Revised February 9, 2005

By Isaac Shapiro and Robert Greenstein [1]


PDF of full report
Press Release: HTM | PDF

View Related Reports

If you cannot access the files through the links, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.

There is a distinct possibility that efforts to reduce the deficit this year will take a large and disproportionate bite out of programs that provide key supports and services to low-income Americans.  This analysis explains that because substantial parts of the budget, including revenues, are expected to be largely or entirely “off the table” when deficit reduction plans are drawn up — and also because low-income programs tend to lack the political support of other programs with more powerful constituencies — a very large share of the budget reductions enacted this year may consist of cuts in programs for low-income families and individuals.  Indeed, when Congress completes work on the budget this year, it is possible that a majority of the cuts will have been made in low-income programs. 

Such an approach would represent unbalanced priorities.

Deficit reduction can be accomplished — and has been in the past — without injuring the most vulnerable Americans.  The bipartisan deficit-reduction package in 1990, negotiated and signed by a Republican President, and the deficit-reduction package enacted in 1993 stand out in this regard.  Both of those measures included a combination of reductions in programs and tax increases (the tax increases primarily affected high-income households), and did much to help move the nation’s fiscal position from one of large, structural deficits to the surpluses that emerged in the late 1990s.  Neither measure contained sizable reductions in programs for low-income families.  To the contrary, both achieved extensive deficit reduction while strengthening programs that assist the working poor, such as the Earned Income Tax Credit.

Why Low-Income Programs May Be Sliced Deeply This Year

The factors indicating that budget cuts could fall heavily and disproportionately on the backs of those most in need include the following.

There appears to be widespread support not to reduce Social Security (at least in the near term), defense, or homeland security spending.[2]  Indeed, the latter two areas may receive further increases under the President’s budget.  In addition, Congressional leaders are expected to shy away from adopting significant cuts in Medicare at the same time that they are considering controversial changes in Social Security and with the Medicare drug benefit legislation about to take effect.  There also appears to be broad consensus on Capitol Hill that scheduled reductions in Medicare payments to physicians need to be reversed; that reversal is likely to consume most or all of any savings made elsewhere in Medicare costs.[3]

Table 1

Low-income Program Cuts Could Be Highly Disproportionate

Low-income programs as a share of:


The change in the deficit since 2001 due to legislative changes


A deficit-reduction plan evenly divided between spending and revenue*


All program spending


Programs “on the table” that might be cut**


Potential cuts, given political realities

More than 50%

*    This scenario assumes half of deficit-reduction would consist of increased revenues and half of reduced spending.  It also assumes all programs would be cut by the same percentage.
**   Excludes defense, homeland security, Social Security, and Medicare

Finally, the Administration and the Congressional leadership are determined not to raise taxes.  This is so even though in 2004, federal tax revenues fell to their lowest level, measured as a share of the economy, since 1959.

The net result is that only a minority of the budget is likely to be considered in deficit-reduction effects this year.  If revenues, Social Security, Medicare, defense, and homeland security are taken off the table for deficit reduction purposes, low-income programs will make up 49 percent — or nearly half — of the part of the budget that remains “on the table” for potential reductions.  Thus, if all of the programs left on the table are cut by the same proportion, low-income programs would bear nearly half of the cuts.

By contrast, programs that provide services or benefits to low-income families and individuals make up 20 percent of all government expenditures.  And increases in these programs accounted for only six percent of the cost in 2005 of legislation enacted since January 2001, the period over which the budget has moved from sizeable surpluses to large deficits.

Moreover, low-income programs may be reduced more sharply, on average, than other programs that are left “on the table.”  Low-income programs tend to have less powerful backers than many other programs, such as farm programs or NASA, to name a few.  Suppose, for example, that Congressional budget writers seek (as they may) to roll back the agriculture subsidies included in the 2002 farm bill and require, through this year’s Congressional budget resolution, that the House and Senate Agriculture Committees reduce entitlement programs under their jurisdiction by a large amount.  Those Committees could respond by shielding farm subsidies to a substantial degree and cutting the food stamp program significantly instead.  The food stamp program provides an average benefit of just 94 cents per person per meal (in 2004).

Media reports indicate that certain low-income programs are likely to be singled out for significant cuts in the President’s budget.  The New York Times has reported that according to Administration officials and Congressional aides, the President’s budget will include large cuts in low-income housing programs.[4]  That would not be surprising; the budget the Administration submitted last year featured deep reductions in the nation’s principal low-income housing assistance program — the Section 8 housing voucher program — with the cuts slated to reach 30 percent by 2009.  News accounts also suggest that the Administration will propose to squeeze the Medicaid program and may propose to cap or scale back federal contributions for large parts of that program, with the result that federal funding would not keep pace with health care costs.  Such a development likely would result in significant reductions in health care coverage or benefits for low-income families over time.

Congress, for its part, is likely to cut low-income programs significantly more than the President proposes.  Media reports indicate that the Administration will propose some scaling back of agricultural subsidies; as noted, Congress could convert some of those reductions into food stamp cuts.  An examination of the House and Senate budget resolutions of the past couple of years provides further indication that Congress may go beyond the White House in seeking cuts in low-income programs.  Last year, the Senate passed a budget resolution that assumed cuts in the Earned Income Tax Credit which the President had not proposed, while the budget resolutions that the House passed in each of the past two years called for substantial entitlement reductions, which appeared aimed in significant part at low-income programs the Administration had not proposed cutting.[5]

Such an Approach Would be Ill-Advised 

A deficit-reduction approach that heavily and disproportionately hits programs that provide services and benefits to the neediest households would be ill-advised.

These and other related issues are examined below.


Low-Income Programs and the Deficit

For the purposes of this analysis, low-income programs fall into three categories. 

Low-income programs make up a relatively small portion of the budget — about one-fifth in 2005.[10]  The majority of expenditures for low-income programs occurs through means-tested entitlement programs; a significant minority occurs through discretionary programs; and roughly one-tenth results from refundable tax credits.

Recent Trends in Low-income Programs

Spending on low-income programs rose from fiscal year 2000 through fiscal year 2005, but the increase was modest in comparison to other factors that influenced the swing in the nation’s fiscal position from one of surpluses to one of deficits.  This can be seen by examining the cost of all legislation enacted since early 2001.  Without this legislation, the nation would currently be running a budget surplus.  In 2005, data from the Congressional Budget Office indicate that the combined costs of this legislation will be $539 billion, substantially larger than the expected deficit of about $400 billion.

Table 2

Cost in 2005 of Legislation Enacted Since January 2001


(in billions of dollars)

As a share of the total cost of such legislation

Tax Legislation






Defense, Homeland Security, and International



Non Low-Income Domestic Discretionary



Non Low-Income Entitlement Legislation



Non Low-income Programs






Low Income Discretionary



Low Income Entitlement Legislation



Low Income Programs






All Legislation



Source:  CBPP calculations from Congressional Budget Office data.

Another way of assessing recent trends that have influenced the deficit is to examine how various categories of spending, as well as the level of revenues, have changed as a share of the economy over the 2000-to-2005 period.  This comparison, too, shows that while there has been growth in the share of the economy devoted to low-income programs, the size of this increase pales in comparison both to the reduction in tax revenues and to increases in spending for other types of programs.

Table 3

Tax and Spending Changes as a Share of the Economy, From 2000 to 2005



Change as a Share of GDP

of Total Change









Low-Income Health Programs



Other Low-Income Programs



Defense (Discretionary)



All Other Programs















Net Effect on the Deficit



*Interest payments declined as a share of GDP, despite the increase in deficits, because of the drop in interest rates.

Source:  CBPP calculations from Congressional Budget Office data.


Longer-term Trends

Not only has growth in low-income programs been relatively modest over the past few years, but the cost of these programs (other than health insurance programs) is expected to decline as a share of the economy in the years ahead.  This anticipated downward trend is due partly to the fact that the number of households with low incomes should decline as the economy more fully recovers.  Enrollment in programs such as food stamps rises during periods when unemployment is elevated and declines when employment growth accelerates.

Adding health insurance programs into the equation changes this outcome only modestly.  Medicaid costs are expected to rise over time as a share of the economy, both because health care costs in general are rising and because the aging of the population means that the number of elderly individuals who live in nursing homes or otherwise need health care will increase.

Nonetheless, in the medium term, the rise in Medicaid expenditures and the increase in expenditures that will result from the low-income provisions of the Medicare prescription drug bill will largely be offset by the projected decline in expenditures for other low-income programs.  If no cuts are made in low-income programs, expenditures for low-income programs in 2015 will equal 3.89 percent of the Gross Domestic Product, hardly larger than the expected 3.80 percent level in 2005.  (This calculation is based on Congressional Budget Office projections of spending levels under current law, as well as its economic forecast.  The year 2015 is used because it is the last year for which detailed spending projections are now available.) 

The appendix includes a full discussion of trends in Medicaid costs.  The appendix shows that the cost of Medicaid coverage is growing much more slowly than the cost of private sector coverage and that Medicaid costs per person are substantially lower than those for private health insurance.  It also notes that states have instituted an unprecedented series of Medicaid budget cuts and cost containment practices in recent years.  Federal cutbacks to Medicaid would lead to further reductions in Medicaid coverage, driving up the number of uninsured and leading, potentially, to a two-tier health system in which Medicaid recipients receive less adequate care than those covered through private plans.


After-tax Income, Poverty, and Health Insurance Trends

In assessing proposed cuts to low-income programs, it is worth reviewing the basic indicators of need to which they respond.  Here we examine recent trends in after-tax income, poverty, and health care in the United States.

Income and Tax Trends

The Congressional Budget Office provides the most comprehensive data available on changes in income over time.  These data have several advantages compared with the more widely-cited Census data.  The CBO data capture a variety of types of income important to low-income households that the standard Census data do not include, such as the Earned Income Tax Credit and food stamps.  The CBO data also capture substantial amounts of income at the top of the income scale that the Census data miss; the Census data do not include capital gains income or wage and salary incomes above $1 million.  Finally, the CBO data measure after-tax income, which is a better yardstick of the amount of income that households have at their disposal than the before-tax income measure that the Census data reflect. 

The CBO data show that in recent decades, the gap between rich and poor in the United States has widened markedly.  (See Table 4.)  Income gains have been spectacular among high-income households but very modest among low-income households.  (This pattern holds throughout the income spectrum; middle-income households have gained more than low-income households, high-income households have gained more than middle-income households, and the highest-income households have gained the most by far.)

Between 1979 and 2001, the first and last years that the CBO data cover:

Table 4

Average After-Tax Income by Income Group

(in 2001 dollars)

 Income Category



Percent Change

Dollar Change

Lowest fifth





Second fifth





Middle fifth





Fourth fifth





Top fifth






Top 1 Percent





Source: Congressional Budget Office, Effective Federal Tax Rates:  1979-2001, April 2004.

The same pattern holds for before-tax income.  When data on before-tax income from a recent National Bureau of Economic Research study
[12] that covers a longer period of time are viewed in combination with the CBO data, it appears that the top one percent of households received a larger share of the before-tax income in the nation in 2001 than at any time since 1936, except for the years from 1997 to 2000.  In other words, except for the recent peak years of the stock market, income was more concentrated at the top in 2001 than in the previous 65 years.

The more current but less comprehensive data from the Census Bureau tell a similar story.  The Census data indicate that in 2003, the share of before-tax income in the nation that went to the bottom 40 percent of households was the lowest on record.  These data go back to 1967.  (When CBO data for 2003 become available, they may or may not show the same result for 2003; the CBO measure for 2003 will show income to be more concentrated at the top than in most other years in recent decades, but it may show income to be somewhat less concentrated in 2003 than it was in 2000.)  The Census data also indicate that incomes fell across the board from 2001 to 2003, but fell fastest among the bottom fifth of households.

It should be noted that because the CBO data that are currently available go only through 2001, they include little of the effects of the recent tax cuts.  The tax cuts have increased income inequality further; they have boosted the after-tax incomes of high-income households — both in absolute dollars and as a share of the after-tax income in the nation — much more than they have increased the incomes of those in the middle or the bottom of the income scale.  Data from the Tax Policy Center show that:

Table 5

Distribution in 2004 of Tax Cuts Enacted Since 2001

Income group

Tax Cut

Share of the
Tax Cuts

Percentage Change in After-Tax Income





Lowest 20 percent




Second 20 percent




Middle 20 percent




Fourth 20 percent




Top 20 percent












Top one percent




Above $1 million




Source:  Urban-Brookings Tax Policy Center

In short, both the income trends of recent decades and the tax cuts enacted since 2001 have decisively favored high-income households.  Given this context, it would be ill-advised for policymakers to adopt deficit-reduction measures that hit low-income households harder than others and essentially used the proceeds to help finance the tax cuts.  Unfortunately, this is what Congress and the President may be poised to do.


Recent poverty data provide further evidence that significant cuts to low-income programs would be problematic.  In 2003, the latest year for which data on poverty are available:

One of every eight people in the nation was poor, with poverty having risen for three consecutive years.  Altogether, there were 35.9 million people living below the poverty line in 2003.  A similar number of people were either hungry or threatened by hunger — a group the government terms “food insecure” — at some point during the year.[14]

Among those who were poor, the depth or severity of their poverty — that is, the degree to which their incomes fell below the poverty line — was greater in 2002 (latest data available) than in any year on record, with these data going back to 1975.[15]  Stated simply, those who are poor have grown poorer, on average.

Children are especially likely to live in poverty.  Some 17.6 percent of children in the United States were poor in 2003.  (By way of comparison, 12.5 percent of the overall U.S. population lived in poverty that year.)

There is little reason to believe these figures have improved significantly since 2003.  The employment rate was essentially the same in 2004 as in 2003; in both years, an average of 62.3 percent of adults were employed.  In addition, wages at the bottom of the economic ladder fell further behind inflation in 2004.  And with the expiration of the temporary federal unemployment benefits program at the end of 2003, some 3.5 million American workers exhausted their regular unemployment benefits in 2004 before finding employment, the largest such number on record.  Significant numbers of these individuals and their families have gone without paychecks or unemployment benefits for extended periods of time and are likely to have fallen into poverty.

Health Insurance

The number of people who lack health insurance coverage throughout the year has risen steadily since 2000; it totaled 45 million in 2003.  In percentage terms, 15.6 percent of Americans — almost one in every six people — were uninsured in 2003.  Although the economy has been slowly recovering since 2001, health insurance coverage has deteriorated.

The principal reason for the decline in health insurance coverage has been the erosion of employer-based insurance coverage, spurred by the escalation in the cost of health insurance and sluggish job growth over the past few years.  The percentage of people with employer-based health insurance dropped from 63.6 percent in 2000 to 60.4 percent in 2003.  This is the lowest level of employment-based insurance coverage since 1993.  Census data show that the recent decline in employer-based health insurance has hit low-income households harder than high-income households.[16]

In response to the loss of employer-based health insurance and the increase in the number of low-income people, enrollment in Medicaid and the closely-related State Children’s Health Insurance Program (SCHIP) rose in 2003.  This development was particularly important for children, for whom the growth in Medicaid and SCHIP coverage was sufficient to offset the loss of private coverage.  The percentage of children who lack health insurance stood at 11.4 percent in 2003, not a significant change from its 2002 level.

For working-age adults (those aged 18 to 64), by contrast, a slight growth in Medicaid coverage was not sufficient to outweigh the larger loss of employer-based coverage.  (In the majority of states, Medicaid eligibility limits for parents are set well below the poverty line, and low-income working-age adults without children generally cannot get any coverage through Medicaid unless they have severe disabilities.)  The number of working-age adults lacking health insurance coverage rose to 36.3 million — or 20.2 percent of such adults — in 2003.  These are the highest levels on record.

These developments are responsible for a portion of the recent increase in Medicaid costs.  Medicaid responds both to weak economic conditions and to the erosion of employer-based coverage among the working poor.  It has prevented even greater increases in the number of people who are uninsured.

International Comparisons

The United States compares unfavorably with other western industrialized nations along all three dimensions just discussed — income inequality, poverty, and health care coverage.  The relative weakness of U.S. government programs, as compared to counterpart programs in other nations, is a major factor behind these unfavorable comparisons.

The best source of information on income and poverty levels across nations comes from the Luxembourg Income Study (LIS).  The LIS is a cooperative research project designed to facilitate comparisons of living standards across industrialized nations.  It compiles and standardizes data from the chief statistical agencies of governments around the world.  Two recent analyses by the director of this study have found:[17]

These analyses also found that government policies in the United States do less to reduce child poverty or income inequality than the policies in almost all comparable countries.

The United States also stands out among high-income nations for the large number of people who lack health insurance.  In all other high-income countries, health insurance coverage is universal.


What Deficit-Reduction Approach Will be Taken?

The information presented here provides some useful guidance regarding efforts to reduce the deficit.  This information indicates that:

These developments suggest that a deficit-reduction approach that takes large and disproportionate slices out of benefits and services for needy households would be both unwarranted and inequitable.  Unfortunately, Congress and the President may be poised to take such an approach this year, in large part because many other ways to reduce the deficit are not likely to be considered.

 Altogether, this means that 58 percent of the budget is likely to be largely or entirely “off the table” when it comes to reducing the deficit.

Other developments also suggest that programs providing benefits or services to needy families and individuals may be a prime target.  These developments include an expected Administration proposal to essentially freeze overall funding for domestic discretionary programs (other than homeland security) for the next five years, without any adjustment for inflation or population growth.  The impact of these cuts would grow with each passing year.  In 2010, such a proposal would mean that discretionary programs would have to be cut by $46 billion below their current level, adjusted for inflation.

The budget that the Administration submitted last February provides a glimpse of what cuts might be in store.  Under last year’s budget, nearly every domestic discretionary budget account in the entire budget (outside of homeland security) was slated for decreases below the budget baseline (i.e., below the 2004 funding level, adjusted for inflation), starting in fiscal year 2006.

Moreover, last year’s budget proposed particularly steep reductions in the housing voucher program, the nation’s principal housing assistance program for low-income families.  By 2009, that program would have been cut $4.6 billion a year, or 30 percent.  A cut of that depth could mean that the number of low-income families provided rental assistance through the program would be cut by 600,000 (30 percent of the number currently assisted) or that rents for the low-income families assisted by the program would be raised an average of $2,000 per family in 2009.  (This is the average rental increase that would be needed to generate $4.6 billion in savings in 2009 if the number of households receiving assistance remained unchanged.)

The Administration also may propose to cap or scale back funding for at least some parts of the Medicaid program, and there may be an effort in Congress to convert the Medicaid program to a block grant.  Over time, such steps would mean that federal funding will not keep pace with increases in health care costs, likely leading to large reductions over time in health care coverage or benefits for low-income families and individuals.  Congress also could seek to cut low-income programs such as food stamps and the Earned Income Tax Credit through the budget “reconciliation” process.

A Different Approach 

A different and more balanced approach can and should be pursued.  Such an approach would begin by putting everything on the table; all aspects of the budget should be considered as possible elements of a deficit-reduction package.  This would include revenues.  For example, it would be sensible to consider repealing two significant tax cuts enacted in 2001 that have not even started to take effect yet and that are slated to phase in starting in 2006.  Virtually all (97 percent) of the benefits of these tax cuts, which concern two little-known tax provisions known as “Pease” and “PEP,” will go to households with incomes exceeding $200,000 a year.  Yet there has been no discussion of canceling or deferring these tax cuts despite the fact that they will worsen the deficit by more than $12 billion a year once they are fully implemented, and thus will place added pressure on Congress to make significant cuts in programs of broad general benefit.

Also on the table should be defense spending that is not related to anti-terrorism efforts.  Observers such as CBO director Douglas Holtz-Eakin have begun to question whether the defense spending increases of recent years are being spent effectively and whether continued defense spending increases are affordable.[19]

On a related front, any changes in Social Security that entailed added costs should be “paid for” contemporaneously.  Such costs should not be financed through borrowing that would increase deficits for several decades to come.

Innovative deficit-reduction approaches also should be considered, such as improving the accuracy of the cost-of-living adjustment used for many programs, including Social Security, and for the tax system.  The current adjustment overstates inflation slightly, resulting in more spending and lower tax collections than would otherwise occur.  Using a more accurate measure of inflation that the Bureau of Labor Statistics has developed could save $35 billion a year by 2014 and larger amounts in years after that.[20]

Two recent efforts demonstrate that it is possible to reduce the deficit substantially while protecting basic supports for needy Americans.  The first was the deficit-reduction package of 1990, which emerged from a bipartisan budget summit in which both revenues and a broad array of programs were put on the table.  The balanced 1990 package was negotiated and signed into law by President Bush’s father.

The second such measure was the deficit-reduction package of 1993.  That package was not enacted on a bipartisan basis but did reduce the deficit substantially.

In both 1990 and 1993, major deficit-reduction was achieved without savaging low-income programs.  To the contrary, in crafting both measures, policymakers took pains to protect poor and vulnerable Americans and the programs that serve them.  The 1990 and 1993 approaches to deficit reduction remain the most appropriate models to follow.


            Future Medicaid Growth Is Not Due to Flaws in the Program’s Design, but to Demographic Trends and General Increases in Health Care Costs[21]

As noted in the text, Medicaid costs in the future are expected to rise as a share of the economy.  This appendix attempts to sort out the reasons for this rise as well as the consequences of attempting to curtail this trend without addressing its underlying causes.

In short, the rise in Medicaid costs is not due to the design of the Medicaid program.  Rather, it is due to two broader trends — increases in health care costs that are affecting the U.S. health care system as a whole, including the private sector, and the aging of the population.  Specifically: 

As this discussion indicates, meaningful relief from rising Medicaid costs rests upon broader efforts to address health care cost increases throughout the U.S. health care system and to close gaps in Medicare coverage.  In the absence of such broader efforts, reductions in the federal contribution for Medicaid costs would have adverse consequences.

End Notes:

[1] Thanks to Ruth Carlitz, Joel Friedman, David Kamin, Richard Kogan, Sharon Parrott, and Dottie Rosenbaum of the Center for their contributions to this analysis.

[2] Some recent media stories have reported on potential “cuts” in defense programs that the Administration may propose.  Even if these “cuts” are offered, however, the Administration still is widely expected to propose an increase in defense spending in 2006, even after taking inflation into account.  The “cuts” that may be proposed are apparently cuts from the Pentagon wish list, not from the budget baseline.  Stated another way, any cuts in particular components of the defense budget are expected to be more than offset by other defense funding increases.  (The forgoing applies to defense spending excluding spending in Iraq, which the Administration is expected to leave out of the budget it presents on February 7 and to handle through a supplemental budget request.)

[3] An article in CQToday January 18, 2005, page 16, that was based on an interview with new Senate Budget Committee chairman Judd Gregg summarized the situation as follows: “…GOP budget writers are likely to largely leave alone the Medicare health care program for the elderly, according to Gregg and senior congressional aides in both parties.  Instead, they are likely to tap the Medicaid health care program for the poor for savings.”

[4] Robert Pear, “Applying Breaks to Benefits Gets Wide G.O.P. Backing:  Bush to Seek Firm and Enforceable Curbs,” The New York Times, January 9, 2005, page 19.

[5] See the following Center on Budget and Policy Priorities’ reports.  David Kamin, Richard Kogan, and Joel Friedman, “House Budget Committee Process Proposal Would Not Restrain Those Areas of the Budget That Have Contributed Most to the Deficits,” March 18, 2004; Robert Greenstein and Richard Kogan, “Comparing the House and Senate Budgets,” April 7, 2003; and Sharon Parrott and John Springer, “The Human Costs of Cuts in Major Low-Income Programs Contained in the House Budget Resolution,” March 28, 2003.

[6] The cost of expansions in refundable tax credits is included in the six percent figure cited for expansions that were enacted in low-income programs, not in the 45 percent figure for the tax cuts.

[7] These estimates include the costs of extending relief from the Alternative Minimum Tax, to the extent that such relief is necessitated by the tax cuts.  (It does not include the cost of AMT relief that would have been needed in the absence of the tax cuts.)  It also includes the associated increases in interest payments on the debt.

[8] This calculation includes the effects of expansions in refundable tax credits, which we count as spending increases elsewhere in this analysis.

[9] The percentage of the population that lacked health insurance in 2003 — 15.6 percent — was tied with the percentage for 1996 as being the third highest percentage on record.  The percentage of working-age adults who lacked insurance was the highest on record in 2003.  On the other hand, the percentage of children without insurance has declined.

[10] More specifically, in 2005, low-income programs — including refundable tax credits — made up 18.9 percent of all federal spending.  If interest payments are excluded from the calculation, low-income programs made up 20.4 percent of federal spending.

[11] This figure excludes the expansions in the refundable tax credits.  It includes the costs of extending tax cuts that existed prior to 2001 but were scheduled to expire, as well as the full costs of all adjustments to the Alternative Minimum Tax.

[12] Thomas Pikety and Emmanuel Saez, “Income Inequality in the United States, 1913-1998,” National Bureau of Economic Research, September 2001.  Tables have been updated through 2000 at

[13] In its analyses, the Tax Policy Center examines the effects of the tax cuts on different “tax units.”  These “tax units” include individuals and married couples who file income tax returns, as well as those who do not file (primarily because their incomes are below the minimum threshold for filing).  For shorthand, this report uses the term “households” instead of “tax units.”

[14] Arloc Sherman, “Hunger, Crowding and Other Hardships Are Widespread Among Families in Poverty,” Center on Budget and Policy Priorities, December 20, 2004.  Specifically, according to the government’s definition, “food insecurity” means that at “some time during the year, these households were uncertain of having, or unable to acquire, enough food for all their members because they had insufficient money or other resources.”  Note that the two populations discussed in the text are not identical; that is, a significant number of the individuals threatened by hunger in 2003, according to government statistics, were not poor.

[15] This finding, which includes the effects of non-cash benefits and the Earned Income Tax Credit, is for 2002.  It means that the average amount (after adjusting for inflation) by which poor households fell below the poverty line was larger in 2002 than in any year on record, back to 1975.  The Census Bureau has not yet released such information for 2003.  Other data suggest that, if anything, the severity of poverty worsened again from 2002 to 2003.

[16] From 2000 to 2003, the share of the poor covered through employer-sponsored health insurance plans fell by one-seventh, while the share of households with incomes above four times the poverty line who are covered through employer-sponsored plans declined by one-fortieth.

[17] Timothy Smeeding, “Public Policy, Economic Inequality, and Poverty:  The United States in Comparative Perspective,” draft revision of a paper presented at the ‘Inequality and American Politics Conference,’ held February 20, 2004 at Syracuse University; and Timothy Smeeding, “Children in America:  A Comparative View of our Nation’s Future,” PowerPoint presentation at a Congressional Seminar, December 9, 2004.

[18] If Medicare is also on the table, low-income programs would make up 37 percent of the programs that are potentially subject to cuts.

[19] Patrice Hill, “CBO raps Bush plan on Social Security,” The Washington Times, front page story.

[20] For a discussion of this adjustment, see Robert Greenstein, “A Simple Proposal That Can Yield Substantial Savings Over Time,” Center on Budget and Policy Priorities, May 18, 2004.  For a detailed discussion of how a more balanced approach to deficit-reduction could occur, see Robert Greenstein and Peter Orszag, “A Broken Fiscal Policy . . . and How to Fix It,” in Mark Green, ed., What We Stand For: A Program for Progressive Patriotism, Newmarket Press, 2004.

[21] Thanks to Leighton Ku and Vikki Wachino for their assistance in writing this appendix.

[22] John Holahan (director of the Urban Institute’s Health Policy Center) and Arunabh Ghosh, “Understanding the Recent Growth in Medicaid Spending, 2000-2003,” Health Affairs, January 26, 2005; Kaiser Family Foundation, news release, “A Sharp Rise in Enrollment During the Economic Downturn Triggered Medicaid Spending to Increase by One-Third from FY 2000-03,” January 26, 2005.

[23] This study also found that people with Medicaid and people with private insurance used health services at roughly comparable levels.  Jack Hadley and John Holahan, “Is health care spending higher under Medicaid or private insurance?” Inquiry, 40:323-42, Winter 2003/2004.  Similar findings were reached by federal researchers: see Edward Miller, Jessica Banthin, and John Moeller, “Covering the Uninsured: Estimates of the Impact on Total Health Expenditures for 2002,” Agency for Healthcare Research and Quality Working Paper No. 04407, November 2004.

[24] SCHIP stands for State Children’s Health Insurance Program; for children in low-income working families, SCHIP complements the Medicaid program.

[25] Oregon Health Research and Evaluation Collaborative, “Research Brief:  Changes in Enrollment of OHP Standard Clients,” January 2004, and “Research Brief:  The Impact of Program Changes in Health Care for the Oregon Health Plan Standard Population: Early Results from a Population Cohort Study,” March 2004.