March 24, 2004
SOURCES AND METHODS USED TO ESTIMATE IMPACT OF
HOUSING VOUCHER PROPOSALS IN ADMINISTRATION’S FISCAL YEAR 2005 BUDGET
By Barbara Sard and Will Fischer
PDF of methodology
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.
This paper describes the methodology we used to derive estimates of the impact of the proposals in the Administration’s fiscal year 2005 budget affecting the “Section 8” housing voucher program. Findings based on this methodology are included in a series of Center analyses (available on the internet at https://www.cbpp.org/housingvoucher.htm) that examine the Administration’s proposals.
As is discussed in detail below, our estimates of effects on particular states and housing agencies assume that cuts would be spread proportionately across housing agencies. The Administration’s proposal does not specify how reductions would be distributed, so the actual cuts for any particular housing agency could be somewhat larger or smaller.
Our estimates are based in part on data reported by individual housing agencies to HUD covering the period through July 2003. These are the most recent national data available, but individual agencies have more current data they could use to update these estimates, based on the methodology described below.
Housingagencies also may be able to use specific information about their voucher programs to estimate effects of the cut on particular subgroups, such as the elderly, people with disabilities, or working families.
We first describe the method used to estimate the national funding shortfalls in the voucher program under the Administration’s budget and the number of vouchers that could be lost nationally. Then we explain how we use these national figures to derive estimates of the funding reductions, numbers of vouchers lost, and possible rent increases for each housing agency and state.
Calculating the Shortfall in Voucher Funding and the National Impacts
Our estimates of the funding shortfalls in 2005 and 2009 rely on Congressional
BudgetOffice assumptions and estimates regarding the needs of the voucher program. Because the type of budget information available for each of the two years differs, we use different methods to estimate the shortfalls.
Shortfall in 2005
The Administration’s budget requests $13.18 billion for the renewal of existing housing vouchers and voucher administrative fees and eliminates the Family Self-Sufficiency program (an asset development and employment counseling program that operates in conjunction with the voucher program). Administration budget documents acknowledge that this funding level would result in a shortfall of $1.63 billion below the amount needed to maintain services under the voucher program at the current level. That estimate, however, is based on projections regarding growth in per-voucher costs and other factors that appear to be unrealistic.
BudgetOffice assumptions regarding per-vouchers costs, the voucher utilization rate, the number of vouchers, and other factors are used to estimate the amount by which the Administration’s budget request falls short of the amount needed, the shortfall comes to $1.92 billion. The shortfall in funding for renewal of existing vouchers alone (that is, excluding administrative fees and the FSS program) under the CBO assumptions will be $1.79 billion. Our analysis examines cuts in voucher subsidies rather than reductions in administrative costs, so we use the $1.79 billion figure in our calculations for 2005.
To estimate the number of vouchers that would be lost in 2005 if the entire $1.79 billion cut were implemented by eliminating vouchers, we divided the $1.79 billion by $6,673, CBO’s estimate of the average cost per voucher with administrative fees excluded. This resulted in a reduction of 268,000 vouchers. For purposes of this analysis, we rounded this estimate down to a reduction of 250,000.
Shortfall in 2009
In 2009, CBO estimates that under the President’s budget, expenditures (or “outlays”) for the Section 8 program — which includes both the voucher program and a smaller “project-based” housing assistance program — will fall $4.56 billion below CBO’s “baseline” outlay level. We assume that this full cut will be applied to the voucher program rather than to the project-based Section 8 program. This appears likely because reducing funding for the project-based program would require changing the laws governing the renewal of expiring contracts between HUD and private owners of subsidized projects. The Administration has not proposed such a change, while it has proposed an overhaul of the laws governing the voucher program.
This analysis uses an estimate that 600,000 vouchers would be eliminated if the entire $4.56 billion shortfall in 2009 were met by reducing the number of families receiving voucher assistance. For two reasons, the process of arriving at this estimate is more complex than the calculation used to arrive at the estimate of the number of vouchers that would be lost in 2005. First, in 2009, unlike in 2005, no information is available indicating whether part of the cut will be applied to administrative funding. If the $4.56 billion reduction were implemented entirely by cutting voucher subsidies with no reduction in administrative funding, the number of vouchers lost could be estimated by dividing the shortfall amount by CBO’s estimated per-voucher subsidy cost in 2009 of $7,409. This results in a loss of 615,000 vouchers. Under this scenario, administrative funding would be substantially higher as a proportion of the voucher program’s overall budget than the Administration has proposed for 2005. If administrative funding in 2009 is assumed to remain at the same proportion of total funding as the Administration proposed for to 2005 (9.8 percent), the cut in subsidy funding would be smaller and the number of vouchers lost would be 583,000.
These estimates of the number of vouchers that could be lost in 2009 under these two scenarios are, however, about 38,000 below the actual number of vouchers that could be lost. This underestimate is a result of technical constraints on CBO’s baseline estimates. When CBO estimates the baseline amount of funding (or “budget authority”) that will be required for the Section 8 program in future years, CBO is required by statute to assume that per-unit costs will grow at the same rate as an inflation index called the “GDP deflator.” The GDP deflator typically grows at a rate slower than the growth in rents and other factors that influence Section 8 costs, so CBO’s budget authority baseline tends to underestimate Section 8 per-unit costs. CBO is permitted to use — and does use — more realistic assumptions about per-unit cost growth in its estimates of the outlay baseline. But it cannot estimate baseline outlays that are above the level that can be paid for with baseline budget authority.
In 2005, the difference between the estimates of cost growth that CBO uses to project budget authority and the estimates of cost growth it uses to project outlays is not large enough to distort its projections. The difference does, however, distort the projections in later years. Under CBO’s assumptions about per-voucher costs and other factors in 2009, the expenditures needed to cover all of the vouchers that will require funding under the current program structure would exceed the funding that will be available from the baseline budget authority. CBO consequently reduced its estimates of baseline outlays — and of the number of Section 8 units that would be funded — to levels that could be covered by baseline budget authority. This process reduced outlays by more than $200 million and reduced the number of Section 8 units funded by 37,700.
If the outlay baseline in 2009 were at the level required to cover all authorized Section 8 units at the estimated utilization rate, the funding shortfall calculated by comparing baseline expenditures to expenditures under the Administration’s budget would be somewhat above $4.56 billion and the voucher cut estimates would exceed 583,000 and 615,000. Balancing these considerations, for purposes of this analysis we use an estimate that the number of vouchers lost would be 600,000. This estimate appears conservative.
Calculating the Impacts of the Funding Cuts at Each
Our estimates of the funding reductions, the reductions in number of families assisted, and the rent increases that the Administration’s proposals could cause at the local level are derived from the national-level cuts using the methods described below. Estimates of state-level funding reductions and reductions in the number of families assisted are the sum of the local-level reductions. State-level rent increases are weighted averages of local rent increases.
Reduction in the Number of Families Assisted
The reduction in the number of families assisted at each local housing agency represents the number of vouchers that would otherwise be in use but that the housing agency would be forced to eliminate if it implemented the cut proposed by the Administration by reducing the number of vouchers. We calculated the number of vouchers that would be eliminated by dividing the estimated national reduction of 250,000 vouchers in 2005 and 600,000 vouchers in 2009 across local agencies in proportion to each agency’s share of vouchers authorized nationally in July 2003. The number of authorized vouchers in this calculation is based on the number of contracted vouchers reported by housing agencies in quarterly data submissions to HUD through July 2003.
The total number of authorized vouchers that agencies reported in these quarterly submissions should have included only Section 8 vouchers, but some agencies may have also reported “five-year mainstream vouchers.” Five-year mainstream vouchers are a form of housing assistance for people with disabilities separate from Section 8 that Congress funds through the Section 811 program. Five-year mainstream vouchers would not be affected by the cuts the Administration has proposed to the Section 8 program, so if an agency included such vouchers in its submission of data to HUD, our calculations would overestimate the cut that that agency would experience. We do not have sufficient information to determine at which agencies this occurred. There are total of approximately 12,000 five-year mainstream vouchers, compared to over two million Section 8 vouchers, so even if a substantial portion of the five-year mainstream vouchers were erroneously included in the Section 8 voucher data the overall impact would be quite small.
Sixteen agencies participating in the Moving-to-Work (MTW) demonstration are not covered in the quarterly data. The number of authorized vouchers at these agencies was determined based on data obtained from HUD’s PIC database on
February 9, 2004or through telephone interviews with agency staff.
Based on the data from these sources, the number of authorized vouchers at individual agencies adds up to 2,062,569. This is below the actual number of authorized vouchers nationally, because these agency totals do not include either Section 8 Moderate Rehabilitation units (which HUD considers as vouchers for budgetary purposes) or “tenant protection” vouchers that HUD plans to award with fiscal year 2004 funds. HUD data indicate that there are 79,908 vouchers in these two categories. Because data are not available showing how these 79,908 vouchers are distributed among local agencies, we allocated the reductions in vouchers in 2005 and 2009 based on each agency’s share of the 2,062,569 vouchers that were reported in data from individual agencies.
In each year, some authorized vouchers are not used, often because some families that are issued vouchers are unable to find housing where they can use their voucher. The “utilization rate” — that is, the percentage of authorized vouchers in use — varies widely from one local agency to another. Since vouchers that are not in use do not incur subsidy costs, it is likely that under the inadequate funding levels the Administration has proposed, agencies with higher utilization rates will (all else equal) experience proportionally larger reductions in the number of families they can assist. It is difficult, however, to predict future trends in utilization rates at individual agencies. Because our analysis distributes the reduction of 250,000 vouchers in 2005 and 600,000 vouchers in 2009 among local agencies based on each agency’s share of the number of authorized vouchers, it effectively assumes the utilization rate will be constant across all agencies.
Reduction in Voucher Subsidy Funding
The estimated reduction in voucher subsidy funding includes only cuts to the funding used to cover the costs of voucher subsidies. (Most housing agencies would also experience sharp cuts in administrative funding.) This figure represents the potential loss to the rental housing sector of the local economy that would result from the Administration’s budget.
We estimate the funding reduction for each local agency by taking the estimated number of vouchers that each agency would eliminate in 2005 and in 2009 if it compensated for all of the funding shortfall in this manner and multiplying this number by the projected per-voucher cost for the agency in those years. We used the agency’s per-voucher subsidy cost in the period from May-July 2003 to project costs for the agency in 2005 and 2009. (For agencies where cost data for May-July 2003 were not available, we projected May-July costs using earlier data.) We assumed that national average per-voucher costs in 2005 and 2009 would be at the levels projected by the Congressional
BudgetOffice: $6,673 in 2005 and $7,409 in 2009 and that per-unit costs at each agency would rise at the rates required to reach these national averages.
For the 16 MTW agencies, we estimated total per-voucher costs (including administrative fees) by dividing the agency’s voucher budget authority in 2003 by its total number of authorized units. To estimate the average per-voucher cost just for the rental subsidy, we deducted an estimated administrative fee of 10.5 percent from the total per-voucher cost. A document provided by HUD to the appropriations committees in November 2003 estimates that administrative fees would make up 10.5 percent of total funding for MTW agencies and Section 8 Moderate Rehabilitation units in fiscal year 2004.
Average Rent Increase
The average rent increase reflects the average increase that housing agencies would be required to impose on households with vouchers if the agencies implemented their share of the cut proposed by the Administration entirely by raising rent burdens. We calculated the rent increase by dividing the reduction in voucher subsidy funding at each agency in 2005 and 2009 by the number of units projected to be in use in that year. The number of units in use was estimated by multiplying CBO’s projection of the national percentage of authorized vouchers that will be in use (96.5 percent in 2005 and 95 percent in 2009) by a projection of the number of authorized vouchers at each agency. As noted above, actual utilization rates at individual agencies may be higher or lower than the national average rate.
The number of vouchers rises each year because some new vouchers are issued to households that lose assistance under project-based Section 8, public housing, or other housing assistance programs. We projected the number of authorized vouchers at each agency in 2005 and 2009 based on CBO’s estimates of the total number of vouchers that will be authorized in those years. We assumed that the number of voucher would increase at the same rate at each housing agency. Of course, some agencies will receive more new vouchers than others, but it is not possible at this point to predict how the vouchers will be distributed.
 In fiscal year 2005, the Congressional
BudgetOffice estimates that 96.5 percent of authorized vouchers will be in use and that the average cost per voucher will be $6,673. CBO documents show 2,176,774 renewal vouchers in 2005, but this total includes CBO’s estimate that Congress will authorize about 35,000 new tenant protection vouchers in 2005 to replace expiring or terminated project-based assistance. During their first year of operation, new tenant protection vouchers are funded through a separate appropriation from existing vouchers. CBO’s estimate of the number of authorized vouchers subject to annual renewal is 2,141,474. This total is equal to CBO’s estimate of the number of renewal vouchers in 2004 plus 3,192 vouchers previously funded under multiyear contracts that are expiring for the first time in 2005. We used the 2,141,474 figure to calculate the shortfall of $1.92 billion. The components of CBO’s budget estimates for the Section 8 program are available on the internet at http://www.cbo.gov/factsheets/2004b/Housing.pdf .
 Agencies that participate in the Moving-to-Work (MTW) demonstration receive voucher funding under a different formula from other agencies, so they are not required to submit the quarterly data that we used to determine the number of authorized vouchers and average costs per voucher at most agencies. Some MTW agencies are permitted to reduce the number of vouchers they administer and use voucher funds for other services or activities. As a result, the number of vouchers that are actually made available to families by MTW agencies may be below the agencies’ authorized number of vouchers. Because we could not use data on actual per-voucher costs or numbers of vouchers to estimate cuts at MTW agencies, the estimates for these agencies are more approximate than the estimates for other agencies.
 HUD data indicate that as of
June 30, 2003there were 48,252 Section 8 Moderate Rehabilitation units, of which more than 41,000 will be subject to annual renewal funding in 2005. These are project-based subsidies administered by PHAs. HUD includes these Moderate Rehabilitation subsidies as vouchers for budgeting purposes. It appears that HUD’s Flexible Voucher Program proposal also would apply to the Moderate Rehabilitation subsidies that require annual funding. We do not have data on the breakdown of Moderate Rehabilitation units by housing agency. The national total also includes the 31,656 tenant protection vouchers that HUD budget documents indicate HUD plans to award in fiscal year 2004.
Home | Background Information | Board of Directors | Center Staff
Search | Publication Library | Job Opportunities | Internship Information
State Policy Work Home Page
Click here to join the Center's e-mail notification list.
To ask questions, or send comments, write to [email protected]
Center on Budget and Policy Priorities
820 First Street, NE, Suite 510
Washington, DC 20002
Ph: (202) 408-1080
Fax: (202) 408-1056