[House Hearing, 110 Congress]
[From the U.S. Government Publishing Office]
MEASURING POVERTY IN AMERICA
=======================================================================
HEARING
before the
SUBCOMMITTEE ON
INCOME SECURITY AND FAMILY SUPPORT
of the
COMMITTEE ON WAYS AND MEANS
U.S. HOUSE OF REPRESENTATIVES
ONE HUNDRED TENTH CONGRESS
FIRST SESSION
__________
AUGUST 1, 2007
__________
Serial No. 110-56
__________
Printed for the use of the Committee on Ways and Means
U.S. GOVERNMENT PRINTING OFFICE
43-757 PDF WASHINGTON DC: 2008
---------------------------------------------------------------------
For Sale by the Superintendent of Documents, U.S. Government Printing Office
Internet: bookstore.gpo.gov Phone: toll free (866) 512-1800; (202) 512�091800
Fax: (202) 512�092104 Mail: Stop IDCC, Washington, DC 20402�090001
COMMITTEE ON WAYS AND MEANS
CHARLES B. RANGEL, New York, Chairman
FORTNEY PETE STARK, California JIM MCCRERY, Louisiana
SANDER M. LEVIN, Michigan WALLY HERGER, California
JIM MCDERMOTT, Washington DAVE CAMP, Michigan
JOHN LEWIS, Georgia JIM RAMSTAD, Minnesota
RICHARD E. NEAL, Massachusetts SAM JOHNSON, Texas
MICHAEL R. MCNULTY, New York PHIL ENGLISH, Pennsylvania
JOHN S. TANNER, Tennessee JERRY WELLER, Illinois
XAVIER BECERRA, California KENNY C. HULSHOF, Missouri
LLOYD DOGGETT, Texas RON LEWIS, Kentucky
EARL POMEROY, North Dakota KEVIN BRADY, Texas
STEPHANIE TUBBS JONES, Ohio THOMAS M. REYNOLDS, New York
MIKE THOMPSON, California PAUL RYAN, Wisconsin
JOHN B. LARSON, Connecticut ERIC CANTOR, Virginia
RAHM EMANUEL, Illinois JOHN LINDER, Georgia
EARL BLUMENAUER, Oregon DEVIN NUNES, California
RON KIND, Wisconsin PAT TIBERI, Ohio
BILL PASCRELL JR., New Jersey JON PORTER, Nevada
SHELLEY BERKLEY, Nevada
JOSEPH CROWLEY, New York
CHRIS VAN HOLLEN, Maryland
KENDRICK MEEK, Florida
ALLYSON Y. SCHWARTZ, Pennsylvania
ARTUR DAVIS, Alabama
Janice Mays, Chief Counsel and Staff Director
Brett Loper, Minority Staff Director
______
SUBCOMMITTEE ON INCOME SECURITY AND FAMILY SUPPORT
JIM MCDERMOTT, Washington, Chairman
FORTNEY PETE STARK, California JERRY WELLER, Illinois
ARTUR DAVIS, Alabama WALLY HERGER, California
JOHN LEWIS, Georgia DAVE CAMP, Michigan
MICHAEL R. MCNULTY, New York JON PORTER, Nevada
SHELLEY BERKLEY, Nevada PHIL ENGLISH, Pennsylvania
CHRIS VAN HOLLEN, Maryland
KENDRICK MEEK, Florida
Pursuant to clause 2(e)(4) of Rule XI of the Rules of the House, public
hearing records of the Committee on Ways and Means are also published
in electronic form. The printed hearing record remains the official
version. Because electronic submissions are used to prepare both
printed and electronic versions of the hearing record, the process of
converting between various electronic formats may introduce
unintentional errors or omissions. Such occurrences are inherent in the
current publication process and should diminish as the process is
further refined.
C O N T E N T S
__________
Page
Advisory of July 25, 2007, announcing the hearing................ 2
WITNESSES
Patricia Ruggles, Ph.D., National Academy of Sciences............ 6
John Iceland, Ph.D., Associate Professor, Sociology Department,
University of Maryland, College Park........................... 17
Nancy Cauthen, Ph.D., Deputy Director, National Center for
Children in Poverty, Mailman School of Public Health, Columbia
University..................................................... 28
Douglas J. Besharov, Joseph J. and Violet Jacobs Scholar in
Social Welfare Studies, American Enterprise Institute for
Public Policy Research......................................... 36
Mark Greenberg, Executive Director, Task Force on Poverty, Center
for American Progress.......................................... 54
SUBMISSIONS FOR THE RECORD
Richard Alarcon, statement....................................... 72
Wider Opportunities for Women, statement......................... 79
MEASURING POVERTY IN AMERICA
----------
WEDNESDAY, AUGUST 1, 2007
U.S. House of Representatives,
Committee on Ways and Means,
Subcommittee on Income Security and Family Support,
Washington, DC.
The Subcommittee met, pursuant to notice, at 10:00 a.m., in
room B-318, Rayburn House Office Building, Hon. Jim McDermott
(Chairman of the Subcommittee), presiding.
[The advisory announcing the hearing follows:]
ADVISORY
FROM THE COMMITTEE ON WAYS AND MEANS
SUBCOMMITTEE ON
INCOME SECURITY AND FAMILY SUPPORT
CONTACT: (202) 225-1025
FOR IMMEDIATE RELEASE
July 25, 2007
ISFS-11
McDermott Announces Hearing on
Measuring Poverty in America
Congressman Jim McDermott (D-WA), Chairman of the Subcommittee on
Income Security and Family Support of the Committee on Ways and Means,
today announced that the Subcommittee will hold a hearing on the
definitions and standards used to measure the number of Americans
living in poverty. The hearing will take place on Wednesday, August 1,
2007, at 10:00 a.m. in B-318, Rayburn House Office Building.
In view of the limited time available to hear witnesses, oral
testimony at this hearing will be from invited witnesses only. However,
any individual or organization not scheduled for an oral appearance may
submit a written statement for consideration by the Subcommittee and
for inclusion in the record of the hearing.
BACKGROUND:
On August 28th, the Census Bureau will release statistics on the
depth and breadth of poverty in 2006. In 2005, nearly 37 million
Americans were officially poor--an increase of 5.4 million since 2000.
In 2005, the poverty threshold for a family of four with two children
was $20,444.
The official poverty rate is a critical indicator of how widely
shared prosperity is in the economy, a key benchmark for targeting
resources towards the most disadvantaged, and a useful measure of the
impact of programs and policies on vulnerable populations.
However, there is a broad consensus that the poverty measurement
has become less accurate in highlighting economic hardship than when it
was created more than 40 years ago. For example, the poverty thresholds
were created in relation to consumption when the average family of
three or more persons spent about one-third of its after-tax income on
food. Today, food demands only one-seventh of that family's budget,
while the share of income devoted to other expenses, such as housing
and health care, has grown. Furthermore, the Federal poverty threshold
for a family of four represented about 50 percent of median income when
first devised, while it now represents only about 30 percent of median
income. Finally, the current poverty measurement fails to count certain
benefits, including the Earned Income Tax Credit and food stamps, as
well as certain work-related expenses, including child care and
transportation.
There have been numerous suggestions for revising the poverty
measure, including recommendations published in 1995 by a National
Academy of Sciences Panel on Poverty and Family Assistance. The panel
recommended both changes in how family resources are calculated and
corresponding adjustments to the poverty threshold. A number of options
were provided, nearly all of which would have increased the number of
Americans considered poor.
In announcing the hearing, Chairman McDermott stated: ``We need a
poverty measurement for 2007, not 1963. Unfortunately, our poverty
measure has not kept pace with societal changes over the past half
century. Improvements are needed so we can fully understand how many
Americans are denied access to a reasonable standard of living, and so
we can target resources to those most in need.''
FOCUS OF THE HEARING:
The hearing will consider the current poverty measure, its
limitations, and possible alternatives.
DETAILS FOR SUBMISSION OF WRITTEN COMMENTS:
Please Note: Any person(s) and/or organization(s) wishing to submit
for the hearing record must follow the appropriate link on the hearing
page of the Committee website and complete the informational forms.
From the Committee homepage, http://waysandmeans.house.gov, select
``110th Congress'' from the menu entitled, ``Hearing Archives'' (http:/
/waysandmeans.house.gov/Hearings.asp?congress=18). Select the hearing
for which you would like to submit, and click on the link entitled,
``Click here to provide a submission for the record.'' Once you have
followed the online instructions, completing all informational forms
and clicking ``submit'' on the final page, an email will be sent to the
address which you supply confirming your interest in providing a
submission for the record. You MUST REPLY to the email and ATTACH your
submission as a Word or WordPerfect document, in compliance with the
formatting requirements listed below, by close of business August 15,
2007. Finally, please note that due to the change in House mail policy,
the U.S. Capitol Police will refuse sealed-package deliveries to all
House Office Buildings. For questions, or if you encounter technical
problems, please call (202) 225-1721.
FORMATTING REQUIREMENTS:
The Committee relies on electronic submissions for printing the
official hearing record. As always, submissions will be included in the
record according to the discretion of the Committee. The Committee will
not alter the content of your submission, but we reserve the right to
format it according to our guidelines. Any submission provided to the
Committee by a witness, any supplementary materials submitted for the
printed record, and any written comments in response to a request for
written comments must conform to the guidelines listed below. Any
submission or supplementary item not in compliance with these
guidelines will not be printed, but will be maintained in the Committee
files for review and use by the Committee.
1. All submissions and supplementary materials must be provided in
Word or WordPerfect format and MUST NOT exceed a total of 10 pages,
including attachments. Witnesses and submitters are advised that the
Committee relies on electronic submissions for printing the official
hearing record.
2. Copies of whole documents submitted as exhibit material will not
be accepted for printing. Instead, exhibit material should be
referenced and quoted or paraphrased. All exhibit material not meeting
these specifications will be maintained in the Committee files for
review and use by the Committee.
3. All submissions must include a list of all clients, persons,
and/or organizations on whose behalf the witness appears. A
supplemental sheet must accompany each submission listing the name,
company, address, telephone and fax numbers of each witness.
Note: All Committee advisories and news releases are available on
the World Wide Web at http://waysandmeans.house.gov.
The Committee seeks to make its facilities accessible to persons
with disabilities. If you are in need of special accommodations, please
call 202-225-1721 or 202-226-3411 TTD/TTY in advance of the event (four
business days notice is requested). Questions with regard to special
accommodation needs in general (including availability of Committee
materials in alternative formats) may be directed to the Committee as
noted above.
Chairman MCDERMOTT. Welcome to the Committee. Last night I
went home at 3:30 in the morning from the Rules Committee
having been talking about SCHIP, and as I got out of my car, I
noticed that the guy next door to me is selling his house. So,
at that hour, you might as well go ahead and find out how much
he's selling it for. I live around the Marine Corps barracks
down in Southeast, and one of those row houses is now selling
for $745,000. I said to myself, I guess this is a timely
hearing, because when you talk about poverty that was once
based on the food basket, housing is now such a question that
it really brings it to my mind. It's part of the reason why I
scheduled this hearing.
At the end of this month the Census Bureau will release
statistics on the number of Americans living in poverty last
year, and I know everyone in the room is hoping that we'll
finally see a decline in poverty, especially after the record
in recent years.
Since 2000, the number of poor has climbed by 5.4 million.
That's an average of almost 3,000 Americans every single day
for 5 years. But even as we wait to see the new Census poverty
figures, we know that will tell really only part of the story
about the depth and the breadth of poverty in the United
States. There's a broad recognition that our poverty measure
has failed to keep pace with changing patterns of consumption,
employment, standards of living and government assistance over
the last 40 years.
Our current poverty definition was developed in 1963, so I
think it's reasonable to think about an update. That's before
the Beatles' first trip to the United States before the
establishment of Medicare, before Martin Luther King, Jr., made
his famous march from Selma to Montgomery to press for civil
liberties. A lot has happened in the last four decades in
America, except in defining poverty.
To set our poverty thresholds, we still rely on the early
1960 cost of providing a minimally adequate food plan to a
family, based on information that families spend about one-
third of their aftertax income on food in the mid 1950s. We
multiplied that amount by three and there's our number. Other
than updating for inflation, we're applying a 1963 standard
based on 1950 data in 2007. It really doesn't make sense to me,
and that is why we are here today.
During this period, the proportion of family income
dedicated to food has significantly declined while the
proportion needed for other necessities, such as housing and
medical care, has grown considerably. Mothers have become an
integral part of the workforce and therefore childcare has
become a major family expense. Food stamps and the Earned
Income Tax Credit were established to provide direct assistance
to low-income households. Of course, standards of living in
America have generally increased over the last four decades.
Our poverty measurement has largely ignored all of these
developments. An accurate reflection of poverty is important
for several reasons.
A valid measure illustrates how many Americans are failing
to meet the basic standard of economic need. It also indicates
the impact that policies and programs have on reducing that
level of need. Finally, it is critically important in the
targeting of resources to individuals, to communities and to
States. In fact, over 80 Federal programs reference some
percentage of the Federal poverty program in determining
individual eligibility or funding allocations for States.
To comprehensively update our poverty measure, we need to
consider what constitutes a minimal level of economic well-
being and what income, benefits and costs should count toward
and against that threshold.
There have been many past recommendations for such a
revision, including the National Academy of Sciences'. The vast
majority of these proposals show more Americans living in
poverty than under the official measure. In contrast, however,
a few have suggested that we look at only half of the equation
adding additional sources of income, such as earned income tax
credit. Taken by itself, such a step would continue to ignore
changes in consumption patterns and in general standards of
living.
It means the basic needs of the poor are permanently fixed
in time. At various times in our history, indoor plumbing,
central heating and refrigerators were not considered a basic
need. Does anyone here want to suggest that they are not a
basic need today?
Now, I look forward to hearing from our witnesses today
about alternative ways to measure poverty to ensure that we are
accurately counting every American who has trouble meeting
those crucial needs.
I now want to yield to my ranking member, Mr. Weller.
Mr. WELLER. Good morning. Thank you, Mr. Chairman. I
welcome the witnesses before our Committee and thank you for
your time in joining us today.
Today's hearing involves how our country measures poverty.
This follows prior hearings in this Subcommittee and the full
Committee in which we explored the cost of poverty; how U.S.
poverty measurement differs from other countries and possible
solutions to poverty. As several members on this side of the
aisle said in prior hearings, one of the first failings of our
current poverty measure is the fact it does not count tens of
billions of dollars in taxpayer funded assistance provided to
reduce poverty for literally millions of families.
This omission limits the usefulness of today's poverty
measure. It also devalues the sacrifices of taxpayers who pay
for those benefits with their hard-earned tax dollars and
increases the apparent number of families in poverty. Several
witnesses today suggest counting the value of more antipoverty
benefits to determine whether families are poor or not.
Major assistance not counted today includes food stamps,
public housing, the earned income tax credit and healthcare
coverage. These also constitute the fastest growing portions of
our nations' effort to help low income families escape poverty.
So, unless we act, more and more of our effort to alleviate
poverty will be ignored each passing year. Consider what this
means for families.
Let's say the Jones family of four has an annual income of
$30,000, all from wages. Current rules count wages as income
for purposes of judging whether a family is poor. Since the
poverty threshold for a family of four is about $20,000, and
the income of the Jones family is above that level, the Jones
family is officially not poor. Now, let's say they have
neighbors. The Smith family is also a family of four. The Smith
family also has a total of $30,000 in annual income, but the
Smith's income comes from multiple sources: $18,000 from wages;
plus a total of $12,000 in housing, healthcare, food stamps,
and earned income tax credit benefits provided by taxpayers.
Under current rules, none of the $12,000 in taxpayer
benefits provided the Smith family is counted as income. So,
since their $18,000 in wages fall short of the $20,000 poverty
threshold for a family of four, the Smith family is officially
poor. This makes little sense.
That's why I introduced a bill yesterday that would provide
more clarity about both the extent of poverty in the United
States and the effectiveness of current antipoverty programs.
My legislation, H.R. 3243, would direct the Census Bureau to
report on poverty as measured three ways. First, Census would
retitle the current official poverty rate as the partial
benefits poverty rate, which is what it is. The second measure,
called a full benefits poverty rate, would count means tested
food, housing and healthcare benefits as income.
The final measure, called the full benefits and taxes
poverty rate would also add in the value of tax credits, like
the earned income tax credit and subtract taxes paid. This
legislation will help us better understand both who is poor and
the effectiveness of current antipoverty benefits. While this
hearing today will help us understand how poverty is measured,
that's not enough. We also need to press on with what works to
reduce poverty.
As we saw in the progress against poverty following the
1996 Welfare Reform law, that starts with promoting more full-
time work instead of welfare dependence and it also means
promoting more healthy marriages, which also reduce poverty and
welfare dependence for the long run.
Mr. Chairman, I look forward to hearing from the witnesses;
and, I again thank them for their participation today.
Chairman MCDERMOTT. Thank you, we welcome all of you, and
as my colleague has said we thank you for coming here and
spending your time trying to educate us about how to make
better public policy.
We will start today with Dr. Ruggles, who is at the
National Academy of Sciences.
Dr. Ruggles.
Dr. RUGGLES. Thank you.
Chairman MCDERMOTT. If you will press that button so you
are on and your full statements will all be in the record. So,
we would like you to try and summarize and stay within 5
minutes, if you can. Somewhere we have a light I guess. It's
over there.
Dr. RUGGLES. All right, thank you.
Chairman MCDERMOTT. Please
STATEMENT OF PATRICIA RUGGLES, NATIONAL RESEARCH, COUNCIL OF
THE NATIONAL ACADEMY OF SCIENCES
Dr. RUGGLES. Well, I am happy to be here today, Mr.
Chairman and Mr. Weller, and I have to say, Mr. Chairman,
you've already managed to say in your opening statement many of
the things that I have planned to say here today.
Chairman MCDERMOTT. That's why I talk first.
Dr. RUGGLES. In my testimony today I plan to first review
the existing measure and its limitations, and then discuss some
alternatives for revising the measure, including the NAS
recommendations. Then finally, I want to consider the
implications of that kind of a revision for public policy
programs.
As you noted, it's been more than 40 years since our
official poverty measure was first designed and most experts
believe that it is now seriously out-of-date. The measure grew
out of a series of studies undertaken by Mollie Orshansky for
the Social Security Administration in the early 1960s.
Orshansky started with a series of food budgets put together by
the Department of Agriculture. These food budgets gave her the
amount of money families of different sizes and types would
need to spend annually to assure a minimally adequate diet.
Then using data from a 1955 survey, she calculated that average
families spend about one-third of their total incomes on food.
Lacking any other data, she just multiplied her food budgets by
three to get approximate levels of basic needs.
Although Orshansky's estimates of basic needs, which she
termed ``poverty thresholds,'' were necessarily very
approximate, they were a considerable advance over earlier work
because at least they recognized that needs varied by family
size and type. In 1969 a slightly modified version of the
Orshansky scale became the official poverty level for the
United States statistical establishment as a whole. Since 1969
the measure has been subjected to considerable criticism, but
it still forms the basis for our official poverty measure.
Much of the discussion about the Orshansky measure focuses
on the specific poverty thresholds that it calculated, but
Orshansky made another choice that's also been very important
for poverty measurement over time. That is she chose to look at
only pre-tax cash income. The official poverty measure actually
consists of two parts. The set of thresholds that she
calculated and the measure of family resources that is compared
to the threshold. In order to decide if a particular family is
poor, its income or consumption level is compared to the
threshold and a level of poverty is determined based on whether
the family's resources fall above or below that threshold, just
as in the example that Mr. Weller gave.
When Orshansky put together her measure, pre-tax, cash
income was all she could measure. That was all we had survey
data on. But we now collect data on a whole lot of other
things, like, for example, the non-cash benefits that families
get, but also things like the work expenses that they need to
pay in order to go to work, for example, as Mr. McDermott
mentioned, many families today include working mothers. Many of
those mothers must pay for childcare in order to earn the money
that they bring home. It doesn't make sense in thinking about
how much money they have available to pay for rent and to buy
food, not to include those childcare expenses as something that
they must pay.
So, it's not just on the addition side that we need to
worry about correctly measuring income. We also need to worry
about things that get subtracted from income. Either way, in
thinking about income, it's important to keep in mind that in
comparing income to thresholds, you can't just change one side
of the equation. If you're going to update the income measure
to take into the account the different kinds of income that
people get today and the joint expenses they have, you also
need to think about adjusting the thresholds.
Most economists believe, as Adam Smith put it as long ago
as 1776, that people can be considered poor if they are unable
to afford the things that, as he put it, ``the custom of the
country renders it indecent for creditable people, even of the
lowest order, to be without.'' In other words, poverty is not
just the inability to afford a subsistence diet. It varies from
time to time and place to place. The implications of this
belief for poverty measurement can be envisioned by imagining a
poverty line set in 1900.
At that time, basic needs would not have included indoor
plumbing, central heating, electricity, or many other needs
that we now take for granted, because the custom of the country
did not include those things, even for most of those who were
well off, let alone of the lowest order. Today, however,
Americans who cannot afford those things would generally be
considered poor. Indeed, a poverty threshold that did not
include those things would be fairly meaningless, since it
would be very difficult for the urban poor in particular to
find housing without them. Failing to include them in budgets,
therefore, would restrict the poor to a standard of housing
that no longer exists in this country and would not be
acceptable to most.
An approach to measuring poverty that takes into account
changes in standard of living over time is often referred to as
a relative poverty measure, as opposed to the absolute poverty
measure embodied in our official thresholds. Some analysts
argue that an absolute measure, one that sets a poverty line at
a fixed level at a particular point in time and then
subsequently adjusts only for changes in prices, is needed to
assess programs because a measure adjusted for changes in
standards of living presents too much of a moving target.
Such analysts assert that if the poor become better off in
material terms because society as a whole is becoming better
of, that's important evidence of public policy success and
should be taken into account in measuring the need for
continued assistance. Most countries do use some kind of
relative poverty measure, but I think there is a general
consensus that that is probably not what we most would like to
do in the United States.
[GRAPHIC] [TIFF OMITTED] T3757A.001
There have been a number of reassessments of the
measurement of poverty in the United States over time and this
chart compares some of them. The black line that goes right
across the middle is our current, official poverty threshold,
which since it is adjusted only for prices when you look at it
in fixed dollars, it is just a straight line.
The other measures that you see here are a relative
threshold; one set at 50 percent of the median income, a
measure based on Gallup Poll results that reports what people
think is the amount that is minimally necessary for people not
to be poor; and then finally that little green line at the end
are the thresholds developed by the National Academy of
Sciences. As you can see, the median family income line and the
subjective line started out below the official threshold, cross
over it just about at the time Mollie Orshansky set her
thresholds and have continued to rise over time.
We do not have the data to calculate the NAS measure back
nearly as far as those other two measures, but you can see on
this chart that it falls somewhere between the official measure
and a purely relative measure.
As I have mentioned, the most serious and comprehensive
reassessment of the U.S. poverty measure to date was undertaken
by the National Research Council of the National Academies of
Sciences in the early 1990s in response to a congressional
mandate. This scientific study by an independent panel of
experts led to the report, ``Measuring Poverty: a New
Approach.'' In setting poverty thresholds, the Academy
recommended an intermediate approach that was neither strictly
absolute nor entirely relevant. Under the NAS approach, a
market basket of basic goods, including food, clothing, shelter
and a small allowance for other needs would serve as the basis
for a new set of poverty thresholds.
Unlike our official thresholds, these thresholds would be
adjusted for differences in cost of living in different parts
of the country. The method for adjusting for differences in
family sizes would be updated to reflect differences in family
needs. Finally, the thresholds would be updated periodically
based on consumption patterns as consumption patterns changed
and ideas about what constituted basic needs changed, the
thresholds would change with them. This type of approach is
more difficult to implement than the present system, because it
requires some judgment calls about what people need, but it has
the advantage of tracking what most people mean by poverty much
more closely than either the official measure or a purely
relative measure. This type of updating is not unprecedented.
Similarly, periodic revisions are used to update the market
basket underlying the consumer price index, for example. A
measure of this type will generally increase more over time
than a strictly absolute measure. Because the standards of
living rise, people generally raise their expectations for even
minimal consumption levels. But the increases will be linked to
changes in consumption levels and costs rather than to changes
in income.
Wrap it up. Okay, well just quickly before we do, the chart
that should be up here and isn't--if you could make that go up.
Chairman MCDERMOTT. Do we put in the cost of technology?
Dr. RUGGLES. There we go. Yes, really.
[GRAPHIC] [TIFF OMITTED] T3757A.002
Dr. RUGGLES. This chart summarizes the differences between
the NAS recommendations and both the official measure and a
purely relative measure. The main differences are the
thresholds, which are based on current consumption rather than
either income or an old consumption survey. It includes some
non-cash income. It deducts out-of-pocket medical. It deducts
work expenses. It deducts taxes. The family size adjustment is
based on something other than the mix of family sizes and types
that was around in 1969.
[The prepared statement of Dr. Ruggles follows:]
Prepared Statement of Patricia Ruggles, Ph.D., National Academy of
Sciences
* Views expressed in this testimony are solely those of the author
and should not be taken to reflect those of her current employer, the
National Research Council of the National Academies.
Mr. Chairman and members of the Committee, I am happy to be here
today to discuss the measurement of poverty. Poverty measurement is
important for public policy, because many of our programs are designed
to help the neediest among us. A good measure of the extent of that
need is essential in designing programs that will alleviate poverty and
that will help to ensure equality of opportunity for all Americans.
Unlike most other countries, the United States has an official
poverty measure, which has been widely used in designing and assessing
programs to combat poverty. That measure is a legacy of the War on
Poverty, and over time it has been very helpful in identifying those
who need help, and in understanding what does and doesn't work in
providing that help. It has been almost 50 years since that measure was
first designed, however, and most experts believe that it is now
seriously out of date.
In my testimony today I will discuss three topics relating to the
poverty measure. First, I will review the existing measure and will
outline its limitations. Second, I will discuss some alternatives for
revising that measure. And finally, I will consider the potential
implications of such a revision for public assistance programs.
The Official Poverty Measure
The current official poverty measure grew out of a series of
studies undertaken by Mollie Orshansky for the Social Security
Administration in the early and mid-1960s. Orshansky did not have the
wealth of studies on needs and consumption that we have today available
to her at that time, so she used the best proxies that she could find.
She started with a series of food budgets put together by the
Department of Agriculture. These were much like today's Thrifty Food
Plan (TFP), the food needs index used to set food stamp levels. These
food budgets gave her the amount of money families of different sizes
and types would need to spend to assure a minimally adequate diet.
Then, using data from a 1955 survey, Orshansky calculated that on
average families spent about one-third of their total incomes on food.
Lacking any other data, she simply multiplied her food budgets by three
to get approximate levels of basic needs for all goods and services.
Although Orshansky's estimates of basic needs--termed poverty
thresholds--were necessarily very approximate, they were a considerable
advance over earlier work in that they at least recognized that needs
varied by family size and other factors. Once Orshansky's scale had
been published, it was widely adopted by other researchers. Finally, in
1969 a slightly modified version of the Orshansky scale was mandated by
the Bureau of the Budget (predecessor of today's OMB) as the official
poverty measure for the government statistical establishment as a
whole. These thresholds were to be updated each year to reflect changes
in prices, but nothing else. Since 1969 the Orshansky measure has been
subjected to considerable criticism, but, with minor changes, it still
forms the basis for the official poverty measure.
Much of discussion about the Orshansky measure focused on the
specific poverty thresholds it calculated, but Orshansky also made
another choice that had important implications for the official poverty
measure. The official poverty measure actually consists of two parts--
the set of thresholds discussed above and a measure of family resources
that is compared to those thresholds. In order to decide if a
particular family is poor, its income or consumption level is compared
to the threshold, and a level of poverty is determined based on whether
the family's resources fall above or below the threshold. In making
this comparison, Orshansky used the only measure of family economic
resources available to her--pre-tax cash income. Although we now
collect data on many other aspects of family needs and resources, the
official poverty definition still bases its estimates on pre-tax cash
income. This measure has become more and more outdated as taxes, non-
cash benefits, and work expenses such as child care have come to play a
larger role in family budgets.
Alternative Measures of Poverty
In considering alternatives to our official poverty measure, we
must examine both the thresholds used to define poverty and the methods
used to compute the incomes that are compared to those thresholds.
Adjusting one side of the measure without changing the other produces
misleading and often contradictory results. For example, adding non-
cash benefits such as child care subsidies to income, without adjusting
the thresholds to reflect the increased need for child care as more and
more mothers of young children work outside the home, will obscure the
very real hardships faced by many working mothers and their children.
Ultimately, any proposed alternative must be assessed in terms of its
treatment of both sides of the measure.
Setting Poverty Thresholds. Two major types of criticism have been
leveled at the thresholds embodied in the official poverty measure.
First, there have been many technical complaints. The adjustments for
family size and type, for example, are based on the distribution of
family sizes in the population in 1969, and bear little resemblance to
today's needs. In fact, the entire needs measure is rather ad hoc, and
many revisions have been suggested. Family needs have also changed as
more family members have gone into the labor force, and as the numbers
of single-parent families has risen.
The second type of criticism of the official poverty line is much
more fundamental. Most economists believe, as Adam Smith put it in
1776, that people can be considered poor if they are unable to afford
the things that ``the custom of the country renders it indecent for
creditable people, even of the lowest order, to be without.'' In other
words, poverty is not just the inability to afford a subsistence diet;
instead, the meaning of poverty varies from time to time and place to
place.
The implications of this belief for poverty measurement can be
envisioned by imagining a poverty line set in 1900, for example. At
that time, basic needs would not have included indoor plumbing, central
heating, electricity, or many other things that we now take for
granted--because the ``custom of the country'' did not include those
things even for most of those who were well-off, let alone those ``of
the lowest order.'' Today, however, Americans who cannot afford those
things would generally be considered poor. Indeed, a poverty threshold
that didn't include them would be fairly meaningless, since it would be
very difficult for the urban poor, in particular, to find housing
without them. Failing to include them in budgets, therefore, would
restrict the poor to a standard of housing that no longer even exists
in this country, and would not be acceptable to most as a basis for
public policy.
An approach to measuring poverty that takes into account changes in
standards of living over time is often referred to as ``relative''
poverty measure, as opposed to the ``absolute'' poverty measure
embodied in our official thresholds. Some analysts argue that an
absolute measure--one that sets a poverty line at a fixed level at a
particular point in time and that subsequently adjusts only for changes
in prices--is needed to assess programs, because a measure adjusted for
changes in standards of living presents too much of a moving target.
Such analysts assert that if the poor become better off in material
terms because society as a whole is becoming better off, that is
important evidence of public policy success and should be taken into
account in considering the need for continued assistance programs. A
relative approach to poverty measurement, in contrast, is seen by such
analysts as biased toward continuing high poverty rates, because
progress against poverty can only be made if the underlying
distribution of resources changes. People who are in favor of helping
those in need do not necessarily favor large changes in the
distribution of income, which supporters of an absolute measure argue
would be necessary to show progress under a relative poverty measure.
Most developed countries use some type of relative measure in
assessing progress against poverty, however. The OECD has published a
measure based on 50 percent of median family income, for example, while
Eurostat uses a measure based on 60 percent of the median. Such
measures increase over time as family incomes increase, but continue to
identify those whose resources fall substantially below the norm.
Interestingly, Orshansky's threshold for a two parent, two child family
fell at about 50 percent of median income when it was first introduced.
Because it has been updated only for price changes over time, however,
the official poverty threshold for such a family has now fallen to
about 35 percent of median family income.
While this type of income-related measure does give policy makers
useful information about relative resources, it has a few drawbacks as
a poverty measure. One of the most important is that when incomes fall
during a recession, the poverty thresholds will fall too, even though
basic needs may not have declined. More broadly, this measure does
indeed track resources rather than needs over time, and as such does
not directly measure how many people have a level of resources that
society might consider ``indecent.''
Various proposals have been made over time to update the official
poverty measure. A thorough assessment by an interagency task force in
1976 concluded that even then the measure was seriously out of date and
needed revision, but the proposed revisions were never implemented.
The most serious and comprehensive reassessment of the U.S. poverty
measure was undertaken by the National Research Council (NRC) of the
National Academy of Sciences (NAS) in the early 1990s, in response to a
Congressional mandate. This scientific study by an independent panel of
experts led to the report Measuring Poverty: A New Approach. This study
made a number of important recommendations relating both to the setting
of poverty thresholds and to the measurement of economic resources.
In setting poverty thresholds, the Academy report recommended an
intermediate approach that was neither strictly absolute nor entirely
relative. Under the NAS approach, a market basket of basic goods
including food, clothing, shelter, and a small allowance for other
needs would serve as the basis for a new set of poverty thresholds.
Unlike our official thresholds, these thresholds would be adjusted for
differences in costs of living in different parts of the country. The
method for adjusting for differences in family sizes would also be
updated to reflect today's family needs. Finally, these thresholds
would be updated periodically as people's consumption patterns--and
thus, their ideas about what constitute ``basic needs''--changed over
time.
This type of approach is somewhat more difficult to implement than
the present system, because it requires some judgment calls about what
people need, but it has the advantage of tracking what most people mean
by poverty much more closely than either the official measure or a
purely relative measure. And this type of updating is not
unprecedented--similar periodic revisions are used to update the market
basket underlying the Consumer Price Index, for example. A measure of
this type will generally increase more over time than a strictly
``absolute'' measure, because as standards of living rise people
generally raise their expectations for even minimal consumption levels,
but the increases will be linked to changes in consumption levels and
costs rather than to changes in income. Research on what the public
thinks is a minimum for ``basic'' needs, based on Gallup poll data,
shows that the level has indeed risen over time, but not as much as
income has risen.
The figure below, put together by John Iceland, shows how poverty
thresholds (adjusted for changes in prices) would have varied over time
under four different measures--the official measure, a relative measure
at 50 percent of median income, a subjective measure based on polling
data, and the NAS-recommended measure. As can be seen, all three of the
alternative measures would have produced higher thresholds than the
official measure, but the NAS measure falls between the official
measure and an entirely relative measure.
[GRAPHIC] [TIFF OMITTED] T3757A.003
Measuring Economic Resources. Poverty thresholds, however, are only
half the story. The official poverty measure also uses a method of
computing family income that is widely considered to be seriously
outdated. When Orshansky undertook her original studies, the only
information on family incomes that was available was about cash income
before taxes. This was not as big a limitation in 1963 as it is today,
because there were few regular sources of non-cash benefits and the
poor paid very little in taxes--and received no refundable credits.
Orshansky also failed to account for work expenses such as day care in
measuring poverty, but again that was not as large a limitation when
most families had only one earner and single mothers were comparatively
rare.
Today, of course, the story is different. Low-income families often
pay substantial payroll taxes, and many qualify for the refundable
Earned Income Tax Credit. Many receive non-cash benefits such as food
stamps and housing subsidies that effectively work like cash in
increasing their ability to meet basic needs. And many working mothers
must pay for child care--leaving their net earnings substantially lower
than their apparent cash incomes. The NAS report recommended taking all
of these changes into account in computing family incomes. The official
measure, however, is still based strictly on pre-tax cash income.
Another change in the computation of income recommended by the NAS
report was that out-of-pocket medical expenses should be deducted from
income in considering whether or not someone is poor. The logic of this
recommendation is that low-income families are likely to have little
choice about paying for medical care and prescriptions, and anything
they must spend on health care comes out of the funds available for
other basic consumption needs.
Over time, the cost of medical care has risen dramatically, and
more recently the number of uninsured has also increased. These factors
have considerably increased the amount that low-income people,
especially the elderly, must spend on their medical needs. It is also
true that the quality of care has improved substantially--many people
are alive today who would have died at a younger age given the medical
standards of the 1960s. But if we do not allow for the costs of today's
medical care in setting poverty standards, many lower-income people
with expensive medical needs will not be treated as poor, even if they
have very low disposable incomes. In effect, most observers argue, not
counting those needs in measuring poverty implies that advances in
medical care since the 1960s should only be available to those with
higher incomes.
Theoretically, the costs of medical care could be incorporated into
the measure of basic needs--the poverty thresholds--rather than being
deducted from income. In practice, however, out-of-pocket medical
expenses vary so much from person to person that a one-size-fits-all
threshold cannot adequately take them into account. It is for this
reason that the NAS recommended that they be deducted from the income
measure instead.
The table below summarizes the differences between the official
poverty measure, a revised measure like the one recommended by the NAS
panel, and a purely relative measure such as 50 percent of the median
income. It suggests that a measure based on the NAS recommendations is
more closely attuned to the specific needs and resources of the low-
income population than either the official poverty measure or a purely
relative measure.
Table 1: Characteristics of Alternative Poverty Measures
----------------------------------------------------------------------------------------------------------------
Official Measure NAS Panel Measure 50% of Median
----------------------------------------------------------------------------------------------------------------
Thresholds Based on 1955 food Based on current Based on median
consumption survey consumption needs family income
----------------------------------------------------------------------------------------------------------------
Non-Cash Income Excluded ``Cash-like'' sources Excluded
included
----------------------------------------------------------------------------------------------------------------
Out-of-Pocket Medical Not considered Deducted from income Not considered
Expenses
----------------------------------------------------------------------------------------------------------------
Work Expenses Not considered Deducted from income Not considered
----------------------------------------------------------------------------------------------------------------
Taxes Not considered Deducted from income Usually not
considered
----------------------------------------------------------------------------------------------------------------
Family size Based on food needs for Based on the relative Usually none
adjustment a 1969 mix of family needs of additional
sizes and types adults and children
----------------------------------------------------------------------------------------------------------------
Implications for Public Programs of Revising the Poverty Measure
In assessing the direct impact on federal programs of changes in
the poverty measure it is important to understand that there are
actually two versions of the measure. In addition to the poverty
thresholds published by the Census Bureau for research and statistical
purposes, there are ``poverty guidelines'' published by the Department
of Health and Human Services (HHS). These are the poverty measures used
for administrative purposes such as determining program eligibility.
The HHS guidelines are closely related to the Census measures, but
have a few differences. First, Census's thresholds are determined for
past years only--when new poverty estimates come out in August or
September of each year, they are based on the previous year's income
data. For example, poverty rates for 2006 will be announced later this
month. Because inflation rates for 2007 are not yet known, Census
cannot yet calculate the poverty thresholds for 2007.
The poverty guidelines, which come out in February, are based on
the thresholds produced by Census the previous fall, but are updated
using preliminary inflation estimates--so that estimated 2007
thresholds are applied to program determinations made in 2008. The
guidelines also use a simplified and more rational set of family size
adjustments than those used in the official measure. Finally, the
guidelines refer only to the poverty thresholds--methods of determining
the income to be compared to the guidelines are left up to individual
program rules. Few programs use income-computing methods anything like
those used by Census for the official poverty measure. For example,
most programs allow for the deduction of work expenses, and many take
into account benefits received from other programs.
The two biggest programs that use the federal poverty guidelines in
their eligibility criteria are the Food Stamp Program (FSP) and
Medicaid. Both programs implicitly recognize that the thresholds
produced under the guidelines are unrealistically low; the FSP sets
gross income eligibility at 130 percent of the poverty guidelines,
while Medicaid uses 133 percent. The school lunch and breakfast
programs use 130 percent of the guidelines to determine eligibility for
free meals, and 185 percent for reduced-price meals. Special subsidies
for prescription drugs are also available under Medicare for
individuals below 130 percent and 150 percent of poverty.
Census poverty estimates (although not generally the thresholds
themselves) are used for some program-related purposes. A typical use
involves poverty estimates for states and localities, which are
considered as one factor in a multi-factor funding--allocation formula.
These estimates are not based on the same data as the national poverty
rates published each year by the Census Bureau; instead, they are
produced using Decennial Census and American Community Survey data,
which are inputs into models of changing local-area incomes over time.
The largest program using such estimates is the Title I grant program
for education.
In other words, although many programs address the problem of
poverty, and others take poverty-related issues into account in
allocating funds, the Census poverty thresholds themselves are not
generally a direct factor in program design. The major programs that
use the HHS guidelines in determining eligibility all use levels that
are well above the Census thresholds, implying that program designers
are aware that current threshold levels are unrealistic. And the
programs that use Census poverty numbers in allocating grants use
estimates produced by a complex model based on data that is several
years old.
Introduction of an alternative Census poverty measure, therefore,
might not have any direct impact on program eligibility, depending on
the relationship between that alternative and the HHS guidelines. If
both the current official measure and the alternative were maintained
for some time, the HHS guidelines could continue to be computed exactly
as they are now. Over time, alternative calculations of the guidelines
could be modeled and program rules could be modified appropriately on
an as-needed basis. And over the longer run, a measure that tracked
changing needs more closely than the official measure does would
presumably improve program eligibility assessments.
Similarly, producing an alternative measure would not necessarily
affect programs with funding allocated by formulas involving poverty,
at least in the short run. Most of these formulas depend on the
comparative poverty status of a specific locality relative to the state
or the nation as a whole. These estimates of comparative poverty status
are typically not very sensitive to small differences in the specific
poverty thresholds used. In fact, any changes allocations resulting
from changes in the thresholds might be hard to identify, because
typically local poverty rates are only one factor in the allocation
formula and the other factors are also changing over time. In any case,
the official thresholds would almost certainly have to be used for
small area poverty estimation in the near future, until new data were
available and the models could be updated. In the longer run, switching
to a poverty measure that tracks real needs more closely might be
expected to improve our ability to allocate funds appropriately.
Conclusion
There is a widespread consensus among experts that a revision of
the official U.S. poverty measure is both feasible and needed. The
current measure is based on underlying data, assumptions, and
measurement methods that were assembled more than 40 years ago. Without
an update, the measure is becoming progressively more removed from a
real measure of needs. No other major statistic produced by the federal
government has been issued over so long a period of time without being
re-benchmarked and having its methodology updated. It is time for a
revision of the poverty measure similar to those that are performed
routinely for other measures such as the Consumer Price Index (CPI) and
Gross Domestic Product (GDP).
The alternative measure produced using the recommendations of the
NAS panel on poverty measurement does a good job of addressing the
shortcomings of the current official poverty measure. The Census Bureau
has already produced a series of estimates of poverty under this
measure, showing that it can be implemented. Some details of the
measure remain to be determined--there are possible variations in data
and methods for measuring some income sources and expenses, for
example. But these issues could be quicklGy addressed either through an
interagency task force or with the help of outside experts, as was done
in the case of the CPI. And as Census's work so far has demonstrated,
all of the variations on this methodology produce a consistent picture
of need in the U.S. that differs in important ways from the picture
obtained using the official measure.
Revising the poverty measure doesn't have to mean radical short-run
changes in federal programs, however. The Census poverty thresholds are
not a direct input into program rules, and they have only an indirect
effect on funding-allocation formulas. Further, as with any change in
methodology, the effects of the revision should be tracked over time.
For the next several years the current measure and the alternative
could both be produced so that the potential impacts of the change can
be assessed. Over the longer run, anti-poverty programs can only be
improved by using a poverty measure that does a better job of tracking
changes in the needs of the poor.
Thank you for this opportunity to testify, and I look forward to
answering any questions you may have.
Chairman MCDERMOTT. Thank you.
Dr. Iceland is Associate Professor of Sociology at the
University of Maryland and I think once worked with the Census.
Dr. Iceland.
Dr. ICELAND. Thank you. Thank you for having me here today.
Chairman MCDERMOTT. Do you want to turn on your microphone
so that the reporter can get it down?
Dr. ICELAND. Sure.
STATEMENT OF JOHN ICELAND, PH.D., ASSOCIATE PROFESSOR OF
SOCIOLOGY, UNIVERSITY OF MARYLAND
Dr. ICELAND. Thank you, Chairman, and members of the
Subcommittee.
I appreciate having this opportunity today to talk about
these issues. In this short presentation, I will talk about
different kinds of poverty measures, focusing on absolute,
relative, and the NAS recommended measure. I will elaborate on
some of the themes mentioned by yourself and Dr. Ruggles. I
will discuss their strengths and weaknesses and show results of
what poverty rates look like when using the different measures.
Overall, here is the story I wanted to leave you with.
While various poverty measures are useful, I believe the NAS
panel-recommended measure is the single-most informative one,
because, first, it measures the ability of families to meet
their basic needs, while second, it also takes into account the
fact that such needs can change over time. Let me take a step
back and talk briefly about what absolute and relative poverty
measures are.
They both involve comparing a family's income to a poverty
threshold to determine whether that family is poor. Absolute
measures, such as the current U.S. official measure, use
thresholds that typically attempt to define a truly basic need
standard. Relative measures, more commonly used in Europe,
explicitly define poverty as a condition of comparative
disadvantage. Thresholds are often pegged at, say, 50 percent
of median household income. The key distinction between the two
is that absolute poverty lines remain constant over time, while
relative ones rise in real dollars as standards of living rise.
The NAS panel proposed a hybrid, sometimes termed ``quasi-
relative poverty measure.'' The thresholds of such a measure
changes as real spending on basic goods change. Because the
poverty lines can change over time in real dollars, the measure
is at least in part relative. However, since it is not based on
mean or median incomes as most relative poverty measures are,
it is not wholly relative either. There are advantages and
disadvantages of alternative measures. The main advantage of
absolute measures is that they are easy to understand. For
example, if there is starvation and hunger, then there is
clearly poverty, regardless of general standards of living.
The main theoretical criticism of absolute measures is that
what people judge to be poor varies across time and place. In
addition, sometimes real needs also rise in richer countries.
For example, a poor family cannot simply build a shack on the
outskirts of a U.S. city without getting in trouble for
violating building ordinances. If you own a car, then you must
pay insurance leaving less money to meet other needs. Such
requirements would be unreasonable in societies with low
standards of living.
I would like to just briefly show the figure also shown in
Dr. Ruggles' testimony to make a couple of additional points
and to highlight the different measures.
[GRAPHIC] [TIFF OMITTED] T3757A.003
[GRAPHIC] [TIFF OMITTED] T3757A.004
[GRAPHIC] [TIFF OMITTED] T3757A.005
[GRAPHIC] [TIFF OMITTED] T3757A.006
Dr. ICELAND. So, again, here from the previous figure we
saw the absolute poverty threshold, which remains the same over
time, being as it is an absolute measure. There were questions
asked by Gallup on what people thought about poverty. Notice
how well it tracks median incomes. This is half the median
income. The last thing I would like to note from here, at the
end we have the NAS threshold, which goes back only to the
early 1990s. The version of the NAS threshold that includes
out-of-pocket medical expenses shown here is a little higher
than the official threshold, though the two thresholds closely
tracked each other for most of the 1990s.
However, the NAS threshold rose modestly more quickly after
1999 when real spending on a basic bundle of items rose faster
than inflation, largely due to relative increases in housing
and medical costs. Thus, the NAS measure again takes into
account real changes in need, while the official threshold does
not.
Figure 2 shows using different poverty measures, so the
poverty rate is highest when using the relative poverty
measure, whose thresholds are higher than those used in the
other measures. The ``disposable income'' measure, which
recently appeared in a Census report, shows the lowest poverty
rate, because it has a more refined measure of family resources
that includes among other things in-kind transfers as it
should.
However, this measure still uses the outdated referenced
family official threshold. The NAS poverty rate is only
modestly above the official rate, because it has a higher
threshold, but takes into account in-kind benefits. My written
testimony provides more details on these issues, including how
poverty rates differ across demographic groups when using
alternative measures.
To conclude, poverty measurement efforts in the U.S. are
approaching a crossroads. A key theoretical issue is whether
poverty should refer to a subsistence standard, such as severe
malnutrition, economic marginalization or something in between.
The NAS panel-recommended measure has the advantage of
increasing, in real terms, spending on basic items increase, as
to reflect changes in real standards of living. Yet, it is not
responsive to changes in spending patterns on other more
discretionary items such as luxury goods that may occur as
median incomes rise and thus reflected in purely relative
poverty measures.
Thus, while all of these measures tell something useful
about people's economic well-being, the NAS measure is in my
view the single most informative poverty measure among them.
While no measure will ever garner universal support, this NAS
measure very much represents a broad consensus among a wide
array of social scientists on how to best measure poverty.
[The prepared statement of Dr. Iceland follows:]
Prepared Statement of John Iceland, Ph.D., Associate Professor,
Sociology Department, University of Maryland, College Park
Chairman McDermott and members of this subcommittee, I thank you
for the opportunity to testify before you. I sincerely applaud your
efforts to revisit poverty measurement issues. In order to best target
our policy efforts, we, as a country, need to have a good yardstick by
which to measure progress.
There are a number of poverty measures one could use to estimate
levels of economic well-being in society. Income poverty measures are
perhaps the most common. They usually involve comparing a family or
household's income to a poverty threshold to determine whether that
family or household is poor. Two basic types of income poverty measures
are absolute and relative measures. Absolute measures, such as the
current U.S. official measure, are ones that typically attempt to
define a truly basic--absolute--needs standard that remains constant
over time and perhaps updated only for inflation, as in the case of the
U.S. official poverty measure. Relative measures, which are more
commonly used by researchers in Europe, explicitly define poverty as a
condition of comparative disadvantage, to be assessed against some
evolving standard of living. The key distinction between the two is
that absolute poverty lines remain constant over time, while relative
ones rise in real dollars as standards of living rise.
In the United States there has been growing dissatisfaction with
the current official poverty measure (described in more detail below).
At the request of Congress, the National Academy of Sciences (NAS)
convened a panel of researchers in the early1990s to review this
poverty measure and make recommendations for a new one. The panel
produced a report, Measuring Poverty, which proposed a hybrid poverty
measure (National Research Council 1995). The thresholds of such a
measure would change as real spending on basic goods (e.g., food,
clothing, and shelter) change. Because the poverty lines change over
time in real dollars, the measure is, at least in part, relative.
However, since it is not based on mean or median incomes, as most
relative poverty measures are, it is not wholly relative either.
Overall, each of these income poverty measures is informative and
they should be viewed as complementary sources of information about
people's economic well-being. My own view, however, is that the hybrid
poverty measure recommended by the NAS panel is the single most
informative income poverty measure because it best measures the ability
of families to meet their basic needs while also recognizing that such
needs can change over time.\1\
---------------------------------------------------------------------------
\1\ Much of the analysis in this testimony is drawn from a recent
paper I wrote (Iceland 2005) that provides greater detail about poverty
measurement issues and challenges.
---------------------------------------------------------------------------
Background
The Current Official Poverty Measure
The current official poverty measure has two components: poverty
thresholds and the definition of family income that is compared to
these thresholds. Mollie Orshansky, an economist at the Social Security
Administration, developed poverty thresholds in 1963 and 1964 by using
the ``Economy Food Plan'' (the lowest cost food plan) for families of
different types and sizes prepared and priced by the U.S. Department of
Agriculture. To arrive at overall threshold figures, Orshansky
multiplied the price of the food plans by three, based on information
from the 1955 Household Food Consumption Survey that indicated that
families of three or more people had spent about one-third of their
after-tax income on food in that year. The thresholds have been updated
yearly for inflation using the Consumer Price Index (CPI).
The definition of family resources used to compare to the
thresholds is the Census Bureau's definition of income--gross annual
cash income from all sources, such as earnings, pensions, and cash
welfare. A family and its members are considered poor if their income
falls below the poverty threshold for a family of that size and
composition.
The current official poverty measure was, for a time, a sensible
indicator of material deprivation in the U.S. At the time of its
initial adoption by the Office of Economic Opportunity in 1965, the
poverty lines were set at a dollar level that coincided with people's
views of poverty. The method of measuring people's resources--gross
cash income--also managed to fairly accurately capture the income
people had to meet their basic needs.
Over the past 40 years, however, the poverty measure has become
increasingly outdated. The poverty lines, originally devised by
multiplying the cost of food needs by three to account for other needs
(such as clothing and shelter), no longer captures families' basic
needs in par because of the rapid growth in housing prices and other
expenditures (such as medical care and childcare) relative to food
prices. Today, people spend closer to one-sixth or one-seventh of their
income on food rather than one-third. While the official poverty
threshold for a four-person family once coincided with people's views
of the dollar amount needed to support such a family in the 1960s--as
reported in public opinion surveys--this was no longer true by the
1990s (National Research Council 1995).
Many have also argued that the definition of money income used in
the official measure--gross cash income--inadequately captures the
amount of money people have at their disposal to meet basic needs. It
has been argued that taxes should be subtracted from income, as this
money cannot be spent to meet basic needs, and that in-kind or near-
money government benefits should be added, such as food stamps, housing
and child care subsidies, and the EITC. The omission of these items
from the official definition of income has become increasingly serious
in recent years because government transfers are now concentrated in
benefits that are not considered part of families' gross cash income.
The unfortunate result is that the current official poverty measure no
longer accurately captures either people's perceptions of poverty or
the effect of various policies on people's resources.
The 1995 National Research Council Report on Poverty and Subsequent
Research
In response to the increasingly apparent weaknesses of the official
poverty measure, the U.S. Congress appropriated funds for an
independent scientific study of the official poverty measure, which led
to the 1995 National Research Council (NRC) report, Measuring Poverty:
A New Approach. The report recommended that a new poverty threshold be
calculated by determining, for a reference family of two adults and two
children, a dollar amount for food, clothing, shelter, and utilities,
and then increasing that dollar amount by a modest percentage to allow
for other needs (such as household supplies, personal care, and non-
work-related transportation). The dollar amount would be scaled down
from the median spending for those four basic items using data gathered
in the Consumer Expenditure Survey (CE). This threshold would then be
adjusted for families of different sizes and types by using an
equivalence scale. Finally, unlike in the official U.S. poverty
measure, the thresholds would be further adjusted for housing cost
variations across regions and metropolitan areas of different
population sizes.
Family resources in the NRC report are defined as the value of cash
income from all sources plus the value of near-money benefits that are
available to buy goods and services covered by the new thresholds,
minus some basic expenses. Cash income sources are the same as those in
the current official Census Bureau poverty measure. The income
definition also includes near-money income: food stamps, housing
subsidies, school breakfast and lunch subsidies, home energy
assistance, assistance received under the Woman, Infants, and Children
nutritional supplement (WIC) program (if the data are available), the
EITC, and realized capital gains (or losses). Basic expenses to be
subtracted include taxes, child care and other work-related expenses of
working parents, medical out-of-pocket costs, and, if the data are
available, child support payments made to another household. Taxes
represent a nondiscretionary expense in that people cannot spend this
money. Child care and other work-related expenses (such as commuting
expenses) are also subtracted because, the panel argued, these costs
are often incurred if parents are to work and earn labor market income.
The release of the NRC report in 1995 was followed by considerable
research activity. The Census Bureau released a few subsequent reports
that were devoted to experimental poverty measures (for example, Short
et al. 1999; Short 2001). Over 50 research papers on experimental
poverty measures have been written by researchers in various other
government agencies, including the Bureau of Labor Statistics, the
Department of Health and Human Services, and the Office of Management
and Budget, to name a few, and by researchers at think tanks and
various universities.\2\ This research has helped identify strengths
and weaknesses in the NRC recommendations. The National Academy of
Sciences convened another workshop in 2004 to review the elements in
the measure. On the whole, workshop participants agreed that the
measure represented a significant improvement over the current official
poverty measure; participants also made additional recommendations on
some of the elements (National Research Council 2005).
---------------------------------------------------------------------------
\2\ Many of these papers are available on a Census Bureau website.
See http://www.census.gov/hhes/www/povmeas/nas.html.
---------------------------------------------------------------------------
Strengths and Weaknesses of Absolute and Relative Poverty Measures
Technical issues aside, there are conceptual advantages and
disadvantages of alternative measures. The main advantage of absolute
measures is that they are conceptually easy to understand and
intuitively appealing. For example, if there is starvation and hunger,
then there is clearly poverty--regardless of how high or low the
overall standards of living in a society. The main theoretical
criticism of absolute measures is that what people judge to be poor
varies across both time and place. Standards of living in the developed
and developing world clearly differ. Even within the U.S., as standards
of living change, so have people's perceptions of what poverty means.
Poverty lines and minimum subsistence budgets devised by researchers
and social workers in the early 1900s were, in inflation-adjusted
dollars, generally between 43 and 54 percent of the subsequent U.S.
official poverty thresholds devised in 1963 (Fisher 1995). Economists
describe this phenomenon as the ``income elasticity of the poverty
line''--the tendency of successive poverty lines to rise in real terms
as the real income of the general population rises. Some people
therefore argue that poverty is by its nature relative; people are poor
when others think of them as poor.
Relative poverty measures address this weakness of absolute
measures. Relative measures are explicitly based on the notion that
poverty is relative to a society's existing standards of living.
Implicit is that people are social beings who operate within
relationships. People whose resources are significantly below the
resources of others, even if they are physically able to survive, may
not be able to participate adequately in social organizations and
relationships. Adam Smith (1776) argued that to be poor was to lack
what was needed to be a ``creditable'' member of society. He noted that
in his day (the 18th century), a man needed a linen shirt if
he was to appear in public ``without shame.''
The most common method of measuring relative poverty is setting a
threshold at a percentage of the national median household income. For
example, analysts comparing poverty across countries in the European
Community and the United Kingdom have often specified a poverty
threshold at half the median income. Other relative thresholds, such as
40 percent of the median or 60 percent of the median household income
(the latter is the official European Union poverty threshold), have
also been used by researchers.
Relative measures also have advantages and disadvantages. On the
positive side, advocates argue that the relative notion underlying
these measures fits with both the historical record and changing views
of poverty. Second, sometimes real needs do indeed rise in richer
countries. For example, while a car may be a luxury in some countries,
in a society in which most families own cars, and where public
transportation services are also poor, a car may often be needed to
find a job and commute to work. Moreover, car owners in many places may
be required to purchase car insurance, thus leaving families with less
disposable income to meet other needs. Such requirements would be
considered unreasonable in societies with lower standards of living.
On the other hand, some have asserted that relative measures
conceptually unappealing, believing that ``poverty'' should refer to a
truly subsistence standard. Relative measures can also, at times,
behave in deceptive ways over the short run, particularly during
periods of economic growth and recession. In particular, relative
thresholds sometimes decline in bad times as median incomes fall. This
could result in a decline in measured poverty rates, even though low-
income people are faring (and feel themselves faring) worse than
before.
Nevertheless, it is no accident that relative measures have become
more common in rich industrialized countries and in analyses that
involve cross-national comparisons in the developed world. The
Organization for Economic Cooperation and Development (OECD) argues,
for example, that absolute poverty lines have little meaning in such
societies, and that poverty should be thought in more in terms of
exclusion from standards of living generally available to others in the
same society (OECD 2001). Measuring poverty in terms of absolute need--
such as starvation--has simply become less meaningful because of how
infrequently it occurs in these contexts.
Empirical Analysis
The analysis below has three parts. I begin by examining how
different poverty thresholds behave over time, including how they
compare to people's subjective notions of poverty. This is followed by
a look at poverty rates derived from different measures. Finally, I
compare official and NAS poverty rates across demographic groups to
show what types of differences they produce.
Trends in Thresholds
As described earlier, research has shown that people's opinions of
what constitutes poverty increase as standards of living increase. A
pair of studies by Denton Vaughan (1993, 2004) examined subjective,
absolute, and relative poverty thresholds in the United States. The
subjective poverty thresholds were essentially based on the following
question last asked by the Gallup Organization in 1993:
People who have income below a certain level can be
considered poor. That level is called the ``poverty line.''
What amount of weekly income would you use as a poverty line
for a family of four (husband, wife and two children) in this
community?
Vaughan estimated a subjective poverty standard back to 1946 based
on a few assumptions (see his papers for details). He found that over
the period, the subjective poverty threshold averaged about 52 percent
of the median four-person family income net of taxes--a figure that
closely approximates the level at which a typical relative poverty
threshold is operationalized (see Figure 1). Despite some fluctuations,
the subjective poverty needs standard declined from about 56 percent of
the median income in the 1947-1950 period to about 49 percent in the
1984-1989 period, perhaps indicating that subjective poverty thresholds
may rise modestly less quickly than relative poverty thresholds
(Vaughan, 2004: 16).
Nevertheless, it is clear that relative thresholds perform better
when evaluated vis-a-vis subjective thresholds than the official U.S.
poverty threshold, which is an absolute threshold. As such, this
threshold remained constant over time in real dollars. Vaughan's (2004)
study indicates that while the official poverty threshold projected
back to 1947 was higher than the subjective poverty threshold in that
year (35 percent greater), these thresholds were about equal at the
time that Mollie Orshansky devised the official poverty thresholds in
1963 and 1964, and by 1989 the official poverty threshold was in fact
only 81 percent of the subjective one. This indicates that the official
poverty threshold has indeed become less socially meaningful since the
1960s, and considerably less meaningful over time than the relative
threshold.
[GRAPHIC] [TIFF OMITTED] T3757A.007
There is no poverty threshold time series based on recommendations
by the National Academy of Sciences that spans enough time to
appropriately evaluate how they compare with subjective notions of
poverty, since no version of NAS thresholds exist before 1989. The
version of the NAS poverty threshold that includes out-of-pocket
medical expenses (shown in figure 1) is a little higher than the
official threshold, though the two thresholds fairly closely tracked
each other for most of the 1990s. However, the NAS threshold rose
modestly more quickly after 1997, when real spending on a basic bundle
of items rose faster than inflation, largely due to relative increases
in housing and medical costs (see also U.S. Census Bureau 2007a).
Poverty Estimates Using Alternative Measures
Figure 2 shows poverty rates using different poverty measures in
2005 (except for the relative measure, which refers to poverty in
2000). In addition to the official poverty measure, the figure includes
two NAS-based measures. One has thresholds that have been updated only
for inflation since 1999, while the other has thresholds that have been
updated for changes in real spending on basic goods using Consumer
Expenditure Survey data, as recommended by the NAS panel (U.S. Census
Bureau 2007a).\3\ The ``disposable income'' measure is a new one
included in a recent Census Bureau Report on income and poverty (U.S.
Census Bureau 2007b). This measure uses the same reference family
threshold as is used in the official poverty measure (though it uses a
different method for adjusting this threshold for families of different
sizes and composition). In addition to money income, this measure adds
capital gains, imputed rental income, and the value of noncash
transfers such as food stamps, housing subsidies, and free lunches. It
subtracts work expenses (though not childcare), federal and state
income taxes, payroll taxes, and property taxes for owner-occupied
homes. The relative poverty measure comes from Luxembourg Income Study
(2007), and it uses a threshold equal to half the median household
income in the U.S.
---------------------------------------------------------------------------
\3\ The inflation-adjusted measure is termed ``MSI-GA-CPI'' while
the other is ``MS-GA-CE'' in the Census Bureau tables.
[GRAPHIC] [TIFF OMITTED] T3757A.008
The poverty rate is highest when using the relative poverty measure
(17.6 percent), whose thresholds are higher than those used in the
other measures. The ``disposable income'' measure shows the lowest
poverty rate (10.3 percent) because it has a more refined measure of
family resources that includes in-kind transfers. However, this measure
still uses the outdated official thresholds (except for including a
more refined equivalence scale).
The figure also shows that the official poverty and the inflation-
adjusted NAS poverty rates are the same, at 12.6 percent each. Note
that the NAS measure with CE-adjusted thresholds is a little higher at
13.5 percent, mainly because spending on basic goods has risen in real
dollars since 1999. The early 2000s witnessed a rise in housing prices
in many areas of the country. Figure 3 illustrates these poverty rate
trends in more detail.
Table 1 shows how poverty rates vary across demographic groups when
using the official poverty measure versus the NAS measure with
inflation-adjusted thresholds.\4\ While the overall poverty rates are
the same, we see that NAS poverty rates are higher for people in
married-couple families and the elderly. These differences have been
noted by previous research (Short et al., 1999). Married-couple
families are less likely to receive government transfers and often
incur higher work-related expenses. The elderly often have higher out-
of-pocket medical expenses. African Americans and people in female-
householder families tend to have lower poverty rates when using the
NAS measure, in part due to greater receipt of non-cash government
transfers.
---------------------------------------------------------------------------
\4\ I would have preferred to use the NAS measure with CE-adjusted
thresholds rather than inflation-adjusted thresholds, but there were no
demographic cross-tabulations with the former available on the Census
Bureau website.
[GRAPHIC] [TIFF OMITTED] T3757A.009
[GRAPHIC] [TIFF OMITTED] T3757A.010
Conclusion
Poverty measurement research efforts in the United States are
approaching a crossroads. Most people recognize the significant faults
of the current official poverty measure, but no new measure has yet
taken its place. There has been increasing use of relative measures
among academic researchers, as is already common in Europe, but not
really among those outside the academy. Some poverty researchers have
also shown interest in the quasi-relative measure recommended by the
National Academy of Sciences panel. A key theoretical issue is whether
``poverty'' should refer to a subsistence standard, such as severe
malnutrition or starvation (a notion associated with absolute poverty
measures), or to economic marginalization (one associated with relative
measures), or to something in between.
I would argue that the NAS panel-recommended poverty measure
addresses many weaknesses of both purely absolute and purely relative
measures. On the positive side, the NAS panel-recommended measure is
technically a more refined measure than the current official poverty
measure in both the construction of the thresholds and the definition
of income used. It is designed to gauge the impact of government
programs on poverty, given that both cash and non-cash government
benefits are taken into account in the measure of family income.
Conceptually, the NAS measure has the advantage of increasing, in
real terms, as spending on basic items increase, as to reflect changes
in real standards of living. Yet, it is not responsive to changes in
spending patterns on other, more discretionary items--such as luxury
goods--that may occur as median incomes rise. I believe that this is a
desirable property of a poverty measure. Relative poverty thresholds,
in contrast, are simply responsive to changes in median income; as such
they less directly measure the ability to attain basic goods or
capabilities. Thus, while all of these measures tell us something
useful about people's economic well-being, the NAS poverty measure is,
in my view, the single most informative poverty measure among them.
References
Fisher, Gordon M. 1995. ``Is There Such a Thing as an Absolute
Poverty Line Over Time? Evidence from the United States, Britain,
Canada, and Australia on the Income Elasticity of the Poverty Line.''
U.S. Census Bureau, Poverty Measurement Working Paper: http://
www.census.gov/hhes/www/povmeas/papers/elastap4.html.
Iceland, John. 2006. Poverty in America (2nd Edition).
Berkeley, CA: University of California Press.
Iceland, John. 2005. ``Measuring Poverty: Theoretical and Empirical
Considerations.'' Measurement 3, 4: 199-235.
Luxembourg Income Study. 2007. ``Luxembourg Income Study (LIS) Key
Figures.'' Available at: http://www.lisproject.org/keyfigures.htm,
retrieved July 27, 2007.
National Research Council. 1995. Measuring Poverty: A New Approach.
Panel on Poverty and Family Assistance, Constance F. Citro and Robert
T. Michael, eds. Committee on National Statistics. Washington, DC:
National Academy Press.
National Research Council. 2005. Workshop on Experimental Poverty
Measures: Summary of a Workshop. Washington, D.C.: National Academy
Press.
Organization for Economic Cooperation and Development. 2001. OECD
Employment Outlook. OECD (June).
Short, Kathleen. 2001. Experimental Poverty Measures: 1999. U.S.
Census Bureau, Current Population Report, Consumer Income, P60-216.
Washington, D.C.: U.S. Government Printing Office.
Short, Kathleen. 2003. ``Material and financial hardship and
alternative poverty measures 1996.'' Paper presented at the annual
meeting of the American Statistical Association, San Francisco.
Short, Kathleen and Thesia I. Garner. 2002. ``A Decade of
Experimental Poverty Thresholds: 1990 to 2000.'' U.S. Census Bureau
Poverty Measurement Working Paper (June). http://www.census.gov/hhes/
www/povmeas/papers/decade.pdf.
Short, Kathleen, Thesia I. Garner, David Johnson, and Patricia
Doyle. 1999. Experimental Poverty Measures: 1990 to 1997. U.S. Census
Bureau, Current Population Report, Consumer Income, P60-205.
Washington, D.C.: U.S. Government Printing Office.
Smith, Adam. 1776. An Inquiry into the Nature and Causes of The
Wealth of Nations. Republished, Oxford: Clarendon Press, 1976.
U.S. Census Bureau. 2007a. ``Official and National Academy of
Sciences NAS Based Poverty Rates: 1999 to 2005.'' U.S. Census Bureau
Tables of Alternative Poverty Estimates, available at: http://
www.census.gov/hhes/www/povmeas/tables.html, retrieved July 27, 2007.
U.S. Census Bureau. 2007b. The Effect of Taxes and Transfers on
Income and Poverty in the United States: 2005. U.S. Census Bureau,
Current Population Report, P60-232. Washington, D.C.: U.S. Government
Printing Office.
Vaughan, Denton R. 2004. ``Exploring the use of the public's views
to set income poverty thresholds and adjust them over time.'' U.S.
Census Bureau Poverty Measurement Working Paper. http://www.census.gov/
hhes/www/povmeas/papers/wkppov20_cen.pdf.
Vaughan, Denton R. 1993. ``Exploring the use of the public's views
to set income poverty thresholds and adjust them over time.'' Social
Security Bulletin 56, 2 (summer): 22-46.
Chairman MCDERMOTT. Thank you, very much.
Nancy Cauthen is a Ph.D. from Columbia University from the
National Center for Children in Poverty, Mailman School of
Public Health.
STATEMENT OF NANCY K. CAUTHEN, DEPUTY DIRECTOR, NATIONAL CENTER
FOR CHILDREN IN POVERTY, MAILMAN SCHOOL OF PUBLIC HEALTH,
COLUMBIA UNIVERSITY
Dr. CAUTHEN. Good morning, and thank you Chairman McDermott
and members of the Subcommittee for this opportunity to
testify.
I want to make three points this morning. First, I want to
emphasize why the subject of this hearing--that is, how we
measure poverty in the United States--is so important. Child
and family poverty exact a high toll on our society. To reduce
both the human and societal costs of poverty, we need a better
measure than we currently have to identify who needs assistance
and what kind of assistance.
An extensive body of research has definitively linked
economic hardship to a range of adverse educational, health,
social, and emotional outcomes for children that place
constraints on their human potential and limit their future
productivity. At the same time, there is compelling evidence
that we can positively affect the developmental trajectories of
children affected by poverty, if we invest adequate resources
and proven strategies and especially if we intervene early.
If, as a nation, we decide to make a commitment to reduce
child poverty, which to date we have not done, it is imperative
that we have the right measures to identify it and quantify its
scope. My second point is that, as the two previous witnesses
have argued, that the National Academy of Science's approach
for updating the poverty measure would be a welcome improvement
over our current measure. But we need to be clear that the NAS
alternative still represents a minimal level of subsistence,
not a decent, modest, standard of living.
The NAS recommendations produce poverty thresholds that are
not substantially higher than the current thresholds as we just
saw on the charts. We are still talking about a poverty level
of roughly $20,000 to $23,000 a year for a family of four,
whether that family lives in New York City or rural Talbot
County, Georgia, for example--localities which obviously have
dramatically different living costs. Research consistently
shows that families with incomes of up to twice the official
poverty level experience many of the same material hardships as
families who are officially poor. These hardships include
things like being evicted from one's apartment, having
utilities shut off, going without needed medical or dental
care, or having unstable or unsafe childcare.
Emerging findings from research on child development
suggest that these types of material hardships are key to
understanding why poverty harms children, and I would be happy
to say more about that later.
My final point is that there has been a considerable amount
of research, especially over the last decade, about what it
really takes for families to make ends meet. Many of us have
adopted 200 percent of the Federal poverty level as a proxy for
low income, that is, a level below which families will have
difficulty meeting basic needs and will face material hardship.
But of course this level varies by region, State and locality.
Family budget research conducted by my organization and others
suggests that families need an income of anywhere between one
and a half to over three times the current poverty level to
make ends meet, and I have provided a number of specific
examples in my written testimony.
[GRAPHIC] [TIFF OMITTED] T3757A.011
But even research on family budgets makes conservative
assumptions about expenses. This is still an approach that
focuses on day-to-day needs. The work on basic budgets includes
housing, food, childcare, health insurance, transportation,
payroll and income taxes, and a small amount for other
necessities. But none of this includes resources for things
like household furnishing, a rainy day fund, disability or life
insurance, or any of the kinds of cushions that would help a
family withstand a major medical crisis, a job loss or other
financial setback.
Family budget approaches are helpful for understanding what
it takes for a family to get by but not what it takes to get
ahead.
In conclusion, I would like to suggest that the United
States, as the wealthiest nation in the world, needs a range of
measures to assess how children and families are doing that go
beyond a minimal level of subsistence. Adopting the NAS
recommendations for a revised poverty measure would be an
important and highly worthwhile first step, but it is not
enough.
Thank you.
[The prepared statement of Dr. Cauthen follows:]
Prepared Statement of Nancy Cauthen, Ph.D., Deputy Director, National
Center for Children in Poverty, Mailman School of Public Health,
Columbia University
Thank you, Chairman McDermott and members of the subcommittee for
this opportunity to testify. I'd like to begin by thanking you for
holding this hearing and addressing the important issue of how we
measure poverty.
My name is Nancy Cauthen, and I am the Deputy Director of the
National Center for Children in Poverty (NCCP). NCCP is a policy
research organization at Columbia University's Mailman School of Public
Health. Our mission is to promote the health, economic security, and
well-being of America's most vulnerable children and families. NCCP
uses research to identify problems and find solutions at the state and
national levels.
My testimony will address the following points:
Child and family poverty exact a high toll on our
society, so it is critical that we measure it in a way that allows us
to best identify who needs assistance and what kinds of assistance.
Although the National Academy of Sciences 1995
recommendations and subsequent refinements for updating the official
poverty measure offer the most promising approach, the thresholds would
still be too low to identify all those who need help.
To improve child and family well-being, we must address
not only income poverty but also material hardship.
Family budgeting approaches provide an alternative way to
understand what it takes for families to meet their basic needs and to
achieve a reasonable standard of living.
What's at Stake: Why Poverty Matters
There is now abundant evidence that not only does poverty create
hardship and adversity for those who experience it, but poverty also
exacts a high toll on our entire society. Testimony presented before
the full Ways and Means Committee in January estimated that child
poverty costs the United States $500 billion per year, which is roughly
equivalent to 4 percent of Gross Domestic Product. These costs are
attributed to reductions in productivity and economic output, increases
in crime, and increases in health expenditures (Holzer, Schanzenbach,
Duncan, and Ludwig 2007). A report prepared by the General Accounting
Office and presented at the same hearing also found that poverty has
large negative economic and social impacts (Nilsen 2007).
These and many other studies point to the seriousness of child
poverty as a longstanding, nationwide problem. Using our current
poverty measure, in 2005, 13 million children--18 percent of children
in the United States--were growing up in families with income below the
federal poverty level, which in 2007 is $17,170 for a family of three
and $20,650 for a family of four (Fass and Cauthen 2006).\1\ But as I
will argue, these figures significantly underestimate the numbers of
children living in families who struggle to make ends meet.
Considerable research indicates that it takes, on average, an income of
twice the federal poverty level to meet basic needs. Using this
definition of low income, 39 percent of children are living in families
that are struggling financially.
---------------------------------------------------------------------------
\1\ These figures refer to the federal poverty guidelines, which
are used for administrative purposes, such as determining financial
eligibility for benefit programs. For statistical purposes, researchers
use a different--but quite similar--version of the federal poverty
measure, the federal poverty thresholds, issued by the U.S. Census
Bureau. Both the guidelines and the thresholds are commonly referred to
as the federal poverty level (FPL).
---------------------------------------------------------------------------
The Effects of Income Poverty on Children
An extensive body of research literature has definitively linked
economic hardship to a range of adverse educational, health, and social
outcomes for children that limit their future productivity (for reviews
of this literature, see Gershoff, Aber, and Raver 2003; Cauthen 2002).
Poverty can impede children's cognitive development and their ability
to learn. It can contribute to behavioral, social, and emotional
problems. And poverty can contribute to poor health among children.
Research also indicates that the strength of the effects of poverty
on children's health and development depends in part on the timing,
duration, and intensity of poverty in childhood. The risks posed by
poverty appear to be greatest among children who experience poverty
when they are young and among children who experience persistent and
deep poverty. The negative effects of poverty on young children,
troubling in their own right, are also cause for concern given that
these effects are associated with difficulties later in life--teenage
childbearing, dropping out of school, poor adolescent and adult health,
and poor employment outcomes.
As discouraging as this research might be, there is compelling
evidence that we can positively affect these trajectories by investing
adequate resources in proven anti-poverty strategies. Research is clear
that we must reach children in poor families when they are very young
and simultaneously address the needs of their parents (Shonkoff and
Phillips 2000).
A holistic approach to reducing child poverty requires increasing
family incomes, improving parental employment outcomes, investing in
high-quality early care and learning experiences, and strengthening
families. I do not mean to downplay the enormity of the task (Haskins
2007)--it would require a huge financial commitment as well as
tremendous political will. But the point is simply that it's possible--
the evidence is clear that in the long term, sound investments in the
healthy development of children can increase economic productivity and
improve overall prosperity, while reducing inequality (Knitzer 2007).
Increasing Family Income Improves Child Outcomes
More than a decade of research shows that increasing the incomes of
low-income families--net of other changes--can positively affect child
development, especially for younger children (for a review, see Cauthen
2002). Experimental studies of welfare programs offer some of the
strongest evidence to date about the importance of income. For example,
welfare programs that increase family income through employment and
earnings supplements have consistently shown improvements in school
achievement among elementary school-age children; other studies have
also shown links between increased income and improved school readiness
in young children.
In contrast, welfare programs that increase levels of employment
without increasing income have shown few consistent effects on
children. Moreover, findings from welfare-to-work experiments show that
when programs reduce income, children are sometimes adversely affected.
Other studies have shown links between increased income and reductions
in behavioral problems in low-income children and youth (Costello,
Compton, Keeler, and Angold 2003). It is not just the amount of income
that matters but also its predictability and stability over time;
research has shown that unstable financial situations also can have
serious consequences for children (Wagmiller, Lennon, Kuang, and Aber
2006).
Reducing the consequences of child poverty will require more than
increasing family incomes. But too often, policy discussions about
reducing child poverty focus only on the symptoms of poverty--low
educational achievement, social and behavioral problems, and poor
health. Yet poverty itself is the single biggest threat to healthy
child development: improving child outcomes requires explicit attention
to lifting families up economically.
Determining the Best Way to Measure Poverty
For quite some time, there has been a consensus among social
scientists that the current poverty measure needs to be improved. The
United States measures poverty by a standard developed more than 40
years ago, using data from the 1950s that indicated families spent
about one-third of their income on food. The official poverty level was
set by multiplying food costs by three. Since then, the figure has been
updated annually for inflation but the methodology has otherwise
remained unchanged. The federal poverty level is adjusted by family
size but is the same across the continental U.S.
The Current Measure
The usefulness of the current measure has declined over time for
two reasons (Cauthen and Fass 2007):
1. The poverty thresholds--that is, the specific dollar amounts--
are too low because they are based on outdated assumptions about family
expenditures.
Food now comprises about one-seventh of an average family's
expenditures--not one-third as was assumed under the original poverty
measure. At the same time, the costs of housing, child care, health
care, and transportation have grown disproportionately. Thus, the
poverty level no longer reflects the true cost of supporting a family
at a minimally adequate level. In addition, the current poverty measure
is a national standard that does not adjust for the substantial
variation in the cost of living from state to state and among urban,
suburban, and rural areas.
2. The method used to determine whether a family is poor does not
accurately count family resources, overestimating resources for some
and underestimating them for others.
When determining whether a family is poor, income sources counted
include earnings, interest, dividends, Social Security, and cash
assistance. But income is counted before subtracting payroll, income,
and other taxes, overestimating how much families have to spend on
basic needs. And the method understates the resources of families who
receive some types of government assistance because the federal Earned
Income Tax Credit is not counted nor are in-kind government benefits--
such as food stamps and housing assistance--taken into account.
Thus, by not reflecting an accurate picture of family expenses and
resources, one unfortunate consequence of the way we currently measure
poverty is that the measure cannot be used to evaluate the
effectiveness of the very programs designed to help alleviate poverty.
The 1995 National Academy of Sciences Recommendations
Social scientists have been debating the usefulness of the current
poverty measure for quite some time. The most extensive effort to date
to address the concerns about the measure began with the work of a
distinguished panel of experts appointed by the National Academy of
Sciences (NAS) at the behest of Congress. In the decade since the
panel's report was released in 1995 (National Research Council 1995),
social scientists at the U.S. Bureau of the Census and the Bureau of
Labor Statistics, as well as at universities and research centers, have
continued to build on the panel's work.
To address the primary concerns about the current poverty measure,
the NAS panel recommended that:
1. The poverty threshold comprise a budget for food, clothing, and
shelter.
The amounts budgeted would be based on expenditure data, and the
figures would be updated annually. The shelter amount would include
utilities, and the threshold would allow a small additional amount for
other common needs (such as household supplies, personal care, and non-
work-related transportation). The panel discussed whether the measure
should be adjusted for regional differences in living costs. This point
has generated considerable debate and contention--the concerns are both
technical and political.
2. The measure of resources include cash and near-cash disposable
income that is available for basic needs that are common to all
families.
The resource measure would exclude certain expenses that are non-
discretionary for the families that incur them (e.g., work-related
expenses such as child care and out-of-pocket medical care expenses).
But it would include in-kind benefits (e.g. food stamps, subsidized
housing, school lunches, and home energy assistance). The measure is
calculated after taxes, so payroll taxes would be excluded, but the
Earned Income Tax Credit and other tax credits would be included in
determining family resources.
Researchers do not agree on all the specific technical aspects of
the NAS and subsequent recommendations. But there is almost universal
agreement among social scientists that the NAS recommendations would
provide the nation with a far more useful poverty measure than the
current one. And pragmatically, the NAS approach is viewed as the most
viable option for creating a bipartisan political consensus around a
new measure.
The NAS recommendations would undoubtedly be an important
improvement over what we have. And they also provide a way to measure
the impact of poverty reduction programs, most of which did not exist
when the original measure was created.
What Are We Measuring?
But even if we reach a consensus on a revised poverty measure along
the lines of the NAS recommendations--and I hope we do--we need to be
clear about what we are measuring. Both the current measure and the NAS
versions attempt to quantify a minimal level of subsistence below which
we have agreed, as a society, that no individual or family should fall.
Any judgment about what constitutes a minimally acceptable level of
subsistence is, of course, normative. Human beings can survive on a
variety of income levels. In 2005, 8 percent of children in the U.S.--
nearly 6 million children--were surviving despite living in households
with incomes of less than half the poverty line, which was just under
$10,000 annually for a family of three. Yet, in the wake of Hurricane
Katrina, many Americans seemed shocked to learn that we still have a
sizable number of desperately poor people in our country.
In short, questions about how we define poverty require value
judgments not only about how to define a minimal level of subsistence
but whether that is in fact a decent and just way to define poverty in
a wealthy society.
Implementing the NAS recommendations produces poverty thresholds
that are not vastly different from the current ones, which means they
do not reflect the substantial improvement in living standards that
have occurred in the U.S. over the last 40-plus years. When the current
poverty measure was developed, the threshold for a family of four
equaled about 50 percent of the median income for a four-person family.
But over time, that percentage has dropped dramatically. Today the
poverty threshold for a four-person family represents only about 30
percent of the median income (Ziliak 2005).
The question becomes: for what purpose are we measuring poverty and
what do we want to do with the information? One of the most compelling
reasons to establish an agreed upon measure of poverty is to identify
who in the population is in need of assistance--and what kind of
assistance--and the scope of that need. To the degree that we want a
poverty measure that can inform policy, especially with regard to
improving the well-being of children and families, we may need
different kinds of measures.
The Difference Between Poverty and Material Hardship
The current poverty line does not accurately predict the likelihood
that a family will experience material hardship (Iceland and Bauman
2007). Examples of material hardships include being evicted, missing
rent payments, having utilities shut off, going without needed medical
or dental care, or having unstable child care. Research consistently
shows that families with income of up to twice the official poverty
level experience many of the same hardships as families who are
officially poor--while families with income above twice the poverty
line are substantially less likely to experience material hardships.
Overall, about two thirds of families with income between 100 and 200
percent of the federal poverty level experience one or more material
hardships such as not having enough food or having utilities turned off
because of inability to pay bills (Boushey, Brocht, Gundersen, and
Bernstein 2001; Amey 2000). Some hardships, such as difficulties paying
for child care and health care, are common among middle-income families
as well.
A critical finding emerging from the child development literature
is that material hardships play an important role in determining
whether or not children will be negatively affected by growing up in a
low-income family. Not surprisingly, facing such hardships is
associated with diminished parental investments in children and
increased parental stress, which in turn negatively affect children
(Gershoff, Aber, Raven, and Lennon 2007). It is now clear that to
reduce the effects of poverty on children, we need to increase family
incomes and reduce the experience of material hardship (Gershoff 2003).
Any new poverty thresholds based on the NAS recommendations would
not be substantially higher than current thresholds. Alternative
poverty levels calculated by the Census Bureau that incorporate many of
the NAS suggestions indicate that the threshold for a two-parent family
with two children would increase by about $3,000 (Bernstein 2007).
Since research indicates that families with incomes of up to twice the
current poverty thresholds face high levels of material hardship, it
seems likely that even with an NAS-based alternative, there will
continue be many families who are deemed non-poor by the new measure
while not being able to meet their basic needs.
Measuring What It Takes to Make Ends Meet
There has been a considerable amount of research over the last
decade about what it takes to make ends meet. One such effort was
spearheaded by Diana Pearce, for Wider Opportunities for Women, who
developed a methodology for creating ``Self-Sufficiency Standards''
(Pearce 2001, 2006). The standards quantify how much money a family
needs to cover basic expenses, such as housing, food, child care,
transportation, health insurance, and payroll and income taxes; a small
amount is also allocated for other necessities (examples include
clothing, diapers, household items, and school supplies). The standards
vary by locality--to account for variations in the cost of living--and
by family type (two-parent or single-parent and the number and ages of
children). The budgets assume that the families receive no public
benefits. Self-Sufficiency Standards have been developed for 36 states.
The Economic Policy Institute (EPI) undertook a similar effort and
created ``Basic Family Budgets'' (Berstein, Brocht, and Spade-Aguilar
2000; Allegretto 2005). The methodology differs somewhat from that for
the Self-Sufficiency Standards, but the concept is the same--what does
it take for different types of families in different localities to
cover the costs of basic living expenses?\2\ EPI has calculated basic
budgets for over 400 localities across the country. The organization
characterizes Basic Family Budgets as providing ``a realistic measure
of the income required to have a safe and decent though basic standard
of living.''
---------------------------------------------------------------------------
\2\ Basic Family Budgets vary based on the number of children in a
family but not their ages.
---------------------------------------------------------------------------
Building on this earlier work, NCCP has created ``Basic Needs
Budgets'' for different family types in over 80 localities in 14 states
plus the District of Columbia.\3\ We developed these budgets in
conjunction with a project, Making ``Work Supports'' Work, that
analyzes the effects of federal and state work support programs--earned
income tax credits, child care and housing assistance, and food
stamps--on the ability of low-wage workers to make ends meet.
---------------------------------------------------------------------------
\3\ For a detailed description of the methodology used to create
NCCP's Basic Needs Budgets, see the User Guide for the Family Resource
Simulator and consult the section on ``Calculating Family Expenses.''
---------------------------------------------------------------------------
Despite some differences in methodology, all three of these efforts
provide additional evidence for the finding that families on average
need an income of twice the current poverty level to cover the costs of
basic expenses. NCCP has found that, depending on locality, this figure
ranges from about 150 to over 300 percent of poverty. For example,
Table 1 shows that it takes an annual income of about $30,000 for a
single-parent family with two children to make ends meet in Atlanta,
Georgia, but a similar family living in Rockville, Maryland would need
over $50,000.
[GRAPHIC] [TIFF OMITTED] T3757A.012
NCCP's Basic Needs Budgets, as well as the Self-Sufficiency
Standards and EPI's Basic Family Budgets, include only the most basic
daily living expenses and are based on modest assumptions about costs.
For example, the budgets in Table 1 assume that family members have
access to employer-sponsored health coverage when not covered by public
insurance, even though the majority of low-wage workers do not have
access to employer coverage. NCCP's Basic Needs Budgets do not include
the cost of out-of-pocket medical expenses for copayments and
deductibles, which can be quite costly, particularly for families with
extensive health care needs. The budgets do not include money to
purchase life or disability insurance or to create a rainy-day fund
that would help a family withstand a job loss or other financial
crisis. Nor do they allow for investments in a family's future
financial success, such as savings to buy a home or for a child's
education. In short, these budgets indicate what it takes for a family
to cover their most basic living expenses--enough to get by but not
enough to get ahead.
Implications
These various measures--poverty measures, measures of material
hardship, basic budgets--are not alternative ways of looking at the
same thing, but rather they provide mechanisms for capturing and
quantifying different phenomenon, which may require different (if
overlapping) policy responses. Given this, what are the implications of
adopting a new poverty measure along the lines of the NAS
recommendations?
First, we would need to acknowledge that the official poverty level
in the United States would remain a measure of deprivation and hardship
rather than a measure of a decent, if modest, standard of living. Such
a measure would still--if more accurately--identify only the most
needy. Many families above this level still need assistance.
Second, we would need to think through the implications for
programs that currently use the poverty level (or a percent of the
poverty level) to determine eligibility. One possibility is to
structure our assistance programs in ways that reflect the fact that
working families with incomes above the poverty level need assistance
with basic needs. The provision of public health insurance is one such
model to build on--for example, providing free health insurance to
families below (or near) the poverty line, and subsidized health
insurance to somewhat higher income families, with premiums and
copayments that gradually rise with family income. Similarly, a child
care program informed by this understanding might provide free or very
low-cost care to families living below the poverty line and reduced-
cost child care to those above poverty but below a basic budget level
(with the government subsidy decreasing as income increases). Cash
assistance programs, on the other hand, would remain targeted at
officially poor families, who have very low (or no) earnings (most
state eligibility limits for cash assistance are currently well below
the poverty level).
Third, I would hope that adopting the NAS approach for measuring
income poverty would be accompanied by government efforts to also
measure hardship, asset poverty, and other measures that inform us
about how families are doing. Too many of our current policies are
``too little, too late.'' We typically wait until children and families
are in deep trouble before we assist them, rather than investing
heavily in prevention--we should help all families succeed instead of
trying to patch them up once they have fallen. But we will need better
measures--and concepts broader than poverty--to do so.
References
Allegretto, Sylvia A. 2005. Basic Family Budgets: Working Families'
Incomes Often Fail to Meet Living Expenses Around the U.S. Washington,
DC: Economic Policy Institute.
Amey, Cheryl. 2000. Families Struggling to Make it in the
Workforce: A Post Welfare Report. Washington, DC: Children's Defense
Fund.
Bernstein, Jared. 2007. More Poverty than Meets the Eye (Economic
Snapshots April 11, 2007). Washington, DC: Economic Policy Institute.
Bernstein, Jared, Chauna Brocht, and Maggie Spade-Aguilar. 2000.
How Much is Enough? Basic Family Budgets for Working Families.
Washington, DC: Economic Policy Institute.
Boushey, Heather, Chauna Brocht, Bethney Gundersen, and Jared
Bernstein. 2001. Hardships in America: The Real Story of Working
Families. Washington, DC: Economic Policy Institute.
Cauthen, Nancy K. 2002. Policies that Improve Family Income Matter
for Children. New York: National Center for Children in Poverty,
Mailman School of Public Health, Columbia University. Available at:
http://nccp.org/publications/pub--480.html.
Cauthen, Nancy K. and Sarah Fass. 2007. Measuring Income and
Poverty in the United States. New York: National Center for Children in
Poverty, Mailman School of Public Health, Columbia University.
Available at: http://nccp.org/publications/pub--707.html.
Costello, E. J, Compton, S., Keeler, G., Angold, A. 2003.
``Relationships between poverty and psychopathology: A natural
experiment.'' Journal of the American Medical Association, 290(15),
2023-2029.
Fass, Sarah and Nancy K. Cauthen. 2006. Who Are America's Poor
Children: The Official Story. New York: National Center for Children in
Poverty, Mailman School of Public Health, Columbia University.
Available at: http://nccp.org/publications/pub--684.html.
Gershoff, Elizabeth T. 2003. Low Income and Hardship Among
America's Kindergartners. New York: National Center for Children in
Poverty, Mailman School of Public Health, Columbia University.
Available at: http:// /nccp.org/publications/pub--530.html.
Gershoff, Elizabeth T., J. Lawrence Aber, and C. Cybele Raver.
2003. Child poverty in the United States: An evidence-based conceptual
framework for programs and policies. In F. Jacobs, D. Wertlieb, & R. M.
Lerner (Eds.), Handbook of Applied Developmental Science: Promoting
Positive Child, Adolescent, and Family Development through Research,
Policies, and Programs (Vol. 2, pp. 81--136). Thousand Oaks, CA: Sage.
Gershoff, Elizabeth T., J. Lawrence Aber, C. Cybele Raver, and Mary
Clare Lennon. 2007. ``Income is not enough: Incorporating material
hardship into models of income associations with parent mediators and
child outcomes.'' Child Development, 78, 70-95.
Haskins, Ron. 2007. Alleviating Child Poverty in the Long Run.
Testimony before the Chairman, Committee on Ways and Means, House of
Representatives. January 24, 2007.
Holtzer, Harry, Diane W. Schanzenbach, Gregory J. Duncan, and Jens
Ludwig. 2007. The Economic Costs of Poverty in the United States:
Subsequent Effects of Children Growing Up Poor. Washington, DC: Center
for American Progress.
Iceland, John and Kurt Bauman. 2007. ``Income poverty and material
hardship: How strong is the association?'' The Journal of Socio-
Economics 36, 376-396.
Knitzer, Jane. 2007. Testimony on the Economic and Societal Costs
of Poverty. Testimony before the Chairman, Committee on Ways and Means,
House of Representatives. January 24, 2007. Available at: http://
nccp.org/publications/pdf/text--705.pdf.
National Research Council. 1995. Measuring Poverty: A New Approach.
Panel on Poverty and Family Assistance, Constance F. Citro and Robert
T. Michael, eds. Committee on National Statistics. Washington, DC:
National Academy Press.
Nilsen, Sigrid R. 2007. Poverty in America: Consequences for
Individuals and the Economy. Testimony before the Chairman, Committee
on Ways and Means, House of Representatives. Washington, DC: Government
Accounting Office.
Pearce, Diana M. 2001. ``The Self-Sufficiency Standard: A new tool
for evaluating anti-poverty policy.'' Poverty & Race, 10 (2).
Pearce, Diana M. 2006. The Self-Sufficiency Standard for
Pennsylvania. Holmes, PA: PathWaysPA.
Shonkoff, Jack P. Deborah A. Phillips. (Eds.). National Research
Council & Institute of Medicine. 2000. From Neurons to Neighborhoods:
The Science of Early Childhood Development. Washington, DC: National
Academies Press.
Wagmiller, Robert., Mary Clare Lennon, Li Kuang, Philip Alberti,
and J. Lawrence Aber. 2006. ``The dynamics of economic disadvantage and
children's life chances.'' American Sociological Review, 71(5): 847-
866.
Ziliak, James. 2006. ``Understanding poverty rates and gaps:
Concepts, trends, and challenges.'' Foundations and Trends in
Microeconomics, 1(3), 127-199.
Chairman MCDERMOTT. Thank you very much for your testimony.
Douglas Besharov is the Joseph and Violet Jacobs Scholar in
Social Welfare Studies at the American Enterprise Institute.
Mr. Besharov?
STATEMENT OF DOUGLAS J. BESHAROV, JOSEPH AND VIOLET JACOBS
SCHOLAR, SOCIAL WELFARE STUDIES AT THE AMERICAN ENTERPRISE
INSTITUTE
Mr. BESHAROV. Chairman McDermott, Mr. Weller, and other
Members of the Committee, it is a pleasure to be here.
My message, reluctantly, is to tell you why you should not
change the official measure, and in fact why you cannot do it.
I come to this conclusion after having gone through a twelve-
month process of exploring all sorts of options with a blue
ribbon group, government officials and outside experts.
In my written testimony, I go through the flaws of the
current official measure, about which you have heard a great
deal. You have mentioned them as well: it does a poor job
accounting for inflation; does not count market income very
well; and it does not take wealth into consideration. (That
$750,000 townhouse doesn't even fit in the measure.) It doesn't
subtract taxes. It doesn't account for changing household
composition or changing consumption patterns which you also
mentioned, Mr. Chairman. It doesn't deal with growing national
affluence. It doesn't even try to deal with geographic
differences in expenditures and costs.
Finally, and most importantly, it doesn't give credit for
our large, maybe insufficient, efforts to alleviate poverty.
First, I think you should not change the official poverty line.
My testimony opens with Ronald Reagan's famous quip, ``We
declared war on poverty and poverty won.'' That's only possible
because the current measure doesn't count all the things we do
for low-income America.
Our committee examined two papers that were quite decisive
in our thinking. One was by Christopher Jencks from Harvard
University. His work, like that of a number of other
specialists, found that the material condition of the poor has
improved tremendously over the last 30 years.
We have not talked about how people are actually doing, but
that's tremendously important. Another paper, which is from the
Democrats on the Joint Economic Committee of 2 years ago,
analyzes the kinds of changes to the poverty measure that many
people have proposed, which suggests that progress and poverty
reduction over the past two decades is much greater than the
official poverty measure would indicate. Anti-poverty programs
like the EITC, combined with changing family formation
patterns, rising teen birthrates, and increases in
cohabitation, resulted in significant decreases in poverty.
My point is not that we shouldn't do everything else that
has been suggested, but it is essential, as a matter of policy,
to be able to understand and accept the credit that the tens of
billions of dollars of Federal aid now make for low-income
people; the analysis from the Joint Economic Committee and
others--on the left and the right--is that we ought to give
credit to past efforts, and raising the thresholds without
doing anything else negates that possibility.
That is why I think you shouldn't change the official
measure. One quick mention about relative measures of poverty;
Western Europe is now in the throes of reconsidering relative
measures, because as an economy gets richer, and when income
dispersion increases as it did in Ireland and is now happening
in England and in Germany, poverty rises even though the
incomes of people at the bottom are also increasing. I would
say this is why there should be many measures. You can have a
relative measure but you have to couple it with an absolute
measure.
You can potentially have a measure about economic
marginalization. You just cannot use it as the measure for
determining eligibility for food stamps, because I think that
we are talking about the true economic margin and
marginalization at $40,000 a year of income. I wouldn't suggest
food stamps for that family. Median family income is about
$49,000, so this family is pretty close to the median.
Before I run out of time, allow me to suggest why you
cannot, in my opinion, change the official measure. I say this
with all due respect, we thought this through a lot, because we
were eager to have a success here. First, technically, there is
absolutely no agreement about how to correct or geographic
differences in the cost of living. Just to give you an example,
the one attempt that was made by the Census Bureau shows that
the poverty rate in California would go from 13.2 percent to
17.8 percent.
There is a vast argument about how to value housing. There
also are large conceptual problems within the NAS approach.
Some people think that subtracting the cost of transportation
from someone's income who lives in West Virginia, and commutes
to Washington so that that family can live in a bigger home
than those of us closer to the city is not a way to measure
poverty. That's also not involved in the Federal poverty
measure. People want to account for the cost of child care, but
rarely want to count the $24 billion we spend a year on child
care subsidies.
I mention these things, not because I think there is a
solution, but because I think there is no solution to the
technical and conceptual problems of changing the poverty
measure in a way that it can be used for setting eligibility
for programs, and that's the key point I want to make here.
Today's debate in the House is about multiples of the
poverty line, not because Congress thinks that the poverty line
itself is magical, but because the Congress, in its political
wisdom, is finding a more absolute number within that.
I have clearly run out of time here. I would be delighted
to answer any questions.
[The prepared statement of Mr. Besharov follows:]
[GRAPHIC] [TIFF OMITTED] T3757A.013
[GRAPHIC] [TIFF OMITTED] T3757A.014
[GRAPHIC] [TIFF OMITTED] T3757A.015
[GRAPHIC] [TIFF OMITTED] T3757A.016
[GRAPHIC] [TIFF OMITTED] T3757A.017
[GRAPHIC] [TIFF OMITTED] T3757A.018
[GRAPHIC] [TIFF OMITTED] T3757A.019
[GRAPHIC] [TIFF OMITTED] T3757A.020
[GRAPHIC] [TIFF OMITTED] T3757A.021
[GRAPHIC] [TIFF OMITTED] T3757A.022
[GRAPHIC] [TIFF OMITTED] T3757A.023
[GRAPHIC] [TIFF OMITTED] T3757A.024
[GRAPHIC] [TIFF OMITTED] T3757A.025
[GRAPHIC] [TIFF OMITTED] T3757A.026
[GRAPHIC] [TIFF OMITTED] T3757A.027
[GRAPHIC] [TIFF OMITTED] T3757A.028
[GRAPHIC] [TIFF OMITTED] T3757A.029
Chairman MCDERMOTT. Thank you. Mark Greenberg is the
Executive Director of the taskforce on poverty, Center of
American Progress Institute.
Mark?
We will probably have another vote, and then we'll have a
little bit, and so understand, we are going to get up and go.
Whatever you have got to say, say it up front.
STATEMENT OF MARK GREENBERG, DIRECTOR, TASK FORCE ON POVERTY,
CENTER FOR AMERICAN PROGRESS INSTITUTE
Mr. GREENBERG. We are aware and, Mr. Chairman and members
of the Subcommittee, first let me thank you for holding this
hearing, for the other hearings you have held over the course
of the year in efforts to bring renewed attention to the
importance of addressing poverty in the United States.
For purposes of my oral testimony right now, I want to do
three things. I want to briefly focus on why, in my view, it is
important to develop and improve measure of poverty. I want to
talk about some principles to guide the process and then make
some suggestions for how you might move forward.
As indicated, I have been serving as the Director of a task
force on poverty for the Center for American Progress, and we
actually spent the better part of the year working on how to
make the case for why the nation should address poverty and
making recommendations for what should be done about it. In the
course of that, we did a lot of consultation around the country
and as we did so, our principal focus was not on the measure of
poverty, but a couple of things kept coming up. First, when we
would talk to people around the country about approaches to
addressing poverty, the recurrent theme is that it is
problematic to start with the existing measure, because the
existing measure is rarely useful as a way of measuring need,
because it is so low in relation to the cost of living.
The existence of things like family budgets, of self-
sufficiency standards, of setting program eligibility as the
multiple of the poverty line, of a set of research that
increasingly treats 200 percent of poverty as a measure of low
income. All those are different ways in which people are saying
we need a better measure to reflect what it costs to make ends
meet, that the poverty line simply does not reflect it.
It is also true that it is a problem in my view that the
existing measure does not well pick up the effects of policy.
When we looked at a set of policies for improving family well-
being, expanding the earned income tax credit, raising the
child tax credit, improving childcare assistance, increasing
housing subsidies, in each case we faced the issue that if you
make a set of changes which will clearly improve well-being, it
is not going to be picked up in the official measure. It is, in
my view, important that we have an effective measure that both
looks at what families need to get by and then does effectively
measure the resources that are available to them to do it.
That really is crucial as a means of developing policies
for moving ahead. Now, I also think that the case for
developing a better poverty measure had been made in a
thoughtful and compelling way by the National Academy of
Sciences panel a dozen years ago, and when they did so, I think
a central insight from their approach is that in thinking about
how to address poverty, the first question should not be what
income do we count, but rather, what is it that we are trying
to measure.
The poverty line needs to be something where we can
articulate what it is trying to measure. It needs to be a
reasonable one, something that is understandable to the public,
something that is broadly acceptable, and the current measure
fails to meet that standard. It does so for many of the reasons
Mr. McDermott talked about in his opening, that when one tries
to explain to anyone that this was a measure developed in the
early 1960s, based upon a food plan at that time multiplied by
three, and we have only since then adjusted for prices, people
do not understand how that is a useful way of thinking about
what it is that families need to get by. So, we need a better
measure to do it.
The National Academy of Sciences also focused on what they
refer to as internal consistency in thinking about the measure,
and essentially what they meant by that is that you want to
start by what it is that you are trying to measure, then
develop a measure of it. That if you are trying to measure the
cost, if you think what should be in the poverty measure is a
set of basic needs, articulate what those are, determine what
the costs of those are, count the resources that are available
to meet them, don't count the resources that are not available
to meet them.
In my view, that is a thoughtful, reasonable way to think
about moving ahead. There can certainly be and indeed there are
real disagreements about what we should think about as the
basic need. In the case of the National Academy of Sciences
panel, the approach they took was to essentially say we should
look at food, clothing, shelter, and a little more. I think a
compelling case can be made, but that is an inadequate way of
looking at basic needs, that we need to have a broader sense of
it.
That it is not that everything beyond that is. If we are
thinking about what it is that children need to grow up in a
healthy, developmentally appropriate way to be contributing
members of society, we all aspire to more than food, clothing
and shelter. So it makes sense to build that in. Similarly, if
we think about the recognized importance of people being able
to save for education or homeownership, for retirement, or for
the future needs of children, we would want to build in things
like that. So, a whole set of things ought to be considered,
but I would most emphasize that I think the NAS overall
framework offers a useful starting point for moving forward.
So, let me just quickly say, in terms of a concrete
recommendation for the panel, while the NAS has a useful
approach, it clearly is a starting point. There are a lot of
both conceptual and technical issues still to be resolved. So,
I would make two broad recommendations to you. One is that I
believe it would be useful to direct the Census Bureau to
report back to you, and in that report for them to address to
what extent it is feasible to replace the existing measure with
something using the NAS as a starting point.
Second, what are the kinds of data limitations that affect
the ability to do that and the ability to measure poverty well.
What would need to be done to address them? What would it
cost to do that? So, that Congress can have that in front of
you; and, then, because it has been a dozen years since the NAS
report, what does the research and experimentation and analysis
since then tell us about whether particular recommendations
ought to be reconsidered or new ones built in.
My final point then is, as several other panelists have
suggested, it is important to have additional measures. No
single measure is going to reflect everything that we would
want to know. Just as we have a poverty measure, it would be
useful to develop this broader measure of a make ends meet or a
decent living standard. It might be useful to have the measure
of relative poverty in relation to median income. It would be
useful to have a measure of asset poverty.
[The prepared statement of Mr. Greenberg follows:]
Prepared Statement of Mark Greenberg, Executive Director, Task Force on
Poverty, Center for American Progress
Mr. McDermott and Members of the Subcommittee:
Thank you for holding this hearing and others this year, bringing
renewed attention to the importance of addressing poverty in America.
In this testimony, I will provide some brief background, and then
discuss why the method for measuring poverty should be updated, some
principles that should guide the effort and recommendations to move the
process forward.
I am the director of the Task Force on Poverty at the Center for
American Progress (CAP), a non-profit, non-partisan public policy think
tank in Washington, D.C. I am on leave from the Center for Law and
Social Policy, where I was the Director of Policy. CAP's fourteen-
member Task Force\1\ was charged with making a case for why the nation
should address poverty and proposing a strategy for how to do so. In
April, CAP's Task Force released its report, From Poverty to
Prosperity: A National Strategy to Cut Poverty in Half.\2\
---------------------------------------------------------------------------
\1\ Task Force members were Angela Glover Blackwell, Founder and
CEO, PolicyLink (co-chair); Peter B. Edelman, Professor of Law,
Georgetown University (co-chair); Rebecca Blank, Dean, Gerald R. Ford
School of Public Policy, Henry Carter Adams Collegiate Professor of
Public Policy, University of Michigan; Linda Chavez-Thompson, Executive
Vice President, AFL-CIO; Reverend Dr. Floyd H. Flake, President,
Wilberforce University; Wizipan Garriott, Law Student and Board
President of the He Sapa Leadership Academy; Maude Hurd, National
President, ACORN; Charles E. M. Kolb, President, Committee for Economic
Development; Meizhu Lui, Executive Director, United for a Fair Economy;
Alice M. Rivlin, Senior Fellow and Director, Greater Washington
Research Program, Brookings Institution; Barbara J. Robles, Associate
Professor, Arizona State University; Robert Solow, Professor Emeritus,
Massachusetts Institute of Technology; Dorothy Stoneman, Founder and
President, YouthBuild USA; and Wellington E. Webb, Former Mayor of
Denver.
\2\ Task Force on Poverty, From Poverty to Prosperity: A National
Strategy to Cut Poverty in Half (Center for American Progress, April
2007), available at http://www.americanprogress.org/issues/2007/04/pdf/
poverty_report.pdf.
---------------------------------------------------------------------------
Our Task Force's principal focus was not on the definition of
poverty, but rather strategies for addressing it. Nevertheless, the
question of how poverty should be defined came up repeatedly in our
efforts, in two significant and related ways.
First, when seeking the views of state and local actors
about strategies to reduce poverty, one of the most common initial
observations was that it was rarely useful to use the official poverty
line as a measure of need, because it was so low in relation to living
costs. In recent years, the increased reliance on approaches like self-
sufficiency standards, family budgets, and setting program eligibility
at some multiple of the poverty line is a direct response to concerns
that the poverty line simply doesn't adequately reflect the amounts
that families need in order to get by.
Second, as our Task Force considered policy responses to
reduce poverty, we faced, in practical terms, an issue that is
routinely recognized in the academic discussions of poverty
measurement. Many initiatives that would clearly improve economic well-
being for low-income families would have no effect on poverty under
official measures, because the official measure does not count the
effects of tax policy and near-cash benefits or adjust for work-related
costs. For example, expanding the Earned Income Tax Credit or Child Tax
Credit would not reduce the official poverty rate (except indirectly if
it affected employment), even though it would increase family
resources. Expanding child care assistance would not reduce the
official poverty rate (except by raising employment) even though it
would defray costs that families face in going to work. Expanded
housing subsidies or improved food stamp participation rates would also
not affect the official poverty rate.
We ultimately addressed the first issue by emphasizing in our
report that while 37 million Americans were living in poverty, a far
larger group faced the challenge of making ends meet, and by developing
policy proposals that were sensitive to and grounded in this reality.
We addressed the second issue by using a modified measure of poverty
when calculating the poverty reduction effects of our proposals,
drawing upon recommendations from the National Academy of Sciences'
Panel on Poverty and Family Assistance: Concepts, Information Needs and
Measurement Methods in Measuring Poverty: A New Approach (National
Research Council, 1995). This modified measure counted the effects of
tax policy, treated food stamps and housing benefits as income, and
deducted out-of-pocket child care expenses from income. Only in doing
so could one fully see the real effects of a set of policies in
improving family well-being. At the same time, we could not readily
incorporate every NAS recommendation into our analysis, and only
adjusted poverty thresholds to the extent necessary to begin our
analysis with the same number of people in poverty as would be the case
under official measures.\3\ Our experience underscored the need for the
federal government to improve and modernize the definition of poverty,
in order to develop both more realistic thresholds, a better measure of
resources, and a more effective way to gauge the effects of government
policies.
---------------------------------------------------------------------------
\3\ For a detailed discussion of the methodology, see Linda
Giannarelli, Laura Wheaton, and Joyce Morton, Estimating the Anti-
Poverty Effects of Changes in Taxes and Benefits with TRIM3 (Urban
Institute, April 25, 2007), available at http://www.urban.org/
publications/411450.html.
---------------------------------------------------------------------------
While my principal focus in this testimony is on the need to
improve the poverty measure, I want to begin by emphasizing that we get
much valuable information from the current one. The current measure is
a useful and reliable indicator of the extent of serious deprivation,
and of the extent of disparities across races, sex, and ages, workers
and non-workers, and other groups. Most importantly, year-to-year
changes help us understand whether more or fewer families are
struggling to get by. Alternative measures--including those based on
the National Academies of Sciences recommendations--show different
poverty levels, but typically reflect quite similar trends because the
largest sources of income and, thus, the largest ``driver'' of poverty
rates will be cash income from sources that are included in the
official measure.
I believe the poverty measure can be significantly improved. Still,
the shortcomings of the current measure should not be used to dismiss
the information provided by the current poverty measure about the state
of our nation.
Why Should the Measure of Poverty be Updated?
There are few, if any, defenders of the current poverty measure. It
remains in place for two principal reasons. First, there are a host of
genuinely difficult conceptual and technical issues to be resolved in
determining how poverty should be measured. Second, adopting any
alternative measure is fraught with political controversy because it
will likely result in either more or fewer people reported as ``poor''
(either immediately or in the long run); greater or lesser measured
poverty rates for particular demographic, racial, and geographical and
other subgroups; and uncertain implications for determining eligibility
and distributing funds for individuals, localities, and states.
Given these challenges, why is it important to update the measure
of poverty?
No single statistic can capture every dimension of need,
consumption, hardship or well-being. An income statistic measures, at
best, income, but not how that income is spent; what would have
happened if it had been spent differently; or whether a family has
greater or lesser needs due to particular individual, family,
neighborhood, or regional factors. An income statistic may provide
little or no insight into factors that affect current and future well-
being such as health, education, social and family relationships,
community conditions, and others. Moreover, any time a line is drawn,
differences between those slightly below and slightly above the line
may be minimal or non-existent. And, even families with identical
incomes and needs may be very differently situated based on the
presence or absence of assets, which are reflected at best only
indirectly through any income-based poverty measure.
Moreover, important dimensions of need cannot be measured by income
alone. Policy efforts in Europe often situate their discussions of
income poverty within a broader context of social inclusion, a term
used to encompass concern about those outside the social mainstream who
are unable to fully participate in the normal activities of citizens.
The idea of social inclusion emphasizes that integration of people into
the social mainstream calls for addressing the range of issues that
prevent full participation in society, and that this necessarily calls
for going beyond a narrow focus on income.
Accordingly, the poverty measure cannot and should not be the sole
measure of need or well-being, but it is important. Research
commissioned for CAP's Task Force on Poverty found that the cost to the
U.S. economy from children growing up in persistent poverty is in the
range of $500 billion a year.\4\ The poverty measure provides a broad
picture of the number and characteristics of Americans who are living
with incomes below a level generally recognized as inadequate to meet
crucial needs. Moreover, ideally the poverty measure would show the
extent of deprivation before and after taxes and government transfers,
so that there is a clear picture of the extent to which government
policy is or is not reducing deprivation. As such, it is in all of our
interests to have a better measure than the current one.
---------------------------------------------------------------------------
\4\ Harry Holzer, Diane Whitmore Schanzenbach, Greg J. Duncan, and
Jens Ludwig, al., The Economic Costs of Poverty: Subsequent Effects of
Children Growing Up Poor (Center for American Progress, January 24,
2007), available at http://www.americanprogress.org/issues/2007/01/pdf/
poverty_report.pdf.
---------------------------------------------------------------------------
What are the principal problems with the current measure? Many of
the difficulties were catalogued in the thoughtful and balanced 1995
report of the National Academy of Sciences panel, Measuring Poverty: A
New Approach. Among the concerns identified by the NAS panel:
the current poverty thresholds cannot be justified as
reflecting contemporary costs for meeting basic needs;
the poverty measure does not reflect the costs of child
care and other work-related expenses;
it does not reflect regional cost variations;
it does not reflect that funds spent to meet health care
costs are not available to meet other needs;
it does not reflect that funds spent to meet child
support obligations are not available to meet other needs;
it does not reflect the impact of taxes; and
it does not reflect the provision of near-cash benefits
such as food stamps and housing assistance.
Sometimes in discussing poverty measurement, an individual may
focus on one particular problem with the measure--e.g., that it is too
low, or does not count taxes or near-cash benefits. An important
insight from the NAS panel is the need to look at the issues together
using an internally consistent approach to measurement. A poverty
measurement effort should be able to articulate what it is seeking to
measure, and its thresholds and rules about which resources are counted
should be consistent with each other and the underlying purpose. For
example, if the goal is to measure whether families have sufficient
resources to meet their food, clothing, and shelter needs, then it
makes sense to set a threshold that reflects the resources needed to do
so, to count resources that are available to meet the needs, and not
count as resources items that are not available to meet those needs.
But if, for example, the threshold is not constructed to include the
amounts needed to pay for medical costs, child care, and work expenses,
then the amounts families must pay for those costs should not be
counted as available to meet other needs. Alternatively, if the goal is
to measure whether families have resources to meet a broader set of
needs, then the thresholds and counting rules should be constructed
consistent with that intent.
Thus, a fundamental problem with the current measure is that it is
not clear what it seeks to measure, the thresholds are not based on the
actual costs of meeting a set of needs in today's economy, and it
brings no consistent approach to when income is counted or excluded in
the measurement. It fails to count resources that are available to meet
basic living costs, and yet counts resources that are not available to
meet basic living costs. The result is a framework that distorts our
understanding of when families are in need, and impairs our ability to
see whether government efforts to provide assistance are improving
family well-being.
What Should the Poverty Line Measure?
The key insight offered by the NAS should be the starting point for
any discussion of measuring poverty: before asking what should be
counted as income, one should begin by asking what the poverty line
seeks to measure. Then, decisions about how thresholds are set and
which resources are included or excluded should be made in a manner
consistent with the decision about what's being measured.
In a broad sense, the current thresholds are often viewed as being
the levels of income that families need in order to meet their most
basic needs. However, the actual dollar figure for the current
thresholds is essentially an arbitrary figure: it reflects an early-
1960s calculation of the cost of a low cost food plan designed for
temporary or emergency use when funds are low, multiplied by three
because food represented about one-third of a family budget in 1955,
and then essentially adjusted only for changes in the consumer price
index. Since that time, there have been dramatic changes in family
budgets and living standards that are not reflected in the measure. As
such, there is no real justification for the current thresholds other
than they continue a historical series and there is not agreement on
what should replace them.
Since the poverty line has only been adjusted to reflect changes in
prices since the 1960s, it has fallen over time in relation to family
median income. The poverty threshold for a family of four was about 49
percent of median income for a family of four in 1959; it was 28.4
percent of median income for a family of four in 2005. Thus, having
income below the poverty line now means that a family is much further
from the mainstream than was the case in earlier decades. Notably,
international comparisons often measure poverty in relation to 50
percent of median income. In the United Kingdom's commitment to end
child poverty by 2020, a principal measure is the share of children in
families below 60 percent of median income.
Evidence from multiple sources suggests that a substantially higher
figure would be used if the goal were to determine the amount that a
family needs to ``get by'' or ``make ends meet.'' The NAS' report
expressly recognized that by 1992, the amount that survey respondents
estimated a family needed to ``get along'' in the community was 76
percent higher than the poverty level. In recent years, a number of
groups have developed various family budgets, generally intended to
reflect a level at which a family can ``make ends meet'' or live
decently. While methodologies differ, the analyses typically find that
the average amount needed to attain such a standard is roughly twice
the current poverty line, with significant regional variation. For
example, the Economic Policy Institute has calculated basic family
budgets for over 400 communities, intended to reflect the income a
family needs to secure safe and decent-yet-modest living standards in
the community. The budget items that are included in the basic family
budgets are: housing, food, child care, transportation, health care,
other necessities, and taxes. EPI concluded that the range of basic
family budgets for a two-parent, two-child family was $31,080 (rural
Nebraska) to $64,656 (Boston, Massachusetts). The median family budget
of $39,984 contrasted with the $19,157 poverty threshold for this size
family.\5\ Wider Opportunities for Women has worked with states,
localities, and community groups in most states to develop self-
sufficiency standards. Self-sufficiency standards are calculated using
a standard methodology that considers the costs of food, housing,
medical care, transportation, child care, miscellaneous costs, and
taxes. These studies routinely find that the amount a family needs to
meet basic costs under such a budget is at least twice the federal
poverty line.\6\ The National Center for Children in Poverty's
estimates, using its Basic Needs Budget Calculator, are that across the
country, families on average need an income of about twice the official
poverty level, or roughly $40,000 for a family of four, to meet basic
needs. In a high-cost city like New York, the figure is over $50,000,
whereas in rural areas, the figure is in the low $30,000s.\7\
---------------------------------------------------------------------------
\5\ Sylvia Allegretto, Basic family budgets: Working families'
incomes often fail to meet living expenses around the U.S. (Economic
Policy Institute, September 1, 2005), available at http://www.epi.org/
briefingpapers/165/bp165.pdf.
\6\ Wider Opportunities for Women, Setting the Standard for
American Working Families (2003), available at http://
www.wowonline.org/docs/FINAL_FESS_report_072103.pdf.
\7\ National Center for Children in Poverty, Measuring Income and
Poverty in the United States, (April 2007), http://www.nccp.org/
publications/pdf/text_707.pdf.
---------------------------------------------------------------------------
Recent public opinion surveys also repeatedly find the public
estimates that amounts needed to live decently are substantially higher
than the poverty level. Note that surveys use a range of wording which
could affect survey results:
In a 2004 survey by Corporate Voices for Working
Families, 59 percent of respondents thought a family of four needed to
earn at least $40,000--an amount over twice the federal poverty line--
``to support a family of four at a decent level.'' Only one percent
thought that income of $15,000 to $20,000 was sufficient to do so.\8\
---------------------------------------------------------------------------
\8\ Corporate Voices for Working Families Survey, (July-August
2004).
---------------------------------------------------------------------------
A 2006 survey for the Catholic Campaign for Human
Development reported that most (55 percent) respondents thought the
amount of income a family of four needed to meet basic needs was
$36,000 or more; 11 percent thought it was $20,000 or less.\9\
---------------------------------------------------------------------------
\9\ Market Research Bureau LLC, Poverty Pulse, Wave VII, (Catholic
Campaign for Human Development, January 2007), available at http://
www.usccb.org/cchd/PovertyPulseVII.pdf.
---------------------------------------------------------------------------
In an April 2007 survey conducted for Northwest Areas
Foundation, 69 percent of a national sample indicated that a family of
four needed to earn $40,000 or more ``in order to make ends meet'' in
their community. In state samples, the percentage indicating $40,000 or
more was needed was 55 percent in Idaho; 57 percent in Iowa; 66 percent
in Minnesota; 51 percent in Montana; 54 percent in North Dakota; 62
percent in Oregon; 50 percent in South Dakota; and 71 percent in
Washington. Seven percent of national respondents, and 5 to 16 percent
of state respondents thought $20,000 or less was sufficient to ``make
ends meet.''\10\
---------------------------------------------------------------------------
\10\ Lake Research Partners, Survey Conducted for Northwest Areas
Foundation, (April 2007), available at http://programs.nwaf.org/pr/
nwaf/info/document/NWAF_topline_natl_and_states.pdf.
Note that the public appears to draw a distinction between the
amounts needed to make ends meet and the appropriate level for a
poverty line. In 2001, an NPR/Kaiser/Kennedy School poll asked what
income level would make a family of four poor. In that year, when the
poverty threshold for four was $17,960, most respondents would use a
higher figure, with 64 percent considering a family with earnings of
$20,000 to be poor, and a substantial group (42 percent) considering a
family with earnings of $25,000 to be poor.\11\
---------------------------------------------------------------------------
\11\ See http://www.npr.org/programs/specials/poll/poverty/
staticresults.html.
---------------------------------------------------------------------------
Thus, if the goal were to set a line for the level at which
families could ``live decently'' or meet all basic needs, it seems
clear that the poverty line would be set substantially higher than the
current one. The NAS took a far more modest approach. It proposed that
the poverty thresholds should represent a budget for food, clothing,
shelter (including utilities) and a small additional amount for other
needs, e.g., household supplies, personal care, non-work-related
transportation. It then proposed that family resources be defined as
the sum of money income from all sources and the value of near-money
benefits that are available to buy goods and services in the budget,
minus expenses that cannot be used to buy these good and services.
Thus, it did not recommend comparing all income to this more modest
level--rather, it recommended excluding income that that was not
available to purchase these basic goods and services, e.g., child care
and work-related expenses, medical expenses, child support paid, and
taxes.
Given the modest approach taken by the NAS panel, the result is a
set of poverty thresholds that are somewhat higher than the official
ones, but far short of the higher figures discussed above.
Specifically, in 2005, when the official threshold for four was
$19,806, the NAS-based threshold calculated by the Census Bureau was
$20,708 if medical expenses were not counted in the threshold and
$22,841 if medical expenses were included in the threshold.\12\ The
threshold with medical expenses was based on three years of consumer
expenditure data, updated to the threshold year, representing about 80
percent of median family expenditures for food, clothing, shelter,
utilities, and medical care, with an additional allowance for other
non-work related expenditures. The approximate breakdown for the
components of the threshold was $6624 for food, $1370 for clothing,
$6395 for shelter, $3198 for utilities, $1599 for medical costs, and
$3655 for all other non-work-related costs.
---------------------------------------------------------------------------
\12\ U.S. Census Bureau, Poverty Thresholds for Two-Adult-Two-Child
Family Following NAS Recommendations: 1999-2005, available at http://
www.census.gov/hhes/www/povmeas/altmeas05/nas_povmeasures2005.xls
---------------------------------------------------------------------------
Is the NAS approach a reasonable one? In my view, it is, but only
if one recognizes that it does not purport to represent a ``decent''
living standard. It leaves out a considerable amount that most of us
take for granted in our daily lives, and allows only a modest residual
sum for the wide array of living costs that do not fall within the
identified needs. It leaves out many of the cultural enrichment
activities that parents would view as essential to healthy child
education and development. The approach was developed at a time when
the Internet was first coming into our awareness, so does not include
costs for a family to own a computer or have Internet access. It does
not explicitly provide room in the budget to save for education, or
retirement, or home ownership, or future needs of children.
Moreover, the original NAS approach opted to treat medical costs,
child care, and other work-related costs as deductions, viewing them as
not available to meet the listed basic needs. An argument can be made
that the costs of each should be included as part of the threshold
instead and treated as comparable to other basic needs--certainly, many
Americans would view access to needed child care and health care as
basic needs. The Census Bureau now calculates NAS-based thresholds with
and without medical expenses. Thus, other approaches could be
considered, but any approach should ensure that it provides the
internal consistency urged by the NAS.
Measuring Poverty Under NAS Measures
Since the release of the NAS report in early 1995, a considerable
amount of research and further study has been undertaken. While there
may never be unanimity on any complex or controversial issue, many
observers have commended the overall NAS approach, but there are still
a number of questions on which researchers disagree, sometimes for
conceptual reasons and sometimes because of data limitations. For
example, many, though not all, would agree that it would be desirable
to build geographical variation into the poverty measurement. However,
there are sharp disagreements as to whether the data are adequate to do
so. Many would agree that it would be important to build an adjustment
for medical costs into the poverty measure, but there are disagreements
as to whether it is better to build an allowance for medical costs into
the threshold, subtract costs from countable income, or do some
combination of the two. It is clear that continued work is needed.
Since the release of the NAS report, the Census Bureau has issued a
set of valuable reports applying and refining NAS recommendations to
provide alternative poverty measures.\13\ The most recent tables,
issued for 2005, show that at a time when the official poverty rate was
12.6 percent, the rate under NAS measures would have been between 12.5
percent and 14.2 percent, depending on how medical expenses and
geographic adjustments are treated in the measure, and depending on
whether the thresholds are computed using the Consumer Expenditure
Survey (as recommended by the NAS panel) or by updating for inflation
using the CPI-U.\14\ Over time, under most NAS measurements, the
numbers in poverty are consistently above the numbers under the
official measures.\15\
---------------------------------------------------------------------------
\13\ See Kathleen Short et al, Experimental Poverty Measures, 1990-
1997, P60-205, (June 1999), available at http://www.census.gov/prod/
99pubs/p60-205.pdf; John Iceland, Poverty Among Working Families:
Findings from Experimental Poverty Measures, P20-203, (September 2000),
available at http://www.census.gov/prod/2000pubs/p23-203.pdf; Kathleen
Short, Experimental Poverty Measures 1999, P60-216 (October 2001),
available at http://www.census.gov/prod/2001pubs/p60-216.pdf; Joe
Dalaker, Alternative Poverty Estimates in the United States: 2003, P60-
227 (June 2005), available at http://www.census.gov/prod/2005pubs/p60-
227.pdf.
\14\ U.S. Census Bureau, Alternative Poverty Estimates Based on
National Academy of Sciences Recommendations, by Geographic and
Inflationary Adjustments: 2004 and 2005, available at http://
www.census.gov/hhes/www/povmeas/altmeas05/
nas_measures_2004_2005_comparison.xls.
\15\ U.S. Census Bureau, Official and National Academy of Sciences
(NAS) Based Poverty Rates: 1999 to 2005, available at http://
www.census.gov/hhes/www/povmeas/altmeas05/nas_measures_historical.xls.
---------------------------------------------------------------------------
In addition to affecting the numbers in poverty, NAS measures also
affect rates for particular groups: poverty rates for married couples
go up, while they go down for female-headed households; they go down
for children and up for the elderly; they go up for whites, down for
African-Americans, and their effect for Hispanics depends on which
measure is used.\16\
---------------------------------------------------------------------------
\16\ U.S. Census Bureau, Alternative Poverty Estimates Based on
National Academy of Sciences Recommendations, by Selected Demographic
Characteristics and by Region: 2005, (2006), available at http://
www.census.gov/hhes/www/povmeas/altmeas05/nas_measures_2005_demog
_and_region.xls.
---------------------------------------------------------------------------
In earlier years, the Census Bureau issued detailed reports
presenting and analyzing the NAS tables. However, the last of these
reports was issued June 2005. Most recently, the NAS tables for 2004
and 2005 were simply posted without narrative or press release and in a
manner such that only specialists were likely to be able to interpret
them.
Beginning in 2006, the Census Bureau has begun reporting a new
series, in which the Census Bureau reports on the effects of various
definitions of income: a ``money income,'' ``market income,'' ``post-
social insurance income,'' and ``disposable income'' definition.\17\
The broadest of the four, ``disposable income'' includes money income,
imputed net realized capital gains, imputed rental income, noncash
transfers, and subtracts imputed work expenses (but not child care) and
taxes. It reports that using its disposal income measure, the poverty
rate for 2005 is 10.3 percent--a level substantially below any of the
NAS measures. Essentially, this happens because it is imputes income
from home equity (even though a prior Census Bureau report had
cautioned about taking such an approach without a corresponding
adjustment to the poverty thresholds), and it does not follow other NAS
recommendations to modify the poverty thresholds, count child care, or
consider medical expenses.
---------------------------------------------------------------------------
\17\ U.S. Census Bureau, The Effects of Government Taxes and
Transfers on Income and Poverty: 2004 (February 14, 2006), available at
http://www.census.gov/hhes/www/poverty/effect2004/
effectofgovtandt2004.pdf; The Effects of Taxes and Transfers on Income
and Poverty in the United States: 2005, P60-232 (March 2007) available
at http://www.census.gov/prod/2007pubs/p60-232.pdf.
---------------------------------------------------------------------------
Recommendations for Next Steps
For at least the past thirty years, there have been discussions
about the need to develop an improved approach to poverty measurement.
While an active research and development agenda should continue, it is
past time to replace the current measure. In my view, the NAS
recommendations offer a valuable starting point for a better measure.
First, I recommend that Congress direct the Census Bureau to issue
a report to Congress, within a specified time period, addressing the
following issues:
To what extent is it now feasible to replace the current
measure of poverty with a measure drawn from the recommendations of the
National Academy of Sciences panel?
Are there data limitations that affect the ability to
implement the NAS recommendations; if so, to what extent could those
data limitations be addressed through improved data collection; and
what would such improved data collection cost?
In light of the research, experimentation and analysis of
the twelve years since the NAS report was issued, are there particular
recommendations that should be reconsidered or additional
recommendations that should be considered in developing an improved
poverty measure?
The Census Bureau should be encouraged to consult with members of
the original NAS Panel and other experts in poverty measurement in the
development of its report.
Second, it is unfortunate that the Census Bureau has seemingly
relegated its NAS analysis to a set of web-only tables, and has
departed from an NAS-based approach in its published narratives. I
recommend that Congress encourage, and if necessary adopt legislation
to direct the Census Bureau to resume fully reporting the NAS measures.
Third, at the same time that the poverty measure is improved, it
would also be valuable for the federal government to begin regularly
reporting a set of additional measures:
Making Ends Meet: It seems clear that the amount of
income a family needs to ``make ends meet'' or have a reasonably decent
standard of living is an amount well above the current poverty line.
Family budget research, polling data, median income data, and other
research all suggest a level roughly twice the current poverty line,
though with substantial geographic variation. It would be valuable to
establish an ongoing research program and methodology for measuring
this concept.
Outside the Mainstream: International comparisons
routinely rely on measuring the share of people below some percentage
of median income, e.g., 50 percent. It is often treated as a measure of
``relative poverty.'' The virtue of such a measure is that it provides
insight into the extent to which a share of the nation's residents is
living with incomes far outside of the social mainstream. In a recent
UNICEF report, the United States ranked 24th of 24 nations on a measure
of the share of children living in families with incomes below 50
percent of median income.\18\ The federal government should track such
a measure, and promote research to better understand the consequences
of being in and growing up in relative poverty.
---------------------------------------------------------------------------
\18\ UNICEF Innocenti Research Centre, Child poverty in
perspective: An overview of child well-being in rich countries, Report
Card 7, (2007), available at http://www.unicef-icdc.org/publications/
pdf/rc7_eng.pdf. Note that in 2005, 50 percent of median income for a
family of four in the United States would have been $35,166, as
compared with the official poverty threshold of $19,157.
---------------------------------------------------------------------------
Asset Poverty: In recent years, we have begun to see
steadily increasing awareness of the importance of assets and asset
disparities. One study estimated that in 2001, about 37.5 percent of
households were ``asset poor,'' meaning they did not have enough liquid
assets to live above the poverty line for three months.\19\ We would
benefit by developing good ongoing measures of assets and asset
poverty.
---------------------------------------------------------------------------
\19\ Robert Haveman and Edward N. Wolff, ``The Concept and
Measurement of Asset Poverty: Levels, Trends and Composition for the
U.S., 1983-2001,'' Journal of Economic Inequality 2 (2): August 2004.
---------------------------------------------------------------------------
Pre- and Post-Tax and Transfer: As the nation moves to a
measure of poverty that includes the effects of tax and near-cash
benefits, it will be important to clearly distinguish pre-tax, pre-
transfer poverty from post-tax, post-transfer. Drawing that distinction
would make explicit the role that tax credits and liabilities and
government benefits play in reducing pre-tax, pre-transfer poverty.
Finally, I want to highlight and underscore one last NAS
recommendation. At present, a large number of government programs
affecting individuals, communities, and states, determine eligibility
or allocate funds on the basis of federal poverty guidelines. For some
purposes, applying the revised threshold and measurement rules may be
appropriate, but in other cases it would not be. For example, with a
new threshold, a program might wish to adjust eligibility accordingly,
or change the percentage or multiple of the poverty line at which
individuals are eligible. Income-counting rules that are appropriate
and feasible for Census Bureau estimates may not be appropriate for
program eligibility determinations. Moreover, if adopting a new measure
of poverty had direct implications affecting program costs and the
allocation of billions of dollars in federal funds, it would likely
paralyze any effort to modify the poverty measure. The NAS panel
expressly recommended that agencies consider the use of the new
measures for purposes of programs they administer, but not that the new
rules be applied automatically when they do not further program
objectives. Such an approach would be essential to efforts to advance a
new and improved poverty measure.
Again, thank you for addressing these important issues.
Chairman MCDERMOTT. We are going to have to go and vote. We
will be over there for about 5 minutes on this vote and then
another 5-minute vote. I think we should be back about 5
minutes after eleven, and I would hope that the panel can all
stay.
We will begin with the questions as soon as we come back.
We thank you all for your participation and you have been a
very good panel, and I think there are some things we want to
get out of you before you get away. So, please wait. We'll see
you then.
Ms. BERKLEY. Mr. Chairman, if I cannot be here, can I
submit my opening statement?
Chairman MCDERMOTT. Sure.
[The opening statement of Rep. Shelley Berkley follows:]
[GRAPHIC] [TIFF OMITTED] T3757A.030
Thank you. We will be back in 20 minutes, or so.
[Recess.]
Chairman MCDERMOTT. If the speakers will take their chairs
again, we have 10 minutes before we are going to have to leave
again. We are in a North Carolina four corners stall in the
Congress and so we are going to have to dodge between votes.
My question, I guess, is this. If we do nothing, what is
the problem, or is there a problem? If we are going to do
something, what is the best way to go about doing it? Now, I
listen to Mr. Besharov and others talk about the things that we
add in. Or even Mr. Weller's idea of taking somebody with
$18,000 income and adding in costs and that brings them up to
$30,000; and, therefore, they are no longer in poverty. Then
they would not qualify for the things that got them out of
poverty.
I mean, that's kind of a catch-22, so what I am interested
in is hearing from you, what is the best way for us to proceed
with this if you think there is some compelling reason why we
should do something? I know you don't think we need to or can
do it, but we are sort of megalomanics in the Congress, so I
think we can do anything.
Go ahead.
Mr. BESHAROV. Well, no. I think you should do things. I
think that what you cannot do successfully is change ``the
official poverty line.'' We are not really talking about the
poverty measure used for program eligibility. That is an
entirely separate decision. I think those are key words:
``entirely separate decision". The considerations that go into
program eligibility, as you just pointed out, are very
different from trying to keep track of economic or material
well-being of low-income Americans.
Quickly, and then I will get off the floor. Here is what I
think you should do in three easy steps. I don't think you can
affect the official poverty measure without having a broad
congressional consensus about raising spending across the board
for means tested programs. Otherwise, you have got to
disconnect the official poverty measure from every program that
exists.
Chairman MCDERMOTT. Can I stop you there? Haven't we
disconnected it now by using 200 percent of poverty all over
the place?
Mr. BESHAROV. Yes and no.
Chairman MCDERMOTT. I mean, that is what I listened to you
and I thought about my days in the State legislature and how
Washington State was always pushing above the poverty line and
no deed goes unpunished.
Mr. BESHAROV. Here is what I would say about that, but you
are the pros. Programs are currently at a multiple of the
poverty lines. You mentioned it is at a few hundred percent of
the poverty line.
State and Federal grant programs that are formula driven,
are often driven by the poverty measure. So, if the poverty
measure is changed, then every Subcommittee, and then every
Committee of the Congress has to disconnect, and redo it. Then,
both houses of Congress have to agree.
Chairman MCDERMOTT. I understand the problems of politics.
Here are some others. How do you think we should do it, because
he does not basically think we should?
Now, you've got a minute.
Mr. GREENBERG. I think the key in moving forward is to say
in doing a revision of the poverty standard and measurements,
it does not automatically lead to changes affecting program
eligibility or allocation of Federal resources. Sever the link
between them and say for each individual program, for each
Federal benefit that is allocated on the basis of the poverty
line, that it will now be appropriate to make a judgment of how
to proceed, but do not do it automatically.
Chairman MCDERMOTT. Yes?
Dr. RUGGLES. I would say that the main thing you want a
poverty line for in terms of Federal programs is figuring out
who you need to serve. Who do you need to serve? Who is in
need?
Chairman MCDERMOTT. Okay, yes.
Dr. RUGGLES. So what you need is the best measure for that
purpose, and in the long run that is going to help you design
better programs. Worrying about how it does or does not affect
existing programs in the short run isn't the first priority.
You really ought to be looking at what you are trying to get
to, not what is the short run effect.
In terms of the short run effect, we revise all of the
other statistics we use for the government. The CPI has gone
through a whole bunch of revisions done by Committee that have
been at least as controversial as anything we are talking about
for poverty, and it effects even more programs than the poverty
line does.
Chairman MCDERMOTT. Are you then suggesting something in
the Census Bureau? I don't know where that function is, but
someplace that would function continually as an annual update,
much as we do with the trustees on Medicare and Social
Security, that same sort of thing.
Is that what you are talking about?
Dr. RUGGLES. Yes, something like that could be done; or,
you could do what the Bureau of Labor Statistics does to update
the CPI where they have a Committee that meets on a periodic
basis, comes up with recommendations and then implements them.
That kind of an approach where you have a built in periodic
review, come up with recommendations, they get implemented.
Chairman MCDERMOTT. Does it require that in doing that you
also say what their goal is? Do we have to set the goal or do
we let the commissions do that?
Dr. RUGGLES. I think you can, to some extent, let the
commission do that, but over all, I think that for policy-
related purposes the goal ought to be hearing out who is in
need.
Chairman MCDERMOTT. So, if you are going to deal with
children's poverty, talk about what that is and then adjust the
poverty line for whatever.
Dr. RUGGLES. Exactly.
Chairman MCDERMOTT. Thank you.
I want to go to my colleague.
Mr. WELLER. Thank you, Mr. Chairman.
First, I would just like to ask unanimous consent for the
copy of H.R. 3243 along with the introduction statement in the
record.
Chairman MCDERMOTT. Without objection.
[H.R. 3243 follows:]
[GRAPHIC] [TIFF OMITTED] T3757A.031
[GRAPHIC] [TIFF OMITTED] T3757A.032
[GRAPHIC] [TIFF OMITTED] T3757A.033
[GRAPHIC] [TIFF OMITTED] T3757A.034
Mr. WELLER. Thank you, Mr. Chairman, and thank you for this
hearing.
Chairman MCDERMOTT. I am glad you put it in, because I
think we need to have some things on the table to start talking
about.
Mr. WELLER. Absolutely, you have got to have ideas to
discuss them, right? I am sorry. I apologize to our witnesses
today for the interruptions here. As a member of the minority
party, all we are asking for is the opportunity to offer
amendments, which is normally a part of the legislative
process. But we are denied that opportunity on the floor today
on the SCHIP legislation. In the Committee on Ways and Means,
we did have that opportunity, but the Commerce Committee shut
them down. So we would like just to have the opportunity that,
when we were in the majority, we gave the minority at that
time.
Mr. Besharov, you are the contrarian today amongst the
group. I've got a copy of a study that was put forward by the
Center on Budget and Policy Priorities, a left of center group
that usually advocates expansion of various government
programs. They noted in 2003 that the earned income tax credit
directly lifted 2.4 million children in working families above
the poverty line. For children, before counting the EITC, 12.6
million were in poverty and after counting the EITC that number
dropped to 10.2 million. In that study their measure of poverty
counted food, housing, and energy assistance benefits as income
and subtracted income and payroll taxes.
As a share of antipoverty benefits, is the portion of
antipoverty benefits that are currently not included in income
when we determine the poverty level rising or falling?
Mr. BESHAROV. It has risen substantially in the last thirty
years. It rose a great deal in the last six, and my guess is
that it will continue to rise. That study, by the way, is
typical of what many researchers do. They use the current
poverty line, left and right, and count all these things that
we don't otherwise count. This is not to say that material
needs no longer exist, but we should see whether, against some
objective benchmark, we've made progress over the last five,
ten or forty years.
What the Center on Budget Priorities report did with the
Joint Economic Committees, Democrats, was show that these
Federal programs reduced measured poverty. That is an important
part of a political debate, and that is why I have taken a
quick look at your bill. I think something like that is
tremendously important. There are some Census Bureau people
here, who have started making that distinction, by just showing
the impact of Federal aid to low-income people. That is, I
think, a central part of understanding how well the government
is doing and what is left to be done.
Chairman MCDERMOTT. You know, if we continue to count, not
count means tested benefits, do you believe the official
poverty rate would actually be more inaccurate over time?
Mr. BESHAROV. There are many things that are making the
official poverty rate an inaccurate measure of where we were in
1968. I wanted to make the distinction. I agree with my
colleagues here that you could do a lot of things to make this
measure better.
But there are many things that make the current measure
inaccurate. I don't think it measures inflation well. It does
not deal with many forms of income that low-income Americans
receive. The problem, I think, will fester and get worse, if we
do increase, for example, the EITC as has been proposed for
single people. That would have a major impact on their poverty
status and official measure would never capture it.
Mr. WELLER. My Chairman slipped me a note that says we have
5 minutes before the vote ends on the floor, so go as long as
you want, he said. So, I guess that is a reminder that I need
to wrap up.
You know, the legislation that I have offered, I would
note, does not determine the eligibility for benefits. It sets
in place three formulas for determining the level of poverty
that should be part of the discussion as we determine how can
we be more accurate.
Since we are out of time, I do want to ask each of the
witnesses to take a look at the legislation introduced as part
of this discussion. I welcome your ideas and my door is open to
discuss them.
So, thank you, Mr. Chairman.
Chairman MCDERMOTT. Thank you, Mr. Weller.
I want to thank all of the witnesses. My view of this is
there is no perfect answer otherwise it would have been done a
long time ago. So I appreciate that not all the good ideas in
the world come from the position politically where I am, nor
where Mr. Weller is. Ultimately, I think in order to have some
consensus as described here, there has got to be consensus in
the Congress and it has got to be reached by us working
together on this issue.
I appreciate all of you coming, giving your ideas. I
suspect staff will be back from both sides to find out what you
think of his bill and whatever we put out, and we thank you
very much for coming.
Thank you.
[Whereupon, at 11:25 a.m., the hearing was adjourned.]
[Submissions for the record follow:]
[GRAPHIC] [TIFF OMITTED] T3757A.035
[GRAPHIC] [TIFF OMITTED] T3757A.036
[GRAPHIC] [TIFF OMITTED] T3757A.037
[GRAPHIC] [TIFF OMITTED] T3757A.038
[GRAPHIC] [TIFF OMITTED] T3757A.039
[GRAPHIC] [TIFF OMITTED] T3757A.040
[GRAPHIC] [TIFF OMITTED] T3757A.041
[GRAPHIC] [TIFF OMITTED] T3757A.042
[GRAPHIC] [TIFF OMITTED] T3757A.043
[GRAPHIC] [TIFF OMITTED] T3757A.044
[GRAPHIC] [TIFF OMITTED] T3757A.045
[GRAPHIC] [TIFF OMITTED] T3757A.046
[GRAPHIC] [TIFF OMITTED] T3757A.047
[GRAPHIC] [TIFF OMITTED] T3757A.048
[GRAPHIC] [TIFF OMITTED] T3757A.049
[GRAPHIC] [TIFF OMITTED] T3757A.050