Revised June 23, 2004


By Robert Greenstein and Richard Kogan

PDF of report

View Related Analyses

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.

An amendment filed by Rep. Mark Kirk (R-Ill.) to the budget enforcement bill coming to the House floor on June 24 will be a major focus of debate.  The amendment would exempt tax cuts from fiscal discipline and allow unlimited future tax cuts, while requiring large cuts in domestic programs.  The amendment, which is in the form of a substitute for the legislation coming to the House floor (H.R. 3973), contains an array of provisions.  Four provisions stand out:

Entitlement Cuts Over 10 Years Under Kirk
Amendment, If All Entitlements Are Cut Proportionately

 (in billions of dollars)



Federal civilian retirement and disability


Unemployment Compensation


Military retirement and disability


Supplemental Security Income


Veterans’ benefits


Earned Income Tax and Child Tax Credits


Food Stamps


Family Support


Child Nutrition


Commodity Credit Corporation


Other federal retirement and disability


TRICARE for Life


Foster Care and Adoption Assistance


Student loans


Universal Service Fund


State Children's Health Insurance


Social services (Title XX, Voc Rehab)


Other miscellaneous







CBO data show that the Kirk entitlement cap would force entitlement cuts of $445 billion over the next ten years.  If all entitlements were cut by the same percentage, Medicaid would have to be cut by $175 billion, military retirement and disability programs by $28 billion, veterans’ benefits by $22 billion, Supplemental Security Income for the elderly and disabled poor by $27 billion, the Earned Income Tax Credit and refundable Child Tax Credit for working poor families by $21 billion, and the school lunch and child nutrition programs by $10 billion.  (If Congress took no action and entitlements were cut entirely through automatic cuts, the specific amounts that would be cut in each program would be somewhat different, but the cuts still would total $445 billion.  It is unlikely that the cuts would be achieved primarily or entirely through automatic cuts; the resulting cuts in some popular programs would be so large as to make heavy reliance on the automatic cuts politically unacceptable.)

This is tantamount to assuming that the wages and the cost of health benefits for the men and women in the armed forces and all federal civilian employees will be frozen for ten consecutive years, along with the salaries and other costs of private contractors who perform work for the government.  If this proposal becomes law, Members who seek simply to keep funding up with inflation so that services in vital programs do not have to be cut will likely be branded as big spenders who are seeking to enlarge the size of the government, since their proposals would increase spending over the official baseline.

These provisions are described in more detail below.


1.         Pay-As-You-Go Rules

The Kirk amendment would impose pay-as-you-go rules for the next three years on entitlement programs while exempting tax cuts from these rules.  By contrast, under the pay-as-you-go rules established on a bipartisan basis in 1990 with the support of the first President Bush, entitlement increases and tax cuts both had to be paid for.  Under the rules that were in place and proved effective in the 1990s, entitlement increases and tax cuts were treated the same — they each had to be offset, and their costs could be offset either through reducing other entitlements or through raising more tax revenue.

By contrast, under the Kirk substitute, as under the House Budget Committee’s bill, tax cuts are protected.  They would not have to be paid for.  Moreover, entitlement improvements could be paid for only by cutting other entitlements.  Measures to close abusive tax shelters or other tax avoidance schemes and use the savings to finance improvements in an entitlement benefit would not be allowed.

This proposal ignores calls from Federal Reserve Chairman Alan Greenspan and a host of budget watchdog groups such as the Concord Coalition and the Committee for Economic Development (an organization of corporate executives) to apply the pay-as-you-go rules to both tax cuts and entitlement increases.  The proposal would leave the door open to unlimited, costly tax cuts that raise deficits to higher levels.

The proposal poses several other serious problems as well:


2.         The Entitlement Cap

The Kirk amendment would establish an entitlement cap, which would limit total expenditures for entitlement programs other than Social Security and Medicare (and interest payments on the debt).  Congress and the President would have to enact legislation that cut entitlement programs deeply enough to fit within the cap.  If they did not, automatic cuts in entitlement programs would be triggered.

The Kirk amendment sets its entitlement caps well below what entitlement programs will cost under current law.  The Congressional Budget Office issues entitlement-cost projections and data that enable analysts to compute what the entitlement caps would be under the Kirk amendment.  These CBO projections and data show that over the next ten years, the entitlement caps would be a total of $445 billion below what the entitlement programs will cost under current law.  In other words, the Kirk amendment mandates cuts of $445 billion in entitlement programs over the coming decade.

The principal reason that the Kirk proposal requires cuts of this magnitude is that the formula it uses to set the caps is seriously flawed with regard to Medicaid.  Under the Kirk proposal, the entitlement cap for each year would equal the sum of the costs of all entitlement programs (except Social Security and Medicare) in the prior fiscal year, with two adjustments.  One adjustment would be made for projected increases or decreases in the number of people eligible for each entitlement program.  The second adjustment would incorporate cost-of-living adjustments required by statute.  If a program has no statutory COLA, an adjustment would be made equal to the projected increase in the Consumer Price Index.

Medicaid has no statutory cost-of-living adjustment.  Its costs rise with increases in the cost of health care.  As is well known, health care costs are rising rapidly in the private and public sectors alike.  The Kirk entitlement cap assumes, however, that Medicaid costs per beneficiary will rise an average of only 2.2 percent per year, the projected rate of increase in the CPI. 

Hardly any employer in America can hold increases in health insurance premium costs to 2.2 percent per beneficiary per year; basic health care costs are climbing much faster than that.  Almost entirely for this reason, the entitlement cap would be set $445 billion below projected entitlement costs.  And as a consequence, not only Medicaid but all entitlements other than Social Security and Medicare would be at risk of large cuts.

The exact size of the cuts in each program would depend on decisions that Congress and the President would make.  In theory, Congress and the President could decline to enact any legislation cutting entitlement programs and let automatic entitlement cuts do all the “dirty work.”  Under the rules for automatic cuts contained in the Kirk amendment, a few programs (such as civil service retirement and food stamps) would be exempt from the automatic cuts, and some other programs (such as veterans’ programs and Medicaid) could be cut no more than two percent per year through an automatic cut.  (Note:  These programs would be cut an additional two percent each time an automatic cut occurred, so the cuts in these programs could mount over time.)

Broad-based Opposition to Entitlement Cap Proposal

A broad array of organizations have expressed strong opposition to entitlement cap proposals.  For example, in a letter to Speaker Dennis Hastert on June 22, the American Legion stated:  “The American Legion opposes any and all entitlement cap proposals.  Although we fully support deficit reduction, we consider an entitlement cap in any form to be the wrong approach, and a potential breach of national trust.”  Similarly, the Paralyzed Veterans of America wrote in a letter to Members of Congress on June 22:  “PVA would also like to urge you to oppose any proposed amendment that would enact caps on entitlement spending.”  In a strong letter sent on June 21, the AARP stated:  “AARP urges you to reject any entitlement caps because they would jeopardize the health and economic security of millions of vulnerable Americans.”

But it is extremely unlikely that the bulk of the cuts would occur through automatic cuts; the automatic cuts are designed to be so unpalatable that Congress and the President would be compelled to enact legislation cutting entitlements in order to avert or minimize the automatic cuts.  Indeed, if all of the reductions needed to comply with the Kirk entitlement cap were made through the automatic cuts, the programs fully subject to the automatic cuts — programs such as TriCare military health benefits, farm-price supports and crop insurance, extended unemployment benefits and trade adjustment assistance, the Earned Income Tax Credit, vocational rehabilitation, child care payments to states, and the Social Services Block Grant (Title XX) — as well as the salaries of Member of Congress and Senators — would have to be cut a whopping 43 percent by 2014.

It is inconceivable that Congress would sit idly by and allow these programs — and Members’ own salaries — to be cut nearly in half.  Congress would seek to spread the pain more broadly by enacting legislation that cut more heavily into entitlement programs that have some protection from the automatic cuts.

The bottom line is that all entitlement programs except Social Security and Medicare would be at serious risk of being subject to substantial cutbacks, since $445 billion in reductions over ten years would be mandated by law.  The table on page 1 shows the ten-year cuts in each program if all entitlements other than Social Security and Medicare are cut by the same percentage.


3.         Lowering the Budget Baseline for Discretionary Programs

For the past 30 years, the Congressional Budget Office and the Office of Management and Budget have taken inflation into account in developing their budget baseline estimates.  Specifically, the baseline for appropriated programs equals the current funding levels for these programs, as adjusted for inflation for future years.

CBO and Members of Congress have long recognized that adjusting for inflation is necessary if programs ranging from defense to health research to child care are to maintain the same level of services from year to year.  This is the case because the salaries and health benefits for military and civilian personnel — and the costs of the goods and services purchased through these programs — rise over time with inflation.

The Kirk amendment, however, would require CBO to stop adjusting for inflation and instead to assume a hard freeze for discretionary programs for ten straight years.  Such a change would be ill-advised.  Adjusting a program’s funding level to keep pace with inflation does not mean the program is expanding, serving more people, or providing richer benefits or services.  Compensating for the effects of inflation simply reflects the basic reality that the cost of delivering government-funded services rises with inflation, just as the cost of providing services in the private sector does.  Costs for the salaries and benefits of those who deliver these services increase with inflation.  So do the costs of materials, supplies, and other goods and services purchased through government programs.

Indeed, some budget experts, such as Urban Institute president and former CBO director Robert Reischauer, have argued that adjusting domestic discretionary programs for both inflation and population growth would provide a better estimate than the current inflation-only adjustment of what it costs to provide the same level of services over time.  Moreover, George W. Bush, as a Presidential candidate, made the same point, arguing that an “honest comparison” of the rate of spending growth in Texas during his tenure as governor must adjust for both inflation and population growth.[1]

Effects of the Kirk Proposal

Even with the low inflation rate that CBO now projects, a baseline with no adjustment for inflation would reflect deep reductions over time in the levels of services that discretionary programs provide.

This reduction of 2.7 percent of GDP would be greater than the entire increase in Social Security costs that the Social Security actuaries and CBO project over the next 75 years.

A Ten-year Freeze Would Sharply Curtail Services Provided By Discretionary Programs

The Congressional Budget Office estimates that if discretionary funding were frozen at the current level for ten years, discretionary programs would be reduced overall by about one-fifth in 2014, compared to the levels reflected in the current, inflation-adjusted CBO baseline.*  At these lower funding levels, services provided by important discretionary programs would have to be cut substantially. 

  • Women, Infants and Children (WIC) nutrition program:  The WIC program provides a package of specific nutritious foods, nutrition counseling, and health care referrals to low-income pregnant women and low-income mothers and their young children.  The cost of providing these services rises as food prices increase and as the salaries of dieticians and other staff rise with inflation.  If funding for the WIC program were frozen for ten years, funding in 2014 would be $867 million below the level in the current CBO baseline.  At this lower funding level, the program would serve an estimated one million fewer women and children than would be served under current baseline levels.
  • Head Start:  The Head Start program provides early childhood education for preschool children as well as some infants and toddlers.  The cost of Head Start rises as inflation raises teachers’ wages and salaries, the cost of materials, and the cost of renting or maintaining facilities.  If Head Start funding were frozen, funding in 2014 would be $1.3 billion below the current baseline level, and an estimated 134,000 fewer children would be served by the program at this freeze level than at the inflation-adjusted level.
  • National Institutes of Health:  The cost of conducting research at the National Institutes of Health rises with the wages and salaries of researchers and the cost of medical technology and procedures.  If the NIH budget were frozen at the current level, funding in 2014 would be $5.7 billion below the level needed to maintain the inflation-adjusted 2004 level, a 17 percent cut in real funding.  At this level, NIH would have to cut back substantially the research that it supports.
  • Education for the Disadvantaged:  Title I of the Elementary and Secondary Education Act provides funding for school districts to improve educational outcomes for low-income and other disadvantaged children.  If funding for this program were frozen at current levels, by 2014 funding would fall $2.6 billion, or 17 percent, below the level needed to support the programs and services now being funded through this program.
  • National Resources and Environment:  This is a broad grouping of programs related to environmental protection, national parks, and the management of the nation’s land, water, and mineral resources.  If funding for these programs were frozen, funding in 2014 would fall $8.7 billion, or 22 percent, below the level needed to continue current programs and services.

*Although the overall reduction is 19.7 percent in 2014, the exact percentage reductions for individual programs would vary somewhat depending on the proportion of spending in each program on federal employee salaries and benefits.  This variation would reflect the different inflation-adjustment measures that CBO uses for personnel versus other types of spending.


Proposed Change Could Affect Budget Debates and Decisions

If this proposal is adopted and a “freeze baseline” becomes the official benchmark for measuring the impact of policy proposals, efforts simply to keep funding for discretionary programs in line with inflation will likely be portrayed as increasing costs and enlarging government, despite the fact that doing so would merely maintain services at their current levels.  Even proposals to set discretionary funding well below the levels needed to keep pace with inflation but modestly above a freeze level — which would require cuts in services over time —would likely be portrayed as expanding government and increasing spending.  Similarly, long-term freezes that entail deep cutbacks in services would be portrayed as not being cuts.

This portrayal of discretionary spending — which reflects the type of presentation of budget trends that various arch-conservative organizations have long advocated — would acquire an official sanction and likely be reflected in media coverage as a result.  Over time, such presentations would likely have an effect on politics and policy, making it more difficult to fund discretionary programs adequately and leading to damaging cuts in effective programs over time.

The proposal also could have another deleterious effect.  Congress would be unlikely to freeze defense, homeland security, veterans health care, and some other popular programs for ten straight years.  Although, as just noted, overall discretionary spending is likely to be pushed downward as a result of the effects of this proposal, a baseline assumption of a ten-year freeze in every area of discretionary spending would be unrealistic.  As a consequence, baseline budget projections that assume a ten-year freeze on all discretionary programs would underestimate discretionary spending — and thereby understate projected budget deficits.  The rosy deficit forecasts that would result could lull policymakers into thinking the deficit problem was smaller and more manageable than would be the case and could thereby encourage more unaffordable tax cuts or other profligate budgetary actions.


4.         Ill-advised Changes in Debt Limit

The Kirk amendment also would change the debt limit.  Federal law has long placed a limit on the size of the federal debt, with the limit applying to gross Treasury debt, the amount the Treasury has borrowed from the public plus the amount the Treasury has borrowed from federal trust funds such as the Social Security and Civil Service Retirement trust funds.

The Kirk amendment would redefine the debt to which the debt limit applies to be the net debt rather than the gross debt.  It would reduce the statutory debt limit accordingly.  The amount the Treasury borrows from federal trust funds would no longer be subject to the debt limit.  Only the debt held by the public — the amount of money the government has borrowed from banks, states, foreign governments, and private and institutional investors — would be subject to the limit.

Looked at purely as a matter of budgetary analysis, the idea of focusing on the net debt — the debt held by the public — has merit.  But it would be a serious mistake for Congress to place a statutory limit on the net rather than the gross debt.  If the definition of the debt subject to statutory limit becomes the net debt, as proposed, the likelihood of an unprecedented default by the federal government would grow. 

Why the New Definition Would Risk Default

For decades, Congress has had difficulty in collecting sufficient votes to raise the debt limit.  (This difficulty should not be confused with fiscal prudence; Congress frequently finds more than enough votes to cut taxes or raise program costs.)  When the gross debt increases and the debt limit is reached before Congress acts to raise it, the Treasury has on occasion found it necessary to use a gimmick to avoid breaching the limit and defaulting.

The gimmick involves temporally “de-investing” a modest part of the Civil Service Retirement trust fund.  This means that the Treasury borrows from that trust fund without counting that borrowing against the debt limit and without paying interest on the borrowing.  The gimmick helps the Treasury continue to operate without defaulting.  The gimmick harms no one, as the Civil Service Retirement trust fund still has more than enough income to pay all benefits on time.  And when the Congress finally raises the debt limit, the trust fund is made whole and the lost interest is restored.

The problem with the proposed change in the definition of the debt that is subject to the debt limit is that the maneuver used to avoid a default when Congress has not acted in time would no longer work.[2]  If the debt limit were reached and the government continued to run a deficit, the Treasury would have no maneuvering room — and would not be able to raise the cash to pay its bills.  For the first time in U.S. history, government checks would bounce, and the U.S. government would default.

If a default occurred, those who deal with the United States in the future — from military contractors to those who lend to the government — would demand higher prices because of the perception that the risk of default, previously thought to be zero, was now recognized as being higher than that.  Even if the increase in interest rates on Treasury securities were as little as 10 basis points (one-tenth of a percentage point), the cost of the default would still be an extra $4 billion per year, and that amount could grow for decades.  This would be a steep price to pay to change the definition of the debt subject to the debt limit.



The Kirk amendment contains a number of deeply troubling provisions that could have far-reaching and deleterious effects on policy.  More balanced and effective approaches to restoring fiscal discipline should be pursued.  As a recent joint statement by the Concord Coalition, the Committee for Economic Development, the Committee for a Responsible Federal Budget, and the Center on Budget and Policy Priorities notes, the first step is to stop “digging the hole deeper” by applying pay-as-you-go rules to entitlement increases and tax cuts alike.

End Notes:

[1] Dallas Morning News, October 28, 1999.

[2] Treasury would no longer be able to use the gimmick of temporarily “disinvesting” part of the Civil Service Retirement trust fund.  Disinvesting means, in effect, canceling the government’s debt by declaring that the securities that have been issued are no longer valid.  While the Treasury can briefly disinvest the Civil Service Retirement trust fund and make it up later, the Treasury cannot cancel securities held by the public.   Doing so would itself constitute a default.