Expansion of HUD’s “Moving-To-Work” Demonstration Is Not Justified
Other Approaches Would Promote Demonstration’s Goals More Effectively
September 28, 2011
Proposals before Congress would sharply expand HUD's Moving-to-Work (MTW) demonstration, which allows HUD to grant broad waivers of federal statutes and regulations to agencies that administer the public housing and Section 8 voucher programs in order to test experimental policies. An expansion of MTW is not justified: it would reduce the number of low-income families receiving housing assistance by shifting funds from the voucher program, expose more families to risky experiments than is needed to test innovative policies, and allow policies to diverge from one community to another to a degree that could weaken housing assistance programs.
In January 2011, Senator Vitter (R-LA) introduced legislation that would require HUD to expand MTW from 35 housing agencies to 250 agencies within three years. In addition, some members of Congress have called for adding an MTW expansion to the Section 8 Savings Act (SESA), a well-crafted package of reforms that the House may consider in the coming months.
Many of the individual agencies that participate in MTW are well run. Some have used their flexibility under MTW to increase efficiency or implement experimental policies that deserve testing. Nonetheless, a major expansion of MTW would very likely result in significant adverse consequences, and there is no persuasive rationale to support it.
Despite its name, MTW is not focused primarily on promoting employment. Some of MTW's most consequential provisions allow participating agencies to shift funds from vouchers to other uses. Agencies have used transferred funds for a range of purposes, including affordable housing development, agency administrative costs, and services for low-income families. These expenditures may have been worthwhile, but they have often come at the expense of assisting as many families as possible with available funds:
- MTW agencies use a much smaller share of the vouchers they were funded to administer than non-MTW agencies. In 2010, MTW agencies left idle 16 percent of the vouchers they could have supported with the funds they received, compared to just 4 percent for non-MTW agencies. As a result, more than 45,000 low-income families at MTW agencies were left without voucher assistance.
- MTW agencies assist far fewer needy families per dollar of federal funding than non-MTW agencies. In 2009, MTW agencies provided housing assistance to about nine families for each $100,000 in public housing and voucher funding, compared to 15 families for non-MTW agencies.
These disparities are particularly harmful at a time when budgets are tight and poverty, unemployment, and homelessness are high. In addition, MTW funding formulas typically provide more funding to MTW agencies than they would otherwise receive, which has reduced funding for non-MTW agencies.
MTW also has allowed agencies to eliminate key federal standards that are fundamental to the effectiveness of housing assistance, such as rules that give families the option to use a voucher to rent housing in a wide range of communities.
Expanding MTW would likely increase the magnitude of these adverse effects. Congress and HUD could further the central goals of MTW expansion -- providing appropriate flexibility and streamlining and testing experimental policies -- more effectively and with fewer risks through other approaches:
- Rigorous, targeted, temporary demonstrations would test policies to promote self-sufficiency and other goals more effectively. MTW was not designed as a research demonstration and has produced few concrete policy findings. HUD could test policies more effectively and with far less disruption if Congress authorized it to conduct small, rigorous demonstrations designed to assess specific types of policies. HUD's 2012 budget, for example, proposed a targeted demonstration to study alternative rent-setting policies.
- Where streamlining is appropriate, Congress should provide it through specific changes applying to all agencies, not sweeping authority for a select few. Streamlining the housing assistance program rules in certain areas would enable agencies to increase efficiency or better meet local needs without reducing the effectiveness of federal housing assistance. Examples include:
- Enactment of SESA, which would, for example, simplify calculation of tenant rents and allow more flexibility in the scheduling of housing quality inspections.
- Legislative and administrative changes to give agencies broader flexibility to support rehabilitation of public housing by converting public housing subsidies to "project-based" vouchers.
- Guidance and technical assistance for voluntary efforts by small housing agencies to ease administrative burdens by consolidating and coordinating operations.
- Added flexibility for agencies that score exceptionally well on HUD performance assessments, such as reduced HUD oversight of financial management at these agencies.
Proposals Would Expand MTW Sharply
Congress established MTW in 1996. In 2008, 25 agencies participated, and Congress has permitted HUD to increase this number to 35 agencies today through a series of expansions attached to appropriations legislation. HUD has also extended MTW to cover all families at some MTW agencies where it had previously only covered a small number of families. As a result of these changes, the number of vouchers and public housing units affected has grown by close to 100,000 in the last three years.
The Senate Appropriations Committee approved a 2012 HUD appropriations bill on September 21 that would continue this expansion, by permitting HUD to add three more agencies to MTW. A House subcommittee opted not to add any agencies when it passed a HUD appropriations bill on September 8, however, so it is uncertain whether an expansion will be included in final 2012 appropriations legislation.
Other expansion proposals would go much farther. In January 2011, Senator Vitter (R-LA) introduced legislation that would require HUD to expand MTW to include 250 agencies within three years. In addition, some members of Congress have called for adding MTW expansion to the Section 8 Savings Act (SESA), a package of well-crafted reforms to the voucher program and other housing assistance programs that the House may consider in the coming months.[i] The House Financial Services Committee in 2009 passed the Section 8 Voucher Reform Act (SEVRA), legislation that was similar to SESA but would have expanded MTW to 80 agencies.
Fewer than 1 percent of state and local housing agencies now participate in MTW, but the agencies administer about 12 percent of vouchers and public housing units. If MTW were expanded to 80 agencies and the new agencies were of about the same size as today's MTW agencies, close to 1 million vouchers and public housing units -- or nearly 30 percent of the national total -- would fall under the expanded demonstration. Senator Vitter's proposal to expand MTW to include 250 agencies could extend MTW to affect close to half of all vouchers and public housing units.
Moreover, even a modest increase in MTW's size would create momentum for further expansion. MTW has significant adverse consequences for agencies that do not participate. The preferential funding arrangements that MTW agencies receive can divert funds from other agencies, as described below, and by exempting a sizeable number of large agencies from many laws governing the housing programs, MTW can make it harder to generate the broad political support needed to improve those laws. If these consequences grow more severe under a larger demonstration, other agencies may feel added pressure to seek admission to MTW in order to enhance their funding prospects and gain special authority to disregard certain statutory requirements.
MTW's Most Consequential Provisions Cover Funding, Not Employment
MTW's name misleadingly suggests that the demonstration is primarily focused on promoting employment. In fact, MTW's most consequential provisions are those that permit HUD to set special MTW funding formulas and allow MTW agencies to shift funds from vouchers to other purposes. These provisions, which have only limited relevance to efforts to support work, have caused a shift in funds to MTW agencies from agencies that do not participate in the demonstration; they also sharply reduce incentives for MTW agencies to assist as many families as they can.
Added Funding to MTW Agencies Has Lowered Funding for Non-MTW Agencies
Many special MTW funding arrangements are highly favorable to participating agencies. As shown in Figure 1, MTW agencies, on average, received 35 percent more funding than other agencies in 2010 for each voucher they were authorized to administer.[ii] In 2009, when Congress enacted a funding policy (a "reserve offset") that reduced funding levels for non-MTW agencies, MTW agencies on average received 52 percent more funding per authorized voucher than non-MTW agencies.[iii]
The added funding for MTW agencies has sometimes come at the expense of other agencies. In five of the last seven years, voucher appropriations have fallen short of the amount for which agencies were eligible. When this occurs, HUD reduces funding for all agencies on a prorated basis. Consequently, each additional dollar that MTW agencies receive directly reduces funding levels for other agencies below the amount they need to fund their vouchers.
In 2009, for example, HUD reduced funding levels by 0.9 percent, forcing non-MTW agencies to assist more than 15,000 fewer families than they could have with full funding. If MTW agencies had been subject to the same funding formula as other agencies, the voucher appropriation would have been adequate to cover the full amount for which all agencies were eligible, and the proration would not have been necessary. This was also true of a 0.5 percent proration applied in 2010, and will likely be true of a 1 percent proration HUD announced it is implementing in 2011.
Some MTW agencies also receive public housing operating funds through special formulas. These are generally block grant formulas that ignore many factors (such as the number of occupied units) that the regular operating fund formula takes into consideration, and they tend to be even more generous than MTW voucher formulas. As Figure 2 shows, the ten agencies with special MTW public housing operating fund formulas received 75 percent more funding per public housing unit in 2010 than other agencies, on average.
Congress fully funded public housing operating subsidies in 2010, but provided inadequate funding in each of the previous seven years. Consequently, HUD was required to cut funding levels by proration in each of these years, and special funding arrangements that provided more funds to MTW agencies reduced funding levels for non-MTW agencies.
Formulas Eliminate Incentives to Assist as Many Families as Possible
In addition to receiving added funding, MTW agencies manage their funds under incentives that are very different from those faced by other agencies. Non-MTW agencies' voucher subsidy funding and administrative funding levels reflect the number of families they assisted in the previous year, so these agencies have a strong incentive to assist as many families as they can. Most MTW agencies, by contrast, are funded under special block grant formulas that give each agency a fixed dollar amount that rises annually by the rate of inflation.[iv] This eliminates the incentive for agencies to provide voucher assistance to as many families as possible, since agencies can leave funds unspent or use them for other purposes with no effect on their funding for the following year. The statute establishing MTW requires agencies to assist "substantially the same" number of families as they would without the funding flexibility MTW provides, but HUD has not enforced that requirement in a meaningful way. [v]
HUD data indicate that in 2010, MTW agencies used more than $400 million in voucher funds for other purposes or left the funds unspent. As a result, agencies participating in MTW used an unusually small share of the vouchers they were funded to administer. As Figure 3 shows, MTW agencies left idle 16 percent of the vouchers they could have supported with the funds they received in 2010; non-MTW agencies left just 4 percent of funded vouchers idle.[vi] More than 45,000 low-income families consequently were left without voucher assistance at MTW agencies, even though funds were available to help them.
Some MTW agencies use most of their voucher funds for voucher subsidies despite the weak financial incentives for doing so. Nine of the 30 agencies that participated in MTW throughout 2010 used at least 95 percent of their funded vouchers. But other agencies left large shares of their vouchers unused. Four agencies (Chicago, Philadelphia, New Haven, and Charlotte) used fewer than 75 percent of the vouchers they could have issued with available funds, accounting for about 25,000 of the 45,000 funded vouchers left unused. Another seven agencies used fewer than 85 percent.
MTW agencies could have left vouchers idle but used the savings to provide housing assistance to additional families through public housing or in other ways. In reality, however, the number of families receiving housing assistance outside the voucher program has often fallen at MTW agencies during their participation in the demonstration. Overall, HUD data and reports submitted by state and local agencies show that MTW agencies assist far fewer needy families per dollar of federal funding than non-MTW agencies.
As Figure 4 shows, MTW agencies in 2009 assisted about nine families for each $100,000 in public housing and voucher funding, while other agencies on average assisted 15 families with the same amount of money. Put differently, MTW agencies received about 60 percent more funds per family assisted. (Complete data are not yet available to make a similar comparison for 2010, but available data suggest that the disparity remained approximately 60 percent.) MTW agencies are located in relatively expensive markets where housing subsidies cost more, but this explains only a small share of the difference; the average HUD-determined Fair Market Rent in the areas served by MTW agencies in 2009 was just 13 percent more than the average for other agencies.[vii]
Many MTW agencies have used voucher funds to rehabilitate or replace public housing or build new affordable housing. Such investments could increase an agency's ability to provide housing assistance in the future to some degree, and the estimates above of the number of families assisted per dollar of federal funding do not take into account added families assisted in the future. It is unlikely, however, that the impact will come close to offsetting the number of families MTW agencies leave unassisted today.
Studies by the Government Accountability Office (GAO) and others have found that even over the long run, developing affordable housing is generally a less cost-effective way to help families afford housing than vouchers.[viii] Moreover, significant amounts of MTW agencies' 2009 housing development expenditures occurred after the units were completed and occupied, to cover debt incurred during construction and rehabilitation. These expenditures are an effective way to finance development, but they will not enable an agency to assist more families than it did in 2009.
Data Do Not Support Claim That MTW Has Enabled Agencies to Help More Families
A report on MTW that HUD submitted to Congress in August 2010 asserts that "most MTW agencies have actually served substantially more families than they would have been able to serve without MTW by streamlining operations and using accumulated funds to administer new assisted housing units."* Data from HUD and state and local agencies flatly contradict this claim. As discussed above, the data show that MTW agencies assisted substantially fewer families per dollar of federal funding, leaving tens of thousands of families unassisted who could have been helped with available funds.
The HUD report contains no analysis of data on the overall number of families assisted by MTW agencies or the cost of assisting them. It does include a table showing a sharp rise in the number of vouchers and public housing units that MTW agencies are authorized to administer. But these data offer no evidence that streamlining at these agencies has enabled them to serve more families with the same amount of funding. Rather, increases in the number of authorized units mainly reflect HUD decisions to award these agencies new vouchers to administer.
Some MTW agencies have streamlined procedures in ways that lowered administrative costs. Limited information is available on the magnitude of the savings, but the bulk likely stemmed from modest, commonsense improvements such as reductions in unnecessary housing quality inspections and tenant income reviews. These efficiencies did not, however, result in an overall increase in the number of families MTW agencies served, because they were outweighed by transfers of housing assistance funds to other purposes. Moreover, MTW expansion is not needed to enable agencies to streamline their operations; targeted legislative changes in SESA would give all agencies the opportunity to carry out many of the most promising efficiencies.
* HUD, Moving to Work: Interim Policy Applications and the Future of the Demonstration, August 2010, http://portal.hud.gov/hudportal/documents/huddoc?id=DOC_10134.pdf, p. 5.
Transfer of Voucher Funds to Other Purposes Raises Concerns
MTW agencies have used voucher funds for a variety of purposes other than development and rehabilitation of public and other affordable housing, including services for low-income families and agency administrative costs. It is difficult to assess these transfers fully, since data on the amount of funds shifted to various purposes are limited and the initiatives funded with the transfers have not been rigorously evaluated. Many of these initiatives appear to be well designed to meet real needs in the community. Nonetheless, the broad authority that MTW gives agencies to shift funds from rental assistance to other purposes raises a series of risks, which are likely to grow if MTW is dramatically expanded.
Leaving Needy Families Unassisted Has Harmful Consequences
Due to funding limitations, only about one in four eligible families receive housing assistance, and most communities have long waiting lists for assistance. A recent HUD analysis of Census Bureau data found that 7.1 million renter households had "worst-case housing needs" in 2009, meaning that they had incomes below half of the local median income, either lived in substandard housing or paid housing costs that exceed 50 percent of their income, and did not receive housing assistance.[ix] MTW policies that result in fewer families receiving assistance aggravate this already deep shortage.
The consequences of leaving needy families unassisted can be severe. A rigorous study has found, for example, that low-income families that were not offered vouchers were substantially more likely to experience homelessness, be compelled to double up with another family, or move frequently from one apartment to another, than similar families who were offered vouchers.[x] Homelessness, crowding, and housing instability, in turn, have been shown to be closely linked to poor health and development outcomes for children. [xi]
Sweeping Flexibility Increases Risk That Some Agencies Will Misuse Funds
Most MTW agencies appear to administer their programs responsibly. But because MTW permits large transfers of voucher funds with only limited reporting on how they are spent, the potential for waste or misspending is high. This potential is lower at non-MTW agencies, which must use their funding for vouchers or see it reduced the next year and must collect, verify, and report to HUD detailed information about voucher holders and their homes. Moreover, it is far more difficult for HUD to oversee the widely varying programs at MTW agencies than to oversee non-MTW public housing and voucher programs, which are largely consistent from one agency to the next.
The experience of the Philadelphia Housing Authority illustrates what can happen when an agency misuses the leeway MTW provides. The authority transferred more than $300 million out of its voucher program from 2005 to 2010, and in 2010 left about 9,300 low-income families unassisted as a result. Beginning in 2010, a series of reports surfaced that substantial funds were used for unnecessary payments to outside law firms, gifts for employees, social events, and unnecessary or excessively expensive improvements to housing developments and administrative buildings. [xii] In March 2011, HUD placed the authority under administrative receivership, and several federal investigations have been completed or are underway. There is no indication that Philadelphia is representative of MTW agencies, but an expansion of MTW would increase the risk that Philadelphia's experience would be repeated.
Funds Could Supplant Other State or Local Funds, Reducing Total Resources for Low-Income Housing
MTW permits housing agencies to shift housing assistance resources into areas that are often funded using state or local resources, including development subsidies and employment or social services. When this occurs, a state or locality could then withdraw state or local funds from these areas (or withhold funds they would otherwise use to increase expenditures in these areas) and use the funds to fill gaps elsewhere in its budget.
The result of this practice, known as supplantation, is that funding for housing assistance drops, but funding for the related areas to which funds are transferred is no higher than it would have been otherwise. Unlike certain other federal programs where a risk of supplantation exists, MTW does not impose a "maintenance of effort" requirement to ensure that states and localities maintain expenditure levels in related areas. In fact, MTW agencies are not required to report whether state and local funds for related purpose have been reduced or even the details of how the agency spent its own funds, so it is impossible to know to what degree supplantation has occurred. The risk of supplantation has grown in recent years, however, as states and local governments have struggled with difficult fiscal environments.
Transfers Could Weaken Efforts to Give All Agencies Adequate Administrative and Public Housing Funding
MTW agencies have used substantial funds that congressional appropriators provided for voucher subsidies to low-income families to instead support activities covered by voucher administrative funds and the public housing operating and capital funds. Public housing has been underfunded for many years, and Congress in 2011 reduced voucher administrative funding to a level that is likely to undermine the voucher program's effectiveness; in both cases, therefore, there is a strong justification for providing added resources. The need for such resources, however, extends to all agencies that administer public housing and vouchers, not just those in MTW.
The best way to address the shortfalls would be for Congress to increase appropriations for voucher administration and public housing, not to expand MTW to grant more agencies flexibility to shift funds away from vouchers for low-income families. Indeed, by granting select agencies a backdoor way to reallocate funding, MTW expansion reduces the pressure on Congress to provide adequate appropriations in these areas and could make it less likely that non-MTW agencies will receive adequate funding.
Large-Scale Funding Shifts Could Undermine Congressional Support for Voucher Program
Over the past decade, Congress generally has proven willing to provide funding to cover needed cost increases for the voucher and Project-Based Rental Assistance (PBRA) programs, which strictly limit expenditures to rental assistance and have funding formulas that are tightly linked to the cost of assisting a specified number of families. By contrast, HUD programs that fund a wide range of activities and whose funding levels Congress sets more arbitrarily each year have tended to see their funding erode or drop.
For example, from 2001 through 2011, funding for the HOME program lost 29 percent of its inflation-adjusted value and funding for the Community Development Block Grant (CDBG) lost 46 percent. Fiscal year 2012 appropriations bills passed by the Senate Appropriations Committee and a House appropriations subcommittee would make deep additional cuts to both programs. Funding for other low-income block grants that support a wide variety of activities has followed a similar pattern; for example, annual funding for the Temporary Assistance for Needy Families (TANF) block grant has lost 28 percent of its value to inflation since 1997.
This trend appears to reflect, at least in part, the fact that underfunding the voucher or PBRA programs directly cuts the number of families assisted, while the consequences of failing to adjust funding for a flexible block grant are less apparent and may be perceived as the responsibility of the state or local agencies that allocate the funds. A major MTW expansion would extend block grant type funding formulas and the authority to transfer funds to cover as much as half of voucher funding, and could raise considerably the share of voucher funds used for activities other than rental assistance. As this share grew, the risk would increase that Congress would treat the program more like CDBG or HOME and freeze or cut funding over an extended period.
MTW Not the Best Means to Streamline Program Rules
Some proponents of large-scale MTW expansion contend that it is needed not just to test specific innovative policies, but to provide broad deregulation of housing agencies to reduce administrative burdens or allow agencies to implement policies suited to their local communities. Eliminating or altering some housing assistance program rules, however, would likely do more harm than good. And while streamlining various other rules or providing new options to housing agencies would indeed be desirable, MTW generally is not the best way to accomplish that.
MTW Permits Changes to Rules That Have Helped Make Programs Effective
Many of the federal statutes and rules governing the rental assistance programs provide important protections for low-income families or help ensure that agencies use federal funds efficiently and effectively. Some rules could be streamlined or otherwise improved. Giving local agencies sweeping authority to alter basic program rules, however, risks causing significant harm.
For example, federal rules that give voucher holders the right to move anywhere in the country where there is a voucher program -- not just within the jurisdiction of the agency that issued the voucher -- play a key role in making vouchers effective. This right allows a worker who is laid off but finds a new job in a different county to use a voucher to move to an apartment within commuting distance of the new employer. Similarly, a victim of domestic violence can flee an abuser, an elderly person or person with a disability can move closer to a needed caregiver, and a family with children can move to an area with better schools -- all without losing their vouchers.
Eight MTW agencies, however, use their flexibility under MTW to limit or eliminate voucher portability. [xiii] Such restrictions may simplify administration for the agency and benefit local landlords, but they also make vouchers less responsive to the employment and other needs of low-income families, deny building owners in other communities the opportunity to rent to the affected families, and overturn a fundamental policy decision Congress made in designing the voucher program.
Complex Patchwork of Local Rules Would Harm Low-Income Families and Private Owners
In some cases, the existence of consistent national rules itself provides important benefits by simplifying the operations of the rental assistance programs. For example, families with housing assistance are generally required to pay 30 percent of their income in rent, after certain deductions. This rule applies (with limited variation) to all of the major HUD rental assistance programs and covers all of the private owners and non-MTW housing agencies that administer the programs. This consistency is a strength of the current system.
Consistent rent rules are critical to voucher holders' ability to move from one community to another. Even when MTW agencies do not explicitly restrict portability rights, it will be far harder for families to exercise those rights if they must sift through widely varying local rent rules to determine which localities they could move to without seeing their rent burden increase sharply.
Maintaining common rules from one locality to another also provides important consistency for the many private entities that participate in housing assistance programs in more than one locality (including property owners who rent to voucher holders, developers who use housing subsidies to back construction or rehabilitation of affordable housing developments, and lenders who finance such developments). Under the current system, these entities can easily assess the value of a housing subsidy in a given locality.
Finally, it is far easier for HUD to monitor and enforce a single set of national rules. HUD reports that the number of errors in tenant rent determinations fell by 65 percent from 2000 to 2009, in significant part due to HUD initiatives to provide technical assistance to agencies and monitor their performance.[xiv] Such quality control initiatives would be far more difficult if a large number of agencies set their own rules and the rules varied from agency to agency.[xv]
MTW permits participating agencies to establish their own rules for setting tenant rents. An expansion of MTW would replace rules that are largely consistent and effective with a complex patchwork of local alternatives whose costs and benefits are unknown. It would be wiser to test alternatives through modest-sized, rigorous experiments; if these experiments identify changes that would clearly improve on the current system, Congress could enact them, enabling all areas of the country to benefit.
Where Streamlining or Flexibility Is Appropriate, All Agencies Should Have It
Local flexibility is desirable in some aspects of the rental assistance programs, and housing agencies already have many options under current law. For example, they can target vouchers on particular subgroups of the population, such as working-poor families or people who are homeless. Such flexibility allows the programs to better respond to local needs and priorities without eliminating key standards that make the programs effective or fostering fundamental program differences that would make it more difficult for a family to move or a lender or owner to operate in multiple communities.
There are strong arguments for streamlining administrative requirements or giving agencies additional options in certain areas. MTW expansion is not the appropriate way to do this, however, since even under the broadest proposed expansion, the demonstration would cover only a small minority of agencies. Expansion would provide those agencies with more deregulation than is needed or beneficial, while doing nothing to help other agencies.
Where flexibility or streamlining is warranted, it should be made available to all agencies (or, in some cases, to all agencies that meet certain performance criteria), not to a select group as under MTW. Congress could achieve this by acting on a series of streamlining and flexibility proposals that have been put forward:
- SESA and SEVRA would give all agencies new flexibility in a number of areas, for example by simplifying calculation of tenant rents and allowing agencies to reduce the frequency of housing quality inspections.
- Giving agencies broader flexibility to convert public housing developments to project-based vouchers would provide more reliable funding, help agencies leverage private investment to rehabilitate public housing, and give low-income residents greater choice about where to live. HUD could also take administrative measures to facilitate conversions under current law.
- Voluntary consolidation or coordination by small housing agencies could also prove useful. About 2,900 of the 3,900 agencies that administer public housing and/or voucher programs administer fewer than 550 units. The large number of small agencies raises local administrative costs, hampers HUD oversight, and makes it more complex for low-income families to apply for assistance (since they must apply separately to each local agency to maximize their chances of receiving assistance). Congress and HUD could take steps to permit local agencies to ease these burdens while still maintaining separate identities and control over many policy and management decisions. For example, HUD could provide guidance and technical support to agencies that wish to enter into consortia to share administrative tasks with other agencies or to consolidate waiting lists to simplify the application process.
- In some areas, it would be sensible to extend added flexibility only to those agencies that perform particularly well on HUD's performance assessments. For example, it may not be necessary for HUD to conduct the same amount of oversight of high-performing agencies' financial management as it does for other agencies.
MTW Has Been Ineffective in Testing Policies in Areas Such as Promoting Self-Sufficiency
MTW was intended to test alternative housing policies, such as changes to rent rules, but has had only limited success in this regard. The lead author of the most complete assessment of MTW to date concluded that while MTW waivers may have positive effects, "there is relatively little hard evidence to show this, or to show that such effects are the result of having greater local autonomy as opposed to other causes." [xvi] Similarly, HUD's 2010 report on MTW acknowledged that "conclusive impacts of many MTW activities, particularly as they relate to residents, cannot yet be known."[xvii]
This is largely because MTW was not designed as an experimental evaluation, in which randomly selected families are assigned to receive housing assistance under alternative policies and are compared to otherwise-similar families who receive assistance under regular program rules. Such a design requires agencies to take on the added task of administering two sets of rules, and requires researchers to monitor the demonstration closely to ensure the design is effectively implemented. But use of random assignment is a standard feature of successful policy demonstrations, because without such a design it is very difficult to determine the actual effects of experimental policies.[xviii]
For example, nearly all MTW agencies have adopted their own rules for determining tenant rent payments. Many of these agencies claim that their policies have been beneficial, but major rent changes carry significant tradeoffs. One alternative implemented by some housing agencies establishes "flat rents" that are the same regardless of a tenant's income. Such rents are meant to encourage work, but they could also increase hardship or even homelessness (if the poorest tenants are charged more than they can afford) or waste money (if higher-income families receive larger subsidies than they need). Without an experimental evaluation across a variety of agencies, it is difficult to determine whether subsequent changes in employment, hardship, or costs stemmed from the flat rent policy or from other factors, such as local economic conditions or changes in the makeup of the agency's caseload.
In addition to being ineffective in generating meaningful research results, MTW is far more disruptive than is needed. MTW covers all families at most agencies, even though this exposes many more families to risky policies than is necessary to test the policies. It gives agencies flexibility to shift housing assistance funds to other purposes even when this is unrelated to the policies being tested. And policies enacted under MTW remain in place far longer than is needed to test them. Agencies were admitted to the demonstration starting in 1998 for five-year periods, but MTW agreements nearly always have been extended. Moreover, HUD has entered into longer agreements in recent years than it did early in the demonstration, with some lasting as long as ten years.
By contrast, housing policy demonstrations such as Moving-to-Opportunity (MTO) and Jobs Plus have generated a far greater quantity of useful findings than MTW, with much less disruption to tenants. These demonstrations involved only about 5,000 families each (across multiple sites and including control groups) and were carried out over five years or less. By comparison, the current MTW demonstration affects more than 400,000 families, and an expansion to 80 agencies would affect as many as a million -- 200 times as many as MTO or Jobs Plus.
A series of demonstrations conducted in the Aid to Families with Dependent Children program prior to the enactment of welfare reform legislation in the 1990s offers another example of rigorously evaluated policy experiments that generated substantial policy lessons. Like MTW, these demonstrations tested a range of largely state- and locally-designed experimental policies, including alternative benefit formulas, time limits, work requirements, and investments in job training and other services. But because the federal government consistently required rigorous evaluation, the experiments produced a wealth of information about the impacts of various policies.
If Congress wishes to identify which self-sufficiency policies work and should be scaled up, it could do so most effectively by creating targeted, temporary, and rigorously evaluated demonstrations -- not by expanding MTW, which has failed for over a decade to generate meaningful policy findings. HUD's 2012 budget proposes a demonstration of alternative rent policies that moves toward this more promising approach.
A major expansion of MTW would fundamentally change the voucher and public housing programs. It would sharply increase the harmful impact of MTW's special funding provisions and expose a much larger number of families to untested policy changes that could reduce the effectiveness of federal housing assistance.
Testing innovative policies and streamlining housing assistance programs are important goals. Substantially expanding MTW, however, is not the best way to achieve them. Congress should limit the scope of any research demonstration -- whether it is labeled an expansion of MTW or not -- to the scale needed for the demonstration, allow untested policies to be tried for the period needed to evaluate them (rather than indefinitely), prohibit transfers of voucher funds, and mandate a rigorous evaluation. And where deregulation is appropriate, Congress should make it available to all agencies nationwide, not a select minority of agencies participating in MTW.
[i] For additional information on SESA, see the testimony by Barbara Sard before the House Financial Services Subcommittee on Insurance, Housing, and Community Opportunity on June 23, 2011, http://www.cbpp.org/cms/index.cfm?fa=view&id=3517.
[ii] The appendix provides an explanation of the sources and methods used for this calculation and others in this analysis. HUD has not released 2011 voucher funding levels for individual agencies, so it is not yet possible to compare 2011 funding for MTW and non-MTW agencies, http://www.cbpp.org/files/9-27-11hous-appendix.pdf .
[iii] A "reserve offset" is a policy of reducing an agency's current year funding level based on unspent funds the agency has accumulated in prior years. In 2009, Congress gave HUD authority to offset reserves at non-MTW agencies but not MTW agencies. Congress did not enact a reserve offset in 2010 or 2011, and a House subcommittee did not include an offset in the 2012 HUD appropriations bill it approved on September 8. The Senate Appropriations Committee did include an offset in the version it passed on September 21, but it gave HUD authority to apply the offset to both MTW and non-MTW agencies -- making it less likely to contribute to the imbalance in funding.
[iv] MTW funding arrangements also typically provide for increases when an agency receives an increase in the number of vouchers it is authorized to administer, as occurs when a public housing development is demolished and vouchers are issued to replace it.
[v] Available data suggest that the majority of MTW agencies would be found to violate the "substantially the same" requirement if it were rigorously enforced. Agencies are required to certify each year that they meet the requirement, but they are not required to demonstrate their compliance. Moreover, HUD has not defined key aspects of the requirement, such as how agencies should calculate how many families they would have assisted without MTW funding flexibility or how much the number of families assisted can decline and still be considered "substantially the same."
Some HUD and agency documents appear to misinterpret the "substantially the same" provision to require that agencies assist a number of families equal to a "baseline" number they served before they joined the demonstration. Since the MTW demonstration began, however, the number of voucher and public housing units that Congress has authorized has grown significantly -- by 20 percent at all agencies and by 47 percent at MTW agencies from 1998 to 2009, according to HUD. Thus, an agency that assisted approximately the same number of families as it did before entering the demonstration will typically be serving far fewer families than it is authorized to administer, and far fewer than it could assist if it used funds for the purposes for which they were appropriated. HUD is in the process of revising requirements for MTW agencies to report data regarding the number of families they assist and in other areas, but it is unclear whether this revision will significantly improve enforcement of the "substantially the same" requirement.
[vi] The percentage of all authorized vouchers left unused in 2010 was 15 percent at MTW agencies and 8 percent at non-MTW agencies. The number of vouchers that an agency could have issued with available funds may be higher or lower than the number of vouchers Congress and HUD have authorized the agency to administer. At certain MTW and non-MTW agencies, some vouchers were left unused because the agency did not receive sufficient funds to support them. In addition, MTW agencies (unlike non-MTW agencies) are permitted to issue more than their authorized number of vouchers if they have funds available. In 2010, 15 MTW agencies received sufficient funds to support more than their authorized number of vouchers as a result of various favorable provisions in their funding formulas.
[vii] The limited evidence available suggests that differences in family size also fail to explain the discrepancy in the number of families that MTW and non-MTW agencies assist per dollar of federal funding. Many MTW agencies report no data on the number of people in the families they assist or the number of bedrooms in the units those families live in. The five largest MTW agencies, however, did report data on unit size in 2009. The data show that those five agencies assisted families living in units that were marginally larger than those typically occupied by families assisted by non-MTW agencies, but the difference was only sufficient to explain a 3 percent difference in costs -- a small fraction of the 60 percent difference in funding per family assisted.
[viii] Government Accountability Office, Federal Housing Assistance: Comparing the Characteristics and Costs of Housing Programs; GAO-02-76, January 2002, http://www.gao.gov/new.items/d0276.pdf. Modest renovation of existing public housing can be less expensive than the development efforts that GAO examined and could consequently prove comparable to vouchers in cost effectiveness over the long run. The available information, however, suggests that MTW agencies have emphasized more extensive renovation and replacement projects, similar to those studied by GAO. A number of MTW agencies, for example, have obtained waivers of HUD development cost limits to enable them to undertake more expensive projects.
[ix] Barry Steffan et al., Worst-Case Housing Needs, 2009: A Report to Congress, Office of Policy Development and Research, Department of Housing and Urban Development, February 2011.
[x] Gregory Mills et al., "Effects of Housing Vouchers on Welfare Families," prepared by Abt Associates for the HUD Office of Policy Development and Research, 2006.
[xi] See, for example, Diana Becker Cutts et al., "US Housing Insecurity and the Health of Very Young Children," American Journal of Public Health, August 2011, p. 1508.
[xii] See for example Philadelphia Inquirer, "PHA Spent Lavishly on Pet Projects, Upgrades," December 19, 2010, http://www.philly.com/philly/news/homepage/20101219_PHA_spent_lavishly_on_pet_projects__upgrades.html ; and HUD Office of Inspector General, The Philadelphia, PA, Housing Authority Did Not Comply with Several Significant HUD Requirements and Failed to Support Payments for Outside Legal Services , Audit Report 2011-PH-1007, March 10, 2011,http://www.hud.gov/offices/oig/reports/files/ig1131007.pdf.
[xiii] One of MTWs three statutory objectives is to "increase housing choices for low-income families." Despite this goal, HUD has approved waivers whose main impact is to reduce housing choice.
[xiv] ICF Macro International, "Quality Control for Rental Assistance Subsidy Determinations: Final Report for FY 2009," prepared for the HUD Office of Policy Development and Research, October 16, 2009.
[xv] Rent errors do not directly affect federal expenditures for agencies with block grant funding formulas, since rent payments at these agencies have no direct effect on agency funding levels. Errors would nonetheless be significant, however, since they could harm families or building owners or cause federal funds to be misspent.
[xvii] HUD 2010, p. 4.
[xviii] HUD has begun to require MTW agencies to report more information on program activities and outcomes, and 2010 appropriations legislation directed HUD to carry out an evaluation of the demonstration. HUD has indicated that it is procuring a third-party contractor to conduct this evaluation. These efforts may generate some useful information, but since they will not alter the basic design of the demonstration at participating agencies, they are unlikely to generate rigorous research findings.