HUD, Congress, voucher program managers, researchers and housing advocacy groups have focused on voucher utilization and the related issues of success rates and program costs for several years. Because under-utilization of vouchers results in fewer families receiving housing assistance each year than could be served with available resources, HUD would like to make all possible efforts to maximize the utilization of vouchers allocated to local programs. Understanding the drivers of utilization can help voucher program administrators determine whether controllable factors (e.g., PHA policies and practices) or uncontrollable factors (e.g., market conditions or waiting list characteristics) are at work when allocations are not fully used. They can then take appropriate actions to improve utilization when needed and when the factors affecting utilization are under their control. Similarly, understanding drivers of program subsidy costs can help program operators and policy makers develop more accurate budget projections and can help them understand potential trade-offs—for example, between the numbers of families served on the one hand and the types of families served and the quality of the housing they rent on the other.

This study is intended to provide insights into the factors that affect Housing Choice Voucher (HCV) program utilization rates and costs in a sample of sites nationwide. The data for the study were derived from existing computerized HUD files, other secondary data sources, and primary data collected on site at a sample of 48 PHAs. The bulk of the information was gathered during on-site interviews with voucher program staff as part of one- to two-day visits we made to each of the study sites between December 2001 and April 2002. While on site, we discussed aspects of each PHA’s local housing market, participant characteristics and PHA policies, to assess their impacts on subsidy costs and voucher utilization. In addition to interviewing key PHA staff in person, we spoke by telephone with local HUD staff, landlords, participants, and community representatives regarding the programs. A sample of participant files was also reviewed on site to determine the time required for each of the several activities that lead to getting a voucher under lease and to assess the completeness of the files.

The sample was selected purposively to include PHAs with high and low utilization rates and PHAs with high and low costs across a range of program sizes and locations. A subset of 28 of the PHAs were selected as pairs. A pair was defined as two PHAs that served either the same or similar housing markets and had at least a 10-point difference in the utilization rate. By looking at pairs, we hoped to separate the factors affecting utilization from the more general market-related factors.

The sample was purposive rather than random, so we cannot derive precise national estimates of the impacts of various factors on program costs and utilization. While the results from this study cannot be generalized to the entire universe of PHAs, they should provide HUD with sufficient information to support program decision-making and help identify areas for technical assistance that can improve utilization rates and assist PHAs in using their increased flexibility to optimize local programs.