Why Fertility Changes

Author(s): Charles Hirschman

Source: Annual Review of Sociology, Vol. 20 (1994), pp. 203-233
Published by: Annual Reviews
Stable URL: http://www.jstor.org/stable/2083365 .
Accessed: 10/01/2011 11:33
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .
http://www.jstor.org/action/showPublisher?publisherCode=annrevs. .
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

Annual Reviews is collaborating with JSTOR to digitize, preserve and extend access to Annual Review of
Sociology.

http://www.jstor.org

Annu. Rev. Sociol. 1994. 20:203-233

Copyright0 1994 by Annual Reviews Inc. All rights reserved

WHY FERTILITYCHANGES
Charles Hirschman
Centerfor Studiesin Demography
andEcology,Department
of Sociology,DK-40,
Universityof Washington,
Seattle,Washington
98195
KEYWORDS: demography,
demographic
transition
theory,fertilitytransitions,
family

Abstract
There is considerable controversy over the causes of the completed fertility
transitionsthat occurredin most industrialcountries from 1870 to 1930 and
the "new" fertility transitionsthat are currentlyunderway in the developing
world. New data and empirical analyses of both historical and contemporary
fertility declines have weakened the standardtheory of the demographictransition, but none of the plethoraof new theoriesof fertilitychangehave emerged
as hegemonic or as alternativeguides to empiricalresearch.The vast body of
empiricalevidence on the origins, speed, and correlatesof fertility declines in
different historical and geographical settings shows more diversity than a
simple theory of fertility change would predict. The challenge for the field is
to develop a common theoreticalframeworkthat will accommodatethe diversity of historical paths from high to low fertility.

INTRODUCTION
Over the course of the last century, changes in death and birth rates have
transformedthe characterof life for virtuallyevery society and family on the
planet.Decreases in mortalityhave led in most partsof the world to reasonable
expectationsthatparentswill see virtuallyall of theirchildrensurvive infancy.
Childbearinghas receded from the center stage of family life and from its
primaryrole in the lives of adult women to become an option that can be
scheduledand sequenced with vocational and lifestyle pursuits.Most parents
can expect to live to see their grandchildren.Although these "new" demographicpatternsare most common in advanced industrialsocieties, they are
on the near-termhorizonfor most societies aroundthe globe. If humanprogress
is to be measuredby longevity and reproductivecontrol, the present century,
203
0360-0572/94/0815-0203$05.00

204

HIRSCHMAN

and the second half of it in particular,has no historical parallel for the advancementof the human condition.
Many people, social scientists included, assume that these demographic
revolutions (including those still in process in the developing world) are
products of the economic and technological changes of the modem era that
have led to economic development,mass communications,effective programs
of public health and curativemedicine, and related social changes. This basic
idea-that lowered mortality and lowered fertility, after some lag period,
follow from socioeconomic development-is widely known as the theory of
the demographictransition.Demographers,however, are not so sure. Broad
empirical generalizationsand theory constructionwere perhapssimpler tasks
in an age with little empirical data. Over the past few decades, intensive
researchon demographicchange in historical and contemporarysocieties has
revealedcomplex patternsthatdo not fit neatly into earliertheoreticalschema.
The fact that fertility transitionsin many developing countries are still "in
process" adds more uncertaintyto the search for explanations.
In recentyears, the field of demographyhas spawneda varietyof new ideas,
conceptual and measurement frameworks, and theories of demographic
change. The debates in the journals are hot with conflicting claims on every
issue from questions of measurementand the relative importanceof causal
forces to the ideological bias of researchersand of the entire field (Thomas
1991, 1993, Cleland 1993). This essay is a criticalreview of some of the recent
researchand the theoreticaldebates on fertility transitionsin different social,
economic, andculturalcontexts. Researchon contemporaryfertilitytrendsand
variationsin low fertility settings (the United States and Europe) falls into a
somewhat differentliteratureand is not directly covered here.
The review begins with issues of measurement of fertility and fertility
change. Next, I summarizethe recent evidence on fertility levels and trends
in historical and contemporarypopulations.The core of the essay consists of
a comparisonof classical and contemporarytheories of fertility transitionsand
a discussion of unresolvedissues in currentresearch.AlthoughI offer my own
evaluation of the relative merits of different approachesin the field, I do not
attemptto provide closure to currentdebates.In lieu of a conclusion, I suggest
an alternativemodel for the field that implies a question very different from
the one posed in the title of this essay.

THE MEASUREMENT OF FERTILITY AND FERTILITY

CHANGE
Questions of theory and interpretationare closely bound up with the measurement of the phenomena of interest. Any discussion of current research on
fertility must be based on a clear understandingof some critical aspects of the
conceptualizationand measurementof fertility.

WHY FERTILITYCHANGES

205

Fertility Rates and Population GrowthRates

It is easy for the nonspecialist to become confused with the variety of demographicmeasuresand rates.For example, populationgrowthrates and fertility
rates are often conflatedin the media (and sometimes by scholars). At present,
populationgrowthratesin many developing countriesremainhigh even though
fertility rates have declined rapidly. This is because mortalityhas decreased
furtherthan fertility, and the age structureof the populationsof many developing countrieshas a disproportionatenumberof persons in the childbearing
ages (this is a byproductof high fertilityin prioryears). The confusion between
populationgrowth rates and fertility rates often leads to frustrationfor those
who just want a simple answer to the question of whether the "population
problem"is getting worse or getting better. The question needs to be more
precisely framed.
For the purposes of this essay, the reader need have only general understandingof one fertilitymeasureand one relatedconcept. The index of fertility
is the "totalfertilityrate"(TFR), which is the averagenumberof childrenborn
to women who survive to age 50 in a population. The additional concept is
"replacementlevel fertility,"which is a TFR of a little more than two births
the reproductivelevel needed to ensure the replacement of the mother and
fatherin the next generation.

Period and CohortPerspectives

Fertility can be viewed as either a life-course phenomenon of women (and
men) over their reproductivecareersor as behavior within a specific interval
(e.g. a single calendar year). This distinction-between cohort and period
perspectives (and schemes of measurement)-is a central one for fertility
analysis. Some measuresof fertility can be constructedwith either cohort and
period data (e.g. total fertility rates, although children-ever-bornis the more
common term for a cohort measure),and care must be taken to be sure of the
basis of measurementand the implications for interpretation.
The conventionaldemographicwisdom until fairly recently was that cohort
measures of fertility were superiorto period measures because they tap the
life experiences of a real group that lives through a particularera of history
together (Hajnal 1947, Ryder 1969, 1983). Ryder (1965) argued that social
change (not just demographic change) occurs primarily via the behavioral
patternsof new generationsthat are exposed to significanthistorical events at
the formative stage of their socialization. Much fertility research, however,
has tendedto rely on period measuresbecause of the natureof available data.
Completedcohortfertilitycannotbe computeduntil the end of the childbearing
span for each generation(until age 45 or 50). The formal interdependenceof

206

HIRSCHMAN

period,cohort,andage has madeit difficult,but not impossible,to estimate

effectsof each.
In spite of the generaltheoreticalpreferencefor cohortmeasures,recent
empiricalstudiesof fertilitytrendshavefoundthatperiodinfluencestendto
be morepowerfulthancohortinfluencesin explainingvariationsin fertility
rates(Foster1990, Ni Bhrolchain1992, Rafteryet al 1993). In modelsof
cyclicalpatternsof fertilitychangein advancedsocieties,the distinctionhas
been criticalbecausehypothesesare typicallyformulatedin terms of the
interplayof cohortandperiodprocesses(Easterlin1976b,1978a,Butz& Ward
1979).Muchless debatehasoccurredoverthe appropriate
frameof reference
for studiesof fertilityin developingcountries,but the greateravailabilityof
dataandnew analyticalmethodswill probablyshiftattentionto this issue in
the comingyears.

The Focus on the Emergence of Fertility Control

In muchof the recentliteratureon historicalandcontemporary
fertilitytransitions,the primaryfocushasbeenon theemergenceof intentionalcontrolof
fertilitywithinmarriage(incontrastto variationsandchangein absolutelevels
of fertility).The reasonsfor this emphasisare foundin severalinfluential
studiesand their links to a particulartheoreticalinterpretation
of fertility
change.BecauseI regardthis emphasisas undulynarrow,I wantbrieflyto
reviewtheconceptualandmethodological
of theseclassicstudies
assumptions
andtheirsubsequentimprinton the developmentof fertilityresearch.First,a
digressionon potentialandactualfertilityis necessaryto set the stagefor the
discussionof the conceptsof "natural
fertility"and"controlled
fertility."
Oneof the nonintuitive
featuresof humanreproduction
is thathigh(seemingly uncontrolled)
levels of fertilityare actuallywell below whatwouldbe
unrestrained
fertility(Bongaarts1975).High-fertilitysocietiesrarelyexceed
an averageof eightor nine live birthsper woman,whichis well below the
biologicalmaximumthat could be achievedin the absenceof social and
culturalnormsthatregulatemarriagepatterns,frequencyof intercourse,and
in noncontracepting
lengthof breastfeeding
societies.The classic articleson
"intermediate
variables"by Davis & Blake (1956) and on "proximatevariables"by Bongaarts(1978,Bongaarts& Potter1983)provideconceptualand
fortheincorporation
of thesesociobiologicalfactorsas
analyticalframeworks
interveningvariablesin the fertilitydetermination
process.Theseconceptual
frameworks,however,do not explainthe originsof the social and cultural
influencesthatconstrainfertilityin high-fertilitysettings.
In oneof themostinfluentialarticlesin moderndemography,
Henry(1961)
presentedthe conceptof "natural
fertility"to characterize
fertilityin societies
wheremarriedcouplesdo notconsciouslytryto limitthe numberof children
theyhave.Variationsin naturalfertilitycanrangeby a factorof two or more

WHYFERTILITY
CHANGES 207
(e.g. froma TFRof 4 or 5 to morethan8). Henrydefined"fertilitycontrol"
behaviorto restrictfertility(breakingwithnaturalfertility)as parity-specific
"whenthe number(of births)reachesthe maximumthatthe coupledoes not
want to exceed"(1961:81).Transitionsfrom naturalfertilityto controlled
maritalfertilityarecriticalhistoricalmoments,accordingto Henry,thatbegan
socialcustomsto the low
the shiftfromthe regulationof birthsby traditional
levels of reproduction
thataretypicalof modemsocieties.
control
Henry'sdefinitionof naturalfertilityas the lack of parity-specific
led to the developmentof a varietyof ingeniousindirectmethodsto measure
demographic
intentions(consciouseffortsto limitfertility)withonly standard
data.Theseinnovativemethodsincludetheinspectionof theshapeof age-specific fertilitycurves (Knodel 1977) and the famous "M & m" indicators
developedby Coale& Trussell(1974,1978)as deviationsfromanempirically
observedset of naturalfertilitypatterns.Gradually,these new methods,and
the assumptions
behindthem,led to a focus on explainingthe emergenceof
fertilitycontrolas the primaryquestionin the field-and a lackof interestin
fertility"populations.
explainingthewidevariationsin fertilityamong"natural
This focus on the emergenceof consciousfertilitycontrolis illustratedin
the analysisandthe conclusionsof the PrincetonEuropeanFertilityProject
(Coale& Watkins1986).Evenwhenone couldnot show directmeasuresof
"parity-specific
control,"patternsof sustaineddeclinesin maritalfertilitywere
interpretedas the productof consciousplanningby couplesto limit their
may well
fertilityafterreachingtheirdesiredfamilysize. This interpretation
was enhancedby the fit with the "new"
be correct,but its attractiveness
from1870to 1930as a product
explanation
of theEuropean
fertilitytransition
of the diffusionof ideasandknowledgeof the meansof fertilitycontrol.
Thereareseveralproblemswiththe exclusivefocuson the transitionfrom
naturalto controlledfertilityas thecentralempiricalquestionin thefield.Most
questionof whetherfertilitycontrolis accubasicis the simplemeasurement
et al (1992)showthatchanges
ratelymeasured
by indirectmeasures.Guinnane
in aggregatefertility(Ig) and"littlemi"maybe poorindicatorsof the origins
of fertilitycontrol(definedas stoppingbehavior).Moreover,fertilitycontrol
andthuspresenta definicouldbe basedon "spacing"ratherthan"stopping"
tional problemfor the distinctionbetweennaturaland controlledfertility
evidencethatlengtheningof birth
(Knodel1983).Indeed,thereis considerable
intervalswas an importantcomponentof historicalfertilitytransitions(Bean
et al 1990:207)andcontemporary
fertilitydeclines(Caldwellet al 1992).
Anotherproblemis thatvariationsin naturalfertility(thatare not subject
to consciousplanning)aredeemedto be theoreticallyunimportant.
Although
therapidpaceof declinein maritalfertilitythateventuatesin verylow fertility
mayrequireconscioususe of birthcontrol,the varietyof socialcustomsthat
spousalseparation,
regulatemarriagebehavior,divorce,widow remarriage,

208

HIRSCHMAN

length of breastfeeding,and coital frequency may well reflect the impact of

economic and environmentalinfluences on reproduction.In other words, consciousness may be an importantintervening variable, but it is not an indispensable ingredientin the story of how fertility is socially regulated.
The other dilemma of the "naturalfertility-to-controlledfertility"scenario
is the need to show that pretransitionsocieties did not know how to regulate
marital fertility (Knodel & van de Walle 1979). If the means of fertility
regulationwere widely known to marriedcouples in pretransitionpopulations,
it is difficult to maintainthe assumptionof naturalfertility in all pretransition
populations.The evidence is, however, mixed, and several well-chosen examples do not establish the patternfor all pretransitionsocieties. Counter-examples of control of maritalfertilityin pretransitionpopulationscan also be cited
(Wrigley 1966, Andorka 1982). Blake (1985) argues that there is more continuity than divergence in the practice of marital fertility control among preand post-transitionsocieties.
Van de Walle (1992) has argued that pretransitionsocieties do not have a
clear concept of desired family size and that there is a lack of conscious
thinking about family size. This may be true in many instances, but it does
not mean that fertility behavior was not "regulated"in many pretransition
societies. My conclusion is simply thatthe focus on the emergence of fertility
control is both too narrow(much of significant fertility variationis ignored)
and unnecessary-individuals and populations can reduce fertility without
necessarilyplanningconsciously a desiredfamily size. The distinctionbetween
naturaland controlledfertility may be an importantelement in the process of
modem fertility transitions,but the currentemphasis seems disproportionate
to its value.

HISTORICALAND CONTEMPORARYTRENDS IN
HUMAN FERTILITY
The assumptionof constant high fertility in pretransitionsocieties is widely
held, but there is considerable evidence of systematic variation. One of the
most significantfindingsof recentresearchis thatfertilityis higherin intensive
agriculturalsocieties than in forager(huntingand gathering)and horticultural
(swidden agricultural)societies (Gillian et al 1993). The reasons for this
difference (and the behavioralmechanisms)are unclear,but possible explanations could be related to higher mortality in agrariansocieties, problems of
caring for too many infants and small childrenin migratorypopulations, and
earlier weaning of infants in settled agriculturalpopulations. Anthony Reid
(1988:158, 162, 1992:461) speculatedthatthe spreadof world religions (Islam
and Christianity)in settled agriculturalpopulationsin SoutheastAsia led to a
rise in fertilityrelativeto thatof migratorypopulationswith traditionalanimist

WHYFERTILITY
CHANGES

209

beliefs. He reasoned that traditionalbelief systems did not forbid premarital

sexuality which coexisted with endemic gonorrhearesulting in high levels of
sterility.
There were systematic oscillations in fertility in preindustrialEurope in
responseto changes in economic conditions,primarilythroughthe mechanism
of marriage (Wrigley 1969, Galloway 1988, Wilson & Woods 1991). The
European(actually WesternEuropean)marriagepattern,which Coale (1973)
labels the "Malthusiantransition,"emergedsometime duringthe Middle Ages.
Therewere two componentsof the Europeanmarriagesystem, the high average
age of people at marriageandthe high proportionof people who never married.
These patternsvaried widely within and between populations, and they rose
and fell in response to economic conditions (Goldstone 1986, Hajnal 1965).
At times of economic crisis, the average age of marriage could rise to 30
(Wrigley 1966) and the proportionof a cohort never-marryingto above 20%
(Goldstone 1986). One major consequence of the Europeanmarriagepattern
was that pretransitionfertility in Europe was at moderate levels relative to
"high fertility"levels elsewhere.
Traditionallevels of fertilityin non-Europeansocieties were higher (approximately 6-8 births per woman compared to 4-5 births per woman in many
pretransitionEuropeanpopulations),but they were also socially regulatedby
"intermediatevariables"other than age at marriageand the proportionevermarried, for instance, spousal separation, breastfeeding, sexual abstinence,
nonmarriageafter widowhood, and so forth. In an interesting illustrationof
the differing mechanisms of fertility regulation, Kumar (1971) showed that
maritalfertility was higher in nineteenth-centurySweden and Finland than in
twentieth century India. It is unclear if fertility fluctuated in response to
economic conditions (wages, harvests) in premodernnon-Europeansocieties
as it did (via marriage)in Europe.
The first "modem" fertility transitionsbegan in early nineteenth century
France and the United States (Coale & Treadway 1986:37, Tolnay & Guest
1984, Sanderson1987, Haines 1989, Bean et al 1990, Gutmann& Fliess 1993).
The rest of Europe followed suit, about a half-centurylater, in the last three
decades of the nineteenthcentury. These trends, which consisted almost entirely of declines in maritalfertility,were largely completedby the 1930s. The
patternsare extensively documentedin the many publicationsof the Princeton
EuropeanFertilityProject(summarizedby Coale & Watkins 1986). If the onset
of a maritalfertility transitionis markedby a decline of 10% from the pretransition plateau, the descent to about 50% of the pretransitionlevel was
largely completed in 30 years (Coale & Treadway 1986:40-41). In contrastto
the fluctuationsin fertilityin earliercenturies,these modem fertilitytransitions
were permanentreductions.
There were "baby booms" with a modest resurgence in fertility in some

210

HIRSCHMAN

Western societies duringthe 1950s (the United States, in particular),but these

were temporaryand did not representa returnto pretransitionfertility levels.
Over the last 20 years, fertility in most industrialcountries has remained at
the lowest levels ever recorded-well below the replacementlevel (Davis et
al 1986). Japan,the most modernnon-Westernsociety, followed the European
patternfairly well. Japan's modern demographictransitionbegan in the late
nineteenth century but was interruptedby World War II and its immediate
aftermath(Mosk 1979, Hanley 1979). The rapiddecline to replacement-level
fertility in Japan in the 1950s is a continuationof trends firmly established
earlier in the century.
In the 1960s, fertility began to decline in some developing countries. Initially, these declines were limited to a small numberof rapidly modernizing
societies in East Asia, and it was unclear whether these trends would spread
to the largerandpoorercountriesin the ThirdWorld.In the last decade modern
fertility transitionshave begun in most countriesin East Asia, SoutheastAsia,
South Asia, and Latin America (McNicoll 1992, Palloni 1990, Feeney et al
1989, Robey et al 1992, Freedman & Blanc 1992, Leete & Alam 1993,
Hirschman & Guest 1990). There are still many areas with high fertility in
South Asia and sub-SaharanAfrica, but recent evidence shows that fertility is
beginning to decline in many of these countries as well (Menken & Phillips
1990, Caldwell et al 1988, 1992, Robinson 1992, Brass & Jolly 1993, Rutenberg & Diamond 1993).
There has been some concern that the rate of fertility decline has slowed
down in some developing countries in the last decade-stalling at a plateau
of moderatefertility in the range of three or four birthsper woman (Bongaarts
1987, Horiuchi 1992). Freedman& Blanc conclude, however, that the slowdown is largely concentratedin East Asia where levels are reaching the replacement floor (1992:45). There is evidence of below-replacementfertility
emergingin a numberof developing countriesincludingTaiwan, South Korea,
Singapore,and Thailand(Rele & Alam 1993, Hirschmanet al 1994).
At present, a crude threefold division of the world fertility regimes might
be: (i) industrialsocieties, that have experienced long-term fertility declines
beginning in the nineteenthor early twentiethcentury and that currentlyhave
fertility at or below the replacementlevel, (ii) developing societies, that have
experiencedsignificantfertilitydeclines over the last 10 to 25 years and where
currenttotal fertility rates are between 2.5 and 4 births per woman, and (iii)
less developed countries, that have yet to experience significant fertility reductions and where average childbearinglevels exceed 5 births per woman.
As soon as such a typology is presented, there are qualifications about the
homogeneity of the categories and the placement of particularcountries or
regions in these categories. A number of developing countries have entered
(or shortly will enter) the first category of below-replacementlevel fertility.

CHANGES 211
WHYFERTILITY
And more andmore countriesaremoving fromthe thirdto the second category.
Any cross-sectional portraitof a rapidly changing historical process is out of
date as soon as it is published and is likely to be a poor image of the future.

THEORIESAND MODELS OF FERTILITYCHANGE

Interpretationsof modem fertilitydeclines rangedwidely in the late nineteenth
and early twentieth centuries, mixing social explanations with speculations
about the impact of industriallife on biological capacities to reproduceand
on declining sexual motivations.Demographers,most notablyWarrenThompson (1929), emphasized the social and economic forces of modem societies
as the basic causes of lowered fertility.These ideas crystallizedover the second
quarterof the twentieth century and emerged in the 1940s as Demographic
Transition Theory (DTT), largely under the pen of Frank Notestein, as a
synthesisof ideas andempiricalcorrelationsbasedon the historicalexperiences
of Western Europe and North America (Notestein 1945, 1953).

Demographic TransitionTheory
Notestein emphasized the changing institutional fabric of urban industrial
society that led to the "emergenceof a new ideal in matters of family size"
(1953:16). Among the motivating factors of modem society were: "reduced
... pressurestoward traditionalbehavior,""educationand a rationalpoint of
view," "the cost of child-rearinggrew and ... economic contributions by
childrendeclined,"and that "women ... found new independencefrom household obligations and new economic roles less compatible with child-rearing"
(Notestein 1953:16). The means for fertility control was contraceptiveuse by
marriedcouples-which had "been widely used for centuries throughoutthe
world,"but was "notwidely used untilthe incentive for birthrestrictionbecame
strong"in industrialsociety (Notestein 1953:16-17).
The corpusof transitiontheory was very broad.Within the same theoretical
tent, Kingsley Davis could minimize the role of religious and culturalvalues
as primarydeterminantsof fertility while Ronald Freedmanwas pointing to
the centralrole of normsfor family size (Davis 1955, Freedman1963). Critics
of the 1980s and 1990s who charge demographictransitiontheory with being
narrowlyeconomistic or with ignoringthe role of ideas have missed the central
point: that transitiontheory had room for every causal variable.
The two modem seminal contributionsin the development of demographic
transitiontheory (and the most widely cited) are Kingsley Davis's 'Theory of
Change and Response in Modem Demographic History" (Davis 1963) and
Ansley Coale's 1973 essay on the demographic transition, which reflected
many of the findings from the Princeton EuropeanFertility Project (Coale
1973). It is hard to imagine two articles more different, yet the field has

212

HIRSCHMAN

acceptedboth as partof a common theoreticalstreamand not as contradictory

theses.
Davis (1963) put the conscious use of contraceptionby marriedcouples as
only one of many possible responses in his "multiphasictheory" of demographicchange.The most importantindependentvariableis the level of household economic strain, which is a function of household size and potential
economic resources. High levels of household economic strain were the motivating factor that caused individuals (and families) to postpone marriage,
never marry, migrate, use abortion,practice infanticide, and restrict marital
fertility by contraception.All of these were mechanisms to maintain (or improve) economic welfare. Davis arguesthatmost societies will use all of these
methods (in varyingproportions)in responseto the populationpressure(manifest as economic strain) caused by mortalitydeclines and the economic opportunitiesthataccompanymodernization.Absolute poverty does not provide
a context for demographicresponses, accordingto Davis; it is the possibility
of betterment(or of holding onto recentimprovements)duringthe modernization process that provides the centralmotivation.
Although Davis's article is widely cited in the demographictransitionliterature,the body of researchthat tests hypotheses from his theory is relatively
modest (Friedlander 1969, 1983, Mosher 1980a, 1980b). One problem of
testing Davis's theoryis the identificationandmeasurementof economic strain
that is a function of potential household size and economic aspirations. If
economic welfare rises fasterthan aspirations,therewould not be any pressure
for changes in demographicbehaviors.Modernizationmay have directimpacts
on the motivation for fertility (and other demographicbehaviors) independently of changes in economic straincaused by declining mortality.
In contrastto Davis's broadeningof the range of demographicresponses,
Coale (1973) focused on the reasons for declines in maritalfertility alone. He
identified three necessary conditions for the decline of marital fertility: (i) a
setting thatallowed for fertilityplanningto be partof the calculus of conscious
choice, (ii) the availabilityof effective informationabout the means to control
fertility, and (iii) clear economic advantages of fertility control. The third
conditionis the standardlink to the traditionalthesis of demographictransition
theory (DTT) that socioeconomic factors change the incentives for childbearing. The first two conditionsareones thatdemographershad prettymuch taken
for granted-that societies adjusteddemographicbehaviorwhen circumstances
warrantedand that some knowledge about fertility control was available in
most historical societies.
The three preconditionsin Coale's frameworkfit well with Henry's (1961)
thesis thatnaturalfertilitywas practicedin pretransitionsocieties. Because the
last condition was already well known, most subsequent attention has been
focused on the observationthat natural-fertilitysocieties lacked legitimating

CHANGES 213
WHYFERTILITY
aboutthemeansto control
culturalvaluesforfertilitycontrolandinformation
were
thatfolkmethodsof contraception
fertility(althoughCoaleacknowledges
see Coale 1979:15).The widelycited
populations;
usedin somepretransition
evidencethatmanywomenin traditionalsocietiescannotanswerquestions
aboutdesiredfamily size (lackingnumeracyaboutfamily size; see van de
Walle1992)illustratedtheculturalrootsof highfertility.Coale'sformulation
representeda shift awayfromthe centralthemeof DemographicTransition
Theory,thatchangesin socioeconomicinstitutionsaretheprimaryprecursors
to fertilitydecline.Theinitialchallengeto DTTwas,however,nottheoretical
evidence,Coale'scategorieswouldprobbutempirical.Withoutdisconfirming
ably have foundtheirplaceunderthe broadumbrellaof DTT, wheremany
variantideaswereoftenexpressedwithlittleinternalconflict.
Beginningin the 1970s and throughoutthe 1980s, empiricalstudiesof
societiesreportedfindingsthat
fertilityof bothhistoricaland contemporary
wereat odds with the expectedassociationsof socioeconomicvariablesand
fertility.The mostfamouswas an articleby Knodel& van de Walle (1979)
the findingsof thePrincetonEuropeanFertilityProjectwith
thatsummarized
in less developed
a discussionof theimplicationsforfuturefertilitytransitions
countries(the articlewas laterpublishedas chapter10 in Coale& Watkins
1986).Knodel& vande Wallereportedthatfertilitydeclinesbeganaboutthe
countriesthatwereatquitedifferentlevels
sametimein a numberof European
culturalsettinganddiffusion
Theyemphasized
of socioeconomicdevelopment.
as the criticalelementsto explainthe spreadof fertilitycontrolin Europe.In
FertilityProject(EFP),Watsummaries
of thePrincetonEuropean
subsequent
kins (1986, 1987)also concludedthatthe resultsdisconfirmedthe empirical
transitiontheory.
predictionsof standarddemographic
Aboutthe same time, resultson fertilitychangefrom the contemporary
countrieswerebeingpublishedfromtheWorldFertilitySurvey
less-developed
(WFS)program(Cleland& Hobcraft1985, Cleland& Scott 1987).Cleland
WFS analysesof maritalfertility
the resultsof the comparative
summarized
as representinga clear refutationof the "demand"(demographictransition
style) theories(Cleland1985,Cleland& Wilson1987).Few of the expected
associationsbetweensocioeconomicvariablesandfertility(withtheexception
of femaleeducation)predictedby DTTwereconsistentlyfoundin the comparativeanalysesof theWFS.
This doublewhammy(fromthe EFP and the WFS) createda crisis for
demographictransitiontheory.It is now the conventionalwisdomin many
transition
andbeyond,thatdemographic
circles,insidethefieldof demography
theoryis neardeath.Intheoverviewtheoreticalchapterin a bookof historical
analysesof Europeanfertilitydeclines,Alterstatesthat"Inthe last 20 years,
transition,has
the-field's dominantconsensus,the theoryof the demographic
beendramatically
shattered"
(1992: 13). Theories,however,rarelypass from

214

HIRSCHMAN

thatprovidesa cleareraccount
the sceneunlessthereis a superioralternative
of researchin the field. The dilemmais that thereis no consensuson an
transitiontheory.A theoreticalvactheoryto replacedemographic
alternative
uumis nottolerableforanempiricalsciencewitha continuousneedfortestable
hypotheses.So thedebatescontinuewitha plethoraof contendingtheoretical
noneof whichhas gainedwide adherence.
frameworks,

Caldwell's Theoryof IntergenerationalWealthFlows

In a seriesof influentialarticlesandbooks,JackCaldwell(1976, 1980, 1982,
Caldwellet al 1988) criticizedthe economisticbiases of traditionaldemotheoryandofferedhis revisionisttheoryof intergenerational
graphictransition
wealthflows.InCaldwell'stheorythereareonlytwostablefertilityregimessocietieswherefertilityis as highas possibleandmodernsocieties
traditional
societies,childrenprovidea positive
is low. In traditional
wherechildbearing
ladderto
up the generational
net flow of resources,services,andstatus-honor
In stablehigh-fertilitysocieties,thereis
parents,especiallyto the patriarch.
alwaysan incentivefor additionalchildren.In modernsocieties,thereis a
time,money,services,and
reversalof the wealthflow, andparentscontribute
supportdownwardto childrenwith minimalexpectationsof any return.Accordingto Caldwell,therearefew economicincentivesfor fertilityin modern
societies.
child-centered
regimesarenota simple
of high-fertility
Thereasonsforthedestabilization
matterin Caldwell'stheoreticalwritings.In some places,he emphasizesthe
familyvia
diffusionof Westernculturalmodelsof the nuclearchild-centered
andthemassmedia.In othercontexts,Caldwell
international
communications
has stressedthe impactof "massschooling"in a community,whichtips the
balanceto loweredfertility.His theoryhas been interpretedas supporting
andculturaltheories.
aspectsof bothstructural
Theprincipalobstaclefor manyscholarsin the field hasbeenthe lackof a
clear model for conductingempiricaltests of hypothesesfrom Caldwell's
theory.Caldwell'sinsistencethatdemographymust shift its focus to more
investigations("micro"approaches)has not clarifiedthe task.
ethnographic
wealthflows (or perceptionsof the flows) over
Measuringintergenerational
problemsinclude
historicaltime is a difficultchallenge.The methodological
componentsof wealthflows, such as
of the nonmonetary
the measurement
andundefinedobligations.Beyond
services,deference,emotionalgratification
Caldwell'sown research,thereare relativelyfew empiricalstudiesof the
wealthflow hypothesis.
intergenerational

Economic Theories of Fertility

to the studyof fertilitychangeand
Therearetwo majoreconomicapproaches
many variants(Sanderson1976). The first is the "new home economics,"

CHANGES 215
WHYFERTILITY
theoryto familyissues,including
whichis the applicationof microeconomic
fertility(Becker1960, 1988, Schultz1981).The secondis the synthesisof
economicand sociologicaltheoriesof fertilitypresentedin the supplyand
demandframeworkof RichardEasterlin.
The firstapplicationsof microeconomictheoryto fertilitybehaviorwere
of consumerchoicetheorywithlittleacknowlrathermechanisticillustrations
edgmentof thesignificantdifferencesbetweentheacquisitionof anautomobile
anda baby.Theseearlyapplicationsof economictheoryto fertilityinspired
JudithBlake'sdevastatingcritique(Blake1968,also see Turchi1975).Over
the years,however,economistshavebecomemoresophisticatedin theirapplicationof economictheorytohouseholdbehavior,andtheyhavealsobecome
of economicbehaviorin theirempirical
morerigorousin the measurement
analyses.Ideasandhypotheseshavealso graduallydriftedacrossdisciplinary
costs of women's
researchthatincludesthe opportunity
lines. Demographic
behavior
timeas well as theimpactof thepricesandincomeson demographic
Thereis someevidencein support
is nowmuchmorecommonin theliterature.
& Roumasset1991),but the approach
of economichypotheses(Hutaserani
remainstoo narrowto be a significanttheoreticalchallengeto demographic
transitiontheory.As Robinson(1992:453)notes,"theproposition(microeconomictheoryof fertility)has not beenproven,only assertedoftenenoughto
gaina certaincredibilityandforcethroughrepetition."
In severalessaysanda majorbook,RichardEasterlinhas madea serious
sociologicalresearchon
efforttojoin economictheorieswithmoretraditional
fertility(Easterlin1969, 1978b, 1983, Easterlin& Crimmins1985). Two
essays on fertilityin developingcountries,sponvolumesof state-of-the-art
soredby theNationalAcademyof Sciences,wereorganizedaroundEasterlin's
of fertility:demand,supply,and
frameworkof the threebasic determinants
thecostsof fertilityregulation(Bulatao& Lee 1983).Demandfactorsinclude
of fertilityfrommodernization
(desocioeconomicdeterminants
the standard
and cultural
mographictransition)theory.Supplyfactorsare environmental
factorsthatconstrainnaturalfertility.Thecosts of fertilityregulationinclude
the monetary,time,andpsychicfactorsassociatedwiththe use of contracepcan
tion.In a seriesof ingeniousgraphs,Easterlinshowshow modernization
leadfirstto a riseandthento a fall in fertilityas theforcesof supply,demand,
andregulationcosts combineto shapefertilitybehavior(1983:566-574).
Theinclusivenessof Easterlin'smodelhas clarifiedsomeempiricalanomaliesin the studyof fertilitytrends.Nevertheless,his modelsharestwo limiFirst,the assumptionof naturalfertility
tationsof priorworkin the literature.
maritalfertility(andmarital
meansthatthe wide variationsin pretransition
behavior)are outsidethe scope of the theory.Second,thereis no effortto
resolvethe theoreticaland empiricalproblemsin the specificationof what
socioeconomicvariablesaccountfor demand.In general,demandfor fertility

216

HIRSCHMAN

does declinewithmodernization,
butthe questionof the manyweakassociationsbetweenthestandard
predictorvariablesandfertilityremainsunresolved.

Ideational Theory
Withinthe dominantbranchesof Americandemography,
therehas been persistentskepticismthatvalues,attitudes,andotherpsychologicalorientations
can explainfertilitytrendsor variationsbetweenpopulations.It is, therefore,
perhapsappropriatethat the new interpretation
of fertilitylinked to deep
culturalvariables,knownas ideationaltheory,was developedprimarilyby
Europeandemographers,
especiallyRonLesthaeghe(Lesthaeghe1980, 1983,
Lesthaeghe& Surkyn1988).Muchof the impetusfor ideationaltheoryhas
been the allegedfailureof demand(socioeconomic)theories(Cleland1985,
Cleland& Wilson 1987). Culturespansa wide varietyof phenomena,and
there are quite variedmeaningsof the term in the demographicliterature
(Hammel1990,Pollak& Watkins1993).Thisgivesriseto notone,butseveral
theoriesof cultureandfertility.
The strongculturalhypothesisis that groupsdiffer in fertilitybehavior
becauseof culturalvalues.Somepopulations
mayhavehigherlevelsof fertility
thanothergroupswithequivalentsocioeconomiccharacteristics
becausetheir
cultureplaces a highervalue on childrenor proscribescertainmethodsof
fertilitycontrol.Lesthaeghe& Surkyn(1988) arguethathistoricalvariations
in Europeanfertilityare closely relatedto nationaldifferencesin religious
andsecularism.Evenif this argumentis accepted,the
beliefs,individualism,
questionof the originsof culturalvaluesremainsunresolved(Preston1986:
186-189).Davis(1963)dismissedall culturaltheoriesbecauseof theinherent
of thebasiclogic, thatis, behavioralpatternsareexplainedin terms
circularity
of culturalpreferencesforthatbehavior.To avoidthisfault,culturalvariables
needto be explainedin termsof structural
conditionsorhistoricalexperiences.
Traditionalvalues are typicallyrootedin ruralenvironmentsand among
recentmigrantsto urbanareas.Culturalvalues,however,may persistlong
afterthestructural
conditionsin whichtheyoriginated
haveeroded.Lesthaeghe
& Surkyn(1988)arguethattherisein fertilityafterWorldWarII (babyboom)
maybe explainedby the orientations
of generationsthatwererearedin traditionalprewarenvironments(the very low prewarfertilitywas a productof
economicconstraints).
Accordingto thisargument,the changein valuesafter
WorldWarII towardgreatermaterialism
andindividualism
led to thelowered
fertilityin the 1960sand 1970s.
A weakerversionof the culturalinterpretation
positsthatideas aboutappropriatefamilysize andmethodsof birthcontrolcan diffusemorequickly
withinculturallyhomogeneouspopulations(Retherford1979, Retherford&
Palmore1983).Thecriticalassumption
is thatthelagperiodbetweenstructural
changesand demographic
responsescan be shortenedor lengthenedby in-

WHYFERTILITY
CHANGES

217

tragroupcommunicationabout fertility ideals, the legitimacy of fertility control, and techniques of birth control and abortion. Interpretationsbased on
studies of the Europeandemographictransitionsuggest that the cultural diffusion of ideas (and knowledge about the practice) of birth control can even
precede the structuralchanges in society. The result is that patternsof fertility
decline are more likely to mirrorgeographic maps of ethnic groups than of
socioeconomic change (Watkins 1986, 1987).
The empirical base for ideational theory seems to rest more on a negative
case for the alternativemodel-the rejectionof socioeconomic explanationsthan on positive evidence (Cleland & Wilson 1987). The claim that fertility
declined at about the same time in a variety of socioeconomic settings does
not mean thatcultureis the primemover (Mason 1992). It is clear thatdiffusion
is an importantprocess in the explanation of fertility, but the links between
cultureand diffusion have yet to be clearly articulatedand empirically tested.

UNRESOLVED ISSUES IN RESEARCHON FERTILITY

The absence of a hegemonic theory of fertility does not mean that there is not
a great deal of knowledge and many empirical generalizationsabout the determinantsof fertility in various settings. The existing base of knowledge,
however, cannotbe summedup in one or two global statements.In this section,
I review some of the majorissues thatrequirefurtherempiricalunderstanding
(and perhapssome new insights) before it will be possible to frame a unified
theory of fertility transitions.

Marriage, Family Structure,and Fertility

In most contexts, human fertility occurs in family units. In fact, the standard
definition and functional explanation of familial institutions are invariably
presentedin terms of childbearingand childrearing.Although there are close
connectionsbetween family structureand fertility, the relationshipis not simple. Prior researchhas examined many strandsof the relationship,including
the timing and prevalenceof marriage,the complexity of household structure,
residence patternsafter marriage,and inheritancepatterns.There seem to be
few universalrelationships.
The idea is sometimes expressed that age at marriagedoes not necessarily
have a strong influence on fertility because there is generally sufficient time
to "makeup" for any delay in childbearingfollowing maritalpostponement.
Although this is logically true for individuals, aggregate patternsalmost invariablyshow that later marriagemeans lower overall fertility (Smith 1983).
Indeed, the principalmechanisms of fertility regulationin premodernEurope
were variationsin marriagetiming and the prevalence of marriage(Coale &
Treadway1986:47-48). The partialbreakdownof the system of maritalpost-

218

HIRSCHMAN

to a risein fertilityin theearlystages

ponementis thoughtto havecontributed
of modernization
(Tilly 1978,Goldstone1986).
mechanismof fertilitycontrolin Europe,and
If marriagewasthetraditional
the modem Europeanfertilitytransitionsfrom 1870 to 1930 were largely
confinedto declinesin maritalfertility,thena reasonablehypothesismightbe
that changesin marriagepatternsdo not covary with changesin marital
fertility.Indeed,thislogic hasbeenimplicitin muchof the historicaldemogTherefore,Coale's(1992)recentpaperwhichshowsa close
raphyliterature.
associationof delayedmarriageand the adventof fertilitycontrolwithin
populations)raisesa number
marriage(forbothhistoricalandcontemporary
of significantquestions.Coalereasonsthatculturalcontextsthatleadto later
marriagearealso favorablefor femaleautonomyandcontroloverreproduction.Thefindingis also consistentwithDavis's(1963)multiphasictheoryof
response.
demographic
In manydevelopingcountries,therehavebeentrendstowarddelayedmarriage(Smith1980)andalso declinesin maritalfertility.Althoughthe largest
componentof Asian fertilitydeclineshas been declinesin maritalfertility
(Retherford& Cho 1973, Hirschman& Guest 1990), changesin marriage
patternshave also been significantin some cases. The negativeimpactof
marriagetimingon fertilityis confounded,however,by the fact thatmarital
often leadsto a shorterfirstbirthinterval(Hirschman1985).
postponement
This apparentanomalyhas been explainedby Rindfuss& Morgan(1983) as
a resultof a trendtowardromanticmarriagesandincreasedcoitalfrequency
Theevidenceis largelyindirect,butthestrengthof thelogic
earlyin marriage.
andreplicationof thebasicfindingssupportRindfuss& Morgan'sinterpretation(Fricke& Teachman1993).Althoughthetrendtowardmaritalpostponementmayleadto shorterbirthintervalsin the shortrun,the long-runexpectationis loweredmaritalfertility.
Anotheremergingpatternis a loweredprevalenceof marriagein societies
thatformerlyhaduniversalmarriage.Recentdatashowdouble-digitpercentagesof womenin theirearlythirtieswhohavenotyet marriedin severalAsian
countries(Limet al 1987,Xenos& Gultiano1992,Guest& Tan 1993).This
will have significantimplications
increasein the numbersof never-married
for futurelevels of fertilityandothersocialpatterns.
of the impactof familystructureon fertilitywas
The classicinterpretation
presentedalmost40 yearsago by KingsleyDavis(1955).He notedthe strong
associationof extendedfamilyhouseholdswithhigherfertilityin less developedareasandarguedthatthe extendedfamilystructureprovidedextensive
supportsfor early marriageand childbearing.Marriagedid not have to be
householdcouldbe set up. The costs of chilpostponeduntilan independent
drenand the burdenof child care are subsidizedby othermembersof the
household,not borneexclusivelyby the parents.In extendedfamilyhouse-

WHYFERTILITY
CHANGES 219
holds,theyoungmarriedcouple,especiallythebride,acquiresrecognitionand
statuswiththe productionof children.Thisthesisconformsto the traditional
contrastof the nuclearfamilysystemin manyindustrialcountrieswith the
traditional
patternof younganduniversalmarriagein otherpartsof theworld
(Hajnal1965,Dixon1971).Buttheremaybe moreflexibilitybetweenfamily
societies.
structureandfertilityin modemindustrializing
RonaldFreedmanandhis colleagueshavebeenmonitoringchangesin the
structureof Chinesefamiliesandfertilityin Taiwanfor severaldecades(for
the latestreport,see Weinsteinet al 1990).Initiallyit was assumedthatthere
wouldbe a close relationshipbetweenextendedfamilyliving arrangements
andfertility-bothdecliningwithmodernization.
TheTaiwanesefertilitytransitionis nowcompletewitha below-replacement
fertilityrate,butco-residence
of marriedcoupleswith husband'sparentsremainsan importantaspectof
familystructurein Taiwan.Aboutone half of adultslive in extendedfamily
units,andmostcoupleslive withthehusband'sparentsfor at leastsometime
aftermarriage.Therehas been some decreasein joint familyliving arrangementsin Taiwan,butthe continuityis mostimpressive.
inTaiwan,
Thepersistenceof traditional
Chinesefamilylivingarrangements
however,has not meantthatfertilityhas remainedhigh.It maybe thattraditional family structuresin traditionalsocieties result in high fertility,but
traditionalfamily formscan accommodatenew content,includinglowered
fertility,in modernizingsocieties.Axinn(1992) reportsthatexposureto extrafamilialactivitiesby husbandsandwives leadsto increasedcontraceptive
in developingcountriesarechanging
useinNepalesesociety.Familystructures
rapidly,as theyarein industrialcountries,andfuturepatternsareunlikelyto
(Thorton& Fricke
fit intoneatcategoriesof traditional
andmodemstructures
1987).Enormousmethodological
problemsconfrontresearchon the dynamic
relationshipsbetween family structureand reproductivebehavior(Burch
1983). Furtherprogressmay requirelongitudinalstudiesthat trackfamily
thatareknitmoreby economic
interactions
andexchangesacrossgenerations
andsocialobligationsthanby commonresidence.

Modernizationand Fertility
Theconceptof modernization
hasundergoneevenmoretrialsthanthe theory
of the demographictransition.Even if synonyms(socioeconomicchange,
canbe substituted
as temporary
remedies,the conceptualcrisis
development)
remains-we do not have a preciseaccountof the social, economic,and
culturalforces that are necessaryor sufficientconditionsto transformlow
income,ruralagrariansocietiesinto high income,urbanindustrialones. As
has
withdemographic
theory,the searchfor a grandtheoryof modernization
oftengivenway to a searchfor empiricalpatternsthatmayvarywithspecific
institutional
andhistoricalcircumstances.

220

HIRSCHMAN

it is
Giventhatit is difficultto specifythe componentsof modernization,
andfertilitychange(denot too surprisingthatthe theoryof modernization
problems.Nevertheless,conclumographictransitiontheory)hasencountered
becausefertilitymay be
sions that socioeconomicforces are unimportant
variablesare premature(Cleland&
weaklycorrelatedwith "modernization"
theory,in my opinion,
transition
Wilson1987).Thebasicflawin demographic
is the assumptionthatthereis a single monolithicpatternof modernization
thatcouldbe indexedby any socioeconomicvariable.This assumptionmade
it possibleto use weakassociations(andoccasionalnegativeevidence)as the
andfertility
basisforclaimsthatalleffortsto developa theoryof modernization
challengeis to specifymoreclearlywhataspects
arefutile.Therealtheoretical
arelinkedto fertilitychange.
of modernization
Theweaknessof someconventionalresearchwithinthe DTTtraditioncan
betweenferbe illustratedby consideringthe expectednegativerelationship
It
factorssuchas incomeor industrialization.
tilityandgeneralmodernization
is truethatbothhigherincomesandgreaterindustrialemploymentarecentral
forcesthathavetransformed
elementsof thebroadcomplexof modernization
the worldover the last century.This does not meanthathigherincomeor
industrialemployment,by themselves,will motivatefamiliesto have fewer
children.Infact,thereverseis equallyplausible.Themostdirectconsequences
of an increasein incomeare higherlevels of consumption.If childrenare
highlyvaluedin a society,economictheorywouldpredicta higherdemand
as to the effect on
for children.Economictheoryis, however,indeterminate
childrenwithgreaterlevelsof investment
fertilitybecausequality(higher-cost
andconsumption)andquantityaresubstitutes.Higherlevels of incomemay
also loosen constraintsto higherfertility,for example,less absenceof the
husbandin searchof employment,lowerlaborforceactivityby the wife, less
mayleadto younger
employment
In a similarfashion,industrial
breastfeeding.
constraintson familyforage at marriage(andhigherfertility)as traditional
mationareeased(Goldstone1986,Haines1979).
Therearefrequentfindingsof a rise in fertilitybeforethe transitionto low
fertilitybegins(Dyson& Murphy1985).Easterlinexplainsthis patternas a
reducedsterility,early
factors(lessbreastfeeding,
productof a risein "supply"
marriage,etc) in societieswheredemandfor fertilityexceedsactualfertility
is a specification
levels(1983:574).Whatis missingin Easterlin'sexplanation
leadto anincreasein supplyandwhatfactors
of whataspectsof modernization
will leadto a decreasein demandin fertility.Davis (1963)arguedthatit was
necessaryto linkthebroadermacrolevelforceswiththemicrolevelmotivations
for loweredfamilysize (withdeclinesin fertilitybeingone of severalmechanismsto this end). In Davis's theory,the criticalmacrolevelfactor was
decliningmortality,whichled to householdeconomicstrainin the contextof
Thelogic of Davis'sarguandnew economicopportunities.
risingaspirations

WHY FERTILITYCHANGES

221

mentis compelling-the expectationis not thatany and all modernization

variableswill correlatehighly with fertilitydecline,only those aspectsof
modernization
thatcreatedirectincentivesfor smallerfamilies.
Freedman(1979)hintedat a new approachin his classicassessmentof the
stateof theoryandresearchon fertility.Afterreviewingrecentresearchthat
showedthe weak empiricalevidencefor traditionaldemographictransition
theory,Freedmannoted that therewere a varietyof sufficientconditions,
exemplifiedby differentcountries(orregionswithincountries),thatcouldlead
of variables
to modernfertilitytransitions.
Certainvariables,or combinations
higherstatusof women,higherlevels of
(effectivefamilyplanningprograms,
socialwelfare,extremepopulationpressure,a strongstate,etc), couldleadto
loweredlevels of fertilityeven in the absenceof othervariablesconsidered
subsequenttheessentialprerequisites
for fertilitytransitions.
Unfortunately,
analysisbeyondthe
oreticalwritingshavenotattempted
to extendFreedman's
discussionof historicalexamplesto thespecificationof hypothesesthatrelate
variables)to fertilitydeclinesin varioussettings.
criticalfactors(modernization
Perhapsthemosteffectivestrategyforthe specificationof a broadertheory
of modernization
andfertilitytheorymightbe to identifythemicrolevelfamily
motivationsfor fewer childrenand then to work outwardto the relevant
forces.The core idea of demographic
transitiontheoryis thatthe
structural
costs andbenefitsof children,bothshort-term
andlong-term,shapemotivationsforchildbearing
(Caldwell1983).All othercausalfactorsmaybe derived
linksto infantandchild
fromthisbasicpremise.Mostobviousarethestandard
mortality(assumingthatthereis a thresholdof family size wherebenefits
decreaseand/orcosts rise),changesin the familyeconomy(less dependence
on childlabor),andrisinglevelsof education(increasingthecostof childrearbutis
ing).Thecost of childrenis notlimitedto directmonetaryexpenditures
If thereis a
also evidentin the timecommitments
necessaryfor childrearing.
high opportunitycost of the mother'stime or an absenceof otherfamily
childcareprovidersorboth,thecostof childrearing
is muchhigher.Extensions
of this basic logic can be appliedto othersocioeconomicvariables,but a
crediblehypothesismustspecifythecausallinksin termsof theincentivesfor
Caremustbe takento considerthebroadersocialandinstitutional
childbearing.
contextthatmay confoundexpectedbivariaterelationships.
Theremay also
be important
thresholds
orflooreffects;forexample,oneortwochildrenmight
be desiredeven if the costs of childrenareveryhigh.
Theseissues can be illustratedwith recentliteratureon the slow-downor
stall in the fertilitytransitionof the Malaypopulationin Malaysia.Rapid
in Malaysiahas broughtverylow fertilityto the Chineseand
modernization
Indianpopulationsthere(almostto the replacementlevel), whereasMalay
fertility,afterexperiencinga modestdeclineduringthe 1970s,has remained
at a plateauof over fourbirthsper womanduringthe 1980s. Leete & Tan

222

HIRSCHMAN

(1993) interpretthese differentialsas evidencethatdemographictransition

theoryis inadequateandthatculturalvariables,mostnotablyIslamicfundamentalism,arenecessaryto explainwhyMalayfertilityhasnotbeenreduced.
It is difficultto evaluateposthoc explanations,
butthelackof similarresponse
amongMalaypopulationsin Singaporeand Indonesiacreatesdoubtsthata
simpleculturalexplanation
is sufficient(Jones1990).Government
policiesof
Malaypreference(affirmativeactionprogramsprovidemoreeducationand
employmentopportunities
forMalaysthanforMalaysianChineseandMalaysianIndians)also seemto be correlatedwithdivergingethnicfertilitydifferentialsin Malaysia,althoughit is difficultto specifythe precisemechanisms
& DaVanzo1992).It is notpossibleto evaluatetheseconflict(Govindasamy
ing claimsempirically,butrival"demographic
transition"
hypothesesarealso
consistentwiththe reportedtrends.MalaysianMalayparentshavebeen shelteredfromthe risingcosts of childbearing
by governmentsubsidiesfor education(includingcollege)andby a greateravailabilityof low-costchildcare
(Hirschman1986). Althoughthese ideas have not been empiricallyproven,
theyillustratethepotentialimportance
of somesocioeconomicvariables(and
not others)as determinants
of fertilitylevels andchange.

Culture,Diffusion, and Fertility

A superficiallook at the recentliterature
on fertilitymaygive the impression
of a battleground
betweenstructural
andculturalinterpretations.
Thisis, however, a mock battleplayedout by a few cases of extravagantclaims and
counter-claims
for the benefitof graduatestudentswho are seekingsimple
conceptualschemato organizethe literature(see Mason1992 for a critique
of someof therhetoricalsleight-of-hand
in Cleland& Wilson1987).Not only
is the evidencemixed,but mostresearchers
acknowledgethe interdependent
role of social structural
conditionsthatmotivatebehaviorandthe spreadof
ideasandinformation
thatreinforcebehavioralchange(Carlsson1966,Friedlanderet al 1991,Gilliset al 1992,Woods1987).Moststructural
hypotheses
areperfectlycompatiblewithconsciousdecision-making
as a potentialinterveningprocess.Theoriesof normativeinfluenceson behaviorrarelydenythat
deepculturalvaluesor innovativeideologiesare rootedin historicalexperiences or materialinterests.Uncertaintiesarisingfrom culturallags, weak
andthe generallackof comprehensive
datahaveled to differing
correlations,
emphasesstressingcertainvariablesas moreimportant,but the numberof
who subscribeto completelymonolithicexplanationsis very
demographers
small.
Indeed,thereis oftenagreementon whichvariablesaremostimportant,
but
of whatthevariablemeans.Oneof themostconsistent
differinginterpretations
findingsin theliterature
is a negativerelationship
betweenwomen'seducation
andfertility,bothat the individualandthe aggregatelevel (Cochrane1979,

CHANGES 223
WHYFERTILUTY
1983,Cleland& Rodriguez1988).Thereareexceptions,butthe relationship
is one of the most robustin the empiricalliterature.Althoughthe general
is thateducationis a socioeconomicvariablethatraisesthecost
interpretation
of children(directlyandindirectlybecauseof thevalueof themother'stime),
Cleland& Rodriquez(1988)arguethattheinfluenceof educationis primarily
ideationalbecausemostof the impactof femaleeducationis not mediatedby
employment.The same argumentcould be developedfor othersignificant
of fertilitywithoutanyclearresolution.Differencesof interpredeterminants
tationare not entirelyemptyrhetoric,but convincingargumentsshouldgo
beyondsimplyclaiminga variablefor one campratherthananother.
Two recentstudiesillustratethe complexityof socialcontextandthe role
of socioculturalprocesseson humanfertility.In one of the most important
FertilityProject,Lesthaeghe&Wilson(1986)found
studiesfromtheEuropean
predictorof thepaceof fertilitydecline
was animportant
that"secularization"
in a numberof Europeancountries.In an imaginativeanalysis,Lesthaeghe&
Wilsonshowedthatboth socioeconomicstructure(the mode of production,
(vote for socialistor
indexedby laborforcecomposition)andsecularization
forcesthatled to morerapiddeclinesin
nonreligiousparties)wereimportant
maritalfertility.Theauthorsreasonedthatthepoliticalandculturalforcesthat
religiouspartieswerepartof a larger
led to a declinein supportfortraditional
bereproductive
shift in the moralcode thatalso legitimatednontraditional
havior(fertilitycontrol).
politicalbehavior
ledtobothnontraditional
thatsecularization
Theargument
fertilitybehavioris convincing.Thequestionremains,howandnontraditional
Secever:whatsocialconditionsfosteredthe developmentof secularization?
ularizationis not simplya productof shifts in economicstructurebecause
wereincludedas predictorsin the
laborforcevariables(modeof production)
models.It seemsprobablethatthe actionsof politicalactivists(local or outof newsevents,thedistribution
side),strikesorotherconsciousness-changing
papersor pamphlets,risinglevels of education,andothersocialandpolitical
activitieswereresponsiblefor the differentialgrowthof secularismin some
or culturalfactors?Changesin the organizational
areas.Are these structural
structureandpoliticalexperiencesof a communityare intimatelyboundup
with changesin consciousnessof the population.The taskis to clarifyhow
these factorsintersectedin differenthistoricalcontextsto change fertility
behavior(andotheroutcomes),not simplyto claimthatit is an "either-or"
contestbetweencultureandeconomicforces.
Anotherrelevantstudyis therecentresearchby Goodkind(1991, 1993)on
of birthsin auspiciousyearsof theChinesezodiacalcalendar
theconcentration
in a numberof EastAsianpopulations.Goodkindshowsthatthe "traditional"
customof havinga baby bornin the year of the dragondid not appearin
Taiwanbefore1976.The explanationthatthe controlof the timingof births

224

HIRSCHMAN

is not possible in naturalfertility populations is disputed by Goodkind, who

shows that the concentrationof birthsin the dragonyear of 1976 was accomplished largely by intermediatevariables other than contraception(marriage
timing, abortion,and coital behavior). If it was always possible to fine-tune
the timing of birthsso that there was a concentrationin auspicious years, why
did this traditionbegin only in the 1970s? It seems that the celebration of
cultural practices is closely intertwined with modernity and not simply a
productof historical continuity. In a survey of the role of culturalfactors on
Chinese fertility, Greenhalgh (1988) argues that cultural factors cannot be
given a transhistoricalrole but must be interpretedwithin particularhistorical
and institutionalcontexts.
The diffusion of patternsof low fertility is often given as a major reason
for the importanceof culturalfactors. Clearly, informationand beliefs spread
more easily within communities that share common cultural and linguistic
characteristics.But rapidchanges in behavior over wide geographic areas do
not necessarilymean thata common culturalsystem is an essential prerequisite
for the diffusion of informationor for changes in reproduction.In an excellent
historical study of a ruralSicilian town, Schneider& Schneider (1992) show
that several decades separatedthe fertility declines of different classes in the
community. The fertility decline of the gentry began around the turn of the
century, the artisanclass developed a patternof lowered fertility in the 1920s
and 1930s, and the peasantryfollowed suit in the 1950s and 1960s. Changes
in fertility seemed to be more closely associated with changes in social and
economic incentives than the diffusion of values and informationin this community. In other contexts, all social classes may face common incentives, and
the pace of change may be determinedsolely by diffusion of informationand
the cost of fertilityregulation.However, diffusion processes have provenmuch
more difficult to model and test than to observe (Rosero-Bixby & Casterline
1993).
The classic idea of the diffusion hypothesis is that it requires much more
motivation to be an earlier innovatorthan to be a later one. Pioneers have to
bearthe full costs of acquiringnew information,independentdecision-making,
breaking with tradition, risking social disapproval, and assuming risks of
uncertaintyof futureoutcomes. After a significant proportionof a population
has already engaged in innovative behavior (e.g. fertility control), the costs
for those that follow are much less. This means that the level of motivation
requiredfor behavioralchange by followers is also much less. Given this logic,
the association of socioeconomic factors (the motivating conditions) and fertility behaviormay loom muchlargerin the initial spreadof a fertilitytransition
than in the latter stages (Hirschmanand Guest 1990).
The incorporationof diffusion processes into studies of fertility decline is
as importantas it is difficult (Casterlineet al 1987, Montgomery& Casterline

CHANGES
WHYFERTILITY

225

1993). The developmentof time series data of detailed geographicalareas and

innovative statisticalapproachesmay be essential prerequisitesfor such work.

Family Planning Programs and Fertility Declines

Twenty-five years ago, there was great uncertaintywhether family planning
programsin developing countrieswould have any impact on fertility declines
(Davis 1967). Over the last 20 years, a series of cross-nationalstudies have
estimatedthe impactof socioeconomic developmentandfamily planningeffort
on fertility levels, fertility change, and contraceptiveuse (Freedman& Berelson 1976, Mauldin & Berelson 1978, Lapham & Mauldin 1984, Mauldin &
Ross 1991). In spite of some thoughtfulskeptics (Hernandez1981, 1984), the
general finding is that both socioeconomic conditions and family planning
programshave significant independenteffects on fertility decline, and there is
a synergistic effect (lowering fertility) of rapid development and an effective
family planning program. A well-designed longitudinal study of treatment
(with a family planning program)and nontreatmentareas in Bangladesh has
found strong evidence that effective family planning programslower fertility
(Phillips et al 1988). A comparablestudy of Egypt, however, found mixed
results of the effects of family planning on fertility attitudes and behavior
(Stycos et al 1988).
In spite of the general consensus on the value of family planningprograms,
there is still much debate in the field over the relative efficacy of investments
in family planning programs(comparedto investments in development) and
over what aspects of programsmake a difference.The main analyticalproblem
is that family planning programs are not random events but are disproportionatelylocated in countrieswhere social andeconomic conditionsare already
favorablefor fertility declines. Skeptics claim that historical fertility declines
in many Western countries occurredwithout organized family planning programs and that the distributionof contraceptivesthrough the private market
might well provide most of the services that public family programscurrently
do. These questions are still debatedand probablycannot be resolved with the
standardmethods of programevaluation.
One of the importantunresolvedquestionshas been the identificationof the
componentsof family planning programsand the relative impact of different
programactivities on contraceptiveadoptionand fertilitydecline. The primary
source of data about the attributesof national family planning programshas
been ratingsof programeffectiveness collected in questionnairesfrom knowledgeableexperts(programadministrators,fundingagency officials, and others
familiarwith family planningprogramsin variouscountries)(Mauldin& Ross
1991). In most studies, the responses from these questionnaireshave been
summarizedinto one global dimension of "programeffort," although there is

226

HIRSCHMAN

clearly a set of underlyingdimensions that characterizethe major features of

family planningprograms(Entwisle 1989).
Perhapsthe most importantfinding for policy makersis that the availability
of family planning services, measuredby the proximity to clinics, does have
measurableeffect on contraceptiveuse (Entwisle et al 1986, Tsui & Ochoa
1992). Researchon the identificationof the impactof specific aspects of family
planningprogramson fertility behavior and the relative cost-effectiveness of
different strategies does not seem to have yielded clear answers (Phillips &
Ross 1992).

OTHERMODELS AND OTHER QUESTIONS

The title of this paperpresumesthat there is an answer to the question of why
fertilitychanges. If the empiricalresearchreviewed in this essay is an accurate
pictureof the state of knowledge on the topic, thereare many possible answers
to this question, usually phrased in terms of particularhistorical conditions.
The model of social causation,implied by the question, does not accord well
with such varied responses. The standardsocial science model is that society
works pretty much like a regression equation:the task is to find the right set
of predictors,solve the equation,and discover what factorsare most important
in predictingsocial outcomes. This frameworkdoes lead to empirical generalizations,but there seem to be endless qualificationsabout the measurement
of variables, the meaning and interpretationof variables, the substitutability
of one variablefor another,and complex interactionswith historical settings.
If science is to discover parsimoniousprinciplesthatexplain complex patterns,
we do not seem to be making progress.
Perhapswe need to reconsiderthe question about fertility change in light
of othermodels of populationand society. The classical model of demography
is the Malthusian equilibrium, where there are fluctuations of growth and
decline aroundthe balance of populationsize and economic resources. Every
studentof demographyquickly learns all the shortcomingsof Malthus's principle of population-Malthus did not anticipatetechnological change nor the
ability of modem societies to control maritalfertility. What is rarely stressed
is that the equilibriummodel did fit rather well with the dynamics of preindustrialsocieties (Wrigley 1969, Grigg 1980, Galloway 1988).
The logic of a homeostaticmodel is a system maintainedby countervailing
pressures.In the Malthusiansystem, constantpressurefor populationgrowth
is fueled by the passion between the sexes, but this is counteredby the negative
feedback loop of limited food supplies on population size. This negative
feedback loop (density dependence) effect leads to the positive check
of increased mortality. Preventive checks (via constraintson marriage)that

CHANGES 227
WHYFERTILITY
slowed populationgrowth were the other mechanismto maintainequilibrium
in the Malthusianmodel.
Ronald Lee (1987) has suggested that Malthusianhomeostatic principles,
along with Boserupianforces that createdpositive feedbacks, were central to
shaping cycles of population growth and decline until about 100 years ago
when rapid technological progress broke the negative feedback loop. There
may be, however, otherequilibratingforces thathave an impact on population
growth in the modem era, especially in countrieswith rapidly growing populations. I am not suggesting that there is an instinct for families to have two
survivingchildren,only thatrapidpopulationgrowthcertainlycreatessystemic
pressures (negative feedbacks) for lowered fertility. A homeostatic model,
incorporatingKingsley Davis's notion of household strain as a central motivatingfactor,mightwell providean alternativeframeworkfor studyingfertility
dynamics as part of an interdependentsystem that creates opportunitiesand
costs for family welfare.
An equilibriumor homeostaticframeworkmightbe usefully appliedto some
of the anomalouspatternsconsideredin the priorreview of the literature(Lee
1987). The first case is the wide variations of fertility in "naturalfertility"
populations.Recall that fertility respondedin a systematic way to variations
in economic conditionsin premodernEuropevia changes in marriagepatterns
(Wrigley 1966, Goldstone 1986). There are also systematic differences between forager and agriculturalpopulations (Gillian et al 1993) and between
agriculturalpopulationsat different levels of density (Easterlin 1971, 1976a,
Firebaugh 1982). The wide variationsin intermediatevariables in many traditional (pre-fertilitytransition)populations(sexual abstinence,breastfeeding
duration,rules on widow remarriage,patternsof male migration,divorce, etc)
seem to be simply culturalcuriosities unless we assume that these practices
were social adaptationsto regulate population size and growth, just as rules
on marriagein premodernEnglandwere responses to economic cycles. Much
of contemporarydemographicwisdom considers these variationsbeyond the
relevance of theories of fertility decline, which are limited to the origins of
conscious controlof maritalfertility.Conscious patternsof fertility control are
an importantpart of the process, but the theoretical framework of fertility
transitionsshould be built on a broaderbase.
If the homeostaticprincipleis to maintaindemographicequilibriumin order
to avoid community and household strain, then changes in fertility are only
one of severalmechanismsthatcan respondto the rapidincreasesin population
growth(as a result of reductionsin mortality)thatbegan in the eighteenth and
nineteenthcenturies in many Europeancountries and aroundthe globe in the
twentiethcentury. The first response was probablymigration:to agricultural
frontiers,to cities, and to settler societies in the New World and Oceania. The
two extreme cases in Europeandemographichistory-why did the fertility

228

HIRSCHMAN

decline begin so early in France and so late in England?-may be explicable

in terms of differentialopportunitiesfor outwardmigrationfrom ruralareas.
But migrationwas only a temporarypalliative;the explosive populationgrowth
that resulted from declining mortality rates led to fertility reductions that
eventually (within a century) resulted in a near demographicequilibrium in
industrialcountries. If currentpatternscontinue, the same outcome is likely
in the contemporarydeveloping countriesin the next 50 years or so.
The wide variations in correlates of the speed of fertility reduction seem
anomalous only if the "true theory" is fixed on a small number of master
determinantsof lowered fertility in all settings. If, on the other hand, the
homeostatic principle is the central idea, the reduction of population growth
by fertility decline could occur differentially across a wide variety of situations-depending on the value and costs of children, the costs of fertility
regulation, access to information, and many other conditions. While it is
importantto map the variationsin timing of fertility transitions(including the
initiationand the pace of change) and the immediatecorrelatesof differential
timing, the centralpoint is thatall of these revolutionarydemographicchanges
spannedthe globe in about a century.
I am not entirely convinced that the alternative model of demographic
homeostasis will lead to directions of day-to-day researchfundamentallydifferent from theories of fertility decline. But it casts the centralquestion rather
differently. Ratherthan asking why fertility declines and expecting a simple
answer, the alternativemay be to ask, how are demographicequilibria reestablished after mortalitydeclines? This question will require a much greater
toleranceof complexity as thereare certainto be many pathsto this end. There
may also be greater variationsin demographicregimes at the conclusion of
fertility transitions,but that is a story for anotherday.
ACKNOWLEDGMENTS

This essay was written while the author was a Fellow at the Center for
Advanced Study in the Behavioral Sciences. I am grateful for financial
support provided by the National Science Foundation (SES-9022192) and
the National Institutesof Child Health and HumanDevelopment (HD21267).
I am grateful to JooEan Tan for bibliographicalassistance in the preparation
of this essay and to BarbaraEntwisle, Kathleen Much, Samuel Preston, and
Ronald Rindfuss for their critical comments and suggestions on an earlier
draft of the essay.
Any Annual Review chapter, as well as any article cited In an Annual Review chapter,
may be purchased from the Annual Reviews Preprints and Reprints service.
1-800-347-8007; 415-259-5017; email: [email protected]

WHY FERTILITYCHANGES

229

LiteratureCited
Alter G. 1992. Theories of fertility decline: a
nonspecialist's guide to the currentdebate.
In The European Experience of Declining
Fertility, 1850-1970: The Quiet Revolution,
ed. JR Gillis, LA Tilly, D Levine, pp. 13-25.
Cambridge:Blackwell
AndorkaR. 1982. Lessons from studies of differential fertility in advanced societies. In
Determinants of Fertility Trends: Theories
Re-Examined,edited by C Hohn, R Mackensen, pp. 31-31. Liege (Belgium): Ordina
Editions
Axinn WG. 1992. Family organizationand fertility limitation in Nepal. Demography 29:
503-22
Axinn WG. 1993. The effects of children's
schooling on fertility limitation. Pop. Stud.
47:481-93
Bean LL, MineauGP, AndertonDL, 1990. Fertility Change on the AmericanFrontier:Adaptation and Innovation. Berkeley: Univ.
Calif. Press
Becker GS. 1960. An economic analysis of fertility. In Demographic and Economic
Changein Developed Countries,pp. 209-31.
NBER Conference Series 11. Princeton:
PrincetonUniv. Press
Becker GS. 1988. Family economics andmacro
behavior.Am. Econ. Rev. 78:1-13
Blake J. 1968. Are babies consumer durables?
A critiqueof the economic theory of reproductive motivation.Pop. Stud. 22:5-25
Blake J. 1985. The fertility transition:continuity or discontinuitywith the past? Proceedings: Int. Pop. Conf, Florence, 1985, pp.
393-405. Liege: Int. Union for the Scientific
Study of Pop.
BongaartsJ. 1975. Why high birthrates are so
low. Pop. Dev. Rev. 1:289-96
BongaartsJ. 1978. A frameworkfor analyzing
the proximatedeterminantsof fertility. Pop.
Dev. Rev. 14:105-32
BongaartsJ. 1987. The transitionin reproductive behavior in the Third World. In World
Population and U.S. Policy, ed. J Menken,
pp. 105-132. New York: Norton
Bongaarts J, Potter RG, 1983. Fertility, Biology, and Behavior: An Analysis of the
Proximate Determinants. New York: Academic
Brass W, Jolly CL, eds. 1993. Population Dynamics in Kenya. Washington, DC: Natl.
Acad. Press
BulataoRA, Lee, RD, eds. 1983. Determinants
of Fertility in Developing Countries.2 vols.
New York: Academic
BurchTK. 1983. The impact of forms of families and sexual unions and dissolutions of

unions on fertility. See Bulatao & Lee 1983,

2:532-61
Butz WB, Ward MP. 1979. Will US fertility
remain low? a new economic interpretation.
Pop. Dev. Rev. 5:663-88
Caldwell JC. 1976. Toward a restatement of
demographic transition theory. Pop. Dev.
Rev. 2:321-66
Caldwell JC. 1980. Mass education as a determinantof the timing of fertilitydecline. Pop.
Dev. Rev. 6:225-56
Caldwell JC. 1982. Theoryof FertilityDecline.
London:Academic
Caldwell JC. 1983. Direct economic costs and
benefitsof children.See Bulatao& Lee 1983,
pp. 370-97
Caldwell JC, OrubuloyeIO, Caldwell P. 1992.
Fertility decline in Africa: a new type of
transition?Pop. Dev. Rev. 18:211-42
Caldwell JC, Reddy PH, Caldwell P. 1988. The
Causes of Demographic Change: Experimental Research in South India. Madison:
Univ. Wisc. Press
Carlsson G. 1966. The decline of fertility: innovation or adjustmentprocess. Pop. Stud.
20:147-74
CasterlineJ, MontgomeryMR, ClarkRL. 1987.
Diffusion models of fertility control: are
there new insights? Pop. Stud. Train. Ctr.
WorklPap. Series 87-06. Providence:Brown
Univ.
Cleland J. 1985. Maritalfertility decline in developing countries: theories and the evidence. See Cleland & Hobcraft 1985, pp.
223-52
Cleland J. 1993. Equity, security, and fertility.
Pop. Stud. 47:345-52
Cleland J, Hobcraft,J. eds. 1985. Reproductive
Change in Developing Countries: Insights
from the WorldFertility Survey. New York:
Oxford Univ. Press
Cleland J, Rodriguez G. 1988. The effect
of parental education on marital fertility
in developing countries. Pop. Stud. 42:41942
Cleland J, Scott, C. eds. 1987. The WorldFertilitySurvey:An Assessment.New York: Oxford Univ. Press
ClelandJ, Wilson C. 1987. Demand theories of
the fertility transition:an iconoclastic view.
Pop. Stud. 41:5-30
Coale AJ. 1973. The demographic transition
reconsidered. In Proceedings: Int. Pop.
Conf, Liege, 1973. 1:53-72. Liege: Int.
Union for the Scientific Study of Pop.
Coale AJ. 1979. The demographictransition:a
summary,some lessons, and some observations. In Fertility Transitionof the East Asia

230

HIRSCHMAN

Populations, ed. LJ Cho, K Kobayashi, pp.

pirationsand resources.Pop. Dev. Rev. 2:
9-23. Honolulu: Univ. Press of Hawaii
415-25
Coale AJ. 1992. Age of entryinto marriageand EasterlinRA. 1978a.Whatwill 1984be like?
the date of the initiation of voluntarybirth
Socioeconomic
of recenttwists
implications
control. Demography29:333-42
in the age structure.Demography15:397Coale AJ, Treadway R. 1986. A summaryof
432
the changing distributionof overall fertility, Easterlin
RA. 1978b.Theeconomicsandsocimarital fertility, and the proportionmarried
ology of fertility:a synthesis.In Historical
in the provinces of Europe. In The Decline
Studiesof ChangingFertility,ed. C Tilly, pp.
of Fertilityin Europe,ed. AJ Coale, SC Wat57-133. Princeton:
PrincetonUniv.Press
kins, pp. 31-181. Princeton:PrincetonUniv. EasterlinRA. 1983.Modernization
andfertilPress
ity:a criticalessay.See Bulatao& Lee 1983,
Coale AJ, Trussell TJ. 1974. Model fertility
pp.562-86
schedules: variationsin the age structureof
EasterlinRA, CrimminsE. 1985.TheFertility
childbearing in human populations. Pop.
Revolution.
Chicago:Univ.ChicagoPress
Index 40:185-258
EntwisleB. 1989. Measuringcomponentsof
Coale AJ, Trussell TJ. 1978. A new procedure
familyplanningprogrameffort.Demografor fitting optimal values of the parameters
phy26:53-76
of a model schedule of maritalfertilityrates. EntwisleB, MasonWM,HermalinAI. 1986.
Pop. Index 44:211-13
Multileveldependence
of contraceptive
use
Coale AJ, Watkins, SC eds. 1986. TheDecline
on socioeconomicdevelopmentandfamily
of Fertility in Europe. Princeton:Princeton
planningprogram
strength.
Demography
23:
Univ. Press
199-216
Cochrane SH. 1979. Fertility and Education: EntwisleB, Casterline
JB, SayedHAA. 1989.
WhatDo We Really Know? Baltimore,MD:
Villagesascontextsforcontraceptive
behavJohns Hopkins Univ. Press
ior in ruralEgypt. Am. Sociol. Rev. 54:1019Cochrane SH. 1983. Effects of education and
34
urbanizationon fertility. See Bulatao & Lee FeeneyG, WangG, ZhouM, Xiao B. 1989.
1983 2:587-626
Recentfertilitydynamicsin China:Results
Davis K. 1955. Institutionalpatternsfavoring
fromthe1987onepercentpopulation
survey.
high fertility in underdevelopedareas. EuPop. Dev. Rev. 15:297-322
genics Q. 2:33-39
Firebaugh
G. 1982.Population
densityandferDavis K. 1963. The theory of change and retilityin 22 Indianvillages.Demography
19:
sponse in moderndemographichistory. Pop.
481-94
Index 29:345-66
FosterA. 1990. Cohortanalysisand demoDavis K. 1967. Populationpolicy: will current
graphictranslation:
a comparative
studyof
programssucceed? Science 158:730-39
recenttrendsin age specificratesfromEuDavis K, BernstamMS, Ricardo-Campbell,R.,
rope and North America.Pop. Stud. 44:287eds. 1986. Below-ReplacementFertility in
315
IndustrialSocieties: Causes, Consequences, Freedman
R. 1963.Normsfor familysize in
and Policies. A supplement to Pop. Dev.
underdevelopedareas. Proc. R. Soc. Series
Rev. Vol. 12
(B) 159:220-45
Davis K, Blake J. 1956. Social structureand Freedman
R. 1979.Theoriesof fertilitydecline:
fertility:An analytic framework.Econ. Dev.
A reappraisal.Soc. Forc. 58:1-17
CulturalChange 4:211-35
Freedman
R, BerelsonB. 1976.Therecordof
Dixon R. 1971. Explainingcross-culturalvarifamilyplanningprograms.
Studiesin Family
ations in age at marriage and proportions
Planning 7(1):1-40
never-marrying.Pop. Stud 25:215-34
Freedman
R, BlancAK. 1992.FertilitytransiDyson T, Murphy M. 1985. The onset of fertion: an update. Int. Family. Planning. Pertility transition.Pop. Dev. Rev. 11:399-440
spec. 18(2):44-50,72
EasterlinRA. 1969. Towards a socioeconomic FrickeT, TeachmanJD. 1993. Writingthe
theory of fertility. In Fertility and Family
names:marriagestyle,livingarrangements,
Planning: A World View, ed. SJ Behrman,
andthefirstbirthintervalin Nepalisociety.
L Corsa, RF Freedman. Ann Arbor: Univ.
Demography30:175-88
Mich. Press
D. 1969. Demographic
Friedlander
responses
EasterlinRA. 1971. Does humanfertilityadjust
andpopulation
change.Demography
6:359to the environment.Am. Econ. Rev. 61:39981
407
Friedlander
D. 1983. Demographic
responses
EasterlinRA. 1976a. Factors in the decline of
and socioeconomicstructure:population
farm family fertility in the United States:
processesin EnglandandWalesin theninesome preliminary research results. J. Am.
teenthcentury.Pop.Stud.20:249-72
Hist. 63:600-14
Friedlander
D, SchellekensJ, Ben-MosheE.
EasterlinRA. 1976b. The conflict between as1991.Thetransition
fromhighto lowmarital

WHY FERTILITYCHANGES
fertility: culturalor socioeconomic determinants. Econ. Dev. Cult. Change 39:331-51
Galloway PR. 1988. Basic patterns in annual
variations in fertility, nuptiality, mortality,
and prices in Pre-industrial Europe. Pop.
Stud. 42:275-302
Gillian RB, GoldbergT, GrazynaJ. 1993. The
fertility of agriculturaland non-agricultural
traditionalsocieties. Pop. Stud. 47:269-81
Gillis JR, Tilly LA, Levine, D. eds. 1992. The
European Experience of Declining Fertility,
185S0-1970: The Quiet Revolution. Cambridge: Blackwell
Goldstone JA. 1986. The demographicrevolution in England:a re-examination.Pop. Stud.
40:5-34
Goodkind DM. 1991. Creating new traditions
in modern Chinese populations. Pop. Dev.
Rev. 17:663-86
GoodkindDM. 1993. New zodiacal influences
on Chinese family formations:Taiwan, 1976.
Demography30:127-42
Govindasamy P. DaVanzo J. 1992. Ethnicity
and fertility differentials in PeninsularMalaysia: do policies matter? Pop. Dev. Rev.
18:243-67
GreenhalghS. 1988. Fertilityas mobility:Sinic
transitions.Pop. Dev. Rev. 14:629-74
GriggDB, 1980. PopulationGrowthand Industrial Change: An Historical Perspective.
Cambridge:CambridgeUniv. Press
Guest P, Tan JE. 1993. Social change and celibacy in Thailand. Unpublished paper. Ctr.
Stud. Demogr. Ecol., Univ. Wash., Seattle
Guinnane TW, Okun BS, Trussell J. 1992.
What do we know about the timing of fertility transitions in Europe? Worn Pap. No.
92-11. Princeton:Off. Pop. Res., Princeton
Univ.
GutmannMP, Fliess KH. 1993. The determinants of early fertility decline in Texas. Demography30:443-47
Haines MR. 1979. Fertility and Occupation:
Population Patterns and Industrialization.
New York: Academic
Haines MR. 1989. American fertility in transition: New estimates of birth rates in the
United States, 1900-1910. Demography26:
137-48
Hajnal J. 1947. The analysis of birth statistics
in the light of the recent international
recovery of the birth-rate.Pop. Stud. 1:13764
Hajnal J. 1965. Europeanmarriagepatternsin
perspective.In Populationin History:Essays
in Historical Demography, ed. DV Glass,
DEC Evesley, pp. 101-143. London:Edward
Arnold
Hammel EA. 1990. A theory of culturefor demography.Pop. Dev. Rev. 16:455-86
Hanley SB. 1979. The Japanesefertilitydecline
in historical perspective. In Fertility Transition of the East Asia Populations,ed. LJ Cho,

231

K Kobayashi,pp. 24-48. Honolulu:Univ.

Pressof Hawaii
HenryL. 1961.Somedataon naturalfertility.
Eugenics Q. 8:81-91
DJ. 1981.A noteon measuringthe
Hernandez
impactof familyplanningproindependent
18:
gramson fertilitydeclines.Demography
627-34
HernandezDJ. 1984. Success or Failure: Family Planning Programs in the Third World.

Greenwood
Westport:
socioeconomic
C. 1985.Premarital
Hirschman
A
rolesandthe timingof familyformation:
comparativestudyof five Asian societies.
Demography22:35-59

C. 1986.The recentrise in Malay

Hirschman
lull in a
fertility:a newtrendor a temporary
23:161-84
Demography
fertilitytransition?
HirschmanC, GuestP. 1990. The emerging
transitionsof SoutheastAsia.
demographic
Pop. Dev. Rev. 16:121-52

A,
HirschmanC, Tan JE, Chamratrithirong
GuestP. 1994.The pathto below replaceInt.Family Planmentfertilityin Thailand.
ning Perspectives. (forthcoming)

HoriuchiS. 1992.Stagnationin the declineof

the worldpopulationgrowthrate.Science
257:761-65
Hutaserani
S, RoumassetJ. 1991.Institutional
changeand the demographictransitionin
ruralThailand. Econ. Dev. Cult. Change 40:

75-100
Jones GW. 1990. Fertilitytransitionsamong
Malaypopulationsof SoutheastAsia: puzPop.Dev.Rev.16:507zlesof interpretation.
17
KnodelJ. 1977.Familylimitationandthe ferEvidencefromage patterns
tilitytransition:
of fertilityin Europeand Asia. Pop. Stud.
31:219-49
KnodelJ, 1983.Naturalfertility:Age patterns,
levels,andtrends.See Bulatao& Lee 1983,
1:61-102
KnodelJ, vande WalleE. 1979.Lessonsfrom
of historicalferthepast:policyimplications
tilitystudies.Pop.Dev. Rev.5:217-45
betweencurrent
KumarJ. 1971.A comparison
Indianfertilityand late nineteenth-century
Swedish and Finnish fertility. Pop. Stud. 25:

269-82
LaphamRJ,MauldinWP. 1984.Familyplandeclinein
ningprogrameffortandbirthrate
developing countries.Int. Family. Planning.
Perspec. 10(4):109-18

Lee RD. 1987. Populationdynamicsof humans and other animals. Demography

24:443-65
Leete R, Alam I. 1993. TheRevolutionin Asian
Fertility: Dimensions, Causes, and Impli-

cations.Oxford:Clarendon
fertility
Leete R, Tan BA, 1993. Contrasting
trendsamongethnicgroupsin Malaysia.See
Leete& Alam1993,pp. 128-147

232

HIRSCHMAN

Lesthaeghe R. 1980. On the social control of

humanreproduction.Pop. Dev. Rev. 6:52748
LesthaegheR. 1983. A centuryof demographic
change in Western Europe: An exploration
of underlying dimensions. Pop. Dev. Rev.
9:411-36
LesthaegheR, SurkynJ. 1988. Culturaldynamics andeconomic theoriesof fertilitychange.
Pop. Dev. Rev. 14:1-46
LesthaegheR, Wilson C. 1986. Modes of production, secularization,and the pace of fertility decline in WesternEurope, 1870-1930.
In TheDecline of Fertility in Europe, ed. AJ
Coale, SC Watkins, pp. 261-92. Princeton:
PrincetonUniv. Press
Lim LL, Jones G, HirschmanC. 1987. Continuing fertility transitionsin a plural society.
J. Biosoc. Sci. 19:405-25
Mason KO, 1992. Cultureand the fertilitytransition. Genus 48(3-4):1-14
MauldinWP, Berelson B. 1978. Conditionsof
fertility decline in developing countries.
Stud. Family Planning 9(5):89-148
MauldinWP, Ross JA. 1991. Family planning
programs:efforts and results. Stud. Family
Planning. 22:350-67
McNicoll G. 1992. Changing fertility patterns
and policies in the ThirdWorld. Annu. Rev.
Sociol. 18:85-108
Menken, J. Phillips JF. 1990. Population
change in a rural area of Bangladesh. Ann.
Am. Acad. Polit. Soc. Sci. 510:87-101
MontgomeryMR, CasterlineJB. 1993. The diffusion of fertility control in Taiwan: evidence from pooled cross-section time-series
models. Pop. Stud. 47:457-79
Mosher WD. 1980a. The theory of change and
response:an applicationto PuertoRico. Pop.
Stud. 34:45-58
Mosher WD. 1980b. Demographic responses
and demographictransition:a case study of
Sweden. Demography17:395-412
Mosk C. 1979. The decline of maritalfertility
in Japan.Pop. Stud.. 33:19-38
Ni Bhrolchain M. 1992. Period paramount?a
critique of the cohort approachto fertility.
Pop. Dev. Rev. 18:599-629
Notestein F. 1945. Population-the long view.
In Food for the World,ed. TW Schultz, pp.
36-57. Chicago: Univ. Chicago Press
Notestein F. 1953. Economic problemsof populationchange. In Proceedings of the Eighth
Int. Conf of Agric. Economists, pp. 13-3 1.
London:Oxford Univ. Press
Palloni A. 1990. Fertilityand mortalitydecline
in LatinAmerica.Ann.Am Acad. Polit. Soc.
Sci. 510:126-44
Phillips JF, Ross, JA eds. 1992. Family Planning Programmes and Fertility. Oxford:
Clarendon
Phillips JF, Simmons R, Koenig MA,
ChakrabortyJ. 1988. Determinantsof repro-

ductive change in a traditionalsociety: evidence from Matlab,Bangladesh. Stud. Fam.

Planning 19:313-34
Pollak RA, Watkins SC. 1993. Cultural and
economic approaches to fertility: a proper
marriageor a mesalliance? Pop. Dev. Rev.
19:467-96
PrestonSH. 1986. Changingvalues and falling
birth rates. See Davis, Bernstam, RicardoCampbell 1986, pp. 176-95
Reid A. 1988. Southeast Asia in the Age of
Commerce, 1450-1680. New Haven: Yale
Univ. Press
Reid A. 1992. Economic and social change,
c.1400-1800. In The Cambridge History of
SoutheastAsia, ed. N Tarling,2 vols. 1:460507. Cambridge:CambridgeUniv. Press
Rele JR, Alam I. 1993. Fertility transition in
Asia: the statistical evidence. See Leete &
Alam 1993, pp. 15-37
RetherfordRD. 1979. A theoryof rapidfertility
decline in homogenous populations. Stud.
Fam. Planning 10:61-67
Retherford RD, Cho UJ. 1973. Comparative
analysisof recentfertilitytrendsin East Asia.
In Proc. Int. Pop. Conf.,Liege, 1973. 2:16381. Liege: Int. Union for the Scientific Study
of Pop.
RetherfordRD, Palmore, JA. 1983. Diffusion
processes affecting fertility regulation. See
Bulatao & Lee 1983, 2:295-339
RindfussRR, MorganSP. 1983. Marriage,sex,
and the first birth interval in Asia: the quiet
revolution.Pop. Dev. Rev. 9:259-78
Robey B, RutsteinSO, MorrisL, BlackburnR.
1992. The reproductiverevolution:new survey findings. Pop. Rep. Ser. M, No.11. Baltimore: Johns Hopkins Univ., Pop. Information Program
Robinson WC. 1992. Kenya enters the fertility
transition.Pop. Stud. 46:445-58
Rosero-Bixby, L, CasterlineJB. 1993. Modelling diffusion effects in fertility transition.
Pop. Stud. 47:147-67
Rutenberg N, Diamond I. 1993. Fertility in
Botswana:the recentdecline andfutureprospects. Demography30:143-57
RyderNB. 1965. The cohort as a concept in the
study of social change. Am. Sociol. Rev. 30:
843-61
Ryder NB. 1969. The emergence of a modern
fertility pattern:United States, 1917-66. In
Fertility and Family Planning: A World
View, ed. SJ Behrman, L Corsa, RF Freedman, pp. 99-123. Ann Arbor:Univ. Michigan Press
Ryder NB. 1983. Cohort and period measures
of changing fertility. See Bulatao & Lee
1983, 2:737-56
SandersonWC. 1976. On two schools of the
economics of fertility.Pop. Dev. Rev. 2:46977
SandersonWC. 1987. Below-replacementfer-

WHY FERTILITYCHANGES
tility in nineteenth century America. 1987.
Pop. Dev. Rev.. 13:305-13
Schultz TP. 1981. Economics of Population.
Reading,Mass: Addison-Wesley
Smith PC. 1980. Asian marriage patterns in
transition.J. Family Hist. 5:58-96
Smith PC. 1983. The impactof age at marriage
and proportions marrying on fertility. See
Bulatao & Lee 1983, 2:473-531
Stycos JM, Sayed HAA, Avery R, FridmanS.
1988. CommunityDevelopment and Family
Planning:An EgyptianExperiment.Boulder,
Colo: Westview
ThomasN. 1991. Land, fertility, and the population establishment.Pop. Stud.45:379-97
Thomas N. 1993. Economic security, culture,
and fertility. Pop. Stud. 47:353-59
ThompsonW. 1929. Population.Am. J. Sociol.
34: 959-75
ThortonA, Fricke T. 1987. Social change and
the family: comparative perspectives from
the West, China, and South Asia. In Demography as an Interdiscipline,ed. JM Stycos,
pp. 128-61. New Brunswick, NJ: Transaction
Tilly C. 1978. The historicalstudy of vital processes. In Historical Studies of Changing
Fertility, ed. C Tilly, pp. 3-55. Princeton:
PrincetonUniv. Press
Tolnay SE, Guest AM 1984. American family
buildingstrategiesin 1900: stoppingor spacing. Demography21:9-18
Tsui AO, Ochoa LH. 1992. Service proximity
as a determinantof contraceptivebehavior:

233

evidence from cross-nationalstudies of survey data.See Phillips & Ross 1992, pp. 22256
TurchiB. 1976. Microeconomictheoriesof fertility: a critique. Soc. Forc. 54:107-25
van de Walle E. 1992. Fertilitytransition,conscious choice, and numeracy. Demography
29:487-502
WatkinsSC. 1986. Conclusions.In TheDecline
of Fertility in Europe, ed. AJ Coale, SC
Watkins, pp. 420-50. Princeton: Princeton
Univ. Press
Watkins SC. 1987. The fertility transition:Europe and the third world compared. Sociol.
Forum 2:645-75
Weinstein M, Sun TS, Chang MC, Freedman
R. 1990. Household composition, extended
kinship and reproductionin Taiwan: 19651985. Pop. Stud. 44:217-40
Wilson C, Woods R. 1991. Fertilityin England:
A long-termperspective.Pop. Stud.43:299316
Woods RI. 1987. Approaches to the fertility
transition in Victorian England. Pop. Stud.
41:283-312
Wrigley EA. 1966. Family limitation in preindustrialEngland. Econ. Hist. Rev. 19:82109
Wrigley EA. 1969. Population and History.
New York: McGraw-Hill
Xenos P. Gultiano S. 1992. Trends in female
and male age at marriageand celibacy. Papers of the East-WestCenter,No 120. Honolulu: East-West Center