Rebuilding the Research Capacity at HUD (2008)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

6 Evaluation of Policy Development and Program Support PD&R has a unique role in the policy development process. Because it is not a program office, it is not committed to a given program or a given recommendation. It is an independent voice and source of policy analysis for the secretary. Major new initiatives are likely to be significant departures from existing programs, and the secretary may often look to PD&R to develop the initiative. Some of these initiatives are proactive, and some are reactive. PD&R has had a particularly significant role when policy makers have felt it necessary to reform programs and have been looking for solu- tions. A number of HUD’s major policy initiatives have been developed in such situations. PD&R staff have an advantage over program offices in designing and managing research projects and in interpreting research results. PD&R staff are likely to be less knowledgeable about a program than the program office staff that actually manage it, but they build up expertise as they conduct research and participate in the policy development process. In addition, individual PD&R staff members sometimes have experience in program offices before coming to PD&R (and the converse is true; mobility occurs in both directions). Thus, individual staff members are frequently extremely knowledgeable about specific programs. Indeed, policy development draws on research in the broadest sense of that term. This chapter focuses on specific instances of policy development, but in describing many of these examples it will be clear that a particular policy initiative draws on previous research and analysis. Many impor- tant policy proposals grow out of research programs that have produced results over a long period of time. However, the committee makes a 93 

94 REBUILDING THE RESEARCH CAPACITY AT HUD distinction between “short-term” policy development, discussed in this chapter, and the long-term interaction between research and policy devel- opment, discussed in Chapter 9. Policy development tends to be thought of in terms of HUD’s programs, but it is also extremely important in the exercise of HUD’s regulatory responsibilities over the housing finance system and the housing market. The committee found it convenient to discuss PD&R’s role in programmatic and regulatory activities separately. Within the programmatic activities, the committee distinguishes between modest changes, major reforms, and broad urban policy concerns that may lie outside the responsibilities of the department. Before considering these activities, we discuss some basic concerns about assessment. THE DIFFICULTY OF EVALUATING POLICY DEVELOPMENT The contribution of HUD’s research activities to the policy develop- ment process is difficult to evaluate. This is inherent in the process. “Policy development” is not the sole preserve of the Office of Policy Development and Research; it involves relevant program offices and other support offices, such as the Office of the General Counsel and the Office of the Chief Finan- cial Officer. More fundamentally, to a large extent the activity leading to any particular policy decision is not public. It is internal to the department or the executive branch; it is pre-decisional and therefore confidential. Some of it is oral rather than written; contributions are made in the course of meetings among senior departmental policy makers. Memoranda and other documents produced during the development process certainly include some of this information, but they do not reflect what happens at the actual decision-making point. Nor do they describe the informal discussions that occur between offices and individuals during the process, in the course of which significant second-order issues may be resolved and major issues clarified. Testimony and speeches by the secretary and other HUD officials do not typically include citations, and seldom if ever include references to the contributions of individuals or offices in HUD. This inherent difficulty can be addressed in several ways. One possibil- ity would be an intensive set of structured interviews asking similar ques- tions with the same basic format with participants in the policy process, such as past HUD secretaries and assistant secretaries, officials at the Office of Management and Budget (OMB) and perhaps elsewhere in the Execu- tive Office of the President, key members of Congress and legislative staff, and perhaps others. Although such an approach was beyond the scope and resources of the committee, several of the committee’s meetings included discussions with former PD&R officials and senior congressional staff, and these provided insight into the policy-making process in general, as well as 

EVALUATION OF POLICY DEVELOPMENT 95 specific examples. In addition, committee members include former PD&R assistant secretaries and deputy assistant secretaries and former visiting scholars, whose combined personal experience at HUD extends over more than half of the history of the office. Apart from the personal experiences of committee members and others with knowledge of specific instances, the quality of PD&R’s contribution to the internal process of policy making, or for that matter the quality of the contribution of any office, has to be inferred from the policy outcome, and attitudes toward a given policy outcome will vary. Moreover, the final decision may primarily reflect the recommendations of some offices rather than others. For these reasons, the committee did not conduct a formal evalua- tion of the policy development function in PD&R. Instead, the committee developed a conceptual framework for categorizing types of policy devel- opment activities and provided a number of illustrations in each category. The committee believes that this approach conveys the broad range and quality of policy development work and the contexts in which it occurs. PD&R has played a positive, essential, and indeed (as stated at the begin- ning of this chapter) unique role in the development of housing and urban policy, across the range of HUD program and regulatory responsibilities and beyond. THE NATURE OF POLICY DEVELOPMENT ACTIVITIES As with research, policy development activities can be categorized in several dimensions. They can serve different purposes, ranging from improvements in the operations of HUD programs, to new initiatives within the general framework of HUD’s authority, and to broad urban policy con- cerns, such as providing support to independent commissions established by the secretary or even the President to address subjects that may be outside the program responsibilities of the department. In the program support cat- egory in particular, there are differences in scope and scale. Policy proposals can be for modest, marginal changes in programs, sometimes conveniently known as “tweaking,” or they can be substantial reforms that affect the fundamental structure and operations of a program, taking several years of work. Some PD&R activities are negative rather than positive—stopping bad ideas from becoming policy. The committee has found this typology—program support, new HUD initiatives, broad urban policy concerns—useful in describing PD&R’s p ­ olicy development activities. In addition, it is useful to consider program- matic and regulatory matters separately. The remainder of this section provides illustrations in each category. 

96 REBUILDING THE RESEARCH CAPACITY AT HUD Program Support Policy development in support of ongoing programs is the most com- mon PD&R activity in the general policy development category. As noted, the scale of effort varies from modest program modifications to substantial improvements that can benefit the federal government or program par- ticipants to the extent of billions of dollars. The range is illustrated in the following examples. TOTAL The TOTAL (Technology Open To All Lenders) Scorecard allows l ­enders approved by the Federal Housing Authority (FHA) to determine whether a specific home mortgage loan will be insured by FHA. It is designed to function as part of the lender’s own automated underwriting process. The TOTAL Scorecard improves FHA’s ability to assess the risk of default at the high-risk end of FHA’s spectrum of borrowers, and thereby to broaden its risk spectrum and serve borrowers it might otherwise not have done. The proportion of minority households tends to increase with risk, so the Scorecard enables FHA to serve more minority households. PD&R staff developed the Scorecard, using private data on credit scores from Fair Isaac and Company, as well as FHA data on borrower characteristics and loan experience, to improve FHA’s ability to assess risk. Work on the Scorecard began in the late 1990s and was completed by 2000. It was made available first through Fannie Mae and then directly to all lenders in 2003. Lenders have reported satisfaction with the Scorecard, especially its compatibility with their origination systems. Through fiscal 2007, about 1.24 million loan applications had been processed through TOTAL, and about 1.1 million mortgages approved for insurance; about 425,000 applications were processed in fiscal 2007 (U.S. Department of Housing and Urban Development, 2007b, p. 9), somewhat more than half of the 770,000 applications received. Quality Control For many years HUD has been sharply criticized for substantial inac- curacies in setting rents for tenants in its assisted housing programs, including both those administered by the Office of Public and Indian Housing (public housing and the voucher and certificate programs) and those administered by the Office of Housing (Section 8 and earlier pri- vately owned subsidized projects). Unlike other assistance programs, such as Food Stamps, HUD lacked a quality control program to identify and reduce payment errors. 

EVALUATION OF POLICY DEVELOPMENT 97 Efforts to reform the process began in the mid-1980s, when Under S ­ ecretary Philip Abrams created a task force to study the application of quality control to HUD programs with income-conditioned subsidies. PD&R staff chaired the task force and played the major role in drafting the report. The Office of Housing was then assigned to develop a quality control program, but a contract to measure the magnitude of errors was not signed until 1988, and data collection did not begin until 1992. After the data collection was completed, PD&R was assigned responsibility for the study, which was completed in 1996 (Loux, Sistek, and Wann, 1996). The study estimated that slightly more than half of assisted households paid either too much or too little; errors totaled about 8 percent of HUD sub- sidies, about $1.4 billion. The errors cut in both directions: some ­tenants paid too much in rent and others too little, relative to their incomes and household size. Specific recommendations to reduce the errors resulted from the research, including better training and monitoring of public housing authorities, sim- plification of the legal and regulatory requirements for calculating rents, and obtaining better information on tenant employment income by matching records of the Public Housing Authority with independent databases. The program offices, working in cooperation with PD&R and the Office of the Chief Financial Officer, began implementing these recommendations in 2001 through the Rental Housing Integrity Improvement Project, and by fiscal 2003 HUD had begun to reduce errors substantially. By fiscal 2005, the error rate had been reduced to about 3 percent, or about $900 million (assisted housing programs were substantially larger in 2005 than in 1993). This continued progress led the U.S. Government Accountability Office to remove the housing assistance programs from its list of high-risk programs. PD&R- s ­ upported external research monitored and verified the improvement. The quality control work took quite a bit longer than TOTAL and even- tually involved in-house and external research as well as policy development. Initially, PD&R’s role was intended to be limited to policy development, and it continues to be involved in policy development and program monitoring. The quality control effort is also an example—not uncommon—of how work begun in one administration continues in successor administrations to the long-term benefit of the department and the taxpayer. SEMAP The Section Eight Management Assessment Program (SEMAP) rates the management of the voucher program by public housing agencies. Origi- nally, some elements of the SEMAP rating were based on data analysis by local housing authority staff, who are typically not trained in sampling and research techniques. A PD&R project found substantial errors in these 

98 REBUILDING THE RESEARCH CAPACITY AT HUD data, large enough to affect the overall SEMAP rating, leading the Office of Public and Indian Housing to revise the SEMAP indicators. Housing Counseling Many policy development activities are much smaller in scope than those just described. The Office of Housing manages the department’s housing counseling program, funded at about $40 million annually in recent years, and has collected data on the activities of counseling agencies. An analysis by PD&R contributed to the design of a new data collection system, with information about individual households, which is currently being implemented. This system is being utilized in the large-scale housing counseling study now under way. Fair Market Rents and the LIHTC For more than 30 years PD&R has had the responsibility for estimating fair market rents for individual housing markets (defined as metropolitan areas or individual nonmetropolitan counties) on an annual basis. Fair market rents originally determined the maximum rent for a unit in the cer- tificate program and, subsequently, the maximum amount that HUD would pay toward the rent of a unit in the voucher program. When the Low- Income Housing Tax Credit (LIHTC) Program was enacted in 1986, HUD was given the responsibility for determining which local areas qualified for higher credit rates because of high construction, land, and utility costs (known as difficult development areas). Within HUD, the responsibility was assigned to PD&R, which used fair market rents as one of the criteria. This is an instance in which PD&R provided program support to another agency (the Treasury Department, which administers the LIHTC). Income Limits PD&R produces median family income estimates for each local housing market. These estimates are then used to determine the income limits for program eligibility (for example, 50 percent of the local median income). The original use of the estimates was in the public housing program; they now apply in 11 HUD assistance and community development programs. In addition, they are used by several other agencies, including the Departments of Agriculture and Veterans’ Affairs. They are used in the LIHTC as one of the criteria for difficult development areas (the criterion is the ratio of the fair market rent to the median income) and in the criteria for qualified cen- sus tracts, which also receive higher credit rates. They are referenced in five 

EVALUATION OF POLICY DEVELOPMENT 99 provisions of the federal tax code, applying to both the LIHTC and state- issued tax-exempt bonds, and in the affordable housing goals for ­Fannie Mae and Freddie Mac, the affordable housing programs of the Federal Housing Finance Board, and the rules for Community Reinvestment Act compliance. Hurricane Katrina and the Northridge Earthquake PD&R has conducted analyses that have contributed to the federal gov- ernment’s responses to natural disasters. Between 1992 and 2005, Congress appropriated 17 supplements to the Community Development Block Grant (CDBG) Program for relief aid in various situations; PD&R was called in to develop the formulas for allocating these funds. In the aftermath of Hur- ricane Katrina and the other storms that damaged Gulf Coast communities in 2005, PD&R’s work included being part of an interagency collaboration that created property damage estimates from government and private data sources (U.S. Federal Emergency Management Agency, U.S. Department of Housing and Urban Development, and U.S. Small Business Administration, 2006). This work became the basis for allocating $16 billion in supplemen- tal CDBG funding; measuring the extent of damage to housing, which was used to justify the second CDBG supplemental appropriation; and analyzing the availability of vacant rental housing within various distances from New Orleans. In 1994 the accumulated evidence from PD&R research about the effectiveness of housing vouchers led HUD to offer emergency ­vouchers to low-income households displaced after the Northridge earthquake. More- over, existing research on program features that discouraged landlord par- ticipation informed the decision to waive some of the standard program regulations so that the emergency vouchers would be more widely accepted in the Los Angeles rental market. PD&R then deployed a team of in-house researchers to monitor and assess the early implementation of the emer- gency vouchers, and, over the following 2 years, funded external research on the Northridge program. New Initiatives New initiatives are inherently less frequent than program support a ­ ctivities. HUD does not offer new initiatives more often than every few years, although the boundary between program support activities and new initiatives is to some extent a matter of individual judgment. Modifications to a large program can be so extensive as to be commonly considered a new initiative. 

100 REBUILDING THE RESEARCH CAPACITY AT HUD Home Equity Conversion Mortgages (HECMs) In 1987 FHA received statutory authority to insure home equity con- version mortgages (HECMs) for elderly homeowners who wanted to use the equity in their homes for living expenses. HECMs were authorized by Congress on a demonstration basis. PD&R had the lead in designing the demonstration, working with the Office of Housing. PD&R’s specific recommendations, incorporated into the program, included a two-part pre- mium structure (200 basis points as an upfront premium and 50 basis points annually), for the first time in any FHA insurance program; a method for calculating the maximum amount of a mortgage, based on the ­borrower’s age and the interest rate; and formulas for borrowers to establish their own plans for receiving payments from the mortgage. The two-part premium subsequently was adopted for the basic FHA home mortgage insurance program and some smaller programs. FHA-insured HECMs grew in popularity gradually until early in this decade, then they began increasing very rapidly. In 2005 Congress made the program permanent. More recently, Congress mandated a study of the insurance premium on the refinancing of HECMs. A PD&R-funded study (discussed in Chap- ter 3) concluded that the premium could be lowered to apply only to the increase in the amount of the mortgage, not the full amount, without affecting the actuarial soundness of the program (Rodda et al., 2003). This proposal was adopted by FHA. Preservation For a decade beginning in the mid-1980s, HUD confronted the question of whether and how to preserve Section 236 and Section 8 new construc- tion projects for their low-income residents, as the original 20-year subsidy contracts expired and project owners could opt out of the program. PD&R conducted a survey and analysis of FHA-insured Section 8 projects as of 1990-1991, which included per-unit estimates of annual subsidy cost, annual accrual of repair needs, and the backlog of needed repairs to bring the units up to market standards (Hodes, 1992; Wallace et al., 1993) The PD&R work provided the most extensive data yet available on the Section 8 inventory and became a basic resource in the policy debates after HUD proposed a major “preservation” initiative in 1995. It was frequently cited by HUD officials during the debates on preservation in 1995-1997. A second survey, updating the data to 1995, also provided useful informa- tion, although the report on this later survey was not published until 1999 (Finkel et al., 1999). 

EVALUATION OF POLICY DEVELOPMENT 101 Stopping Bad Ideas The converse of providing program support and developing major initia- tives is critically analyzing proposals and pointing out their weaknesses as well as their strengths. To the extent that the proposals are internal to HUD and PD&R’s critique results in their not being adopted, this role does not receive much public attention; PD&R’s work is pre-decisional and the deci- sion is negative. But while much of PD&R’s activity does not see the light of day, some instances have become public and can serve as illustrations. Program Rents The formulas to calculate tenant rents in assisted housing programs are complicated, and there are frequently proposals to simplify them. PD&R has the responsibility for estimating the budgetary impact of these p ­ roposals, both overall and for particular categories of assisted house- holds. Frequently the analysis shows that the proposal will have substantial budgetary costs or have particularly severe impacts on some demographic groups. Such findings result in a decision to not proceed with the proposal, or to withdraw it if it has already been introduced in Congress. Changing Fair Market Rents Fair market rents have been set at the 40th percentile of the rent dis- tribution for recent movers living in decent housing. PD&R analyzed a proposal in 2000-2001 to raise the standard to the 50th percentile; the analysis showed that the cost was high and unnecessary in most markets; eligible families could find decent housing at the 40th percentile of the dis- tribution (or lower). The proposal was modified to apply only to areas for which the data showed substantial concentrations of voucher recipients in particular neighborhoods. Coinsurance In 1983 HUD created a coinsurance program for multifamily projects as a new activity for FHA, although PD&R had advised against the program. By 1990, the program had incurred substantial defaults and insurance claims, and the coinsurance feature was proving to be illusory: the lenders’ reserves were inadequate to cover their share of the losses and FHA was, in fact, bear- ing a much larger share of the losses than anticipated in the program design. These were concerns raised by PD&R in advance. In this case, the quality of PD&R’s contribution to policy making was high, though not followed. 

102 REBUILDING THE RESEARCH CAPACITY AT HUD When Things Go Wrong Sometimes a major policy initiative results from a program failure. As noted earlier, PD&R can be especially useful when things go wrong with a program. The Quality Control research and policy development mentioned earlier could be regarded as an example; the cost of assisted housing programs was rising rapidly and HUD lacked a means of deter- mining the accuracy of subsidy payments. Coinsurance is another example. When the problems became apparent in 1989, PD&R was given the lead responsibility for developing a response. PD&R’s analysis contributed to the secretary’s decision to terminate the program, rather than attempting to redesign it, and also helped to persuade Congress that the program should be terminated. Reinventing HUD The most extensive effort to address HUD problems since the National Housing Policy Review was the “Reinvention Blueprint” of the 1990s. Reinventing HUD was a prominent element in the Clinton administration’s National Performance Review, chaired by Vice President Gore. The initial impetus for reinvention was reinforced after the 1994 midterm elections, when the newly elected congressional Republican majority began to call for dismantling HUD. PD&R played a major role in envisioning a radically different depart- ment and streamlined approach to delivering housing and community development resources. It had the lead responsibility in crafting the foun- dational document, A Place to Live Is the Place to Start, released in January 1995, a statement of principles which formed the basis of the department’s sub­sequent major legislative proposals (U.S. Department of Housing and Urban Development, 1995). The specific proposals were described in “HUD R ­ einvention: From Blueprint to Action,” issued in March 1995; they included plans to consolidate 60 major HUD operating programs into three performance-based funds by 1998; to fundamentally change the nature of federal subsidies to public housing authorities by providing tenants with portable rental certificates they could use in place or to rent housing in the private market; and to convert the Federal Housing Administration into a government-owned corporation that would run more like a modern insur- ance company than a bureaucracy. PD&R again had the lead in developing the detailed proposals. Reinventing Public Housing As part of the HUD reinvention, the public housing proposal was especially far-reaching. The plan called for changing public housing to a 

EVALUATION OF POLICY DEVELOPMENT 103 tenant-assisted, market-based system in an environment in which most federal regulations would be eliminated and operating subsidies terminated. It assumed current public housing residents would receive a fully funded housing certificate or voucher that would enable them to remain in their current apartment, rent another available public housing unit or, if they preferred, use the portable subsidy to rent housing in the private market. The local housing authority would become a supplier of affordable hous- ing in the larger housing market, competing for both residents and rental revenues (U.S. Department of Housing and Urban Development, 1996b, p. i). In addition, the plan included the demolition of 80,000-100,000 of the most severely distressed public housing units, with the displaced residents being given housing vouchers. This radical reinvention of public housing had to be grounded as much as possible in operational realities. It fell to the small, but agile, staff of PD&R’s Division of Policy Studies, working with the cooperation of the Housing Authority of Baltimore City (HABC), to conduct a “quick- t ­ urnaround” field study in Baltimore of the public housing reinvention plan, which would enable the department to answer as many specific questions as possible about what would happen to families, properties, and housing authorities under the department’s proposals. The 5-month exercise involved the systematic comparison of every HABC public hous- ing property with private unsubsidized rental housing as of June 1995. The market value of each HABC property was estimated by a professional real estate market firm under contract with PD&R. Property-specific comparisons, along with estimates of tenant move-outs and neighborhood resident move-ins to the now-unsubsidized public housing properties, given prevailing operating costs, led to the finding that about one-half of HABC’s developments, representing about one-third of all units, would operate at a net surplus if they were marketed in an as-is condition. At the other extreme, just three family-designated high-rise developments were responsible for 55-67 percent of the total financial deficit incurred in Baltimore’s public housing program (Abravanel et al., 1999, p. 84). The study further found that the near-term fiscal consequences of marketing properties in an as-is condition under competitive conditions would result in HABC incurring annual costs of $90.4 million. Moreover, it would take an investment of $500 million for HABC to bring all of its properties to a market-competitive standard. These findings had the effect of causing HUD to propose stretching out its timeline for fully implement- ing the reinvention plan and to provide additional capital funds for hous- ing authorities to renovate their salvageable properties to make them more competitive (U.S. Department of Housing and Urban Development, 1996b, p. 5). Although the department’s proposals were not enacted into law, the seeds of future public housing policies—such as mandatory conversion of 

104 REBUILDING THE RESEARCH CAPACITY AT HUD obsolete properties, project-based budgeting, and a greater focus on asset management—were planted in this work. FHA Mutual Mortgage Insurance (MMI) Fund In the immediate aftermath of the savings and loan industry crisis that culminated in the enactment of the Financial Institutions Recovery, Reform, and Enforcement Act in 1989, the financial status of the FHA single-family home mortgage insurance program became a major concern. HUD Secre- tary Kemp ordered a systematic analysis of the Mutual Mortgage Insurance (MMI) Fund (the fund into which mortgage insurance premiums are paid and from which claims and other expenses are paid). This was the first independent actuarial analysis of the MMI Fund. PD&R had most of the relevant expertise in the department and was assigned the responsibility for managing the contract and working with the accounting firm perform- ing the analysis. The study found that the MMI Fund was solvent, but not actuarially sound. PD&R then assumed the lead role in developing the proposed reforms to restore soundness, and worked with Congress, the U.S. Government Accountability Office, and OMB until the reform was enacted in late 1990 (see Weicher, 1992). The fund has met and exceeded the statu- tory targets for reserves since then, and continues to do so. RESEARCH AND ANALYSIS IN THE REGULATORY PROCESS Policy development includes regulatory as well as programmatic issues. PD&R’s Office of Economic Affairs has a specialized role in developing regulations. Federal agencies are required to prepare economic and regula- tory analyses for major regulations, assessing their impact on the economy in general and specific industries or sectors, such as small businesses. The requirement for these analyses dates back to the Inflation Impact Analysis of the mid-1970s; PD&R has had a role in conducting or reviewing the analyses for HUD rules since that time. The rules mainly concern HUD programs, but the department also has had certain responsibilities for regulating the activities of private entities. Until the passage of the Housing and Economic Recovery Act in August 2008, HUD was the “mission regulator” for Fannie Mae and Freddie Mac, and the secretary had general regulatory authority over these government- sponsored enterprises. Financial safety and soundness regulation was the responsibility of the independent Office of Federal Housing Enterprise Oversight. HUD also regulates all home purchase and mortgage transac- tions under the Real Estate Settlement Procedures Act (RESPA). Both gov- ernment sponsored enterprise (GSE) and RESPA regulation have involved in-house and external research that fed into the regulatory process, as well 

EVALUATION OF POLICY DEVELOPMENT 105 as preparation of supporting economic analyses. HUD also establishes and administers the building code for manufactured housing (the only U.S. national building code) under the Manufactured Housing Construction and Safety Standards Act. Finally, it regulates the sales of subdivision lots to consumers under the Interstate Land Sales Full Disclosure Act. Affordable Housing Goals Beginning in 1992 and until the passage of the Housing and Economic Recovery Act, HUD has had the statutory mandate to establish “affordable housing goals” for Fannie Mae and Freddie Mac. The goal categories were established by legislation in that year; HUD was charged with quantifying the goals (e.g., what share of GSE mortgage purchases should be for low- and moderate-income housing). The numerical targets have been revised by regulation every 4 or 5 years since 1995. Each proposed regulation has included an exceptionally extensive analysis of the mortgage and housing markets and the GSEs’ business, which have been prepared by PD&R’s Office of Economic Affairs. (The most recent, in 2004, was more than half the length of War and Peace.) To support these analyses, PD&R has conducted more than a dozen in-house studies and overseen several exter- nal research projects, making use of loan-level data from the GSEs, data produced by HUD, such as the American Housing Survey, and mortgage market data from the Federal Reserve Board, other government agencies, and private entities. RESPA HUD has the statutory responsibility to issue regulations under RESPA, governing the provision of settlement services and disclosures to people who are buying homes or refinancing their mortgages. Regulatory reform proposals have been developed several times over the last 20 years, each requiring an extensive economic and regulatory analysis that has been prepared by PD&R’s Office of Economic Affairs in cooperation with the Offices of Housing, the General Counsel, and Fair Housing and Equal Opportunity. In addition to preparing these analyses, PD&R has funded much of the economic research on RESPA, dating back to the earliest s ­ tudies in the late 1970s. Manufactured Housing Wind Safety Standards Manufactured housing regulation has been primarily the responsibility of a specialized staff in the Office of Housing since passage of the Manufac- tured Housing Construction and Safety Standards Act of 1974. PD&R has 

106 REBUILDING THE RESEARCH CAPACITY AT HUD always had the role of preparing the regulatory impact analyses of manu- factured housing rules, and these analyses have sometimes had a significant effect on a proposed rule, and thus on the cost of manufactured housing. In 1993, in the aftermath of Hurricane Andrew, HUD revised the Manu- factured Housing Wind Safety Standards to provide better protection for manufactured housing residents in a severe storm. Originally, the proposed rule would have applied the new standard on a broad geographical basis, but the regulatory impact analysis showed that the benefits of applying the higher standard in areas where hurricanes were not likely to occur were outweighed by the costs. The final rule was revised to change the standards only in hurricane-prone areas. PD&R has also had a broader role in manufactured housing policy. The PD&R assistant secretary served as a member of the National Commission on Manufactured Housing, by appointment of the HUD secretary. The commission was created by Congress to develop legislative and regulatory proposals and met during 1993-1994. BROADER URBAN ISSUES AND CONCERNS A common vehicle for identifying urban policy issues that go well beyond current program concerns is a commission of nongovernmental experts, appointed by the secretary or the President, to take an overview of a subject and develop legislative recommendations. These commissions typically consider very broad issues, often reaching beyond HUD’s pro- grammatic responsibilities and cutting across several government agencies, and their recommendations frequently become the basis for major new policy initiatives. Besides supporting HUD’s program and regulatory activities, PD&R has often provided staff and information to such independent commissions. PD&R’s role results from the breadth of its staff expertise and its ability to provide and analyze housing and urban data from within and outside the department. Its work is sometimes acknowledged explicitly in the com- mission report and sometimes in the citations in the tables and charts that support the recommendations. Millennial Housing Commission The most recent example is the Millennial Housing Commission, appointed by Congress in 2000 to explore a broad range of matters concern- ing affordable housing. The commission’s report in 2002 recognized several PD&R staff members “for their ongoing involvement and support, which was especially important to the completion of the Commission’s work” ( ­ Millennial Housing Commission, 2002, p. 122). The commission also made 

EVALUATION OF POLICY DEVELOPMENT 107 use of data provided by PD&R for many of the tables in its report, and PD&R also contributed to the commission’s definition of housing needs. President’s Commission on Housing The President’s Commission on Housing in 1982 similarly acknowl- edged the assistance of HUD, “particularly the office of Policy Development and Research” (U.S. President’s Commission on Housing, 1982, p. viii). A large number of the tables and figures in the commission’s report were drawn from PD&R’s external and in-house research and from the American Housing Survey and other PD&R data. The commission’s recommenda- tions included establishing the voucher as the main housing assistance program, which was enacted in two stages between 1983 and 1987. Much of its work, however, concerned the U.S. housing finance system and the problems of thrift institutions, which were not and have not been HUD responsibilities, but were regulated by the Federal Home Loan Bank Board and other financial regulators. Regulatory Barriers PD&R provided the staff for the Secretary’s Advisory Commission on Regulatory Barriers to Affordable Housing. The commission’s 1991 report was widely praised and its recommendation for a Regulatory Reform Clear- inghouse on state and local building codes, land-use regulations, and the permitting process was adopted by HUD and became the basis of a perma- nent departmental interest in the issue. The department’s current regulatory reform effort, the America’s Affordable Communities Initiative, draws on the work of the commission. State of the Cities In 1970 Congress mandated a biennial report on urban issues as part of the Housing and Urban Development Act of that year. Originally known as the National Growth Report, it became the National Urban Policy Report in 1977 and most recently the State of the Cities report (an annual report) in 1997. The most recent of these reports was published in 2000. The first report in 1972 was prepared by the Domestic Policy Council in the Executive Office of the President; the second and later reports have been prepared by HUD, with participation by other government departments. Responsibility in HUD for the reports was first assigned to the Office of Community Planning and Development, but since 1982 PD&R has had the   There was no report in 1990, and the 1994 report appeared in 1995. 

108 REBUILDING THE RESEARCH CAPACITY AT HUD lead role. The reports vary widely in length and scope; they have generally contained the administration’s assessment of urban conditions and policy recommendations and sometimes provided extensive and useful urban data. In support of these reports, PD&R has established the State of the Cities Data Base containing census and other data on individual cities and metro- politan areas, dating back to 1970. Enterprise Zones Providing support to independent commissions and reporting on urban conditions are not the only ways in which PD&R has supported broad initiatives that go beyond HUD’s programs. After the 1992 Los Angeles riots, the administration proposed legislation to create enterprise zones in distressed urban and rural areas. The proposal was developed jointly by PD&R and the Treasury Department’s Office of Tax Policy. It consisted primarily of tax incentives (the responsibility of the Treasury Department) targeted to areas designated by formula (the responsibility of HUD). PD&R developed the formula and, as the HUD office with the most tax expertise, also worked with Treasury on the tax incentives. The proposal was not enacted, but became the forerunner of legislation to create “enterprise com- munities” and “empowerment zones” in the next administration. Home Ownership Initiatives Promoting home ownership, a major goal of housing policy since the 1930s, has been a priority in the last two administrations. President Clinton announced a national home ownership strategy in 1995, with the goal of achieving the highest home ownership rate in U.S. history by the year 2000. The strategy was a public collaboration between HUD and 50 public- and private-sector organizations. It included 100 specific actions to address the practical needs of potential home buyers: moderate-income families that had not been able to save enough for a down payment, lower-income working families ready to assume the responsibilities of home ownership but held back by mortgage costs that were just out of reach, and families who had historically been excluded from home ownership. HUD Secretary Henry Cisneros had the lead in developing the strategy, and he received substantial support from many parts of the department, but none greater than PD&R. PD&R economists collaborated with their counterparts at Fannie Mae, the National Association of Home Builders, and other organizations, to provide the analytical justification for the goal. PD&R also played a major role in compiling the compendium of action elements that should be taken to remove market barriers and increase 

EVALUATION OF POLICY DEVELOPMENT 109 home buying opportunities for low- and moderate-income and minority families. In 2002 President Bush established a policy objective of promoting home ownership among minority households. He called housing industry and advocacy organizations to a White House conference to identify actions that each would take in support of the initiative, and established a numeri- cal target for additional minority homeowners over the rest of the decade. PD&R again provided the analysis that led to the specific target, and con- tinues to monitor minority home ownership trends and measure progress toward achieving the target. QUANTIFYING THE VALUE OF POLICY DEVELOPMENT On balance, the policy development work of PD&R has been valu- able to the taxpayer and the participants in HUD programs. Creating an overall cost-benefit analysis is well beyond the charge to the committee, but a review of a handful of activities is more than enough to establish that PD&R’s work has produced savings to the taxpayer and benefits to pro- gram participants many times greater than its cost. We offer two examples of actual and potential large savings from PD&R’s policy development activities. They are not intended as a formal cost-benefit analysis of the specific activities or the work of the office as a whole, but as illustrative of the value of the work. The first, more recent example is the quality control project to reduce errors in rental assistance payments to lower-income households. Housing assistance has accounted for about 75 percent of the HUD budget in recent years. As discussed above, PD&R efforts to measure errors and establish a quality control system date back to the 1980s, and PD&R assumed respon- sibility for the research and analysis from the Office of Housing in 1993. The benchmark study of payment errors, for the year 2000, estimated gross payment errors of $3.2 billion and net payment errors of $2.0 billion, con- sisting of $2.6 billion in overpayments on behalf of tenants and $0.7 billion in underpayments (ORC/Macro, 2001). The subsequent Rental Housing Integrity Improvement Project, in which PD&R was a major participant, adopted the major recommendations of the quality control reports, with major savings. As of 2004-2005, annual gross errors averaged $1.3 billion and annual net errors $0.6 billion, consisting of $0.9 billion in overpay- ments and $0.3 billion in underpayments (ORC/Macro, 2005, 2006). The annual net savings of $2.0 billion in gross errors and $1.4 billion in net   Net errors are the savings to taxpayers; gross errors measure the amount by which program resources have been misallocated, with some low-income households receiving a larger subsidy than they are entitled and others receiving less than they are entitled. 

110 REBUILDING THE RESEARCH CAPACITY AT HUD errors measure the benefit of the project. The annual savings, either gross or net, are many times PD&R’s average annual budget of $48 million over 2001-2005; the savings over 2 years are more than the total outlays by PD&R since its creation in 1970—about $3.2 billion in 2006 dollars. The second example is of a loss to the federal government because PD&R’s advice was not followed. In the early 1980s, PD&R recommended against establishing a multifamily coinsurance program, but HUD did establish such a program in 1983. Between 1983 and 1990, FHA insured more than 1,500 apartment projects, containing more than 350,000 apart- ments, with a total mortgage amount of over $10 billion. By the time the program was terminated, losses to FHA, and through FHA to the gov- ernment, were projected at $3.7 billion; losses in 1989 alone were about $2.5 billion. These losses are net of the amount recouped when FHA sold the project after default and foreclosure (Gerth, 1991; U.S. Government Accountability Office, 1992, pp. 20-22). The loss is a cumulative value for the entire program, rather than an annual flow of savings. CONCLUSIONS AND RECOMMENDATIONS PD&R has clearly produced important and valuable policy develop- ment work across a broad range of HUD program responsibilities and urban policy concerns. It has perhaps been most valuable when fundamen- tal reforms are necessary, but it has also contributed to numerous large and small policy initiatives and program improvements. At the same time, the committee has identified two significant limita- tions on PD&R’s ability to contribute effectively to the policy development process at HUD. The first is procedural. Each year all cabinet departments go through the process of preparing their portion of the budget plan that the President submits to Congress for the next fiscal year. The process is coordinated by OMB, and the budget is submitted to Congress, usually in February of each year. In each cabinet department, the process of pre- paring future budgets is almost continuous but usually involves intense departmental activity over the summer and fall of the year preceding the beginning of any fiscal year (e.g., the departments’ fiscal 2008 budget must be developed over the summer of 2006, approximately 16 months before the beginning of the fiscal year). While this process is commonly called “the budget process,” it also involves consideration of any change in policy that the President and the cabinet secretary wish to propose. As such, the president’s budget includes any legislative proposals that would be neces- sary to change existing laws or establish new programs and authorities to carry out the new policies. Just as the names of the offices vary among the cabinet departments (almost always containing the words policy, evaluation, research, or plan- 

EVALUATION OF POLICY DEVELOPMENT 111 ning), the roles of these offices vary in how they have traditionally been involved in the budget process. By design, this committee has representation from past assistant secretaries that headed two of these offices, the Office of the Assistant Secretary for Policy and Evaluation (ASPE) at the Depart- ment of Health and Human Services (HHS) and PD&R at HUD, so our comparisons are primarily between these two offices. At HHS, ASPE has traditionally worked closely with the Office of the Assistant Secretary for Management and Budget (ASMB) to develop the annual budget proposal. While ASMB is in charge of the process, ASPE has been assigned the task of working with the various operating divisions of the department to develop the legislative proposals that were part of the department’s budget submis- sion to OMB. Policy research and evaluation were always an integral part of designing the provisions of any legislative proposal, the rationale for any new or changed policy, and the estimates of the costs or savings that could be expected from the change in the policy. The planning process for new legislation typically involves issues about the behavior of consumers and sellers and how they respond to economic incentives. This means that policy research and empirical studies play an essential role in answering the crucial questions about how to design more efficient policies. As such, offices such as ASPE and PD&R that conduct policy-related research and keep up with current research by others have, or should have, a crucial role to play in the annual budget process. Yet, at HUD, PD&R has a minimal role in the design and preparation of the department’s budget proposal. The second limitation involves capacity. In policy development, as in research, the recent staff and funding reductions have made it more diffi- cult for PD&R to continue some valuable work. The staff reductions have resulted in the elimination of two divisions since the mid-1990s, the Policy Support and Program Demonstration Divisions. The former provided staff support for the Secretary’s Advisory Commission on Regulatory Barriers to Affordable Housing and conducted the study of voucher concentration in Baltimore neighborhoods. The latter conducted the Project Self-Sufficiency Demonstration in the mid-1980s, the first HUD effort to combine housing assistance with other services in an effort to help low-income families move from welfare to the labor market, and the forerunner of other, larger HUD initiatives with the same purpose. There are now no entities in PD&R in a position to conduct similar activities. The loss of these divisions is only part of the larger decline in staff for policy development. As reported in Chapter 2, the total staff in the Office of Policy Development has been cut in half since 1989, from 35 to 17 positions. The total staff in the Office of Economic Affairs (OEA), which also has major policy development functions particularly in the area of housing finance, has been cut by one-third, from 32 to 23 positions, while at the same time acquiring new responsibilities to support the secretary’s 

112 REBUILDING THE RESEARCH CAPACITY AT HUD much more extensive regulatory authority over the housing finance system. The transfer of those responsibilities to the new Federal Housing Finance Agency includes a transfer of OEA staff, which is likely to leave OEA with about half the staff it had in 1989, and the same responsibilities it had at that date. The loss of staff capacity in both offices is effectively greater than these figures, because they include the Research Utilization Division in the case of the Office of Policy Development and the Economic and Market Analysis Division in the case of the Office of Economic Affairs. The former unit does not have policy development responsibilities and the latter has a mixture of policy development and program operational responsibilities; their staffs have remained essentially constant. In addition to the loss of these specific abilities, PD&R has been ham- pered for almost a decade by the nearly continuous absence of a deputy assistant secretary for policy development. This position has been central to the conduct of many of the major policy initiatives since the creation of PD&R; it has most commonly been held by political appointees, some with personal ties to the secretary of the time. Major Recommendation 4: Formalizing what has been an informal practice over most administrations, the secretary should give PD&R’s independent, research-based expertise a formal role in HUD’s processes for preparing and reviewing budgets, legislative proposals, and regulations. Recommendation 6-1: The loss of staff capacity in offices and divisions that specialize in policy development should be reversed. Recommendation 6-2: The appointment of a deputy assistant secretary for policy development should be routinely given a high priority.   he T position is now filled by a long-time career employee.