Analysis of changes enacted during the 104th Congress shows that a highly disproportionate share of the budget cuts made over the last two years has fallen on programs for low-income families and individuals. This analysis is based on a examination of Congressional Budget Office estimates of the fiscal effects of all legislation enacted in the past two years.

• Entitlements: More than 93 percent of the budget reductions in entitlements have come from programs for low-income people. CBO estimates that legislation enacted during the 104th Congress reduced entitlement programs by $65.6 billion over the period from 1996 to 2002. Of that amount, almost $61 billion comes out of low-income entitlement programs. The food stamp program and the Supplemental Security Income program for the elderly and disabled poor bore the largest reductions.

Although entitlement programs for people with low incomes absorbed 93 percent of the entitlement reductions, they accounted for only 37 percent of total expenditures for entitlement programs other than Social Security and only 23 percent of all entitlement spending. Entitlement programs for individuals with low incomes thus bore a highly disproportionate share of the entitlement reductions enacted during the 104th Congress.

• Programs Other than Entitlements: Similarly, low-income programs that are not entitlements absorbed a disproportionate share of the reductions in the non-entitlement side of the budget. Slightly more than one-third of the reductions in nondefense programs that are not entitlements — 34 percent — came from non-entitlement programs for people with low incomes. These low-income programs accounted, however, for only 21 percent of overall funding for nondefense discretionary programs at the start of the Congress. They thus bore more than one and one-half times their share of the nondefense discretionary cuts.

These reductions in nondefense programs that are not entitlements (i.e., nondefense discretionary programs) are measured by comparing the fiscal year 1995 funding levels for these programs when the 104th Congress took office, as adjusted for inflation since 1995, to the funding levels for these programs in fiscal year 1997. This comparison shows that overall appropriations for discretionary programs for low-income households were reduced more than 10 percent over the past two years, while appropriations for other nondefense discretionary programs shrank five percent.

The Congressional leadership measures changes in discretionary programs in a different manner — by comparing funding levels in one year to funding levels in a subsequent year without any adjustment for inflation. This approach does not count as a reduction the erosion in services that occurs when inflation shrinks the purchasing power of the funds on which a program runs. Under this approach to measuring program cuts and increases, all of the reductions in nondefense discretionary funding during the 104th Congress came in low-income programs. Total funding for nondefense discretionary programs for households with low incomes was lower in fiscal year 1997 than it had been in fiscal year 1995 when the 104th Congress took office. But overall funding for other nondefense discretionary programs was slightly higher in fiscal year 1997 than it had been in fiscal year 1995.

• Combining entitlement and non-entitlement programs: At the start of the 104th Congress, CBO projected that low-income programs — including both entitlement and non-entitlement programs — would constitute 23 percent of the total nondefense budget for the period from 1996 through 2002. But low-income programs absorbed 53 percent of the overall reductions in nondefense programs for this period if the cuts in discretionary programs are measured from the 1995 level adjusted for inflation. (If the cuts in discretionary programs are measured from the 1995 level without adjustment for inflation, the low-income programs absorbed 89 percent of the overall reductions in nondefense programs.)
  
• Low-Income Households Affected: The disproportionate reductions in programs for low-income families and individuals will affect virtually all types of low-income households. This includes low-income working families with children and the low-income elderly and disabled.

Entitlements

Legislation enacted during the 104th Congress reduced entitlement programs for which households must have low incomes to qualify by $61 billion over the period from 1996 through 2002. Entitlement programs not targeted on low-income households were reduced only $4.6 billion.¹ (See Table 1.) Thus, 93 percent of the entitlement reductions were absorbed by low-income programs. In fiscal year 2002, some 86 percent of the entitlement reductions will be borne by the low-income programs.

These reductions reflect Congressional Budget Office estimates of the effects of changes in legislation enacted during the 104th Congress. The CBO estimates represent the amounts by which legislation enacted during the past two years reduced the amounts these programs otherwise would have expended.

The most dramatic reductions in entitlements resulted from the welfare legislation. Congress and the President also enacted more than 40 other bills, however, that changed entitlement law and affected U.S. households. More than 25 changes in entitlement programs were accomplished through "legislative riders" attached to appropriation bills. Table 2 shows the changes made in the entitlement programs for low-income people.

It is difficult to quantify precisely the effects of such a sweeping array of changes in a large number of programs that provide benefits to low-income families and individuals. A noted Urban Institute analysis of legislation very similar to the welfarebill subsequently enacted into law, however, forecast that the legislation would increase the number of poor children by 1.1 million and the number of poor people of all ages by 2.6 million. The analysis also estimated the legislation would make large numbers of poor families poorer, with 8.7 million families with incomes below 150 percent of the poverty line losing an average of more than $1,000 a year each in assistance.

The Urban Institute analysis also sheds light on the extent to which the reductions in low-income entitlement programs are likely to be spread across various categories of low-income households, including low-income working families. For example, the Institute's analysis found that 70 percent of the 1.1 million children whom the welfare legislation was expected to add to the ranks of the poor would be children in working families. The study also projected that several million low-income families that work and receive no cash welfare assistance would lose an average of $970 a year due to the legislation.

The Urban Institute's analysis found the legislation would affect significant numbers of the elderly and disabled, as well as families on public assistance. All of these groups would be affected by such changes as the welfare bill's across-the-board reductions in food stamp benefits. In addition, the bulk of the reductions the welfare bill makes in assistance for legal immigrants will affect immigrants who are elderly or disabled.

These figures are conservative for several reasons. First, Table 2 is limited to changes in entitlement programs that are primarily or exclusively for households with low incomes. Some of the reductions made in programs not classified as low-income entitlements in this analysis, however, also affect poor people. For example, federal expenditures for Federal Housing Administration single-family mortgage assistance were reduced $1 billion by restricting the amount of forbearance the government grants a household that falls behind its mortgage payment schedule. Since the FHA is targeted on households with more modest means than most homeowners — and since those who cannot meet their mortgage payments are often those who have lost their jobs and suffered a reduction in income — a significant portion of the cut in this program likely affects low-income families. Nevertheless, Table 2 does not treat reductions in FHA mortgage assistance as a reduction in a "low-income" program.

In another example, the Contract with America Advancement Act denied various benefits to low-income individuals disabled by substance abuse. Table 2 includes the resulting reductions in SSI, Food Stamps, and Medicaid. Some of these individuals also became ineligible for Social Security disability insurance and Medicare benefits. The benefit reductions in Social Security and Medicare are not counted here as reductions in low-income programs; to the contrary, they are counted as reductions in non-low-income entitlements. (The elimination of Social Security and Medicare for individuals disabled by substance abuse will result in a reduction in expenditures for those programs of $2.7 billion over the period from 1996 to 2002. About half of those affected by this policy change have low incomes.)

Furthermore, the figures cited here reflect only reductions in federal funding for these programs. Among the most profound changes the welfare bill makes is that it enables states to withdraw substantial amounts of state funding from income support and work programs for poor families with children without such action triggering any loss of federal funds for these states. States can withdraw a total of up to $32 billion in state funding from these programs over the next six years without losing any federal money.

Many experts believe this provision of the welfare legislation will result in a substantial withdrawal of state funds over time. In what may be a harbinger of future developments, New York Governor George Pataki earlier this month proposed a welfare plan under which New York would withdraw substantial amounts of state money. The plan includes deep reductions in public assistance benefit levels; some families ultimately would face benefit cuts of as much as 45 percent. This analysis does not include any reductions in state funding for low-income programs induced by legislation enacted during the 104th Congress.

Finally, the figures in Table 2 reflect changes only in entitlement benefits and not in tax benefits. If legislation enacted in the past two years that changed tax benefits also is taken into account, the highly unequal burden of deficit reduction over the past two years becomes slightly more pronounced. Over the period 1995 through 2002, the share of net reductions borne by low-income households rises from 93 percent when only entitlement are considered to 95 percent when both entitlements and tax benefits are considered. (See Appendix A.)

Discretionary Spending

At $51.7 billion, funding in fiscal year 1997 for discretionary programs that primarily or exclusively serve low-income families and individuals is $5.9 billion below the 1995 funding level in place at the start of the 104th Congress, as adjusted for inflation.³ This represents a funding reduction of 10.2 percent for low-income discretionary programs. By comparison, other non-defense discretionary programs were reduced an average of 5.2 percent. (See Table 3.) ⁴

The low-income programs constituted 21 percent of all non-defense discretionary appropriations at the start of the 104th Congress. ⁵ But the $5.9 billion removed from these programs represents 34 percent of the $17.3 billion total that was taken from non-defense discretionary appropriations as a whole.

If the changes in appropriations for discretionary programs are measured from the original 1995 level without any adjustment for inflation, the degree to which the low-income reductions are disproportionate appears still greater. Measured in this fashion, the reductions in funding for low-income discretionary programs account for all (i.e., 100 percent) of the reductions in nondefense discretionary appropriations during the 104th Congress. (Total fiscal year 1997 funding for nondefense discretionary programs other than the low-income programs is slightly higher than the 1995 level when no adjustment for inflation is made. If inflation is ignored, total funding for the low-income discretionary programs declined during this period, while total funding for the other nondefense discretionary programs slightly increased. See Table 4.)

Some types of low-income discretionary programs were hit considerably harder than others. Low-income housing and employment and training programs were cut most deeply. Funding for low-income education programs registered a modest increase. (See Appendix B for an analysis of the changes in various categories of low-income discretionary programs.)

As in the entitlement area, the reductions in low-income discretionary programs affect a wide array of low-income people. For example, reductions in the summer jobs program affect low-income youth. Reductions in the low-income home energy assistance program affect all groups of the poor, especially the low-income elderly who rely disproportionately on this program; in many states, that program also serves large numbers of working poor families and families on public assistance. Reductions in homeless assistance programs affect those already homeless or at risk of becoming so.

The reductions in low-income discretionary programs resulted from either the rescissions bill enacted in August 1995 or the fiscal year 1996 appropriations legislation enacted in April 1996. The appropriations legislation for fiscal year 1997 reversed a small portion of the previous reductions made in low-income discretionary programs.

Congressional Budget Office figures show that when the 104th Congress started, low-income entitlement and discretionary programs combined were projected to account for 23 percent of all nondefense spending over the period from 1996 through 2002. These programs will absorb 53 percent of the nondefense reductions over this period that result from changes enacted during the 104th Congress.

The figures just cited measure changes in discretionary programs from the original 1995 level, adjusted for inflation. If the changes in discretionary programs are measured from the 1995 level without an adjustment for inflation, the reductions in low-income programs account for 89 percent of all of the nondefense reductions. ⁶

If the defense budget is added to the equation, both the proportion of the budget that low-income programs constitute and the proportion of the cuts that they absorbed decline somewhat. But the basic story remains unchanged — no matter how the changes are measured, low-income programs bore a highly disproportionate share of the reductions enacted during the past two years.

CBO projected at the start of the 104th Congress that low-income programs would constitute 19 percent of the total federal budget — including defense spending and all other discretionary programs and entitlements — during the period from 1995 through 2002. But the low-income programs will absorb 37 percent of all cuts in all parts of the budget, including the defense budget, that will occur during this period as a result of decisions made over the past two years.

Here, also, these figures measure changes in discretionary programs from the original 1995 level adjusted for inflation. If no inflation adjustment is made, the low-income programs will absorb 100 percent of the reductions in federal programs enacted during the 104th Congress. The reason that the percentage of the cuts borne by the low-income programs becomes so much larger when the reductions in discretionary programs are measured from the 1995 level without an adjustment for inflation is that defense appropriations for fiscal year 1997 are higher than they were for fiscal year 1995 when the 1995 level is not adjusted for inflation, but lower than the 1995 level when that level is adjusted to reflect inflation over the past two years. Stated another way, defense funding increased slightly during the 104th Congress under one approach to measuring changes in funding for discretionary programs and decreased slightly under the other approach to measuring such changes.

Legislation enacted by the 104th Congress changed tax law as well as entitlement law. Although no general changes were made in tax rates, a number of special exemptions or deductions from the tax code were increased or decreased.

Legislation of this sort effectively changes "tax benefits" relative to existing law. The net effect of changes in tax law enacted in the past two years has been to grant additional tax benefits.

Table 5 displays the effects of tax legislation enacted during the past two years. It shows that the one new tax provision designed primarily to assist low-income households — the reinstatement of the Work Opportunities Tax Credit — provides less in tax benefits than tax provisions enacted during this period that benefit other households. The tax provisions benefitting other households are principally found in the Self-Employed Health Insurance Act of 1995.

This appendix discusses the changes in some of the major low-income discretionary programs. The table at the end of this appendix provides a comprehensive listing of all changes in funding levels for low-income discretionary programs since fiscal year 1995.

Employment and Training Programs

FY 1997 appropriations for employment and training programs targeted on low-income individuals total $3.632 billion. This level is 22 percent — or approximately $1 billion — below the FY 1995 appropriations level at the beginning of the 104th Congress, as adjusted for inflation.

• The Adult Training Grants Program under the Job Training and Partnership Act (JTPA) is funded at $895 million in FY 1997. This level is 20 percent below the FY 1995 pre-rescission level, adjusted for inflation, and represents a substantial reduction in financial resources for employment and training opportunities for disadvantaged adults. Under the provisions of the JTPA, 90 percent of the funds for Adult Training Grants are targeted on welfare recipients.
  
• The Summer Youth Program is funded at $871 million in FY 1997, or 22 percent below the FY 1995 level before recessions, as adjusted for inflation. Like some other programs, this program was cut more deeply than this in 1996; a portion of the cut was restored in the FY 1997.
  
• The year-round Youth Training Grants Program is funded at $127 million in fiscal year 1997. This is 80 percent below the FY 1995 pre-rescission level, as adjusted for inflation.
• The Job Corps program was spared the sharp cuts made in other major employment and training programs. The FY 1997 appropriation of $1.154 billion for this program is less than one percent — or $10 million — below the FY 1995 pre-rescission level, as adjusted for inflation.
  
• Funding for the Older Americans Employment Program was increased to $463 million in FY 1997. This is six percent above the FY 1995 pre-rescission level, as adjusted for inflation.

Housing Programs

Overall FY 1997 appropriations for low-income housing and housing-related programs (such as the low-income energy assistance program) are 25 percent below the original fiscal year 1995 level, adjusted for inflation.

• Funds for the repair and modernization of public housing projects were decreased from $3.7 billion to $2.5 billion in FY 1997, a reduction of approximately 36 percent, when inflation is taken into account. FY 1997 funds for operating subsidies for local public housing authorities are six percent below the original FY 1995 appropriation, adjusted for inflation.
  
• A total of $550 million is provided in the FY 1997 appropriation for revitalization of severely distressed public housing, including the demolition of dilapidated public housing, revitalization of sites, and relocation of the displaced tenants. This is an increase of four percent — or $20 million — above the FY 1995 pre-rescission level, adjusted for inflation.
  
• Appropriations for various other housing-related programs also were cut. For example, funding for Homeless Assistance Grants was reduced 31 percent.
  
• In addition, the FY 1997 funding for the Low-Income Energy Assistance Program (LIHEAP) — $1 billion — is 36 percent beneath the FY 1995 pre-rescission level, after adjusting for inflation.

Education Programs

Under the FY 1997 appropriations legislation, total funding for education programs for children and youth from low-income families was increased to $20.4 billion. This represents a five percent increase above the FY 1995 pre-rescission level, as adjusted for inflation. Head Start and Pell Grants received most of the increased funding.

• The FY 1997 appropriations legislation increased funding for Head Start to $3.981 billion. This is six percent above the FY 1995 pre-rescission level, as adjusted for inflation.
  
• The appropriation for the Pell Grant program in fiscal year 1997 was increased to $5.919 billion, 13.2 percent above the FY 1995 pre-rescission level, as adjusted for inflation.⁷ The maximum grant award is set at $2,700 in the FY 1997 appropriations bill. This represents a nine percent increase, after adjusting for inflation, above the level provided in the FY 1995 appropriation legislation.
  
• Fiscal year 1997 funding for other student financial assistance — $1.641 billion — is six percent greater than the FY 1995 pre-rescission level, as adjusted for inflation.
  
• The $7.698 billion included in the FY 1997 appropriation for Compensatory Education for the Disadvantaged represent a 0.5 percent increase — or $38 million — above the FY 1995 pre-rescission level, as adjusted for inflation.

Nutrition Programs

Low-income discretionary programs include two nutrition programs of significant size — the Special Supplemental Food Program for Women, Infants, and Children (WIC) and the Commodity Assistance Program.

• The FY 1997 appropriation for WIC is $3.730 billion. This represents a 1.5 percent increase above the FY 1995 pre-rescission level, as adjusted for inflation.
  
• At $166 million, the FY 1997 appropriation for the Commodity Assistance Program constitutes a 17 percent reduction below the FY 1995 pre-rescission level, as adjusted for inflation. ⁸

Health Programs

A total of $4.434 billion was appropriated for low-income discretionary health programs in FY 1997. This is four percent lower than the FY 1995 pre-rescission level, as adjusted for inflation.

• At $802 million, the FY 1997 appropriation for Consolidated Health Centers is essentially the same as the FY 1995 level, adjusted for inflation. According to U.S. Department of Health and Human Services data, health centers funded under this appropriation provide health care to more than eight million people, including approximately 3.5 million children.
  
• The Maternal and Child Health Block Grant, which supports health care services for low-income mothers and children, is funded at $681 million in FY 1997. This is a six percent reduction below the FY 1995 inflation-adjusted level.
  
• At $1.806 billion in FY 1997, the Indian Health Services Program is essentially level-funded when compared to the FY 1995 appropriation, as adjusted for inflation.

Other Programs

Among other low-income discretionary programs for low-income families and individuals that experienced significant changes in funding levels are the legal services program and the Community Services Block Grant.

• The FY 1997 appropriation for the legal services program is $283 million. This represents a 36 percent reduction from the original FY 1995 level, as adjusted for inflation.
  
• On the other hand, at $490 million, the FY 1997 funding level for the Community Services Block Grant program is 18 percent above the original FY 1997 level, as adjusted for inflation.

(1) Student Financial Assistance — Pell Grants: in 1995, Congress appropriated $6.244 billion for Pell Grants, of which approximately $1.304 billion was unused at the end of the year. The amount actually spent in fiscal year 1995 was $4.940 billion. This is the number reflected in the table for FY 1995, prior to the rescission bill. In 1996, Congress appropriated $4.914 billion for the program; an additional $869 million was carried over from prior years, making a total of $5.783 billion available in FY 1996. As shown in the table, an estimated $5.407 billion of the available funds were actually spent in FY 1996. In 1997, Congress appropriated $5.919 billion. The actual amount ultimately spent in FY 1997 may be either higher or lower than the appropriated amount, depending on the number of eligible applicants and the amount of funding carried over from fiscal year 1996.

(2) Summer Youth Employment: prior to the enactment of the rescission bill in August 1995, this program was funded on a "forward funding" basis. The original pre-rescission level for 1995 shown in the table — $1.056 billion — represents the amount originally appropriated for the program. The rescission bill changed this program from a forward-funded program to a current-year-funded program. All of the forward funding for fiscal year 1996 was rescinded. This left available for FY 1995 some $682 million that had been appropriated in FY 1994 as forward funding for FY 1995, plus $185 million not rescinded from the 1995 appropriation, for a total of $867 million. Starting in 1996, the amount appropriated for this program represents the amount of funds available to the program for the current fiscal year.

(3) The figures for Section 8 housing assistance exclude the amounts appropriated for contract renewals in fiscal years 1995, 1996, and 1997 because the year-to-year fluctuations in the appropriation levels for that program are largely unrelated to program increases or decreases and reflect the timing of the expiration of existing contracts. In addition, the appropriations for tenant-based certificates and vouchers in fiscal years 1995 and 1996 provided for multi-year contracts, while the fiscal year 1997 appropriation provides only for one-year contracts. To make a valid comparison of funding levels for these three years, the amounts appropriated for fiscal years 1995 and 1996 for this program were adjusted downward to reflect the amounts that would have been needed if all vouchers and certificates had been funded for a one-year period. Had this adjustment not been made, the reduction in this program would have been overstated.

Appropriations for housing for the elderly and disabled have been excluded. Here, also, changes in the funding levels for these programs over the past several years primarily reflect factors not related to increases or decreases in the programs. The exclusion of these programs is likely to result in the table slightly understating the reductions in low-income discretionary programs.

(4) HOME Investment Partnership: includes a $15 million set-aside for Housing Counseling Assistance in FY 1997. In FY 1996, a $12 million set-aside for this activity was included in Community Development Block Grant. In FY 1995, Housing Counseling Assistance was originally funded at $50 million in a separate appropriation account. This was later reduced to $12 million by the FY 1995 rescission bill.

(5) Low Income Energy Assistance (LIHEAP): in addition to the regular LIHEAP appropriation, Congress has established a contingency fund that can be tapped at the Administration's request in cases of emergency. In FY 1995, an additional $100 million in contingency funds were used for summer cooling. In FY 1996, an additional $180 million in contingency funds were used for winter heating. It is unclear whether contingency funds will be used in FY 1997, and if so, in what amount. To date, the Administration has not submitted a contingency request for FY 1997. The table does not reflect contingency funds for any year.

(6) Community Development Block Grant (CDBG): the FY 1997 appropriation for CDBG includes a $30 million set-aside for Youthbuild, a $60 million set-aside for Lead-based Paint Hazard Reduction, a $60 million set-aside for Self-Sufficiency, and a $5 million set-aside for the Tenant Opportunity Program. In FY 1996, the CDBG included a $20 million set-aside for Youthbuild, $12 million for Housing Counseling, $53 million for Self-Sufficiency, and $15 million for the Tenant Opportunity Program; $65 million for Lead-based Paint Hazard Reduction was included in a separate appropriation account. In FY 1995, funds for Youthbuild, Lead-paint Hazard Reduction, and the Tenant Opportunity Program all were included in separate appropriation accounts.

(7) Commodity Assistance Program: the FY 1997 does not include $100 million in mandatory funding that the new welfare law provides for The Emergency Food Assistance Program (TEFAP). This $100 million is intended to provide a small offset to the much larger reductions in food stamp benefits contained in the welfare law. The increase in mandatory funding for TEFAP is included in the portion of this analysis relating to changes in entitlement programs.

(8) Child Care and Development Block Grant: the amounts appropriated for the child care block grant for each fiscal year are generally used in the succeeding fiscal year.

1. This analysis excludes the budgetary effects of asset sale legislation and of legislation changing the stream of income that federal deposit insurance agencies receive from premiums. Changes in the Earned Income Tax Credit (EITC) are officially classified partly as outlays and partly as revenues; all such EITC changes are shown as outlay changes in this analysis.

2. The figure shown here for food stamps includes changes in the program's rules contained in the welfare law's food stamp title as well as those in the title restricting assistance to legal immigrants. It does not include the welfare law's $600 million in new spending for commodity purchases; since that money does not go to the food stamp program, it is included in the "other" line on this table. This figure also does not include the $3.3 billion in additional food stamp costs that will result from the law's reductions in cash assistance, SSI, and child support. (For every $10 by which an eligible household's net income is reduced, its food stamp benefits rise by approximately $3.) These increases in food stamp benefits are shown in a separate line in this table. Finally, this figure does not include $345 million in food stamp cuts that Congress enacted simultaneously in the welfare law and the Agriculture Appropriations Act for Fiscal Year 1997. Those reductions are included in the figures for appropriations bills. If that cut were instead attributed to the welfare law, the food stamp reductions would total $27.7 billion and the reductions in the welfare law would total $54.7 billion.

3. The FY 1995 levels used here are those in effect when the 104th Congress convened in January 1995. Those levels were subsequently reduced by the fiscal year 1995 rescissions bill enacted in August 1995.

These levels — and all other figures in this analysis pertaining to funding for discretionary programs — exclude several low-income housing program whose year-to-year appropriations levels can fluctuate for reasons largely unrelated to program increases or decreases. Appropriations for renewals of section 8 contracts, housing for the elderly, and housing for the disabled are excluded from the analysis for this reason. In addition, the amounts appropriated for incremental section 8 vouchers and certificates for fiscal years 1995 and 1996 are reduced in this analysis to reflect the amounts that would have been needed if all vouchers and certificates had been funded for a one-year period rather than a multi-year period. Since all vouchers and certificates funded in the fiscal year 1997 appropriation are for one year, this adjustment to the 1995 and 1996 levels is necessary to make them comparable to the 1997 level and to avoid overstating the level of reductions in housing assistance. Had appropriations for these various housing programs not been excluded from the analysis or adjusted in this manner, the reductions in low-income discretionary programs — and the disproportionate degree to which budget reductions over the past two years were concentrated in low-income programs — would appear to be larger than this analysis finds. (The exclusion of figures for housing for the elderly and disabled may result in some understatement of the reductions made in low-income housing programs.)

4. The changes discussed here in discretionary programs between 1995 and 1997 reflect changes in appropriation levels. (Appropriation levels include obligation levels for transportation trust funds.) Changes in appropriation levels rather than in expenditure levels are used for two reasons. First, changes in appropriation levels over the past two years reflect the policy changes made by Congress and the Administration. This is less true of trends in expenditure levels for discretionary programs between 1995 and 1997 because the expenditure levels for many discretionary programs in a given year reflect funding decisions made in previous years; recent expenditure trends thus do not necessarily provide an accurate picture of recent funding decisions. Second, projected fiscal year 1997 expenditure levels are not currently available for many discretionary programs.

5. The low-income programs now constitute 20 percent of non-defense discretionary funding.

6. These calculations combine the funding reductions in non-entitlement programs with the outlay reductions in entitlement programs. It is ordinarily preferable to use outlay reductions for both entitlement and non-entitlement programs for reasons of strict comparability, but as discussed in footnote 4, outlay projections for 1997 and beyond based on the FY 1997 funding of non-entitlement programs are not available. Over long periods, however, outlay levels for non-entitlement programs will be quite similar to the funding levels for these programs.

7. The FY 1995 pre-rescission level for Pell Grants excludes $1.304 billion of funds that the states did not use in that year.

8. This level does not include an additional $100 million in mandatory funding for The Emergency Food Assistance Program (TEFAP) provided in the new welfare legislation. These additional funds are intended to provide a small offset to the much larger reductions in food stamp benefits contained in the new welfare law. The new mandatory funding for the TEFAP program is included in the entitlement portion of this analysis.