Revised July 21, 2003
NEW HUD DATA SHOW FAMILIES WilL LIKELY LOSE HOUSING VOUCHERS
IF CONGRESS Approves president’s budget requesT
Reduction of 184,000 in Number of Households Assisted Would
Primarily Affect Low-Income Working Families and Elderly and Disabled People
By Barbara Sard and Will Fischer
PDF of this full report
HTM of press release
PDF of press release
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.
Data collected by the Department of Housing and Urban Development (HUD) in April 2003 from state and local housing agencies that administer nearly all federal housing vouchers show that the Administration’s budget request for fiscal year 2004 is inadequate to fund all housing vouchers likely to be in use when the fiscal year starts in October 2003. Analysis of these data, which were not available at the time the President’s budget was prepared, shows that if the necessary funding above the President’s current request is not included in the appropriations bills this year and additional funds do not become available from other sources, approximately 184,000 vouchers in use serving low-income families at the beginning of the fiscal year likely will not be funded. Such a reduction would be unprecedented in the history of the voucher program. Efforts to avoid this problem would be greatly facilitated if the Administration, using this more recent data, submitted a budget amendment to Congress requesting the necessary funds.
If the projected funding shortfall occurs, it is likely that some families that now rely on vouchers to help pay their rent will lose assistance, placing these families at a high risk of eviction and, in some cases, homelessness. These cuts would fall primarily on low-income working families, the elderly, and the disabled, since these groups make up 70 percent of the population that the voucher program serves.
Many housing agencies will not impose their share of the 184,000 voucher reduction immediately. Agencies that do not impose the full reduction at the beginning of the year will have to reduce the number of families served by an even greater number later in the year to achieve the required cut in the average number of families served over the full year that is required by the level of funding requested by the Administration. If the overall reduction is imposed evenly over the course of the fiscal year the number of families with vouchers would be 368,000 lower by the end of the fiscal year than when the fiscal year begins. (The 184,000 figure represents the amount by which the average number of families assisted over the course of the fiscal year will be below the number assisted at the start of the year.)
The housing voucher program (sometimes referred to as the “Section 8” voucher program) is the largest federal low-income housing assistance program. Families use vouchers to rent modestly-priced apartments in the private market. Typically, a family pays 30 percent of its income for rent and utilities, and the voucher covers the rest.
Voucher Utilization and Costs in Fiscal Year 2003 Exceed Expectations
During the six months from August 2002 to January 2003, the number of families receiving housing voucher assistance rose by 60,000. In January 2003, local and state housing agencies were using 94.9 percent of authorized vouchers. If the rate of increase in voucher utilization evidenced in this six-month period (the latest period for which data are now available) continues for the remainder of fiscal year 2003, some 95.9 percent of authorized vouchers will be used by families for the fiscal year as a whole. This 95.9 percent average fiscal year 2003 utilization rate is more than two percentage points higher than the 93.5 percent rate that Congress had assumed would be the rate achieved this fiscal year.
The average cost per voucher also has exceeded Congressional estimates. The average reported cost per voucher in January 2003 was $549 per month, or $6,588 on an annual basis (including administrative fees). We estimate, based on the trend in costs during the six-month period, that the average cost per voucher for fiscal year 2003 as a whole will be $6,692. In finalizing the appropriations for fiscal year 2003, Congress had assumed that the average voucher cost this year would be $6,372. (The Congressional Budget Office had estimated the average cost at $6,608, close to what our analysis now estimates it will be.)
Based on the available data concerning the number of vouchers in use and their average cost, we estimate that $13.1 billion will be needed in fiscal year 2003, about $700 million more than Congress appropriated. This $700 million in funds will be drawn from unspent funds from prior years (including funds previously allocated to state and local administering agencies in addition to the carryover funds discussed in endnote 1) so there will be no shortfall in fiscal year 2003. But because the Administration’s budget relied on the availability of these carryover funds to cover a portion of the voucher program’s cost in fiscal year 2004, the use of some of these funds in 2003 will increase the size of the 2004 shortfall.
The shortfall in the Administration’s 2004 budget request for the voucher program is unrelated to the Administration’s proposal to convert the voucher program to a block grant to the states in fiscal year 2005. If the final funding level for fiscal year 2004 is similar to what the Administration requested in February, the level of assistance that the voucher program provides will have to be cut in 2004 regardless of whether the block grant proposal is approved.
Approximately $1.26 Billion Short of Amount Needed to Avoid Cutting Assistance Budget Request Falls
The annual funding needs of the voucher program depend on two factors:
- The average cost per voucher. Per-voucher costs are heavily influenced by market rents and by economic conditions that affect the incomes of voucher holders and therefore the amount that they are able to contribute toward rent. For example, when unemployment rises, the average income of voucher holders tends to decline, causing voucher costs to increase (since vouchers generally cover the difference between rental costs and 30 percent of a tenant’s income).
- The proportion of vouchers that are in use. At any given time, some vouchers are not in use, often because families are unable to find suitable housing where they can use their vouchers. The percentage of vouchers that are in use is sometimes referred to as the utilization rate.
The President’s budget for fiscal year 2004 requested $13.05 billion to fund existing housing vouchers. At the time that the Administration’s budget was released, no complete, up-to-date data on per-voucher costs or utilization rates were available. Evidence that was available at the time suggested that the Administration might have underestimated the average cost per-voucher and the utilization rate nationally in calculating its request for voucher funding, and that there was significant risk the funding level would be inadequate. But without more current data, it was difficult to estimate reliably the program’s funding needs or the size of a shortfall, if one existed.
In April 2003, HUD required state and local housing agencies that administer the voucher program to report data on both voucher costs and utilization for the six-month period from August 2002 to January 2003. HUD recently released these data to Congressional staff. These data are critical to estimating funding needs for fiscal year 2004. An analysis of these data shows that:
- The number of vouchers in use when fiscal year 2004 begins in October 2003 that will require renewal funding will be approximately 2,011,000, or 75,000 more than the 1,936,000 vouchers that the Administration’s budget request assumes will be in use when the fiscal year begins; and
- The average cost of these vouchers in fiscal year 2004 will be an estimated $6,871, including administrative fees. (This estimate is very similar to the estimate of $6,842 for fiscal year 2004 that the Congressional Budget Office used in its March 2003 baseline.)
- As a result, the amount of funding required to fund the vouchers that will be in use when the fiscal year begins will exceed all of the funding requested by the Administration for voucher renewals by $1.26 billion, equivalent to approximately 184,000 vouchers. 
- The level of funding in the Administration’s budget would be sufficient to fund only 86.7 percent of authorized vouchers, substantially below the 96.7 percent utilization rate indicated by the newly available data.
In addition, without more funding than this, Congress will not be able to continue an integral component of the new method for funding vouchers that Congress adopted this year: a central fund to cover unanticipated program costs and to enable state and local agencies to make fuller use of the number of vouchers that they have been authorized to administer. In recent years, Congress has repeatedly urged HUD and state and local housing agencies to increase the proportion of vouchers that are in use, so that vouchers do not sit unused while families remain on waiting lists for assistance. The newly released HUD data show that the voucher utilization rate rose during the period the data cover. This increase appears to be caused in substantial part by the success of new policies put in place at the national, state, and local level in recent years designed to raise the proportion of vouchers in use and by a loosening of housing markets in some sections of the country that has made it easier for families to find housing where they can use their voucher. It is likely that the trend toward higher utilization will continue in fiscal year 2004 if adequate funding is available. Moreover, HUD has the authority to put to use more of the vouchers that Congress has authorized by “reallocating” vouchers from agencies that have consistently failed to use them to agencies whose track record demonstrates they would be able to put the vouchers to use promptly.
If the President’s funding request is not increased, however, no further progress could be made in putting authorized vouchers to use. As a result, an additional approximately 95,000 authorized vouchers that are not likely to be in use at the start of fiscal year 2004 — but that potentially could be used to serve families if utilization continues to rise — would be defunded. These 95,000 vouchers are in addition to the 184,000 vouchers in use at the start of the fiscal year that, as mentioned earlier, also would have to be defunded.
Comparison of Voucher Funding Needed in Fiscal Year 2004
Cost per voucher in FY 2004
Funding available for voucher renewals
Number of renewal vouchers
Number of renewal vouchers not funded
Number of vouchers in use in October 2003 not funded
Percent of authorized vouchers that would be
Shortfall if FY 2004 utilization is at its expected level in October 2003
CBPP estimate based on new HUD data
and Local Agencies to Take Vouchers Away from Families that Rely on Them to Pay Rent May Force State
A shortfall of this magnitude will result in a substantial cut in voucher assistance to low-income families. The number of families assisted through the voucher program would drop from approximately 2.02 million in October 2003 to an average of approximately 1.84 million during fiscal year 2004. Moreover, if the reduction in families served were imposed evenly over the course of the year rather than all at once at the beginning of the year, the number of families served would need to be reduced to 1.65 million by September 2004 to achieve an average of 1.84 million during fiscal year 2004. (Some families leave the program each year as their incomes rise or for other reasons, and agencies may try to meet a portion of the required cut by gradually reducing the number of families served through such turnover.) In other words, the number of families assisted at the end of fiscal year 2004 would be 368,000 fewer than the number being assisted when the fiscal year began.
The great majority of the households with vouchers — seven out of ten according to the most recent available HUD data — are either working families with children or elderly or disabled households. As a result, the impact of such cuts in the voucher program would fall most heavily on these groups. If cuts were applied proportionately across all groups, the average number of households assisted during fiscal year 2004 would fall below the number assisted in October 2003 by the following amounts.
- 56,000 fewer working families would be assisted;
- 41,000 fewer disabled households would be served; and
- 31,000 fewer elderly households would be served.
For a list of the cuts that would be required in each state under the Administration’s budget, see table 2.
This reduction in the number of low-income households receiving housing assistance would come at a time when the need for such assistance is great. The economy remains weak, and in most locations, there are long and growing waiting lists for the voucher program. In the event of a funding shortfall of the size described here, it is likely that state and local housing agencies would not be able to serve any families from their waiting list. Under normal circumstances, some families leave the voucher program each month and are replaced with families on the waiting list. If families leave the program but are not replaced by new families from waiting lists, the number of families receiving assistance will steadily dwindle.
In addition, to accomplish a reduction of this magnitude it is likely that many state and local agencies would have no choice but to terminate the voucher subsidies for some families currently using vouchers to help pay their rent. Few of these households are likely to be able to retain their housing without voucher assistance, and few will be able to find other decent housing that they can afford. The small landlords who own the bulk of the apartments rented to voucher holders and depend on the income from these apartments also will be harmed economically.
Projected State Level Reductions in Voucher Assistance
If Funding Equals Administration’s Request
Number of Vouchers Projected to Be in Use in October 2003
Projected Reduction in Average Number of Households Assisted During Fiscal Year 2004 if President’s Request is Approved
District of Columbia
Note: Projections assume that reductions in assistance would be distributed proportionately among states based on the number of vouchers in use in each state, and that cuts within states would be distributed based on the proportion of voucher holders in different demographic groups nationally. State projections do not add up to national totals because national projections include some vouchers that were missing from the recently released HUD data and therefore could not be assigned to any state. Estimates of reductions in each state are based on the number of expiring vouchers needing renewal that are likely to be in use in each state in October 2003.
 This amount includes $1.07 billion that the Administration expected would be carried over from funds appropriated for the voucher program in previous years. (For an explanation of the amount requested by the President to renew expiring housing vouchers, see
Barbara Sardand Will Fischer, “President’s Budget Requests Insufficient Funding for Housing Vouchers in 2004,” Center on Budget and Policy Priorities, revised April 24, 2003, available on the internet at < http://www.centeronbudget.org/3-27-03hous.pdf>.) In calculating the potential fiscal year 2004 shortfall in voucher funding we have assumed that the amount of carryover funds available to renew expiring housing vouchers in fiscal year 2004 is equal to the amount projected by the Administration in its fiscal year 2004 budget request, minus the $500 million of those funds that were rescinded from program funding by the Fiscal Year 2003 Omnibus Appropriations Act, which was enacted after the Administration’s budget was released. That is, we assume that $570 million rather than $1.07 billion in prior years’ funding will be available, in addition to the new funding the President has requested, making a total of $12.55 billion available for the renewal of housing vouchers. The precise amount of carryover funds (or other unspent funds from within the HUD budget) available could be somewhat larger or smaller than this level, but the Administration has given no indication that this will be the case. Indeed, there may be little if any carryover funding available to supplement appropriations for fiscal year 2004; the new HUD data indicate that voucher utilization and per voucher costs substantially exceed the figures on which Congress based the fiscal year 2003 appropriation, with the result that much of the carryover funding could be used up in this fiscal year. If the amount of additional funds available is lower than assumed, the budget shortfall and resulting cuts in the voucher program will be greater than projected in this analysis. If the amount of additional funds available is larger than assumed, the shortfall and resulting cuts will be smaller. (See box on page 2 for a related discussion.)
 HUD compiled data from 2,449 of the approximately 2,550 local and state agencies that have voucher contracts with HUD. These agencies administer 1,942,000 of the approximately 2,066,000 vouchers awarded to local and state agencies as of
January 2003 (including vouchers funded under multi-year contracts). An additional 94,955 vouchers are administered by 16 agencies that receive voucher funding under different rules, as part of the Moving to Work (MTW) demonstration. Using other data sources, we have been able to include these MTW vouchers in this analysis. For the approximately 29,000 authorized vouchers for which we have no data – only 1.4 percent of the total number of authorized vouchers – we applied the utilization and cost trends in the HUD data. For a description of our methodology see.“Estimating The Shortfall In Requested Voucher Funding For Fiscal Year 2004,” July 21, 2003, available on the internet at < https://www.cbpp.org/housingvoucher.htm>.
 This conclusion assumes that: 1) the cost and utilization trends for the six-month period for which HUD has just provided data will continue through the end of fiscal year 2003 (but no agency will be funded for more than its authorized number of vouchers over the full year), and 2) during fiscal year 2004, utilization will not increase further and per-voucher costs will increase at the same rate (2.5%) as HUD projects that the fair market rents used to determine how much rent a voucher can cover will increase from 2003 to 2004.
 Based on the HUD data concerning the leasing of vouchers under contract to housing agencies through
January 2003, plus the 94,955 MTW vouchers discussed in endnote 2, we estimate that 96.7 percent of these vouchers will be in use in October 2003. Some or all of the 43,000 tenant protection vouchers initially funded in fiscal year 2003 also are likely to be under contract to housing agencies in October 2003. We have assumed that only half of these additional vouchers will be in use in October. Additional tenant protection vouchers will, however, be likely to be in use as the fiscal year progresses, so that with only a minimal increase in voucher utilization by other agencies the average voucher utilization for fiscal year 2004 as a whole will reach the 96.7 percent level indicated by the available data.
 We use HUD’s estimate that funding for 2,106,233 authorized vouchers expires in fiscal year 2004.
 The shortfall estimates are based on the estimate that 2,011,000 vouchers requiring renewal will be in use in October 2003.
 As explained in footnote 1, some $500 million in carryover funds assumed by the President’s budget to be available for use in fiscal year 2004 were rescinded after the budget was issued. (If $1.07 billion in carryover or other unspent funds from prior years were available to support housing vouchers in fiscal year 2004 as the President’s budget assumed, for a total of $13.05 billion, the consequences of the President’s inadequate funding request would still be stark: on average over the fiscal year 110,000 vouchers likely to be providing rental assistance to families at the beginning of the fiscal year would not be funded, rather than 184,000, and an additional approximately 95,000 vouchers that Congress has previously authorized, as discussed above, would be defunded.)
 To accomplish a reduction of 368,000 households assisted from the start to the end of the fiscal year, as would be required if cuts were implemented evenly over the course of the fiscal year, the number of working families served would need to be reduced by 112,000, the number of households with disabilities by 81,000, the number of elderly households by 63,000, and the number of other households by 112,000.
An average of 56,000 additional households that are not elderly or disabled and whose income is not derived primarily from earnings — including unemployed households — also would likely be deprived of voucher assistance due to the funding shortfall in fiscal year 2004. Calculations of voucher loss in particular demographic groups are based on the followings estimates of breakdown of voucher holders: 30.5 percent working families, 22 percent disabled households, 17 percent elderly households, and 30.5 percent other households. The proportion of households in each category is based on HUD Multifamily Tenant Characteristics System data for the 18-month period ending September 2000, as reported in Housing Choice Voucher Location Patterns: Implications for Participants and Neighborhood Welfare, HUD Office of Policy Development and Research,
January 2003, pp. 92-93, Tables A-5 and A-6, on similar data gathered by HUD’s Office of Policy Development and Research in a report for the Millennial Housing Commission entitled Basic Facts on Work Participation and Length of Tenure in HUD-Assisted Housing, and on similar data for the period ending in late 1999 as reported in Recent Research Results: New Facts about Households Assisted by HUD’s Housing Programs, HUD Office of Policy Development and Research, October 2000. Working families are defined as families with a non-elderly, non-disabled head of household whose income is reported by HUD to be “primarily from earnings.” Most of these households have children, but some do not. (Some additional elderly and disabled households also derive their income primarily from earnings.) Households that are both elderly and disabled are counted as elderly rather than as disabled.