the truth of the
second premise, I turn now to consider the first premise of
the moral misanthropic argument for anti-natalism:
We have a (presumptive) duty to desist from bringing into
existence new members of species that cause (and will likely
continue to cause) vast amounts of pain, suffering, and death.
The first thing to note about this premise is what it does

not claim. It does not claim that we should cull members
of dangerous species. Nor does it claim that we have a duty
to prevent others from bringing new members of dangerous
species into existence. The claim is a much more modest
one. It says that one should oneself desist from bringing
such beings into existence.
For this premise to be true it does not have to be the
case that every single member of the species will cause pain,
suffering, and death. To see why this is so, consider another
presumptive dutythe duty not to drive through red traffic lights. We have such a duty because driving through red
traffic lights is dangerous, even though not every instance
of such conduct results in harm.
The normative premise is neutral between whether the
species in question is ones own or another. Here it is important to note how widely the premise would be accepted
if the species were not human. Imagine, for example, that
some people bred a species of nonhuman animal that was
as destructive (to humans and other animals) as humans
102 |
actually are. There would be widespread condemnation of

those who bred these animals. Or imagine that some scientists replicated, and released, a virus that caused as much
suffering and death as humans cause. Again, there would
be little hesitation in condemning such behavior.75
The question, then, is whether it makes any difference
whether the highly destructive species is our own. In offering an affirmative answer to this question, some people
might suggest that there is something paradoxical about
claiming that we have a duty to desist from bringing into
existence members of a species that is harmful to itself.
There is, on this view, something odd about citing the harm
caused to humans by humans as a reason to desist from
creating humans. In other words, the misanthropic argument seems to be in conflict with the philanthropic ones.
If humans are worth protecting from harm then they are
not so bad that we should not replicate the species. And if
they are as bad as the second premise of the moral misanthropic argument suggests, then we should not count the
harm done to them as relevant in the first premise.
This line of argument fails. First, the harm that humans
do to humans is only part of the harm humans do. We are
also extremely harmful to other species. Thus, even if we
could not cite the harm that humans do to other humans
for the purposes of the moral misanthropic argument, the
argument could still be carried on the strength of the harm
that humans do to animals. This does assume, of course,
that animal interests count morally. However, there are
very powerful arguments for this conclusion and I shall not
rehearse them here.76
Second, it is a mistake to muddle our attitudes to victims and our attitudes to perpetratorseven when the
| 1 0 3
victims are also perpetrators. The recommendation that

we should keep these attitudes separate is not uncommon.
In civilized societies it is agreed that there are limits on
what we may do to even the worst perpetrators, let alone
lesser perpetrators. Those who torture and rape their victims before murdering them are not subjected to similar
treatment by the state (at least in civilized societies). This
is because the perpetrator remains morally considerable
despite his perpetration and, on this view, there are limits
on what we may do to morally considerable beings. The separation of attitudes is not restricted to the context of punishment. A woman may be guilty of physically assaulting
her child, but that does not mean that we should be unconcerned about the physical assault her husband inflicts on
her, or that we should be not be concerned about the violence he suffers at the hands of others. We should be concerned about the harm inflicted even on those who inflict
harm on others. This point is even more important when
greater harms are inflicted on lesser perpetrators. Thus,
the philanthropic and misanthropic arguments are not incompatible. We can believe both that it would be better if
humans never suffered the harms of existence and that it
would be better if there were no humans to inflict harms.
Now, it may be suggested that what is odd about the
moral misanthropic argument is the particular way it recommends preventing harm. It seeks to prevent harm to
humans by preventing humans. This objection would have
more force than it does if there were reasonable prospects
of reducing human destructiveness to negligible levels
fairly promptly and then ensuring that they do not rise
again. If that were the case then it could be argued that
instead of preventing humans we should rather reduce
104 |
their destructiveness. In fact, however, we cannot expect

that human destructiveness will ever be reduced to such
levels. Human nature is too frail and the circumstances
that bring out the worst in humans are too pervasive and
likely to remain so. Even where institutions can be built
to curb the worst human excesses, these institutions are
always vulnerable to moral entropy. It is nave utopianism
to think that a species as destructive as ours will cease, or
all but cease, to be destructive.
Am I being overly pessimistic here? After all, it has
been argued that rates of violence have been steadily diminishing and are now much lower than they were in
prehistoric times.77 This trajectory does not supplant the
pessimism implicit in the misanthropic argument. Insofar as violence has decreased it is only the rate of violence
that has declined. People are now less likely to suffer violence than they were before.78 However, the total amount
of suffering and death that is inflicted has increased, primarily because there are now many more people to inflict
harm and to suffer harm at the hands of others. Desisting
from creating new humans would mean that there would
be fewer humans to be harmed and thus less total harm.
While rates of violence are important, the total amount of
violence is at least as important a consideration in deciding
whether to create new people. There would be less violence
if there were fewer people.79
Even if we restrict our attention to the rate of violence,
the rate could still increase. Given human nature, we cannot
assume that the trend toward reduced rates is inexorable.
However, even if we set that concern aside, the current
rates are far from negligible despite the reduction. Even if it
were not nave to think that, in the very long term, human
| 1 0 5
destructiveness could be reduced to negligible levels, it

would still be indecent to create beings that in the interim
would cause massive pain, suffering, and death.
A PRE SUM PTIVE DUTY

If my argument so far is correct then we have a presumptive duty to desist from bringing new humans into existence. Can the presumption be defeated?
Those who think it can might suggest that while the
destructiveness of humans does create a presumption,
the presumption can be defeated because of the good that
humans do. One version of this view maintains that the
good is sufficiently widespread that the presumption can
regularly (even though not always) be defeated, and I shall
consider this version first.
The more regularly a presumption can be defeated the
less clear it is that the presumption really is a presumption. However, the presumption against creating new
members of a species that is as destructive as ours must
surely be a strong one. Thus those who would suggest that
it is regularly defeated must bear the burden of proof and
demonstrate that humanity does enough good to outweigh
all the harm it does. I am not optimistic that this burden
can be met.
Certainly in the case of the treatment of animals,
the scales are heavily weighted against us. Although it is
true that some humans do some good for animals, much
of this is merely rescuing animals from the maltreatment
of other humans. At the level of the human species such
benefits cannot be used to offset the harms. If there were
106 |
no humans to inflict the harms, these benefits would not

be necessary. Of course, humans do bestow some other
benefits, such as veterinary care for their companion animals. However, the number of animals affected and the
amount of good done is massively outweighed by the harm
the human species does to nonhuman animals.
Humans do bestow more benefit on other humans
than they do on animals. Nevertheless, it seems clear to
me that the good humans do is not sufficient to outweigh
the presumption against creating new people. There may
well be no definitive argument to prove this to those who
think otherwise. However, there are a number of considerations that can be offered in support of my assessment. At
the very least, these considerations show that those who
think that the presumption is defeated cannot demonstrate that it is.
First, the benefits humans provide to other humans
have to offset not only the harms done to humans but
also those done to animals, and these harms are colossal.
When the levels of destruction are this great, the amount
of benefit one is going to have to demonstrate in order to
defeat the presumption is immense. If pro-natalists think
that the good humans do does indeed outweigh the terrible destruction I have described, then we need to hear
some explicit details. Approximately how much good outweighs the dismemberment of a living being? How much
outweighs mass rape? How much outweighs the Rwandan
genocide or Joseph Stalins purges? It is when one actually keeps the atrocities in mind rather than speaking
about them abstractly as the evil humans do that the
claim that these atrocities are outweighed is shown to be
indecent.
| 1 0 7
Second, we need to understand what it means for the

good to outweigh the bad. This may not be as straightforward as it sounds. For example, imagine you knew that if
you conceived a child it would later, as an adult, murder
somebody. How many lives would that potential person
have to save during the course of his life (that would not
otherwise be saved) in order to override the presumption
against bringing him into existence? I doubt that that
number is two or even anything close to that.
What this case suggests is that the notion of good outweighing bad is more complicated than may first appear.
Now perhaps it will be suggested that this particular case
is a poor analogy for the matter at hand. Whether or not
that is true, the very same point can be made about the
case at hand. Thus a species that kills n-billion humans
and animals over some specified period would not redeem
itself by saving n-billion + 1 lives over that same period.
Perhaps there are some utilitarians who would, in each
of the cases just mentioned, assert that saving an additional
life would indeed be sufficient to offset the lives taken. However, utilitarians are not committed to such a view and any
form of utilitarianism that did adopt it would be a simplistic
one. A more nuanced view would recognize that taking lives
typically (even if not always) has worse secondary effects
than failing to save lives. A murderer, for example, arouses
more fear than a person who fails to save some lives that he
could otherwise have saved.
Non-utilitarians would have further reason for accepting the more complicated conception of what it takes for
good to outweigh bad. For them, considerations such as the
violating of rights could be a moral cost that is not offset
by saving a few more lives than are lost by the violation of
108 |
rights. On at least some non-utilitarian views, there could

be a threshold of harm beyond which no amount of benefit
can compensate. If there is any such threshold then human
destructiveness arguably exceeds it.
Third, some benefits will be moot in determining whether
the presumption is defeated. To understand which these are,
consider two intersecting distinctions:
1(a) benefits to those humans who already exist; and
(b) benefits to those future people who will be
brought into existence only if the presumption
against creating them is defeated.
2(a) benefiting by preventing harm; and
(b) benefiting by bestowing some (intrinsic) good.
At least those benefits at the intersection of 1(b) and 2(a)
are moot.80 This is because the benefits that fall in this intersection can be achieved in two ways: (i) by overriding
the presumption and creating the people who will prevent
the harm; and (ii) by deferring to the presumption and not
creating the people who will suffer the harm. Because of
this, these benefits are not net benefits of creating new
humans. That is to say, they are not an advantage over the
situation that would result from following the presumption against creating new humans. Thus they should not be
factored into a decision whether or not the presumption is
defeated.
Fourth, at least under current conditions the creation
of each new human or each new cohort of humans does not
produce benefits at the same rate that it produces harms.
Given the current size of the human population and the
| 1 0 9
current levels of human consumption, each new human

or cohort of humans adds incrementally to the amount of
animal suffering and death and, via the environmental
impact, to the amount of harm to humans (and animals). The
additional harm caused by each additional human may be
imperceptible but it is nonetheless an addition that, when aggregated with other imperceptible additions, becomes perceptible. However, it is not the case that the addition of every new
human or cohort of humans adds benefits. Much of the good
that humans do could be done by fewer rather than by more
humans. Thus, even if it is not always the case that creating additional humans is a net harm, it certainly is a net harm when
the human population is sufficiently large (and destructive).
For these reasons I reject the suggestion that the presumption against creating new people can regularly be defeated. In response, those who think that the presumption
can be defeated could fall back on a less ambitious version of
this viewnamely, that the presumption can occasionally
be defeated. Thus particular potential procreators might
agree that humanity is in general a very dangerous species.
However, they might suggest that the odds are that their
own potential offspring are much more likely to do enough
good and little enough bad to defeat the presumption.81
Depending on what we take to be enough good and
little enough bad, this may well be true of some (small
number of) potential people. However, we can fully expect
that most potential procreators will be very poor judges of
whether their potential offspring are likely to fall into this
category. The optimism bias, coupled with a tendency to
rationalize that the action one wants to perform will serve
the greater good, will lead the vast majority of potential
110 |
procreators (or at least the vast majority of those who

think before procreating) to the conclusion that the presumptive duty of the misanthropic argument is defeated
in their case. The overwhelming majority of them will be
wrong. Those who doubt this should consider the average
persons destructive effect on, at least, animals and the
environment.
We saw earlier that well in excess of 166 billion animals
are killed every year for human consumption or in industries providing for this consumption. The overwhelming
majority of humans on the planet are contributing to this
killing and the prior suffering. With the exception of India,
where a significant proportion of the population is vegetarian,82 only a very small proportion of people in other
countries are either vegetarian or vegan.83 This suggests
that, on average, each flesh-eater is responsible for the
deaths (and suffering) of at least 27 animals per year84
which amounts to at least 1690 animals over the course of
a lifetime.85 This is an underestimation, but it is nonetheless a lot of destruction for a single individual.
Each new person also has an impact on the environment and thereby on those sentient beings affected by environmental damage. In developed countries, the impact of
each person is massive. In the United States, for example,
the average person produces 28.6 tons of CO2 emissions per
year.86 In developing countries the per capita emissions are
typically lower, but they are not zero. In Bangladesh and
India, the annual average emissions of CO2 per person are,
respectively, 1.1 and 1.8 tons.87 Thus each new child contributes to environmental damage. Perhaps a pro-natalist
will want to argue that we cannot expect the production of
new people to have no impact on the environment and that
| 1 1 1
some such impact is acceptable. However, whatever force

this argument has is weakened as the number of people increases. The more people there are the less justifiable it is to
add further increments of environmental damage. Developing countries often have higher birthrates than developed ones. Individuals within such countries are going to
have a difficult time justifying their repeated procreation.
Humanity is a moral disaster. There would have been
much less damage had we never evolved. The fewer humans
there are in the future the less damage there will still be.
C O N C LU S I O N
Anti-natalist arguments vary in the scope of their conclusions. At its most extreme an anti-natalist conclusion
opposes all procreation, but milder versions oppose only
select cases of procreation.
The philanthropic arguments generate an extensive conclusion. They suggest that coming into existence is always a
harm. Because that harm is actually severe, it is, at least on
some views, always wrong to have children. (Other views might
allow some procreation as part of a plan to phase humans out
of existence.88)
The conclusion of the moral misanthropic argument is
that it is presumptively wrong to have children. It is possible that this presumption could sometimes be defeated.
I have argued that people will think it is defeated much
more often than it actually is and that it is very difficult,
if not impossible, to know when the presumption is indeed
defeated. However, it remains possible that there are
some circumstances in which a new human would produce
112 |
sufficient good to offset the harm that that particular

human would cause.
When the misanthropic argument is considered in conjunction with the philanthropic ones we find that the case
against procreation, and especially in our current circumstances, is almost always overdetermined.89
NOTE S
1. There are also aesthetic considerations against procreating.
For a discussion of these see the appendix to The Misanthropic Argument for Anti-Natalism in Permissible Progeny?,
eds. Sarah Hannan, Samantha Brennan, Richard Vernon
(New York: Oxford University Press), forthcoming. The chapter to which that appendix is attached is the source for the
current chapter in Debating Procreation.
2. A variant of this argument would focus on human nature
rather than on what humans do. There is obviously a close
connection between these two versions. Human behavior
could be taken as evidence of humanitys flawed nature, and
its flawed nature partially explains its bad behavior. Thus
the difference between the two versions is primarily one of
focus or emphasis.
3. This is not to deny that the religious worldview also recommends species humilitybut typically only relative to
God.
4. Some animals may surpass us in some cognitive capacities,
but none do all things considered.
5. See, for example, Dan Ariely, Predictably Irrational. Revised
and Expanded Edition (New York: HarperCollins, 2009);
Daniel Kahneman, Thinking Fast and Slow (London: Penguin, 2011).
6. Consider the case of Joe DiMaggio, of whom it has been
asked how he, a high school dropout whose favourite reading material was Superman comics, a man who was a lousy
father, an unfaithful husband and a wife beater . . . someone
| 1 1 3
who never did a meaningful days work in the last 47 years

of his life, who was monumentally vain and cheap and
mistrustfulbecame an American hero? The answer, we
are told, is simple: He could hit and throw and run with a
gliding grace, and when he could no longer do those things
he . . . well, he looked great in a suit. (Daniel Okrent, Say It
Aint So, Joe, Time, November 20 2000, 74.)
7. At the time of writing, the latest instance is the doomsday hysteria from those who feared that the world would
come to an end on December 21 2012. Russias government
resorted to issuing a statement, saying that it had access
to methods of monitoring what is occurring on the planet
Earth, and that it could say with confidence that the world
was not going to end in December. See Ellen Barry, In Panicky Russia, Its Official: End of World Is Not Near, New
York Times, December 1, 2012, accessed December 2, 2012,
http://www.nytimes.com/2012/12/02/world/europe/
mayan-end-of-world-stirs-panic-in-russia-and-elsewhere.
html?pagewanted=print
8. Solomon E. Asch, Studies of Independence and Conformity:
I. A Minority of One Against a Unanimous Majority, Psychological Monographs 70, no. 9 (1956): 170.
9. Rod Bond and Peter B. Smith, Culture and Conformity: A
Meta-Analysis of Studies Using Aschs (1952b, 1956) Line
Judgment Task, Psychological Bulletin 119, no. 1 (1996):
111137.
10. Steven T. Katz, The Holocaust in Historical Context, vol 1: The
Holocaust and Mass Death Before the Modern Age (New York:
Oxford University Press, 1994), 403404.
11. Sharon LaFraniere, African Crucible: Cast as Witches, Then
Cast Out, New York Times, November 15 2007, accessed
November 15, 2007, http://www.nytimes.com/2007/11/
15/world/africa/15witches.html?_r=2&th=&oref=slogin&
emc=th&pagewanted=all&.
12. Stanley Milgram, Obedience to Authority: An Experimental
View (New York: Harper and Row, 1974).
13. Ibid., p. 54.
14. Hannah Arendt, Eichmann in Jerusalem: A Report on the Banality of Evil (New York: Viking, 1965).
114 |
15. Philip Zimbardo, The Lucifer Effect: Understanding How Good

People Turn Evil (New York: Random House, 2007).
16. Of course, psychologists and others who have studied the
human mind are aware of the flaws to which I have pointed.
The problem is that their findings have not shaken the dominant self-congratulatory view of humanity.
17. We are in denial about our destructive capacities. When
humans behave in all the appalling ways I shall describe,
we characterize them as inhuman. Yet such behavior is so
rampant in human history that the negation signaled by the
prefix in in inhuman is, in fact, a negation of our pretensions (or, more charitably, our aspirations) about what
it means to be human. It does not signal a deviation from
the way the species actually conducts itself. It is similarly
ironic that when humans behave badlyand even when
they behave badly toward animalswe say they are behaving like animals, namely brutally. When I use the terms inhuman and brutal below, I am using them ironically.
18. Population Control, Marauder Style, New York Times, November 6, 2011, accessed November 7, 2011, http://www.
nytimes.com/imagepages/2011/11/06/opinion/06atrocities_
timeline.html?ref=sunday. Source: Matthew White, The Great
Big Book of Horrible Things (New York: W.W. Norton, 2012).
19. Adam Hochschild, King Leopolds Ghost (Boston: Houghton
Mifflin, 1998), 234.
20. Julie Flint, Observer, March 3 1991. Cited by Jonathan
Glover, Humanity: A Moral History of the Twentieth Century
(New Haven: Yale University Press, 2001), 32.
21. Nicholas Kristof, The Grotesque Vocabulary in Congo, New
York Times, February 11, 2010, accessed February 11, 2010,
http://www.nytimes.com/2010/02/11/opinion/11kristof.
html?th=&emc=th&pagewanted=print.
22. Jeffrey Gettleman, The Worlds Worst War, New York
Times, December 15 2012, accessed December 16 2012,
http://www.nytimes.com/2012/12/16/sunday-review/
congos-never-ending-war.html?pagewanted=2&pagewante
d=print&_r=0.
23. Ibid.
| 1 1 5
24. Neil G. Boothby and Christine M. Knudsen, Children of the

Gun, Scientific American 282, no. 6 (2000): 6065.
25. Midnights Children, Harpers Magazine 329, no. 1970
(August 2004): 23.
26. Neil G. Boothby and Christine M. Knudsen, Children of the
Gun, Scientific American 282, no. 6 (June 2000): 6065.
27. Adam Hochschild, Bury the Chains: The British Struggle to
Abolish Slavery (London: Macmillan, 2005), 7980. (One of
the slaves survived by grabbing onto a rope from the ship
and getting himself back on board without being noticed.)
28. Lydia Polgreen, Court Rules Niger Failed by Allowing Girls
Slavery, New York Times, October 28, 2008, accessed December 24, 2012, http://www.nytimes.com/2008/10/28/
world/africa/28niger.html?_r=3&pagewanted=print.
29. Nicholas Kristof, If This Isnt Slavery, What Is? New
York Times, January 4, 2009, accessed December 24, 2012,
http://www.nytimes.com/2009/01/04/opinion/04kristof.
html?pagewanted=print.
30. Report of the Sierra Leone Truth and Reconciliation Commission, Witness to Truth, 2004, Appendix 5, 17, accessed
November 23, 2014, http://www.sierra-leone.org/TRC
Documents.html.

31. Jo Thomas, Whats Furry, Literate and, Judging by
Events, Indispensable? New York Times, December 11,
1998, accessed December 25, 2012, http://www.nytimes.
com/1998/12/11/us/what-s-furry-literate-and-judgingby-events-indispensable.html?pagewanted=print&src=pm.
32. Ibid.

33. Michael Barbaro, Attention, Holiday Shoppers: We
have Fisticuffs in Aisle 2, New York Times, November 25,
2006, accessed November 27, 2006, http://www.nytimes.
com/2006/11/25/business/25shop.html?_r=0.
34. Robert D. McFadden and Angela Macropoulos, Wal-Mart
Employee Trampled to Death, New York Times, November
29, 2008, accessed December 1, 2008, http://www.nytimes.
com/2008/11/29/business/29walmart.html?_r=0.
35. Steven Nadler, Spinoza: A Life (New York: Cambridge University Press, 1999), 306.
116 |
36. Ibid.
37. For an example of a more recent mob killing see Murder
Most Pointless, Time, July 5 1999, 14.
38. The balance (32%) were imprisoned, executed, assassinated,
killed in battle, or took their own lives. Matthew White, The
Great Big Book of Horrible Things (New York: W.W. Norton,
2012), 534.
39. See, for example: Nicholas Kulish, Speculation Surrounds
Case of Albanian Whistle-Blowers Death, New York
Times, October 8, 2008, accessed October 8, 2008, http://
www.nytimes.com/2008/10/08/world/europe/08albania.
html?ref=europe&_r=0; Deaths in Moscow, The Economist,
January 24, 2009, 57.
40. For example: Simon Romero and Taylor Barnes, In Brazil,
Officers of the Law, Outside the Law, New York Times,
January 9, 2012, accessed January 10, 2012, http://www.
nytimes.com/2012/01/10/world/americas/in-parts-of-brazilmilitias-operate-outside-the-law.html?pagewanted=print.
41. Ruth Richardson, The Making of Mr Grays Anatomy: Bodies,
Books, Fortune, Fame (New York: Oxford University Press,
2008).
42. Peter Pringle, Notebooks Shed Light on an Antibiotics
Contested Discovery, New York Times, June 11, 2012, accessed
June 14, 2012, http://www.nytimes.com/2012/06/12/
science/notebooks-shed-light-o%20. . .%20mentors-betrayal.
ht m l?_ r =1 & p a r t n e r =r s s & e m c=r s s & p a g e w a nt e d =
print.
43. This data is from the Food and Agriculture Organization
of the United Nations (FAO). According to its statistics for
2010 the number of animals in these and related categories
that were slaughtered numbered 63,544,184,849. Food and
Agriculture Organization of the United Nations, Statistics
Division, faostat.fao.org (Accessed December 4, 2012.)
44. This figure is for 2009 and is based on the FAOs estimate
that 145.1 million tons of aquatic animal life was harvested that year (The State of the World Fisheries and Aquaculture, Rome: FAO Fisheries and Aquaculture Department,
2010, 3) and assumes that the average marine animals
| 1 1 7
weight is about 1.4 kg (see ADAPTT, Animal Kill Counter,

http://adaptt.org/killcounter.html, accessed December 4,
2012). It is noteworthy that official agencies such as the FAO
do not calculate the number of animals, but estimate only
the weight of the catch. Aquatic animals are, on this view,
not individual sentient beings but rather mere biomass.
45. Michael C. Appleby, Joy A. Mench, and Barry O. Hughes,
Poultry Behaviour and Welfare (Wallingford, UK: CABI,
2004), 184.
46. This figure is cited in a European Union Memo: Questions
and Answers on the Proposal for the Protection of Animals
at the Time of Killing, MEMO/08/574, Brussels, September 18, 2008, accessed December 19, 2012, http://europa.
eu/rapid/press-release_MEMO-08-574_en.htm.
47. How Many Dogs and Cats Are Eaten in Asia?, Animal People,
September 2003, accessed December 19, 2012, http://www.
animalpeoplenews.org/03/9/dogs.catseatenAsia903.html.
48. The State of the World Fisheries and Aquaculture, 8384, cites
a figure of 7 million tons. Using the average marine animal
weight of 1.4 kg., this amounts to about 5 billion animals.
49. Appleby, Mench and Hughes, Poultry Behaviour and Welfare,
184.
50. Ibid., 184186.
51. Victoria Braithwaite, Do Fish Feel Pain? (Oxford: Oxford
University Press, 2010), 177.
52. Ibid. For the pain inflicted by hooks, see pp. 164168.
53. I am skeptical about the value of animal experimentation,
but even if one thought that this practice were justifiable
on account of a benefit to humans, it would still be the case
that if there were no humans these harms would not be inflicted on animals.
54. Katy Taylor et al., Estimates for Worldwide Laboratory
Animal Use in 2005, Alternatives to Laboratory Animals 36,
no. 3 (2008): 327342.
55. Ingrid Torjesen, Animal Experiments Rose in 2011 Despite
Coalition Pledge To Reduce Them, British Medical Journal
345, no. e4728 (2012).
118 |
56. See, for example, Kathryn Shevelow, For the Love of Animals:
The Rise of the Animal Protection Movement (New York: Henry
Holt, 2008), 144.
57. Deborah Blum, The Monkey Wars (New York: Oxford University Press, 1995), 82.
58. Ibid., 83.
59. Ibid., 90.
60. Ibid., 91.
61. Mark E. Gurney et al., Motor Neuron Degeneration in Mice
that Express a Human Cu,Zn Superoxide Dismutase Mutation, Science 264, no. 5166 (1994): 17721775.
62. Alun Anderson, Oncomouse Released, Nature 336, no.
6197 (1988): 300.
63. See, for example, Peter M. Grace et al., A Novel Animal
Model of Graded Neuropathic Pain: Utility To Investigate
Mechanisms of Population Heterogeneity, Journal of Neuroscience Methods 193, no. 1 (2010): 4753.
64. Juliana Casals et al., The Use of Animal Models for Stroke
Research: A Review, Comparative Medicine 61, no. 4 (August
2011): 305313.
65. See, for example, Kathryn L. Gatford et al., Acute Ethanol Exposure in Pregnancy Alters the Insulin-Like Growth
Factor Axis of Fetal and Maternal Sheep, American Journal
of PhysiologyEndocrinology and Metabolism 292, no. 2, Part
1 (2007): E494E500.
66. See, for example, Kelly J. Clemens et al., Repeated Weekly
Exposure to MDMA, Methamphetamine or Their Combination: Long-term Behavioural and Neurochemical Effects in
Rats, Drug and Alcohol Dependence 86, no. 2 (2007):183190.
67. See, for example, Walt Bogdanich, Joe Drape, Dara L Miles
and Griffin Palmer, Mangled Horses, Maimed Jockeys, New
York Times, March 24, 2012, accessed March 25, 2012, http://
www.nytimes.com/2012/03/25/us/death-and-disarray-atamericas-racetracks.html?_r=2&hp=&pagewanted=print.
68. Adam Hochschild, King Leopolds Ghost, 196.
69. Pet-Abuse.com, http://www.pet-abuse.com/cases/10007/
EN/UK/ (Accessed December 17, 2012).
WI/US/ (Accessed December 17, 2012).
| 1 1 9

WA/US/ (Accessed December 17, 2012).
72. It is very difficult to determine how many millions of dogs
and cats are killed each year by animal shelters because
those animals cannot be homed. The American Humane Association estimates that the number in the United States is
about 3.7 million animals per year and says this number
represents a generally accepted statistic that is widely
used by many animal welfare organizations, including the
American Society for the Prevention of Cruelty to Animals
(ASPCA). (see Animal Shelter Euthanasia, American
Humane Association, http://www.americanhumane.org/
animals/stop-animal-abuse/fact-sheets/animal-sheltereuthanasia.html, accessed December 4, 2012.) However,
these figures, which are based on a 1997 survey conducted
by the National Council on Pet Population Study and Policy,
explicitly note that it is not possible to use these statistics
to estimate numbers of animals entering animal shelters in
the United States, or the numbers euthanized on an annual
basis because the reporting Shelters may not represent
a random sampling of U.S. Shelters. (see http://www.pet
population.org/statsurvey.html, accessed December 4, 2012).
73. Nicola Rooney and David Sargan, Pedigree Dog Breeding in
the UK: A Major Welfare Concern? Royal Society for the Prevention of Cruelty to Animals (RSPCA), West Sussex, UK,
2009, accessed November 23, 2014.
74. There are some environmental ethicists who believe that
damage to the environment is morally relevant in its own
right, independent of the effects on humans and animals.
That is a controversial position, which I do not accept. Those
who do endorse it can, again, add it to the list of (morally
relevant) harm that humans cause.
75. Some might argue that these other species do not do as
much good as humans do and are thus not analogous. I shall
later consider whether the good that humans do can defeat
the presumption against creating more of them.
76. These arguments are advanced in dozens of books and articles. See, for example, Peter Singer, Animal Liberation.
New Revised Edition (New York: Avon, 1990) and David
120 |
DeGrazia, Taking Animals Seriously (New York: Cambridge

University Press, 1996).
77. The evidence and arguments for this have been presented
in impressive detail in Steven Pinker, The Better Angels of
our Nature: The Decline of Violence in History and its Causes
(London: Allen Lane, 2011).
78. The same may not be true of animals. Although some kinds
of animals, such as dogs and cats, are now treated better
than they were in the past, other animals, most notably farm animals, are treated worse today than they were
before. The intensive farming and associated suffering is a
product of the last few decades.
79. One does not have to be a consequentialist to be concerned
about the total amount of violence. Non-consequentialists
can think that it is better to have less violence rather than
more, at least on condition that lowering violence does not
violate any non-consequentialist principles, such as rights.
80. If one accepts the basic asymmetry then the benefits at the
intersection of 1(b) and 2(b) are also moot.
81. It is curiously rare for people to think that chances are that
their own children would produce a net harm but that the
children of others would produce a net benefit.
82. There are varying estimates of the number of vegetarians in
India. Here I shall assume a generous estimate of 42% of the
population (see R. Mehta et al., Annex II: Livestock, Industrialization, Trade and Social-Health-Environment Issues
for the Indian Poultry Sector, Food and Agricultural Organization, June 2002, accessed January 15, 2013, http://
www.fao.org/WAIRDOCS/LEAD/X6115E/x6115e0c.htm)
83. There are no data for the proportion of vegetarians worldwide. There are data for some developed countries, where the
proportions are between around 1% and 9%. (See Matthew
B. Ruby, Vegetarianism: A Blossoming Field of Study, Appetite 58, no. 1 ( 2012): 142.) For my purposes, I shall assume
that 5% of the world (outside of India) is vegetarian or vegan.
This is likely an overestimate as the rate of vegetarianism in
the developing world (outside of India) is likely to be lower.

84.
This calculation assumes a world population of
7,021,836,029 and the population of India as 1,205,073,612
| 1 2 1
[The World Factbook, Central Intelligence Agency, Washington DC, accessed January 15, 2013, https://www.cia.gov/
library/publications/the-world-factbook/geos/xx.html).
85. This assumes an average life expectancy of 67.59 years,
which is the current average globally (https://www.cia.gov/
library/publications/the-world-factbook/geos/xx.html,
accessed December 27, 2012), with meat consumption beginning at age five. (Of course omnivorous humans begin
eating meat earlier than age five, but their consumption
levels are low at the earliest stages.).
86. Edgar G. Hertwich and Glen P. Peters, Carbon Footprint of
Nations: A Global, Trade-Linked Analysis, Environmental
Science and Technology 43, no. 16 (2009), 6416.
87. Ibid.
88. For more on this, see David Benatar, Better Never to Have
Been (Oxford: Oxford University Press, 2006), 182193.
89. I am grateful to Anna Hartford for research assistance for
this chapter, and to participants in the Permissible Progeny
workshop (London, Ontario, June 2013) for their comments.
Contra Procreation
IN THE PREVIOUS CHAPTERS I presented various positive
arguments for anti-natalism. I shall now respond to arguments for the opposite position. Pro-natalists argue in favor
of procreation. Unsurprisingly, such arguments are typically
not grounded in the interests of the people who would be
brought into existence. Instead, most of them are grounded,
one way or another, in the interests of existing people.
DIVINE COMMANDS
One exception is the argument that we have a presumptive duty to procreate on account of Gods commandment
to be fruitful and multiply.1
The difficulty with responding to this sort of argument
is that its assumptionsthat God exists, that he commands
us to procreate, that the command creates a moral dutyare
articles of faith. Responses to it are thus unlikely to persuade
C ontra P rocreation
| 1 2 3
those who are convinced by it. That does not mean that the
argument is a good one. Precisely because its assumptions
are so controversial, the argument will have no force for
those who, quite plausibly, reject these assumptions.
However, the argument is controversial even for theists
and even for those who believe that the Bible is the word of
God. There are, after all, such people who believe that they
are under no such obligation. Catholic priests and nuns, for
example, do not take their vows of celibacy to constitute a
violation of the commandment to be fruitful and multiply.
Celibacy is required for all Shakers, not only the clergy. Required celibacy entails the absence of a duty to procreate.
In Judaism, the commandment is understood as applicable only to males.2 The rationale for this is that because
women are endangered by pregnancy and childbirth they
cannot be obligated to put themselves in this danger. For
obvious reasons, therefore, men are unable to fulfill their
purported duty without the cooperation of those who have
no such duty.
Thus we see that there is considerable scope for the interpretation of biblical commandments. There is no overwhelming reason why other religious people could not
believe that also exempted are those men who do not wish
to put their wives at the very risk that exempts women
from the obligation. In this or some other way, the duty
could be voided entirely.3
PA R E N TA L I N T E R E S T S
A second pro-natalist argument is grounded in the interests
of the prospective parents. According to this argument,
124 |
procreation and subsequent parenthood are important

features of a good life, at least for many people.4 Parents
lives, it is said, are improved by the loving relationships
they have with their children. Children are also seen as a
kind of insurance policy because they can care for a parent
during the latters old age. Because parents usually predecease their children, parents also often see their children
as a means for transcending their own deaths.
The first thing to note about these considerations is
that at most they can generate prudential reasons for procreating. They do not seem to generate a duty to procreate.
One has no duty to improve ones life in this particular way
or to secure this insurance for ones old age or to attempt to
transcend ones mortality in this way.
Moreover, some of the supporting considerations for
this more limited pro-natal conclusion are contested. For
example, although it seems, intuitively, that having children makes ones life go better, it is far from clear that this
is actually true. Some studies have supported the intuitive
conclusion,5 but others have found either no difference6 or
that having children has a negative impact on well-being
or happiness.7 Among those studies that found that children do increase parents happiness, some found that this
was not true of all parents. For example, they found it to
be true only of married or partnered parents.8 There have
also been conflicting findings on the role that a parents
sex plays. Some studies have found that while fathers are
significantly happier than childless men, mothers are neither happier nor less happy than childless women.9 By
contrast, others have found that female life satisfaction
increases more than male life satisfaction with the presence of children.10
| 1 2 5
Even if we assume that rearing children does make

ones life go better, it is possible for many people to derive
these benefits without procreating. There are millions of
unwanted children who could be adopted and with whom
those seeking a loving parent-child relationship could find
such a relationship. Adopting a child is typically much
harder than having one the more traditional way. Nevertheless, because creating a child causes considerable harm
to that child, whereas adoption characteristically benefits
the adopted child, adoption is the preferred way to enter
into a parenting relationship.
Adopted children are not ones genetic own and many
people seem to see this as a disadvantage.11 However, children do not need to be genetically ones own in order to
enjoy a loving relationship with them. Similarly, adopted
children can care for one in ones senescence. What about
the claim that children enable one to transcend ones own
death? This idea is inchoate. However the claim is understood, the transcendence is obviously of a limited kind.
Children do not enable one to survive ones death in anything like as a robust a way as people would like to survive
their death. However, children can, for example, continue ones way of life, protect ones legacy, and keep ones
memory alive, at least for a while. No genetic connection
is required for these means of transcending death. However, if one is interested in a genetic transcendence of ones
death then adopted children clearly cannot provide this.
That said, it is unclear why it is so important to leave genetic traces after one dies. This preference seems to be a
kind of genetic narcissism.
Even if we accept that adoption is preferable to procreation, there are clearly more people who want children
126 |
than there are unwanted children available for adoption.

Thus, even if people were to prioritize adoption, there
would be many people who would be unable to gain the
benefits of rearing children without procreating. Let us
assume then that for the vast majority of people, the benefits of parenting require procreation. This is not sufficient
to justify procreation. As important as these benefits may
be to many people, it is hard to see how they may justifiably be purchased at the cost of serious harmor even the
significant risk of serious harmto those who are brought
into existence.
GROUP INTERE STS

Other pro-natalist arguments are grounded in cultural continuity or in national interests.12 Such arguments, like the
religious and parental interest arguments, do not appeal to
the interests of those people who would be brought into existence. This is because future merely possible people have
no interest in the nation or the continuity of the culture
into which they would be born. Invoking cultural continuity or national interests to justify a duty to procreate ultimately appeals to the interests of existing people. It is they
who may have an interest in their culture or nation.
Cultures and other ways of life cannot survive without new generations to replace those cultural members
who die. Existing people from other cultures could make
the transition into the culture, but this is not a sustainable
mechanism for cultural continuity and in any event is parasitic on the procreation of others. Thus it is unsurprising
that (many of) those committed to particular cultures and
| 1 2 7
wanting those cultures to continue would see an imperative to reproduce. The imperative might be thought especially pressing when a cultural group is small and under
threat.13
Although these thoughts are understandable, they fail
to ground a moral duty to procreate. Cultural attachments
may be valuableeven immensely valuablefor some
people. However, the continuity of the culture beyond
those peoples own lives has primarily what we might call
a reflected value. Put another way, the value lies not so
much in the actual continuity of the culture beyond their
death, but more in their thinking that the culture will continue then. Of course, there need to be real prospects for
the actual continuity of the culture in order for people to
think that the culture will continue, and thus the distinction is heuristic. What it helps to illuminate is what kind
of value we are weighing up against the very real and considerable harms that the procreative expression of the cultural continuity imperative has for those people who are
brought into existence. It is very hard to see how the comforting thought that ones culture will continue beyond
ones death can justify the massive harms caused by bringing somebody into existence.
Cultural and national interests in procreation sometimes amount to more than a comforting thought. For
example, new generations of a group might be needed to
defend the group militarily against aggressors. Or there
might be a demographic struggle, in which the less fecund
group will be swamped, even if only via the ballot box, by
the more rapid breeders. In such cases the impact on the
less populous group may be significant. In some situations
their very lives may be in danger.
128 |
Such weighty considerations are thought to be operative much more often than they actually are. Many states
that adopt pro-natalist policies in the name of the national
interest say that they need troops for defensive purposes,
but more often than not the purpose is actually aggressive.
Where the need for defensive demographics is both
real and significant, there are sometimes non-procreative
solutions. These include emigration, immigration, and various political means, including protections for minorities.
Where procreation is the only solution, we are faced
with a tragic situation: the only way that serious harms to
existing people can be avoided is by creating new people
on whom we thereby inflict all the serious harms that are
attendant upon existence.
Such a tragic dilemma could, at least in principle, arise
in other circumstances too. If, for example, humans were
a generation away from extinction, the final people could
be expected to suffer terribly as they aged and there were
no younger people to produce food, provide health care, do
the policing necessary for ensuring personal security, and
bury the dead.14 If procreation could save the ageing generation from this apocalyptic finale then the dilemma would
arise. This, of course, is not the situation in which current
procreators find themselves, given the vast amount of
procreation that is taking place. Instead it is a theoretical
possibility.
Whether one thinks that procreation is permissible, let
alone required, in such tragic circumstances, will depend
heavily on ones moral theoretical commitments.15 For example, it will depend in part on what view one takes about
whether one may seriously harm some in order to prevent serious harm to others. However, if procreation is justified (or
| 1 2 9
required) under some such circumstances it must surely be as

an interim measure only. To save the lives of members of vulnerable groups it must be a stopgap until a non-procreative
solution can be found. To avoid the suffering of the final
people, it must be a mechanism for phased extinction.
The reason why it cannot be a more enduring solution is that continued procreation in order to save existing
people from harm is a giant procreative Ponzi scheme.16
Each generation has to procreate in order to save itself
from the fate of the final generation, thereby creating a
new generation that must procreate in order to spare itself
the same fate. Like all Ponzi schemes, it cannot end well. It
merely delays the inevitable. However, unlike other Ponzi
schemes, the procreative one also causes vast amounts of
suffering before the bubble bursts or the pyramid crumbles.
Although sparing any generation from the misery of
being the final generation is arguably the most powerful
consideration in favor of continuing humanity, it is by no
means the only one. Just as many people are invested in
the continuity of some subspecies grouping, such as a cultural group or a nation, after their own deaths, so many
humans are invested in the continuity of the species as a
whole. They want human life to continue even after they
themselves die. They want there to be humans and human
civilization in the future.
However, even if it is true, as has recently been argued,17
that the continued existence of the species gives purpose
and meaning to the projects of many currently existing
people, it is hard to see how these could be sufficiently
weighty considerations to warrant the severe harms and
risks of being brought into existence. And even if one did
think that they were sufficiently weighty, one would again
130 |
become trapped in the futile finite regress that is the procreative Ponzi scheme. As a species we can tread the perilous waters of purposelessness by procreating, but only for
so long. The final peoples problems of purpose will be no
different whether the final people are the current generation or some distant future one. The difference between
the two scenarios is how many suffering generations there
will be between now and then.
C O N C LU S I O N
Far from being the innocent, or even noble, activity that
it is often taken to be, procreation is an inherently problematic practice. In creating a child, one is creating a new
center of consciousness, a new subject of desire. One must
know that that child will experience considerable unpleasantness, pain, and suffering during the course of its life.
The font of desire one creates will regularly be thwarted
and frustrated when one seeks to satisfy those desires. This
is all unnecessary in the sense that the conscious existence
subtending these experiences was unnecessarily created. It
could have been avoided. In procreating one creates a being
for whom things canand, sooner or later, willgo very
badly. Every birth is a future death. Between the birth and
the death there is bound to be plenty of unpleasantness.
Procreators inflict these harms on their progeny obviously without the latters consent. Nor can they inflict
those harms for the sake of the children they create. None
of the reasons for procreating have anything do with the
interests of the beings that are brought into existence.
Procreation serves the interests of othersthe parents,
| 1 3 1
grandparents, already existing siblings, the nation or state,

or the broader human community into which the child is
born.
Inflicting serious harmor even the risk of iton
one person, without his or her consent, in order to benefit others, is presumptively wrong. The misanthropic argument for anti-natalism deepens the presumption against
procreation. None of the reasons for procreating are sufficiently strong to defeat this presumption.
NOTE S
1. Genesis 1:28
2. Sefer HaChinuch, Mitzva 1. See Sefer haHinnuch, Vol. 1 (Jerusalem and New York: Feldheim, 1978). The English translation and annotation was done by Charles Wengrov. The
authorship of the book is in dispute, but the common attribution is to Aaron HaLevi.
3. Some might reply that any move that renders a commandment void takes Biblical hermeneutics too far. However,
there is precedent for interpreting a commandment in such
a way as to void it. For example, the Talmud applies numerous restrictive interpretations on the commandment
to execute a stubborn and rebellious son (Deuteronomy
21: 1821) and records a view that there was never and will
never be a case of somebody being executed under this law.
(Sanhedrin 71a.)
4. This argument has been advanced or considered by Saul
Smilansky, Is There a Moral Obligation To Have Children?
Journal of Applied Philosophy 12, no. 1 (1995): 46; and Marshall Missner, Why Have Children? The International Journal of Applied Philosophy 3, no. 4 (1987): 2, 4.
5. See S. Katherine Nelson et al., In Defense of Parenthood:
Children Are Associated With More Joy Than Misery, Psychological Science 24, no. 1 (January 2013): 310.
132 |
6. See Tanya Rothrauff and Teresa Cooney The Role of Generativity in Psychological Well-Being: Does it Differ for Childless Adults and Parents? Journal of Adult Development 15,
no. 34 (2008): 148159.
7. Sara McLanahan and Julia Adams, The Effects of Children
on Adults Psychological Well-Being: 1957-1976, Social
Forces 68, no. 1 (1989): 124146.
8. Nelson et al., In Defense of Parenthood, and Luis Angeles,
Children and Life Satisfaction, Journal of Happiness Studies 11, no. 4 (2010): 423538. Others did not find that increased happiness was restricted to partnered or married
parents but did find that the happiness levels were greater
for this group of parents. (See Arnstein Aassve, Alice Goisis,
Maria Sironi, Happiness and Childbearing Across Europe,
Social Indicators Research 108, no. 1 (2012): 6586.
9. Nelson et al., In Defense of Parenthood, 9.
10. Angeles, Children and Life Satisfaction, 532.
11. Smilansky, Is There a Moral Obligation To Have Children?
47.
12. Ibid., 4748. The cultural argument is also considered by
Missner, Why Have Children? 2.
13. By contrast, the imperative may either weaken or even
vanish for a given person if enough other people in the
group are procreating.
14. See Benatar, Better Never to Have Been, 182186.
15. For a much more detailed discussion of this, see Better Never
to Have Been, 186193.
16. Although it has been drawn to my attention that Mark
Johnston has also used the term Ponzi scheme with
reference to procreation, albeit in a slightly different context (Is Life a Ponzi Scheme? Boston Review, January 2,
2014, accessed October 1 2015, http://www.bostonreview.
net/books-ideas/mark-johnston-samuel-scheffler-deathafterlife-humanity-ponzi-scheme), I employed the term
not only unaware of his review article but also before it was
published.
17. Samuel Scheffler and Niko Kolodny, Death and the Afterlife
(New York: Oxford University Press, 2013).
PART II
PRO-NATALISM
DAV I D WA S S E R M A N
Better to Have Lived and Lost?
IF NO ONE IS BORN, no one suffers. If any possible birth is
foregone, then someone who would have enjoyed the benefits of existence will not exist. These two propositions are
the fixed points for the debate on the ethics of child creation. On the one hand, most of us do not believe that the
inevitable suffering of a future child gives its potential creators an obligation or strong reason to desist if they expect
it to have a good life on balance. On the other hand, most of
us do not believe that the prospect of a good or even great
life gives fertile couples or other potential child makers an
obligation or strong moral reasons to have children.
In my contribution to this volume, I will focus on recent
attempts to develop an ethics of procreation that supports
both beliefs: that it is often permissible to have children despite the serious harms they will suffer, and that it is never
a requirement, despite the great benefits they may enjoy. I
will begin by framing the challenge for such an ethics as one
of establishing a middle ground between those who reject
136 |
the very idea that procreation needs defense and those who
argue that it is categorically or presumptively indefensible.
For some philosophers writing from a religious background, it reveals a loss of ethical moorings to even question
the reasons or justification for having children. Elizabeth
Anscombe began her essay Why Have Children? by asserting that this very title tells of the times we live in.1
Describing those times as ones in which the creation of a
child is little more than a vanity purchase, she concluded
It is distressing to live in a world where this question ...
so intelligibly presents itself . . . Just over twenty years
later, reviewing a book by Christine Overall with (perhaps
ironically) the same title as Anscombes essay, Gilbert Meilaender expressed similar, if slightly gentler, doubts about
the very question:
If we ourselves are grateful to be alive, there must be some
instances (probably many) in which it would be good to
transmit life to the next generation. And if, on the other
hand, we are not prepared to affirm the goodness of our
own existence, its hard to know why we should take an interest in the question posed by Overalls book.2
For Meilaender, it is not only ironic but perverse that the

question of what justifies procreation should arise at a
time, and in places, where having and rearing children was
never safer, easier, or more assured of material success.
Another challenge to the very idea that having a child
requires justification comes from a very different, secu
lar direction. David Heyd argues that we cannot harm or
wrong individuals by the acts responsible for their existence. Because existence is not a predicate, it cannot be regarded as a harm or benefit to those receiving it.3 For this
B etter to H ave L ived and L ost ?
| 1 3 7
reason, progenitors are not morally accountable to the children they bring into existence for its unavoidable bads or
goods, whatever their balance. Heyd would find it reasonable for prospective parents to ask why have children? because he would hold them accountable to individuals and
institutions that exist independently of their procreative
decisions. It is unclear whether the religious traditions that
Anscombe and Meilander draw upon impose similar accountability. But Heyd defends a view that is certainly congenial to theirs: that progenitors never can wrong a child
merely by creating him and need not justify their decision
to that future childa convergence suggesting that procreation makes strange bedfellows.
At the other extreme, a number of philosophers over
the past decade have not only raised the question of justification dismissed by Anscombe and Meilaender, but
answered it, more or less categorically, in the negative. Several rely on moral differences between imposing harms and
conferring benefits on a future child. David Benatar presents the most categorical argument, based in part on the
asymmetrical valuation of the absence of bads and goods in
nonexistence.4 Others argue that extreme risk aversion is
appropriate in deciding whether to have a child, either because (1) unlike harms imposed to prevent greater harms,
those imposed only to confer pure benefitsthose that
do not prevent or mitigate harmscannot be justified
without a consent that is unobtainable prenatally5; or (2)
the worst possible consequences for the child created are
so much greater than the consequences for that child of
not being created.6
For Benatar and some other anti-natalists, categorical arguments are fortified by claims about how we should
138 |
assess the lives we actually lead and what conclusions about

their goodness or badness a proper assessment would yield.
Thus Benatar argues that we typically see our lives through
rose-colored glasses, and that when we remove them, we
see those lives, or almost all of them, as very bad when
viewed in their totality. His position rests in part on a normative claim about the how we should weigh goods and
bads postnatally, and in part on empirical claims about the
frequency, duration, and intensity of those goods and bads.7
I will suggest that these anti-natalist positions, though
quite distinct, give some kind of priority to the avoidance
of harms over the conferring of benefits. This priority assumes different forms, and plays different roles in their
arguments. It may ultimately reflect a bedrock conviction
rather than a debatable proposition. The responses to antinatalism challenge the specific priority claims that are
made, but they do not, and perhaps could not, refute the
underlying conviction. My own response, concerning the
moral complaint children could have against their progenitors, does not question the priority of harm over benefit
as an axiological claim, although I do not accept it. It just
maintains that this priority itself would not provide the
basis for a complaint.
Other philosophers make more contingent or qualified
anti-natalist claims. Some consequentialistswho judge
the rightness of actions solely by their outcomesargue
that the massive cost of bearing and raising children in
wealthy societies is unacceptably high, in terms of lost opportunities to give desperately needed aid.8 This judgment,
however, is contingent. If procreation increased rather than
reduced total or average good, then be fruitful and multiply
would be a moral imperative.9 Other non-consequentialist
| 1 3 9
philosophers argue that procreation is presumptively impermissible when there are needy children available for
adoption, since most prospective parents can fulfill their interests in rearing children without bearing them.10 Like the
consequentialist objection, this one is contingent: it would
permit procreation it if there were no adoptable children.
There is, or appears to be, a vast middle ground between
those who regard procreation as categorically or presumptively wrong and those who hold that procreationin all,
most, or many circumstancesrequires no defense. Most
philosophers writing on the subject reject all forms of
anti-natalism, but do regard procreation as requiring a defense.11 It has long been recognized that unfettered procreation could cause harm to third parties, from other family
members to the global communityconcerns emphasized
by the more contingent anti-natalist arguments. The categorical anti-natalist arguments make the stronger claim
that is it never possible to justify procreation to the child
created, a child who always faces a life with serious harms
and grave risks.
Most arguments for procreation have been negative;
they claim that the burden of justification is not as great
as anti-natalists claim; nonexistence either cannot be
compared to existence or need not be better; subjecting a
child to the inevitable harms of existence does not require
a consent that is impossible to obtain prospectively; and,
although the worst outcomes of existence are awful, it is
permissible to risk them if their probability is sufficiently
low.12 Many pro-natalists then proceed to offer minimum
standards for permissible procreation: for avoiding wrong
to the child, to third parties, or to the world at large.13 Many
also argue that adults have a moral as well as legal right to
14 0 |
procreate because of the highly personal character of reproductive decisions and their central importance to those
making them. Procreative liberty insulates individuals
and couples from moral reproach as well as legal sanction,
as long as long as they do not fall below the prescribed minimum standards.14 A similar but perhaps more comprehensive immunity is recognized by those who maintain that
the decision to become a parent is, like the decision to gestate, an intimacy of the first order, which cannot be the
subject of moral duty.15
Yet it is not enough, or so I will claim, to reject antinatalism and propose minimum standards for procreation.
The arguments against procreation may fail, but some of
them raise concerns about the awful, unpredictable, unpreventable fates of some humans that resonate even with
the most resolute defenders of procreation. There appears
to be a dearth of strong secular reasons to have children
that would justify exposing individuals to the risk, however slight, of such fates. The good for existing individuals of another sibling, child, grandchild, or playmate not
only seems less compelling; it provides reasons that seem
to justify the creation of a future person as a mere means
to the good of others. The survival of humanity, which
manythough by no means allphilosophers regard as a
critical good,16 is just marginally advanced by the creation
of a single child. Finally, it is difficult to justify procreation in terms of the good lives of future children without
acknowledgingas most of us are reluctant to dothat
we have a strong moral reason to create children. In the
absence of good justifications for having children, it may
appear that procreation is a marginally permissible but
highly problematic enterprise.
| 1 4 1
Although few philosophers have attempted to provide secular reasons for procreation, many have proposed
standards that serve, in effect if not intent, to mitigate the
concerns raised by anti-natalist critiques. Thus some claim
that if we are to procreate we should try to select those
children least likely to suffer the worst harms that life
brings, to suffer less harm than other possible children, or
to enjoy the greatest possible benefits.17 Other argue that
since children are demanding beings who consume substantial resources and require substantial investment, we
should select children who will impose the least burden,
contribute the most benefit, or do the most good.18 These
considerations seem to suggest that procreation, if it is to
be more than barely acceptable, needs to be a highly selective enterprise.
Such selectivity is becoming possible because of modern
reproductive technology. Our possibilities for choosing
among future children were extremely limited before the
advent of prenatal and pre-implantation genetic diagnosis.
With the availability of technologies that enable us to predict an increasing number of features genetically, we can
make a stab at picking the best, as well as excluding the
worst. These capacities pose hard questions for those who
previously found it largely unproblematic for parents to
bear a couple of children. Are prospective parents obliged to
generate large numbers of fertilized eggs to raise the odds
of finding one with great potential for happiness or achievement? Would it be acceptable to settle for a child expected
to be merely happy or modestly successful, when, by going
a few more cycles, the parents would be likely to get an
ecstatic or brilliant one? Or is any such selectivity incompatible with the duties of prospective parents in bearing
142 |
children and forming a family; in establishing uniquely intimate relationships and uniquely inclusive associations?
These are among the questions that arise if the categorical
anti-natalist challenge is rejected.
In chapter 7, I will review what I regard as the three
principal anti-natalist arguments: comparative, consent,
and risk. The arguments claim, respectively, that existence involves harms that cannot be morally outweighed
by its benefits, given the alternative of nonexistence; that
the benefits of existence, however great, cannot justify its
inevitable harms without a consent that is impossible to
obtain; and that prospective parents must seek to avoid the
worst possible outcome for their child, which they can only
do by not procreating. I will contend that none of these arguments succeed in establishing that bearing a child necessarily wrongs her. They do, however, show that procreation
is an activity fraught with risk, which should be undertaken only with great caution and serious reflection.
In chapter 8, I claim that even if the categorical antinatalist arguments are mistaken, they suggest that we
must have a good reason to bring a child into the world,
given the serious harms and grave risks she will inevitably face. That reason cannot concern the impersonal good
of a new lifethe net good it brings to the worldor it
will give all of us a reason to procreate whenever adding
a child would make a positive contribution. Rather, I will
consider the reasons that a parent can offer the child for
having brought her into a difficult and dangerous world
reasons that must in some way concern her own expected
good. I will argue that such reasons can justify that decision to her without implying any moral reason to have
created her, or to procreate again. This justification can
| 1 4 3
be found in two closely related reasons, often expressed

by prospective parents: to give the goods of a life initially
lived with the parents and their family to a now-unknown
or unknowable child; and to form an intimate relationship, in which those goods initially will be conferred, with
a now unknown or unknowable child. To be able to offer
this justification, these must have been among the reasons for having the child. But these reasons do not require
the creation of a new being; they can also be reasons for
adoption. They are similar to the reasons people can, and
perhaps should, have for seeking a variety of intimate relationships with unknown partners.
This account does not directly address anti-natalist
claims that the expected goods of a future life, whether
intended or not, cannot be balanced against the expected
bads. But it offers prospective parents a kind of justification many pro-natalists have been reluctant to acknowledge: that the risks and costs of procreation to the future
child, if not to other individuals, can be offset by the value
to the child of the goods it is intended and expected to
enjoy, and the value of the parent-child relationship it is
intended and expected to enter.
I will begin chapter 9 by distinguishing two approaches
to setting moral standards for procreation. The first treats
same-number choices by prospective parents about which
child to have as governed by impersonal considerations
considerations about increasing the good or reducing the
bad in the world at large. The second approach treats procreative choices as grounded in the birthrights of children or
the role-based duties of prospective parents. Most writers
who adopt this approach see those rights or duties in person-affecting terms: held by or owed to present or future
14 4 |
individuals, whose violation wrongs those individuals. I

will focus on the first approach in chapter 9. I will argue
that defining the domain of procreative decisions in terms
of same-number choices is untenable, because that is a
domain that is difficult to define, arbitrary in scope, and,
ultimately, unable to treat procreation as a strictly permissive enterprise.
In chapter 10, I will consider birthright and rolebased accounts. I will begin by examining the former accounts, which impose a minimum standard of expected
well-being as both a necessary and sufficient condition for
permissible procreation. I will then consider two kinds of
role-based accounts. The first claims that prospective parents have a role-based duty to select the future child expected to enjoy the most well-being, or experience the least
suffering. I will then present and defend a second kind of
role-based account, which denies that selection among
future children is morally required, as long as their lives
are expected to be worth living, and holds that such selection may be incompatible with role of prospective parents.
I will conclude chapter 10 by considering the moral
force of family, community, and global interests as constraints on procreative decisions. No plausible account of
the duties of prospective parents can ignore the actual
and opportunity costs of having and raising children.
These concerns are raised by the decision to have, or risk
having, more children or particular kinds of children, such
as those expected to have expensive medical needs. I will
argue that in assessing costs, it is critical to distinguish
the roles of prospective parents from those of public officials. The latter have role-based duties that may conflict
with those of prospective parents. Potential conflicts are
| 1 4 5
limited by the fact that, in a just society, parents should

bear some of the costs of the children they have chosen to
bring into the world, and are entitled only to limited partiality toward those children. Those conflicts can also be
mitigated by the availability of means for containing costs
that do not directly infringe the procreative liberty of
prospective parents.19
NOTE S
1. Elizabeth Anscombe, Why Have Children? in Proceedings
of the American Catholic Philosophical Association 63 (1989):
4853, 52.
2. Gilbert Meilaender, The Blessing of Children, The New Atlantis [pdf version 9298, 92, originally appearing Summer
2012], accessed April 18, 2014, http://www.thenewatlantis.
com/publications/the-blessing-of-children.
3. David Heyd, Genethics: Moral Issues in the Creation of People
(Berkeley: University of California Press, 1994).
4. David Benatar, Better Never to Have Been (Oxford: Oxford
University Press, 2006).
5. Seana Valentine Shiffrin, Wrongful Life, Procreative Responsibility, and the Significance of Harm, Legal Theory 5,
no. 2 (1999): 117148.
6. Matti Hayry, A Rational Cure for Prereproductive Stress
Syndrome, Journal of Medical Ethics 30, no. 4 (2004):
377378.
7. Christopher Belshaw argues that the uncomprehending
pain and distress we, or our biological precursors, suffer in
gestation and early infancy are great enough to outweigh
whatever good follows (or to make us guilty of exploiting
those precursors to experience that good). See A New Argument for Anti-Natalism, South African Journal of Philosophy
31, no. 1 (2012): 117127. Although he does not offer this as
an empirical claim about the badness of gestation and early
14 6 |
infancy for those undergoing them, his position appears to

rest on empirical assumptions about the badness of animal
and early infant lifeassumptions that we cannot verify
and that many would likely find implausible.
8. Stuart Rachels, The Immorality of Having Children, Ethical Theory and Moral Practice 17, no. 3 (2013): 116.
9. Indeed, total consequentialists must accept what Derek
Parfit has called the Repugnant Conclusion: that if it continued to maximize the good, we should continue having
children with lives of steadily diminishing well-being indefinitely, or until everyone had a life just barely worth living.
Reasons and Persons (Oxford: Clarendon, 1984), 381390.
10. Daniel Friedrich, A Duty to Adopt? Journal of Applied Philosophy 30, no. 1 (2013): 2539; Christina Lyn Rulli, The
Duty to Adopt (PhD diss, Yale University, 2012); Preferring a Genetically-Related Child, Journal of Moral Philosophy
(2014) online, http://booksandjournals.brillonline.com/
content/journals/10.1163/174552434681062 (accessed
February 5, 2015).
11. David DeGrazia, Creation Ethics: Reproduction, Genetics, and
Quality of Life (Oxford: Oxford University Press, 2012); Christine Overall, Why Have Children? The Ethical Debate (Cambridge, MA: MIT Press, 2012).
12. DeGrazia, Creation Ethics, ch.5.
13. DeGrazia, ch. 6; Bonnie Steinbock, Wrongful Life and Procreative Decisions, in Harming Future Persons, eds. Melinda
Roberts and David Wasserman (London: Springer, 2009),
155178.
14. John A. Robertson, Procreative Liberty and Harm to Offspring in Assisted Reproduction, American Journal of Law &
Medicine. 30, no. 1 (2004): 730.
15. Margaret Olivia Little, Abortion, Intimacy, and the Duty
to Gestate. Ethical Theory and Moral Practice 2, no. 3 (1999):
295312.
16. Yes: Samuel Scheffler and Niko Kolodny, Death and the Afterlife (New York: Oxford University Press, 2013); No: Mark
Johnston, Is Life a Ponzi Scheme? Boston Review, January 2, 2014, accessed April 18, 2014, http://bostonreview.
| 1 4 7
net/books-ideas/mark-johnston-samuel-scheffler-deathafterlife-humanity-ponzi-scheme.
17. Allen Buchanan et al., Reproductive Freedom and the Prevention of Harm, in Bioethics, eds. Nancy Jecker, Albert
Jonsen, Robert Pearlman (Sudbury, MA: Jones and Bartlett
Learning, 2011): 416425; Julian Savulescu and Guy
Kahane, The Moral Obligation to Create Children with
the Best Chance of the Best Life. Bioethics 23, no. 5 (2009):
274290.
18. Jakob Elster, Procreative BeneficenceCui Bono? Bioethics 25, no. 9 (2011): 482488; Thomas Douglas and Katrien
Devolder, Procreative Altruism: Beyond Individualism in
Reproductive Selection, Journal of Medicine and Philosophy
38, no. 4 (2013): 400419.
19. I thank my older son, Jacob, for suggesting the title of this
chapter (and theme for this essay). The question mark is
mine.
Against Anti-Natalism
THE CLAIM THAT IT IS categorically or presumptively
wrong to create new human lives has been based on several

kinds of arguments. All concern the moral significance of
harm, but none rest exclusively on empirical claims about
how harmful life is or would be in particular times and
places. Explicitly or implicitly, all understand harm, or specific kinds of harm, not as the mere absence of good, but in
terms of discrete bads such as pain, suffering, injury, frustration, loss, and death. For Benatar, the absence of pain is
good whether or not there is someone for whom it is good,
whereas the absence of pleasure is merely not-bad unless
there is someone for whom it is a deprivation. This asymmetry gives nonexistence an advantage in comparison with
any existence containing even the slightest pain.1 For Shiffrin, the unconsented infliction of harm can be justified
to the individual harmed only by the avoidance of greater
harm, not by the bestowal of pure benefitsbenefits
A gainst A nti - N atalism
| 1 4 9
that do not prevent or ameliorate harm.2 For Hayry, the

realistic possibility, however slight, of catastrophic harm,
has greater prudential and moral weight than the likelihood of benefit, since the most reasonable rule to adopt
for a future child is one of avoiding the worst possible
outcomea rule that precludes procreation.3 I will refer to
these, respectively, as the comparative, consent, and riskbased arguments.
The three arguments differ in the reliance they place
on facts or predictions about the world into which future
children will be born. The expectation of any serious harm
suffices to make procreation presumptively wrongful for
the consent argument, whereas the risk argument requires
a realistic possibility of disastrous harm. According to the
comparative argument, the certainty of some harm makes
existence disadvantageous for the individual created,
while the near certainty of grave harm makes her creation
wrongful.
The comparative argument differs from the other two
in a significant respect. It concerns what is bad or harmful in procreation, whereas the other two concern what is
wrong in it. Neither of the latter two claims that it is necessarily disadvantageous to come into existence; neither
rests on the net badness of existence. Criticism of the comparative argument has focused on the claim that bringing
someone into existence is always a net bad or harm for that
individual. I am more interested in how such a harm would
constitute a wrong, a claim that is developed in Benatars
philanthropic and misanthropic arguments. I will conclude
that none of the anti-natalist arguments establish even a
presumptive wrong in procreation.
150 |
B E N ATA R S C O M PA R AT I V E , Q U A L I T Y
OF LIFE, AND MISANTHROPIC
ARGUMENTS
Benatars Comparative Argument
According to Benatar, it is against the interest of anyone
to come into existence, because the advantage of moving
from the not-bad of absent pleasures in nonexistence to
the good of pleasure in existence is outweighed by the
disadvantage of moving from the good of absent pain in
nonexistence to the bad of pain in existence. The latter
outweighs the former because the absence of pain is good,
even if there is no one for whom it is good, whereas the
absence of pleasure is bad only if it is a deprivation for
someone.4 Several critics have questioned the coherence
of this claim on the grounds that there can be no subject
for the good of avoiding bad by nonexistence, since good
cannot be attributed to merely possible people.5 Other
critics have maintained that even if such an attribution
can be made, Benatar is mistaken in the way he values
absent pain and pleasure in nonexistence.6 Although I
think there is much to these objections, I will not at present dispute Benatars claim that existence is a bad bargain for the person created, in the sense that it is not in
her best interests. My concern is rather with whether
a person who is living a good life, as conventionally assessed, would have grounds for complaint against his parents for this bad bargainwith whether such a person
could be regarded not only as disadvantaged by this bad
bargain, but as wronged.
| 1 5 1
Benatar himself invites this separate appraisal by asserting that [t]he conclusion that existence is always a
harm tells us nothing about the magnitude of that harm7
(emphasis in text). There is no reason to think that if the
harm were slight, it could not be counterbalanced by the
benefits of life, or many lives, so as not to constitute an allthings-considered wrong.
Of course, Benatar argues that the harm is almost
always greata claim I will address in the next subsection. Here, I want to examine whether, by itself, the harm
claimed by Benatars wide-comparison argument could
give rise to a moral complaint on the part of the individual
harmed; whether, for the person created, the benefit of a
good life could weigh against the harm sufficiently to nullify or trivialize any complaint.
Benatar does not appear to deny that this is a formal
possibility. He does not regard his comparative argument
by itself as making the case that all procreation is wrongful. That case also requires his philanthropic argument,
which is designed to show just how bad our lives really are.
But if it is the magnitude of the harm that gives rise to a
complaint, not the conclusion that life is always a harm,
then it is not clear what role the comparative argument
plays in reaching that conclusion.8 If careful scrutiny and
critical assessment could show that life was very harmful
overall for everyone, or almost everyone, then why would
it matter for purposes of a moral complaint that it was also
disadvantageous in comparison with nonexistence? The
extremely high odds of a very bad existence would make
procreation wrongful on any reasonable decision rule for
risk or uncertainty. Benatars philanthropic argument
would then serve a similar function to the risk argument
152 |
discussed below, showing that procreation was unacceptably risky. It would, however, be more robust than the risk
argument in relying not on the controversial maximin rule
but on a bleak appraisal of outcomes and a rough estimate
of probabilities.
Before assessing the philanthropic argument, I want
to consider what sort of complaint a child with a great life,
conventionally assessed, would have against his parents
if the comparative argument were correct. It is instructive to compare the complaint Unlucky Child (UC), with
an awful life, could make against his life, within which
the bads vastly outweigh the goods, with the complaint
available to Lucky Child (LC) for whom the opposite is the
case.9 Benatar would appear to hold that UC has a complaint against his parents for the poor quality of his life
that LC lacks. Admittedly, LC, like UC, has been disadvantaged just by coming into being. As Benatar states, although [LC] may be lucky relative to other children (who
suffer more than he does), [he] is not lucky enough if he
is actually brought into existence (emphasis in text).10
His interests would have been better served by not being
created.
But even if this claim grounds an intelligible complaint,
it seems a pretty feeble one. LCs parents, whom we may
assume have been conscientious as well as lucky, could respond, Sorry to set back your interests, but we really wanted
a child and we expected that you would flourish. Sure, life is
full of risks, but we had good reason to expect that yours
would be highly rewarding on balance. Benatar would think
the parents badly self-deceived about their childs prospects,
but thats a separate issue. If their assessment is reasonable
(or reasonably mistaken), this appears to be an adequate
| 1 5 3
response to his complaint. (I will argue later that they can

also claim to have acted at least in part for his good in creating him, even if they were to accept that nonexistence would
have been more advantageous for him in Benatars sense.)
Indeed, the contrast between the two children suggests that
Benatar displays an insufficiently robust appreciation of the
vast difference in their quality of life, and, more generally, of
the vast differences in how well the lives of different people
go. Even if we have a strong tendency to ignore, forget, or
discount the pain and suffering in our lives, these differences remain profound. If UCs parents had good reason
to expect his life to be as awful as it is, then he has a very
serious complaint against them for a life they expected, or
should have expected, to be awful.
Benatar does recognize that UC suffers two kinds of
disadvantage, LC only one. First, the pains within UCs life
vastly outweigh the pleasures; second, his very existence
gives him good and bad instead of the good and not-bad of
nonexistence. LC, in contrast, suffers only the latter disadvantage. Although I have suggested that LCs complaint on
that basis is frivolous, I have assumed that it is coherent;
one can always complain that one has not received what is
in ones best interests. But it seems a perverse complaint for
LC to make, one that would be belied by his attitude toward
the life he leads. He presumably loves that life and craves
more happy years. It is an understatement to say that from
his ex post perspective he has no regrets about having been
created. Not only did he lack an ex ante perspective; he does
not see his ex post perspective as one he is merely resigned
to, or incapable of escaping. If, like the protagonist of countless time-travel stories, he could make it the case that his
parents never conceived him, he emphatically would not do
154 |
so. Benatar might respond that to do so would merely be

a form of preemptive suicide, given that he already exists,
or else that that this was a thought experiment he could
not perform. But it remains the case that even if Benatar
is correct that existence was not in his best interests, he
has given LC no other, better reason to feel the slightest
regret about it. So it could be argued that LCs complaint
would be in bad faith, because he is wholeheartedly glad
that this parents did not act to prevent it. In contrast, UCs
complaint could be made in perfectly good faith.
This distinction between LC and UC would collapse if
there were an independent wrong in creating LC as well
as UC; if the bounty of life was his only because his parents had violated some right of his in providing it. There
need be no inconsistency in enjoying the windfall but still
regarding its source as a rights-violation. But I have suggested that Benatar does not make a case in his comparative argument that causing the harms of existence is a
wrong, as opposed to a harm. It would be a wrong if (prospective) parents had a duty to maximally satisfy the interests any child they might create, which for Benatar would
require them not to create any children. But apart from
familiar issues about the identity of the rights-holder, it
seems odd to claim that (prospective) parents have a correlative duty. Even the most spoiled child should recognize
that his parents do not have a duty to maximally satisfy
his interests; that their own interests, as well as those of a
myriad of others, appropriately limit their duty to satisfy
his. If bringing someone into existence is always a wrong,
it can only be so because it is always a net harm. But I have
argued that the lucky victim of this putative wrong could
not reasonably regard it as such.
| 1 5 5
Benatars Quality-of-Life Argument

Benatar would argue, however, that few, if any, children are
actually lucky. He regards existence as massively harmful
to almost all individuals, and attributes the widely shared
impression that it is not harmful to recalcitrant Pollyannaism and profound self-deception.11 If he is correct, then
LC is wronged as well as UC because his parents could have
had no reason to believe he would beat the overwhelming
odds and have a great lifeeven if, however improbably,
he actually has or will have had such a life. No genetic constitution or external environment can be proof against the
slings and arrows of such pervasive misfortune.
As well as claiming that life is truly awful for a great
many existing peoplea claim that few pro-natalists
would denyBenatar insists that it is very bad for almost
all of us. The second claim is critical to his version of antinatalism, since the first alone does not show that procreation is wrongful. Many of Benatars critics have claimed
that a lot of prospective parents, especially in more developed nations, have good reason to believe that their
children are very likely to lead good lives, conventionally
assessed.12 They know, of course, that their lives may go
very badly, as badly as those of the worst off in a developing
country, but they may reasonably believe the odds of such
a dreadful life are very slight. Without a radical reappraisal
of lives conventionally regarded as good overall, Benatar
could only reject procreation on the basis of a maximin
rulea rule that requires the avoidance of the worst possible outcome, however improbable it may be.13 And that
rule, as I will argue in discussing Hayry, is unreasonably
risk averse.14
156 |
Benatar takes up this challenge, arguing that our lives

those of well-off people in more developed countries in the
early 21st centuryare really much worse than we think
they are on any plausible theory of well-being. He provides
an error theory for our mistakenly optimistic judgments in
the adaptive value of those false judgments for evolutionary purposes. Basically, he argues that we have far more and
more intense aversive mental states and unfulfilled desires
than we acknowledge, and that our objectively valuable activities and achievements fall far short of any reasonable
threshold for a good life.
Concerning aversive experience, Benatar argues that
[w]e tend to overlook the extent to which we experience
negative mental states. Among the most common (even if
not always the worst) [are] hunger, thirst, bowel, and bladder distention (as these organs become filled), tiredness,
stress, thermal discomfort, (that is, feeling too hot or too
cold) and itch. . . .15 These are certainly common states, but
Benatar appears to overlook the extent to which they are
locally balanced out, or even mildly pleasant in context. In
our comfortable lives, we experience these states, to the
extent we experience them at all,16 with the reassuring or
even pleasing anticipation of their relief. It is not even clear
that it is at all aversive to be hungry just before a meal,
thirsty after a vigorous workout, etc. Our knowledge of the
almost certain, immediate, and pleasurable alleviation of
these sensations may make their very experience innocuous or pleasurable. And their alleviation usually is as well,
canceling out any residue of unpleasantness (Of course, it
is possible to dwell on the aversive experience, as in the
joke about the old man who repeatedly declares Boy, am
I thirsty! then, after gulping down a tall glass of water,
| 1 5 7
repeatedly declares Boy, was I thirsty! But his theatrical

preoccupation with a mildly aversive and easily alleviated
state is amusing because it is perverse.) It is a very different, far more aversive experience to be thirsty in a desert
or hungry in a faminebut these are rare occurrences in
our privileged world.
More serious aversive states cannot be regarded the
same way. I doubt many asthmatics enjoy their breathing
difficulties in anticipation of their vaporizers, or migraine
sufferers their headaches in anticipation of eventual relief.
But such unquestionably aversive conditions (not to mention far worse ones) may heighten the appreciation of
those experiencing them of the long periods in which they
are absent. And many chronically painful or uncomfortable conditions are subject to a kind of adaptation that reduces their aversive character, somewhat like adjusting to
a bright or dark room. This is not Pollyannishness or selfdeception, but an actual change in experience, a kind of
natural anesthesia. Of course, not all injuries and diseases
are subject to this benign process. But those suffering traumatic injury often experience a state of numbness before
the pain sets in, and their pain can often be preempted by
powerful analgesics.
A second problem with Benatars hedonic assessment is
that we experience our lives at various removes from particular experiences, and it is not clear why one perspective
should be privileged over another. Why shouldnt the perspective of tranquil recollection count as much or more in
a hedonic calculus as that of the untranquil states that are
recollected? Certain sequences of sensations are meant to
be experienced as a whole as well as in part, retrospectively
as well as concurrently. Why should we give more weight to
158 |
the average or sum of mental states during a roller-coaster

ride (assessed during the ride) than to the experience of
relief and exhilaration immediately afterwards, in which
the anxiety, terror, and pleasure of the ride play an integral role? This is a question that, writ large, needs to be
addressed on any hedonic account.
Benatar succinctly states his case for thinking that life
in generally bad in terms of a second account of well-being,
based on the satisfaction of desires:
[T]he general pattern is a constant state of desiring punctuated by some relative short periods of satisfaction. Therefore, there is very good reason to think we spend more time
unsatisfied than satisfied.17
In its emphasis on comparative duration, this sounds like a

hedonic version of a desire-satisfaction account. What matters on standard accounts, though, is merely the satisfaction or frustration of desires, whether experienced or not,
rather than feelings of desire, satisfaction or frustration.
In the terms of a standard account, Benatar asserts that
most people have many more unsatisfied than satisfied desires. But he does not attempt to survey people, count their
desires, or suggest how desires might be individuated; his
argument does not rest on such a quantitative assessment.
Rather, it appeals to our yearning for the impossible and/
or improbable, to the undeniable fact that our reach is very
often much farther than our grasp.
It is difficult to deny that the loftier our desires, the
more likely they are to remain unsatisfied. But if a desiresatisfaction account leads to the conclusion that it is bad to
live a life of self-conscious striving for unattainable goods,
then so much the worse for desire-satisfaction theories. As
| 1 5 9
John Stuart Mill famously asserted in Utilitarianism: It is

better to be a human being dissatisfied than a pig satisfied;
better to be Socrates dissatisfied than a fool satisfied.
Benatar gives the verdict of a desire-satisfaction account intuitive appeal with his hedonic inflection: the frustration of desires is a highly aversive experience, while their
satisfaction gives only transient pleasure. The experience
of frustration can indeed be deeply unpleasant, especially
conjoined with jealousy or humiliation. But the tenacious
pursuit of a goal that one that expects to be unattainable
need not produce frustration in this sense; rather, it may
lead to the satisfaction of a higher-order desires to be resourceful and persevering in trying to satisfy first-order
desires, and to pride in whatever partial successes one has
achieved.
If Benatar sometimes treats desire-satisfaction in hedonic terms, he also offers an objective appraisal of our
desires as overly modest and unimaginative: [W]e would
have even more unsatisfied desires if we did not restrict
some of our desires to the realm of the possible.18 He gives
as an example the fact that most of us do not desire to live
to 1000 years, but at most to the current limits of human
longevity. Benatar thinks we constrain such desires from
a lack of imagination or to avoid inevitable frustration. It
is, however, not clear that a desire-satisfaction account
has any business prescribing desires; at most, it considers the desires we would have if we were fully informed.
But unless full information is stretched to include the
appreciation of the myriad benefits of conditions that are
currently impossible to achieve, then our lack of desire to
enjoy those benefits does not seem attributable to our ignorance. Rather, Benatar appears to be claiming that our
160 |
lives would be more valuable if these conditions obtained.

If so, Benatars claim does not concern what we do or
would desire, but what is really desirableit is a claim
about objective well-being. And Benatar goes on to propose
an objective account which virtually guarantees a massive
shortfall.
His claim is that in assessing the quality of human
lives for purposes of deciding how good human life in
general is, it is appropriate to assess human lives not in
comparison with other human lives, but from the point
of view of the universe. The quality of human life is then
found wanting.19 This raises two questions: First, is it even
meaningful to assess human life, or any life, from that
viewpoint; second, even if it is meaningful to do so and
life is found wanting, why would that verdict mean that
human life was objectively bad?
Concerning the first question, it can be argued that the
universe does not have a point of view; to attribute one to
it is to anthropomorphize or deify it. It does not value at
alleven to attribute indifference to the universe is to humanize it. We might be able to adopt the view of some superior species that landed on Earth, so as to perceive our
lives as sadly wanting. Since we have not had the dubious
fortune to encounter such a species, however, why should
we be tempted to adopt their viewpoint? Moreover, as Bernard Williams has pointed out, we might discover a nowdormant partiality toward our biological humanity, and
prefer a life with human imperfections to life in which we
would become smarter, stronger, kinder, and longer-lived,
at the price of estrangement from our human ways of life.20
Admittedly, we can imagine tremendous improvements in
our capacities that would leave us recognizably human.
| 1 6 1
And we can reasonably conclude that our lives might have

been much better with such improvements. But, As Aaron
Smuts observes, Benatar fails to see that the opposite of
better is not always bad; sometimes, it is just less good.21
An ordinary human life in a developed country in
the early 21st century may well lack many of the valuable experiences and achievements of ordinary lives a
century laterat least if the world as a whole does better
than Benatar thinks it will. But if it is arbitrary to assess
our lives objectively against existing norms or averages,
it seems no less arbitrary to assess them against the
norms of any more advanced society. We might hold that
above some minimum level of objective functioninga
minimum that is highly debatablewe may reasonably
choose our own comparison class, or if possible, avoid
comparisons altogether. No doubt, it might be devastating to acquire and then abruptly lose some of the physical
and mental enhancements that contemporary visionaries
seek.22 But although we may be able to imagine having
some of them, and can even yearn for them, their absence
does not make our lives go badlyunless, like some of
those visionaries, we obsess over their absence. Benatars
assessment of humanity from what amounts to a visionary perspective is an effective antidote for the complacency expressed in the old beer commercial that concludes
it doesnt get any better than this. Benatar reminds
us that it could indeed get much better. But that hardly
means that the imperfect lives we lead in the shadow of
such possibilities are bad.
No one reading Benatars account of the suffering experienced by individuals who die of fairly common diseases, such as cancers, can fail to take seriously his claim
162 |
that even the best lives can end very badly. Prospective
parents should certainly be aware that their children are
likely to die from some disease that may have very painful
symptoms, and they would be naive to assume that medical or palliative breakthroughs will dramatically improve
their childrens odds of living longer and dying well. Nor
should they discount the disturbing prospect of social and
economic upheavals that will make their childrens lives
worse than their own. But they need not be either insensitive or naive to reject Benatars conclusion that almost all
lives, realistically appraised, will be too harmful to start.
This is so for several reasons.
First, as I have stated before, there is no authoritative
vantage point from which to assess the most physically or
psychologically painful experiences in life. Benatar implicitly privileges the testimony of individuals currently
in or just out of the throes of suffering. He certainly would
be warranted in doing so against a claim that such experiences really werent so bad made by someone who
had never gone through them or largely forgotten their
intensity. He jests at scars that never felt a wound is a
fair reproach to someone who naively belittles the experience of suffering, but not to someone who gives it less
weight in the overall evaluation of a life than Benatar
does. Indeed many in the throes of a painful final illness
still regard their lives as well worth living, even if they do
not die surrounded by loved ones or narcotized into bearable pain without loss of awareness. Benatar may think
they are mistaken and self-deceived. But it is not clear on
what basis he can invoke their authority about their experiences while dismissing the weight they give them in the
overall evaluation of their lives.
| 1 6 3
Or consider an even more routine experience: giving

birth. Despite the early onset of labor amnesia, many
women recall it being the most painful experience of their
lives, for which nothing could have prepared them. Some
declare in the throes of labor that they wish they had never
decided to get pregnant; that it is simply not worth the
pain. But few conclude better never to have given birth,
even though most are aware of the (undoubtedly adaptive)
amnesia that makes their worst pain inaccessible in a way
that other awful pain may not be. Many go on to have other
children, aware as they are of how painful the final hours
of pregnancy can be. Why should their experience in labor
be authoritative, or have decisive weight, in determining
whether having a child contributes to their well-being?
Now that women in developed countries rarely die in childbirth, weespecially menmay unduly discount labor
pain as insignificant in the larger scheme of things. But
almost no one gives the most painful experiences of labor
the authoritative weight in evaluating an individuals life
that Benatar gives the most painful experiences of cancer.
This may be because even the most protracted labor is comparatively brief, because it is overshadowed by the frequent
joys of childbirth, because there is obvious adaptive value
in discounting its weight, or because it comes in the middle
of a life, not at its end. But Benatar needs to defend his selective invocation of immediate experience in his qualityof-life argument.
More broadly, there is a great deal of philosophical debate
about how to weigh the goods and bads of a life, and its various parts, in assessing life as a whole.23 Benatar must argue
for a particular approach to such weighing in order to categorically reject those who, at or near lifes end, judge their
164 |
lives to have been well worth living despite long and intense
periods of pain and suffering at various stages.
Second, individuals have very different experiences
of the physical conditions Benatar describes, and, if those
conditions are not terminal, live with them in very different ways. Thus Benatar treats paralysis as a terrible fate,
and for some people it undoubtedly is. But for others it is
definitely not. They find ways to live rich lives with and despite their limitations, and not only because they may have
a robust euphoria. Thus consider this short description of
her life from Connie Panzarino:
At 42 years of age I am mostly paralyzed; have full feeling; I cannot swallow food unless it has been pureed in a
blender, use a BiPAP for respiratory problems; use a puff
n sip wheelchair; take medication for my heart, stomach,
and body pain; and must be repositioned by my PCAs every
20 minutes. I also run my private psychotherapy/art therapy practice; own my own home and van; serve on several
boards; maintain my sexual relationship with my lesbian
lover; pet my cat with my chin; take my blender out to
dinner with friends (and blend lobster, or whatever I like);
travel; show the artwork I make by mouth or computer;
write; read; plant a garden, and on and on. I have made a
choice to live as fully as I can.24
In this passage, there is no reliance on feelings of pleasure

or satisfaction, no claim that those feelings are impervious to the biomedical realities of Panzarinos life. Rather,
there is a detailed description of valuable activities and
achievements coexisting with those realities. Some will
undoubtedly find Panzarinos appraisal of her life, as presented in this passage, unconvincing as evidence of wellbeing. In any case, I am not claiming that her account has
| 1 6 5
more authority than those of people in similar physical

conditions with similar histories who regard their lives as
constricted and demeaning. My claim is merely that in the
absence of an authoritative vantage point, prospective parents in a privileged society like ours can reasonably take a
rosier view of their childrens prospects than Benatar does
without being at all Panglossian.
Of course, even in the most privileged society, lives can
go horribly, and no prospective parents can ensure that the
lives of their own children will not. But they can reasonably conclude that the odds of such fates are low enough
to be acceptable. To deny that such an assessment is reasonable, Benatar would either have to defend a specific calculus of risk that would reject it, or else adopt a principle
of extreme risk-minimization for procreation, a position
I criticize below. Benatars risk argument gains spurious
strength, I suspect, from the Russian roulette simile he
employs, which has prospective parents pointing a gun at
the head of their future child; a gun with a high proportion of chambers loaded.25 This simile is misleading. As the
literature on the ethics of risk imposition and distribution
points out, it matters a great deal if the threatened harm
will be imposed intentionally. Shooting someone with a
loaded gun is intentionally harming him, even if the discharge of a bullet had been far from certain. As David DeGrazia points out in discussing Shiffrins harm argument,
parents do not impose harm on their children so much as
expose them to it, while making a concerted effort to avoid
or mitigate it.26 And that is a far different matter for a nonconsequentialist than imposing harm.
I believe there is an unavoidably comparative aspect
to acceptability judgments. Prospective parents may find
166 |
appropriate reinforcement for their decision to have children in the expectation that those children will have lives
that are no harder or riskier than their own. One of the
legitimate concerns for baby boomers deciding whether
to have children in the past two decades is the lack of such
assurance. Arguably, prospective parents should not create
children they expect to have lives substantially harder or
riskier than their own, or at least lives they would find unacceptable for themselves. Such a constraint would be akin
to that imposed on combat officers not to impose risks and
hardships on the soldiers they command, especially conscript soldiers, that are any greater that those they assume
or are willing to assume themselves. Although I lack the
space to examine this suggestion, it raises a difficult question about the moral constraints on prospective parents, a
subject I will address in the last chapter.
Benatars Misanthropic Argument
Benatar has recently advanced a misanthropic argument
against procreation, based on the incalculable harm that
humans inflict on each other and other species.27 I can only
respond briefly, in part because I strongly disagree with Benatars weighing of the suffering of minimally-sentient animals,
a disagreement we cannot resolve here. I certainly do not
doubt the enormous cruelty that almost all humans are capable of, and that many continue to exercise, with ever more
lethal effects. But unlike Benatar, I think this aspect of humanity is well appreciated, at least by regular viewers of the
evening news. I also disagree with Benatar in believing that
many of those viewers recognize that they themselves would
be capable of engaging in similar atrocities under different
| 1 6 7
circumstancesthe vivid demonstrations of this capacity by

Milgram, Zimbardo, and others have become well known in
our society.28 Where I disagree most sharply with Benatar,
however, is on the implications of human destructiveness
and cruelty for individual procreative decisions.
Setting aside the important but disputed issue of harm
to animals, I believe that many prospective parents can
reasonably conclude that their own children are unlikely to
commit, or be significantly complicit in, the kind of harms
that would (or should) make them ashamed of having borne
them. I think that in reaching that conclusion they should
recognize that some prospective parents who reached it
have been and will be mistaken. But this recognition should
merely make them more cautious; it should not be decisive
unless they themselves harbor grave doubts about their capacity to create and nurture minimally decent children.
As I will argue in the last chapter, I do no think prospective parents must universalize about the likely consequences if everyone judged or acted as they do. I think
the concerns about the consequences if everyone did it
have far more relevance for policymakers than prospective parents.29 Although the latter must be cautious in
their predictions about their children and may reasonably
have concerns about the fairness and cumulative impact
of similar decisions, I believe that they do not need to give
this the same weight in their decisions as policymakers or
other impartial third parties should in theirs.
At the same time, Benatars arguments rightly emphasize the moral importance of serious reflection and commitment on the part of prospective parents. They must be
prepared not only to teach their children to respect their
peers and their superiors, but to oppose them if they suggest,
168 |
initiate, or demand immoral conduct. Prospective parents

must also be prepared to inculcate the kind of discernment
necessary to recognize at an early age when conduct is immoral. It is not unrealistic to expect this effort. One has only
to look at the large number of young Americans who, in the
past fifty years, actively protested their countrys foreign interventions or fought for minority civil rightsoften with
the active support of their parents.
THE CONSENT ARGUMENT

Shiffrin argues that prospective parents cannot presume
the consent of any child to being born.30 She observes that
all human life contains serious harms, such as pain, loss,
frustration, and death. We can presume consent for imposing any such harms only to avoid greater harms, not for
harms imposed to permit pure benefitsthose that do
not involve harm avoidance. Since there is no harm in not
coming into existence, procreation confers a pure benefit.
Shiffrin does not claim that serious harms cannot be outweighed by benefits, but that the balance cannot be struck
vicariously. The only fact about human existence on which
Shiffrins argument relies is the same one on which Benatars comparative argument relies: all lives contain some
significant harm.
In illustrating the moral difference between conferring
pure benefits and avoiding harms, Shiffrin offers examples
of postnatal interventions. A doctor who comes upon an
unconscious accident victim can amputate his arm to save
his life or legs, but not to endow him with greater strength
or intelligence. We would object to the latter at least in part
| 1 6 9
because we expect people to differ far more in the tradeoffs

they would be willing to make between harms and benefits
than between different similar harms of different magnitude.31 Amputating for pure benefits would be presumptuous and paternalistic. But procreation does not face a
similar objection: a merely possible person has no values or
preferences to override.32
Shriffrin attempts to fortify her case with a more fanciful hypothetical, in which a wealthy man can enrich the
residents of an island, with whom he cannot communicate,
only by air-dropping gold cubes, predictably breaking an
islanders arm. She maintains that the wealthy man acts
wrongly, although his victim will be crying all the way to
the bank.33 The infliction of benefits, however great, cannot
justify the infliction of harms, however minor, without
actual consent. I suspect many people will agree, or at least
share the intuition that the air drop is morally problematic.
Two factors, however, make it problematic in ways that
bringing a child into existence is not. First, bombing the
unlucky islander appears to violate his rights, however
much it enriches him. We recognize rights against the imposition of certain kinds of harm, such as battery, though
not (except in special contexts) against the net loss of benefits. But there is no reason to assume that people have
rights against the infliction of all non-comparative harms.
The actions of the wealthy man would have been less objectionable if he had found means of enriching the islanders
that did not involve such a clear-cut rights violation. Say,
for example, that he had dropped the gold in an uninhabited location, knowing that the sudden, influx of wealth,
even if distributed fairly, would cause slight social and cultural disruption. This harm would be inflicted on all the
17 0 |
residents who benefited from the increased wealth, whether

or not they tacitly consented to the benefit by appropriating the gold.34 But I suspect that few people would think
that he acted wrongly, so long as the disruption was minor
and all of the islanders were better off despite the disruption. Any remaining misgivings would arise, I believe,
from a concern that the disruption would actually not be
minor: that the influx of wealth would violate strong cultural or religious norms or erode social relationships and
institutions.
The second distinguishing feature of the air drop is
that the islanders are stipulated to enjoy modest comfort
already, so that the benefit of great wealth seems slight or
uncertain. It is by no means clear that the additional wealth
would improve the lives of the inhabitants in any but a
narrow material waywe have many anecdotes and much
evidence that would lead us to doubt this. Indeed, it may be
that in purifying benefits to be free of a harm-prevention
or harm-reduction role, Shiffrin has reduced them to mere
luxuries. Perhaps, as Rivka Weinberg suggests (in her personal correspondence) it is difficult to make benefits more
substantial without giving them a harm-prevention aspect.
In the developed world, at least, significant increases in
wealth come with significant increases in health and longevity, with greater avoidance of disease and death. The
one substantial good of which this is clearly not true, if it
is a benefit at all, is existence. So the difficulty in purifying
postnatal benefits without making them luxuries does not
threaten Shiffrins basic claim. It does, however, threaten
her reliance on postnatal benefit cases to defend it.
In contrast to the wealthy gold-bomber, it is not clear
that parents would violate any right by bringing a child
| 1 7 1
into existence, although existence brings unavoidable harms.

Unlike in postnatal interventions, there is neither a physical
incursion on an individual with a right against such incursion,
nor a disruption of plans and attachments already formed.
Moreover, in contrast to the slight, uncertain, or two-edged
benefits from the islanders sudden enrichment, the benefits
of life can be enormouseven if the harms can be as well.
Another significant contrast is that, unlike the goldbomber, parents do not generally cause or impose the more
serious harms their children suffer, but merely expose
them to those harms. DeGrazia, who makes this distinction, argues that parents faultlessly expose children to
harm all the time, often as part of creating opportunities
for greater benefits.35 He gives the example of sending
children to school, to which they often vocally dissent, and
which exposes them to a variety of significant harms for
the sake of rich opportunities.
Shiffrins understanding of the moral gravity of harm
does have powerful intuitive appeal. Certain experiences
and conditions are bad regardless of the social context
or overall eudaimonic status of the individual suffering
them. Acute physical pain, the death of a loved one, the
destruction of ones life work, rejection or betrayal by
those one most loves or respectsthese are bad whatever
else is bad or good, unfortunate or privileged, about ones
life. This does not mean that one may never inflict such
non-comparative harm on one person for the sake of some
benefit to another, or for some good or imperative not reducible to harm or benefit to particular people, such as corrective justice. But it does suggest a presumption against
causing harm, and a priority to avoiding it, however elusive
that presumption or priority may be as a guide to action.
17 2 |
One consideration supporting this priority, which seems

to have influenced Shiffrin, is epistemic. In effect, she reverses Tolstoy, who claimed that all happy families were alike
while each unhappy one was miserable in its own way. Shiffrins assumption seems to be that while there are an indefinite number of ways to flourish, there are a limited number
of ways to suffer; whereas we have an irreducible plurality of
conceptions of the good, we recognize only a few basic forms
of badness. But apart from epistemic considerations, it may
appear that duties to prevent or not inflict harm are stronger
than duties to confer or not prevent the receipt of benefits.
It is hard, though, to be more precise about this difference,
in part because of doubts that there can be a common metric
of goods and bads; doubts dismissed by skeptics about the
special moral status of harm like James Griffin.36 Indeed
merely adopting a single metric, as Griffin does, assumes
what many proponents of a priority would deny: that harms
and benefits can be so readily compared.
We are often willing to trade off harms against benefits, as Griffin notes, even when the harms and benefits
go to different people. But we generally regard such tradeoffs as constrained by rightsrights more often against
harms than to benefits. Intuitions, however, vary widely
on what those constraining rights are, and even if there
were greater agreement, any priority given to harm avoidance would be complex and context-dependent.37
THE RISK ARGUMENT

Hayry gives priority not to harm in general, but to the risk
of the worst possible harm.38 He argues that in deciding
| 1 7 3
whether to create a life, the only rational strategy is Rawlss

maximin, which requires maximizing the positions of the
worst-off people. As adapted by Hayry, this rule requires
minimizing the worst possible outcomes, and thus prohibits procreation, which can have disastrous consequences
for the child created. Since the maximin takes no account
of the probability of being in the worst-off group, the infrequency or slight likelihood of such a great misfortune is
irrelevant. And since, like Benatar and Shiffrin, Hayry sees
no harm in nonexistence, that is clearly the better alternative. The decision to have a child, with a very slight but
foreseeable possibility of a life not worth living, is morally
wrong as well as irrational. . . .
[S]ince potential parents cannot guarantee that the lives of
their children will be better than non-existence, they can
be rightfully accused of gambling on other peoples lives,
whatever the outcome. Because of the uncertainties of
human life, anyones child can end up arguing that it would
have been better for them if they had not been born at all.
The probability of this outcome does not necessarily matter.
It is enough that the possibility is real, which it always is.39
It is hard to deny the possibility that anyones child could

have a life that she reasonably regarded as not worth living.
And it would certainly seem ad hoc to assign a numerical
threshold to the expected odds of such a disastrous outcome, above which it would wrong the child to create her.
But this hardly implies that any odds are too great. That
requires a further argument. Hayry does not supply one,
except to condemn progenitors for gambling on other peoples lives. However evocative the phrase, it does not do the
work Hayry requires unless any gamble is unacceptable (or
174 |
any gamble with disaster as a possible outcome). As DeGrazia notes in response to Shiffrin, we do not look askance
at many significant gambles parents take with their childrens lives, as long as a reasonable parent could conclude
that the payoffs were worth the risks.40
Perhaps Hayry would argue that many postnatal gambles seem acceptable only because there is a real possibility of disaster no matter what option is chosen. Take your
child out and you risk her being fatally struck by a car; leave
her at home and you risk her dying in a fire. Because there
is nowhere to run, nowhere to hide, the one safe haven
is nonexistence. Since the parents could have chosen that
uniquely safe option, they cannot excuse their risky postnatal gambles by claiming, correctly, that any choice would
be a gamble.
This reconstructed argument, however, needs the maximin to rescue it from the charge of implausible risk aversion. When we condemn people for gambling with others
lives, we condemn them for imposing unacceptable risks. I
do so when I drive drunk from a party, but not when I drive
sober to a store. As countless authors have argued, we individually and collectively make many decisions that impose
very small risks of disastrous outcomes on unknown individuals, from driving cars to building roads.41 And our
tradeoffs are hardly limited to risk vs. risk; we trade off
risks with security, comfort, and mobility. Even when the
trade-offs are explicit, we choose within a range of options, none of which is anywhere close to the safest possible option. Transportation statisticians produce tables
showing how many lives we would expect to lose or save
by raising or lowering the present highway speed limit. The
differences are substantial, but a 20 mph limit would have
| 1 7 5
few takers, let alone a ban on nonemergency driving. Even

if we cannot justify the trade-offs we make except procedurally (i.e., democratic decision making with adequate
information, participation, and deliberation), we do not
regard them as unprincipled, just imprecise. We also recognize that reasonable people can have different risk tolerances in different times, places, and contexts.
So if zero tolerance is to be the only moral option in
procreation, we need an argument as to why that is so. Although Rawls argues that the strong risk aversion of the
maximin is appropriate for decisions that affect future
generations, we cannot derive an argument against procreation from his rationale for applying the maximin to decisions about the basic structure of society behind a veil
of ignorance. Rawls suggests that a rule of minimizing the
worst outcome (i.e., choosing the arrangement in which the
worst outcome is better than in any others) is appropriate
when 1) uncertainty about outcomes is too great to assign
probabilities; 2) the affected individuals have conceptions
of the good such that they care very little about outcomes
above the minimum and very much about outcomes below
it (i.e., It is not worthwhile for them to take a chance for
the sake of a further advantage, especially when [they] may
lose something important to [them] . . .) and 3) the rejected alternatives involve grave risks ... that they ...
can hardly accept.42 Although 1 and 3 do seem to apply
to creation, 2 clearly does not. Of course, possible people
do not have their own conceptions of the good, although
Hayry could safely assume that they would if they were
made actual. But on almost all reasonable conceptions,
the goods foregone through non-conception are goods
that any actual person would have reason to care about
17 6 |
immenselyall the goods of life. Those goods would be unimportant from a pre-conception perspective only if, per
Benatar, their mere absence counted as not-bad. Creation
certainly involves grave risks, with possible outcomes that
are unacceptable. But Rawlss rationale for adopting the
maximin rule does not justify the refusal to weigh the benefits of existing against its risks. As Weinberg suggests, applying Rawlss condition to procreation would require not a
categorical ban, but a reasonable assurance of a minimally
good lifea condition I will return to in the final chapter.
C O N C LU S I O N
The plausible idea on which Benetar, Shiffrin, and Hayry
all rely is that the failure to create a life worth living is not
a harm to anyone. For Benatar, treating nonexistence as
a harm would preclude his claim that coming into being
was always against the individuals interests. For Shiffrin,
treating nonexistence as a harm might give prospective
parents a duty to bring people into existence to avoid that
harma duty that, however weak, presumptive, or prima
facie, few of us recognize. For Hayry, treating nonexistence
as a harm would deny it the status of an safe haven that is
has in his risk argument.
The shared weakness in all their arguments, I suggest, is their reliance on the special urgency or priority of
avoiding harm as grounds for claiming the opposing duty
to not bring anyone into existence. For Shiffrin, that reliance is explicit in her prohibition of vicarious trade-offs
between harms and benefits; for Benatar, it is reflected
in his attempt to explain the priority of avoiding harm
| 1 7 7
over conferring benefits in terms of his own asymmetry

between existence and nonexistence. It is also implicit, I
think, in the appraisal of lifes goods and ills he makes in
his philanthropic argument. And for Hayry, the priority
of harm avoidance supports the maximin rule he believes
that prospective parents are required to adopt.
NOTE S
University Press, 2006). For other purposes, as in assessing
how well a life goes, Benatar adopts a (more) comparative
notion of harm, treating it as the mere absence of benefit.
Although it may be appropriate to treat harm comparatively
in that context, since the lack of benefit can be seen as a
deprivation, Benatar seems to elide these two notions in
claiming that the benefits of existence can never outweigh
the benefit/good of nonexistence so as to make existence a
net benefit.
2. Seana Valentine Shiffrin, Wrongful Life, Procreative Responsibility, and the Significance of Harm, Legal Theory 5,
no. 2 (1999): 117148.
3. Matti Hayry, A Rational Cure for Prereproductive Stress
Syndrome, Journal of Medical Ethics 30, no. 4 (2004): 377.
For Belshaw, the net harm of infancy makes life not worth
living, however beneficial it is thereafter. I will discuss only
the first three anti-natalist arguments, because Belshaws
appears to involve an asymmetry of life stages as much or
more than an asymmetry of harm and benefit. (Christopher
Belshaw, A New Argument for Anti-Natalism, South African Journal of Philosophy 31, no. 1 (2012): 117127.
4. It does seem that, in treating existence as a harm to the
one created, Benatar is shifting from a noncomparative to
a comparative sense of harm. He implicitly uses harm
in the former sense in framing his asymmetry in terms of
17 8 |
the good and not-bad of absent harms and benefits; he uses

harm in the latter sense in claiming that it is not in the
interests of an individual to be, or to have been, created.
5. David DeGrazia, Is It Wrong To Impose the Harms of
Human Life? A Reply to Benatar, Theoretical Medicine and
Bioethics 31, no. 4 (2010): 317331; Elizabeth Harman,
David Benatar. Better Never To Have Been: The Harm of
Coming into Existence, (Oxford: Oxford University Press,
2006), Nos 43, no. 4 (2009): 776785.
6. Harman, Benatar; David Boonin, Better To Be, South
African Journal of Philosophy 31, no. 1 (2012): 1025; Aaron
Smuts, To Be or Never to Have Been: Anti-Natalism and a
Life Worth Living, Ethical Theory and Moral Practice 17, no.4
(2014): 711729.
7. Benatar, Better Never To Have Been, 61.
8. As Rivka Weinberg points out (in personal correspondence)
he strengthens his case by defending an asymmetry between avoiding suffering and experiencing happiness. But
even with that further claim, he would need more to establish that procreation is not merely non-negligibly harmful
but wrongful.
9. The names come from Boonin, Better To Be.
10. Benatar, David. Every Conceivable Harm: A Further Defence of Anti-Natalism, South African Journal of Philosophy
31, no. 1 (2012): 128164, 138.
11. Benatar, Better Never to Have Been, ch. 3.
12. For example, DeGrazia, Is It Wrong? 324329; Smuts, To
Be or Never To Have Been, 1017.
13. Actually, maximum requires the maximization of the minimum outcome, which is a stronger requirement implying
the avoidance of the worst possible outcome.
14. Jason Marsh has recently made a strong case for what he
calls procreative skepticism (e.g., Quality of Life Assessments, Cognitive Reliability, and Procreative Responsibility, Philosophy and Phenomenological Research 89, no. 2
(2014): 436466. He argues that a life may still be good despite its inevitable bads, but that we have no basis for confidently concluding that this is the case for any given life,
| 1 7 9
present or future. He criticizes Benatar for inferring from

our strong evaluative biases that our lives are bad, or almost
all bad, rather than inferring that they are not as good as
we typically think they are. Because of our biases, though,
and the difficulty comparing various goods and bads in our
lives, we cannot have reasonable confidence that any future
life will be good. Marsh hesitates to offer moral prescriptions to prospective parents based on his analysis. But it
clearly supports risk aversion in deciding whether to have
children and a greater willingness to decline on the basis of
unacceptable prospects.
15. David Benatar, Still Better Never to Have Been: A Reply
to (More of) My Critics, The Journal of Ethics 17, no. 12
(2013): 121151, 141.
16. DeGrazia, Is It Better, 325.
17. Benatar, Still Better,143.
18. Still Better.
19. Benatar, Better Never to Have Been, 86.
20. Williams, Bernard. The Human Prejudice In Philosophy as a Humanistic Discipline, eds. Bernard Williams and
A.W. Moore (Princeton: Princeton University Press, 2006),
135152.
21. Smuts, To Be or Never to Have Been, 17.
22. Max More and Natasha Vita-More, eds. The Transhumanist Reader: Classical and Contemporary Essays on the Science,
Technology, and Philosophy of the Human Future (New York:
John Wiley & Sons, 2013).
23. For a recent discussion, see Joshua Glasgow, The Shape of
a Life and the Value of Loss and Gain. Philosophical Studies
162, no. 3 (2013): 665682.
24. Connie Panzarino, What Is Choice, and Who Is Choosing?
Roll Call 1990 (October/November), 7.
25. Benatar, Better Never to Have Been, 92.
26. DeGrazia, Creation Ethics: Reproduction, Genetics and Quality
of Life (Oxford: Oxford University Press), 153154.
27. This volume, ch. 5.
28. Harry Perlstadt, Milgrams Obedience to Authority: Its Origins, Controversies, and Replications, Theoretical & Applied
180 |
Ethics 2, no. 2 (2013): 5377; S. Alexander Haslam and Stephen D. Reicher, Contesting the Nature of Conformity:
What Milgram and Zimbardos Studies Really Show, PLoS
Biology 10, no. 11 (2012): e1001426.
29. I am speaking here of consequentialist univeralization,
which is quite distinct from the Kantian injunction to act
under maxims acceptable as universal law. The latter is
claimed to be a general requirement or characterization of
acting morally. I lack the space or expertise to adequately
explain the distinction.
30. Shiffrin, Wrongful Life.
31. The contrast would not be as great if the surgeon amputated the arm of an unconscious adult to prevent a different kind of harm (e.g., to prevent his false conviction of a
crime that could only have been committed by a two-armed
perpetrator).
32. Wrongful Life, 126127. This is not to deny that we would
object to a mothers arranging the amputation of her fetus
to endow it with great intelligence. But I think we would
object far less strongly than in the case of an adult, and not
only because of the lesser moral status of the fetus. That
intervention would lack the objectionable paternalism in
Shiffrins original case.
33. Wrongful Life, 127128.
34. Thanks to Sean Aas for this point. It poses a defense against
the example DeGrazia attributes to Frances Kamm of islanders suffering psychological stress over how to distribute the gold they appropriate (Degrazia, Creation Ethics,
162, n.28).
35. Degrazia, Creation Ethics, 153.
36. James Griffin, Is Unhappiness Morally More Important
than Happiness? Philosophical Quarterly 29, no. 114 (1979):
4755.
37. One difference between harms and benefits, though not a
categorical one, may lie in the comparative burdens of the
duties to avoid or alleviate harm and to confer benefits. On
a global level, this distinction has lost force as we come to
recognize the wide range of cumulative harms to which we
| 1 8 1
contribute (Judith Lichtenberg, Distant Strangers: Ethics,

Psychology, and Global Poverty (Cambridge, Cambridge University Press, 2014) ch 4.). But if we restrict those duties
to a much smaller population (e.g., of family and friends)
the duty to alleviate harm to the extent we can becomes, at
least in most circumstances, much less oppressive. In contrast, the duty to confer benefits remains oppressive, since
we can almost always confer additional benefits on family
and friends who have quite modest projects and appetites.
It may be that the duty to avoid as much harm as possible,
with suitable population restrictions, demands much less
of us than the duty to confer as much benefit as we can on
the same population, and that for this reason, we regard the
former as a more stringent duty. But this is hardly a difference Benatar or Shiffrin could exploit to defend the priority
of avoiding harm, since as they (differently) interpret that
priority, it imposes an onerous presumptive duty not to
have childrena duty that many would find more demanding than any duty of beneficence.
38. Hayry, A Rational Cure.
39. A Rational Cure, p. 378.
40. DeGrazia, Creation Ethics, 153154.
41. See, for example, David McCarthy, Rights, Explanation,
and Risks, Ethics 107, no. 3 (1997): 205225.
42. John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1979), 154155.
The Good of the Future

Child and the Parent-Child
Relationship as Goals
of Procreation
BENATAR AND SHIFFRIN STRESS a point acknowledged by
many pro-natalist philosophers: that there is no basis for

claiming a harm to a possible person in failing to make
her actual.1 And if there is no harm in not coming into
existence, the avoidance of harm cannot help to justify the
harms and risks of existence. Nor is there any other moral
duty or reason to bring future people into existence, no
matter how bright their prospects.
This last claim is one side of what Jeff McMahan has
called The Asymmetrythe widely shared, deeply held
conviction that that there is a strong moral reason not to
have a child expected to lead a miserable life, but no moral
THE GOOD OF THE FUTURE CHILD
| 1 8 3
reason to have a child expected to lead a great life.2 This

position, however difficult to defend, appears to be a tenet
of the commonsense morality of procreation.
If, however, we have no moral reason, let alone duty,
to bring potentially happy children into the world, can
we still have good reasons for doing so? Even if we confidently reject the anti-natalist arguments that it is categorically wrong to do so, we should be keenly sensitive to
the concerns they raise about the unavoidable harms and
grave risks facing a child in even the most safe and comfortable environment. What can the parents of any child
say to justify their decision to bring him into the world if
they cannot claim to have had a moral reason, let alone a
duty, to do so? Their own desire for a child, and the desire
of others, may not be bad reasons, but if they are they only
ones, prospective parent s face the charge of having created
the child as a mere means to their own fulfillment.
In this chapter, I will argue that prospective parents
can, and often do, have a different, albeit closely related
reason for having children, a reason that concerns the
good of the future children as well as their own. I think
that reason can go along way to meeting the charge that
they created their child as a mere means, or with insufficient sensitivity to the risks and harms it would face. But
in order to offer that reason as justification to an actual
child, or so I will argue, it actually must have been among
their reasons for bringing a child into existence.
This reason concerns the value for the child of the kind
of life they intend and expect it to live, and the value for
the child and themselves of the kind of relationship they
intend and expect to establish. Much of this chapter will be
devoted to defending three claims: (1) prospective parents
184 |
can act in part on this reason, (2) if they do, they can adduce
the value of the childs life for that child, and the value of
the parent-child relationship for all of its participants, in
justifying risks and harms to the child, and (3) if they do
not act on this reason in having a child or continuing a
pregnancy, they cannot adduce those values to justify the
risks and harms the child faces. I will address the obvious
objection that the only values prospective parents can seek
concern their own good, the good of others, or the impersonal good. None provide an adequate justification to the
child facing the harms and risks of any human life. I am
more confident of the first two claims that the third. That
claim requires a broader defense that I can give here of the
necessity of appropriate intentions for the permissibility
of a great many morally significant actions.
Relational and Child-Centered Reasons
forHaving Children
I am, of course, hardly the first to suggest that the good of a
future child, nested in a parent-child relationship, can be
a reason for procreation. Susanne Gibson held that:
. . . reasons for having child may be judged morally desirable or undesirable according to the extent to which they
enhance or detract from the particular kind of relationship
with that child. The goals of this relationship will be many,
although one of the most important goals will be to aid the
child in developing a sense of her own value ... 3
This passage is vague about how reasons for having

children can enhance or detract from the parent-child
relationshipby their psychological effects on it, or by
| 1 8 5
providing an appropriately child-centered basis for establishing it. Christine Overall (from whom I take the Gibson
quote) moves closer to the second view. Although denying
that prospective parents can create a child to advance her
interests, because she does not preexist her own conception,4 Overall holds that they can seek a relationship in
which the good of the child is an integral part:
The best reason to have a child is simply the creation of the
mutually enriching, mutually enhancing love that is the
parent-child relationship. In choosing to become a parent,
one sets out to create a relationship, and in a unique way,
one also sets out to create the person with whom one has
that relationship.5
Overall views the choice to become a parent as entirely optional; the commonsense view that is part of the Asymmetry. Gilbert Meilaender, however, takes her as treating
procreations merely as a personal project:
Having children is for her an entirely individual project
intended to deliver a product. It will require some collaborators, of course, but a parent is simply a person who undertakes and manages such a project.6
Overall may not regard a child, like Meilaender, as a mysterious gift and blessing,7 or the attempt to have a child
as an expression of gratitude to a Higher Power for ones
own life. But there are alternatives to his theological view
besides the consumerist one Meilander attributes to her.
An individual seeking to create a child with whom she will
have a unique kind of mutually enriching, mutually enhancing relationship is hardly a project manager trying
186 |
to make a product. Not all people have such a yearning,

and not all people should become parents. But it verges
on caricature to treat the often-difficult decision to have
a child or continue an accidental pregnancy as the choice
of a project, akin to a couples New Years resolution to enliven their social lives by making more friends in their new
neighborhood.
Overalls best reason for having a child is reflected
in the views of many prospective parents. The motivational pull of doing good for a future unknown child is
made explicit in a recent study of the reproductive attitudes and preferences of young men, which found the
five most influential reasons for fatherhood to be as
follows:8
I want to share what I have and what I know with a child.
I want the special bond that develops between a parent and
a child.
Raising a child would be fulfilling.
I want to give love and affection to a child.
I would give a child a good home.
Only the third of these reasons sounds even remotely like

it treats procreation as an entirely individual project, let
alone as treating the future child as a product. The selfish,
relational, and child-centered aspects are all woven into
one prospective fathers expression of interest:
Being a father will really feel wonderful, to have someone
who I helped create and to have my own child means that I
would have someone to protect and be there for.9
| 1 8 7
None of these remarks suggest that the young men were

motivated by a desire to bring new people into existence
except as a precondition of nurturing them. I am not claiming that most prospective parents would express or act on
the reasons they gave, but that those reasons are perfectly
intelligible. They require neither a metaphysically premature
attachment to a possible person nor recourse to a theological posture of gratitude. As I will discuss later, those reasons
could equally motivate prospective adoptive parents.
But this leaves the question of how, or whether, the
future child can be treated as an end in herself in trying
to create her. Overall insists, and I agree, that prospective
parents should aim at a relationship in which the actual
child is treated as an end in herself. Can a stronger claim be
made? In seeking to establish such a relationship, can they
act for reasons that concern the good of the future child?
A number of philosophers have doubted that prospective
parents can act for such a reason, even in an attenuated
sense.10 In the next section, I will attempt to analyze and
address the most plausible grounds for those doubts. I will
not claim that prospective parents can treat future children as ends in themselves in the same way they can treat
existing children. I will merely argue that they can act for
reasons that concern the good of those future children,
reasons they can adduce to justify to them their decision
to bring them into existence.
How Can There Be Child-Centered Reasons
forProcreation?
I will defend a justification for having children that many
pro-natalists have been reluctant to recognize: not merely
188 |
that prospective parents can act permissibly in bearing a

child they expect to have a worthwhile life, but that in justifying that decision to the child who results from it, parents can adduce the good of the life and relationship they
sought with an unknown and unknowable child. Like the
reasons that often motivate the attempt to establish other
respectful and mutually beneficial relationships, these reasons concerns the good of the unknown partner, but give
rise to no duty to create that relationship in order to confer
that good.
That justification may fail, however, for any number of
reasons. The prospect of giving the child a good life may be
slight because of an extremely harsh or inhospitable environment. The prospect of forming a mutually rewarding
parent-child relationship may be slight because external circumstances are likely to disrupt it, or because the prospective
parents are unlikely to sustain an adequate commitment. In
either case, prospective the parents acted negligently toward
the child they had, even if they hoped to beat formidable
odds and were lucky enough to do so.
Further, it is not enough that the justification available
to prospective parents is adequate for the (future) child. It
may still leave them open to the charge that they breached
a duty to others or displayed a lack of more general beneficence. They may violate duties to specific existing individuals, especially their existing children and other individuals
whose welfare depends on their limited resources, time, and
effort. Procreation also may have considerable opportunity
costs, requiring the prospective parents to forego valuable
projects, whether humanitarian, intellectual, or artistic.
Often, prospective parents could have conferred the
goods of a shared life and a loving relationship by adoption,
| 1 8 9
which would also have rescued an actual child from the

prospect of a grim life.11 Admittedly, adoption may also
pose substantial risks of harm, both to the adopted child
and its biological parents. The child may have been better
off living with a poor, even dysfunctional birth family than
torn away and placed in a wealthier, better-functioning
adoptive one. The judicial termination of their parental
rights may have been based on error or prejudice. The biological parents may have been coerced into giving up a child
they loved and sought to nurture, or exploited by monetary
incentives too good to refuse in their desperate circumstances. Nonetheless, there may be means of protecting
against these risks, and a strong case can be made that prospective parents should consider adoption seriously before
deciding to procreate.
Thus I am hardly offering a complete or unqualified defense of procreation in claiming that prospective parents
who seek the good of a future child can adduce that good to
justify the risks and harms it will face. I will argue in the
final chapter that prospective parents are partially exempt
from some of the moral demands faced by citizens in general. For example, they need not take account of the social
costs of having a medically complex child to the extent
they should take account of the social costs of buying a gasguzzling SUV. The adverse impact likely to result if everyone acted as they did is, I will claim in chapter 10, less of
a moral constraint on prospective parents.12 But that is not
to deny that they are morally accountable to many others
beside the future child, or to claim that a justification adequate for that child will be adequate for those others.
My focus here, however, is not on the accountability of
prospective parents to others, but to the child they create.
190 |
Creating a child for reasons that concern its own good is as

close as they can come to treating a future child as an endin-itself, and not as a mere means. I will therefore refer
to such reasons as respectful, using the Kantian idiom
without making any claims about the direct applicability
to procreation of Kants third formulation of the Categorical Imperative: Act in such a way that you treat humanity,
whether in your own person or in that of another, always
at the same time as an end and never merely as a means.13
As I will argue in chapter 10, the claim that prospective parents can and do seek the good of future children
in having them does not suggest that they should seek the
child for whom they can do the most good. Nor does it lend
support to the view, discussed in chapter 9, that those bearing children have a pro tanto duty to bear a child with the
best possible life.14 It might well be perverse, and disrespectful, for prospective parents to seek a child who will have a
less happy life than others they might have.15 But, as I will
argue in chapter 10, that does not imply that they must seek
a better-off child. It is reasonable for prospective parents to
be neutral among future lives expected to exceed a minimal
threshold of net good or well-being.16 This neutrality is not
moral laxity; it reflects a lively appreciation of the separateness of persons, future as well as actual.
Here is an example to give some flesh, and plausibility, to the idea that prospective parents can create children for reasons that concern the good of the children,
or at least their shared good. Consider a couple who very
much want children and decide to adopt. They are normally
fertile, but are moved by the need to find homes for the
many orphaned children in their country now housed in
institutions. This, however, is not their primary reason for
| 1 9 1
adopting; it merely tips the balance. Their reasons include

wanting the fulfillment of raising a child from a young age,
seeking the uniquely intimate relationship that a child develops with its parents, and giving the child a good home
among the reasons given in surveys of prospective parents.
They regard these as reasons that could be served equally
well by adoption or conception. Just as they are going to
start visiting orphanages, their government prohibits
adoptionorphans and abandoned children will be wards
of the state, with temporary foster parents in special cases.
The couple is very disappointed but quickly decides to go
with Plan Bthey conceive a child for the same reasons.
The point of this example is not just to illustrate that
adoption and procreation may be done for similar reasons.
As important for my purposes, it suggests the limited role
that the actual vs. contingent existence of the child may
play in the sorts of reasons prospective parents have. The
couple in my example starts by seeking to find a child of
their own who already exists, or whose existence is not contingent on their actions. Barred from doing so, they shift
to creating a child. But their reasons for doing the latter
are largely the same as their reasons for having sought the
former. The desire to help existing needy children was just
a tiebreaker.
The limited motivational role of actual existence is
suggested by comparing procreation to the pursuit of other
intimate relationships, where the existence of the other is
assumed. You can seek a close, caring relationship with an
unknown adult in part for that persons goodthat relationship will be good for that other was well as for you.
People in search of (ISOthe phrase in pre-internet personal ads) friends and lovers are often motivated in part
192 |
by the belief that they have so much to give, as well as

to receive, in a friendship or romance. You can likewise
seek a relationship with a future child in part for its own
goodyou want to give a good life and a loving, nurturing
relationship to a new being. Why should it matter that at
the time you decide to procreate, the barrier to being motivated by the good of another particular individual is not
only epistemic but metaphysical?
Jeff McMahan makes perhaps the strongest argument
against the view that the good of a future child can serve
as a reason for bringing him into existence.
There may be good for the individual who comes to exist
but it cannot be ones reason for acting, or ones intention
in acting, to bestow that good on that individual. At most,
ones reason might be to create additional noncomparative
good which would then necessarily attach to and be good
for someone. But the creation of that good would not be
better for that individual, nor would not being created have
been worse for him or her.17
The two italicized demonstratives (that) in the quoted

passage reflect, I think, two related problems, concerning
the benefit and the beneficiary McMahan finds in the view
that we can intend to benefit an individual by creating him.
With respect to the first, McMahan claims that existence
itself can be good for the individual who receives it, but not
better for that individual than nonexistence. With respect
to the second, McMahan contends that there is no individual to whom ones beneficent intentions could attach
at the time the procreators act. I will concede both claims
but deny that they support the conclusion that we cannot
intend to benefit children in creating them.
| 1 9 3
The Benefit: I think that McMahan is mistaken in assuming that the benefit sought is, or must be, existence
itself. The reasons I have defended for bearing a child do
not treat existence as an initially free-floating good waiting to attach to a possible being. It may be that procreating is the only feasible or acceptable way to give a child
the goods of a particular kind of life and create a parentchild relationship. But if not, these are equally good reasons for adoption. The goods sought are among those that
the would-be adoptive parents in my hypothetical initially
sought. When they went to Plan B, they need not have seen
themselves as giving an additional gift, existence; they just
may have seen that they could not give a child the goods
of a particular kind of life, and an intimate relationship
with them, unless they brought one into existence. They
still shared with prospective adoptive parents the end of
giving love, nurture, and a home to a child.
Those goods, unlike existence itself, can be better for an
actual child (better, say, than the alternative of a loveless
home or an orphanage), whether it exists independently of
its adoptive parents efforts, or came into existence because
of the procreative acts of its biological ones. For both adoptive and biological parents, the childs bare existence is
a necessary condition for fulfilling their primary end but
is not, or need not be, a primary end in itself. (Nor, in an
ordinary sense, is causing the childs existence a means to
that end; existence is a presupposition or foundation of,
not a way or means to, love and nurture.)
Importantly, the goods that prospective biological
parents seek to bestow are not ones they have any moral
reason to bestow. They can be seen as playing what McMahan calls a canceling role,18 outweighing the harms and
194 |
burdens of life. In that limited role, they (or their expectation) are merely necessary for an act of procreation to
be permissible. Although a parent-child relationship can
be immensely valuable to its participants, and should be
valued by their society if it is, prospective parents have no
more moral reason to seek it than they have to seek any
other valuable intimate relationship. It is only if they are
seeking such a relationship that they have a moral reason
to confer the goods associated with, or integral to, that relationship on their future, unknown partner.
Prospective parents then are only seeking to give
the goods of life to a future child in the context of a specific kind of relationship, however much those goods will
exceed the scope and duration of that relationship. They
seek goods that their children will enjoy apart from their
relationship and care deeply about how their lives will go
when they are no longer around or alive. But these are the
concerns of future parents; concerns associated with a particular kind of intimate relationship. The provisions they
make in response to those concerns are those of a future
parent, not those of an impersonal benefactor seeking to
bring a little more person-affecting good in the world.
Prospective parents can certainly seek the bare existence of a biological child to prove their fertility, to keep
up with their peers, to please their relatives, to perpetuate
their family lines, to help raise their countrys dwindling
population or workforce, or even, I suppose, to increase the
total good in the world. But none of these reasons concern
the good of the future child. The fact that parents often act
on such other reasons hardly means, though, that they do
not also (often in the same cases) create a child in order to
love and nurture it, a reason that does concern its good.
| 1 9 5
There is a corresponding distinction from the childs

perspective. Most children, to the (limited) extent they express gratitude to their parents, express it for their love,
nurturing, and support, not (or rarely) for their very existence. Adoptive children may also, if they are aware of
their origins, be extremely grateful for their rescue; biological children are unlikely to be grateful to anyone for
having been rescued from the limbo of nonexistence. But
this merely reflects one difference in parental reasons for
adoption and procreation, noted earlier. Otherwise, prospective parents have many or most of the same reasons to
procreate and adopt: adopted and biological children have
many or most of the same reasons to be grateful.
At the same time, prospective parents do not create
children in order to provide them with discrete goods like
a trust fund or a comfortable house. Rather, they seek to
bestow the more encompassing goods of family and home,
material and psychological securitygoods that will
enable the child to grow and flourish. Overall observes that
it is unsurprising that many people want to pass on their
property and money to their children, but to have children
only for this purpose puts the cart before the horse. Handing down an inheritance benefits the children, but to have
children in order to hand down an inheritance means that
one is having a child in order to benefit the inheritance.19
This might be the case if the funds would otherwise escheat to the state; a child created solely to prevent that
outcome would be created as a mere means to wealth preservation. But creating a child to give it a nurturing environment, a home and family, is not creating her in order
to confer a discrete gift; it is placing her in a network of
mutually rewarding intimate relationships.
196 |
The Beneficiary: If the benefit prospective parents

typically seek to confer is not existence, but goods that presuppose existence, what of the beneficiary? If they seek to
confer more abundant and secure goods, they face a moving
target. Pursuing many of the goods prospective parents
seek to bestow is identity-affecting, since it involves delaying conception. Were they seeking to bestow existence on
a particular possible child, their actions, by delaying conception, would be self-defeating. Prospective parents who
wait to have children until they get tenure, make partner,
or move to a comfortable house and neighborhood make
it the case that they will create different children than if
they had not waited. They seek to do the best for any child
they have.
Clearly they cannot do so with the particularized intention they would have to give the same benefits by the
same actions to an existing child. But their reasons can
be child-centered even when the only child who will benefit is a future one, whose identity is contingent on their
actions. This is apparent from the fact that some parents
who seek such goods for future children also seek them
for existing onesthey seek a more comfortable home or
better schools both for children to whom they have a particularized attachment and for intended future children,
to whom they cannot have such attachment. These reasons
may have greater psychological strength and moral weight
for existing than future children, but they seem child-
centered in both cases.20
Although prospective parents do not and could not
seek to bestow a family and home on any particular future
child, they seek to give those goods to their child, to a child
with whom they will have a lifelong intimate relationship.
| 1 9 7
They might be equally willing to bestow those goods on a

biological or adoptive child, but they rarely want them to
attach to someone elses childto a child, whatever its
origins, who is to be raised by strangers. They also want to
be the ones who bestow the good. Indeed what they typically want to bestow requires that they rear the child until
it is ready to care for itself, or to go off to some educational
institution that will help care for him or her.
The fact that prospective parents want to make it the
case that they give myriad goods to their children may mean
that there is an inextricably selfish aspect to their benevolence. But this feature hardly distinguishes them from many
other benefactors we regard as acting for the good of the
beneficiaries. Parents of existing children usually want to be
the primary source of the tangible and intangible goods they
receive, and they want to ensure that their own children and
not others are the recipients.21
In sum, prospective parents want to share their lives,
fortunes, and goods with someone, now unidentified or
unidentifiable, who will become their child. Since they
are motivated by the goal of giving and sharing, they
can be said to act in part for the good of a future child.
The good they seek for themselves is not only compatible
with the good they seek for the child; both are aspects of
a mutually rewarding parent-child relationship. As Adrienne Asch describes the aspirations prospective parents
have for almost any future child: we seek to give ourselves to a new being, who starts out with the best we can
give, and who will enrich us, gladden others, contribute
to the world, and make us proud.22 In seeking such a relationship, prospective parents are seeking the good of a
future child.
198 |
This anticipatory concern for the good of an unknown

other is reflected in the pursuit of a variety of other intimate relationships, with persons whose existence is not
contingent on the search. In some of these relationships,
one or both parties are expected to provide material and
psychological security. Provisions for future intimate relationships with unknown partners were once common;
in many parts of the English-speaking world, unmarried
women would fill hope chests with important household items in expectation of eventual marriage to a thenunknown man. They did not seek a husband in order to
furnish their household; they sought an intimate relationship of which their furnishing a household was a part. Is
it different when prospective parents buy a house with an
extra room for a child they hope to have? Or can they act
for the good of the child or the shared goods of an intimate
relationship only when they can identify or apprehend the
child as a particular individual? Can they act for the good of
a future child when they choose its room before it has been
conceived? Or can they do so only when they choose the
rooms color based on the childs sex, ascertained in utero23
(a practice that may someday go the way of hope chests)?
Like the goods set aside in a hope chest for a future
husband, the material provisions made for a future child do
not provide reasons to establish the relationship in which
they will be conferred. The intangible goods of matrimony
or child-rearing do, or can. But those seeking such a relationship have a moral or prudential reason to acquire, give,
or share those goods, tangible or intangible, only if they
form such a relationship. And they have no moral reason to
form such relationships rather than forming other kinds
of relationships or conferring other kinds of goods.
| 1 9 9
I do not want to press the analogy to other prospective

relationships too far. In attempting to establish a mutually
rewarding relationship with an unknown child or adult who
exists, or will exist independently of ones efforts, one is
seeking to make someones life better. This cannot be said of
prospective parentsthey only seek to make someones life
good, not bettereven if, as I have argued, they can seek
that good for the sake of their uncreated child. As in the
comparison to adoptive parenting, this does mark a significant moral difference I would not want to deny or obscure.
Are Child-Centered Reasons Necessary
for Permissible Procreation?
There is an ongoing debate in moral philosophy about the
relevance of intentions to permissibility; about whether or
when bad intentions can make an otherwise good or neutral action impermissible, or vice versa. Most of this debate
has focused, as Matthew Liao points out, on whether bad
intentions can make an otherwise morally required act
wrongful. But there has also been some discussion about
whether certain acts must be done for specific reasons or
with specific intentions to be permissible. There is broad
agreement in the case of some types of acts. Thus most
people agree that it is wrong to make a promise intending
to break iteven if the occasion never arises. Many also
hold that the promise also must be made with the intent to
fulfill it, which requires the expectation that the promisor
be able to do so.
Similarly many people also would agree that it is wrong
to have a child intending to sell it into slavery, to reduce
the inheritance of ones younger sibling, or even, perhaps,
200 |
to save ones marriage. My claim is stronger and similar to

the second claim about promisingthat procreating can
be wrongful not only if it is done for bad reasons, or with
bad intentions; it must be done for good reasons, or with
good intentions. Those good reasons, as I have suggested,
concern the good of a particular kind of life for a future
child, and the good for her and her prospective parents of
the relationship she will enjoy with them.
I can offer no general account of why certain types of
action morally require certain kinds of intentions or reasons; I am not sure there is one. But procreation, I believe,
is subject to such a requirement because the actual child
is entitled to a respectful reason for having been brought
into a world where she is exposed to the harms and risks
so vividly described by the anti-natalists. Her progenitors
must have intended to have a child in part so that he or
she could enjoy a life whose goods would outweigh those
bads. If bringing about a good life and a loving relationship was not part of their reason for having a child, only a
side-constraint they respected, prospective parents would
be disingenuous to offer it as justification to their future
child. As I stated in an earlier paper:
All prospective parents should expect their children to
face significant hardshipsdeath, loss, frustration, and
pain. . . . They must be able to justify the decision to subject
their children to those hardships, and they can do so only
if part of their reason for having those children is to give
them lives good and rich enough to offset or outweigh those
hardships.24
In the final chapter, I will argue that this justification does

not require the prospective parents to select a future child
| 2 0 1
expected to have the most favorable balance of goods to

bads; indeed, that they may have a moral reason not to
engage in such selectivity. For now, I want to focus on the
justification owed any to child, regardless of whether or not
a choice among future children was possible or available.
A skeptic might question why the mere expectation
of a good life would not suffice for justification. Clearly
the expectation is necessary. Prospective parents who intended a good life for a future child but had no confidence
they could secure it would be acting irresponsibly in most
circumstances. (Indeed it is not clear they could even have
that intention if they believed that the odds of bringing
about what they sought were truly negligible.) But some
people may also find the expectation sufficient.25
I think that mere expectation is inadequate as a personal
justification to a child exposed to the serious harms and
risks of any life. Suppose the childs parents had her solely
for reasons that were not child-centered but not obviously
inconsistent with respect for the future child: for example,
to please impatient relatives or to avoid the stigma of childlessness. They also took precautions to ensure, to the extent
possible, that the child would have a good life, and were committed to establishing a loving parent-child relationship.
Imagine that the child, facing some of the predictable harms
or risks of her life, demanded to know why her parents, by
creating her, had exposed her to them. It would not appear
to be an adequate response to say that they merely had been
confident that she would have a good life overall. She would
not, in a robust sense, have been a wanted child, although
she would not have been an unwanted one either.
This is not an argument, but let me offer a partial analogy. There is a debate in criminal law about whether the
202 |
intent to protect oneself or another against a deadly threat

is necessary for the killing of the aggressor to be justified.
Proponents of an intent requirement argue against the
claim that knowledge suffices with cases like this: A woman
is sitting at a bar. She sees a man walk in with a gun and announce that he is going to kill another patron over a trivial
slight. She is indifferent to the fate of the prospective victim,
but sees an opportunity to try out her new gun with legal impunity. As the aggressor is about to shoot, she shoots him.
Although most people would agree that the law should not
make a finely grained inquiry into her reasons or intentions
for shooting a lethal aggressor, many believe that she would
lack a moral justification for doing so, because her reason
was not to prevent a clearly wrongful homicide. They argue
that to enjoy a justification for conduct that would otherwise be a serious criminal offense, the agent must act for
the right reason. And, although this is not a point stressed
in the debate, she displayed profound disrespect for the man
she killed by treating his lethal aggression as a pretext to
use him for live target practice. Arguably, she displayed profound disrespect for the target of the aggression as well, by
treating his rescue as a mere pretext.
As I said, this is only a partial analogy. The good of the
future child need not be used as a pretext by prospective parents who do not have it as among their reasons, in contrast to
the good of defending against lethal aggression. Moreover, the
actual reasons for bearing her are hardly as disrespectfulif
they are disrespectful at allas the actual reason for killing the aggressor. But the shared claim is that in cases where
the consequences of an action are as weighty as starting or
ending life, the agent must act for the right reason.
| 2 0 3
The analogy might be made closer by taking on the

moderate anti-natalist claim that the harms and risks of
existence make procreation presumptively wrongful, like
homicide. The shared claim in the cases of procreation and
defensive homicide would be that the act was only justifiable if done for the right reasons. This is not the condition
Shiffrin herself imposes for overcoming the presumption
of wrongful procreation, though it is consistent with that
conditionthat the prospective parents take on the duties
of nurturing the vulnerable child they create.
But we may be reluctant (I certainly am) to regard procreation as presumptively wrongful. Moreover, a requirement that prospective parents must intend the benefits
which they expect to outweigh the harms in the childs life
is still unresponsive to Shiffrins claim that such benefits,
however great, cannot justify the imposition of serious
harms without consent.
Another basis for requiring child-centered procreative
reasons may be found in a comparison to the formation
of other intimate relationships. As I argued earlier, people
ISO significant others can have the good of those unknown
others as one of the reasons for their search. A personals
ad that mentioned only the benefits the ad-placer sought
to gain would be unlikely to elicit favorable responses.
More important, it would reflect a morally deficient view
of the intimate relationship sought. Of course we accept
that many loving relationships are initiated with selfish
intentions. Many loving romantic relationships began as
singly or mutually exploitative ones; many deep friendships began opportunistically. Why set the bar higher for a
parent-child relationship?
204 |
Two points, neither conclusive, are worth noting. First,

even if we accept that morally attractive relationships can
evolve from unattractive ones, we do not regard the exploitative genesis of a relationship as morally innocent. Let
me illustrate with an extreme case, loosely adapted from
David DeGrazia26:
B initiated his romantic relationship with A to win a bet
with a friendthat he could score with the first young
woman who came around the corner on which they were
standing in the next five minutes. At the point of forming
his intention, A had no idea whom his target would be, or
even if there would be one. I think B would be wronged by
his reasons for acting, even if he planned and executed the
gentlest seduction and soon proceeded to fall deeply in love
with her.
Of course, A exists at the time B forms his intentions, so

shes around to be disrespected. Their future child, ABey,
isnt around to be disrespected at the time that A and B
conceive her for disrespectful reasonssay, to ensure a
large inheritance. But I fail to see why that gives ABey a
lesser moral complaint after hes born. Indeed it appears
that he has a stronger complaint, since he owes his very
existence to intentions that treated him as a mere means,
while A owes only her relationship to disrespectful intentions. Relatedly the emergence of the intimate relationship is unilateral for AB, interactive for A and B. A could
have recognized B as a cad from the outset; she could even
have made a complementary bet with her own friend. At
the very least, I think A and B would owe ABey as profuse
an apology as B owes her for the disrespectful inception of
their relationship.
| 2 0 5
But even if this comparison is apt, it only makes a case

that it is wrong to procreate with intentions that are disrespectful toward a future child, not that it is disrespectful to
fail to have its own good as one of the reasons for bearing it.
The more difficult claim to defend is that it is disrespectful
to create a child only for reasons that are otherwise morally
innocent, but that do not concern the good of the future
child. I have suggested that having such reasons is the closest that prospective parents can come to treating a future
child as an end but not a mere means, so that the failure to
act on such reasons would treat the future child with disrespect. But the skeptic could respond that the positive requirement of treating individuals as ends cannot apply to
future individuals, at least those whose existence depends
on the procreative acts in question.27 Without a resolution
of that issue, or a general account of when intentions make
actions permissible, I can only suggest that one factor is
the magnitude of the harms that must be justified. In the
case of existence, those harms, as the anti-natalists powerfully remind us, are great indeed.
NOTE S
University Press, 2006); Seana Valentine Shiffrin, Wrongful Life, Procreative Responsibility, and the Significance of
Harm, Legal Theory 5, no. 2 (1999): 117148. Melinda Roberts
suggests that a possible person is harmed in not being made
actual, because he suffers a loss. But she denies that that loss
has any moral significance, so it cannot be a wrong. See The
Asymmetry: A Solution, Theoria 77, no. 4 (2011): 333367.
2. Jeff McMahan, Problems of Population Theory, Ethics 92,
no. 1 (1981): 96127.
206 |
3. Susanne Gibson, Reasons for Having Children: Ends,

Means and Family Values, Journal of Applied Philosophy 12,
no. 3 (1995): 231240, 238.
4. Christine Overall, Why Have Children? The Ethical Debate
(Cambridge, MA: MIT Press, 2012), 215.
5. Overall, Why Have Children?, 217.
6. Gilbert Meilaender, The Blessing of Children, The New Atlantis [pdf version Summer 2012, 9298, 94] accessed April
18, 2014, http://www.thenewatlantis.com/publications/
the-blessing-of-children.
7. Meilanender, The Blessing of Children.
8. Rachel Thompson and Christina Lee, Fertile Imaginations:
Young Mens Reproductive Attitudes and Preferences, Journal
of Reproductive and Infant Psychology 29, no. 1 (2011): 4355.
9. Fertile Imaginations, 50.
10. For example, David Benatar, Better Never to Have Been
(Oxford: Oxford University Press, 2006), 34.
11. Daniel Friedrich, A Duty to Adopt? Journal of Applied Philosophy 30, no. 1 (2013): 2539; Christina Lyn Rulli, The
Duty to Adopt, PhD diss., Yale University, 2012.
12. Chapter 10, Sec. iv, this volume.
13. Immanuel Kant, Grounding for the Metaphysics of Morals,
trans, by James W. Ellington. 3rd ed. (Cambridge, MA:
Hackett: [1785] 1993), 43. (reference courtesy of http://
en.wikipedia.org/wiki/Categorical_imperative,
accessed
October 10, 2014). I am grateful to Rivka Weinberg for
pressing me on this way of understanding child-centered
procreative reasons, and for the insights she and her coauthor offer on the subject in Paul Hurley and Rivka Weinberg, Whose Problem is Non-Identity? Journal of Moral
Philosophy (2014) 132.
14. Julian Savulescu and Guy Kahane, The Moral Obligation
to Create Children with the Best Chance of the Best Life.
Bioethics 23, no. 5 (2009): 274290.
15. Chapter 9, Sec. iv., this volume.
16. Chapter 10, Sec. iii, this volume.
17. McMahan, Asymmetries in the Morality of Causing People
to Exist, in Harming Future Persons, eds. Melinda Roberts and
David Wasserman (Netherlands: Springer, 2009): 4968, 52.
| 2 0 7
18. Jeff McMahan, Asymmetries, 4968.

19. Overall, Why Have Children? 63. Similarly, Mianna Lotz
argues that because prospective parents cannot act for the
good of a particular future child, the provisions they make
for future children cannot provide a reason for creating
them: we might regard inheritance of the family fortune
as compatible with the good of any child, even if we cant
thereby count it as a reason or bringing the child into existence. (Procreative Reasons-Relevance: On the Moral
Significance of Why We Have Children, Bioethics 23, no. 5
(2009): 291299. 294).
20. Melinda Roberts suggests (in correspondence) that a biological or adoptive parents reason for action is child-based
only given the supposition that the child does or will exist.
This is true in the sense that the parents must presuppose
the future childs existence to have or act on the reason.
But it is false if it implies that the characterization of their
reason as child-based depends retroactively on whether
they actually succeed in creating a child. Consider prospective parents who set up a trust fund for a future child. Their
reason may clearly be child-centeredto provide economic
security for a child who does not yet exist, even if doing
so imposes significant financial hardship on them at present. Their reason may only makes sense on the supposition
that the child does or will existor at least is likely to
exist. If, however, that supposition turns out to be false
if say, the parents are infertile and cant adopt, or die in a
car accidenttheir reason simply loses its point; it does
not become, ex post, parent-centered rather than child-
centered because there is or will be no child.
21. Indeed, it is only in more impersonal forms of beneficence
that the connection of donor and recipient matter less. If
I give money through a reputable organization to a designated child or village, I do not mind permitting my donation to be directed elsewhere because the donations of
others have already met the need of that individual or place.
22. Adrienne Asch, Reproductive Technology and Disability,
in Reproductive Laws for the 1990s, eds. S. Cohen and N. Taub
(Clifton, NJ: Humana Press, 1990), 69124.
208 |
23. David Velleman in effect takes this position, holding that

love in a full-bodied sense requires the acquaintance-based
knowledge that comes with seeing or touching. Persons
in Prospect, Philosophy and Public Affairs 36, no. 3 (2008):
221288, 269.
24. David Wasserman, The Nonidentity Problem, Disability,
and the Role Morality of Prospective Parents, Ethics 116,
no. 1 (October 2005): 132152, 135136.
25. Prospective parents, especially first-time ones, often worry
about whether they will be able to love the child they are
creating, and whether they will be able to form a truly intimate parent-child relationship. Should we regard such
parents as acting impermissibly when they go ahead despite
their doubts?
26. David DeGrazia, Procreative Responsibility in Light of
What Parents Owe Their Children, in Oxford Handbook of
Reproductive Ethics, ed. Leslie Francis (New York: Oxford
University Press, forthcoming).
27. Sam Kerstein, whom I thank for a discussion of this issue,
expressed skepticism about the applicability of the Categorical Imperative to future people.
Impersonal Constraints
onProcreation
BENATAR ACCEPTS THE ASYMMETRY described at the start
of the last chapter; the view that there is a strong moral

reason not to have a child expected to lead a miserable life,
but no moral reason to have a child expected to lead a wonderful life.1 In fact, he argues that this asymmetry is best
explained by his own, about the harm of coming into existence, fortified by his assessment of the magnitude of that
harm.2 Defenders of procreation who reject Benatars arguments need to provide standards for when it is morally permissible to create a child, standards that do not imply that
it is ever morally required to do so. There have been two
general approaches to setting such standards. One is based
on the claim that since we cannot have personal duties to
merely possible people, impersonal standards must play a
significant role in selecting which of them will become
actual. Our choices should be guided, in other words, by
210 |
the good or value they realize, or contribute to the world.

In having children, we should seek those who will be best
off or most altruistic, or who will increase or promote the
good in other ways.
The second approach holds that those intending to
bear and raise children are constrained by the rights to the
actual children they may bear, or by the duties specific to
their roles as prospective parents. I will refer to these respectively, as birthright and role-based views. The duties
they posit do not exclude all considerations about the
broader impact of their choices, but for most such views,
those considerations are secondary. Moreover, these duties
may conflict with impersonal duties claimed to govern
procreation.
The two approaches are, somewhat confusingly, combined in arguments that the role-based duty of prospective parents is to select the child with the highest expected
welfare, as the closest approximation to parental concern
they can now achieve. I will review these maximizing arguments in their impersonal form in this chapter, and in
their role-based form in the next one.
Perhaps the clearest way to distinguish the two approaches is that impersonal standards, with one notable
exception to be discussed below, are comparative: one
should choose the child expected to be better off, less vulnerable or limited, more beneficent, etc. This comparative
feature sometimes calls for maximization, requiring the
selection of the best, least, or most. But even then the required maximizing is constrained by the burdens on the
agents and third parties. Accounts based on birthrights
or role-based duties are, with the exception of those that
treat the role in maximizing terms, non-comparative. They
I mpersonal C onstraints on P rocreation
| 2 1 1
require the individual or couple to have a child only if they

expect it to meet the proposed standard, not to select the
child they expect to most exceed that standard. On these
accounts, prospective parents owe it to a future child not
to create it unless they expect it to have a sufficiently good
life. If they violate that duty, they wrong the actual child
they create. But they have no duty, to the future child or
anyone else, to instead create a different child expected to
have an even better life.
The two approaches both treat prospective parents or
guardians as the appropriate agents for selecting future
children. It is they who have the duty to satisfy the applicable standard, however impersonal or demanding. As far as
I know, no one defending an impersonal standard claims
that the decision should be made by a more impersonal
agent like the state. I will argue, however, that recognizing
prospective parents as the agents for selection raises a serious problem for impersonal accounts.
The impersonal approach regards the domain in which
special moral rules apply as one of same-number choices.
That term comes from Derek Parfit, who defines such a
choice as one that will ultimately result in the same number
of people coming into being.3 Birthright and role-based accounts focus on the choices of couples and individuals who
have already decided to bear a child, about which future
child to select or whether to select at all. Those choices
often appear to be same-number ones, but they need not
be. Prospective parents who seek a boy for family balance
may try to have an additional child if they do not succeed
that time; fertile couples who opt for IVF and PGD over
standard procreation may greatly increase the probability
of having more than one child.
212 |
The impersonal approach faces a threshold problem in

defining its domain. As McMahan argues, its not clear that
we can regard any choice as same number.4 The problem is
not just that we can never be sure that a particular choice
among embryos, or time for conception, will not affect the
population of the universe. It is that we can be fairly sure
that it will, since even slight differences among individuals
are very likely to ramify over time in population-affecting
ways. So if Parfits definition were strictly applied, a samenumber choice would be a rare occurrence, and one that
could never be ascertained. There are other ways of defining
same number (e.g., in terms of the proximate outcomes
or the parents intentions). But these also have problems,
since the former would exclude unexpected twinning from
the domain of impersonal criteria, while the latter would
face the uncertainties arising from indeterminate or conflicting parental intentions.
Even if same-number choices can be clearly defined,
an impersonal approach must explain why they provide a
special occasion for a beneficence that is not generally required. If, for example, we are required to seek the happiest
child when we are having a child, why should we not be required to have some, or indefinitely many, children? In contrast, those who hold that the prospective parents merely
have a duty not to create children whom they expect to fall
below some threshold of well-being are not committed to
having any children at all.
A related challenge for impersonal approaches lies in explaining why the comparative criterion applies only to prospective parents. Why does it not apply to choices involving
the same number of mammals, or living beings, or choices
| 2 1 3
between creating a human or some other living being? As McMahan observes, it is unclear why it does not permit a couple
to select a child with a life expected to be just better than
barely worth living over one expected to flourish, but does
permit that couple to breed goldfish rather than bear children, when any goldfish will have a life with less well-being
than a child with a minimally adequate life.5 This question
does not arise on birthright and role-based accounts. Those
who chose to bear children have duties specific to that role,
but the decision to assume that role, rather than (or along
with) that of goldfish breeder, does not need to be justified
in terms of the comparative well-being of different species.
These problems with the impersonal approach are
best illustrated by looking at specific accounts. The four I
will consider are the Non-Person Affecting Principle (N),
which asserts an impersonal pro tanto duty not to have a
child expected to experience serious suffering or limitation6; Procreative Beneficence (PB), which claims a strong
moral reason to select the best-off child7; General Procreative Beneficence (GPB), which claims a strong moral
reason to select the child expected to contribute most
to total welfare8; and Procreative Altruism (PA), which
claims a strong moral reason to select the child most likely
to do good for others.9 Of the four accounts, N is the most
modest in requiring only the prevention of serious harms,
but not the minimization of lesser harm. PB and N are
more modest than PA and GPB, limiting themselves to the
welfare of the future child, not the world at large. Greater
modesty has a price, however. It requires some justification of the narrower scope of the impersonal criterion;
some explanation of why prospective parents should be
214 |
concerned with one aspect of impersonal goodness but

not others.
All four proposals claim that prospective parents act
wrongly in failing to adopt their proposed selection criteria; none claim that this failure wrongs the child selected. I will argue that none provide a plausible basis for
restricting beneficence to the context of child-bearing,
and that that context is particularly inhospitable to impersonal considerations. I will then argue, drawing on
McMahan and Christopher Belshaw,10 that the judgments
relied upon to support PB, PA, and GPB support or imply
a duty to have more children as well as better ones. This is
a problem that birthright and role-based accounts do not
confront.
B U C H A N A N E T A L . S N O N - P E R S O N AFFECTING PRINCIPLE (N)

N is worth reviewing in some detail, because it attempts to
resolve some of the problems confronting a context-specific
impersonal morality, and because its ambiguities, largely
acknowledged by its authors, suggest the challenges that
confront any such principle. N is presented as a hybrid: a
role-specific but impersonal duty of harm-prevention, applicable only in same-number choices but analogous to the
duty (M) caretakers have to prevent harm to actual dependents. It has the obvious advantage of avoiding the oppressive demands of a maximizing morality: individuals can
always produce more happiness, but N requires only that
they prevent a high degree of suffering. Unlike the other
impersonal principles, it is non-comparative.
| 2 1 5
Here, then, is Buchanan et al.s non-person affecting

principle N:
Individuals are morally required not to let any child or
other dependent person for whose welfare they are responsible experience serious suffering or limited opportunity or
serious loss of happiness or good, if they can act so that,
without affecting the number of persons who will exist and
without imposing substantial burdens or costs or loss of
benefits on themselves or others, no child or other dependent person for whose welfare they are responsible will experience serious suffering or limited opportunity or serious
loss of happiness or good.11
N faces the same limitation as any same-number principle:

It does not cover harms that are caused by acts that affect
the number of people who will exist. A decision to create
and implant more embryos, at great risk to each, would not
be covered. A problem share by other impersonal principles but well-illustrated by N is the mismatch between the
holders of the duty it imposes and the impersonal content
of the duty. Why should the impersonal duty to prevent
suffering or lost opportunity be imposed on present and
future caregivers? A consequentialist might regard caregivers as having such a duty because they were best positioned to minimize the suffering or limitation of those in
their care. But this is not the basis most of usincluding
Buchanan et al.accept for imposing special duties on
caregiverstheir responsibility does not rest, at least primarily, on the practical value of the arrangement, but on
their special relationship with the present and future objects of their care. Indeed, the special duties of caregivers
are usually regarded as quintessentially personal, often
216 |
exempt from, rather than embodying, impersonal imperatives. Yet under N, individuals in these roles have a special
impersonal duty to make the world a less bad place.12
Buchanan et al. do not explain why N should have
roughly the same content as the person-affecting duty (M)
they impose on parents to prevent harm to their actual
children. N is drafted to cover both the failure to prevent
harms to existing children and the failure to prevent them
by substitutionby selective abortion or implantation.
But this parity produces some anomalous results. Most of
us would consider parents failure or refusal to prevent their
infants painless loss of a toe by easy safeguards as culpable,
if relatively minor, neglect. In contrast, I suspect that many
who believe that parents have a duty to select against serious impairments would deny they had a duty, even a pro
tanto one, to select against minor ones like a missing toe.13
The lack of parity in the duty to prevent minor impairments
may suggest that N is simply narrower and weaker than M.
But it may also suggest that there is something fundamentally misguided in treating the failure to prevent harm by
substitution as the moral equivalent of failing to prevent it
in existing children.
A similar absence of parity suggests that Ns burden
exemption may be too permissive in some contexts. We
do expect the caregivers of existing children to incur
substantial burdens or costs to avoid or alleviate harm
to their charges. We may regard the parents in Lorenzos
Oil, who devoted their lives to finding an elusive cure for
their sons fatal illness, as having displayed supererogatory
devotion, but we do expect a great deal. We would regard
middle-class parents as neglectful if they refused to travel
across the country to obtain the only therapy available to
| 2 1 7
treat their childs serious but non-fatal illness that would

leave it with a decent life. In contrast, few of those who
support selection against serious disability would regard
middle-class parents as neglectful if they failed to undertake the same travel and incur the same costs to obtain
IVF and PGD to avoid bearing a child that would have that
disease, even if it could not be treated postnatally.
In conclusion, it is mysterious that prospective parents
should face even a modest impersonal duty, and one that
only applies in same-number cases. Not only is it is a duty
that prospective parents could circumvent by implanting
different numbers of embryos, it is a duty that specifically
applies to roles we usually regard as less subject to impersonal imperatives.
P R O C R E AT I V E B E N E F I C I E N C E ( P B )
As most recently formulated by Julian Savulescu and Guy
Kahane, the principle of Procreative Beneficence holds
that:
If couples (or single reproducers) have decided to have a
child, and selection is possible, then they have a significant
moral reason to choose the child, of the possible children
they could have, whose life can be expected, in light of the
relevant available evidence, to go best or at least not worse
than that the others.14
Several objections to this controversial proposal have

been raised that I will not consider here. In particular, it
has been argued that PB, even qualified as above, is too
demanding on prospective parents, and that there is no
218 |
generally accepted way to assess well-being, let alone the

well-being of a future child.15 Although I sympathize with
these objections, my concern is with the success of PB in
limiting maximization to same-number choices.
Like N, PB is circumscribed to apply only to the choice
of a single child, if selection is possible. Kahane and Savulescu also note that although PB is a maximizing principle, it may often clash with total act consequentialism,16
which would dictate the choice of a less well-off child if that
would increase overall welfare or good. They present PB instead as simply the application of a general constraint on
practical reason . . . Roughly, we have reason to choose what
is good, and we have more reason to prefer what is better. If
A and B are identical in all regards save one, and A is superior in that regard to B, we have a reason to choose A.17 But
this can hardly be a general constraint on practical reason.
If A and B were twins competing for scarce lifesaving treatment, and the only regard in which they were not identical
was that A was more productive than B, it would hardly
be practically irrational to flip a coin instead of selecting
A. Obviously embryos lack the moral rights of people; my
point is only that the claim of practical rationality is untenable. The fact that embryos A and B would become different
people matters greatly if we believe that the separateness
of future persons limits the relevance of impersonal comparisons. The fact that As life would be better for A than
Bs life for B has limited moral significance if A and B are
numerically different individuals.
If PB cannot be regarded as a constraint on practical rationality, why should it provide even a strong moral
reason for selection? Savulescu and Kahane suggest, like
Buchanan et al., that their principle can be seen as based
| 2 1 9
on the role-specific concerns of prospective parents: To

the extent that parents have reasons to care about the expected well-being of their future children, these reasons
can be seen as extensions of parents special relations to
their children, not as the external demands of an impartial
morality.18 But this suggestion fares no better than the
argument from practical rationality. Prospective parents
have special reason to care about the well-being of their
future child, whoever it will be. But they can also recognize
that possible children A and B will have different lives. If B
will live a good life, it is not clear why concern for his welfare gives them a reason to select A, who will lead an even
better one. (Some arguments that it does not are discussed
in the next chapter.) Given the difficulties faced by both
the practical-rationality and parental-role rationales, it is
not clear how PB can justify its limitation to same-number
cases or to procreative decisions.
G E N E R A L I Z E D P R O C R E AT I V E
BE NE FICE NCE (G PB) A ND
P R O C R E AT I V E A LT R U I S M ( PA )
Other philosophers have questioned the limits and selection
criteria that PB imposes on procreative choice. Rebecca Bennett has argued that it fails to take the non-identity problem
seriouslythe problem in the moral evaluation of procreative acts arising from the fact that they often produce different individuals. Bennett suggests that PB just appeals to
the intuition that it is wrong to choose a worse outcome, an
intuition that the fact of non-identity challenges.19 There
is also criticism from the other directionthat PB has an
220 |
unmotivated concern with the best-off child, as opposed to

the best outcome. Given that neither the appeal to practical
rationality nor parental role can justify PB, this argument is
complementary to the one I raised abovethat if prospective
parents should procreate beneficently, there is no apparent
reason for them to limit their beneficence to the well-being
of the future child. Jacob Elster argues on this ground for a
position that Savulescu and Kahane explicitly rejectthat
prospective parents should choose the child likely to make
the world best off, not the child likely to have the best-off life.
Elsters alternative, however, raises the question arises
of why prospective parents should limit their beneficence to
procreation, if other activities could produce even greater
good20 a point made by consequentialist critics of procreation, as noted in chapter 6. Perhaps GPB could serve
as a reasonable compromise for generally consequentialist
agents who regarded childbearing as a central project. It
would, however, be just one of many such compromise positions, including PB itself, and there does not seem to be a
good reason to accept the particular line that either draws.
Tom Douglas and Katrien Devolder offer a friendlier
amendment to PB. They suggest that parents have moral
reason to select for the most altruistic as well as best-off
child.21 Clearly, these desiderata can conflict, except on an
implausible theory of well-being or in the particular circumstances where individual well-being is tied to an altruistic
project. Douglas and Devolder merely regard the expected
altruism and well-being of the future child as two of the factors that have moral weight in procreative decisions. They
share with Kahane and Savulescu the view that parents
should, all else equal, choose children highest on all relevant factors over those lower.
| 2 2 1
C A N I M P E R S O N A L C O M PA R AT I V E
ACCOUNTS PR E SE RV E TH E
ASYMMETRY?
All three comparative impersonal principles may be in tension with the Asymmetry mentioned earlier: that while we
have a strong moral reason not to create a miserable child,
we have no moral reason to create a happy one. (This tension
does not appear to arise for N, which is non-comparative,
in that it does not require the selection of a happier or less
unhappy child among those not expected to face serious
suffering . . . .) The conviction that none of us has a duty to
create a child, no matter how well off it is expected to be, appears to conflict with the claim that if we are choosing a child
we should choose the better off. The problem is not just that
restricting benefit-maximization or harm-minimization to
procreative choices seems ad hoc or perverse; it is that such a
restriction may imply a duty to have a well-off child.
Belshaw argues for this incompatibility by comparing
two propositions:
(A) Given a certain number of lives, then (other things being
equal), their quality should be as high as possible and
(B) Given a certain high quality of lives, then (other things
being equal), their numbers should be as high as possible.
(A) is the demand of procreative beneficence; (B) is the

demand that the Asymmetry explicitly rejects. Belshaw
argues that although the former doesnt straightforwardly
imply the latter . . . they are intimately connected.22
For when its made clear that (A) is held to be true even when
different lives are involved, such that no one is made better
222 |
off by selecting the higher-quality lives, then it seems to me

that (A)s only hope of defense goes via a commitment to
something like total utilitarianism, which has, of course,
implications for the acceptance of (B).23
McMahan goes further, arguing that the choice of better

implies, on what he regards as plausible assumptions, the
choice of more24:
1) (General Assumption): Creating A, who will live to
eighty, is better than creating B, who will live to
sixty (if these are the only choices).
2) (General Assumption): Creating A, who will live to
eighty, is neither better nor worse than creating no
one.
3) Since the actual consequences of creating A are
the same in 1 and 2, they are morally equivalent.
4) Given 3, then if creating B is worse than creating A,
it must be worse than creating no one.
5) But this is absurdcreating a person who will live
sixty good years is definitely not worse than creating
no one. (If you think sixty is a sadly truncated lifespan these days, substitute eighty and one-hundred
as the two lifespans.)
6) There are two ways of avoiding this unacceptable
conclusion:
i/Deny 1, that it is better to create A than B (or
worse to create B than A);
ii/Deny 2, that it is no better to create A than no
one.
McMahan finds ii more plausible, since he is willing to
accept that the noncomparative good (good that doesnt
| 2 2 3
involve making things better for someone) of bringing a

happy person into existence has some moral weight, even if
less weight than comparative good (good that is better for
someone); in support of this view, he adduces the widely
accepted imperative to avoid human extinction. But if we
accept ii, then we have at least a weak moral reason to create
new people.25 And this, as McMahan goes on to show, has
some very counterintuitive implications in choices between
creating and saving lives (e.g., it will sometimes be morally
required to create new ones rather than save existing ones).26
I find i more plausible for prospective parents, in part
because of the implications of ii, in part because I think
that prospective parents have no duty or moral reason to
select among future children whose lives are all expected
to be acceptable. They have no moral reason to pick a child
expected to live to eighty over one expected to live to sixty,
as long as they regard the latter lifespan as acceptable
(except for reasons concerning the well-being of third parties who may be affected by their choices).
Another way to challenge McMahans conclusion is to
reject 3. The actual consequences are not all that matter
in bringing new people into being. Although a couple or
individual deciding whether to start a family could have
a perfectly good reason for not wanting a child at all, its
hard to think of a good reason why they would want a child
who would live to sixty rather than eightyeven if with sufficient effort we might fill in details to make that preference reasonable. But it would be reasonable, as I argue in
the final chapter, for prospective parents to flip a coin between embryos A and B in an IVF array, recognizing that
they are, or would become, distinct individuals who would
both have good lives.27
224 |
I believe that an argument similar to McMahans can

be made against GPB: if it is worse to select the child that
would produce less good overall when either would produce
a substantial amount, but not worse to have no child than
one who would produce the most good, then it must be
worse to select a child who that would produce less but still
substantial good than to create no life at all. PA, despite its
pluralistic character, faces a similar objection, because it
regards it as better to select a child with the most of some
weighted combination of altruism, well-being, and other
desiderata than one within less of that combination, while
not worse to have no child than one with the best combination. That implies that it would be worse to choose a child
with some lesser combination than not to have a child at
all. Any maximizing principle for child selection will confront a similar problem.
The difficulties in defining same-number choices and
in restricting maximizing duties or reasons to such choices
are reflected in the ongoing discussion of what Derek Parfit
has termed Theory X: the theory that will best integrate
the same-number intuition stated in McMahans proposition 1 above with other deeply-held moral principles.
Parfit generalizes that intuition as the Same Number Quality Claim or Q:
If in either of two possible outcomes the same number of
people would ever live, it would be worse if those who live
are worse off, or have a lower quality of life, than those who
would have lived.28
Parfit suggests the need for a Theory X that will tell us how
Q should be justified, or more fully explained and would
as well apply to different-number choices.29 He concedes in
| 2 2 5
Reasons and Persons that he failed to find such a theory.30

In the thirty years since the publication of that book, no
generally accepted theory has emerged.
NOTE S
1. Jeff McMahan, Problems of Population Theory, Ethics 92,
no. 1 (1981): 96127.
2. David Benatar, Better Never to Have Been (Oxford: Clarendon), chs. 2,3.
1984), 356.
4. Jeff McMahan, Preventing the Existence of People with
Disabilities in David Wasserman, in Quality of Life and
Human Difference: Genetic Testing, Health Care, and Disability,
eds. Robert Wachbroit and Jerome Bickenbach (New York:
Cambridge University Press, 2005), 146.
5. Jeff McMahan, Wrongful Life: Paradoxes in the Morality
of Causing People to Exist. in Rational Commitment and
Social Justice: Essays for Gregory Kavka, eds. Jules Coleman
and Christopher Morris (New York: Cambridge University
Press, 1998), 208247.
6. Allen Buchanan et al., From Chance to Choice: Genetics and Justice (Cambridge, MA: Cambridge University Press, 2002), 247.
to Create Children with the Best Chance of the Best Life,
Bioethics 23, no. 5 (2009): 274290.
8. Jakob Elster, Procreative BeneficenceCui Bono? Bioethics
25, no. 9 (2011): 482488.
9. Thomas Douglas and Katrien Devolder, Procreative Altruism: Beyond Individualism in Reproductive Selection,
Journal of Medicine and Philosophy 38, no. 4 (2013): 400419.
10. Jeff McMahan, Causing People to Exist and Saving Peoples
Lives, The Journal of Ethics 17, no.12 (2013): 535; Christopher Belshaw, More Lives, Better Lives, Ethical Theory and
Moral Practice 6, no. 2 (2003): 127141.
226 |
11. Buchanan et al. From Chance to Choice, 249.

12. David Wasserman, The Nonidentity Problem, Disability,
and the Role Morality of Prospective Parents, Ethics 116,
no. 1 (2005): 132152, 141142.
13. Indeed, such an exclusion might even be regarded as violating what we could call, inverting Frances Kamm (The
Choice Between People, Commonsense Morality, and Doctors. Bioethics 1, no. 3 (1987): 255271), the principle of
irrelevant disutilitiesa principle condemning the use
of relatively trivial considerations to make decisions with
grave consequences.
14. Savulescu and Kahane, The Moral Obligation, 274.
15. Jonathan Glover, Choosing Children: Genes, Disability, and
Design (Oxford: Oxford University Press, 2006), 51; Michael
Parker, The Best Possible Child, Journal of Medical Ethics
33 (2007): 279283.
16. Savulescu and Kahane, The Obligation, 283.
17. Obligation, 281282.
18. Obligation, 283, (citing Wasserman, 2005, n. 14).
19. Rebecca Bennett, When Intuition is Not Enough. Why
the Principle of Procreative Beneficence Must Work Much
Harder to Justify Its Eugenic Vision, Bioethics 28, no. 9
(2013), 447455.
20. Elster, Procreative Beneficence: Cui Bono?
21. Douglas and Devolder, Procreative Altruism.
22. Belshaw, More Lives, 132.
23. Belshaw, More Lives.
24. McMahan, Causing People to Exist, 2526.
25. McMahan, Causing, 26.
26. Causing, 3134.
27. Savulescu and Kahane might also try to reject 3, for a very
different reason. They might argue that 1 involves an interpersonal comparison, while 2 does not. 1 claims that it
is better to create more person-affecting value than less. 2
claims that it is neither better or worse to create personaffecting value than not to, but this does not mean or imply
that the outcomes are equal; they may be incommensurable.
And if they are incommensurable, then we cannot conclude
| 2 2 7
that it is worse to create a sixty-year-old than to create no

one. I do not feel equipped to evaluate this response, only to
claim that it places a great deal of weight on the notion of
incommensurability.
28. Parfit, Reasons and Persons, 360.
29. Parfit, Reasons, 361.
30. Parfit, Reasons, 451.
10
Alternatives to Impersonal
Approaches: Birthrights
andRole-Based Duties
THE PREVIOUS CHAPTER CONSIDERED the challenges
faced by impersonal accounts in limiting the imperative to

maximize happiness or prevent serious suffering to same
number choices. Drawing on arguments by McMahan,
I concluded that it was very difficult to limit impersonal
considerations to such choices. First, it is unclear whether
there are such choices; second, a duty to choose better
rather than worse lives appears to imply a duty to choose
more rather than fewera prescription most philosophers
would find hard to swallow.
In this chapter, I want to consider moral constraints on
procreative choice that are based on the rights of children
against their procreators or on the role-based duties of prospective parents. Although these rights and duties apply for
A LT ER N AT I V E S T O I M PER S ON A L A PPROAC H E S
| 2 2 9
the most part to choices that will yield the same number
of children, that feature has no special moral relevance.
The decision to have a child is itself a different-number
choicea choice to have one or more children rather than
none. Prospective parents may then make what appears to
be a same-number choice by deciding when to have a child,
by deciding whether to terminate a pregnancy and start
over, by choosing embryo(s) to implant from an IVF array,
or by selecting an embryo or gamete provider for collaborative reproduction. But these are not strictly same-number
choices (if any choices are), since they may not be willing or
able to have a child after a long delay or an abortion, or may
have multiple children after IVF.
I will begin by briefly reviewing several claims that children have against their parents; rights to be created with the
expectation that their lives will meet a minimum standard
of well-being. These rights belong to actual children, and
give them a complaint against parents who did not expect
their lives to meet that minimum standard. Although proponents of these birthrights sometimes invoke the role
of prospective parents, their standards appear to apply to
any agent bringing a new human life into being. All would
condemn the creation of a child expected to have a life of
unremitting suffering devoid of pleasure or joy, regardless
of whether the progenitors were prospective parents or synthetic biologists. It is less clear how much the non-comparative birthrights really differ in the standards they impose
for responsible procreation.
I will suggest the most plausible standards are really
robust requirements for a life worth living. Although they
explicitly demand more than that, I will maintain that
their doubts about the permissibility of creating lives
230 |
barely worth living mainly reflect concerns about misfortunes and tribulations impossible to predict by prospective
parents. I will suggest that their additional requirements
for a minimally decent life are either insurance against
the risk that the childrens lives will not be worth living
or requirements that specific individualstheir parents
commit themselves to loving, nurturing, and protecting
them. While conceding that some degree of risk aversion
may be appropriate in procreative decisions, I claim that
the best understanding of procreative constraints is found
in the latter requirement, involving the role-based duties
of prospective parents. That is the subject of the rest of the
chapter.
M I N I M U M S TA N D A R D S A S B I R T H R I G H T S
Of the minimum standards that have been proposed, some
are comparative in the sense of requiring a life expected
to be as happy, or no less difficult, than the population
average. But all are non-comparative in another way: the
permissibility of having a child who meets the standard is
not affected by the possibility of instead having one who
exceeds it. Furthermore, these standards underwrite a personal complaint by the child created. It is not only wrong
for prospective parents to fail to meet the prescribed standard; it wrongs the child they create.
There appears to be widespread divergence about what
higher standard, if any, is appropriate. Some regard a life
barely worth living as acceptable, especially if it is the best
the prospective parents can expect to provide, given their
circumstances and endowments. Most, however, reject this
| 2 3 1
standard as implausibly permissible in allowing parents

to bear children whose lives are expected to be at the very
margins of acceptability.
This general view is sometimes expressed as the claim
that children have a right to a decent or minimally good life,
not one that is barely worth living. This standard may be
seen as a birthright for the child or as a duty for the prospective parents. Perhaps the most demanding standard is
Michael Tooleys, which holds that children have a right to
a life expected to be of at least average well-being.1 Under
this standard, about half of all births in any comparison
group would violate the childs right, except perhaps in
Lake Wobegon, where all children are (and are presumably
expected to be) above average. Several philosophers have
proposed less demanding standards. These include Elizabeth Harman2 and David Benatar3 (as an alternative for
skeptics of his view that procreation is always wrongful),
who both propose that it would be wrong to create children
expected to face usually severe hardships. (As discussed
in the last chapter, Buchanan et al. proposed such a standard, although they frame it in impersonal terms.) If such
a right were understood to trump even a highly favorable
balance of goods to bads, it would condemn the birth of
Stevie Wonderif blindness were regarded as a hardship
and if his parents had reason to expect both his blindness and his long, rich career. Even more problematically,
it would permit many or most births in many developing
countries, if it permitted any, only on the grounds that the
hardships imposed by their difficult or hostile environments were not unusual.
Two more flexible and appealing standards are proposed by Bonnie Steinbock and David DeGrazia. Steinbock
232 |
defends a decent minimum standard, which she spells

out as follows:
A decent minimum is reached only if a life holds the reasonable promise of containing the things that make human
lives good: an ability to experience pleasure, to learn, to
have relationships with others. If someones life will be
irreparably and irremediably bereft of these goods, we do
that person no favor by bringing him into existence; indeed,
knowingly and voluntarily to conceive a child a person
under such conditions is a harm and a wrong to the person.
. . . In addition, the ability to be a good enough parent is also
part of the decent minimum. I maintain that it is wrong,
irresponsible procreation to have a child if one knows that
one lacks either the ability to love the child or the capacity
to care properly for him or her.4
In a similar spirit, DeGrazia presents a tentative list of what

he calls essential interests or basic needs of all children.
These include minimal material and medical provisions, as
well as freedom from slavery, other forms of wrongful coercion, and physical abuse; education and adequate stimulation; opportunities to play and experience enjoyment;
the opportunity to develop interests and gradually find
their own path; . . . and the love, kindness, and attention
of at least one reasonably competent parent.5 He claims
this list approximately captures a norm that is far more
demanding than a worthwhile life but less perfectionist
than the very best an advantaged parent can provide.6 DeGrazia is willing to let parents off the hook for failing to
meet most of the individual items on his list if that failure
is due to circumstances beyond their control. The one nonnegotiable item is parental love, care, and commitment.7
| 2 3 3
Both these standards are fairly permissive. DeGrazia

believes that most of the worlds prospective parents could
satisfy his standard with sufficient commitment to their
children.8 Steinbock recognizes that her decent minimum
may be achieved, and expected to be achieved, even by a
child with a fairly serious impairment.9 Although I find
these standards attractive, I am wary of the effort to set a
threshold above a life expected to be judged worthwhile by
the person living it (or by some more objective standard).
Those who propose a higher threshold face two challenges:
justifying the specific threshold they set, and justifying
a higher threshold at all. I do not want to press the first
challenge, in part because the standard for a life that is
just worth living is itself quite vague, if not arbitrary in
the same way as a specific higher threshold. But I think the
second challenge is formidable: Why would the mere fact
that a child had only a worthwhile life give it a complaint
against loving and committed parents, if that was the best
possible life it could have and its parents had displayed no
indifference to its expected hardships in conceiving it? And
if such a child had a complaint, what about one who had
only a minimally decent life when her parents could have
had a child expected to lead a wonderful life?
I am inclined to regard Steinbock and DeGrazia as instead having offered plausible specifications of a worthwhile life, in terms of objective goods, rather than setting a
higher threshold. Prospective parents should not have children they expect to be unable to experience what DeGrazia
calls undeluded joy in and about their livesat least at
some points during those lives. But arguably an individual
who could not do so would not have a life worth living. And it
is unclear how one could have a wretchedas opposed to
234 |
very difficultlife that was worth living (as opposed to not

ending). Even if one could, it is not clear how that judgment
could be made before life had unfolded, let alone prenatally.
Indeed, recent work on well-being and the shape of a life10
argues that the value of a life for the person living it is often
largely determined by its unpredictable
vicissitudesa
wretched life redeemed by heroic acts or great achievements; a wonderful life undone by betrayal or disaster. But
such vicissitudes can hardly be predicted, unlike many of
the items on Steinbocks and DeGrazias lists.
I suspect that some of the intuitive appeal in setting
a margin above the minimum may reflect one of two reasonable but misplaced concerns: that a life without such a
margin runs too much risk of not being worth living; and
that it would be perverse for parents to prefer a life at just
over the threshold to one significantly above it. (I am not
claiming that the proponents of these standards rely on
these concerns in making their arguments.) As to the first,
it is difficult, as I have argued, to assess a life prospectively,
and it is not meaningful to assign numerical probabilities
to most outcomes. Although I have rejected the maximin
rule imposed by Hayry,11 some degree of risk aversion does
seem warranted in procreation. It is easier to prescribe risk
aversion, however, than to exercise it in this context, for
the reasons given in the last paragraph. The development
of increasingly comprehensive and sophisticated prenatal
genetic tests is unlikely to make more than a marginal
contribution to risk assessment. With the exception of a
small number of genetic mutations that virtually guarantee intense pain and early death, the risks posed by most
genetic variations will not even be quantifiable, in part
because the outcome depends on their complex and often
| 2 3 5
unpredictable interactions with a myriad of other genetic

and environmental factors.
As to the second concern, it would be perverse to seek a
child with a worse life than others one might have. But prospective parents can act in ways they know will increase
the odds of having a child with poorer prospects for many
reasons besides such a preference. First, they may simply
refuse to choose among future children. The choice to risk
having a child with worse genetic prospects seems less perverse if prospective parents have the option of not choosing at allif, for example, they can refuse PGD and choose
randomly among their IVF embryos. Indeed, as I argue
below, a refusal to select may be the most appropriate posture for prospective parents, expressing a benign neutrality among possible future children with worthwhile lives.
Moreover, prospective parents rarely if ever act with
the intention of choosing a worse-off child. The refusal to
choose a child with better genetic prospects appears less
perverse if it is embedded in a decision about the timing
of childrento have them now, when the couple is poorer
but more energetic, or later, when they are wiser and more
prosperous, but less energetic and perhaps more likely to
conceive children with genetic challenges. In such a family-making context, prospective parents can expect to have
children with worse genetic prospects without intending
that result. They may be balancing the interests of other
family members against that expectation, and hope that
it is not fulfilled. Finally even selection for disability does
not reflect a preference for a worse over a better life. It is
based either on the belief, however reasonable, that a child
with that disability will have a better life with these parents than a child without it, or a preference for a child who
236 |
shares a parents disabilitya preference that may be selfish but not perverse.
I think there is an implicit recognition of the importance
of the reasons or attitudes of prospective parents in DeGrazias willingness to let parents off the hook for failing to
meet some of their childrens basic needs except for love and
commitment. Take one such need, freedom from slavery.
Consider slave parents, several of whose basic needs are unsatisfied, who chose to have children they know would be
similarly enslaved. DeGrazia would appear to regard their
choice as permissible if they expect to love and protect their
children to the best of their limited abilities, and if they
expect that their childrens lives would be at least tolerable.12 The parents would not commit a wrong excused by the
added burden of childlessness, but rather make a permissible choice under extremely harsh circumstances beyond
their control. In contrast, free prospective parents would act
reprehensibly if they contracted for a large fee to bear a child
to be consigned to a far more benign slavery (an example
adapted from Greg Kavka), a child expected to enjoy greater
well-being than the child of the enslaved parents. To explain
this difference, I will argue, it is necessary to look beyond
the expected well-being of the child to the parents reasons
for having a child, or one child rather than another.
A ROLE-BASE D DUTY TO MAXIM IZE

THE EXPECTED WELL -BEING
OF THE CHILD?
In chapter 8, I challenged the assertion made by several
philosophers that we cannot create a child for reasons that
| 2 3 7
concern the expected well-being of the future child. Here,

I will reject the claim made by several philosophers, including some of he same ones, that we can have reasons
for choosing one child over another that concern the expected well-being of the future child. Like the claim about
the decision to have a child, the claim about the decision
to have one rather than another rests on a mistaken inference from the undeniable fact that prospective parents
cannot relate to future children as particular individuals.
Two approaches with radically different implications
have been proposed about how the identity-affecting character of prenatal selection affects the duties of prospective
parents. The first relies on the impossibility of apprehending a future child as a particular individual to treat it as
a single, generic individual, or as a role that can be filled
by different individuals with different characteristics. The
duty of prospective parents toward their future child is
to increase its happiness or decrease its suffering in the
only way they canby selecting a happier or a less unhappy
child. I will argue that it is simply mistaken to assume that
because prospective parents cannot relate to the future
children they are choosing among as particular individuals, they must treat them as, or as if they were, the same
child, who can be made better or worse by selection. This
false assumption reflects a failure to adequately respect the
separateness of future persons.
The second approach, which I will defend, also recognizes the impossibility of apprehending a future child as a
particular individual. But it does not regard that limitation
as supporting any kind of selectivity. Because each future
child would be a distinct individual, there is no reason to
choose a happy or less unhappy one for the sake of the
238 |
child, since no actual child would be made any happier or

less unhappy by that choice. Rather, prospective parents
have no reason to choose a more over a less happy child,
as long as the latter is expected to have a life that would
be acceptable, subjectively or objectively. Indeed, favoring
a happier child may be in tension with duties arising from
their role in creating a certain kind of relationship and
association.
In making this last claim I will suggest that a plausible
source of constraints on selection by prospective parents
arises from the kind of relationship and association they
should be seeking to establish: a loving and respectful
relationship that lasts a lifetime, with many years of intense care and nurturing, in a family whose members are
expected to be highly partial and deeply devoted to each
other, and in which membership, once established, is virtually unconditional. These are, of course, only ideals; I
offer them not as standards for assessing the performance
of actual parents, but as a source of moral reasons that prospective parents have for refusing to choose among future
children.
As mentioned in the previous chapter, the first approach is suggested by some philosophers who defend
an impersonal approach to procreative choice. Thus, Buchanan et al. argue that parents and caregivers have the
same type of duty toward future as present childrento
prevent unnecessary suffering and limitation. For future
children such prevention requires selection. Savulescu and
Kahane justify their principle of procreative beneficence in
similar terms. They suggest that the special bonds of parenthood give prospective parents reason to seek the bestoff children as well as the best for their children.13 Several
| 2 3 9
philosophers have gone beyond such suggestions to argue

that prospective parents have role-based duties toward
their future children to select the best off or least badly off.
Casper Hare offers an approach to procreative choice
that attempts to transmute impersonal duties into duties
that, if not quite personal, are owed to a specific class of
people: the currently unknowable individuals who will be
affected for better or worse by those who have a special responsibility for them. This latter class includes prospective
parents. Hare claims that in certain contexts, we expect
a person to display a psychological attitude that involves
caring, not that the occupant of a certain role be as well
off as possible, but that a certain role be filled by someone
as well off as possible. These contexts have two features:
First, it is appropriate to expect the person in question to
be partial . . . toward a group picked out by a definite description. Second, because of the causal circumstances the
person finds herself in, that partial concern has no de re
expression.14 Prospective parents are expected to be partial toward their future children. But because they cannot
express this partiality toward particular individuals, they
must settle for expressing it toward whoever is picked out
by a definite descriptionour nth child.
As Ive argued elsewhere, the inability to show such partiality toward future children provides at most an explanation, not a justification, of the attempt to have the role of
ones next child filled with someone as well off as possible.15
Hare, I contend, confuses the claim that we should select
the best-off children with the claim that we should do as
much as we can to make our n
ow-unidentifiable future
child better off. Doing the latter often involves actions
that delay procreation and thereby alter the identity of the
240 |
children. But it does not consist in the deliberate selection

of some possible children over others. It is done with the
intention of doing more for our childrenwhoever they
may benot with the intention to have children who will
be better off irrespective of the quality of care and nurture
we provide for them.
Perhaps the most forceful defense of the view that
prospective parents must apprehend their future children
as occupants of a role, and select accordingly, comes from
David Velleman. Velleman gives a poignant description of
parents and their severely disabled child, both doomed
to love the child and his regrettable life.16 The child,
like the victim of a cruel seduction, cannot help loving a
personhimselfwho will bring him only sadness and
misfortune. The parents are also in a tragic bind, but it is
of their own making if they could have prevented the existence of a child with a severe disability. The parents cannot
invoke their present love for the child to justify their failure to prevent his existence, since they could not choose
to create him in particular, considered demonstratively, as
he would subsequently be loved.17
Vellemans last claim is hard to dispute; it echoes Robert
Adams observation that only God can love in prospect the
particular beings we love at present.18 But although parents cannot love a particular child before he is conceived,
they can certainly recognize that any future child has, so
to speak, but one life to live; that a child cannot live a life
without the unfortunate or regrettable aspects that are
necessary for his actual existence. Velleman may also be
correct that prospective parents should not doom their
future child to a love that will be unrequited by the goods
of lifesuch as rich sensory experience, deep friendships
| 2 4 1
and other commitments, satisfying personal achievements,

and extended periods free of physical pain and emotional
suffering. Such a life would not be, in a robust sense, worth
living; it would not be a life that could be joyously affirmed
despite its many adversities. But few severe disabilities
would condemn a child to such a life. In another passage,
however, Velleman appears to hold prospective parents to
a higher standard:
In creating human lives, we must take care that they afford
the best opportunity for personhood to flourish. We are obligated to give children the best start that we can give to
children, whichever children we have, so we are obligated
to have the have those children to whom we can give the
best start. A child to whom we give an initial lesser provision will have been wronged by our lack of due concern for
human life in creating himour lack of concern for human
life itself, albeit in his case.19
This passage is breathtaking in its ambiguity. Does Velleman think that prospective parents can be faulted for
failing to aim for the child who can most fully realize
the attributes of personhood; the child who will be most
rational, self-conscious, or empathic? Or does he merely
claim that parents should ensure that their children
meet the minimum standards for personhood, whatever
they are? Like Hare, he appears to elide distinct duties
in his claim that parents must give their children the
best start or provision. They may be obliged to give their
children the most material and psychological support
that they can give (consistent with their other duties
and their own needs) and that those particular children
can receive, but that is a far cry from choosing children
242 |
whose genetic endowment will give them the best start

and provision. It also seems permissible for prospective
parents to select against children for whom they reasonably doubt that they could adequately provide, whether
because those children are likely to have extraordinary
talents they could not nurture or severe, complex medical conditions for which they could not adequately care.
But unless the flourishing of personhood requires the
greatest excellence prospective parents can bring about
by nature and nurture, it is difficult to see how they can
be held to anything more than a strong commitment to
help their children achieve a good life.
Contra Hare and Velleman, the fact that prospective
parents cannot identify or love their future children (except,
possibly, if they select them from an IVF array) hardly requires them to express their anticipatory concern and partiality by choosing the best off ones. The fact that some
prospective parents believe that are acting for the welfare of
their child not only in delaying conception, but in selecting
among embryos or aborting and starting over, just goes to
show that theyno less than some philosophersare capable of overlooking obvious, morally relevant considerations.
A ROLE -B A SE D DU T Y AG A I NST
SELECTION?
The recognition of the distinct identities of future possible
children informs the second approach to prenatal selection
by prospective parents. In emphasizing the separateness of
future children, proponents of this approach reject not only
comparative standards like Savulescus and Buchanans, but
| 2 4 3
higher-than-minimal non-comparative standards. They require only the expectation that the child have a life that is,
or that she would consider, worthwhile.
One version of this approach emphasizes the role of
prospective parents and the moral character of the relationship and association they are, or should be, seeking
to create. The family, in its contemporary form, is in one
respect the most inclusive human association, encompassing members who may differ widely in many of their
most significant attributes. In contrast to other associations, family members cannot be excluded except for the
most egregious misconduct. Moreover, the children in a
family are expected to be loved not only unconditionally
but equally, even if actual parents rarely live up to either
ideal. Although unconditional love is a demanding ideal,
parents are expected to continue to love, and love equally,
children who become very difficult or severely impaired, or
who display unexpected musical or mathematical genius.
The virtue that would help satisfy these ideals has been
called acceptingness by Rosalind MacDougall.20
Asch and Wasserman propose that an ideal of unconditional welcome for prospective parents best reflects the
moral character of the parent-child relationship and the
family. That ideal enjoins prospective parents not to condition their willingness to bear and raise a child on the
expected presence or absence of virtually any trait. They
argue that this posture distinguishes prospective parents
from prospective friends and lovers, who should exercise
some measure of selectivity in forming intimate relationships and whose commitment to each other is more dependent on valued attributes than is the commitment of
parents to their children.21
244 |
Admissions criteria affect the moral and psychological

character of any association. Even within exclusive associations, with high standards and competitive admission,
there is a tension between those admission standards and
the equality expected among members, expressed in an
ideal of collegiality. Many a member of an academic search
committee has remarked, only half in jest, that he or she
would never have been admitted under current standards.
More poignantly, institutions that raise their standards
often relegate their older members to a kind of secondclass citizenship. This kind of tension may be well worth
bearing to maintain and enhance the excellence of academic institutions. But it would be toxic in a family.
Imagine a family formed during an era of rapidly improving genetic testing, in which the capacity to detect
predispositions to disease and disability grew ever more
comprehensive, discriminating, and accurate. The prospective parents avail themselves of the latest selection technology in having children spaced two or three years apart.
Each child, if he knew about the selection process, would
have the disturbing awareness that he might well have
been selected against on the latest roundthat he would
no longer pass muster. This would not, of course, preclude
the parents from displaying unconditional and equal love
for all their actual children, to the extent any parents can.
But it would lead to a profound moral, if not psychological,
unease.22
A different version of the second approach is offered
by Peter Herissone-Kelly. He argues that to assess the
prospects for future children, a prospective parent must
imaginatively inhabit their expected lives. Like an actual
parent, she should identify with each of their lives, an
| 2 4 5
identification that bars selection against any life the child

herself would find acceptable.23 Herissone-Kelly calls this
an incumbent model of procreative choice, since prospective parents are to take the perspective of each potential
child as if she were the actual one24 He subjects this approach to the Principle of Acceptable Outlook, holding
that the future life must be expected to be acceptable to
the individual living it.25 This proviso appears merely to express in subjective terms a life-worth-living requirement.
The internal perspective otherwise provides no moral basis
for choosing among future children.
Asch, Wasserman, and Herissone-Kelly argue that prospective parents have a distinct moral role, with constraints
and prerogatives different from those of citizens or public
officials. This position could have a number of theoretical
foundations. It could be based on a notion of role-moralities embedded in, or distinct from, ordinary morality: It is
widely recognized that actual parents have duties and prerogatives toward their present children that others do not
have to those children and that they do not have to other
children. Prospective parents arguably have similar, if attenuated duties and prerogatives, to have children in ways
consistent with the moral character of the families they
seek to form. Such an account, however, faces the daunting task of explaining how such a role-based morality is derived from, or relates to, the more general moral claims to
which the individuals in the role are subject. A role-based
morality might also rest on a qualified extension of widely
accepted notions of parental partiality to prospective parents. As we will see, however, such an extension may yield
a different set of duties and prerogatives than those put
forward by Asch, Wasserman, and Herissone-Kelly. Less
246 |
plausibly, the special duties and prerogatives of prospective parents could be justified on the grounds that they are
welfare-maximizing.
Hallvard Lillehammer extends parental partiality
from present to future children, arguing that prospective
parents are governed by a fundamentally different morality than the state. There are incommensurable impartial
and partial moralities, the former requiring the maximization of value for an indefinite number of actual or possible
objects; the latter, the maximization of value for a subset
of those objects. Lillehammer holds that under a partialist morality, prospective parents may choose children
whom impartial considerations would disfavor.26 They may
choose to have children who will be their own genetically,
even if they could have had healthier or smarter children
with third-party gametes. And they may choose a child
with genetically based affinities, like deafness, that they
expect to enrich their relationship with her, even if the affinities are disadvantageous, viewed impartially. Lillhammer denies that partiality can be justified by or grounded
in some deeper morality; as Niko Kolodny has held with
respect to the relationship between parents and existing
children, partiality is basic.27
Lillehammer shares the view that a state or society
may have moral reasons prospective parents do not for
promoting the creation of certain kinds of peoplebetter
off, longer-lived, or more talented.28 But his approach diverges from the other two in recognizing a partiality-based
prerogative to choose among possible children based on affinity and other reasons with little impersonal weight.29
For Herissone-Kelly, Asch, and Wasserman, the appropriate moral posture toward possible children is benign
| 2 4 7
neutrality. In contexts where a choice has already been

presented, that neutrality would be expressed by random
selection.30 (Such neutrality would also be consistent with,
but not required, by views like Steinbocks and DeGrazias,
which find it permissible to bear any child expected to have
a good-enough life.) On their views, selecting for deafness because the prospective parents were partial to deaf
childrenLillehammers examplewould be problematic
for the same reasons as selecting against deafness.
REASONS FOR SELECTION NOT

BASED ON THE CHILDS EXPECTED
WELL -BEING
I have argued that prospective parents do not (except in
the case of lives not worth living) have reasons to delay
child bearing based on the genetic constitution of the child
they will actually create, because they should recognize
that the children they will create in selecting or delaying
will be different children. On my account, the expected
well-being of a child is relevant for the permissibility of
creating it mainly in an indirect way: the poorer the childs
prospects, the harder it may be for prospective parents to
claim appropriate, respectful reasons for seeking to bring
it into being. But except in the case of lives expected to be
utterly awful, prospective parents may have respectful reasons for having a child they expect to be less happy than
another they might have had. They may, for example, have
family reasons for having a(nother) child sooner or professional reasons for having a(nother) child later. And they
may also have appropriate reasons for selection based on
248 |
their perceived duties and on the kind of relationship they

seek to have.
Prospective parents also may have two related, arguably child-centered reasons to favor some future children
over others. The first is that they may more adequately fulfill their duties as parents if they acquire children at some
stages of their lives rather than others. This may be a good
reason for Parfits fourteen-year-old girl to wait31not because her child will be better off if she waits, but because
she will be a much better mother to any child she has if she
waits than she would be to any child she would have if she
didnt. And this also may be a reason for a sixty-year old
man not to waitbecause he is unlikely to be there for his
child as long as he should if he conceives one at seventy.
A related reason for delay concerns the prospective
parents capacity to shield their future children from a
hostile environmentsocial, physical, or uterine. Because
parents have a duty to protect their children from harm,
prospective parents have an anticipatory duty to avoid, if
they can, environments in which the harm will be too great
for them to do so effectively. They should perhaps wait to
have a child until they can flee a country where their ethnic
group is brutally persecuted or lives in a region too arid to
feed its population, or until they can modify a uterine environment that will seriously damage any fetus.
This also suggests a moral difference between two
kinds of reasons for selecting against a child with mild
cognitive impairments. Even if such a child is unlikely to
lead as rich or good a life as a child with normal cognitive capacities (itself a debatable claim), it is likely to live
well in many developed countries. Selecting against such
a child to prevent the harm of its existence or impairment
| 2 4 9
is, I have argued, misguided. But there is another reason

prospective parents may have for their reluctance to have
such a child. One of the most common concerns parents
have about such children is that when they become adults,
they will not be able to care for themselves, and be left to
an uncaring bureaucracy. Prospective parents may believe
that they would have a lifelong duty of care to such children, one that they could not fulfill. It is not responsive
to their concern to insist that in a better society, children
with such impairments would be meaningfully employed,
decently housed, and respectfully treated. Their fear, perhaps exaggerated but not unreasonable, is that this is not
the case in their actual society; that as parents, they would
have failed in their duties of care. At the least, these are
the right kind of reasons for selecting against a child with a
particular kind of impairment.
A second child-affecting reason for timing or selection may be more problematic: to have children who are
best able to enjoy or appreciate the goods they will offer
thema reason based on a kind of smart philanthropy.
On this view, prospective parents should seek children
whom they will be best equipped to raise and who be the
most receptive beneficiaries of their gift of love, nurturing,
and family. This reason would not favor the selection of the
best-off child if the prospective parents were more qualified and motivated to raise children who were less well off,
and if those children could benefit more from their rearing
than other children. Indeed, it might favor the deliberate
selection of children who would be much worse off than
others they might have had.
Consider an example adapted from Robert Adams.32
Most people would find it wrong for prospective parents
250 |
to seek a child with profound intellectual impairments,

even with reasonable confidence that they would be wonderful caregivers for such a child and would develop a deep
and loving relationship with him. Yet the good of such a
child appears to be part of their reason for seeking to have
one. Indeed, they might insist that they could do far more
good for such a child that for an intellectually average child.
Their course of action would clearly be wrong if what
they sought was a relationship that, on its own terms,
would be disrespectful to the child. Parents who intended
to have a child they would make permanently dependent
on them would treat that child with disrespect, whether
he was intellectually impaired or not. But the prospective
parents in Adams case might plausibly deny that this was
the kind of relationship they sought. They might insist that
they would aim to make that child as autonomous as possible given its cognitive limitations. Although I am troubled
by the prospective parents course of action, I find it difficult to identify the source of disrespect. It may be that despite their intention to form a respectful relationship with
such a child, that prospect is precluded by the very fact that
they sought a child who would be more dependent on them
that the average childhowever much they would strive
to reduce that dependency. What would prevent a respectful relationship is not a psychological barrier but a kind of
moral taint. Because they sought dependency in order to
confer greater good on the child, they could not have the
respectful relationship that parents could have with such a
child had they had her unintentionally.
In general, I do not think prospective parents should
select a future child based on the extent of the good they
can give it. It may be reasonable for them to select against
| 2 5 1
future children who would be utterly impervious to their

love and nurturing because of the most severe cognitive
and psychiatric impairments. But they should not be like
maestros, selecting students who will most benefit from
their master classes. Nor should they be like social workers, seeking the neediest clients.
Parents can be loving and effective nurturers for a wide
variety of children. There may be some children a prospective parent reasonably believes he cannot nurture, and
some children who cannot benefit at all from parenting, or
even form parent-child relationships. Prospective parents
reasonably decline to have children impervious to nurturing, whether because they lack the biological capacity, like
anencephalic infants, or, more fancifully, because they are
born fully mature, like Athena emerging as an adult from
the head of Zeus. But excluding such future children hardly
implies or encourages a more general selectivity.
TENSIONS BETWEEN DIFFERENT ROLES

I N T H E P R O C R E AT I V E C O N T E X T
In responding earlier to Benatars Misanthropic argument, I suggested that prospective parents, like actual
parents, are entitled to discount and even ignore some consequences of their decisions for third parties. Specifically,
neither are required to universalize in making those decisions; to be constrained by the cumulative impact if everyone chose as they do.33 Although I cannot offer a complete
argument for those prerogatives here, I think they are critical to the practice of parenting as we understand it. In an
earlier article, I considered a moral and social division of
252 |
labor that separated progenitors from parents, with the

former having a more impersonal role, much like that envisioned by the accounts discussed in the last chapter. I
argued that although such a role as impersonal gatekeeper
would be coherent, it would be at odds, both morally and
psychologically, with the parental role that succeeded it.
If the same individuals occupied both roles, the transition
would involve a jarring discontinuity, and would give rise
to the tensions about selectivity that I discussed in the sectionbefore last. If different individuals occupied the two
roles, it would replace the paradigm of parenting, in which
the same individuals create and rear children. Although
that paradigm should certainly be relaxed to encourage
adoption, and perhaps other forms of acquiring children,
its wholesale replacement would involve a radical break
from our existing moral and social division of labor.
The existing prerogatives of prospective parents are especially strong in choices about what kind of child to have,
or whether to choose at all; the prerogatives are weaker
in choices about how many children to have, beyond the
choice of whether to have them at all. The ideal of unconditional welcome described earlier applies only to decisions about whether to have a particular kind of child, not
to decisions about how many to have. A parental decision
to have no more children is not like a national decision to
close the borders to immigrants. No one is excluded by the
former; no one is disrespected. Parents and families are
hardly expected to hold up a welcoming beacon to the infinite masses yearning to exist.
Constraints on the number of children have two
sources: the duties of (prospective) parents to other dependents and societal needs that may override their
| 2 5 3
prerogatives. The former constraints arise because of the

risk of harm to other family members. Since they are engaged in forming and maintaining a family, prospective
parents have duties to any existing children that may constrain the number, and even the kind of future children
they bear. Parents lacking the resources to feed their existing children should wait to have any others; parents at risk
of having children with medically complex diseases should
consider (though not exaggerate) the impact of having such
children on their ability to care for their existing children.
This last constraint may conflict with the ideal of unconditional welcome. Imagine parents who want one more child,
but judge that they can only afford an initially healthy one,
not one expected to have costly medical needs from birth.
To avoid conditioning their welcome, they might decline
to have another child at all (a decision often considered by
parents who have children or close relatives with a severe
medical condition34).
Some of the interests of the larger society (e.g., to
correct a sex imbalance or expand the skilled workforce)
appear less appropriate for prospective parents to take
into account. Yet some consideration of the social costs of
having more children, or particular kinds of children, may
be incorporated into their role-based duties. The codes of
most professions include at least some provision for societal interests, though how much weight they should be
given is a matter of protracted debate. For prospective parents in a reasonably just society in conditions of modest
scarcity, it may be unfair to impose on that society the costs
of more than one or two children, or children with predictably expensive medical needs.35 Few actual societies, however, may be sufficiently just to demand such self-restraint,
254 |
especially when so much of the cost of having additional,

disabled, or medically complex children can be attributed
to inadequate health- and child-care and public education,
inflexible work arrangements, and exclusionary structures
and social practices.
To the extent that concerns about justice and social
impact are external to the roles of prospective parents,
their role-based reasons will be exclusionary in Joseph
Razs sense,36 barring or sharply discounting consideration
of the impact of their decisions on third parties outside the
family. If that is the case, prospective parents whose decisions threaten substantial adverse social effects may have
an all-things-considered reason to break role in deference
to their duties as citizens. Some insight into how to understand such conflicts may be found in debates over how
to take account of third-party interests in institutionally
defined roles. For example, some regard certain role-based
duties, such as patient confidentiality, as absolute, but hold
that the individual occupying the role may have an overriding moral reason to violate those duties in the case of
imminent serious or catastrophic harm to third parties.
This is an important, unresolved issue for the morality of
both professional roles and social roles like parents and
prospective parents. But I will not pursue it further here.
Agents of the state and other collective institutions
also may have duties that conflict with those of actual and
prospective parents. The former duties are, in general, consequentialistconcerned with the aggregate health, welfare, or good of the population.37 As public servants, and
even as citizens, individuals have such strong impersonal
duties; as parents and prospective parents they do not.
In the latter roles, I do not think they must take account
| 2 5 5
of concerns about population size or age distribution, or

about population health, in their procreative decisions.
But they are surely permitted to consider take population size; it is not wrong for prospective parents to decide
to have only one child out of concern for overpopulation
or many children out of concern for underpopulation (if
they have justly acquired the resources to support a large
family). And it is clearly justifiable for the state to attempt
to manipulate population size in its tax policies, as long as
it does not impose relatively stronger (dis)incentives on its
poorer citizens.
The potential for conflict arises over concerns about
population health. It is the responsibility of the United
States Centers for Disease Control to reduce the incidence
of genetically based diseases and disabilities associated
with intense suffering or tremendous medical costs. But
it would be inappropriate for prospective parents to act as
agents of the CDC by selecting against such conditions in
order to reduce aggregate suffering or medical costs. Further, it would interfere with the exercise of their role for
the state to offer them incentives for selection, such as tax
breaks or subsidies for using IVF and PGD for selection.
Similar objections have been raised by proponents of procreative autonomy on the basis of protecting the negative
liberty of prospective parents against state interference.
Those objections, although compatible, have a different
focus: not on the different kinds of moral considerations
that are relevant to state agents and prospective parents,
but on the protection of the latters rights against state
interference.
Ironically, some state actions to improve population
health by eliminating or reducing genetically based disease
256 |
and disability may be less problematic than selectivity by

prospective parentswhat Troy Duster has called back
door eugenics.38 On the one hand, deliberate state action
would be explicitly eugenic, regardless of the absence of
coercion. Even if it did not pressure or coerce prospective
parents to select against undesirable traits or for desirable
ones, its actions would clearly send a message about the
desirability of certain traits and undesirability of others. It
would have an expressive significance that, as several philosophers have argued, cannot be found in the decisions
of couples or individuals to select for or against specific
traits.39
On the other handand I think this argument is
strongerit is the states role to promote the aggregate
welfare of its citizens. It is permissible, even mandatory,
for the state to promote the genetic health of the population in developing noncoercive public health and reproductive policies. Because the public health goals the state may
justifiably pursue are not an appropriate concern for prospective parents, the state must be wary of pursuing those
goals in ways that make them their concern, and that seek
to influence their procreative decisions.
Thus, it might be less problematic for the state to reduce
the incidence of genetically based diseases by interventions
that affect the identity of future children but do not operate directly through parental choice. To take an extreme
example, imagine that a drug company developed Leonard
Flecks Omega Pill, which would suppress the operation
of the mechanism that causes up to 80% of severely deformed fetuses to miscarry.40 It would likely place a severe
strain on families, communities, and the society as a whole
if the number of children born with severe, costly genetic
| 2 5 7
disorders dramatically increased. But it would be more objectionable to ban the possession or use of the Omega Pill
by prospective parents than to ban its production and sale.
The former would coerce prospective parents, the latter
would not. To take a more realistic case, it would be more
acceptable for the state to reduce the incidence of birth
defects by putting chemicals in the water that would immobilize mutation-bearing sperm than by offering prospective parents financial incentives for gamete screening
or prenatal testing.
More broadly, conflicts between the duties and prerogatives of prospective parents and public officials can be
mitigated by two kinds of state action: (1) measures that
reduce the incidence of congenital or genetically based
disability that are not directed at parental choice, such as
public health interventions to reduce environmental mutagens; (2) measures that promote the integration of individuals with such disabilities into society, such as inclusive
education and universal design in construction.
C O N C LU S I O N
I have offered a piecemeal defense of procreation: a critique
of arguments that procreation is categorically or presumptively wrong; a proposal of my own intended to strengthen
the justification for having children despite the undeniable
harms and risks of bringing them into the world; and a critical assessment of various standards for permissible procreation. I have not addressed, except in passing, the more
contingent arguments against procreation that challenge
one of the basic prerogatives I have just defendedto have
258 |
a modest number of children. Those arguments claim that

in a world with millions of adoptable children, dire poverty,
and fragile ecosystems, it is unacceptably self-indulgent for
well-off people in developed countries to tie up the considerable resources, and to impose the significant externalities, required to create and rear even a single child.
To the extent these contingent arguments do not rest
on a consequentialist calculus, but on the moral claims of
individuals, I find them more plausible and powerful than
the categorical ones. But I have two reasons, or excuses,
for not having given them closer attention. First, I think it
is important to first challenge the categorical arguments,
which, if correct, would render the contingent ones superfluous. Second, I am ill equipped to take on the latter without going well beyond the scope of this book. Although I
have argued, or perhaps asserted, that prospective and
actual parents have strong prerogatives against certain
kinds of third-party claims, I have not attempted to defend
a specific view of the content and scope of their prerogatives against the urgent needs and putative rights of existing people in dire straits.
Moreover, the issue of how much well-off people owe
badly off people, especially from different countries, is
hardly limited to the context of procreation. It is an general
issue in ethical and political philosophy, and it can only
be responsibly engaged by addressing sharply contested
factual and policy claims.41 What, for example, are the
harms and risks to individual biological parents, family
structures, and support services of adopting children from
poor neighborhoods or poor countries? How could the resources saved by not bearing children be used in ways that
help rather than harm badly off people in impoverished
| 2 5 9
communities or less developed countries? How much of an

incremental burden to the environment is an additional
child in a developed country with low birth rates and increasingly widespread and effective green policies and
practices? Clearly, collecting and evaluating the conflicting
evidence on any of these questions would require a separate essay.
A more general reason for the piecemeal character of
my defense is that it is difficult or foolhardy to attempt
a wholesale defense of practices as deeply embedded in
our way of life as bearing and rearing children. This is
not to deny, of course, that many aspects of those practices should be subject to close scrutiny. I believe the antinatalists are correct on this important point. But it does
mean that we should distrust our capacity to criticize
those practices from an external perspective, from outside of a human community that takes its perpetuation as
an unquestionable good. Of the anti-natalist arguments I
have addressed, only Benatar seeks to do so, and I am not
the first to criticize him for adopting a perspective that
is beyond his, and our, reach. But most of his arguments,
and almost all of the other anti-natalist arguments, do not
adopt such a perspective. They seek to persuade us that
procreation is wrong because it threatens bedrock values,
which cannot receive, and do not need, further argument
to accept: the wrongfulness of harming other people without adequate justification; the importance of consent in
imposing certain kinds of harms and risks, etc. Like most
defenders of procreation, I have adduced the same values
in defending it against anti-natalist challenges, and in defending my own view on the correct standards for permissible procreation.
260 |
NOTE S
1. Michael Tooley, Abortion and Infanticide (Oxford: Clarendon,
1983), 272.
2. Elizabeth Harman, Can We Harm and Benefit in Creating? Philosophical Perspectives 18, no. 1 (2004): 89113.
3. David Benatar, The Wrong of Wrongful Life, American Philosophical Quarterly 37, no. 2 (2000): 175183.
4. Bonnie Steinbock, Wrongful Life and Procreative Decisions, in Harming Future Persons: Ethics, Genetics, and the
Nonidentity Problem, eds. Melinda Roberts and David Wasserman (New York: Springer, 2009), 155178.
5. David DeGrazia, Creation Ethics: Reproduction, Genetics, and
Quality of Life (Oxford: Oxford University Press, 2012), 168.
6. DeGrazia, Creation Ethics.
7. Creation Ethics, 169.
8. Creation Ethics.
9. Steinbock, Wrongful Life, 165.
10. Joshua Glasgow, The Shape of a Life and the Value of
Loss and Gain, Philosophical Studies 162, no. 3 (2013):
665682.
11. See ch. 9 this volume, sec. iii.
12. This may no longer be DeGrazias position. See David DeGrazia, Procreative Responsibility in Light of What Parents
Owe Their Children, in Oxford Handbook of Reproductive
Ethics, ed. Leslie Francis (New York: Oxford, ms. forthcoming), 1718.
to Create Children with the Best Chance of the Best Life,
Bioethics 23, no. 5 (2009): 274290, 283.
14. Caspar Hare, Voices from Another World: Must We Respect
the Interests of People Who Do Not, and Will Never, Exist?
Ethics 117, no. 3 (2007): 498523, 518519.
15. David Wasserman, Hare on De Dicto Betterness and Prospective Parents, Ethics 118, no. 3 (2008): 529535.
16. David Velleman, Persons in Prospect, Philosophy & Public
Affairs 36, no. 3 (2008): 221288, 272.
17. Velleman, Persons In Prospect.
| 2 6 1
18. Robert Adams, Must God Create the Best? Philosophical

Review 81, no. 3 (1972): 317332.
19. Velleman, Persons In Prospect, 276.
20. Rosalind MacDougall, Parental Virtue: A New Way of
Thinking about the Morality of Reproductive Actions, Bioethics 21, no. 4 (2007): 181190.
21. Adrienne Asch and David Wasserman. Where Is the Sin
in Synecdoche? In Quality of Life and Human Difference:
Genetic Testing, Health Care, and Disability, eds. David Wasserman, Robert Wachbroit, and Jerome Bickenbach (Cambridge: Cambridge University Press, 2005), 172216. This
ideal of unconditional welcome is fully consistent with
limits on numbers. Prospective parents have limited psychological, social, and economic resources, and they should
hardly be required to exhaust them on additional children.
But setting limits on the numbereven if those limits are
affected by the costs of raising a particular child already in
the familyis not a form of trait-based selectivity.
22. There would be the same kind of tension in a family formed
by the kind of criteria, like psychological compatibility,
used in forming other kinds of intimate relationships.
Each child would, if he understood the selection process,
be aware that he might not have been found sufficiently
compatible by the more refined screening used to select
his younger siblings. A standard of expected compatibility
may conflict with the inclusive character of the family even
more than a standard of expected well-being. A child aware
that he might have been excluded as an insufficiently good
match for his parents might reasonably take his hypothetical rejection more personally than a child who might have
been excluded because of insufficiently good prospects for
flourishing.
23. Peter Herissone-Kelly, Two Varieties of Better-For Judgments, International Library of Ethics Law and the New
Medicine 35 (2009): 249264. Herissone-Kelly grounds his
incumbent model in the duty of parents not to favor one
of their actual children over another. An unqualified equation of the roles of prospective and actual parents, though,
262 |
would be oppressive and possibly incoherent. For those who

deny that embryos have the same rights or moral status as
children, it is untenable to require prospective parents to
have an unconditional commitment to all the embryos in
an IVF array. Actual parents would be obliged to overload
a lifeboat to include all their present children; prospective
parents are hardly obliged to implant all of their IVF embryos, at substantial risk to the survival and development
of each of them. Those who implant even a large number,
like the so-called Octomom, are condemned for irresponsible procreation. The notion of parental commitment is
even harder to apply to future children not embodied in
embryos. It is difficult to understand how prospective parents could feel even attenuated love and devotion to the
possible embryos they might conceive at a given time, let
alone toward the possible children they might have together or with others.
24. Herrisone-Kelly, Two Varieties, 258259.
25. Two Varieties, 259261.
26. Hallvard Lillehammer, Reproduction, Partiality, and the
Non-Identity Problem, in Harming Future Persons, eds. Melinda Roberts and David Wasserman (New York: Springer,
2009), 231248; Benefit, Disability, and the Non-Identity
Problem in Philosophical Reflections on Medical Ethics, ed. Nafsika Athanassoulis (New York: Palgrave McMillan, 2005),
2443, 3439.
27. Niko Kolodny, Which Relationships Justify Partiality? The
Case of Parents and Children, Philosophy & Public Affairs 38,
no. 1 (2010): 3775.
28. None of these views preclude preferences for a more viable
embryoone more likely to become a childand Lillehammers, as noted, permit a preference for an embryo from the
couples own gametes, rather than one created from thirdparty gametes. Herissone-Kelly and Asch and Wasserman
do not address the issue of third-party gametes, but their
views would be incompatible with their use for trait selection by fertile prospective parents.
29. Lillehammer, Benefit, 3439, suggests that prospective parents may select children likely to share their own
| 2 6 3
weaknesses or impairments, if those features did not preclude a worthwhile life or impose ruinous costs on third
parties. Wasserman himself entertains the idea that prospective parents can show partiality among their possible
children, however suspect their preferences (Ethical Constraints on Allowing or Causing the Existence of People
with Disabilities, in Disability and Disadvantage, eds. Kimberley Brownlee and Adam Cureton (New York: Oxford
University Press, 2009), ch. 11.) He recognizes that this position is in tension with the unconditional welcome he and
Asch argue for (Asch and Wasserman, Where is the Sin?)
For Herissone-Kelly (Two Varieties), there seems to be no
room in parental role-morality for favoring some possible
children over others if all are expected to have lives they
would judge to be acceptable.
30. Random selection would not require an actual equiprobable
lottery, in any sense of probability, since possible children and
IVF embryos do not have a right to equal chances of conception.
1984), 358359.
32. Robert Adams, Must God Create the Best? 317332.
33. Chapter 8, 1920, this volume.
34. Felicity Boardman, The Expressivist Objection to Prenatal
Testing: The Experiences of Families Living with Genetic
Disease, Social Science and Medicine, 107 (2014): 1825.
35. Stephen John, Efficiency, Responsibility and Disability:
Philosophical Lessons from the Savings Argument for PreNatal Diagnosis, Politics, Philosophy and Economics (2013):
1470594X13505412.
36. Joseph Raz, The Morality of Freedom (New York: Oxford,
1986).
37. Robert Goodin, Government House Utilitarianism, in The
Utilitarian Response: The Contemporary Viability of Utilitarian Political Philosophy, ed. Lincoln Allison (London: Sage,
1990), 140160.
38. Troy Duster, Backdoor to Eugenics (New York: Routledge, 1990).
39. Nelson, James Lindemann. Testing, Terminating, and Discriminating, Cambridge Quarterly of Healthcare Ethics 16, no.
4 (2007): 462468; Adrienne Asch and David Wasserman,
264 |
A Response to Nelson and Mahowald, Cambridge Quarterly

of Healthcare Ethics 16, no. 4 (2007): 468.
40. Leonard Fleck, Abortion, Deformed Fetuses, and the
Omega Pill, Philosophical Studies 36, no. 3 (1979), 271283.
41. Judith Lichtenberg, Distant Strangers: Ethics, Psychology,
and Global Poverty (New York: Cambridge University Press,
2014).
INDEX
abortion, 16n2, 63, 89, 216, 229

accommodation. See adaptation
Adams, Robert, 240, 249, 250
adaptation (psychological trait),
424, 73n8, 157
adoption, 5, 12, 16n4, 1256, 139,
187, 18891, 193, 195,
197, 199, 207n20, 252
animals, 13, 49, 56, 81, 84, 85,
93102, 1057, 10910,
166, 167
Anscombe, Elizabeth, 136, 137
anti-frustrationism. See
frustrationism
anti-natalism, 1, 4, 1317, 37,
38n5, 38n6, 40, 62, 789,
101, 111, 112n1, 122,
131, 13743, 14881,
183, 200, 203, 205, 259.
See also comparative
argument; misanthropic
argument; quality-of-life
Asch, Adrienne, 197, 2436,
261n21, 262n28, 263n29
asymmetries, 137, 148, 177,

177n3, 177n4, 178n8,
182, 185, 205n17, 209,
221. See also axiological
asymmetry; empirical
asymmetries
axiological asymmetry, 3,
4, 15, 2138, 40, 48,
67, 120n80. See also
comparative Argument
Belshaw, Christopher, 145n7,
177n3, 214, 221
Benatar, David, 15, 1378,
14868, 173, 176,
177n1, 177n4, 179n14,
181n37, 182, 209, 231,
251, 259
benefit
existence not (intended as) a,
18799
intrinsic, 31, 33, 34
Pure, 70, 137, 148, 16872
relief, 304, 45, 51
266 |
I N D E X
Bennett, Rebecca, 219

biological drives, 11, 13
birthrights. See rights
Boonin, David, 178n9
Buchanan, Allen, 2147, 218,
231, 238, 242
comparative argument (Benatar),
4, 14854. See also
axiological asymmetry
comparison (psychological trait),
434, 53, 75n26
conformity, 823
consent argument (Shiffrin), 4,
701, 1489, 16872
consequentialism, 138, 139,
146n9, 215, 218, 220,
254, 258. See also
utilitarianism
consequentialist. See
consequentialism
consumption, 82, 93, 99, 109,
110
crowds. See mobs
death, 14, 48, 67, 689, 79, 858,
90, 92, 93100, 104,
1245, 130, 148, 168,
170, 171, 200, 234
DeGrazia, David, 171, 174,
178n12, 180n34, 204,
2316, 247
Devolder, Katrien, 220
disability, 3, 217, 2356, 2401,
244, 2547
disease, 157, 1612, 170, 217,
244, 253, 2556. See also
quality-of-life
Douglas, Thomas, 220
Duster, Troy, 256
duties
impersonal, 6, 20911, 2167,

254
role-based, 1434, 21011,
2134, 228, 230, 236,
23945, 2534
Eichmann, Adolph, 84
Elster, Jakob, 220
empirical asymmetries, 378,
4854
endowment effect, 81
enhancement, 601, 161
environment, 85, 99100,
10911
evolution, 11, 80
extinction, 13, 75n27, 100,
1289, 223
Fehige, Christoph, 32
Fleck, Leonard, 256
Framing effect, 81
frustrationism, 323
genitalia, 7, 13, 87
genocide, 86, 91, 106
Gibson, Susanne, 1845
Goodin, Robert, 263n37
good, noncomparative, 41,
177n4, 192, 222
habituation. See adaptation
Hare, Caspar, 23942
harm
distinguished from wrong,
149, 1505
(not) outweighed by good, 4,
18, 24, 35, 1067, 142,
150, 1523, 168, 177n1,
1934, 200, 203
priority of avoiding, 176177
INDEX
Harman, Elizabeth, 231

Hayry, Matti, 4, 149, 155, 1726,
177, 234
Herissone-Kelly, Peter, 2447,
261n23, 262n28,
263n29
Heyd, David, 1367
holocaust. See genocide
Homo infortunatis, 578
Homo perniciosus, 85100
Homo sapiens, 136, 5860, 128,
129. See also qualityof-life; misanthropic
argument
humanity, 140, 1601, 166. See
also Homo sapiens
impersonal views, 56, 21, 356,
1434, 207n21, 20927,
228, 231, 2389, 246,
2524
injury, 49, 53, 54, 62, 69, 98, 148,
157
injustice, 92
intentions (procreative) and
permissibility, 3, 199205
Kahane, Guy, 21718, 220,
226n27, 238
Kamm, Frances, 180n34,
226n13
Kantian. See Kant, Immanuel
Kant, Immanuel, 180n29, 190
Kerstein, Sam, 208n27
killing, 14, 83, 8590, 939, 107,
110, 202
Lichtenberg, Judith, 215n37
life (not) worth living, 1921,
35, 173, 176, 177n3, 229,
2334, 245
| 2 6 7
Lillehammer, Hallvard, 2467,

262n29
Lotz, Mianna, 207n19
Marsh, Jason, 178 9n14
Maslow, Abraham, 51
McMahan, Jeff, 182, 1924, 212,
214, 2224, 228
Meilander, Gilbert, 1367, 185
Mere Addition problem, 36
Milgram, Stanley, 167
minimum standards for
permissible procreation,
13940, 144, 229, 2306
misanthropic argument, 4,
5, 156, 78121, 149,
1668, 251
mobs, 901
non-identity problem, 1821,
346, 219
obedience, 834, 167, 179n28
optimism, 12, 22, 412, 43, 44,
45, 48, 51, 53, 5461, 66,
71, 81, 91, 104, 109
Overall, Christine, 136, 1857, 195
paradox of future individuals. See
non-identity problem
parents
adoptive. See adoption
biological, 5, 27, 189, 1935,
197, 207n20, 258
prospective, 3, 5, 6, 7, 12, 28,
68, 70, 137, 139, 1415,
154, 155, 162, 1656,
1678, 1767, 179n14,
182208, 2101,
2134, 217, 219, 220,
223, 22864
268 |
I N D E X
Parfit, Derek, 36, 146n9, 2112,

224, 248
person-affecting theories, 345,
103, 194, 214, 216,
226n27
pessimism, 5, 57, 68, 104, 105
pollution. See environment
pollyannaism, 155, 157. See also
optimism
Ponzi scheme, procreational,
12930
population, 347, 99, 1089, 110,
181n37, 194, 212, 230,
248, 2546
post-humans. See
transhumanism
procreative selection, 6, 141,
144, 201, 20911, 214,
2168, 21922, 2245,
235, 2378, 240, 2425,
2479, 251, 252, 2556,
261n21, 261n22, 262n28,
263n29, 263n30
quality-of-life, 34, 5, 15, 4077,
153, 15566, 224
Rawls, John, 173, 1756
Raz, Joseph, 254
reasons for procreation
child-centered, 3, 18499
and permissibility of
procreation, 199205
relationships, 5, 31, 467, 1245
parent-child and other
intimate, 3, 208, 215,
238, 24350, 261n22
religion, 13, 25, 54, 80, 87, 89,
1223, 126
Repugnant Conclusion, 356,
146n10
rights, 6, 14, 1078, 13940,

143, 154, 16972, 189,
210, 218, 22831, 255,
258, 262, 263n29
risk argument (Hayry), 4, 142,
1489, 1512, 1726
Roberts, Melinda, 205n1,
207n20
Robertson, John, 146n15
Russian Roulette, procreational,
656, 165
same-number choices, 1434,
2112, 2145, 2179,
224, 229
Savulescu, Julian, 2178, 220,
238, 242
Scheffler, Samuel, 146n17
Schopenhauer, Arthur, 32, 49
Shiffrin, Seana, 4, 701, 148,
165, 16672, 1734, 176,
180n32, 181n37, 203
slavery, 86, 87, 889
Smuts, Aaron, 161
speciecide, 14
Spinoza, Baruch, 90
Stanford prison experiments, 84
Steinbock. Bonnie, 2314, 247
stupidity, 46, 812, 89
subjective well-being. See
well-being
suffering. See quality-of-life
suicide, 14, 712
theodicies. See optimism
Tooley, Michael, 231
transhumanism, 601
unintentional procreation,
1112, 186
utilitarianism, 25, 107, 120n79
INDEX
Velleman, David, 208n23,

2402
Wasserman, David, 1, 2, 4, 5, 6,
75n28, 2437, 261n21,
262n28, 263n29
Weinberg, Rivka, 7n4, 170, 176,
178n8, 206n13
| 2 6 9
well-being, 6, 144, 146n8,

156, 158, 160, 163, 190,
21213, 21820, 221,
2234, 229, 231, 234,
2367, 247, 261n22.
See also quality-of-life
Zimbardo, Philip, 167

(Debating Ethics) Benatar, David - Wasserman, David-Debating Procreation - Is It Wrong To Reproduce - Oxford University Press (2015)

Uploaded by

Copyright:

Available Formats

(Debating Ethics) Benatar, David - Wasserman, David-Debating Procreation - Is It Wrong To Reproduce - Oxford University Press (2015)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(Debating Ethics) Benatar, David - Wasserman, David-Debating Procreation - Is It Wrong To Reproduce - Oxford University Press (2015)

Uploaded by

Copyright:

Available Formats

DEBATING PROCREATION

Oxford University Press 2015

Better to Have Lived and Lost?

8. The Good of the Future Child and the

READERS OF A BOOK DEBATING the ethics of procreation

Unlike the majority of pro-natalists, he rejects the idea

deal of technical criticism. David Wasserman contends that

will make the world best off, or some combination of these

IT IS ALWAYS DIFFICULT TO convince people that a wide-

spread practice in which they participate is morally wrong.

more commonly, a mere impulseto have sex.3 A small

I ntroducing A nti - N atalism

The parents themselves often seem hugely impressed with

This outcome is certain. The uncertainty concerns when

I ntroducing A nti - N atalism

I ntroducing A nti - N atalism

5. It is unlikely that this exhausts the number of unwanted

The Asymmetry Argument

MANY PEOPLE THINK THAT COMING into existence is a

harm only if the person brought into existence has a life

THE NON-IDENTITY PROBLE M

The problem arises in those circumstances in which a

into existence unless his life is not worth living. We should

the life was not good enough to be worth starting, it is

THE AXIOLOG ICAL ASYM METRY

no reference to the interests of the person who either is

figure 2.1 The Axiological Asymmetry

Some people think that the way to determine whether

itself provides insurmountable evidence for the axiological

about war-torn and poverty-stricken parts of the planet

However, if it is the sacrifice involved in procreation

accepting the implications of it, because it is far too easy to

PROBLEMS WITH SYMMETRY

One way is to suggest that the absence of benefit in never

SYMMETRY AND RELIEF BENEFITS

assuagement of guilt, and emergence from ignorance. The

of harms. For example, continuous hydration can prevent

need for oxygen. Fulfillment of a need is good only if the

thus the cost-benefit calculation is much less likely to come

there will be the residual problem of explaining how bad

The impersonal average view, according to which the

THE AXIOLOG ICAL ASYM METRY

support for axiological asymmetry. Given the empirical

7. Genesis 1:28. Note, however, that not all religious people

The Quality-of-Life Argument

THE ASYMMETRY ARGUMENT IS sufficient to reach the con-

T he Q uality - of - L ife A rgument

lives. The common view, however, is that the quality of some

WHY PEOPLE S JUDG MENTS ABOUT

commonly experienced level, suggesting, in the words of

T he Q uality - of - L ife A rgument

but the altered self-assessment is not stable. As we adjust

T he Q uality - of - L ife A rgument

objective level. Even if ones overly optimistic subjective

quite a bit of the day tired. Indeed, many people wake up

T he Q uality - of - L ife A rgument

spouse. For some this desire is never fulfilled. For others

Some deaths are better than others, but there is always

T he Q uality - of - L ife A rgument

Better to Have Lived and Lost?

8. The Good of the Future Child and the