S. HRG.

110118

MEETING THE CHALLENGE OF INCOME

INSTABILITY

HEARING
BEFORE THE

JOINT ECONOMIC COMMITTEE

CONGRESS OF THE UNITED STATES
ONE HUNDRED TENTH CONGRESS
FIRST SESSION

FEBRUARY 28, 2007

Printed for the use of the Joint Economic Committee

U.S. GOVERNMENT PRINTING OFFICE

WASHINGTON

34852 PDF

2007

For sale by the Superintendent of Documents, U.S. Government Printing Office

Internet: bookstore.gpo.gov Phone: toll free (866) 5121800; DC area (202) 5121800
Fax: (202) 5122250 Mail: Stop SSOP, Washington, DC 204020001

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00001

Fmt 5011

Sfmt 5011

C:\JEC\34852.TXT

CELINA

PsN: CELINA

JOINT ECONOMIC COMMITTEE

[Created pursuant to Sec. 5(a) of Public Law 304, 79th Congress]
SENATE
CHARLES E. SCHUMER, New York, Chairman
EDWARD M. KENNEDY, Massachusetts
JEFF BINGAMAN, New Mexico
AMY KLOBUCHAR, Minnesota
ROBERT P. CASEY, JR., Pennsylvania
JIM WEBB, Virginia
SAM BROWNBACK, Kansas
JOHN E. SUNUNU, New Hampshire
JIM DEMINT, South Carolina
ROBERT F. BENNETT, Utah

HOUSE OF REPRESENTATIVES
CAROLYN B. MALONEY, New York, Vice Chair
MAURICE D. HINCHEY, New York
BARON P. HILL, Indiana
LORETTA SANCHEZ, California
ELIJAH E. CUMMINGS, Maryland
LLOYD DOGGETT, Texas
JIM SAXTON, New Jersey
KEVIN BRADY, Texas
PHIL ENGLISH, Pennsylvania
RON PAUL, Texas

CHAD STONE, Executive Director (Acting)

CHRISTOPHER J. FRENZE, Minority Staff Director

(II)

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00002

Fmt 5904

Sfmt 5904

C:\JEC\34852.TXT

CELINA

PsN: CELINA

CONTENTS
OPENING STATEMENT

OF

MEMBERS

Statement of Hon. Charles E. Schumer, Chairman, a U.S. Senator from

New York ..............................................................................................................
Statement of Hon. Carolyn B. Maloney, Vice Chair, a U.S. Representative
from New York .....................................................................................................
Statement Hon. Robert P. Casey, Jr., a U.S. Senator from Pennsylvania .........

1
3
4

WITNESSES
Statement of Hon. Peter R. Orszag, Director, Congressional Budget Office ......
Statement of Dr. Lael Brainard, Vice President and Director, Global Economy
and Development, The Brookings Institution ....................................................
Statement of Maurice Emsellem, Public Policy Director, National Employment Law Project .................................................................................................
Statement of Lily L. Batchelder, Assistant Professor of Law and Public Policy,
New York University School of Law ...................................................................
Statement of Dr. Bradley R. Schiller, Professor, School of Public Affairs,
American University ............................................................................................
SUBMISSIONS

FOR THE

5
13
15
18
19

RECORD

Prepared statement of Senator Charles E. Schumer, Chairman .........................

Prepared statement of Representative Carolyn B. Maloney, Vice Chair ............
Prepared statement of Hon. Peter R. Orszag, Director, Congressional Budget
Office .....................................................................................................................
Letter to Director Orszag from Hon. Charles E. Schumer and Hon. Jim
Webb ..............................................................................................................
Response of Director Orszag to Hon. Charles E. Schumer and Hon. Jim
Webb ..............................................................................................................
Attachment, analysis, Congressional Budget Office, Health and
Human Resources Division ...................................................................
Prepared statement of Dr. Lael Brainard, Vice President and Director, Global
Economy and Development, The Brookings Institution ...................................
Prepared statement of Maurice Emsellem, Public Policy Director, National
Employment Law Project ....................................................................................
Prepared statement of Lily L. Batchelder, Assistant Professor of Law and
Public Policy, NYU School of Law ......................................................................
Prepared statement of Dr. Bradley R. Schiller, Professor, School of Public
Affairs, American University ..............................................................................

29
30
31
40
41
43
74
77
81
91

(III)

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00003

Fmt 5904

Sfmt 5904

C:\JEC\34852.TXT

CELINA

PsN: CELINA

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00004

Fmt 5904

Sfmt 5904

C:\JEC\34852.TXT

CELINA

PsN: CELINA

MEETING THE CHALLENGE OF INCOME

INSTABILITY
WEDNESDAY, FEBRUARY 28, 2007

CONGRESS OF THE UNITED STATES,

JOINT ECONOMIC COMMITTEE,
Washington, DC
The Committee met at 9:35 a.m., in room 562 of the Senate Dirksen Building, the Honorable Charles E. Schumer, Chairman of the
Committee, presiding.
Senators present: Casey, Klobuchar, Schumer, and Webb.
Representatives present: Maloney.
Staff present: Katie Beirne, Chris Frenze, Nan Gibson, Rachel
Greszler, Colleen Healy, Asrael Klein, Michael Laskawy, Frank
Sammartino, Chad Stone, Annabelle Tamerjan, Robert Weingart,
Adam Wilson, and Jeff Wrase.
OPENING STATEMENT OF HON. CHARLES E. SCHUMER,
CHAIRMAN, A U.S. SENATOR FROM NEW YORK

Chairman Schumer. Good morning. I would like to thank our

witnesses and guests for attending today. Today were at a critical
juncture in America. In 2007, we are at a critical juncture in U.S.
economic policy. We know that the upheavals caused by technological change and international competition most acutely affect
those who are gaining least economically, the middle class and
those who are aspiring to get there.
Yet, in order for us to expand trade and make significant technological investments to help grow the economy, the middle class
must feel that they will benefit. Right now, too many of them dont.
They feel theyre being left out of the party. Everyone knows that
America is growing. Everyone knows the macrostatistics are pretty
good. But your average middle-class person says Im not doing so
well. Im not doing terribly, but Im not doing so well. Whats all
this talk about prosperity?
Working in a large corporation for 30 or 40 years that takes care
of you and your family for a lifetime is rapidly becoming a thing
of the past. Employers are now shifting the high costs of health
care and the burden of saving for retirement on to families, and increasingly jobs are being automated away by technological advancements or moving overseas, leaving many displaced workers
and their families behind at various levels of the economy.
Meanwhile, official numbers on the economy have been positive,
or at least until very recently. But we must face the reality lurking
behind the official numbers in order to address anxiety on Main
Street. Not only have wages significantly lagged behind produc(1)

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00005

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

2
tivity over the past two decades, but they are increasingly more
volatile as workers bounce in and out of jobs. Even those in highend jobs, computer programmers and others, cannot be assured the
same job security they had 10 or 15 years ago.
So theres much more volatility in the job market. Between 2003
and 2005, nearly 4 million workers were laid off from jobs they
held for more than 3 years. About half of those workers and their
families took a pay cut, and nearly one-third lost 20 percent or
more of their prior earnings.
And if the recession in the manufacturing sector that hit our
radar screens this week spreads throughout our economy, the economic roller-coaster for families will only get worse. Income volatility can cause major upheavals for families on top of the changes
theyre facing in the workplace. They can be forced to sell their
homes, for instance, or discontinue their health care. Income volatility also leaves families feeling unsettled about their families and
their childrens economic future.
We need a new policy direction to meet the challenge of income
instability. That would be the topic sentence of this first hearing
and something this Committee hopes to pursue. We must start by
strengthening the safety net that helps displaced workers rebound
from job losses that occurred through no fault of their own.
Weve asked our witnesses on the second panel to share their recommendations for doing just that. This morning our experts will
explore new policies like wage insurance and income averaging, as
well as ways to strengthen our existing unemployment insurance
and trade adjustment assistance programs. We also need to do everything we can at the Federal level to spur the development of
high quality, high paying jobs to replace the jobs lost in declining
segments of the economy or through advancements in technology.
We need to make serious investments in our most promising industries for future growth, such as renewable energy and life sciences,
and there are many others.
We need to help our displaced workers acquire the skills and experience theyll need to succeed in the new jobs created. Well investigate opportunities for creating good jobs in more detail in a series of JEC hearings in the coming months. But right now middleclass families need help dealing with the tectonic shifts technology
is causing. They need help dealing with the forces beyond their
control that are changing their lives. They dont want handouts,
but they certainly need a hand.
I know well have some disagreements over particular solutions
to this problem of income instability, but I hope that we will all
prioritize the need to help our families mitigate the new risks they
face and achieve their aspirations, and I look forward to working
closely with all of you to do just that. As Ive said before, the JEC
will seek insight and advice from the best. Thats what we have to
offer here again today.
With that, let me turn it over to the Vice Chair of the Joint Economic Committee, my friend and colleague from New York, who
has done an excellent job on so many different issues, Chairwoman
Maloney.
[The prepared statement of Chairman Schumer appears in the
Submissions for the Record on page 29.]

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00006

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

3
OPENING STATEMENT OF HON. CAROLYN B. MALONEY, VICE
CHAIR, A U.S. REPRESENTATIVE FROM NEW YORK

Representative Maloney. I thank the Chairman for yielding,

and I thank him for his leadership on the issues before this Committee today. Ive known him for 20 years, and hes always been
a champion of the middle class and working for economic opportunities and for jobs for working men and women. I thank you,
Chairman Schumer.
Im pleased to welcome Director Orszag on our panel of witnesses
to talk about the critically important issue of income instability
and what we can do to help families manage the economic shocks
that we may experience. As Director Orszag points out, wild swings
in the overall economy have been tempered, but the same cannot
be said for the economic circumstances of families trying to adapt
to a dynamic global economy.
The Congressional Budget Office has looked at this issue and
found that households experience significant ups and downs in
their earnings and income from year to year, and the downside
problem may be getting worse due to the forces of globalization and
technological change. But not surprisingly, the income roller coaster is a particularly rough ride if you are less educated.
Our second panel of witnesses will touch on various proposals to
address income instability. I know that they, like the members of
this Committee, will be coming at these issues from different perspectives, but I look forward to a very serious policy discussion and
competition among ideas. One idea we will focus on today is wage
insurance. Our Chairman is planning to introduce legislation on
this issue which will no doubt generate further useful debate about
what is the best way to deal with the adverse side effects of economic change. Dr. Brainard has offered a wage insurance proposal
with her Brookings colleagues to provide economic incentive for
more rapid re-employment and on-the-job training.
I certainly agree with the goal, but not necessarily the game plan
for getting there. As Dr. Brainard observes, our nations safety net
hasand I quotemore holes than nettingend quote. Which is
why I think we should mend it before we make it bigger. Mr.
Emsellem urges, wage insurance may well have a role to play, but
implementing it should not come at the expense of shoring up the
unemployment insurance system or trade adjustment assistance,
both of which are in dire need of reform.
Finally, CBO data show how the Tax Code can exacerbate income volatility, especially for low income taxpayers. Professor
Batchelder proposes novel changes to the Tax Code so that low income families, whose incomes tend to fluctuate the most, could average their income over 2 years to smooth out variability and enjoy
similar tax advantages as businesses in their ability to shift unused deductions and exemptions.
As an aside, I want to note that CBO examined earnings and income volatility using the Survey of Income and Program Participation, the so-called SIPP, a leading source of comprehensive data on
the income well-being of American families. Last year, there was
an effort by the Administration to eliminate this very important
program that gathers this important information between the

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00007

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

4
haves and the have-nots without having an adequate plan in place
to collect this invaluable information.
The status of the SIPP remains up in the air. I hope, Director,
you will join the growing chorus of researchers and academics who
have called upon the Administration to preserve this survey until
a better one can be designed and implemented.
I look very much forward to the testimony of our witnesses and
their thoughts on policies that can help Americas working families
better manage income instability, and I congratulate the Chairman
for his leadership on this issue, and on his book.
[The prepared statement of Representative Maloney appears in
the Submissions for the Record on page 30.]
Chairman Schumer. Thank you, Congresswoman Maloney.
Senator Casey, would you like to make an opening statement?
STATEMENT OF HON. ROBERT P. CASEY, JR., A U.S. SENATOR
FROM PENNSYLVANIA

Senator Casey. Just very briefly, Mr. Chairman. Thank you for
putting this hearing together and bringing together the great witnesses.
Director Orszag, we appreciate your appearance here, as well as
the others.
And I also want to commend Vice Chairman Maloney for her
leadership of this Committee as well.
I dont have a long statement other than to say what were talking about here today I think is a reality for a lot of families across
the country. Certainly in the state that I represent, Pennsylvania,
weve got a lot of families that struggle with this issue. Very rarely,
if ever, has it been dealt with here in Washington. I think often
the message from Washingtoncertainly the message the last 5
years in my judgmentI hate to say this, but I think its the
truththe last 5 years what families have heard, working families
have heard from Washington is unfortunately you are on your own.
We tell them theyve got to achieve a high level of education; we
dont help them enough to get there. We tell them theyve got to
train themselves and re-train and get new skills. The Federal Government doesnt do near enough. We tell them to go to college; we
dont help them with that.
This government has not helped them on the issue of health care.
So over and over again weve told families, this government has
told families youre on your own. This hearing today is one example
of where the Federal Government can respond positively and constructively to the reality that families face in terms of their income
instability. Thats why Im grateful for the appearance of our witnesses and the testimony and scholarship theyll bring to the recommendations that they make and the assessment of the problem.
And I want to thank Chairman Schumer and Vice Chair Maloney
for bringing this together. Thank you.
Chairman Schumer. Thank you, Senator Casey. Thank you for
the outstanding job youre already doing here.
Lets go right now to our first witness on our first panel. We welcome CBO Director Peter Orszag. He needs no introduction to anyone whose followed economic policy in this country over the past
decade. Peter is the new Director of the CBO. Before joining CBO,

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00008

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

5
Dr. Orszag was the Deputy Director of Economic Studies at the
Brookings Institution and Director of the Hamilton Project. He also
served as Special Assistant to President Clinton for Economic Policies, Senior Advisor to the National Economic Council and Senior
Economist on President Clintons Council of Economic Advisors.
We know you have a tight schedule and stretched it to be here.
We very much appreciate your being here. Well try to keep our
questions to a minimum so you can get onto your next subject.
Director Orszag.
STATEMENT OF HON. PETER R. ORSZAG, DIRECTOR,
CONGRESSIONAL BUDGET OFFICE

Director Orszag. Thank you very much, Mr. Chairman, Vice

Chair Maloney, other members of the Committee.
My testimony this morning examines volatility at both the macroeconomic and the household income levels, along with the role of
the Tax Code in smoothing income for both the macroeconomy and
for households and workers. It makes three main points:
First, macroeconomic volatilitythat is, the ups and downs of
the overall economy, overall economic growth and inflationhas
declined and is now relatively low. In particular, year-to-year fluctuations in the economy have become smaller than in the past.
You should have a handout with some charts. The first chart
shows that economic growth volatility has declined, basically falling in half over the last 20 years relative to the 1950s, 60s and 70s.
Several potential explanations have been put forward for this socalled great stabilization. Among the leading explanations are that
a more flexible economy, itself reflecting developments such as improvements in production processes and investments in information
technology, have made it possible for the economy to adjust more
smoothly to changes in the availability of goods and services. As a
result, the macroeconomy can more easily adapt to shocks, such as
the energy price shock of 2004 and 2005, without large changes in
output or large jumps in inflation.
Another explanation is that financial innovation since the 1970s
has provided alternatives to lending by banks, broadened opportunities for financial intermediation between borrowers and lenders
and enhanced risk management. The result has been more stable
financing for both businesses and households and more resiliency
in the financial system, also leading to a greater middle-class stabilization.
The second main point of my testimony is that, despite the relatively modest volatility that we now see in the overall economy,
workers in households still experience quite substantial variability
in their earnings and income from year to year. CBO has undertaken new empirical analysis to explore earnings and income volatility. Between 2001 and 2002, and, after adjusting for inflation,
one in four workers saw his or her earnings increase by at least
25 percent while one in five, 20 percent of workers, saw his or her
earnings decline by at least 25 percent.
The second table in my handout shows that workers with less
education tend to experience more volatility in their earnings than
do workers with more education. That is a pattern that has been
shown in other studies as well.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00009

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

6
Such fluctuations in earnings can result from many sources, including job changes, voluntary exits from the labor force for reasons such as to care for children or other family members, changes
to the number of hours worked per year or changes in the wage
rate received by workersand more work is needed in order to
parse out the causes of this volatility.
The third point of my testimony is that the figures I just referred
to are for before-tax earnings and income. The Tax Code can play
a very important role in smoothing income variability, both at the
macroeconomic level and at the microeconomic level. At the macroeconomic level, economists have long referred to the tax system as
an automatic stabilizer. That is, the Tax Code offsets part of shifts
in demand in the economy and thereby helps to smooth overall economic performance.
At the household level, the final chart that Ive given you provides an example of how the Tax Code helps to smooth income.
Consider a single worker earning $45,000. The worker owes roughly $5,700 in Federal income taxes and about $3,400 in payroll
taxes, leaving about $36,000 in after-tax income. If that workers
wages were to fall by $9,000, after-tax earnings would fall by
$6,672, a lot less than the $9,000. That illustrates the role of the
Tax Code in offsetting part of the income volatility that exists on
a pre-tax basis.
This insurance provided by the tax system, though, can come at
a price for households with variable income, in particular households whose income varies year by year will wind up paying more
in taxes over their lifetime than other households who have the
same lifetime earnings but steady income from year to year. Various options for changing the tax system would alter the tradeoff
between the income insurance provided by the Tax Code and the
price thats paid by households with variable income for obtaining
that insurance.
In addition to this tradeoff between the insurance provided and
the price paid by households with variable income, any risk sharing
benefits that the tax system generates must be weighed against potential costs that it can impose on the economy at large.
Comparing the costs and benefits is difficult, and a complete accounting has not yet been achieved. Nonetheless, some recent studies have attempted to assess the importance of the insurance benefit of the tax system in smoothing income volatility. Those studies
have found that, when compared with other alternatives, the insurance benefits of the current tax system may be larger than the
costs it imposes on the economy, for example, by distorting decisions to work and save.
A reasonable conclusion from this new research is that the income smoothing insurance provided through the tax system could
be quantitatively important and should be taken into account in
any analysis of the relative costs and benefits of different tax systems.
Thank you very much, Mr. Chairman.
[The prepared statement of Director Orszag (including the charts
and tables referenced above) appears in the Submissions for the
Record on page 31.]

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00010

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

7
Chairman Schumer. Thank you, Director Orszag. As usual,
your testimony is to the point, succinct, and focused on the questions we are facing. Thank you.
One of the points you make in your testimony is that part of the
reason we have greater stability in the economy as a whole is
theres greater economic flexibility. One of the consequences of that
flexibility is that for many workers employment becomes less stable, both in the amount of money earned and in actually having a
job. In other words, we have a conundrum: we have greater stability in the broader economy but greater instability for workers
and their families. Its almost as if the middle class and working
class workers are the shock absorbers to keep the economy stable.
They sort of suffer, as opposed to in the old days where they sort
of stayed pretty even and the economy went up and down more.
Given all of that, Id be remiss if I didnt ask you to comment
on the most recent economic news. In the past week, weve seen the
following: Alan Greenspan predicted the economy could slip into recession by the end of the year, we had a massive sell-off on Wall
Street yesterday, highlighting our vulnerability to market events in
China, we have caught wind of a recession in the manufacturing
sector, which declined more than anyone expected, and the recently
revised GDP number that came out just today shows our economy
is not growing as fast as we thought.
My question is two-fold: one, are you suggesting that our ability
to withstand bad economic developments requires that workers and
their families pay the price of greater instability and, two, isnt the
implication of what youre saying to us that in this new economy
were going to have to develop new policies to help families deal
with increased economic instability.
Director Orszag. Senator, with regard to your first question, I
wouldnt say that stability at the macroeconomic level requires instability for families. But what I would say to that is it is certainly
plausible that many of the forces that have created stability at the
macroeconomic level are also partially responsible for the instability at the microeconomic level. The challenge becomes how can
you preserve the benefits of macroeconomic performance, including
overall economic growth and overall stability, while perhaps reducing volatility at the household level. On that front, I would say
there are two broad approaches: one is to try to intervene in some
way before market outcomes are determined that directly affect
someones wage or directly affect a particular sector or a particular
firm. Another approach is to intervene after market outcomes are
determined with things like the earned income tax credit or with
wage insurance or other approaches like that.
My read of the evidence suggests that the more you try to do on
the direct market interventions, the more costly that macroeconomic implications are. The evidence broadly suggests that intervening through social insurance and the Tax Code is a better
tradeoff in terms of delivering better outcomes at the household
level without harming macroeconomic performance.
Chairman Schumer. In other words, acknowledge economic
forces, but then deal with them and their consequences to other
people, rather than change the economic forces.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00011

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

8
Director Orszag. Thats correct. Your second questionwhat
was your second question, again?
Chairman Schumer. It was the opposite side of the same coin,
which you sort of answered.
One other question: Its clear that one of the major differences
between the economy today and the economy of 50 years ago, is
that there is a big increase on the return of intellectual capital.
I always use the example of Henry Ford and Bill Gates. Each one
created enormous wealth, each dramatically changed the economy.
Each one should have become rich, but one of them needed many
more workers, if you will, to share his idea, than today, because the
ideas are more intangible, because we have different ways, quicker
ways of communicating, selling, et cetera.
Henry Ford created a million jobs, each making $10,000 a year.
Bill Gates created 10,000, each making a million dollars a year.
Thats a rough cut, but it has some truth to it.
Given all that, given that wealth agglomerates, it seems to be
agglomerating to the top. I should say this: Its not the fault of
George Bush or anybody else. Hes doing nothing to help it, but its
not his fault.
Its a fundamental condition of the economy. Doesnt it make
sense that we ought to look at further progressivity in the Tax
Code?
Director Orszag. Let me say a couple of things about the Tax
Code, again reinforcing the point that my testimony makes, the
Tax Code can play an important role in smoothing economic volatility at the household and worker level.
It is generally true that a more progressive Tax Code provides
more insurance at the household level; in other words, helps to
smooth income more than a less progressive Tax Code.
Obviously, a choice as to the appropriate level of progressivity,
is up to policymakers like you, but the point of my testimony is
that you should take into account, another important factor.
Traditionally, people look at fairness or equity versus incentive,
in measuring progressivity of the Tax Code. I think that in an environment in which household income volatility is much higher, you
also need to take into account, this role of the Tax Code in helping
households and improving economic efficiency.
Furthermore, you also need to consider what the alternatives
would be, if household anxiety rises to such a level that there are
demands for other steps.
Returning to our previous discussion about pre-tax versus posttax, that it generates demands for other steps that would be more
costly, thats another factor that should be taken into account in
evaluating the appropriate level of progressivity.
Chairman Schumer. Thank you, Director, Orszag. Senator
Casey?
Senator Casey. Thank you, Mr. Chairman. Mr. Director, I was
struck by some of the information in your testimony, as well as the
prepared testimony, with regard to risk-sharing.
I wanted to have you speak to that. I know that in one part of
your written testimony, on page 11, you say that the predictability
of household income will affect how much value they place on the
insurance provided through the tax system.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00012

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

9
I just wanted you to speak to that in terms of how this issue not
only has a direct impact on families income and their ability and
their instability, I should say, with regard to income, but, specifically, how that affects the economy overall, their ability, for example, to take risks to make an investment, to invest in more training
or invest in a family need.
If you could just speak to that question of risk?
Director Orszag. Let me make a few comments. With regard to
the specific comment in the testimony, it obviously matters, if your
income is going to fall by a certain percentage; it matters whether
you can anticipate that and know that ahead of time.
For example, in taking this job, my income has declined. I knew
that that was going to happen. Thats a different thing than having
that happen in an unexpected way.
In terms of an insurance system, what you want is to provide
protection to people experiencing unexpected events. That was the
point of that sentence.
I would also say, Senator, that as we look at increased risk, higher levels of volatility in the economy, again, exploring the role of
the tax system in offsetting that volatility, is critically important,
and thats a basic point of my testimony.
Finally, I would just say one further thing: The role of insurance
in encouraging certain activities, the fact that youre protected
against downside risk. We know that in the corporate sector, limited liability likely played an important role in encouraging the
vast expansion of corporate activity.
The same logic may well apply at the household level. If you
know that you are protected against certain really bad outcomes,
you may be more willing to take risks in the first place.
I think that insight can apply across a wide variety of settings.
For example, it is well known that the return to college, the economic benefits of going to college, have gone up, on average.
It is also the case, and much less remarked upon, that the return
to education, the return to going to college, has become much more
risky; in other words, its become much more variable.
Some people wind up doing really well, and some people earn a
lower return from having gone to college.
Whether that affects the incentives to enroll and take the risk of
going to college, in a sense, is, I think, an important topic. I think
youre hitting on a potentially quite important issue, which is
whether this increased volatility and increased risk at the household level, is impairing some steps that would actually improve
long-term economic performance, because people dont take the
kinds of risks that they otherwise would.
Robert Schiller is a professor at Yale and has written basically
a whole book on exactly that point, that because there are not
enough different kinds of risk protection in the economy, people
dont take the risks that will lead to better economic performance.
Senator Casey. I have one more question. With regard to the
chartIm not sure everyone can see this, but the chart that you
gave us with regard to volatility as it pertains to GDP growth, for
those who cant see it, the overall umbrella, is the question of volatility.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00013

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

10
There are two time periods: 1950 to 1984 and then 1985 to 2005.
But I was really struck by both the GDP growth line and the inflation line.
Could you just talk about the 1950 to 1984s 3.1 versus the 1985
to 2005, 1.4? Theyre striking numbers.
Im not sure that when you were giving the overview of your testimony, we didnt stop and highlight them. I just want to have you
comment on the significance of just that GDP growth, that particular line of the chart.
Director Orszag. What that tells you, is, in terms of year-toyear ups and downs in the overall economy, they are much less severe now than they were in the past. That is reflected in a variety
of things.
Recessions tend to be less severe than they were in the past;
booms tend to beyou know, there tends to be less movement from
one year to the next in overall economic performance. This is referred to as the great stabilization.
Theres an ongoing question as to exactly why it has occurred,
but the evidence does suggest that at least over the past 20 years
or so, economic performance is more stable than it was historically.
Senator Casey. Thank you very much.
Chairman Schumer. Senator Klobuchar?
Senator Klobuchar. In your testimony, you talked about how
the household income has become more volatile, but that this volatility may be the price were paying for relative macro economic
stability.
At the same time, you talk about how its easier now for families
to get credit. The question I had, just from cases that Ive seen
when I was a prosecutor and things that Ive seen in our own community isis it beneficial to the economy as a whole or to families
to offer credit to low- and middle-income families, thus increasing
their debt levels when they have little chance of repaying the debt?
Director Orszag. The spread of financial products allows households to adjust in ways they werent able to before, but whether
that opportunity is used well and used prudently and used in a
sound way, is a harder question. I think we would all, if our income
falls by 40 percent, welcome the opportunity to be able to cushion
the blow, at least temporarily, by having access to borrowing.
Whether, however, borrowing is done over a longer period of
time, in an unsustainable way, is a harder question.
The concerns Senator Schumer raised about developments recently, which may be tied to the mortgage market, is a reflection
of ongoing concerns about the degree to which those financial possibilities are being exercised in a sound, prudent, and responsible
way.
Senator Klobuchar. How about the debt? What does this mean
for the long-term stability of our families?
Director Orszag. A few comments: The financial obligation
ratio, that is, basically the payments that households have to make
on debt relative to income, as calculated by the Federal Reserve,
is significantly higher than it was in the past. On the other hand,
net worth is also higher, relative to income, than it was in the past.
I think that the key question on debt, is whether there is a significant mismatch between debt and assets and income that can fi-

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00014

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

11
nance that debt. Obviously, there are concerns about that mismatch.
What I say, more broadly, though, not just at the household
level, or net national savings rate is now bouncing around somewhere around zero and 2 percent or zero to 2.5 percent. That level
of net national savings, necessarily implies one of two things: Either were only going to be investing that amount domestically, or
were going to be borrowing the difference from abroad.
Increasingly, what we are doing, is the latter, borrowing the difference from abroad. That is not, however, free money. We are running a very large current account deficit and accumulating significant liabilities to foreigners.
That effectively imposes a burden on future American generations, because they will not enjoy the full returns of investments
that are being made today.
Senator Klobuchar. I think, if I remember the statistics, 1 out
of 12 Federal tax dollars is going to interest on the debt and a lot
of it is going to foreign entities, foreign countries.
Again, it concerns me for the long-term stability of the country
and for our families.
Last, in your written testimony you mentioned that research on
the rise of wage volatility and income instability is currently lacking and that this lack of research makes it hard to reach firm conclusions about the significance of these trends.
Is this something that this Committee could help to push along?
Director Orszag. Yes. Let me sort of reinforce that. I think that
a variety of evidence suggest that volatility at the household level
is now higher than it was in the 1970s. Exactly what the time pattern was, whether it went like this or went like that, I dont think
we know enough to conclude.
The Congressional Budget Office will be looking into this issue,
and I would welcome your continued interest in that topic. We have
the ability to use other datasets and do other work than some private academic researchers do. Im very much eager for us to do
that.
Senator Klobuchar. Thank you.
Chairman Schumer. Senator Webb?
Senator Webb. Thank you, Mr. Chairman. Im trying to find a
mike here. Youve got me sitting in a place where I cant be heard.
Chairman Schumer. Not by design, I assure you.
Senator Webb. Im not so sure.
[Laughter.]
Senator Webb. Im sorry to have arrived at the end of your testimony or after your testimony was given, Director Orszag.
I have one question about your chart, and then just a general
question. This distribution of changes in workers annual earnings
that you use in your chartand, by the way, as a writer, I know
a great deal about income instability. Some years are good and
some years arent.
But you use changes in earnings from 2001 to 2002. Im wondering if those years are representative of whats going on now and
what was going on before then.
This was the year of 9/11, and there was an enormous jolt in
many sectors of the economy, and the absorption of that jolt was

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00015

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

12
felt in different places in the economy. So Im wondering why you
used those 2 years and whether they are representative.
Director Orszag. We also did the analysis for 2 years in the
mid- to late 1990s and found similar results, so were more confident that these numbers are representative of some recent level
of volatility.
Obviously, as we try to expand our work in this area, well be
able to add more years and more information, including from other
datasets, but I think, again, a variety of evidence from different
data sources and from different years, suggests that at the household and worker level, there is a significant amount of volatility.
Senator Webb. Do you have numbers at hand from years after
2002?
Director Orszag. No, I dont have that.
Senator Webb. It will be interesting to see those, because, as
you know, there were some huge job shifts that took place into the
government sector, regionally, and otherwise. It would be interesting to see how those impacted with respect to not only wage levels, but education levels, different kinds of jobs, not only the jobs
that came out of the homeland security environment and the spinoffs, but jobs that were lost because of the internationalization of
corporate America.
Were seeing in this country, perhaps a steady line in terms of
numbers of jobs, but, really, its a different situation in terms of the
quality of a lot of those jobs.
Theres no data that you have, that would be able to reflect
whats been going on in the last 3 or 4 years?
Director Orszag. We will be able to provide you updated information, and, as was earlier mentioned, these data are based off of
a particular dataset and there are some lags involved, so that we
dont have instantaneous access to the most recent information.
Nonetheless, we can provide something thats more recent than
this, and part of our expanded activity will be do exactly that prior
to giving you those results.
Senator Webb. Someone may have asked this question before I
got here, but I have heard it said that a lot of the proponents of
wage insurance, argue, either directly or implicitly, that it would
encourage workers to take lower-paying jobs.
Would you say thats correct?
Director Orszag. Theoretically, it could have that effect. The
evidence that we have from an experiment in Canada, however,
suggest that, in practice, it either has limited or very minimal significance.
The basic theory that if youre cushioning the blow, people may
be more willing to take a lower-paying job, seems correct, but in
practice, it doesnt seem to be that consequential.
Senator Webb. Thank you.
Chairman Schumer. Thank you, Director Orszag. I would
askI dont know if we need to send a letter to CBO, but following
up on Senator Webbs questionI have the same oneis it possible
now for you to do an analysis of some of the later years, in terms
of volatility?
[A letter to Director Orszag from Chairman Schumer and Senator Webb appears in the Submissions for the Record on page 40.]

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00016

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

13
Director Orszag. I think so.
Chairman Schumer. The Committee is going to request that
you do so, and you can get us an answer, either with numbers or
saying you cant do it, until you have the following data. Thank
you.
[A response from Director Orszag to Chairman Schumer and
Senator Webb appears in the Submissions for the Record on page
41.]
Director Orszag. Thank you very much.
Chairman Schumer. Thank you, Director Orszag. Youve done
a great job, as usual. Let me call our next panel forward.
Thank you all for coming. In our second panel, were going to be
hearing from Lael Brainard, vice president and founding director
of the Global Economy and Development Center at the Brookings
Institution. Dr. Brainard served as deputy national economic advisor and chair of the deputy secretarys Committee on International
Economics during the Clinton administration.
As deputy director of the National Economic Council, she helped
build the new White House organization to address global economic
challenges such as the Asian financial crisis, and Chinas WTO
entry.
Maurice Emsellem is the public policy director of the National
Employment Law Center. Mr. Emsellems areas of specialization
are: government systems of support, including unemployment compensation, workforce development programs, and the welfare system. Hes published extensively on the unemployment system.
Professor Lily Batchelder is assistant professor of law and public
policy at NYU School of Law. Her current areas of research include: tax incentives, wealth transfer taxation, income volatility
and social insurance. She previously practiced law at Skadden
Arps, as well as a Wiener Fellow at the Wiener Center of Social
Policy at the Kennedy School.
Dr. Bradley Schiller is a professor at the School of Public Affairs
at American University here in Washington. He not only teaches
economic theory to students and practitioners in public policy, but
practices it as well. His specializations include public policy analysis in economic policy as a consultant to governments and major
corporations. Hes designed, evaluated, and even operated scores of
employment training and welfare programs.
Dr. Schiller also lectures extensively on Social Security reform
and the Federal budget.
I want to thank all the witnesses for being here. We should have
a lively discussion. You each have 5 minutes, but your entire statements will be included in the record. Dr. Brainard, you may begin.
STATEMENT OF DR. LAEL BRAINARD, VICE PRESIDENT AND
DIRECTOR, GLOBAL ECONOMY AND DEVELOPMENT, THE
BROOKINGS INSTITUTION, WASHINGTON, DC

Dr. Brainard. Chairman Schumer, Senators Klobuchar, Casey,

and Webb. Its a pleasure to be here. Its a pleasure to be focusing
on an issue which I think is a reality that confronts more and more
middle-class Americans and hasnt received much attention in the
last few years.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00017

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

14
Ill only spend a minute and a half on the backdrop, some of the
forces, maybe, that are leading to increased insecurity for many
Americans today, and then talk about one possible proposal to add
to the arsenal to help Americans cope with uncertainty.
In terms of the backdrop, one of the inevitable forces affecting
the economy today, is that we are, I think, experiencing a new
wave of globalization. Weve experienced waves of globalization before.
This one, I think, has familiar elements, but the scale, the scope,
and the speed, are something we have not experienced before, partly because China is pursuing, at a scale that has never been done
before, a growth strategy thats very export-led, and foreign directinvestment-fed, which means that it has repercussions for every
economy in the global marketplaceand especially in manufacturing.
Chinas entry is confronting higher-wage competitors with a
stark choice: You either move up the value chain or cut costs. That
challenge, I think, is made much more complex by the concurrent
emergence of India.
India is a very different story, very much more consumption-focused, but confronting white collar workers with foreign low-wage
competition for the first time, a very different playing field than
theyre accustomed to.
And then if you look at the scale of those two economies together,
essentially in a very short period of time, were being asked to absorb a 70 percent expansion of the global labor force. Any textbook
model would tell you that that kind of expansion, while capitaland investment- and technology-adjusted, is going to put a little bit
of pressure, at minimum, on wage earners at the middle.
In fact, thats what were starting to see; were starting to see the
evidence that inequality is increasing and that the earnings at the
middle, the gap between the middle and the top, is increasing.
Whats the answer to some of this? Well, a piece of the answer
has to be to strengthen social insurance. Despite the fact that, as
Director Orszag was testifying earlier, in the U.S. context, theres
a lot of flexibility, there is a lot of churn at the worker level, and
our social safety nets are the weakest among the rich economies.
If you look at what happens to permanently displaced workers,
they can experience average earnings decline of 16 percent when
they are reemployed. For manufacturing workers, that average
earnings drop can be 20 percentvery substantial.
In import-competing industries, the numbers are even more
stark. What are the social programs that we have to address that
difficulty?
The reality is, theyre pretty thin and theyre pretty old, so, unemployment insurance, as everybody knows, has a lot of holes in
it, to the extent that only about 40 percent of workers are actually
in that system.
The other system that we have developed was back in the 1960s.
President Kennedy developed trade adjustment assistance. It had
laudable goals, but the reality is that, unfortunately, for complicated reasons, most workers dont actually qualify for benefits,
and those that do experience long periods out of work and the same
kind of income declines when they come back into work.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00018

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

15
What else could we possibly add to the arsenal? I think one thing
we should be taking a serious look at is wage insurance, and let
me just state right up front that the intent here is not to replace
other programs, but to augment the existing safety net.
Vice Chair Maloney was talking earlier about mending the existing programs. I think we need to mend and expand. I dont think
its an either/or; I dont think it should be.
The point of wage insurance is to essentially help workers by insuring them against that kind of steep decline in earnings following permanent displacement, and the kind of displacement that
we would be talking about would be factors outside of a workers
control.
The idea would be to essentially subsidize their initial salary on
the new job for some period of time, while their attachment to the
new job improved, while they might acquire on-the-job skills to
make them more productive, more valuable to their new employer.
There are different ways that you can structure such a program.
We have a policy brief that Ive put on the back table that gives
you a variety of different parameters, but, obviously, the key parameters are how much of the earnings loss you are going to replace, what kind of a cap you put on that compensation each year
and that, of course, affects who is more likely to qualify for these
benefitsand, of course, what the duration of that is.
Let me just give you an example of a program that would replace
about 50 percent of an earnings loss, up to a maximum of about
$10,000 a year. That kind of a program, we estimate, would cost
on the order of about $3.5 to $4 billion per year.
And if you think about an average trade-displaced worker who
earned $37,000more than that in 2004their experience, generally, in 2004, was a 26 percent drop in earnings, so if you insured
half of that, youd essentially be receiving $33,500, rather than
$37,000. Thats $6,000 a year, a big change in terms of smoothing
that income on a per worker basis.
Its an insurance program that would essentially amount to
about $25 per worker per year.
There are a variety of ways you could implement it through the
existing unemployment insurance program, or through the refundable tax credit. There are a variety of ways you could do it.
So the bottom line, I think, here, is wage insurance is a potentially important tool to be added to the arsenal of available tools.
I dont see, again, any reason that there has to be a tradeoff between existing, improving existing and expanding.
[The prepared statement of Dr. Brainard appears in the Submissions for the Record on page 74.]
Chairman Schumer. Thank you, Dr. Brainard. Well now turn
to Mr. Emsellem, who has a different point of view. Id ask each
witness to try to stay within the 5 minutes, so get to the heart of
your arguments.
STATEMENT OF MAURICE EMSELLEM, PUBLIC POLICY DIRECTOR, NATIONAL EMPLOYMENT LAW PROJECT, OAKLAND, CA

Mr. Emsellem. Thank you, Senator, members of the Committee.

Our organization, the National Employment Law Project, specializes in economic security programs. We have a long history of serv-

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00019

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

16
ing families hard-hit by economic downturns, by helping them access the benefits and by promoting innovative policies that deliver
on the promise of economic opportunity.
Wed like to offer our perspectives on proposals to create a new
program of wage insurance for dislocated workers, then highlight
other options for Federal reform to protect communities struggling
to cope with the realities of todays unstable economy.
Like the AFL-CIO and several major unions that have expressed
serious concerns with wage insurance, we believe there are far too
many unanswered questions that convince us its not the right time
to move ahead with a national wage insurance program.
First, its important to ask the question that Senator Webb
posed, whether wage insurance will promote more downward mobility for the nations most vulnerable workers, since, by definition,
wage insurance is only available for jobs that pay less than they
earned before and are less likely to provide health insurance and
other critical benefits.
We believe the limited Federal resources devoted to the economic
security of Americas workers, should promote good employment
outcomes and quality jobs, but thats not the case with wage insurance.
Were also not aware of any empirical evidence that wage insurance jobs will promote transferable skills or meaningful training.
Workers are usually employed full-time to qualify for wage insurance. The program may actually preclude workers from pursuing
education and training they need to compete for better jobs in todays economy.
Second, does the experience with actual wage insurance programs make a convincing case that now is the time to create a new
national program?
What we know from the only major evaluation of a wage insurance program, the Canadian pilot program, is that it failed in most
areas to achieve its intended results, thus, the Canadians never
adopted wage insurance.
I could say more in response to Senator Webbs question about
the issue of downward mobility and what they found there. They
found that at the low end of the wage scale, in fact, folks who accepted wage insurance, did take lesser-paying jobs.
It evened out, when you looked at the various wage scales, but
thats all we know, that limited information about the impact on
wages, of wage insurance. Theres no information on the impact of
benefit and other critical criteria.
Were still waiting for the results from the U.S. pilot program
serving trade-impacted workers over age 50, although we know
that participation has been limited.
Another question that has not received enough attention, is,
what impact will the program have on other workers who are competing for similar jobs, with those collecting wage insurance?
A leading researcher with the Upjohn Institute found that, quote,
Virtually all the employment gains experienced by dislocated
workers, as a result of the wage subsidy, came at the expense of
other workers. Will this crowding-out effect be even more severe
in those communities in the Midwest and elsewhere, where there
are already large concentrations of dislocated workers?

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00020

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

17
In addition to the research questions, theres also a concern that
wage insurance could undermine those Federal programs that now
provide some measure of economic security to U.S. workers. For example, will major funding and support for wage insurance, take
precedence, in reality, here in Congress, over long-delayed reforms
of the TAA and UI programs?
Certainly, if wage insurance is funded by the Federal UI payroll
tax or other similar sources of revenue, it will effectively compete
for funding at a time when unemployment insurance desperately
needs to be modernized.
Were also concerned with the precedents wage insurance will set
when hostile groups like the Heritage Foundation, are on record
strong supporting wage insurance as a rapid reemployment substitute to dismantle the TAA program. Will wage insurance set the
stage for more attacks on TAA, which is up for authorization this
year and the UI program?
These are some of the questions that leave many of us who work
with these programs convinced that a national program of wage insurance could do more harm than good.
So, what are some other priorities for Federal reform to create
a reemployment system that promotes quality jobs? First, we urge
Congress to fulfill the promise of economic security to the nations
workers, who have suffered major job losses due to Federal trade
policy.
That means reforming the TAA program, starting by removing
the $220 million cap on training funds that has left many states
forced to suspend or deny enrollment to thousands of eligible workers. In fact, we have an office in Michigan and there we know that
just a few months into the funding cycle now for TAA, they have
already obligated the State of Michigan all its current TAA training funds due to the massive layoffs in the auto industry. That
should be a priority.
Second, its time to finally modernize and expand unemployment
insurance by making Federal incentives available to states that fill
the gaps in the program, which now deny benefits to most lowwage and women workers and to support education and training
with the help of extended unemployment benefits.
Were not talking about training, just for the sake of training,
but training of the sort that many states are pursuing, like industry-sector initiatives that have proven successful in improving employment outcomes and making their industries more competitive.
Finally, Congress should support or replicate some of the most
promising innovative state strategies, like self-sustaining home
protection funds that prevent foreclosures, healthcare coverage for
those who qualify for UI benefits, and model training partnerships
with business and labor to help save good jobs. Thank you for your
interest and commitment on these issues.
[The prepared statement of Mr. Emsellem appears in the Submissions for the Record on page 77.]
Chairman Schumer. Thank you, Mr. Emsellem, thank you for
being right on point in answering the questions directly. Professor
Batchelder?

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00021

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

18
STATEMENT OF LILY L. BATCHELDER, ASSISTANT PROFESSOR
OF LAW AND PUBLIC POLICY, NEW YORK UNIVERSITY
SCHOOL OF LAW

Ms. Batchelder. Thank you very much for the opportunity to

testify today. What I would like to do, is briefly discuss who is burdened by income volatility and how income tax simultaneously
helps and hurts these families, then outline two proposals that
could be implemented on a revenue-neutral basis, each of which
could increase the way that the tax system helps families cope with
income fluctuation, and decrease the ways in which it penalizes
them.
As the other panelists have discussed, household income volatility is extensive, and the evidence to date suggests that it has
been rising. This is a serious problem and can increase families
stress; it can force families to incur additional living expenses, for
instance, if they move more often, and these hardships are greatest
for middle- and low-income households, because they typically have
less savings which can serve as a buffer, and, relatedly, they typically have less access to low-cost borrowing.
Unfortunately, as you can see in the first figure in my written
testimony, these are also the families that face the widest swings
in their income. Income volatility is essentially disproportionately
affecting the families that are least able to cope with it.
Currently, the income tax simultaneously helps and hurts these
families. It helps because a progressive income tax results in families paying more tax in relatively good years, and less tax in relative bad ones, so it essentially softens income fluctuations on an
after-tax basis.
It hurts, because, over time, it imposes higher tax burdens on a
household that has relatively volatile income than it does compared
to another household that has the same average income, but a
household that earns it more smoothly.
I refer to these higher tax penalties as fluctuation penalties.
They basically result from families with unstable income, being
bumped into higher tax brackets that would never have applied, if
their earnings were more stable.
Like income volatility, these penalties are largest for middle- and
low-income families. They can actually be immense for low-income
households who have the most volatile incomes.
I think we can do better than this in the tax system. We should
be trying to make the income tax impose more equal tax burdens
on households with volatile incomes, and we should deliver that tax
relief in years when their income has actually declined.
So, as Director Orszag mentioned, one way you could view what
the tax system is currently doing, is these fluctuation penalties are
essentially like premium payments for the income insurance that
its providing by smoothing income on an after-tax basis.
Our goal should be to increase these income-smoothing benefits,
while decreasing the premium payments. One promising way that
we could do this, is a limited form of income-averaging, which I
refer to as targeted averaging, which would allow families to carry
back the standard deduction and personal and dependent exemptions for 1 year, and also to average their income over 2 years for
purposes of calculating their earned income tax credit.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00022

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

19
What this would mean, is, if a familys breadwinner lost his or
her job or was a writer and had a bad year, and their income fell
to a point that they couldnt use all of their personal deductions,
they could apply them to the previous year and receive a refund
in the current year for that amount. They also might be eligible for
a larger EITC.
This is a relatively modest and administrable proposal, but the
simulations Ive done, suggest that it actually would eliminate a
substantial share of the fluctuation penalties that the income tax
creates, and it would make the tax system better at smoothing
after-tax income, because families would generally only benefit
from this in years when their income has declined.
The second complementary, but much broader approach that I
would like to mention, is the possibility of transforming individual
tax incentives into uniform refundable tax credits.
Currently, we provide about $500 billion a year in tax incentives
intended to encourage households to spend or invest their money
in ways that we consider socially valuable.
I want to pause and emphasize, as you grapple with on a much
more daily basis than me, that $500 billion is a really big number.
Its close to 4 percent of GDP, about equal to our outlays for the
Department of Defense last year. Its about half of individual income tax revenue raised last year, and the vast majority of these
tax incentives are structured as deductions or in other ways where
their value rises, the higher income you are. It depends on your
marginal tax rate.
What these types of tax incentives do, is actually create fluctuation penalties, beyond those that exist in the regular income tax.
Whats worse, is, they mean that people get the biggest incentive
in their relatively good years and the smallest incentive in their
relatively lean years.
So, unlike the rest of the income tax, these types of tax incentives arent sort of simultaneously helping and hurting families
with unstable incomes; theyre only hurting them.
If they were structured as uniform refundable tax credits, which
could be done on a revenue-neutral basis, they wouldnt generate
these fluctuation penalties, because they would be worth the same
amount every year, and they wouldnt provide smaller benefits in
years when a familys income decline.
I also think they would be more fair and efficient for other reasons I can go into. In short, I think both of these proposals are
worth serious consideration as ways that the tax system deals with
families with unstable income and I look forward to your questions.
[The prepared statement of Ms. Batchelder appears in the Submissions for the Record on page 81.]
STATEMENT OF DR. BRADLEY R. SCHILLER, PROFESSOR,
SCHOOL OF PUBLIC AFFAIRS, AMERICAN UNIVERSITY

Dr. Schiller. Thank you for this opportunity to testify on this

important subject. In addition to being a professor at American
University, Im sure Chairman Schumer will take note that Im the
author of three very expensive college textbooks and so will be regarded as a hostile witness here.
Chairman Schumer. Perhaps we can make them deductible.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00023

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

20
Dr. Schiller. That would be nice. Maybe the hostility level will
rise when I offer my message, which is to praise income volatility,
not to bury it. The United States would be a far less vibrant economy if we go too far in trying to make incomes stable. We would
look more like the French than the entrepreneurs of Silicon Valley.
Let me remind you of the French riots of last spring. French
workers have always had tremendous wage insurance; in fact,
theyve had pretty much lifetime income security through a package of income growth, guaranteed fringe benefits and generous pensions. What sparked the riots in Paris last year was a proposed increase in resource mobility, specifically a very modest proposal that
would allow employers to fire newly hired workers within 1 year
for any reason for workers under age 26. Well, the French youth
viewed that as a threat to their own income security and took to
the streets.
The important point I want to make is many of those French
youth have stayed in the streets because the French system of
wage insurance and guaranteed lifetime incomes puts an enormous
cost on employers, and employers are very reluctant to hire, therefore, new entrants. Youth unemployment in France hovers around
24 percent, more than double the U.S. levels. The French economy
is growing half as fast as the United States and the French middle
class has incomes 25 percent below American levels. So I ask you
at the beginning how many Americans do you think would trade
American prosperity for French income stability? The reality is the
resource mobility, specifically labor mobility, is a critical factor in
the advance of the U.S. economy.
Wal-Mart hires dozens of new workers every day. And I know
youre not fans of the Wal-Mart employment model, but how about
Google? Google hired 2,000 workers last year, Genentech hired
2,000 workers last year. XM and Sirius satellite hired over 1,500
workers in the last 2 years, many of them engineers. The
healthcare industry has created 3.5 million jobs in the last 10
years. Colleges and universities and high schools have created 2
million more jobs.
So who are filling all of these jobs? We have 2 million new entrants into the labor market every year. Those are mostly teenagers and immigrants. They may be getting the jobs at Wal-Mart;
theyre not getting the jobs at Google, Genentech or SM and Sirius
satellite. Employers for those large corporations want workers with
experience, skills and employment references. Where are they getting the workers?
The answer is theyre getting the workers from firms and industries that are in decline. The telephone industry has shed tens of
thousands of workers. The auto industry has done the same thing
and is now embarked on another wave of dislocations. Real estate
brokers and mortgage bankers are now looking around for new
jobs.
So the point is the American economy thrives because were able
to move people out of declining industries into expanding ones. Is
this job mobility good for the economy? Absolutely. Its what makes
us so responsive to changes in technology, changes in trade, and
changes in consumer taste. Without such mobility, our incomes
might be more stable but theyd also be lower.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00024

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

21
So to get down to the policy implications here, Im not denying
that theres economic deprivation that results from income instability, but I would emphasize that economic depravation tends to
be a relatively brief experience. Contrary to what youve heard so
far, there is no evidence that income stability at the household
level has increased. What you heard from Director Orszag is that
macroinstability has decreased and we now have some numbers on
household instability. But there is no evidence that household instability, income instability, has increased.
Most of our instability, particularly the downward dislocations,
are of relatively brief duration and our safety net programs are, for
the most part, time limited. Unemployment insurance tops out at
26 weeks, our welfare benefits top out after 5 years, our trade adjustment assistance is time-limited as well, and I think these are
appropriate responses to the instability that exists.
Im not saying that we should not take any further steps to improve the social safety net, but I do want to advocate some cautionblink a yellow lightand point out that any further expansions of the social safety net carry a risk. They impose higher costs
on employers which will make them more reluctant to hire displaced workers and new entrants to the labor market and they may
create additional disincentives for the workers themselves to take
on new jobs in expanding industries. Its far better for the worker
to grasp toe-holds in an expanding industry than to cling to a job
thats in decline.
Thank you.
[The prepared statement of Dr. Schiller appears in the Submissions for the Record on page 91.]
Chairman Schumer. Thank you. I want to thank all of our witnesses. We had a multiplicity of viewpoints here, and everyone got
to their points. It was great.
My first question goes to Mr. Emsellem about wage insurance.
First, the bulk of your argument seems to be that this will take
away from other programs, particularly trade adjustment assistance which has been notoriously poor. We had the exact same experience in New York. Workers are laid off, they qualify for trade adjustment TAA, but they dont get any money. This has happened
over and over again, in Syracuse and Rochester and Buffalo.
But just for the sake of argument, lets say that we had a good
trade adjustment assistance program. Lets say we had a good unemployment assistance program. Would you then object to a wage
insurance program to augment those rather than replace those in
terms of either substance or dollars?
Mr. Emsellem. At this point, yes. As I mentioned, there are a
lot of other unanswered questions. The point is, we should be pursuing a good job strategy, that is really where were coming from.
We should be pursuing strategies that promote quality jobs, and
there is no evidence that wage insurance does that. With $3.5 billion, you can put a lot of money into an initiative to promote good
jobs.
Chairman Schumer. Isnt training on the job the best training?
Its been shown over and over again that employers who hire somebody and then train somebody, thats what really advances work
more thanweve had lots of job training programs, Ive supported

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00025

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

22
many of them, where they train people for jobs and the jobs dont
exist or they cant work out in the jobsomeone on the job.
Again, this is notyou seem to be saying do it the old way weve
done it all along, put more money into all of them and heres a new
idea. That seems to me to make a good deal of sense. Not as a replacement; Im wary of that. Im careful of that admonition. But
youre saying well, its simply going to soften the blow of people,
theyre not going to take low wage jobs because of this. If someone
can findsomeone loses a $40,000 a year job and can find another
$40,000 a year job, theyre going to take it. But if the only job they
can find is a $25,000 a year job and you can say OK, were going
to say your salary is going to be $33,000 for 2 years, that seems
to me all to the good.
Ill bet if you asked average workers, theyd all be for this kind
of program. Im going to go ask some of them in Syracuse and Buffalo whove been laid off at the $40- or $45,000 jobs. Again, I dont
get the objection when we have such a problem here of saying well
were not sure it will work. Were sure a lot of other programs dont
work.
I could make the argument trade adjustment assistance doesnt
work. Were never going to get the adequate funds for that. That
pays people money without a job. What about paying people money
with a job?
Mr. Emsellem. Senator, you asked a lot of questions there.
[Laughter.]
Chairman Schumer. Im going to give you a chance to answer,
I just dont quite get it.
Mr. Emsellem. Let me try to answer a couple of your points.
First, training. There is no evidenceand thats my other point,
there are a lot of unanswered questions about this program. There
is no evidence that folks who take wage insurance because theyre
taking lesser paying jobs will receive any meaningful training or
transferable skills.
As for on-the-job training, the fact is that there are a lot of other
options out there. But were talking about jobs that involve lesserpaying work. We dont know, if you take a job at Wal-Mart
Chairman Schumer. Question: Do you think anyone takes a job
for lesser-paying work if they can get a job for the same or more
paying work?
Mr. Emsellem. Thats a different point.
Chairman Schumer. Thats the problem in the economy we
face.
Mr. Emsellem. I think its more than that. I think its a problem
with what options are available to workers today. If the only option
is wage insurance, thats one issue. If we can create better jobs,
more quality jobs, and put a real initiative into resources that do
thatand there are good training examples, lots of them in New
York; I come from New York originally, I used to live there for a
lot of yearsa lot of good examples of training that have shown
limited results on a national scale.
But if you look at the good things that states are doing, sector
kind of training, the sort of things that bring employers and business together

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00026

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

23
Chairman Schumer. Sir, I asked you a question. If we had unlimited resources, OK, and we put money into job trainingthe
kind of training before people get jobs, OK. Why wouldnt we do
this as well?
Mr. Emsellem. Thats another question. I hope Ive responded to
the question.
Chairman Schumer. But my general question is rather than
say
Mr. Emsellem. If we had unlimited resources, Senator, I think
the money could be much better spent than on wage insurance on
an initiative that promotes good jobs along the lines of what the
states are doing. Build up the sectors, the industries that are more
competitive, put the money into those programs. That helps employers and helps workers and saves jobs. Wage insurance does
none of that.
Chairman Schumer. Give me one example of where you would
put a large amount of money? I dont mean a small program that
works somewhere, but when weve spent large amounts of money
on job training the results are mixed at best.
Mr. Emsellem. Thats at a national scale, not at the individual
level where youand Im not talking about small programs. Im
talking about where you take a look at what the interesting states
are doing. In Wisconsin, theres a regional partnership set up between business and labor thats all about retooling the manufacturing industry. That has had major consequences for saving jobs
and creating training that helps these workers and helps the states
stay competitive.
Chairman Schumer. Has the number of manufacturing jobs in
Wisconsin stayed constant?
Mr. Emsellem. In those areas theyve put the resources into,
yes, its made a big difference, Senator. Let me just say in terms
of what workers think its an excellent question. Of course thats
the first question that comes to your mind.
In Canada, what we know is that 2 out of the 10 peopleonly
2 out of 10 people who were in the control group where they could
collect wage insurance took part in the program. When they did follow-up interviews to ask those folks what they thought about wage
insurance, they said its not relevant to my situation. You and me
and most folks, we have the same reaction: I want to spend my
time collecting my unemployment benefits and finding the best job
that I can find. Wage insurance, you cant collect wage insurance
if youre still on unemployment insurance.
Chairman Schumer. But each worker could have an option.
One worker might want to train and collect unemployment insurance, and another might want to take the $25, he may have real
needs, or she. They may have to fund some medical illness, they
may have to fundand youre telling them you have to take unemployment insurance or a job training program. And some of them
not all, maybe not mostwould say its better for my situation to
take the lower-paying job and at least make up half the difference.
Mr. Emsellem. Im saying the money should be spent on more
quality jobs and just also a question.
In Canada, what they found was 44 percent of the people who
took wage insurance were still on wage insurance 2 years after the

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00027

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

24
program, so they hit the max. Thats a lot of people. What happens
to those people when wage insurance is over? What have we done
to improve their situation?
Chairman Schumer. We should do all other things. But youre
saying 56 percent werent, thats pretty good.
I look at a lot of job training programs and if they have a 50 percent or over rate of success, I think theyre pretty good.
Do you want to comment on what Mr. Emsellem had to say, and
then Ill yield to Senator Webb. Dr. Brainard.
Dr. Brainard. I have to say Im a little puzzled as well by the
line of reasoning. The reality is people are taking lower-paying
jobs. Thats the reality today. Two million displaced workers losing
16 percent of their income when they are re-employed. If you look
at manufacturing, 20 percent of their incomemaybe 75,000 workers a year, maybe, end up in TAA.
The GAO just did a good report on five different locations where
TAA was administered. The portion who were in relevant training
ranged from 9 to 39 percent at each site. Unfortunately, the workers who entered trainingand Im reading herereplaced slightly
less in their wages than workers that didnt. Unfortunately, much
as I would hope that TAA would be the answer and that training
would lead to full replacement of earnings, the reality today is that
people are taking lower paid jobs.
In terms of the evidence, the evidence to me, the way I read the
evidenceand again, you know, I agree, we dont have a lot of evidence out there, so I dont want to overstretch the amount that we
can predict about what this program would do. The evidence was
that the search was more intense, that people were motivated essentially but did not provide any evidence that I could see that
they actually were motivated to take the first job that came along.
People take jobs for very complicated reasons. They look at a
whole host of factors. I dont read the evidence as suggesting that
they took a lower-quality jobagain, unfortunately I do read the
evidence for the Nation as a whole that a lot of people are being
forced into taking lower-quality jobs.
Chairman Schumer. Senator Webb.
Senator Webb. Thank you, Mr. Chairman. I have moved up to
a better mike here.
Professor Batchelder, you understand why you probably didnt
want me near a microphone.
I have a question for the Chairman. We have this bottled water
here that has no label on it. I was wondering if it was part of the
new ethics law that were not allowed to do product placement. I
find it very curious.
Chairman Schumer. Just make sure its water.
Senator Webb. I would, first of all, like to thank all the panelists for their testimony. Id like to take this in a little bit different
direction.
First of all, before I do that, Professor Batchelder, I agree with
you from a totally different set of experiences that income averaging has its benefits, even if were not looking at the way that incomes can fluctuate as a result of the sorts of issues were talking
about today.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00028

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

25
Again, as I said before, as someone thats in a profession where
your income does vary greatly year by year, I found that when we
did this that it was quite beneficial to be able to do basic planning,
basic financial planning.
But Dr. Brainard, I believe that you are at the right starting
point here. When were looking at these issuesand I appreciate
all your comments talking about the impact of globalization on
what were doing. Wherever people come down on these issues, I
dont think that theres a great deal of dispute in this country
about the facts, specifically what has happened with the internationalization of corporate America. The question is where do we
go for the answers on this issue?
My view is that we have to come up with some sort of a competitive economic model that includes a strategy for good jobs for
American workers so that they benefit and get a fair share of the
growth of this economy. The data doesnt really show that happening.
Mr. Emsellem, I take your point when youre talking about how
Federal trade policy has contributed to this situation. If I were
going to start with tax fairness, I would start with tax fairness in
corporate America. Thats not a slam on corporate America, just
that corporate America has been able to take advantage of the
trade practices that have gone through our government and, in
many cases, American workers were not included in terms of being
protected as we moved into the WTO.
And another thingand this goes basically to the question Id
like to throw at the panelit seems to me that we need as a nation
to come up somehow with a different way to make adjustments for
how were paying for pension and medical expenses that allows us
to be more competitive and, over time, actually give a different
kind of relief for our corporations.
I am mindful of an article that was in The Economist last summer. As you know, its very pro-business but they did a 19-page
special survey on the impact of globalization. The bottom line in
The Economists article was the United States is absorbing
globalization differently than even other first-world nations because of a lot of the things that I just mentioned. And that if we
dont find some solutions that were going to end up with protectionism on the one hand and potentially social unrest on the other.
So where do we really find these answers?
Im not taking a position even on the issue of wage insurance per
se. I think its a small piece of what we need to be looking at, and
Im interested in your-alls reactions. Well just start with Dr.
Brainard.
Dr. Brainard. I think youre exactly right. My hope is that this
Committee will be looking at some of the other pieces of that puzzle, too. A piece of the answer is that we have to do a better job
of smoothing incomes and protecting those who might be shouldering, almost certainly are shouldering a disproportionate share of
that burden of adjustment.
But another big piece which I dont think has moved along as
much as one would have liked, and weve beenover the last 6
years weve seen very stark changes in the international economic
landscape. We havent seen very robust policy responses to it on

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00029

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

26
areas like competitiveness and asking those kinds of questions
about how we are preparing ourselves to continue to be among the
highest wage earners in the international competitive arena.
How are we going to do that? Were going to do it by investing
in certain areas of value, and that means our workforce has to be
prepared to be able to do that. It does go to education and workforce training and it goes to innovation policy and it goes to a
whole host of things. It goes to infrastructure. Those questions I
think are part and parcel of it.
And of coursewhich you didnt raise, but asking questions
about the extent to which the international rules are being advanced to favor U.S. interests I think is another really important
question. Are we enforcing the rules that weve got and are we
pushing forward those rules that are most consequential for us economically, as opposed to a set of foreign policy trade agreements,
for instance.
Mr. Emsellem. My field of expertise, is not globalization, but I
know a little bit about how workers deal with these issues. I guess
my response would be to obviously pursue the issue of healthcare
and health insurance, our experience there.
There are some states that are really doing some really novel
things there to fund health insurance for folks who qualify for unemployment benefits.
That makes a big difference in peoples livesjust paying out the
COBRA coverage and all that, takes a huge chunk of funds. If it
was the same amount of wage insurance money, it could go toward
that, and that would, we know, make a big difference.
In the last recession, it was very interesting when Congress got
back to the business of putting together an extension of the unemployment benefits for the workers who were laid off in the last recession.
There was a lot of talk, and President Bush supported the idea
of providing health insurance to unemployed workers. There is an
opportunity there.
We ended up with a little piece of a program in TAA, that has
a lot of problems. That is a starting point, but it has a lot of problems. I think thats the area, from the perspective of workers who
are struggling with these issues, that would be very important to
pursue.
Ms. Batchelder. Its a very interesting question, and, I think,
first, we need to think seriously about moving away from a purely
employer-based model for providing both pension coverage and
health insurance.
Also, the proposal to transform tax incentives into refundable
credits, I think, would help a lot at the worker level. Right now,
our tax incentives for both retirement savings and health insurance, are, of course, worth much more to a higher-income worker.
For instance, the retirement savings incentive for the bottom 40
percent of the income distribution, they gain about 2 percent of all
of the value of retirement savings incentives.
The problem with this is not only that these are the workers that
need incentives the most, but also that theyre kind of being wasted, because were often buying out the base, as we call it in tax
lingo, where people who would already have health insurance, or

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00030

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

27
would already be saving adequately for retirement, were giving
them an incentive they dont need.
In the retirement savings context, often they will just shift their
existing savings into the tax-deferred vehicle, and not necessarily
increase their savings. So we can provide incentives a lot more efficiently.
The other point I would make, is that its really important to leverage insights from behavioral economics in both of these areas.
In particular, looking at default rules, trying to make it easier for
people to save, eliminating some of the administrative hassle and
can often have a much greater impact than any tax incentive or financial incentive on increasing retirement savings.
Dr. Schiller. Let me say that Im concerned about the perceptions about globalization. Globalization has been a boon to the
American economy. All this conversation focuses on import competition, but the export sector has been the fastest growing sector
in the U.S. economy for the last 15 years. Its been a tremendous
engine of growth.
We point to the Chinese economy and people suggest that they
are outdoing us. The Chinese have located four million manufacturing workers in the last 10 years, while their economy has
grown, so its technology thats improving so fast and dislocating
workers in manufacturing; its not trade, per se.
Trade is an engine of growth for the U.S. economy.
Senator Webb. If I may, on a couple of points: One is, we had
a trade deficit last year of $832 billion, and just to nail down something we were talking about with Mr. Emsellem, the great concern
I have when I look at our movement toward globalization through
the different agreements, ending up with the WTO, is that there
were no provisions that were directly protecting or standardizing
the workforce.
We could beginits very difficult to do with fast-track trade legislation, but we could begin to be asking for equal workplace environments in other countries, competitive countries, as a starting
point on trying to protect American workers.
Its no accident that China and India have the largest greenhouse effects in the world right now, because there are no standards in their workplaces like we enforce and demand in our own.
There are ways that American workers are being affected, that
do not go to the quality of the work or even the inequality of economic systems, that could be approached. Thank you, Mr. Chairman.
Chairman Schumer. Thank you. I want to thank our panel.
Certainly, I agree with Senator Webb, that we have to look at the
macro picture, but I think that in this changing world, even if
trade were not an issue, technology would cause lots of displacement here within America.
We also have to look at the fact that individual workers are buffeted about more than they used to be. I think this was a very
helpful hearing.
Professor Batchelder, your ideas, I love the idea of making a
credit. In fact, Ive just proposed the college tuition deductibility,
which is a law I authored. We combine all of those programs and

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00031

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

28
make them into a credit, which makes a great deal of sense, and
the idea of income-averaging, seems to make sense, as well.
So I thank all of you. Wage insurance is one idea. It will be debated here. Well see where to go with it. Were looking for other
ideas. The Committee is going to look at both ends of this new
international economy, the macro picture, which Senator Webb focused on, as well as more of the individual worker picture.
One thing I would say to you, Dr. Schiller, is that the French
students who were demonstrating, did not have wage insurance,
because they werent working.
In any case, I thank all of you for being here, and would ask
unanimous consent because Senator Brownback would like to submit some questions for the witnesses. Congresswoman Maloney
would also like to submit questions, so, without objection, well ask
that you respond to those questions within a week, if thats OK
with you.
[The material referred to was unavailable at press time.]
Chairman Schumer. Thank you all for being here. It was a
very informative panel. The hearing is adjourned.
[Whereupon, at 11:05 a.m., the hearing was adjourned.]

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00032

Fmt 6633

Sfmt 6633

C:\JEC\34852.TXT

CELINA

PsN: CELINA

Submissions for the Record

PREPARED STATEMENT

OF

SENATOR CHARLES E. SCHUMER, CHAIRMAN

Good morning. I would like to thank our witnesses and guests for attending today,
and I want to welcome the new Vice Chair, my colleague from New York, Mrs.
Maloney. I look forward to working closely with her to use this Committee as an
engine for generating economic policies that will work to deliver the benefits of economic growth to all Americans.
Today, we are at a critical juncture in U.S. economic policy. We know that the
upheavals caused by technological change and international competition most acutely affect those who are gaining the least economicallythe middle-class and those
who aspire to get there. Yet in order for us to expand trade and make significant
technological investments to help grow the economy, the middle-class must feel that
they will benefit. Right now, too many of them dont.
Working at a large corporation for thirty or forty years that takes care of you and
your family for a lifetime is becoming a thing of the past. Employers are now shifting the high costs of health care and the burden of saving for retirement onto families. And increasingly, jobs are being automated away by technological advancements or moved overseasleaving many displaced workers and their families behind.
Meanwhile, official numbers on the economy have been positiveat least until
very recently. But we must face the reality lurking behind the official numbers in
order to address anxiety on Main Street.
Not only have wages significantly lagged behind productivity over the past two
decades, but they are increasingly more volatile as workers bounce in and out of
jobs. Between 2003 and 2005, nearly 4 million workers were laid off from jobs they
held for more than 3 years. About half of these workers and their families took a
pay cut, and nearly one-third lost 20 percent or more of their prior earnings. And
if the recession in the manufacturing sector that hit our radar screens this week
spreads through our economythe economic roller coaster for families will only get
worse.
Income volatility can cause major upheavals for families, on top of the changes
they are facing in the workplacethey could be forced to sell their homes, or to discontinue health care coverage. Income volatility also leaves families feeling unsettled about their familys and their countrys economic future.
We need a new policy direction to meet the challenge of income instability. We
must start by strengthening the safety net that helps displaced workers rebound
from job losses that occur through no fault of their own.
We have asked our witnesses on the second panel to share their recommendations
for doing just that. This morning, our experts will explore new policies like wage
insurance and income-averaging, as well as ways to strengthen our existing unemployment insurance and Trade Adjustment Assistance programs.
We also need to do everything we can at the federal level to spur the development
of high-quality, high-paying jobs to replace the jobs lost in declining segments of the
economy or through advancements in technology. We need to make serious investments in our most promising industries for future growth, like renewable energy
and life sciences.
(29)

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00033

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

30
And we need to help our displaced workers acquire the skills and experience they
will need to succeed in the new jobs created. We will investigate opportunities for
creating good jobs in more detail in a series of JEC hearings in the coming months.
But right now, middle-class families need help dealing with the tectonic shifts
that technology is causing; they need help dealing with the forces beyond their control that are changing their lives. They dont want handouts, but they need a hand.
I know we will have some disagreements over particular solutions to this problem
of income instability, but I hope that we will all prioritize the need to help our families mitigate the new risks they face and achieve their aspirations. And I look forward to working closely with all of you to do just that.
Ive said before that the JEC would seek insight and advice from the best and
thats what we have to offer here again today. I will now introduce todays panelists.

PREPARED STATEMENT

OF

REPRESENTATIVE CAROLYN B. MALONEY, VICE CHAIR

Thank you, Chairman Schumer. I am pleased to welcome Director Orszag and our
panel of witnesses to talk about the critically important issue of income instability
and what we can do to help families manage the economic shocks they may experience.
As Director Orszag points out, wild swings in the overall economy have been tempered, but the same cannot be said for the economic circumstances of families trying
to adapt to a dynamic global economy. The Congressional Budget Office (CBO) has
looked at this issue and found that households experience significant ups and downs
in their earnings and income from year to year, and the downside problem may be
getting worse due to the forces of globalization and technological change. Not surprisingly, the income roller coaster is a particularly rough ride if you are less educated.
Our second panel of witnesses will touch on various proposals to address income
instability. I know that theylike the members of this committeewill be coming
at these issues from different perspectives, but I look forward to a serious policy discussion and competition among ideas. One idea that we will focus on today is wage
insurance. Our Chairman is planning to introduce legislation on this issue, which
will no doubt generate further useful debate about what is the best way to deal with
the adverse side effects of economic change.
Dr. Brainard has offered a wage insurance proposal with her Brookings colleagues
to provide economic incentives for more rapid reemployment and on-the-job training.
I certainly agree with the goal, but not necessarily the game plan for getting there.
As Dr. Brainard observes, our nations safety net has more holes than netting,
which is why I think we should mend it before we make it bigger, as Mr. Emsellem
urges. Wage insurance may well have a role to play, but implementing it should not
come at the expense of shoring up the Unemployment Insurance system or Trade
Adjustment Assistance, both of which are in dire need of reform.
Finally, CBO data show how the tax code can exacerbate the income volatility,
especially for low-income taxpayers. Prof. Batchelder proposes novel changes to the
tax code so that low-income families, whose incomes tend to fluctuate the most,
could average their income over two years to smooth out variability, and enjoy similar tax advantages as businesses in their ability to shift unused deductions and exemptions.
As an aside, I want to note that CBO examined earnings and income volatility
using the Survey of Income and Program Participation, the SIPP, a leading source
of comprehensive data on the economic well-being of American families. Last year
there was an effort by the Administration to eliminate the SIPP without having an
adequate plan in place to collect this invaluable information. The status of the SIPP
remains up in the air, and I hope, Director Orszag, that you will join the growing
chorus of researchers and academics who have called on the Administration to preserve this survey until a better one can be designed and implemented.
I look forward to the testimony of our witnesses and their thoughts on policies
that can help families better manage income instability.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00034

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

31
PREPARED STATEMENT

OF

HON. PETER R. ORSZAG, DIRECTOR, CONGRESSIONAL

BUDGET OFFICE*

Chairman Schumer, Vice-Chair Maloney, Congressman Saxton, and Members of

the Committee, I appreciate the invitation to participate in todays hearing. My testimony makes four main points:
First, macroeconomic volatilitythe ups and downs of overall economic growth
and inflationhas declined and is now relatively low. In particular, year-to-year
fluctuations in the economy have become smaller than in the past.
Second, despite the relatively modest volatility in the overall economy, workers
and households still experience substantial variability in their earnings and income
from year to year. The Congressional Budget Offices (CBOs) analysis shows, for example, that between 2001 and 2002, one in four workers saw his or her earnings
increase by at least 25 percent, while one in five saw his or her earnings decline
by at least 25 percent. Some of that variability stems from voluntary actions, such
as a decision to stay home and rear children, and some stems from involuntary
events, such as the loss of a job. Earnings volatility is somewhat higher for people
with less education.
Third, although earnings and income volatility is substantial, more research is
required to determine how and when that variability has changed over the past few
decades. The existing evidence suggests that annual earnings have tended to fluctuate more, on a percentage basis, over the past 25 years than they did during the
1970s. The number of studies on the topic is limited, however, and they have somewhat different results. Therefore, it is too early to reach firm conclusions about the
precise timing or magnitude of any increase. Given their importance, trends in earnings and income volatility seem to warrant significant research attention.
Fourth, many observers are accustomed to thinking about the Federal tax system as an automatic stabilizer that helps to reduce variations in national income.
The tax system, though, also helps to smooth out variability at the level of households by reducing year-to-year fluctuations in their after-tax income. That insurance
effect of the tax system is potentially significant, given the substantial variation in
households earnings and income. At the same time, however, the tax system levies
higher average rates on households whose income is more variable and imposes
costs on the economy by distorting the decisions that households make about how
much to work, how much to save, and how to receive their compensation for doing
so. In evaluating different tax structures, policymakers need to weigh the role of the
tax system in smoothing income against its other effects on households and the
economy.
MACROECONOMIC VOLATILITY

Macroeconomic volatility has been significantly lower during the past 20 years
than in preceding decades. Although recessions can still be quite painful for particular sectors and workers, recessions have been less severe overallin duration,
frequency, and magnitudethan they were between 1950 and the mid-1980s. The
quarter-to-quarter fluctuations in real (inflation-adjusted) gross domestic product
(GDP) have also become smaller (see the top panel of Figure 1). In addition, the
level and volatility of inflation over the past 20 years have also been relatively low
(see the bottom panel of Figure 1). Volatility in more recent years has been less
than half that of the previous period (see Table 1). The corresponding reduction in
peoples uncertainty about prices allows them to plan better for the future. Volatility
has declined not only in the growth of overall GDP and inflation but also in virtually all of the major components of GDP and in aggregate unemployment, wages,
and income.

* Some of the figures in this testimony use shaded vertical bars to indicate periods of recession. (A recession extends from the peak of a business cycle to its trough.)
Numbers in the text and tables may not add up to totals because of rounding.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00035

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

32

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00036

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

33
Table 1.Changes in Macroeconomic Volatility
Volatility
19501984

GDP Growth .........................................................................................................................................

Inflation ...............................................................................................................................................

3.1
2.9

19852005

1.4
1.0

Sources: Congressional Budget Office; Department of Commerce, Bureau of Economic Analysis.

Note: Volatility is measured as the standard deviation of the change from the previous year in gross domestic product (GDP) per capita
(for GDP growth) and in the chained price index for personal consumption expenditures (for inflation), in each case using quarterly data.

Although there is no conclusive explanation for the decline in the volatility of

GDP growth and inflation, numerous reasons have been advanced, many of which
are closely interrelated. The proposed explanations fall into four broad categories.
A More Flexible Economy. Improvements in production processes and investments in information technologies (such as those that facilitate just-in-time inventory management), increases in temporary and flexible work arrangements, and the
deregulation of many industries (especially in the transportation sector) have made
it possible for the economy to adjust much more smoothly to changes in the availability of, or demand for, goods and services. The economy an more easily adapt to
shocks, such as the energy price shock of 2004 and 2005, without large changes in
output or large jumps in inflation.1
Improvements in Financial Markets and Institutions. Financial innovations
since the 1970s have enhanced businesses and households access to credit and
thereby enabled them to borrow more readily when their income turns down. Those
innovations include improved assessment and pricing of risk (including the development of credit derivatives and interest rate swaps) and the greater use of financial
markets in supplying credit (through securitization, for example).2 In addition,
changes in government regulations have allowed more diversification in banking
and made housing financing more stable. Even though those changes in capital markets seem esoteric, they appear to have broadened and deepened access to credit for
both businesses and households and to have improved the resiliency of the financial
system by spreading the risk of default more widely and efficiently.
Management of Monetary Policy. During the past two decades, the Federal Reserve has shown a strong commitment to keeping inflation low and stable. Its actions to reduce and contain inflationary pressures seem, in turn, to have stabilized
firms and households expectations of future inflation. As a result, the Federal Reserve has not needed to respond as forcefully as in the past to dampen swings in
expectations of inflation or to bring inflation down from a high level. The result may
be reduced macroeconomic volatility.
Fewer Shocks to the Economy. This explanationthat fewer shocks to the economy, particularly the worldwide economy, have occurredwas proposed before the
rapid rise in oil prices from 2004 to mid-2006. Given the mild effect of that oil price
shock on economies worldwide, the explanation now seems less persuasive. Moreover, overall U.S. economic growth was little affected by other major shocks during
the past 20 years, such as the Asian currency crisis of 1997, the Russian debt crisis
of 1998, and the terrorist attacks of September 11, 2001.
WORKERS EARNINGS AND HOUSEHOLDS INCOME

The story at the level of the individual worker or household is different from the
story at the macroeconomic level. Individual earnings tend to rise over time, but the
data suggest that workers and families experience substantial volatility year to year
around that underlying trend.
To examine earnings and income volatility, CBO analyzed recent data from the
Survey of Income and Program Participation (a data set collected by the U.S. Census Bureau). The analysis focused on workers who were 25 to 55 years old and not
in school, so it therefore does not capture changes in earnings associated with grad1 See Congressional Budget Office, The Economic Effects of Recent Increases in Energy Prices
(July 2006). See also Lawrence F. Katz and Alan B. Krueger, The High Pressure U.S. Labor
Market of the 1990s, Brookings Papers on Economic Activity, no. 1 (1999).
2 Securitization involves the conversion of cash flows into securities; credit derivatives are financial instruments designed to transfer credit risk from one party to another; and interest rate
swaps are exchanges of two series of payments based on different interest rates, which entities
undertake to manage their exposure to changes in rates.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00037

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

34
uating from school or leaving work for school.3 Even so, the analysis shows substantial variation in workers before-tax earnings from 2001 to 2002. After an adjustment for inflation, one in four workers saw his or her earnings increase by at least
25 percent, while one in five saw his or her earnings decline by at least 25 percent.
A substantial portion of workers, 11 percent, saw their earnings decline by at least
half (see Figure 2).

Workers with less education tend to experience more volatility in their earnings
than do workers with more education (see Table 2). For example, from 2001 to 2002,
16 percent of workers without a high school education had their earnings decline
by 50 percent or more, compared with 10 percent of workers with more than a high
school education.

3 For a discussion of wage trends in low-wage labor markets, see Congressional Budget Office,
Changes in Low-Wage Labor Markets Between 1979 and 2005 (December 2006).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00038

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

35
Table 2.Distribution of Changes in Workers Annual Earnings from 2001 to 2002, by
Educational Attainment and Age
Decrease in Earnings of
At Least
50 Percent

All Workers ...........................................................

Educational Attainment
Less than high school ....................................
High school .....................................................
More than high school ....................................
Age
25 to 30 ..........................................................
31 to 40 ..........................................................
41 to 55 ..........................................................

Changes in
Earnings of
Less Than 25
Percent

25 Percent

Increase in Earnings of
At Least
25 Percent

50 Percent

10.7

19.8

55.5

24.7

14.2

15.6
11.6
9.5

26.0
19.8
18.8

47.9
55.0
57.0

26.0
25.2
24.2

16.4
14.8
13.6

11.4
10.7
10.5

20.0
19.8
19.7

53.8
54.5
56.7

26.2
25.7
23.6

14.6
14.9
13.7

Source: Congressional Budget Office based on data from the 2001 panel of the Bureau of the Censuss Survey of Income and Program Participation.
Note: The sample consists of individuals ages 25 to 55 in 2001 who had positive earnings in 2001 and were not enrolled in school that
year or in 2002. Earnings are inflated to 2002 dollars using the research series of the consumer price index for urban consumers.

Such fluctuations in earnings can result from many sources, including job
changes, job losses, job gains, voluntary exits from the labor force to care for children or other family members, changes in the number of hours worked per year,
or changes in the wage rate received by workers. Among workers who experienced
at least a 50 percent drop in earnings, most did not work at least a month and typically did not work eight months in 2002. When asked why they were not working,
the most common responses were that they were caring for a child or other family
member or were pregnant; were not able to find work or had been laid off; were
unable to work because of disability, illness, or injury; or were not interested in
working or were retired.4 The responses appear to be split evenly between those
suggesting that the departure from the labor force was voluntary and those suggesting that it was not.
Total household income consists not only of the earnings of household members
but also other sources of cash income such as unemployment insurance, retirement
income, dividends, and interest. Compared with earnings, it thus represents a
broader measure of the economic resources available to individuals.5 Like workers
earnings, household income can vary from year to year, though it tends to be less
variable than individual earnings. First, if an individual worker in a household with
multiple earners loses a job, the earnings of the other members may partially mitigate the consequences of the job loss. Second, a loss in earned income may be alleviated by an increase in other sources of income, like unemployment insurance, payments from a retirement plan, or disability insurance. Neither the mitigating effects
of the presence of other earners in the household nor the potential for increases in
nonlabor income is captured in the more narrow measure of individual earnings.
To be sure, household income can vary from changes in the composition of households. Households are not fixed entities: They often evolve, as couples marry, separate, or divorce and working children move out of or into the house.
According to CBOs analysis, the growth of before-tax income varied substantially
among households between 2001 and 2002 (see Figure 3). Nearly one in four households experienced an increase in income of at least 25 percent, virtually identical
to the number of individuals who experienced a similar percentage increase in earnings. Fewer households, one in seven, experienced a decrease in income of at least
25 percent. And one in 25 households experienced a decrease in income of at least
50 percentcompared with one in nine individuals who experienced such a decline
in earnings. Unlike the variability of earnings, however, the variability of household
income seems similar across education levels (see Table 3).

4 Only those individuals who had at least four consecutive months without a job responded
to the question.
5 Household income, as reported here, is before-tax income and excludes capital gains and
losses.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00039

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

36

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00040

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

37
Table 3.Distribution of Changes in Households Annual Income from 2001 to 2002, by
Educational Attainment and Age of the Head of the Household
Decrease in Income of
At Least
50 Percent

All Households .....................................................

Educational Attainment of the Head of the
Household
Less than high school ....................................
High school .....................................................
More than high school ....................................
Age of the Head of the Household
25 to 30 ..........................................................
31 to 40 ..........................................................
41 to 55 ..........................................................

Changes in
Income of
Less Than 25
Percent

25 Percent

Increase in Income of
At Least
25 Percent

50 Percent

4.3

14.2

62.2

23.6

12.5

4.3
4.2
4.3

14.6
13.8
14.3

62.1
61.9
62.3

23.3
24.2
23.3

12.6
12.6
12.4

4.2
4.3
4.8

14.8
14.7
15.1

59.3
59.6
61.2

26.0
25.7
23.7

13.8
13.6
12.1

Source: Congressional Budget Office based on data from the 2001 panel of the Bureau of the Censuss Survey of Income and Program Participation.
Note: The sample consists of households in January 2001 that were surveyed for all of that year and 2002. Income, which is before taxes,
includes earnings, unemployment compensation, workers compensation, Social Security benefits, Supplemental Security Income, public assistance, veterans payments, survivors benefits, disability benefits, pension or retirement income, interest, dividends, rents, royalties, income
from estates or trusts, alimony, child support, financial assistance from outside the household, and other cash income. Income is inflated to
2002 dollars using the research series of the consumer price index for urban consumers.

For another point of comparison, CBO conducted a similar analysis using data
from 1997 to 1998a period of relatively rapid economic growth, in contrast to the
relatively slow growth from 2001 to 2002and found similar results.6 Thus, substantial variability in workers earnings and income can occur in periods of both
strong and weak economic growth.
Using surveys to measure the year-to-year variability in earnings and income is
complicated by the fact that individuals responses are often in error (which could
either overstate or understate the actual changes in earnings or income).7 In addition, while the surveys are intended to be nationally representative, they may not
include undocumented workers and can be subject to biases because some people either refuse to respond at all or drop out of the surveys before their completion. An
important question, then, is whether, over longer periods of time, earnings and income volatility has increased. According to most studies on the topic, earnings have
tended to fluctuate more, on a percentage basis, over the past 25 years than they
did during the 1970s.8 Relative to other questions about income and earnings, however, the trend in their volatility has received relatively little research attention.
More research is therefore needed before firm conclusions about the precise time
trend in earnings and income volatility can be reached.
To the extent that variability in earnings and income has increased, the phenomenon may be consistent withand indeed perhaps part of the explanation of
the decreased macroeconomic volatility described earlier. For example, more-flexible
labor markets could enable the economy to adjust to changes in the economic environment more quickly but also could mean that individuals change jobs and have
their wages change more frequently.
RISK SHARING, INCOME FLUCTUATIONS, AND TAXATION

Economists have long noted that the tax system serves as an automatic stabilizer
that offsets at least part of demand shocks to the economy.9 A decline in aggregate
6 The data are from the 1996 and 2001 panels of the Survey of Income and Program Participation, conducted by the U.S. Census Bureau.
7 See John Bound and Alan B. Krueger, The Extent of Measurement Error in Longitudinal
Surveys: Do Two Wrongs Make a Right? Journal of Labor Economics, vol. 9, no. 1 (January
1991), pp. 124.
8 See, for example, Peter Gottschalk and Robert Moffitt, The Growth of Earnings Instability
in the U.S. Labor Market, Brookings Papers on Economic Activity, no. 2 (1994); Costas Meghir
and Luigi Pistaferri, Income Variance Dynamics and Heterogeneity, Econometrica, vol. 72, no.
1 (2004), pp. 132; Maury Gittleman and Mary Joyce, Earnings Mobility in the United States,
196791, Monthly Labor Review, vol. 118, no. 9 (September 1995), pp. 313; and Peter
Gottschalk and Robert Moffitt, Trends in the Transitory Variance of Earnings in the United
States, Economic Journal, vol. 112, no. 478 (2002), pp. 6873.
9 See Alan J. Auerbach and Daniel Feenberg, The Significance of Federal Taxes as Automatic
Stabilizers, Journal of Economic Perspectives, vol. 14, no. 3 (Summer 2000), pp. 3756; and
Continued

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00041

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

38
before-tax income of one dollar generates a decline in aggregate after-tax income of
less than one dollar. As a result, the tax system helps to stabilize demand for goods
and services, which in turn helps to reduce fluctuations in the overall economy.10
In addition to its well-recognized role as a macroeconomic automatic stabilizer,
the tax system can serve as a microeconomic automatic stabilizer by helping to
smooth out variability at the level of workers earnings and households income.11
The tax system automatically reduces the tax burden when before-tax income declines and automatically raises the burden when before-tax income rises. After-tax
income therefore tends to vary less than before-tax income.12 In that way, the tax
system provides a form of after-tax earnings or income insurance, which complements the social insurance provided through a variety of government programs.
(Although the Federal tax system generally works to smooth out fluctuations in income, that attribute does not apply for each and every taxpayer.13)
The risk-sharing features of the tax system can be illustrated in a simple example
(see Table 4). Consider a single worker earning $45,000 in 2006 with no other
sources of income. At that level of income, the worker would owe $5,695 in Federal
income taxes and $3,443 in payroll taxes and would therefore have $35,863 in aftertax income. If the workers earnings fell by 20 percent, to $36,000, after-tax earnings would decline to $29,491. Although before-tax earnings fell by $9,000 (20 percent), after-tax earnings declined by only $6,372 (18 percent).

Table 4.Effect of Taxes on the Variability of Income: An Example

(Dollars)
Initial
Wages

Lower
Wages

Before-Tax Wages .................................................................................................

Income Taxes ........................................................................................................
Payroll Taxes .........................................................................................................

45,000
5,695
3,443

Total taxes ........................................................................................................

After-Tax Wages ....................................................................................................

9,138
35,863

Change in Wages
Dollars

Percent

36,000
3,755
2,754

-9,000
..............
..............

-20
..............
..............

6,509
29,491

..............
-6,372

..............
-18

Source: Congressional Budget Office.

Note: Based on the tax schedule for a single worker in 2006.

The predictability of households income will affect how much value they place on
the insurance provided through the tax system. To the extent that swings in earnings or income are unpredictable, households will tend to value the insurance more.
However, the value of that insurance will be smaller for households whose earning
or income swings are largely expected or stem from intentional decisions about how
much and when to work.
The insurance provided by the progressive tax system to households with variable
income comes at a price: it can reduce average after-tax income for such households.
Consider two people who have the same amount of lifetime earnings; one has steady
earnings and the other, large swings in earnings. Under a progressive tax system
Thomas J. Kniesner and James P. Ziliak, Tax Reform and Automatic Stabilization, American
Economic Review, vol. 92, no. 3 (June 2002), pp. 590612.
10 The stabilizing effect of the tax system on the overall economy reached a peak around 1980
and by 1995 had declined to about the same level as in the 1960s. Since 1995, according to
CBOs estimates, there has been relatively little change. Those movements mirror the increase
and then the decline in effective tax rates. See Auerbach and Feenberg, The Significance of
Federal Taxes as Automatic Stabilizers.
11 See Hal R. Varian, Redistributive Taxation as Social Insurance, Journal of Public Economics, vol. 14, no. 1 (August 1980), pp. 4968; Jonathan Eaton and Harvey S. Rosen, Labor Supply, Uncertainty, and Efficient Taxation, Journal of Public Economics, vol. 14, no. 3 (December
1980), pp. 365374; Jonathan Eaton and Harvey S. Rosen, Taxation, Human Capital, and Uncertainty, American Economic Review, vol. 70, no. 4 (September 1980), pp. 705715; Jonathan
Eaton and Harvey S. Rosen, Optimal Redistributive Taxation and Uncertainty, Quarterly
Journal of Economics, vol. 95, no. 2 (September 1980), pp. 357364.
12 Variability of income can be measured in different ways. Some analysts measure it as the
change in dollar income; other analysts measure it as the percentage change in income. A pure
proportional tax system can reduce the dollar amount of variability but does not affect the variability in percentage terms; a progressive tax system can reduce variability by both measures.
13 See Robert Moffitt and Michael Rothschild, Variable Earnings and Nonlinear Taxation,
Journal of Human Resources, vol. 22, no. 3 (Summer 1987), pp. 405421. For example, the payroll tax for the Old-Age, Survivors, and Disability Insurance program does not apply to earnings
above the taxable maximum ($97,500 in 2007). As a result, when earnings fluctuate across that
threshold, after-tax earnings can be more variable in percentage terms than before-tax earnings.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00042

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

39
based on annual income, the steady earner pays less in taxes over a lifetime even
though both people have the same total amount of earnings. Thus, progressive taxation combined with an annual accounting period fails to treat people in similar circumstances in the same way. Various options for changing the tax system would
alter the tradeoff between the income smoothing insurance provided and the average cost imposed on households with variable income.
In addition to that tradeoff between the insurance provided to and the price paid
by households with variable income, any risk-sharing benefits that the tax system
generates must be weighed against the potential costs that it imposes on the economy at large. Marginal tax rates affect households decisions about how much to
work and save, as well as the form in which to receive compensation for doing so,
and those distortions reduce the efficient operation of the economy. The implicit insurance that the government provides through the tax system may have other effects, such as changing the types and forms of insurance products offered by the private markets or encouraging people to take risks they would not take in the absence
of that implicit insurance.14
Comparing the various costs and benefits is difficult, and a complete accounting
of all of those effects has not yet been achieved. Nonetheless, some recent studies
have found that, compared with some alternatives, the current tax system may provide insurance benefits that are larger than the costs that it imposes on the economy by distorting decisions about working and saving.15 However, those analyses
depend on many assumptions, and alternative assumptions could yield different estimates, so the studies should be viewed with caution. Despite those caveats, a reasonable conclusion from this new research is that the income-smoothing insurance
provided through the tax system could be quantitatively important and should be
taken into account in any analysis of the relative costs and benefits of different tax
systems.
Finally, it is important to note that the benefits of risk sharing and the costs of
distortions are not captured by changes in GDP. Although GDP is a useful summary
measure that may be related to households well-being, it does not measure the
value that households place on smoother incomes or the cost of distorted decisionmaking. Instead, GDP is merely a measure of how much output the market economy
produces using its capital, labor, and technology. It does not measure what ultimately matters and what needs to be measured: changes in the well-being of households.
CONCLUSION

The U.S. economy has become less volatile: Macroeconomic fluctuations are now
much milder than they were in the past. At the same time, however, households
continue to experience substantial variability in their earnings and income, and that
variability may now be greater than in the pastperhaps contributing to anxiety
among workers and families. The tax system can help to smooth fluctuations in income not only at the macroeconomic level but also at the level of workers and households. The income insurance provided as a result may be quite valuable but needs
to be weighed against the other effects of the tax system.

14 See Dirk Krueger and Fabrizio Perri, Public Versus Private Risk Sharing (working paper,
December 2005).
15 See Shinichi Nishiyama and Kent Smetters, Consumption Taxes and Economic Efficiency
with Idiosyncratic Wage Shocks, Journal of Political Economy, vol. 113, no. 5 (October 2005),
pp. 10881111; Juan Carlos Conesa and Dirk Krueger, On the Optimal Progressivity of the Income Tax Code, Journal of Monetary Economics, vol. 53, no. 7 (October 2006), pp. 14251450.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00043

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

40

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00044

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

41

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00045

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

42

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00046

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

43
ATTACHMENT, ANALYSIS, CONGRESSIONAL BUDGET OFFICE, HEALTH
RESOURCES DIVISION

AND

HUMAN

Trends in Earnings Variability Over the Past 20 Years

(April 2007)
Contents

Page

Methodology ..............................................................................................................................................................
Analysis of Variability Using Administrative Data ..................................................................................................
Analysis of Variability Using Survey Data ...............................................................................................................
Conclusion ................................................................................................................................................................
Appendix: Alternative Measures of Earnings Variability .........................................................................................
Tables
1. Percentage of Workers for Whom Total Wage Earnings Dropped or Rose by 50 Percent or 25 Percent, by
10-Year Age Category .....................................................................................................................................
2. Distribution of Changes in Workers Annual Real Earnings, by Educational Attainment and Age, 2001 to
2002 ................................................................................................................................................................
Figures
1. Percentage of Workers for Whom Total Wage Earnings Declined by 50 Percent or More Over the Previous
Year, by Sex ....................................................................................................................................................
2. Percentage of Workers for Whom Total Wage Earnings Declined by 25 Percent or More Over the Previous
Year, by Sex ....................................................................................................................................................
3. Percentage of Workers for Whom Total Wage Earnings Rose by 50 Percent or More Over the Previous
Year, by Sex ....................................................................................................................................................
4. Percentage of Workers for Whom Total Wage Earnings Rose by 25 Percent or More Over the Previous
Year, by Sex ....................................................................................................................................................
5. Standard Deviation of the Percentage Change in Workers Total Wage Earnings Over the Previous Year,
by Sex ..............................................................................................................................................................
6. Percentages of Workers Without Any Total Wage Earnings in the Previous or Subsequent Calendar Years
7. Percentage of Workers in the Bottom Two-Fifths of the Earnings Distribution for Whom Annual Social
Security Taxable Earnings Declined by 50 Percent or More Over the Previous Year, by Sex .......................
8. Percentage of Workers in the Bottom Two-Fifths of the Earnings Distribution for Whom Annual Social
Security Taxable Earnings Declined by 25 Percent or More Over the Previous Year, by Sex .......................
9. Distribution of Changes in Workers Annual Real Earnings, 2001 to 2002 ..................................................
A-1. Percentage of Workers for Whom Total Wage Earnings Declined by 50 Percent or More Over the Previous Five Years, by Sex .................................................................................................................................
A-2. Percentage of Workers for Whom Total Wage Earnings Declined by 25 Percent or More Over the Previous Five Years, by Sex .................................................................................................................................
A-3. Percentage of Workers for Whom Total Wage Earnings Rose by 50 Percent or More Over the Previous
Five Years, by Sex ...........................................................................................................................................
A-4. Percentage of Workers for Whom Total Wage Earnings Rose by 25 Percent or More Over the Previous
Five Years, by Sex ...........................................................................................................................................
A-5. Standard Deviation of the Percentage Change in Workers Total Wage Earnings Over the Previous Five
Years, by Sex ...................................................................................................................................................
A-6. Percentage of Workers for Whom the Log of Total Wage Earnings Declined by 50 Percent or More Over
the Previous Year, by Sex ...............................................................................................................................
A-7. Percentage of Workers for Whom the Log of Total Wage Earnings Declined by 25 Percent or More Over
the Previous Year, by Sex ...............................................................................................................................
A-8. Percentage of Workers for Whom the Log of Total Wage Earnings Rose by 50 Percent or More Over the
Previous Year, by Sex ......................................................................................................................................
A-9. Percentage of Workers for Whom the Log of Total Wage Earnings Rose by 25 Percent or More Over the
Previous Year, by Sex ......................................................................................................................................
A-10. Standard Deviation of the Difference in the Log of Workers Total Wage Earnings from the Previous
Year, by Sex ....................................................................................................................................................
A-11. Percentage of Workers for Whom Total Age-Adjusted Wage Earnings Declined by 50 Percent or More
Over the Previous Five Years, by Sex .............................................................................................................
A-12. Percentage of Workers for Whom Total Age-Adjusted Wage Earnings Declined by 25 Percent or More
Over the Previous Five Years, by Sex .............................................................................................................
A-13. Percentage of Workers for Whom Total Age-Adjusted Wage Earnings Rose by 50 Percent or More Over
the Previous Five Years, by Sex ......................................................................................................................
A-14. Percentage of Workers for Whom Total Age-Adjusted Wage Earnings Rose by 25 Percent or More Over
the Previous Five Years, by Sex ......................................................................................................................
A-15. Standard Deviation of the Difference in the Log of Workers Total Age-Adjusted Wage Earnings from
the Previous Five Years, by Sex ......................................................................................................................

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00047

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

44
45
47
47
57

47
48

49
50
51
52
53
54
55
56
57
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74

PsN: CELINA

44
TRENDS

IN

EARNINGS VARIABILITY OVER

THE

PAST 20 YEARS

In response to a request from Senators Charles Schumer and Jim Webb, the Congressional Budget Office (CBO) analyzed the extent to which workers earnings vary
from year to year and whether that variability has increased over the past 20 years.
To analyze those issues, CBO used data and techniques it has developed for projecting individual earnings in its long-term model for Social Security and Medicare.1
Understanding past trends in variability is key for projecting future earnings patterns, and those patterns are an important input into CBOs projections for Social
Security and Medicare (because revenues and outlays are directly tied to individual
workers earnings through tax and benefit formulas).
For its analysis, CBO used data from the Social Security Administrations Continuous Work History Sample (CWHS) and the U.S. Census Bureaus Survey of Income
and Program Participation (SIPP). Although the use of the CWHS allows for a more
accurate picture of the extent of earnings variability than do survey data, the analysis based on the CWHS is limited in several ways. Most notably, aside from age
and sex, no information on workers characteristics is available. Nor is any information available on the reasons for changes in workers earnings. CBO therefore supplemented administrative data from the CWHS with data from the SIPP, which contains information on workers levels of education and the reasons for which many
workers experience large declines in earningssuch as illness, unemployment, or
exiting the labor force to have or care for children.
METHODOLOGY

In its analysis of administrative records, CBO looked at a sample of workers

whose earnings information was collected by the Social Security Administration between 1980 and 2003. The measure of annual total wage earnings available for this
analysis includes wage and salary earnings, tips, and some other sources of compensation; it excludes self-employment earnings and deferred compensation. The
measure also includes earnings above the maximum amount subject to the Social
Security payroll tax. Earnings are indexed to 2006 dollars using the research series
for the consumer price index for all urban consumers. The analysis focuses on workers who were between 22 and 59 years old at any time during the 19802003 period.
For each worker, CBO calculated the percentage change in earnings from one year
to the next.2 CBO then calculated, in each year from 1981 to 2003, the fraction of
workers whose earnings fell by at least 50 percent from the previous year, the fraction whose earnings fell by at least 25 percent, the fraction whose earnings increased by at least 25 percent, and the fraction whose earnings increased by at least
50 percent.3 Because the extent to which earnings vary from year to year within
those categories is also important, CBO calculated the standard deviation of the
one-year change in earnings. The standard deviation can be used to construct an
interval (from the average percentage change minus the standard deviation value
to the average percentage change plus the standard deviation value) within which
roughly 80 percent of workers fall.4 CBO also calculated two additional measures
of variability: the fraction of workers in each year who had no earnings at all in
the previous calendar year, and the fraction of workers in each year who had no
earnings in the subsequent calendar year.
In its analysis of survey data from the 2001 panel of the SIPP, CBO focused on
the annual earnings of workers between the ages of 22 and 59 in 2001 and 2002.
The 2001 panel of the SIPP is the latest available from which the annual percentage
change in workers earnings can be calculated. Because the survey collects demographic information on workers, CBOs analysis was able to determine how the
changes in earnings varied with the workers education level and age. Finally, CBO
used information on the reasons for which individuals were not working to help provide insight into the causes of large declines in earnings.
1 See Congressional Budget Office, Projecting Labor Force Participation and Earnings in
CBOs Long-Term Microsimulation Model (October 2006).
2 Most existing studies adjust for workers ages. CBOs analysis does not; that is, a portion
of the trends in variability may be the result of the aging of the workforce. An analysis that
does account for age is presented in the appendix to this report.
3 Individuals with no earnings in both years of a two-year pairing are excluded from the analysis. Workers with no earnings in the first year and positive earnings in the second year of a
two-year pairing are coded as having a 100 percent increase in earnings; the percentage increase
in earnings for those workers would otherwise not be defined. The analysis of the trends in
earnings volatility is not sensitive to that choice. See the appendix for a discussion of how CBOs
analysis is related to that used in other studies.
4 CBO calculated this statistic on the basis of the empirical distribution of the one-year percentage change in total wage earnings in the CWHS.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00048

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

45
ANALYSIS OF VARIABILITY USING ADMINISTRATIVE DATA

Individual earnings tend to rise over a workers lifetime.5 From year to year, however, there is substantial variability in those earnings, according to data from the
CWHS. For example, between 2002 and 2003, one-in-five workers saw his or her
real (inflation-adjusted) earnings increase by at least 25 percent, and roughly the
same share of workers saw his or her earnings decline by at least 25 percent. A
substantial portion of workers, about one-in-seven, saw their earnings decline by at
least half.
Relatively little research to date has explored whether earnings variability has
risen over the past 20 years. Resolving questions about those trends is important
not only to inform policymakers, but also to allow CBO to construct more accurate
long-term projections of earnings for its analyses of the Social Security and Medicare programs.
To examine trends in earnings variability, CBO used administrative data from its
long-term Social Security model. Administrative data have advantages over survey
data because the administrative records yield very large samples of workers, allowing for more precise statistical analyses. Furthermore, administrative data more accurately measure year-to-year variability in earnings, because individuals responses
to surveyswhich rely on the respondents recallare often in error. Such error
could lead researchers to either overstate or understate workers actual changes in
earnings.6
Analyses using administrative data are also limited in a number of ways, however; the primary limitation is that, beyond the age and sex of the worker, little or
no demographic information is available. Moreover, the administrative data only reflect workers earnings: No information on workers family income or assets is available. Therefore, the analyses cannot examine how changes in a workers earnings
might be offset by changes in other sources of family income or by the existence of
financial assets. Furthermore, the analyses do not account for the impact of income
or payroll taxes. The tax system can help to smooth fluctuations in incomesometimes quite significantlyso workers after-tax income can vary less from year to
year than their pretax income does.
CBOs analysis of the CWHS administrative data indicates that, since 1980, the
trend in year-to-year earnings variability has been roughly flat. That finding is consistent with the results of existing studies, which tend to show more variability in
earnings in the 1980s and 1990s (on a percentage basis) than in the 1970s but relatively stable trends in earnings variability since about 1980.7
Although the trend in earnings variability has been roughly flat since 1980, it
does appear to vary with the business cycle; large declines in total wage earnings
were more frequent in years in which the growth rate of gross domestic product
(GDP) was relatively low. Between 1980 and 1981, for example, when the U.S. economy was in a recession and GDP growth was slowing, nearly one-in-five workers
experienced a 50 percent drop in earnings, and nearly one-in-four experienced a 25
percent drop in earnings, adjusted for inflation (see Figure 1 on page 9 and Figure
2 on page 10). By 1983, when the economy had recovered somewhat, only one-infive workers experienced a decline in earnings of at least 25 percent from one year
to the next and only 15 percent experienced declines of at least 50 percent. Since
2000, earnings variability has increased slightly: By 2003, almost one-in-five work5 For a discussion of trends in hourly wages, hourly wage dispersion, and earnings dispersion,
see Congressional Budget Office, Changes in Low-Wage Labor Markets Between 1979 and 2005
(December 2006); and Jonathan A. Schwabish, Earnings Inequality and High Earners: Changes
During and After the Stock Market Boom of the 1990s, Congressional Budget Office Working
Paper 200606 (April 2006).
6 See John Bound and Alan Krueger, The Extent of Measurement Error in Longitudinal Surveys: Do Two Wrongs Make a Right? Journal of Labor Economics, vol. 9, no. 1 (January 1991),
pp. 124; and Julian Cristia and Jonathan A. Schwabish, Measurement Error in the SIPP: Evidence from Matched Administrative Records, Congressional Budget Office Working Paper 2007
03 (January 2007).
7 See, for example, Peter Gottschalk and Robert Moffitt, The Growth of Earnings Instability
in the U.S. Labor Market, Brookings Papers on Economic Activity, no. 2 (1994); Steven Haider,
Earnings Instability and Earnings Inequality of Males in the United States: 19671991, Journal of Labor Economics, vol. 19, no. 4 (2001); Maury Gittleman and Mary Joyce, Earnings Mobility in the United States, 196791, Monthly Labor Review, vol. 118, no. 9 (September 1995),
pp. 313; Robert Moffitt and Peter Gottschalk, Trends in the Transitory Variance of Earnings
in the United States, Economic Journal, vol. 112, no. 478 (2002), pp. 6873.
Gottschalk and Moffitt (1994) examine earnings variability through 1984. Haider (1991) and
Gittleman and Joyce (1995) examine earnings variability through 1991. Finally, Moffitt and
Gottschalk (2002) examine earnings variability through 1996. Each study finds relatively stable
trends in comparable measures of variability after 1980.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00049

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

46
ers experienced at least a 25 percent drop in earnings and one-in-seven workers experienced a 50 percent drop.
The percentage of workers who experienced at least a 50 percent increase in earnings from one year to the next declined somewhat between 1981 and 2003from
about 23 percent to 16 percentand the percentage of workers who experienced at
least a 25 percent rise in earnings declined slightly, falling from 27 percent to 22
percent (see Figure 3 on page 11 and Figure 4 on page 12). Between 1980 and 2003,
women were more likely to have experienced large changes in earnings than men
were, although the difference between the two sexes narrowed over that period.
That narrowing occurred during a period in which the participation rate of women
in the labor force increased substantially.
The measures of earnings variability displayed in Figures 1 through 4 rely on
changes in earnings that are greater or less than prespecified amounts. An alternative measure, which incorporates changes of any size, is the standard deviation
of the one-year change in inflation-adjusted earnings. Unlike the other measures,
which generally show stable levels of variability since 1980, the measure of variability based on the standard deviation has declined somewhat over the 19812003
period (see Figure 5 on page 13).
CBOs analysis of earnings includes the variability that stems from transitions between years in which workers had no earnings and years in which they had positive
earnings. Both the percentage of workers in each year who did not have any earnings in the previous calendar year and the percentage of workers who did not have
any earnings in the subsequent calendar year have declined over the 19802003 period (see Figure 6 on page 14). In 1981, for example, 11 percent of workers had no
earnings in the previous year (1980) and 12 percent had no earnings in the subsequent year (1982). In 2002, by contrast, 5 percent of workers had no earnings in
the previous year (2001) and about 6 percent had no earnings in the subsequent
year (2003).
There was no increase in the level of earnings variability in selected years between 1980 and 2003 for workers of different ages or in the overall population. In
general, younger workers (those ages 22 to 29) tend to experience more variability
in earnings than do older workers (see Table 1 on page 7). Because older workers
have more stable earnings than do younger workers, earnings variability among all
workers should decline somewhat as the workforce ages. Indeed, the declines in variability observed in Figures 3 through 6, in part, are the result of that aging.8
In addition to analyzing the trends since 1980 in workers total wage earnings,
CBO analyzed the trend in variability since 1960 in the earnings on which workers
paid Social Security taxes. That measure of earnings is more limited than the measure of total wage earnings, because if a workers earnings exceed the Social Security
maximum taxable income, only that maximum value is reported. That maximum
was relatively low in the 1960s, so the analysis examines the fraction of workers
in the bottom two quintiles (or fifths) of the earnings distribution who experienced
large declinesof 25 percent or 50 percentin their Social Security taxable earnings. The changes in the maximum taxable income would not be expected to affect
those workers because the maximum is above the 40th percentile of annual earnings
throughout the 19602003 period.
Between the early 1960s and the early 1980s, the fraction of male workers in the
bottom two quintiles of the earnings distribution who experienced at least a 50 percent decline in their Social Security taxable earnings over the previous year increasedfrom roughly one-in-six workers in 1961 to one-in-four workers in 1982
(see Figure 7 on page 15). Between 1982 and 2003, by contrast, there was little
change in earnings variability for male workers (although it did vary with the business cycle, increasing slightly during the 1991 and 2001 recessions).
The pattern differs significantly for female workers. Between the early 1960s and
the mid 1980s, the percentage of female workers who experienced 50 percent or
greater declines in earnings fell from 30 percent to less than 25 percent. Since 1984,
earnings variability among female workers has been roughly constant. For all workers in the bottom two quintiles of the earnings distribution, there has been little
change in this measure of earnings variability over the entire 19602003 period.
For workers in the bottom two quintiles whose Social Security taxable earnings
fell by at least 25 percent from one year to the next between 1961 and 2003, the
trends are similar to those displayed in Figure 7. The overall trend in earnings variability between 1960 and 2003 for all workers has been roughly flat (see Figure 8
on page 16). The results for male workers are consistent with most existing studies
8 See the appendix for a discussion of an analysis that more closely follows that of Gottschalk
and Moffitt (1994). In particular, that analysis controls for workers ages and excludes workers
who transition between years of no annual earnings and years with positive earnings.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00050

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

47
that find less earnings variability in the late 1960s and 1970s than in the 1980s
and early 1990s. They do suggest, however, that there may have been a decline in
earnings variability among women that offset the increase among men.
ANALYSIS OF VARIABILITY USING SURVEY DATA

To determine how changes in earnings varied by workers characteristics and to

examine potential reasons for large changes in workers earnings, CBO analyzed recent data from the Survey of Income and Program Participation. The analysis focused on workers ages 22 to 59. As with the analysis based on administrative data,
this analysis showed substantial variation in workers earnings from 2001 to 2002.
Over that one-year period, one-in-four workers saw his or her earnings increase by
at least 25 percent after inflation, while one-in-five saw his or her earnings decline
by at least 25 percent. A substantial portion of workers, 11 percent, saw their earnings decline by at least half (see Figure 9 on page 17).
Workers with less education tend to experience more volatility in their earnings
than do workers with more education (see Table 2 on page 8). For example, from
2001 to 2002, 16 percent of workers without a high school education had their earnings decline by 50 percent or more, compared with 10 percent of workers with more
than a high school education.
Such fluctuations in earnings can result from many sources, including job
changes, losses, or gains; voluntary exits from the labor force, perhaps to care for
children or other family members; changes in the number of hours worked per year;
or changes in the wage rate received by workers. Most workers who experienced at
least a 50 percent drop in earnings between 2001 and 2002 were not working for
at least one month and typically did not work for nine months in 2002. When those
survey respondents were asked why they were not working, the most common answers were that they were caring for a child or other family member or were pregnant; were not able to find work or had been laid off; were unable to work because
of disability, illness, or injury; or were not interested in working or were retired.9
The responses appear to be split evenly between workers suggesting that their departure from the labor force was voluntary and those suggesting that it was not.
For another point of comparison, CBO conducted its analysis using data from
1997 to 1998a period of relatively rapid economic growth, in contrast to the relatively slow growth from 2001 to 2002and found similar results.10
CONCLUSION

CBOs analysis finds that a significant number of workers experience substantial

variability in their total wage earnings from year to year. An examination of trends
over the past 20 years shows little change in such earnings variability for both men
and women. The reduction in macroeconomic volatility over the past several decades
does not appear to have translated into lower levels of variability in workers earnings. CBO will be examining trends in family income variability in its future work.

Table 1.Percentage of Workers for Whom Total Wage Earnings Dropped or Rose by 50 Percent
or 25 Percent, by 10-Year Age Category
(Percent)
Total Wage Earnings
50 Percent
Drop

Ages 20 to 29
1983 ..............................................................................................
1993 ..............................................................................................
2003 ..............................................................................................
Ages 30 to 39
1983 ..............................................................................................
1993 ..............................................................................................
2003 ..............................................................................................
Ages 40 to 49
1983 ..............................................................................................

25 Percent
Drop

25 Percent
Rise

50 Percent
Rise

17.5
16.0
16.4

23.2
22.4
23.7

36.0
33.2
32.6

29.6
25.6
24.8

15.1
14.1
13.8

19.9
19.4
19.9

29.5
23.0
22.2

24.4
17.5
16.2

13.7

18.1

25.4

20.9

9 Only those survey respondents who had at least four consecutive months without a job were
asked this question.
10 The data are from the 1996 and 2001 panels of the SIPP, the latest panels available for
which the annual percentage change in workers earnings can be calculated.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00051

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

48
Table 1.Percentage of Workers for Whom Total Wage Earnings Dropped or Rose by 50 Percent
or 25 Percent, by 10-Year Age CategoryContinued
(Percent)
Total Wage Earnings
50 Percent
Drop

1993 ..............................................................................................
2003 ..............................................................................................
Ages 50 to 59
1983 ..............................................................................................
1993 ..............................................................................................
2003 ..............................................................................................
All Workers Ages 22 to 59
1983 ..............................................................................................
1993 ..............................................................................................
2003 ..............................................................................................

25 Percent
Drop

25 Percent
Rise

50 Percent
Rise

12.0
11.9

16.8
17.1

18.5
17.8

14.0
12.8

15.1
14.6
13.1

19.7
19.7
18.6

21.7
15.7
14.2

18.1
12.0
10.3

15.5
14.1
13.6

20.5
19.5
19.5

29.4
23.1
21.3

24.2
17.7
15.7

Source: Congressional Budget Office based on data from the Social Security Administrations Continuous Work History Sample.
Note: Total wage earnings include wages and salaries, tips, and other forms of compensation; they exclude self-employment earnings and
deferred compensation. Workers without any earnings in the previous calendar year are included, and their percentage change in earnings is
coded as 100.

Table 2.Distribution of Changes in Workers Annual Real Earnings, by Educational Attainment

and Age, 2001 to 2002
(Percent)
Decrease in Earnings of
At Least
50 Percent

All Workers Ages 22 to 59 ..................................

Educational Attainment
Less than high school ....................................
High school .....................................................
More than high school ....................................
Age
22 to 30 ..........................................................
31 to 40 ..........................................................
41 to 59 ..........................................................

Changes in
Earnings of
Less Than 25
Percent

25 Percent

Increase in Earnings of
At Least
25 Percent

50 Percent

11.3

20.2

52.2

27.6

17.4

15.9
12.4
10.1

25.9
20.8
19.0

43.8
51.7
53.7

30.3
27.6
27.2

21.5
17.5
16.8

12.8
11.0
10.9

21.4
19.7
19.9

45.3
52.7
54.9

33.3
27.6
25.2

22.1
17.1
15.6

Source: Congressional Budget Office based on data from the 2001 panel of the Bureau of the Censuss Survey of Income and Program Participation.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00052

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

49

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00053

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

50

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00054

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

51

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00055

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

52

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00056

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

53

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00057

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

54

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00058

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

55

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00059

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

56

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00060

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

57

APPENDIX:
ALTERNATIVE MEASURES

OF

EARNINGS VARIABILITY

The results presented in the main text are based on the methodology used by
Peter Gottschalk and Robert Moffitt in their paper titled The Growth of Earnings
Instability in the U.S. Labor Market, which was published in the Brookings Papers
on Economic Activity series in 1994 (no. 2, pp. 217272).
The Congressional Budget Offices (CBOs) primary analysis uses administrative
data from the Continuous Work History Sample (CWHS). Those data are provided
by the Social Security Administration to CBO so that CBO may closely examine patterns in earnings over time and continue to improve the accuracy of its long-term
models of the Social Security and Medicare programs.
The use of the CWHS involves trade-offs.1 On the one hand, administrative data
are well-suited to an examination of year-to-year variability in earnings, as the data
are not subject to the same measurement error as are survey data, which rely on
the survey respondents recall. The presence of that measurement error may cause
one to overstate or understate the actual change in earnings from year to year.2
Furthermore, the CWHS data contain a large number of observations, allowing for
relatively precise statistical analyses. On the other hand, the CWHS is limited in
scope in that it only contains reliable data on an individuals earnings, birth year,
1 For a comparison of CWHS data to survey data from the Current Population Survey, see
Jonathan A. Schwabish, Earnings Inequality and High Earners: Changes During and After the
Stock Market Boom of the 1990s, Congressional Budget Office Working Paper 200606 (April
2006).
2 See John Bound and Alan Krueger, The Extent of Measurement Error in Longitudinal Surveys: Do Two Wrongs Make a Right? Journal of Labor Economics, vol. 9, no. 1 (January 1991),
pp. 124; and Julian Cristia and Jonathan A. Schwabish, Measurement Error in the SIPP: Evidence from Matched Administrative Records, Congressional Budget Office Working Paper 2007
03 (January 2007).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00061

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

58
and sex. There is no additional information on the individual, such as education,
nor is there any information on the individuals family members. Using those data
alone, one cannot examine the circumstances under which a change in earnings occurredwhether it is the result of a job change, job loss, job gain, or changes in
hours worked or wages paid at the same job. Nor can one examine whether a
change in earnings was mitigated or exacerbated by changes in the earnings of
other family members. In addition, there is no information on other sources of income or assets, both of which could serve as important buffers against the consequences of changes in earnings (especially a decline in earnings).
Earnings in the CWHS are total wage earnings; they include wages and salaries,
tips, and other forms of compensation and are not subject to top-coding. Self-employment earnings and deferred compensation are excluded. The earnings are pretax;
the mitigating effect of the tax system on the consequences of changes in earnings
cannot be captured here. Finally, earnings are inflation-adjusted, using the research
series for the consumer price index for all urban consumers.
The sample consists of males and females ages 22 to 59, which results in the (intentional) exclusion of many transitionsfrom school to work, for example, or from
work to retirementfrom the analysis.
The results presented in Figures 1 through 5 in the main analysis are based on
one measure of earnings variability: the inflation-adjusted percentage change in a
persons earnings between a given year (et) and the previous year (et 1), calculated
as

That measure is undefined for individuals with earnings of zero in both years;
those individuals are excluded from the analysis. The treatment of workers with
positive earnings in one year and zero earnings in the other is asymmetric, as those
individuals who transition from positive to zero earnings have a calculated change
in earnings of -100 percent. For workers who transition from zero to positive earnings, the percentage change in earnings is undefined. To capture those transitions
symmetrically in Figures 1 through 5, CBO assigned those workers moving from
zero to positive earnings a percentage change in earnings of +100 percent.
Gottschalk and Moffitt (1994) measure the percentage change in earnings somewhat differently. Instead of comparing earnings in a given year with earnings in the
previous year, they compare earnings in a given year with a five-year moving average of earnings around that year. To determine whether the results presented in
its main analysis are sensitive to such a distinction, CBO examined the percentage
change in a workers earnings between a given year (et) and the average earnings
of that worker over a five-year period (et 4 to et), calculated as

The measure is undefined for individuals with no earnings in all five years; those
individuals are excluded from the analysis.
The results presented in the main analysis are robust to that slight change in
methodology. The fraction of workers experiencing a 50 percent or 25 percent decline in their earnings remains relatively stable over time (see Figure A-1 on page
23 and Figure A-2 on page 24), while the fraction of workers experiencing a 50 percent or 25 percent increase in their earnings trends slightly downward over time
(see Figure A-3 on page 25 and Figure A-4 on page 26). That downward trend in
the measure of variability remains, even when CBO examined the standard deviation of the percentage change (which captures the entire distribution of changes)
rather than focusing on single points in the distribution of changes (see Figure A5 on page 27).
Another difference between the methodology used in this analysis and that used
by Gottschalk and Moffitt (1994) and in many other studies is that those studies
examine variability in the natural logarithm of earnings and also control for the age
of the worker. Using the natural logarithm of earnings in place of the level of earnings eliminates workers with any years of zero earnings; thus, changes in earnings
between years of zero earnings and years with positive earnings would not be included in this measure of variability. As shown in Figure 6 of the main analysis,
roughly 6 percent of workers in the latter part of the period had no earnings in either the prior or subsequent year.
To determine whether the results in the main analysis are sensitive to those differences in specification, CBO first conducted its analysis using the natural logarithm of earnings and, second, estimated a fixed-effects model in which the natural

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00062

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

59
logarithm of earnings for all individuals in all years is regressed on a quartic in age.
The residuals (t) from that regression were calculated for each individual. For a
given individual, the five-year moving average of those residuals was used as the
basis of the percentage difference, calculated as
The findings using the natural logarithm of earnings are presented in Figures A6 through A-10. Comparing those results with the results in the main analysis, the
trends over time in the fraction of workers experiencing a 50 percent or 25 percent
decline in earnings remains relatively stable (see Figure A-6 on page 28 and Figure
A-7 on page 29). Eliminating transitions between years of zero earnings and years
of positive earnings eliminates any downward trend in the fraction of workers experiencing a 50 percent or 25 percent increase in earnings over time (see Figure A8 on page 30 and Figure A-9 on page 31). And, finally, examining the standard deviation (and thus capturing the full distribution of changes over time), a small portion
of the downward trend seen in Figure 5 is eliminated (see Figure A-10 on page 32).
Adopting the natural log specification and controlling for workers age results in
even flatter trends over time than were observed in the previous two specifications
(see Figures A-11 through A-15). The consistent flattening of the trends in earnings
variability after controlling for age suggests that a portion of the decline in the variability in earnings seen in Figures 1 through 5 in the main analysis is probably because of the aging of the population. As the population of workers ages, older workers, who tend to have less-variable earnings, make up a larger fraction of the overall
population. As a result, workers overall have less-variable earnings.
The results presented in this report are consistent with those of Gottschalk and
Moffitt (1994) for the early 1980s (the only years for which the two analyses overlap). Both show relatively stable levels of earnings variability during that period.
The results presented in Figure A-10 are consistent with the findings of other studies that use more-formal statistical models of earnings dynamics. Those studies include later work by Moffitt and Gottschalk (Trends in the Transitory Variance of
Earnings in the United States, published in The Economic Journal in 2002) as well
as work by Steven Haider (Earnings Instability and Earnings Inequality of Males
in the United States: 19671991, published in the Journal of Labor Economics in
2001). Haider examined earnings variability through 1991, and Moffitt and
Gottschalk captured variability in earnings through 1996.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00063

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

60

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00064

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

61

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00065

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

62

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00066

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

63

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00067

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

64

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00068

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

65

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00069

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

66

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00070

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

67

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00071

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

68

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00072

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

69

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00073

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

70

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00074

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

71

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00075

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

72

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00076

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

73

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00077

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

74

PREPARED STATEMENT OF DR. LAEL BRAINARD, VICE PRESIDENT AND DIRECTOR,

GLOBAL ECONOMY AND DEVELOPMENT, THE BROOKINGS INSTITUTION
Mr. Chairman, Members of the Committee, thank you for the opportunity to testify before your committee. Your focus on income fluctuations is all too real for many
American middle class families today and is likely to be a reality for many more
in coming years. It is worth spending a minute on some of the likely economic drivers before turning to one of the promising policy responses.
A NEW WAVE OF GLOBALIZATION

A new wave of globalization has reached our shores. Although the individual elements feel familiar, the combined contours are unprecedentedin scope, speed and
scale.
Because China is successfully pursuing at a scale never seen before a growth
strategy that is export-led and foreign direct investment fed, its rise is sending
waves to the farthest reaches of the global economy. China is already deeply embedded in global manufacturing supply chains, confronting higher wage producers with
the difficult choice of moving up the value chain or lowering costs.
Indias concurrent economic emergence has complicated the challenge. While India
is pursuing a growth strategy more reliant on domestic consumption and investment
than China, nonetheless its success in exporting higher skilled knowledge services
such as software programming has expanded the scope of globalization. Many Americans in white collar occupations are confronting the reality of low wage foreign competition for the first time.
The current episode of global integration dwarfs previous expansions: the entry
of India and China into the global labor force amounts to an expansion of roughly
70 percentconcentrated at the lower end of the wage scale. Textbook economics

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00078

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

75
would predict a squeeze on wage earners until capital and technology investments
adjust. Indeed, the data suggests inequality is once again on the rise in many of
the worlds richer economies.
In the United States, profits are capturing a larger share of income and wages
a lower share than at any time in the last 50 years. Moreover, economists David
Autor, Larry Katz, and Melissa Kearney have pointed out that the gap between the
middle and top of the U.S. wage distribution (between the 90th and 50th percentile)
appears to be widening today, in contrast to earlier decades, where the focus was
on the gap between the bottom and middle (between the 50th and 10th percentiles).
A WEAK SAFETY NET

In the face of accelerated job losses in manufacturing and white-collar offshoring

in services, an ever-broader pool of American workers is finding that the nations
safety net has more holes than netting.
Despite the fact that the U.S. labor market ranks second to none when it comes
to job turnover, the nations safety net for easing job transitions remains one of the
weakest among the wealthy economies. Not only do U.S. unemployment benefits
have a short duration, but Americas heavy reliance on an employer-based system
of insurance means that displaced workers face the prospect of losing health and
pension benefits along with income. For permanently displaced workers, average
earnings in the new job are 16 percent lower than earnings in their previous job,
while displaced manufacturing workers generally face a 20 percent drop in earnings.
The consequences of job loss are particularly damaging in import-competing industries, where displaced workers face longer spells of unemployment and greater permanent wage declines than do workers in other industries.
Americas safety net is miserly in comparison with those of almost every other advanced economy. The main federally mandated unemployment insurance (UI) program contains so many restrictions that today only about 40 percent of all jobless
workers receive benefits.
The last serious overhaul of the nations employment safety net was in 1962,
when President John F. Kennedy established the TAA program to compensate workers who suffer job loss as a result of trade liberalization. But workers have long
found it difficult, time-consuming, and expensive to prove that they are entitled to
extended unemployment benefits under the nations Trade Adjustment Assistance
(TAA) program.
In 2002 Congress overhauled and expanded TAA, adding a health care tax credit,
doubling the training budget, and substantially raising budget outlays for income
support. But the TAA program continues to disappoint. Participation has remained
surprisingly low, thanks in part to confusing Department of Labor interpretations
and practices that ultimately deny benefits to roughly three-quarters of workers
who are certified as eligible for them. TAA has helped fewer than 75,000 new workers per year, while denying more than 40 percent of all employers petitions. And
remarkably, the Department of Labor has interpreted the TAA statute as excluding
the growing number of services workers displaced by trade.
Despite its laudable goals, the TAA program has repeatedly failed to meet expectations. Between 2001 and 2004, an average of only 64 percent of participants found
jobs while they participated in TAA. And earnings on the new job were more than
20 percent below those prior to displacement.
THE CASE FOR WAGE INSURANCE

With workers more likely to face permanent displacement and experiencing average income declines of 16 percent when they are reemployed following displacement,
the time has come for the Federal Government to augment existing programs by
adopting a new insurance program that insures against wage loss, not just unemployment, for permanently displaced workers.1 Wage insurance would smooth income fluctuations while encouraging displaced workers to broaden their employment
search. It also defrays the cost to employers of hiring and providing on-the-job training to new employees from different sectors. On aggregate, wage insurance could
lead to shorter spells of displacement and more efficient reskilling for workers.
A chief goal of wage insurance is to smooth the incomes of workers who suffer
permanent displacement and declines in their earnings. Wage insurance is most
likely to have overall positive economic benefits if it targets workers whose earnings
would otherwise fall dramatically as forces outside their control devalue their skills.
By replacing some of the lost earnings, wage insurance encourages more rapid reem1 Jeffrey Kling, Lori Kletzer, Robert Litan, and Howard Rosen have put forth a variety of proposals for wage insurance.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00079

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

76
ployment; a Canadian pilot wage insurance program reduced unemployment durations by 4.4 percent on average.
Wage insurance can act as a subsidy of on-the-job training for the workers new
employer. Generalized retraining programs not only fail to guarantee a worker a job
but also cost the worker the wages that he or she could earn by accepting new employment sooner. The retraining that a displaced worker receives on a new job provides new skills that contribute directly to his or her performance in the new job
and is thus directly useful not only to the worker but also to the new employer.
Finally, evidence suggests that wage insurance encourages workers to consider
different types of jobs and sectors of employment and, therefore, broadens the job
search. This is particularly important for displaced workers whose firm-specific
skills have declined in value.
Most programs designed to ease job transitions entail a tradeoff between the degree of eligibility targeting and participation rates. While targeted programs should
be more cost-effective in principle, targeting requires burdensome eligibility and
compliance requirements that sharply lower participation rates and sometimes introduce stigma. The TAA experience argues strongly for a less targeted program implemented through an existing system with proven efficacy, such as the UI system
or though the tax system as a refundable tax credit.
Moreover, if the goal is to provide some degree of insurance against extreme income fluctuations, wage loss insurance should be available to all permanently displaced workers, who have at least 2 years of tenure at the previous job. It might
also make sense to restrict the program to workers displaced from full-time jobs and
reemployed full-time, so as to avoid any possible incentive to reduce hours of work.
Further, the compensation period would be limited to some initial period, perhaps
2 years, long enough to help strengthen the new employment relationship during
the period when on-the-job-training is arguably most concentrated.
The wage loss replacement rate, the duration of benefits, and the annual cap on
compensation determine the kinds of workers who would benefit most from the program. A high replacement rate combined with a low annual compensation cap would
provide the greatest cushion to lower-income workers suffering steep losses in earnings, while a lower replacement rate combined with a high annual cap would tilt
compensation toward higher income earners.
According to our estimates, a wage insurance program that replaces 50 percent
of earnings losses for long tenure full-time displaced workers up to a maximum of
$10,000 per year for up to 2 years would cost roughly $3.5 billion per year, using
a conservative estimate of offsetting savings in other unemployment and training
programs. On a per worker basis, this cost falls midway between the current unemployment and retraining benefits available under UI and Worker Investment Act
(WIA) programs and the comprehensive cost of TAA benefits.
Under such a program, an average trade-displaced worker, who earned $37,382
in 2004 and was reemployed with a 26 percent loss rate at $27,662 would instead
receive $33,522 for the first 2 years after reemployment, thus enabling them to
smooth their income while becoming more valuable in the new job.
Of course, the costs can be substantially reduced by offering more modest benefits.
For a high-unemployment year such as 2003, costs could range from a low of $1.6
billion for a 1-year program with a 30 percent replacement rate and a $10,000 cap
to a high of $7 billion for a 2-year program with a 70 percent replacement rate and
a $20,000 annual cap.
How do we think about the price tag? For a relatively robust program, the net
cost of $3.5 billion per year amounts to an insurance premium of roughly $25 per
worker per year. One simple way to finance the uncovered costs of wage insurance
would be through a modest increase in the current Federal unemployment tax
(FUTA) with the incidence split between employers and employees.
Wage insurance could provide an important tool in a broader set of policies designed to help American middle class families insure against disruptive income fluctuations, while preserving the benefits of a dynamic economy. For the price of $25
per worker per year, the Nation reaps economic benefits in the form of less income
volatility and more rapid reemployment. Wage insurance could be an important policy tool to help make work pay following displacement; the intention is to augment
the insurance available to middle class Americans facing the possibility of greater
income volatility, to augment the programs current availablenot to replace them.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00080

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

77
PREPARED STATEMENT

MAURICE EMSELLEM, POLICY DIRECTOR, NATIONAL

EMPLOYMENT LAW PROJECT

OF

Chairman and members of the Committee, thank you for this opportunity to testify today on the critical subject of economic insecurity in the United States and
offer our perspective on proposals to create a new program of wage insurance and
other options for Federal reforms.
My name is Maurice Emsellem, and I am the Policy Director for the National Employment Law Project (NELP), a non-profit research and advocacy organization that
specializes in economic security programs, including unemployment insurance,
Trade Adjustment Assistance (TAA) and the workforce development system. Our organization has worked in the states and with Congress to protect the nations economic security programs against serious attacks in recent years and successfully
promote reforms that deliver on the nations promise of economic opportunity.
We worked with Members of Congress to advocate for the extension of unemployment benefits during the last recession and for major improvements in the Federal
program of benefits provided to the families left jobless by Hurricanes Katrina and
Rita. We also have a special project working with state officials in the Midwest to
help those workers laid-off from the auto industry to better access trade act benefits
and other programs. Thus, we have a long-standing interest and commitment to
policies that serve the interests of families hardest hit by economic downturns in
the U.S. and the fallout from globalization.
Today, we hope to call attention to some key unanswered questions about wage
insurance given the interest in possible Federal legislation. Like the AFL-CIO and
several unions that have expressed concerns with wage insurance,1 we believe that
there are important questions that remain unanswered given the limited experience
with the program. We are especially concerned that wage insurance will also promote more downward mobility, not good jobs, by subsidizing mostly low-wage employment. If adopted in the U.S., wage insurance could also undermine funding and
support for existing economic security programs, including unemployment insurance
and Trade Readjustment Assistance.
As described below, there are other immediate Federal priorities, including reform
of the TAA program and an expansion of the unemployment insurance system,
which could go a long way to promote economic opportunity and support the families
hardest hit by long-term layoffs. In conclusion, we also highlight some of the most
promising state innovations that could be incorporated into Federal law to protect
working families against major economic hardship and help rebuild their communities.
A. KEY WAGE INSURANCE QUESTIONS

1. Does Wage Insurance Promote More Downward Mobility?

By definition, wage insurance compensates workers who take lesser paying jobs,
which are the same jobs that are less likely to pay benefits, like health insurance,
that are critical to working families in todays unstable economy. Most economists
who support wage insurance also argue that it creates an incentive for workers to
be re-employed faster and thus reduces the period they collect their unemployment
benefits.
We are especially concerned that wage insurance promotes more downward economic mobility rather than new labor market policies that support quality jobs with
benefits. In other words, wage insurance is not merely added income to help families
get by during hard times. Nor is it like universal insurance promoted by Professor
Jacob Hacker, which provides compensation to those who suffer major economic
hardships. Instead, wage insurance is expressly contingent upon the worker accepting a lesser-paying job.
If the goal is to support reform of low-road jobs that increasingly dominate the
economy, then our reemployment strategies should do everything possible to promote good jobs. Federal policy can play a critical role but first Congress must not
to endorse rapid reemployment proposals like wage insurance that encourage more
low-road employment, or at least fail to distinguish between good and bad employment outcomes.
1 Testimony of Bill Samuel, AFL-CIO Legislative Director, Hearing on Unemployment Compensation Aspects of U.S. Department of Labor Fiscal Year 2007 Budget: Hearing Before the
House Ways & Means Committee, Subcommittee on Income Security and Family Support, 109
Cong. (2006).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00081

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

78
2. Does the Available Research Make A Convincing Case for Wage Insurance?
Despite all the attention generated in support of wage insurance by economists
and others, there has been remarkably limited scrutiny of the research on wage insurance. We believe the available evidence raises fundamental questions about the
merits of wage insurance that should be more closely evaluated before pursuing
Federal legislation.
First, other than two pilot programsone in Canada that produce limited results
and another in the U.S. that is still pendingwage insurance is not a program that
has existed on any large scale. Indeed, we question whether it is premature to create a new national program of wage insurance in the U.S. when the 2003 pilot, the
Alternative Trade Adjustment Assistance (ATAA) program, has not yet issued its
final findings. If it turns out that wage insurance is not working for the targeted
group of trade impacted workers age 50 and over who are having the hardest time
finding a new job at comparable pay, then why expand the program to those younger than 50 and to all dislocated workers as some have proposed?
Second, what do we know about the impact of wage insurance on others who will
be competing for the same lesser-paying jobs with those who are collecting wage insurance? According to a leading Upjohn Institute researcher who simulated the impact of a 2-year wage insurance program covering dislocated workers at half their
prior pay, virtually all the employment gains experienced by dislocated workers as
a result of the wage subsidy come at the expense of other workers. 2 Will this
crowding out effect be even more severe in communities hardest hit by job losses,
as in the Midwest, where large concentrations of dislocated workers are now competing with other workers for the same jobs?
Third, if wage insurance encourages workers to take a job sooner, will they also
end up taking lower paying jobs than they could have found if they kept looking
for work with the help of their unemployment benefits? This gets at the critical
tradeoff that laid-off workers constantly have to make, which is whether to take a
lesser paying job or collect unemployment benefits and continue looking for a better
job that will also increase their productivity. We know, for example, that workers
who collect UI have an increased likelihood of finding a new job that will have employer-sponsored health insurance.3 In addition, at least one study has found that
workers who receive unemployment benefits receive higher pay as well by a factor
of $240 a month compared to those who do not collect UI benefits.4
Fourth, will workers who take lesser paying jobs with wage insurance benefit
from any training that will improve their long-term productivity or would they be
better off pursuing other forms of education and training? While some have argued
that wage insurance leads to valuable training,5 we are not aware of any empirical
evidence suggesting that workers who find jobs at half their prior pay are likely to
receive substantial training that will significantly increase their earnings potential.
In fact, wage insurance will often interfere with valuable education and training,
including some community college programs that have produced major gains in income.6 Notably, the ATAA pilot program precludes the workers from collecting wage
insurance while participating in training.
Finally, what are the major lessons learned from the only empirical experience
with wage insurance, the Canadian pilot program of the 1990s? The Canadian program, called the Earnings Supplement Project, was evaluated by a leading research
organization in a random assignment study (comparing a group that could collect
wage insurance replacing up to 75 percent of their prior wages with a control group
that could not). On nearly every measure they evaluated, focusing on the impact on
2 Davidson, Woodbury, Wage-Rate Subsidies for Dislocated Workers (Upjohn Institute Staff
Working Paper 9531, January 1995), at page 22.
3 Boushey, Wenger, Finding the Better Fit: Receiving Unemployment Insurance Increases
Likelihood of Re-Employment with Health Insurance (Economic Policy Institute: April 2005).
4 Kiefer, Neumann, An Empirical Job Search Model with a Test Constant Reservation Wage
Hypothesis, Journal of Political Economy, Vol. 87, No. 1, 89107.
5 Brainard, Litan, Warren, Insuring Americas Workers in a New Era of Offshoring (Brookings Institution, Policy Brief #143, 2005), at page 3 (Wage insurance also serves as a training
subsidy for the workers new employer. Generalized retraining programs not only fail to guarantee a worker a job but also cost the worker the wages that he or she could earn by accepting
new employment sooner. The retraining that a displaced worker receives on a new job is the
best kind: it provides new skills that contribute directly to his or her performance in the new
job and is thus directly useful not only to the worker but also to the new employer.)
6 Trutko, Barnow, Farrell, Glosser, Final Report: Earnings Replacement Outcomes for Dislocated Workers: Extent of Variation and Factors Accounting for Variation in Earnings Replacement Outcomes Across State and Local Workforce Investment Boards (Capital Research Corporation: March 2005), at page A-8 (summarizing the results of various community college programs on dislocated worker post-displacement earnings, including Pennsylvania where men
earned $1,047 more per quarter by attending community college and woman earned $812 more.)

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00082

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

79
employment and unemployment benefits, wage insurance fell far short of expectations. Thus, the Canadians did not continue the program.
Of special significance to the U.S., the study found that of those assigned to the
group who could collect wage insurance, only about 2 out of 10 actually did so. When
follow-up interviews were conducted to better understand this result, the researchers found that one of the most striking findings from the non-recipient groups was
the limited perceived relevance of the supplement offer . . . 7 Quoting one participant that typified the concerns they found, It [the supplement] was secondary. It
was a not a priority. The priority was to get a job. I would like a good fit considering
my background and my education so I wasnt willing to settle. It wasnt a money
issue really. It may be that the low take-up rate in the ATAA program reflects a
similar concern with wage insurance.
The results of the Canadian program also showed virtually no difference in the
duration of [UI] benefits paid to recipients (22.1 weeks for supplement group members versus 21.9 weeks for control group members). 8 This finding conflicts with the
claims of some researchers that wage insurance in the U.S. will produce savings
based on reduced reliance on UI benefits. Late into the period when the workers
started collecting UI, there was a modest impact on how many more workers found
full-time work when they collected wage insurance. However, that impact was reduced in half when the study counted those in the control group who found parttime work (bringing the employment rate to 50.7 percent for those who could collect
wage insurance compared to 48.4 percent for the control group).9
3. Will Wage Insurance Undermine Existing Economic Security Programs?
We are also concerned that a new national program of wage insurance for dislocated workers could undermine funding and support for necessary reforms of existing economic security programs, especially unemployment insurance and Trade Adjustment Assistance. At a time when economic security is a growing reality for
working families from all walks of life, the existing economic security programs are
struggling from limited resources and years of neglect and hostile oversight by the
Bush Administration.
Take the case of the unemployment insurance program. Today, only 36 percent
of unemployed workers collect jobless benefits due in large part to the major gaps
in the program that leave out large numbers of low-wage, part-time and women
workers. Meanwhile, Federal funding for administration of the program has declined compared to the increased demand for services, which has caused states to
severely cut back on UI services. The states have also cut UI payroll taxes to record
low levels, creating more pressure to deny benefits and take out loans from the Federal UI trust funds. Despite the new pressures on the Federal trust funds, Congress
has also failed to increase the $7,000 tax base on Federal UI payroll taxes for nearly
25 years.
The Trade Adjustment Assistance (TAA) program serving trade impacted workers
has also been severely compromised, both by the Bush Administrations attacks on
the program and by limited funding and program restrictions imposed by Congress.
Despite the record trade deficit and major manufacturing layoffs, Congress has
capped TAA training funds at just $220 million, thus providing training to fewer
than 38,000 workers in 2005. As a result of the funding limits, 19 states also suspended enrollment in training at some point between Fiscal Years 2001 and 2003.10
And this Fiscal Year, Michigan has already been forced to suspend enrollment in
TAA training despite devastating layoffs in the auto industry.
Given these sobering realities, our concern is that the funding (estimated at $3.5
billion) and support for wage insurance will take precedence over long-overdue reforms of the TAA and UI programs. Whatever the ultimate source of revenue to pay
for wage insurance, whether it is generated from increased Federal UI payroll taxes
or new employer taxes (some have also suggested that employee taxes help pay for
the program), it will effectively compete with funding for the UI program. And if
the Canadian experience holds true in the United States, that wage insurance did
not result in reductions in UI benefits, then the funding constraints will be even
more severe.
In addition to the funding threat, there is a potential substantive threat to existing economic security programs created by wage insurance. Specifically, wage insur7 Bloom, et al., Testing a Re-employment Incentive for Displaced Workers: The Earnings Supplement Project (Social Research & Demonstration Corporation: May 1999), at page 39.
8 Id. at page 53.
9 Id.
10 U.S. General Accountability Office, Reforms Accelerated Training Enrollment, But Implementation Challenges Remain (GAO041012), September 2004, at page 32.
Insert GAO Report.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00083

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

80
ance promotes the work first agenda of the Heritage Foundation and other groups
that are working hard to dismantle the TAA program. According to the Heritage
Foundation, If the aim of such programs is to help workers find new jobs, then the
TAA should be eliminated over time and replaced by a program that provides incentives, not disincentives, for workers to do just that. Wage insurance is one such proposal that has won widespread support. 11
B. FEDERAL ECONOMIC SECURITY PROPOSALS

These are tough times for many more working families, full of concern that they
will not share in the promise of the American dream, or worse, that they will end
up destitute despite a lifetime of hard work. What follows are several proposals for
Federal policies that we believe will help create a reemployment system driven by
the creation of quality jobs that will also restore confidence in the nations workers
that their government is there to support them and create new opportunities especially in times of special financial need.12
1. Honor the Promise of Economic Security to Trade-Impacted Workers
The first priority of the 110th Congress should be to fulfill the promise of economic security to the nations workers and their communities that have suffered
major job losses due to Federal trade policies. Given the record trade deficits and
the devastating loss of good-paying manufacturing jobs resulting from Federal trade
policies, Congress should move boldly to create a more robust TAA program.
Congress should start by establishing an entitlement to TAA training, thus removing the $220 million cap on funding that now deprives training to thousands
of deserving workers who have been certified as TAA eligible. The entire TAA program is funded at $1 billion a year, which compares with the $3.5 billion in funding
being proposed to create a new wage insurance initiative. A serious new investment
of funding in the TAA program could also pay for coverage of service workers, a new
system of TAA certification that applies to whole industries and regions suffering
dislocations due to trade, and other necessary reforms.
2. Modernize and Expand the Unemployment Insurance System
Recognizing the changing nature of unemployment in todays economy, with far
more long-term joblessness and increasing turnover of low-paying service sector
jobs, it is also time to modernize and expand the nations unemployment insurance
system.
The 110th Congress should make Federal incentive funds available to the states
to support innovative reforms that fill the gaps in the program that deny benefits
to low-wage, part-time and woman workers. Federal funding should also target
states that support training and education with the help of extended unemployment
benefits and that increase the duration of unemployment benefits recognizing the
new realities of long-term unemployment.
In addition, the states should be more adequately compensated for the administration of their UI programs and Federal standards should be created to promote
the solvency of state UI trust funds. Equally significant, the Federal system should
be better prepared to provide far more adequate benefits in times of recession, major
disasters like Hurricane Katrina and terrorist events like the September 11th attacks, which produce widespread devastation and threaten the nations economy.
3. Model New Federal Policies on Innovative State Reforms
Over the past decade, many states have been at the forefront of new economic security reforms that could help shape bold new Federal policies.
Of special note, in response to the record rates of foreclosures, some states have
created home protection funds providing revolving loans that save homes from
foreclose and preserve the fabric of their communities. Others have created special
training funds created from an offset of their UI payroll tax, often designed to make
local and regional industries more globally competitive. One state has taken the
lead in creating broad health care coverage for jobless families. And perhaps most
significant, California has recently established the nations first program of paid
family and medical leave running along side the state unemployment insurance system.
11 Denise Fronig, Trade Adjustment Assistance: A Flawed Program (The Heritage Foundation: July 31 2001).
12 For more detail on these and other Federal proposals, see Emsellem, Innovative State Reforms Shape New National Economic Security Plan for the 21st Century (National Employment
Law Project: December 2006).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00084

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

81
Congress can play a critical role supporting innovative state reforms by creating
new financial incentives and providing pilot program funding to expand these and
other initiatives. The more the states are successful in creating and sustaining such
programs, the stronger the case that can be made in Congress that these innovative
state reforms should make their way into Federal law and policy.
PREPARED STATEMENT OF LILY L. BATCHELDER, ASSISTANT PROFESSOR
PUBLIC POLICY, NYU SCHOOL OF LAW1

OF

LAW

AND

HOUSEHOLD INCOME VOLATILITY AND TAX POLICY: HELPING MORE AND HURTING LESS

Good morning, Mr. Chairman, Vice Chair Maloney, Ranking Member Saxton, and
Members of the Committee. My name is Lily Batchelder and I am an assistant professor at NYU School of Law. Thank you for the opportunity to testify before you
today on potential tax policy responses to household income instability. My testimony makes three main points:
First, income volatility, especially when it involves income declines, imposes significant hardships on American families. It heightens stress about finances and may
increase household living expenses. These hardships are most pronounced for
middle- and low-income families, whose incomes tend to be more volatile, and who
tend to have less access to low-cost borrowing.
Second, the income tax system currently simultaneously helps and hurts families trying to cope with these burdens. It helps in that it softens annual income fluctuations on an after-tax basis by timing tax payments so that a larger share of a
familys income is due in taxes in its higher-income years, and smaller share in its
lower-income years. It hurts because over time it imposes higher average tax rates
on households with relatively volatile incomes than it does on others whose income
is the same but more stable.
Finally, I will discuss two potential reforms to make the tax system help more
and hurt less when a familys income fluctuates. The first is a limited form of income averaging. It would permit taxpayers to elect to carryback unused standard
deductions and personal and dependent exemptions for 1 year, and to average their
income over 2 years when calculating the Earned Income Tax Credit. The second
is a much broader proposal, which would involve converting the roughly $500 billion
per year that we spend on tax incentives into uniform refundable tax credits. These
reforms could be implemented on a revenue-neutral basis. Both would reduce the
penalties that the tax system currently imposes on families with volatile incomes,
and would provide relief from these penalties in the years when families need it
most-when their income has fallen.
I. BACKGROUND ON HOUSEHOLD INCOME VOLATILITY

Household income volatility is pervasive. The evidence to date suggests that on

average family income tends to vary by roughly 30 percent from its mean.2 While
further research is needed, there is also mounting evidence that household income
volatility has been increasing over the past several decades as a percentage of
household income.3 The source of this apparent rise in household income volatility
is unclear. It likely results in part from increases in labor market flexibility and
capital mobility that stem from legal changes and globalization. Both may have increased the variability of individual earnings. It also likely reflects the increasing
presence of women in the paid labor force. As a result, couples now face a greater
1 The views expressed in this testimony are those of the author alone and do not necessarily
represent those of NYU School of Law. Portions of this testimony draw upon joint work with
Fred Goldberg and Peter Orszag. My co-authors also should not be held responsible for the
views expressed in this testimony. I am grateful to David Kamin for excellent research assistance.
2 See, for example, Lily L. Batchelder, Taxing the Poor: Income Averaging Reconsidered, 40
Harvard Journal on Legislation 395, 446 (2003); Jeffrey Liebman, Should Taxes Be Based on
Lifetime Income: Vickrey Taxation Revised fig. 5 (July, 2002).
3 See Jacob S. Hacker, The Great Risk Shift: The Assault on American Jobs, Families, Health
Care, and Retirement and How You Can Fight Back 27, 20304 n. 39 (2006); Peter G. Gosselin,
The Poor Have More Things TodayIncluding Wild Income Swings, L.A. Times, Dec. 12, 2004,
at A1; Lily L. Batchelder, Taxing the Poor: Income Averaging Reconsidered, 40 Harvard Journal
on Legislation 395, 446 (2003). There has been more research on earnings volatility, which appears to have also risen. See, for example, Ann Huff Stevens, Changes in Earnings Instability
and Job Loss, 55 Industrial & Labor Relations Review 60, 60 (2001); Peter Gottschalk and Robert Moffitt, The Growth of Earnings Instability in the U.S. Labor Market, in 2 Brookings Papers
on Economic Activity 217 (1994).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00085

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

82
combined risk of job loss or wage declines, and no longer have a back-up potential
worker if the primary earner loses his or her job.4
Income volatility is a serious social policy concern because it imposes both psychological and economic costs on families, especially when it involves sudden income
declines. Economic insecurity can heighten anxiety and family conflict.5 It creates
incentives not to take on risky jobs or invest in goods, like higher education, that
may generate an uncertain but greater expected return for the individual and society.6 In addition, families with relatively volatile incomes likely incur additional expenses as a result of the unplanned changes in their standard of living. For instance, they may move more often or incur high-interest debt in order to keep up
with relatively fixed expenses, like mortgage payments and utility bills.
The economic costs associated with income fluctuations are largest for middle- and
low-income families, and those that are relatively disadvantaged. Unlike more
wealthy families, these families typically have little savings and few assets against
which they can borrow.7 Downward income shocks for these families are also more
likely to result in earnings reductions that persist over a long period of time and
are passed on to their children.8
Unfortunately, these families are also precisely the ones that face the widest
swings in their income. As illustrated in Figure 1, the annual income of a family
in the bottom quarter of the income distribution on average varies about 44 percent
from the familys average income over a 6-year period. By contrast, the comparable
figure for families in the top quarter of the income distribution is about 18 percent.9

4 Elizabeth Warren and Amelia Tyagi Warren, The Two-Income Trap: Why Middle-Class
Mothers and Fathers Are Going Broke (2003).
5 See, for example, Patricia Voydanoff, Economic Distress and Family Relations: A Review of
the Eighties, 52 Journal of Marriage and the Family 1099 (Nov. 1990).
6 See, for example, Kathryn L. Shaw, An Empirical Analysis of Risk Aversion and Income
Growth, 14 Journal of Labor Economics 626, 626, 64142 (1996); Orley Ashenfelter and Cecilia
Rouse, Schooling, Intelligence, and Income in America: Cracks in the Bell Curve (National Bureau of Economic Research Working Paper No. 6902, Jan. 1999).
7 Edward N. Wolff, Recent Trends in Wealth Ownership, 19831998 (Levy Economics Institute
Working Paper No. 300, Apr. 2000).
8 Philip Oreopolous, Marianne Page and Ann Huff Stevens, The Intergenerational Effects of
Worker Displacement 14 (National Bureau of Economic Research Working Paper No. 11587,
2005).
9 See also Peter G. Gosselin, The Poor Have More Things TodayIncluding Wild Income
Swings, L.A. Times, Dec. 12, 2004, at A1; Jeffrey Liebman, Should Taxes Be Based on Lifetime
Income: Vickrey Taxation Revised fig. 5 (July, 2002); Christopher D. Carroll and Andrew A.
Samwick, The Nature of Precautionary Wealth, 40 Journal of Monetary Economics 41, 47
(1997). While the Congressional Budget Office has found an association between greater individual earnings volatility and lower education levels, it has not found this relationship with respect to household income. Peter R. Orszag, Director, Congressional Budget Office, Economic
Volatility: Statement before the Committee on Ways & Means, January 31, 2007, at 7. This may
result from the fact that, unlike the other studies cited here, the CBO estimates are based on
data from the Survey of Income and Program Participation and look at volatility over a 2-year
period, not multiple years.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00086

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

83

These broad movements in many households income are not simply due to families making their way steadily up the economic ladder over time, or transitioning
from school to work, or from work to retirement. Figure 1 is based on households
in which the head was 44 to 55 years old, the point in life when earnings are typically the most stable.11 Moreover, the Congressional Budget Office recently estimated that between 2001 and 2002, 24 percent of households with a head aged 25
to 55 experienced a 25 percent increase in income, but 14 percent experienced a 25
percent decline.12
Sometimes these downward income shocks are planned or the family can insure
against them. For instance, a young family can save so that one parent can take
a couple of years off to care for a newborn, and workers can often purchase disability insurance through their employer. Frequently, however, large declines in
household income are unexpected and private insurance for the relevant risk-such
as unemployment or a wage reduction after a layoffis unavailable. In these situations, the case is strongest for the government stepping in to cushion the decline.
II. INCOME TAX EFFECTS OF HOUSEHOLD INCOME FLUCTUATIONS

Currently the income tax system simultaneously helps and hurts families facing
income fluctuations. It helps because the progressive nature of our income tax results in families paying relatively more tax in good years and relative less in bad
ones.13 It hurts because the fact that we levy taxes on annual income results in a
10 Unless otherwise noted, all estimates are taken from Lily L. Batchelder, Taxing the Poor:
Income Averaging Reconsidered, 40 Harvard Journal on Legislation 395 (2003). The estimates
of income volatility are based on Panel Survey of Income Dynamics (PSID) data from 1987 to
1992. They exclude income from transfers, capital gains and inheritances. The measure of income volatility is the coefficient of variation over that period, which roughly speaking is the percentage by which a households income varies from its mean. The estimates of the tax effects
that follow are based on PSID data from 1968 to 1992 for households with at least 10 years
of income data during which their filing status was unchanged. In both sets of estimates, income
groups are based on average income over the period.
11 Don Fullerton and Diane Lim Rogers, Who Bears the Lifetime Tax Burden? 117 tbl.42
(1993).
12 Peter R. Orszag, Director, Congressional Budget Office, Economic Volatility: Statement before the Committee on Ways & Means, January 31, 2007, at 9.
13 Thomas J. Kniesner and James P. Ziliak, Tax Reform and Automatic Stabilization, 92
American Economic Review 590, 590 (2002). A progressive tax system also helps smooth macroeconomic fluctuations. See, for example, Alan J. Auerbach & Daniel Feenberg, The Significance
of Federal Taxes as Automatic Stabilizers, Journal of Economic Perspectives 38, 48 (Summer
2000).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00087

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

84
family whose income is relatively volatile paying more tax over time than a family
whose income is more stable.
To illustrate these countervailing effects, suppose two families both earn an average of $80,000 each year. Family A earns $100,000 in the first year and $60,000
in the second. Family B earns the same amount in both years. Both are composed
of a married couple with two children. Table 1 shows that the rising marginal rates
of the income tax provide a form of insurance for the family with fluctuating income.
Family A owes a smaller share of its income in taxes in the tough year in exchange
for paying a larger share in the more comfortable one. The decline in its income is
therefore smaller on an after-tax basis (about $32,500) than it is on a pre-tax basis
($40,000).

Table 1.Example of Income Smoothing and Fluctuation Penalties (2006 Law)14

Average Tax Rate
Income

Tax Due
Per Year

Family A ..............................................................................
Family B ..............................................................................

Year
Year
Year
Year

1
2
1
2

.....
.....
.....
.....

100,000
60,000
80,000
80,000

10,240
2,720
5,720
5,720

10%
5%
7%
7%

After-Tax
Income

Over 2
Years

8%
............
7%
............

89,760
57,280
74,280
74,280

At the same time, though, Family A owes about $1,500 more in taxes over the
2 years than Family B. Its average tax rate over the 2-year period is 8 percent,
while Family Bs is 7 percent. This fluctuation penalty arises because we tax annual income, not average income. As a result, in its good year Family A is pushed
into higher tax brackets that would never apply if, like Family B, it earned the
same amount of income more evenly.
The penalties that the tax system imposes on families with fluctuating incomes
can be substantial. They are also more pronounced for middle- and low-income
households.15 Figure 3 provides rough estimates of the average increase in households tax rates under 2001 (post-EGTRRA) law as a result of paying tax on their
annual income instead of on their average income over the 10 to 25 years for which
I had data on individual households. It shows that fluctuation penalties are much
larger for families in the bottom quarter of the income distribution. This pattern is
a product of two factors: lower-income families experience wider income swings as
a proportion of their income and marginal tax rates rise more rapidly at the lower
end of the income distribution, especially because of the earned income tax credit
(EITC).

14 The table assumes that each family claims the standard deduction, personal and dependent
exemptions, and child tax credit, but no other tax benefits.
15 See also Jeffrey Liebman, Should Taxes Be Based on Lifetime Income: Vickrey Taxation Revised fig. 7 (July, 2002).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00088

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

85

At all income levels, fluctuation penalties are also much larger for the roughly 10
percent of families that experience the widest income fluctuations. As illustrated in
Figure 4, depending on their income level, the average tax rate on these households
is between about two and twelve percentage points higher as a result of not being
able to average their income over the period. Moreover, in the individual years when
these households experience the largest income shocks, the increase in their average
tax rate is much greater. These estimates are also based on data on the earnings
patterns of individual households from 1968 to 1992. To the extent that household
income volatility has increased, the percentage of families facing these large fluctuation penalties has likely grown.

If anything, the hardships created by income volatility suggest that we should impose smaller, not larger tax burdens on households with wide income fluctuations.
Instead we are currently doing just the opposite. Direct spending programs like un-

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00089

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

86
employment insurance, welfare, and food stamps may ameliorate these tax penalties
imposed on income fluctuations to some degree, but it is unlikely that they offset
them completely. All three programs are time limited. Food stamps and welfare are
restricted to very low-income households and only between 20 percent and 60 percent of households eligible for these benefits actually claim them.16 Meanwhile, unemployment insurance covers an increasingly small share of workers, and only provides benefits to workers who are unemployed, not to those who experience earnings
declines as a result of underemployment or reemployment at a lower wage.
Thus, absent large new transfer programs, the onus is on the tax system to soften
downward income shocks more and impose more equal burdens on households with
volatile incomes. Essentially the tax system needs to simultaneously become more
part of the solution, and less part of the problem.
III. REFORMS WORTH CONSIDERATION

Any tax reform to address household income volatility should further these twin
objectives of concentrating tax payments in higher-income years and reducing tax
penalties on income fluctuations. Fluctuation penalties can be seen as premium payments for the income insurance that the income tax effectively provides by smoothing after-tax income. Meeting these twin objectives will result in greater income insurance benefits and smaller premium payments at the same time.
The result will be a fairer and more efficient tax system. Reducing fluctuation
penalties will mean that taxpayers with the same income over time owe more equal
amounts of tax. It will also reduce the disincentives for risk-taking that these penalties create. Heightening the income smoothing effects of the tax system can further enhance efficiency by addressing the failure of private markets to offer income
insurance. If a tax reform can further these objectives while minimizing administrative and compliance costs and strengthening the fairness and efficiency of the tax
system in other ways, all the better.
Targeted Averaging
One promising approach for ameliorating the hardships associated with income
volatility is to implement a limited form of income averaging, which I refer to as
targeted averaging. This approach would have two components.
First, if a family has less income in a given year than the standard deduction and
personal and dependent exemptions to which they are entitled, the family could
elect to carry these excess personal deductions back to the prior year. What this
means in practice is that they could recalculate their tax liability for the previous
year as if they used these excess personal deductions in that prior year. The family
would then receive as a refund the difference between the amount of taxes they actually paid in the previous year, and the smaller amount due after the recalculation.
Second, under targeted averaging, families could elect to average their income
over 2 years for purposes of calculating the EITC. The EITC effectively provides an
earnings subsidy of up to about $4,500 for a low-income households with two or
more children and with income under $38,000.17 In doing so, it offsets work disincentives for such families that are created by transfer programs and the payroll
tax. In practice, it has successfully induced more work, especially among single
mothers.18 However, families with income in the range of the EITC frequently face
the largest fluctuation penalties. In part this occurs because the EITC creates rapidly rising marginal tax rates.19 It also occurs because, on average, the incomes of
lower-income families are the most volatile.
Targeted averaging would eliminate a significant portion of the penalties that the
tax system imposes on households with unstable incomes. Figure 5 shows the average tax rate increase for households over 10 to 25 years as a result of families not
being able to claim the benefits of the targeted averaging proposed here. When compared to Figure 4, it shows that targeted averaging would eliminate roughly a quar16 See, for example, Robert Moffitt, The TANF Program, in Means-Tested Transfer Programs
in the United States 291, 309 (Robert Moffitt, ed., 2003); Allen L. Schirm and Laura A. Castner,
Reaching Those in Need: State Food Stamp Participation Rates in 2000 (U.S. Dept. of Agriculture, Dec. 2002).
17 The maximum credit is about $2,700 for households with one child and about $400 for
households with no children.
18 Joseph Hotz and John Karl Scholz, The Earned Income Tax Credit, in Means-Tested Transfer Programs in the United States 141, 16984 (Robert Moffitt ed., 2003); Bruce Meyer and Dan
T. Rosenbaum, Welfare, the Earned Income Tax Credit, and the Labor Supply of Single Mothers, 116 Quarterly Journal of Economics 1063 (2001).
19 For families with two or more children, when it is phasing in, the implicit marginal tax
rate is up to 40 percent, and as it phases out at higher incomes, it imposes an implicit marginal
tax rate of up to 21 percent.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00090

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

87
ter of fluctuation penalties and, like more comprehensive averaging, would provide
the greatest benefits to low- and middle-income households.

In addition, targeted averaging would enhance the income smoothing benefits of

the tax system. Taxpayers would only benefit from carrying back personal deductions in years when their income had declined to the point that that they could not
use personal deductions. They wouldnt benefit in years when their income had
risen.20 Similarly, 2-year EITC averaging would provide benefits more frequently to
households that had experienced a downward income shock than to those whose economic circumstances had improved. All households with income up to $76,000 could
potentially benefit from EITC averaging. A small share of householdsthose earning less than the phase-out amount of about $16,000 in the first yearwould receive
benefits if their income rose as a result of working more or at a higher wage. But
all others would only benefit from the reform if and when their household income
had fallen.
More comprehensive averaging of all income or over a longer timeframe could reduce fluctuation penalties and heighten income smoothing still further.21 However,
it could significantly increase administrative and compliance costs as taxpayers and
the government would need to recalculate innumerable items on multiple prior returns. Moreover, the theoretical and practical benefits of averaging decline as the
averaging period is extended. Most income volatility appears to be short-lived, with
a large share disappearing after 1 year.22
Targeted averaging is a modest step that simply narrows the tax differential between families with stable and unstable incomes. In a relatively administrable way,
it would simultaneously put families with volatile incomes on more equal footing
20 Conversely, allowing taxpayers to carry forward personal deductions would provide benefits
only in years when their income had increased.
21 Another way to heighten income smoothing through the tax system would be to allow taxpayers to defer taxes owed with interest so that theoretically they could concentrate all of their
tax payments in their more prosperous years. Given the current high levels of household debt
and rising personal bankruptcy rates, this approach could create more problems than it solves.
See Daniel Shaviro, Beyond the Pro-Consumption Tax Consensus 14 (January, 2007).
A further possibility would be to deliver more extensive income insurance through the tax
code in a manner that results in families with unstable incomes paying less tax over time than
families with stable incomes, not more. For instance, if a family was eligible for a $2,000 increase in its refund due to targeted averaging, that amount could be doubled to $4,000.
22 See, for example, Jeffrey Liebman, Should Taxes Be Based on Lifetime Income: Vickrey
Taxation Revised fig. 7 (July, 2002).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00091

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

88
relative to those with smoother incomes, and provide a cushion in the years when
they need it most.
Converting Household Tax Incentives into Uniform Refundable Tax Credits
A second complementary but more far-reaching approach to address household income volatility would be to transform individual tax incentives into uniform refundable tax credits. Targeted averaging focuses on the income side of the tax ledger,
but the tax treatment of household expenses and investments is almost equally important.
Currently the individual income tax provides about $500 billion per year in tax
incentives intended to encourage people to spend or invest their money in ways that
are considered socially valuable, such as on homeownership, retirement savings,
charitable contributions, health insurance and education.23 Id like to pause and emphasize, as you are well aware, that $500 billion is a big number. It is close to 4
percent of GDP.24 It is about half of the revenue raised by the individual income
tax,25 and equals our total outlays for the Department of Defense last year.26

As illustrated in Figure 6, about 80 percent of these tax incentives are currently

structured as deductions, exclusions, exemptions, and non-refundable tax credits.
Whenever a tax incentive is structured in these ways it is upside-down. Deductions, exclusions, or exemptions (all of which I will refer to as deductions for simplicity) are worth more to higher-income households because the value of these
types of incentives is the amount deducted times the taxpayers marginal tax rate.
Indeed, itemized deductions are typically worth nothing for middle- and lower-income taxpayers because about two-thirds of taxpayers do not itemize.27 Similarly,
non-refundable credits are worth nothing to the roughly 40 percent of households
with no income tax liability.28
23 Staff of the Joint Committee on Taxation, 109th Congress, Estimates of Federal Tax Expenditures for Fiscal Years 20062010 (Comm. Print 2006). The estimated cost in 2006 was
$510 billion for tax expenditures for homeownership, charitable contributions, health insurance,
education, retirement savings, life insurance, annuities, state and local bonds, and work incentives (the EITC and child tax credit). This figure may be substantially higher or lower due to
interaction effects. In addition, some may disagree with the characterization of these tax benefits as designed to promote certain behavior, rather than, for example, being designed to measure ability to pay more accurately.
24 Bureau of Economic Analysis, Current-Dollar and Real Gross Domestic Product (Jan. 1,
2007), http://www.bea.gov/national/xls/gdplev.xls.
25 Office of Management and Budget, Analytic Perspectives: Budget of the U.S. Government,
fiscal year 2008, 239 (2007).
26 Office of Management and Budget, Budget of the U.S. Government, fiscal year 2008: Department of Defense (2007), available at http://www.whitehouse.gov/omb//budget/fy2008/defense.html.
27 Internal Revenue Service, 2003 Statistics on Income, Individual Income Tax Returns Publication 1304 (Complete Report) tbl. 1.2 (2003).
28 See Scott A. Hodge, Number of Americans Paying Zero Federal Income Tax Grows to 43.4
million, Fiscal Facts 54 (Tax Foundation, March 30, 2006); Peter R. Orszag & Matthew G. Hall,
Nonfilers and Filers with Modest Tax Liabilities, 100 Tax Notes 723 (2003).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00092

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

89
The upside-down incidence of these types of tax incentives can be seen in Figure
7, which summarizes estimates of the average value of the home mortgage interest
deduction and tax incentives for retirement savings among claimants.29 Since higher-income households are more likely to claim these benefits, the actual distribution
among the entire population is even more skewed that Figure 7 suggests.

Fred Goldberg, Peter Orszag and I have argued that, purely on efficiency grounds,
these tax incentives should generally be restructured into uniform refundable tax
credits on a revenue-neutral basis, if they are to continue.30 For example, instead
of providing a deduction for up to $10,000 spent on some good, we could match a
percentage of spending on that good, up to $10,000. In the context of retirement savings, Bill Gale, Jonathan Gruber and Peter Orszag have estimated that the revenueneutral match would be 28 percent.31 At current levels, allowing refundability would
be roughly akin to allowing averaging of expenses eligible for tax incentives because
the vast majority of households with zero tax liability or negative tax liability in
1 year have positive tax liability over time.32
In a nutshell, the efficiency argument for uniform refundable credits is that since
tax incentives are intended to encourage certain expenditures and investments generating social benefits, the default policy should be that all taxpayers are eligible
for the same subsidy. We should only deviate from a uniform subsidy if there is evi29 Calculations based on Staff of the Joint Committee on Taxation, 109th Congress, Estimates
of Federal Tax Expenditures for Fiscal Years 20062010 (Comm. Print 2006), and Leonard E.
Burman et al., Distributional Effects of Defined Contribution Plans and Individual Retirement
Accounts, in The Distributional Effects of Government Spending and Taxation 69 (Dimitri B.
Papadimitriou ed., 2006). The estimates for retirement savings incentives are only for new contributions to these vehicles in 2004.
30 Lily L. Batchelder, Fred T. Goldberg, Jr., and Peter R. Orszag, Efficiency and Tax Incentives: The Case for Refundable Tax Credits, 59 Stanford Law Review 23 (2006). Similar proposals have been made historically by Stanley Surrey and more recently by Bill Gale, Jonathan
Gruber, Laurence Seidman, and Jason Furman, among others. See Jason Furman, If Youre
Going to Do Social Policy Through the Tax Code, Do It Right (Center on Budget and Policy Priorities, Jan. 24, 2007); Laurence S. Seidman, Pouring Liberal Wine into Conservative Bottles
2027 (2006); William G. Gale, Jonathan Gruber, Peter R. Orszag, Improving Opportunities and
Incentives for Saving by Middle- and Low-Income Households (Hamilton Project Discussion
Paper 20062, April 2006); Stanley S. Surrey, Pathways to Tax Reform: The Concept of Tax Expenditures 98100 (1973).
31 William G. Gale, Jonathan Gruber, Peter R. Orszag, Improving Opportunities and Incentives for Saving by Middle- and Low-Income Households (Hamilton Project Discussion Paper
20062, April 2006).
32 Lily L. Batchelder, Fred T. Goldberg, Jr., and Peter R. Orszag, Efficiency and Tax Incentives: The Case for Refundable Tax Credits, 59 Stanford Law Review 23, 68 (2006).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00093

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

90
dence that certain groups of taxpayers are more responsive to the incentive, or that
their expenditures and investments generate more social benefits. In my view, this
conversion would not only improve economic efficiency, it would also be more equitable.
Most importantly for todays topic, though, this transformation would have very
positive effects on household income volatility. The estimates of fluctuation penalties
that I have discussed are all based on the assumption that taxpayers dont claim
tax incentives. But tax incentives structured as deductions and non-refundable credits create fluctuation penalties as well, and often quite sizable ones.33 For instance,
suppose that two families would both like to claim $30,000 in home mortgage interest deductions each year, and both earn the same income over time and are composed of a married couple with two children. Family A earns $150,000 in the first
year and $90,000 in the second. Family B earns $120,000 in both years. Table 2
shows that these tax incentives will be worth almost $2,000 less for Family A whose
income is unstable.

Table 2.Example of $30,000 Home Mortgage Interest Deduction Penalizing and Exacerbating
Income Fluctuation (2006 Law) 34
Value of
Deduction
Income

Family A .............................................................................................................
Family B ............................................................................................................

Year
Year
Year
Year

1
2
1
2

...........
...........
...........
...........

150,000
90,000
120,000
120,000

Per
Year

Over 2
Years

9,084
5,020
8,000
8,000

14,104
............
16,000
............

Whats worse, structuring tax incentives as deductions also results in taxpayers

reaping the largest tax benefits in their most prosperous years, not the years in
which they are most financially stressed. These types of tax incentives sharpen
rather than smooth income fluctuations.35 The value of Family As deduction falls
by about $3,000 in the year when its income plummets. Thus, in the context of tax
incentives, our tax system doesnt help and hurt families with volatile incomes. It
only hurts.
Transforming tax incentives into uniform refundable credits would reverse both
of these adverse effects on families with unstable incomes. The value of tax incentives would then not vary depending on degree of income volatility that a family
experiences. In addition, relative to current law, the benefits of tax incentives would
be much more concentrated in a households relatively lean years. Structuring tax
incentives as uniform refundable credits might increase administrative and compliance costs slightly as more taxpayers claim them. However, this seems to be a desirable result if it means that more households spend and invest their money in the
ways that tax incentives seek to encourage. Uniform tax incentives might also spur
more families to respond positively to the incentive because the value of the incentive would be clearer.
This second proposal is clearly an extremely ambitious one. But it is worth serious
consideration. The current structure for the lions share of our tax incentives is fundamentally flawed. Deductions, exclusions, exemptions and non-refundable credits
all penalize families with fluctuating incomes at their most vulnerable points. They
are inefficient and unfair. It is rare that tax reform can improve the tax code in
so many ways, and with so few substantive drawbacks.
IV. CONCLUSION

The tax system can play an important role in addressing the serious hardships
that sudden income declines create for American families. However, to date, the tax
system has been both a help and a hindrance for families experiencing wide income
33 This conclusion only holds if the marginal income tax rate schedule is concave, meaning
that marginal tax rates rise more quickly at the low end of the income distribution and more
slowly at the high end. While there are some exceptions, this assumption generally holds.
34 The table assumes that each family claims the personal and dependent exemptions and the
child tax credit, and has other itemized deductions equal to the standard deduction.
35 For an example, see Appendix 2 in Lily L. Batchelder, Fred T. Goldberg, Jr., and Peter R.
Orszag, Efficiency and Tax Incentives: The Case for Refundable Tax Credits, 59 Stanford Law
Review 23 (2006).

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00094

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

91
swings. It has taxed families more if they experience income fluctuations, but it has
imposed these taxes disproportionately in the years in which they are better able
to pay.
Fortunately, two concrete reformstargeted income averaging and transforming
tax incentives into uniform refundable tax creditsdo not exhibit this tradeoff. Both
would increase the tendency of the tax system to smooth income fluctuations. At the
same time, they would reduce the extra taxes that families with unstable incomes
currently pay. Together, they could be implemented on a revenue-neutral basis.
Moreover, they would improve the equity and efficiency of the tax system in other
ways.
As income fluctuations appear to be rising over time, the economic lives of American families, and especially those that are middle- or low-income, are increasingly
unstable and insecure. The time is ripe to make the tax system more of a cushion
and less of a disproportionate burden on these families that are already vulnerable.

PREPARED STATEMENT

OF

DR. BRADLEY R. SCHILLER, PROFESSOR, SCHOOL

AFFAIRS, AMERICAN UNIVERSITY

OF

PUBLIC

I come before you today to praise income instability not to bury it. The United
States would have a far less vibrant economy if incomes were stable. Wed look and
behave more like the French than like the entrepreneurs of Silicon Valley.
Members of Congress and academic faculties are fearful of income instability. We
both occupy positions with effective lifetime tenure. Members of Congress have a 96
percent re-election rate and so, implicit lifetime tenure. University professors have
explicit tenured positions. This gives us both a lot of license. Members of Congress
can take bribes, evade taxes, and even commit manslaughter without losing their
job, their income, or their pensions. University professors can do pretty much the
same thing. Thats one of the virtues of income and job stability.
Theres a downside to income stability too, though. Job security often dulls efficiency, innovation, and entrepreneurship. People in secure positionswith guaranteed incomes and benefitsdont have to be as responsive to their customers or
their constituents. They are also less likely to take chances on new ideas, new products, or new technologies. A stable income, after all, implies not only little risk
of income loss, but also little prospect for income gain. So why pursue a new idea
if theres no payoff? Just stay put, follow the established order and you can count
on job stability and income security.
Lets look outside the halls of Congress and the University to see how the rest of
America grapplesand prospers fromincome instability.
JOB FLOWS

Wal-Mart hires dozens of new workers every day. Maybe youre not a fan of the
Wal-Mart employment model. Well, then, how about Google? They hired over 2000
new workers last year alone. Genetech also hired 2000 workers last year. XM and
Sirius Satellite have taken on over 1000 workers in the last couple of years. The
healthcare industry as created 3.5 million new jobs in the last 10 years; schools and
colleges have added another 2 million jobs.
So who filled all these jobs? A couple of million workers enter the labor force every
year. But most of these labor-market entrants are teenagers and immigrants. They
might get some of those jobs at Wal-Mart, but they probably didnt fill many of
those jobs at Google, Genentech or XM Satellite. Those companies want employees
with experience, demonstrable skills, and employment references.
So where do growing firms and industries get the workers they need? For the
most part, from firms and industries that arent doing so well. Workers have lost
thousands of telephone company jobs in the last 10 years. The auto industry is now
shedding tens of thousands of workers. With the downturn in housing, a lot of realestate brokers and mortgage lenders are re-thinking their career choices.
Is all this job mobility good for the economy? Absolutely. Consumer tastes, production technologies, product innovation, and global competition are always changing.
To respond to those changes, weve got to be fast on our feet. Specifically, weve got
to be able to move capital and workers out of one set of industries and into another
set of industries. That resource mobility is a prerequisite for productivity advance
and output growth. Without such mobility, our incomes might be more stable, but
theyd also be lower.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00095

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

92
FEAR OF FALLING

I know youve worried about the individuals who are part of this processthe
workers who lose their jobs as a result of plant closings and job layoffs. Thats a
legitimate concern for public policy. But we should keep our eye on the big picture
even as we reach out to help displaced, dislocated, and otherwise unemployed workers. For the most part, the workers who move from one industry to another end up
better off in the long run. Holding onto a job in a declining industry isnt the path
to prosperity. Far better to get a toehold in an industry where jobs and wages are
growing. As we seek to provide a safety net for unemployed workers weve got to
be sure were not discouraging workersor their employersfrom grasping that toehold.
Remember the French riots of last Spring? French workers have always had
something akin to job tenure. Even new entrants into a firm are pretty much guaranteed a lifetime package of income growth, fringe benefits, and a generous pension.
What sparked the riots in Paris and its suburbs last year was a proposal for more
resource mobility. Specifically, the proposed law would have given French employers
the legal right to fire newly hired workers under age 26 for any reason within the
first 2 years of employment. French youth viewed this as a threat to their income
securityand took to the streets. A good many of them have stayed in the streets,
since French employers are reluctant to shoulder the upfront cost of hiring young
workers. Youth unemployment in France hovers around 24 percent, more than twice
U.S. levels. The French economy is growing half as fast as the U.S. economy, with
average incomes 25 percent below American levels. How many Americans would
trade American income prosperity for French income stability?
UPWARD MOBILITY

Income instability sounds pejorative. But we mustnt forget that instability includes both upward movement and downward movement. Winning a Powerball jackpot generates an enormous amount of income instabilityand an ocean of envy. The
high school dropout who advances from a minimum-wage job at McDonalds to a better job at UPS also experiences welcome income instability. So does the welfare
mom who becomes a sales clerk at Wal-Mart.
So the concern over income instability isnt really about instability per se, but
instead about the single dimension of income losses, i.e., downward instability. The
issue boils down to the adequacy and efficiency of the social safety net that is intended to cushion income falls.
TIME-LIMITED AID

For the most part, the U.S. social safety net is woven from time-limited income
transfer programs. Regular, unemployment insurance benefits are available for a
maximum of 26 weeks. TANF welfare benefits are available for a lifetime maximum
of 5 years. By putting time limits on such benefits we are implicitly recognizing the
importance of keeping people in the job market, where the best chances for upward
mobility reside. Providing wage insurance, unemployment benefits, trade adjustment assistance, or welfare for longer periods reduces incentives for seeking new opportunities in the labor market. Such extended benefits are an important explanation for the higher unemployment and lower average incomes in France and most
of Europe. Our shorter time limits and lower benefits strike a more dynamic balance
between equity (safety net features) and efficiency (economic incentives).
BUSINESS INCOME INSTABILITY

If were going to worry about income instability, we ought to look also at the dynamics of business instability. Over 50,000 new businesses are started each year in
the United States. These startups are the wellspring of some of our greatest innovations, new products, and technological advance. Most of these startups are little
more than the inspirations of a lone entrepreneur or the aspirations of an ambitious
household. A good many of these upstarts will fail, often with devastating financial
results for their owners and investors. Should we be extending profit insurance
to entrepreneurs? Probably not. Collectively, we seem comfortable with the notion
of business income instability. We even seem to regard that income instability as
a productive source of innovation and growth.
MIDDLE CLASS DYNAMICS

Much of the concern for income instability originates in perceptions of middleclass stagnation. The media ceaselessly depicts a disappearing middle class, the
result of a surge in inequality that leaves America a divided nation of rich and poor.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00096

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

93
The rich get richer while everyone else gets poorer is a popular mantra. That perception is not entirely consistent with the facts, however.
If you look only at median household incomes, its easy to see why people get the
wrong impression about the middle class. According to the Census Bureau, the inflation-adjusted median income for U.S. households was
$46,326 in 2005
37,599 in 2000
43,366 in 1990
39,739 in 1980
These numbers suggest that middle class incomes have fallen over the last 5
years and risen by only 0.50.6 percent annually over the last 1525 years.
There is no dispute about the Census statistics themselves. What is controversial
is what the numbers tell us about the typical household. Is the typical U.S. household just barely clinging to its middle class existence? Or are there other forces at
work here?
POPULATION DYNAMICS

One force that helps explain the income statistics is population growth. Just since
2000, the U.S. population has increased by over 18 million people. Nearly half of
that growth comes from immigrants, both legal and illegal. According to the U.S.
Labor Department, nearly half of the growth in the U.S. labor force has come from
foreign born workers, most of whom take low-wage jobs. What this means is that
the flow of new households is heavily concentrated in the lower end of the income
distribution. This bottom-heavy population growth puts a damper on the level of
median household income.
As a result of this bottom-heavy population growth a stagnant median income
need not imply stagnant or falling individual incomes. Think of the people lined up
for concert or baseball tickets. Individuals move up the line as tickets are purchased. But new people keep coming. So the line never gets shorter, even though
individuals are advancing.
Something similar happens with the distribution of income: People keep entering
the distribution line from the bottom. Even though individuals are moving up the
line, the middle of the line never seems to move. Hence, an unchangedor even recedingmedian marker could co-exist with individual advancement. The people who
were at the middle marker before have moved up the distribution line.
The same thing happens at colleges that open their doors wider. As enrollments
grow, the median SAT score may decline, even though no student is less accomplished than he or she was before. The same thing happens when a Harvard student
transfers to American University and the average SAT score rises at both schools.
The change in the median tells us nothing about changes in individual performance.
CHANGING HOUSEHOLD COMPOSITION

Another factor distorting our collective view of income dynamics is the changing
composition of American households. The Census Bureau defines a household as one
or more persons living under the same roof and sharing kitchen facilities. In 1980,
74 percent of all households were actually families of two or more persons. Today,
only 59 percent are families. Economic growth over the last 25 years has enabled
GenXers to move out of the family home and establish their own household. Rising
incomes and employment opportunities for women have also encouraged delayed
childbirth, fewer children, and single-parent households. Senior citizens too, have
used rising income and asset values to establish their own residences. These residence shifts depress median incomes. But those same shifts are a symptom of affluence, not of income deterioration.
These demographic changes suggest that even an actual decline in median or average household income need not signify lower living standards. When you look at
the big picturethe really big pictureit is apparent that living standards are rising. Just since 2000, real GDP has risen by 18 percent while the population has
grown by 6 percent. So per capita incomes have clearly been rising.
Some people would have you believe that all of this added income was funneled
to the rich. But the math doesnt work out. The increase in nominal GDP since 2000
amounts to nearly $4 trillion. If you assume that all that money went to the
wealthiest 10 percent of U.S. households, that bonanza would come to a whopping
$350,000 per household. Yet, according to the Census Bureau, the top 10 percent
of households has an average income of only $200,000 or so. Where is the extra
$350,000 they allegedly got? The implied bonanza is so absurd that the notion that
only the rich have gained from the economic growth can be dismissed out of hand!

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00097

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

94
Clearly, there is a lot of economic advancement across a broad swath of the population.
RISING CONSUMPTION

That broad swath of economic advancement shows up in personal consumption.

According to the Labor Department personal consumption spending has risen by
$2.5 trillion since 2000. More Americans own new cars and homes today than ever
before, despite modest slowdowns in both industries. Ipods, camera phones, and flatpanel TVs are fast becoming necessities rather than luxury items.
The point of all these observations is that the average American household is
doing pretty well. Certainly well enough to reject the notion of income stagnation
across the vast middle class and also well enough to appreciate the phenomenon of
upward income instability.
I dont mean to suggest here that everything is coming up roses for every American household. Inequality and income deprivation are still very real problems for
millions of American households. But its better to approach these problems from
a factual perspective than the hyperbole of middle-class stagnation.
POVERTY DYNAMICS

We should shed the same factual light on the hyperbole concerning Americas poverty population. Here again, the facts do not match popular perceptions. The notion
that the poor are getting poorer seems etched into the medias internal processor.
The foundation for that perception is Census data that reveal a shrinking income
share for low-income households. The bottom 20 percent of households got
4.2 percent of total income in 1980
3.8 percent in 1990
3.6 percent in 2000
3.4 percent in 2005
Evidence on the shrinking incoming share of the poor should not be confused with
receding income levels. Even if one accepts the Census data at face value, they do
not depict worsening deprivation. Although their percentage share of the pie may
be shrinking, the size of the slice received by the poor keeps getting larger. In 1980,
4.2 percent of Americas $5.16 trillion output (in constant dollars of 2000) amounted
to $217 billion. In 2005, the smaller 3.4 percent share amounted to $375 billion. So
the absolute size of the low-income slice grew by 73 percent. Over the same period,
the population of the lowest quintile grew by only 30 percent. Here again, the math
is compelling: living standards have risen substantially among low-income households, despite increases in income inequality. So we must reject the notion that the
poor are getting poorer.
INCOME MOBILITY

The economic situation among low-income households is not adequately conveyed

by this increase in statistical averages. A much more meaningful picture emerges
from the observed mobilityincome instability, if you willof individual families.
The same kind of population dynamics that affect measured median incomes also
impact poverty statistics. In fact, the impact may be greater. Think about the families that were counted as poor in 1980. Where are they now? Most of the elderly
poor from that year are now dead. The younger families of that year have changed
as well. The children have grown up and established their own households. The
teenage moms of 1980 are now middle-aged, with few if any children to care for.
Life goes on. In the process, the composition of the poor population changes.
Just because the same number of people are poor each year doesnt mean the
same people are poor every year. The poverty statistics are similar to emergency
room statistics. Every time you visit the emergency room youll see people bleeding.
But that doesnt mean the same people are bleeding continuously. People move in
and out of emergency rooms just as households move in and out of poverty.
Even over very short periods of time there is tremendous turnover in the poverty
population. Census data reveal that 1 out of 6 Americans will experience poverty
for at least two consecutive months over a 4-year period. But fewer than 1 out of
50 Americans will stay in poverty for as long as 4 years. Hence, persistent poverty
is by far the exception rather than the rule. Close to half the people in poverty in
a given year wont be poor in any of the following 3 years. Thankfully, most of the
patients bleeding in the emergency room dont come back.
If households are exiting from the poverty population with such frequency, how
come the poverty rolls dont shrink? Census data show that the official poverty population has been in the narrow range of 3237 million people for the past 25 years.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00098

Fmt 6601

Sfmt 6621

C:\JEC\34852.TXT

CELINA

PsN: CELINA

95
So the number of people entering poverty must roughly match the number exiting
from poverty each year. Where are they coming from?
Weve got a constant flow of immigrants, for starters. Well over a million immigrantsboth legal and illegalenter the country each year. Most come in at the
lowest rungs of the economic ladder, working for the minimum wage or less. The
household poverty rates among immigrants are twice as high as those of non-immigrants.
Then weve got 3 million or so low-achieving kids dropping out of high school each
year. And more than a million births a year to single moms, about a third of whom
are teenagers. On top of that, add more than a million divorces every year that
often devastate someones finances. Then there are the persistent scourges of death,
disability and illnessall of which throw families into poverty, often without warning. Finally, theres the economy, in which constantly shifting demands, costs and
technology create a continuous profusion of winners and losers. So theres always
a flow of new faces into the poverty ranks.
The reality of our poverty population is constant churn. Sure, this reflects a lot
of income instability. But the net change is positivethat is to say, there is net
movement out of poverty and up the income ladder. This has to be regarded as a
good thing. Moreover, unless we learn how to control all of lifes vicissitudesbirths,
illnesses, divorces, job layoffs, etc.such income instability is also inevitable.
MINIMUM WAGE WORKERS

Perhaps no group manifests the virtues of income instability better than minimum-wage workers. Most minimum-wage workers are young people taking their
first paid job. New immigrants also gravitate toward minimum-wage jobs. But neither group stays at minimum-wage jobs very long. Minimum-wage jobs have two salient characteristics. The first, and most obvious characteristic, is low wages. Wages
so low that they cant possibly support a family. But there a second characteristic
that is relevant hereturnover. Ask any fast-food manager or other low-wage employer what their greatest labor problem is and the answer is always the same:
turnover. Once minimum-wage workers accumulate some job experience (including
a resume and employer references), they move on to better jobs. Its the emergencyroom phenomenon again. We may have a constant stock of minimum-wage jobs, but
a stream of different workers keeps flowing through them.
Research shows how brief most minimum-wage experiences are. One out of three
minimum-wage entrants moves entirely into higher-wage strata within the first
year. Sixty percent surpass minimum-wage thresholds within 2 years. Only 1 out
of 6 minimum-wage entrants still have any minimum-wage experience after 3 years.
Here again, upward mobility is pervasiveand welcome.
POLICY IMPLICATIONS

These observations about the middle class, the poor, and minimum-wage workers
all have a common themenamely, that income instability is a common phenomenon and that it might not be as devastating as presumed. For the most part,
the economic deprivation that can result from income instability tends to be a relatively brief experience. Moreover, the patchwork of safety-net programs now in
place appear appropriately targeted to those time-limited problems. No, we havent
solved all our poverty and inequality problems. But before anyone jumps on the income instability bandwagon, we should exercise some caution. In particular, we
should ask whether any new policy responses to income instability might impose unintended costs. Of special concern are programs or policies that raise hiring costs
for employers or reduce work incentives for workers. Either phenomenon may increase income stability but reduce income levels.

VerDate 11-MAY-2000

11:56 Aug 29, 2007

Jkt 034852

PO 00000

Frm 00099

Fmt 6601

Sfmt 6611

C:\JEC\34852.TXT

CELINA

PsN: CELINA