“SEVRA” Housing Voucher Reform Bill Would Update and Streamline Program
But Deregulation Provision Holds Significant Risks
May 13, 2010
The Section 8 Voucher Reform Act (SEVRA), which the House Financial Services Committee approved in July 2009, contains a series of important, carefully crafted measures to strengthen the housing voucher program. Most significantly, it would help and encourage state and local housing agencies to assist more needy families within the available funds, an important improvement at a time of rising poverty and homelessness. In addition, the bill would reduce administrative burdens for housing agencies and private owners, strengthen work supports, and provide more flexible and effective assistance to low-income families.
Vouchers, the nation’s most widely used form of low-income housing assistance, help roughly 2 million low-income families afford modest rental units of their choice in the private market. Studies have shown that vouchers are highly effective in reducing homelessness and housing instability (both of which have been linked to a range of developmental problems among children) and help families move to lower-poverty neighborhoods with better schools and less exposure to crime.
Some of SEVRA’s key provisions would:
- Establish a stable, efficient voucher funding system that would allow agencies to help more needy families with the funds they receive, control program costs, and reduce the risk that agencies will be forced to cut assistance in response to funding shortfalls;
- Simplify rules for setting tenant rent payments, while continuing to cap rents at 30 percent of a tenant’s income;
- Help develop and preserve affordable housing through broader use of “project-based” vouchers (which, unlike more widely used “tenant-based” vouchers, can be tied to a particular housing development); and
- Protect tenants whose landlords face financial difficulties by giving agencies new tools to ensure that buildings are kept in habitable condition.
However, one problematic section of SEVRA threatens to undercut the improvements in the remainder of the bill: a sharp expansion of HUD’s “Moving-to-Work” (MTW) demonstration, which the bill would also rename the “Housing Innovation Program” (HIP). Under MTW, HUD can permit housing agencies to operate their voucher and public housing programs without regard to many federal statutes and regulations.
MTW has enabled some agencies to implement innovative, promising initiatives. It also has permitted policies with the potential to harm vulnerable families, such as large rent increases and time limits on assistance, even for working-poor families that cannot afford housing on their own. In addition, MTW has enabled some agencies to divert funds away from the voucher program, leaving tens of thousands of families without assistance even though Congress provided adequate funds to help them. Finally, the demonstration has not been subject to rigorous evaluation and consequently has generated few useful policy lessons.
SEVRA’s sweeping expansion of the demonstration could expose close to 750,000 families to the riskiest experimental policies — far more than is necessary for research purposes — and make it difficult for HUD to effectively evaluate and monitor the results. It will be important that Congress add key protections to SEVRA’s HIP provisions, including a limit on the number of families that can be exposed to policy experiments and firm restrictions on the diversion of voucher funds.
In addition, it will be important for Congress to revise or eliminate a provision of SEVRA that would require all adults in households that receive or apply for vouchers to provide certain specified forms of identification. While ostensibly aimed at preventing undocumented immigrants (who are ineligible for vouchers) from receiving assistance, the provision’s largest impact would likely be to prevent a significant number of eligible U.S. citizens and legal immigrants from receiving needed housing assistance because they are unable to produce the required documents quickly enough.
Congress should act expeditiously on SEVRA. More than a decade has passed since the last major authorizing legislation that affected housing vouchers, the 1998 Quality Housing and Work Responsibility Act. Vouchers have emerged as the centerpiece of federal low-income housing policy during that period, as policymakers have increasingly emphasized the voucher program’s market-based approach. There are substantial benefits to be reaped from updating the voucher program to reflect its expanded role and the lessons that have been learned over time.
In addition, SEVRA is urgently needed to help low-income families cope with widespread foreclosures and rising homelessness and poverty during the economic downturn. Particularly timely are the bill’s provisions enabling agencies to assist more families within the available resources and to protect tenants whose landlords face financial difficulty. The sooner SEVRA is enacted, the sooner it will begin to help families that are at risk of losing their homes or are already doubled up with friends or relatives, living in shelters, or on the streets.
New Funding System Would Assist More Families Without Adding to Federal Costs
The most important improvements in SEVRA would establish a stable, efficient system of distributing funds to renew vouchers to the more than 2,400 state and local agencies that administer the program. SEVRA’s funding provisions would provide agencies with tools and incentives to assist more families with the funds they receive.
In addition, the provisions would reduce the chances that agencies will need to abruptly cut assistance because the funding available to them is inadequate to cover voucher costs. Many agencies faced shortfalls in 2009, in part because the economic downturn pushed down voucher holders’ incomes. (Voucher holders generally pay 30 percent of their income in rent and the voucher covers the rest, so when families’ incomes decline in a recession, voucher program costs increase.) As a result, some agencies were compelled to cut costs by placing additional rent burdens on needy families or shelving vouchers as participants left the program rather than reissuing them to other eligible families. Harsher measures — such as cutting off assistance to families already using vouchers — were avoided in 2009 only because Congress and HUD acted to make added funds available to agencies facing shortfalls. SEVRA would lower the odds that future shortfalls will inflict difficult cuts on families or force urgent mid-year action to avert them.
Importantly, the funding formula changes would enable agencies to help more families and cushion the program against shortfallswithout adding to federal costs. These changes would enable the program to use the funding Congress provides more effectively. 
Protection Against Economic Volatility
Two key changes in SEVRA would address the effects of volatility in market rents, tenant incomes, and other economic factors on the voucher program.
Congress sets the amount of funding for vouchers each year in appropriations acts. It is difficult, however, for federal policymakers or agency managers to predict the exact cost of state and local voucher programs. The number of vouchers in use at a given time depends on the share of families issued a voucher that successfully use it to lease housing (which in turn may reflect vacancy rates in the local rental market) and the number of families whose incomes rise enough that they no longer need assistance or leave the program for other reasons. Also, the average cost of a voucher can vary due to fluctuations in market rent and utility costs (within caps established by voucher program rules) and the incomes of low-income families.
This unpredictability can have two adverse consequences. First, when voucher costs turn out to exceed the funding available, an agency may be compelled to make abrupt cuts, like those that occurred in 2009 in some areas. Second, agencies may deliberately leave some vouchers and voucher funds unused out of concern that factors beyond their control will temporarily push up their expenses and force them to make painful cuts. SEVRA would address both of these risks through measures to help agencies respond when voucher costs are higher than expected, without raising overall program costs:
- Temporary advances for agencies that exhaust their voucher funds. SEVRA would create an advance-funding mechanism that would work like overdraft protection. An agency that has insufficient funds in the last quarter of the calendar year to make all of the rent payments due to owners could borrow a small portion of its funding for the following year, which would then be subtracted from the funding allocated to the agency a few months later.
- Modest reserve funds. SEVRA would allow agencies to accumulate 6 percent of their voucher funds as a cushion against unexpected costs. (These reserves are not a substitute for the advance mechanism, since agencies that used all their funds in previous years would not have accumulated reserves.)
More Stable, Efficient Funding Formula
SEVRA would establish, as a permanent part of the statute that authorizes the voucher program, a formula that bases each agency’s voucher funding for a year on the cost of the vouchers it used in the previous year, adjusted for inflation and certain other factors. This would control costs by forcing agencies to manage within a fixed budget (with reserves and advances providing only a limited cushion) while also ensuring that each agency’s funding level matches its actual needs.
From 2004 to 2006, HUD allocated voucher funds using a series of inefficient formulas that gave some agencies less funding than they needed to maintain assistance to the families they were serving, but provided other agencies with more funds than they could use. As a result of these flawed formulas, together with inadequate overall funding, about 165,000 vouchers were taken out of use. Since 2007, Congress has instituted a funding formula similar to that in SEVRA on a year-by-year basis in annual appropriations legislation. This change has enabled agencies to restore about 100,000 of the vouchers that had been lost. SEVRA would build on this progress (and help agencies recover from any cuts resulting from the 2009 shortfalls) by making the formula permanent, thus providing agencies with greater confidence that they will receive funding based on their needs.
Additional Incentives to Serve More Families
SEVRA would also establish other incentives for agencies to assist as many families as possible:
- “Use it or lose it” provision. While SEVRA would allow agencies to retain modest reserves, it would also make clear that any unused funds beyond the permitted reserve amount would be reallocated to other agencies that can use them. This balanced approach is important, since without a limit on reserves some agencies may accumulate large amounts of unspent funds (as many did between 2004 and 2006) rather than use the funds to assist needy families.
- Elimination of cap on number of families agencies can assist with available funds. Recent appropriations acts have prohibited agencies from subsidizing more vouchers than they are authorized to administer, even if an agency has funds available that could be used for this purpose. SEVRA would allow agencies to assist families beyond their authorized level and generally would include the cost of those added vouchers in determining an agency’s funding for the next year. This would encourage agencies to keep per-voucher costs low, since they could use any savings to assist more families, even above their authorized voucher cap.
- More administrative funding for agencies that use more vouchers. From 2004 to 2007, HUD distributed funds to cover the cost of administering vouchers without regard to how well an agency performed. SEVRA, in contrast, would encourage agencies to use as many vouchers as possible by requiring HUD to allocate administrative funds primarily based on the number of vouchers the agencies put to use. (Appropriations bills from 2008 to 2010 have adopted this policy, but SEVRA would make clear that Congress intends to retain it going forward.) SEVRA also would allow HUD to add incentives for agencies to perform well in other areas of program administration.
Other Reforms Would Make Program More Effective and Efficient
SEVRA would make a series of other important reforms as well, including:
- Streamlining calculation of tenant rent payments. Tenants in HUD’s rental assistance programs generally pay 30 percent of their income for rent, after certain deductions. SEVRA would streamline the calculation of tenant incomes and deductions to reduce administrative burdens. These changes would apply to public housing and the separate project-based Section 8 program, in addition to the voucher program.
For example, SEVRA would replace the complex earned income disregard now in place with a simple provision under which 10 percent of a family’s first $9,000 in earnings would be deducted from the family’s income. It also would allow agencies and owners to review the incomes of tenants with fixed incomes — such as elderly people and people with disabilities who receive Supplemental Security Income (SSI) — every three years instead of every year and to assume that in the intervening two years the tenant’s income rose at the rate of inflation. (SSI, Social Security, and certain other such benefits are adjusted annually for inflation.)
- Modifying housing inspection rules. HUD requires agencies to inspect apartments rented with vouchers to certify that they meet federal quality standards. SEVRA would allow agencies to make certain changes in their inspection rules to ease administrative burdens and encourage landlords to rent apartments to voucher holders. For example, it would allow agencies to conduct inspections every two years instead of annually.
- Helping develop and preserve affordable housing through expanded use of “project-based” vouchers. SEVRA would expand agencies’ authority to enter into agreements with owners for a share of their vouchers to be used at particular housing developments. Such “project-basing” can promote key policy goals. For example, an agency could partner with a social service agency to provide supportive housing to formerly homeless people or could subsidize mixed-income housing in a neighborhood with good schools and low crime but a tight rental market.
- Directing HUD to remove barriers to voucher “portability.” Voucher holders generally have a statutory right to move from one agency’s jurisdiction to another’s without losing assistance. Such “portability” can enable a family to relocate to a neighborhood with good schools and lower crime, a worker to move closer to a job or to take a new job in another community, an elderly person or person with a disability to move close to family or a needed caregiver, or a domestic violence victim to flee an abuser. However, many voucher holders fail to benefit from this option because agencies currently face disincentives to help families relocate to other jurisdictions.
The biggest disincentive is that the agency that first issues a voucher to a family must cover the cost of the voucher indefinitely after the family moves, unless the agency in the destination community voluntarily accepts the cost by “absorbing” the voucher. This is administratively cumbersome and can carry added costs for the issuing agency if the community to which the family moves has higher rents than the community the family left. For their part, destination agencies are often reluctant to absorb portability vouchers because that would divert scarce resources away from families on the agency’s own waiting list.
SEVRA would direct HUD to issue regulations that minimize the extent to which one agency must pay another for the voucher of a family that has relocated, without preventing either agency from assisting families from its own waiting list. HUD could achieve this, for example, by requiring destination agencies to absorb the voucher while providing those agencies funds to cover the costs. This would treat both agencies equitably and ensure that the portability process is not unnecessarily cumbersome.
- Protecting tenants when property owners face foreclosure or other financial difficulty. Owners who rent to voucher holders sometimes fail to maintain the units in decent condition or to pay utility bills for which they are responsible. Currently, if an owner does not make needed repairs or utility payments within a reasonable time, the agency has no choice but to terminate the subsidy payment and require the family to move. SEVRA would encourage owners to repair properties that fail to meet standards by allowing voucher families to remain temporarily while suspending subsidy payments until repairs are made. In addition, it would allow agencies to use the suspended payments to make or contract for repairs or to cover the security deposit and moving expenses if a family ultimately is forced to move to another unit.
Similarly, when owners fail to make utility payments, agencies could use subsidies that would normally be paid to the owner to pay the utility bills.
SEVRA would also make permanent several protections Congress enacted in May 2009 to prevent voucher holders from being displaced when the unit they rent is subject to foreclosure. Those protections would otherwise expire at the end of 2012.
- Stabilizing funding for the Family Self-Sufficiency program. The Family Self-Sufficiency program (FSS) encourages work and savings among voucher holders and public housing residents through financial incentives and employment counseling and services. HUD distributes funding for agency staff to counsel participants and coordinate employment services through annual competitions, but because HUD has changed the criteria for these competitions repeatedly in recent years, many agencies have experienced abrupt funding cutoffs and overall FSS enrollment has declined. To reverse this decline, SEVRA would provide stable funding for FSS staff, as well as a permanent mechanism to provide added funding to cover costs from FSS financial incentives.
Sweeping Expansion of “Moving-to-Work” Demonstration Poses Risks
SEVRA would rename HUD’s “Moving-to-Work” (MTW) demonstration the “Housing Innovation Program” (HIP) and permit HUD to raise the number of participating agencies from 33 today to 80, while placing a number of new requirements on participating agencies.  Established by legislation in 1996, MTW permits HUD to grant agencies broad waivers that allow experimentation with a wide range of policies. Agencies can use these waivers to test promising, innovative policies, as well as policies with serious risks for low-income tenants such as sharp rent increases or time limits on assistance even for working-poor families that cannot afford housing without ongoing help. MTW also allows HUD to grant waivers that let agencies dispense with most federal rules, with deregulation as an end in itself. 
Demonstration’s Record Raises Concerns
MTW was intended to test innovations in housing policy, but its success in this regard has been limited. The demonstration’s design did not provide for careful evaluation of the policies with which agencies experimented. As a result, MTW has generated a wealth of anecdotal reports but few firm, objective findings. Targeted, rigorously evaluated housing policy demonstrations such as Moving-to-Opportunity, Jobs Plus, and the Welfare-to-Work Voucher program have generated a far greater quantity of useful findings than MTW, with much less disruption to tenants.
MTW also does little to guarantee that agencies will be held accountable for the policies they adopt or fully disclose how they have used their flexibility under the demonstration. HUD’s Office of the Inspector General has issued a series of sharply critical reports on MTW that have noted flaws such as ineffective oversight by HUD and poor use of funds.
MTW agencies are permitted not only to experiment with new policies, but also to shift funds between the voucher and public housing programs and to use funds for some activities that neither program normally supports. Generally, HUD also has permitted MTW agencies to accumulate unlimited amounts of unspent funds as reserves, unconstrained by the limitations that apply to other agencies.
From 2005 to 2008, MTW agencies used more than $1 billion in voucher funds for purposes other than voucher assistance or accumulated the funds as reserves. In 2008 alone, MTW agencies diverted about $400 million. MTW agencies left about 30,000 vouchers unused in 2008 that could have been used with funding they had available.
There is no comprehensive information on how MTW agencies have used the diverted voucher funds, but at least some have used substantial amounts to renovate or maintain public housing developments. Agencies likely felt pressure to make such transfers because public housing capital funds and operating costs have been deeply underfunded in recent years (although this underfunding now appears to be easing considerably ). An agency may be particularly reluctant to allow public housing developments to fall into disrepair, since an agency may believe it would face greater public criticism for broken windows or boarded-up units at public housing projects than for the higher rent burdens and longer waiting lists that would result from cuts to the voucher program.
However, diverting large amounts of voucher funds to renovate existing public housing units or for other permitted purposes (such as to cover administrative costs or accumulate large reserves) is unlikely to help substantially more needy families in the near term — and certainly not nearly enough additional families to offset the tens of thousands of vouchers left unused. The net result of the diversion of voucher funds has been to significantly reduce the number of low-income families assisted. Moreover, large-scale diversion of funds from vouchers to public housing usurps the authority of congressional appropriators to determine funding levels for the different housing programs.
Expansion Provision Could Undermine Improvements in SEVRA
Several of SEVRA’s HIP provisions would improve on MTW by placing needed requirements on participating agencies. Agencies would have to consult with residents and the public about HIP plans and issue periodic, publicly available reports on the activities carried out under the demonstration. HIP would also be subject to stronger evaluation requirements than MTW. In addition, SEVRA would retain a number of important statutory provisions that HUD was permitted to waive under MTW, such as portability (see above) and prohibitions against unjustified evictions. These changes could help the demonstration generate meaningful policy lessons while holding agencies accountable and reducing risks for tenants.
Unfortunately, the Financial Services Committee approved an amendment that removed or weakened further improvements to the demonstration. The most important of these would have required agencies to continue to serve at least 98 percent of the families they served at the time they entered the demonstration (adjusted to reflect any new vouchers issued to the agencies or funding shortfalls that force them to cut back on assistance) and prohibited agencies from transferring funds out of the voucher program except to test a specific policy innovation. Without these requirements, the bill offers little protection against the harmful diversion of voucher funds that has left a substantial number of needy families without assistance. 
Moreover, even the significant improvements the bill does contain could be undermined by the SEVRA provision expanding the demonstration to 80 agencies. The bill requires that HUD admit “both large and small” agencies to HIP, but this vague standard would allow HUD generally to favor the largest agencies for admission, as it has under MTW. Less than 1 percent of the nation’s housing agencies participate in MTW today, but these agencies administer about 11 percent of vouchers and public housing units. If agencies admitted to HIP are of approximately the same average size as the agencies HUD has opted to admit in the past, close to 1 million families — or nearly 3 in 10 voucher holders and public housing residents nationally — could be subject to the expanded demonstration. This would pose several problems:
- It would expose far more families than necessary to untested policies that could have adverse effects. Under SEVRA, HUD could grant 60 agencies broad authority to experiment with the riskiest policies. (SEVRA would reserve the other 20 of the 80 HIP slots for agencies participating in a “HIP-lite” component that would provide narrower flexibility.) This could expose close to 750,000 families to untested measures such as time limits or alternative rent policies that result in rent increases. Some limited testing of these policies may be justified to better understand their impact, but the proposed scope of HIP is much greater than needed.
- Losses of vouchers due to fund transfers could rise sharply. If HIP agencies divert voucher funds at the same rate MTW agencies do today, as many as 75,000 vouchers could be left unused and the number of families assisted could drop significantly. This would run counter to SEVRA’s central goal of assisting as many needy families as possible with available funds.
- The demonstration could be too large for HUD to effectively monitor and evaluate. Since each MTW agency undertakes different activities and is subject to different rules, it takes more resources for HUD to effectively oversee those agencies than it does to oversee agencies following normal federal rules and standards. The serious problems that HUD’s Office of the Inspector General has identified with MTW oversight stemmed partly from lack of staff resources for monitoring and oversight. Similarly, effectively evaluating participating agencies is a substantial task. Even with HIP’s improved transparency and evaluation requirements, HUD likely would struggle to oversee and evaluate a demonstration that is nearly three times MTW’s current size.
These risks would be substantial even if HUD sought to limit the number of families exposed to HIP or to discourage housing agencies from implementing policies with serious risks for tenants. HUD likely would face considerable political pressure (for example, from mayors of large cities) to admit a number of large agencies to the demonstration. HUD would have some authority to limit a participating agency’s flexibility to operate outside normal rules, but HUD could face pressure to exercise that authority sparingly.
Moreover, SEVRA would allow HUD in the future to exempt a HIP agency from additional standards and rules, beyond those HUD approves to at the time the agency enters the demonstration. As a result, regardless of the types of flexibility HUD might emphasize immediately after SEVRA’s enactment, a future administration with different priorities could allow agencies to implement very different policies, such as sharper rent increases or more stringent time limits.
Changes in HIP Provisions Could Strengthen Demonstration and Protect Families
HIP has the potential to serve as an effective testing ground for future housing policies while at the same time minimizing risks of harm to low-income families or diversion of voucher funds. But achieving these goals will require changes to the provisions now in SEVRA. Two improvements are particularly crucial:
- SEVRA should cap the number of families exposed to the riskiest policies, not just the number of participating agencies. Such a cap would not only limit the number of low-income families exposed to untested and potentially harmful policies, but also ensure that small- and medium-sized agencies — which have been underrepresented in MTW — would have the opportunity to participate.
- SEVRA should place firm restrictions on diversion of voucher funds. Such diversions can harm needy families and seldom serve a useful research purpose. Congress should either prohibit diversion of funds entirely, or, at a minimum, require that an agency suspend diversion if the number of families it assists falls significantly below the number it assisted when it entered the demonstration.
Identification Requirement Would Effectively Deny Vouchers to Many Eligible Families
The House Financial Services Committee amended SEVRA to require all adults in households that receive or apply for vouchers to provide identification documents from a very restrictive list (for example, a Social Security card together with a government-issued photo ID, or a passport). Proponents framed this measure as an effort to prevent undocumented immigrants from receiving assistance, but the voucher program already prohibits this and there is no evidence a significant number of undocumented immigrants have ever used vouchers.
The provision’s most likely impact instead would be to deny or terminate vouchers for U.S. citizens and legal immigrants who are otherwise eligible for vouchers but cannot readily produce the required forms of identification. Indeed, when a similar requirement was enacted in Medicaid in 2006, studies by the Government Accountability Office, Congressional Research Service, and a number of states indicated that the primary impact was on U.S. citizens, tens of thousands of whom had Medicaid coverage denied or delayed. In addition, the costs and burdens of administering the requirements proved to be substantial.
Studies show that disproportionate numbers of African Americans and senior citizens lack a government-issued photo ID, so these groups may be among those that have the most difficulty complying with the SEVRA requirement. In addition, people who have a mental illness, have experienced bouts of homelessness, or lack safe access to their former homes because of domestic violence may struggle to produce the documents and could be denied vouchers even if they have a pressing need for housing assistance.
SEVRA would build on the voucher program’s many strengths through a series of measured, targeted improvements to make the program more effective and efficient. Congress could improve the legislation by placing some additional restrictions on its HIP provisions and revising or eliminating its problematic identification requirement. Those improvements would involve only modest changes to the bill as a whole.
 For a detailed comparison of SEVRA’s provisions with current law, see http://www.cbpp.org/files/9-22-09hous-appendix.pdf.
 A separate provision of SEVRA would authorize the creation of 150,000 new “incremental” vouchers in 2010 to meet a portion of the large unmet need for housing assistance. That provision would expand the voucher program at an increased cost, but only if congressional appropriators separately provided funds for the voucher expansion.
 In fiscal year 2010 appropriations legislation, Congress permitted HUD to admit three more agencies to MTW, which would result in a total of 36 agencies.
 For additional analysis of the proposed MTW expansion, see Will Fischer, “Sharp Expansion of HUD’s ‘Moving-to-Work’ Demonstration Raises Serious Concerns,” Center on Budget and Policy Priorities, January 22, 2010.
] The Office of the Inspector General conducted an audit of HUD’s oversight of the Philadelphia MTW program in 2004, a general audit of MTW design and implementation in 2005, an audit of HUD’s decision to admit the Housing Authority of Baltimore City to MTW in 2006, and a number of audits of individual agencies’ MTW programs. The reports are available at http://www.hud.gov/offices/oig.
 These estimates are based on HUD data reporting the amount of voucher renewal funds provided to MTW agencies and the amount the agencies spent on voucher subsidies. Some MTW agencies receive voucher administrative funding (which for non-MTW agencies is provided through a separate budget account) and subsidy renewal funds together in a single funding stream. In these cases, we estimated the amount that was intended as administrative funding and deducted it from the agency’s funding level before calculating the amount of funds left unspent.
 The economic recovery legislation enacted in February 2009 provided a major infusion — $4 billion — of supplemental public housing capital funds. In addition, Congress approved a public housing operating fund level for fiscal year 2010 that HUD estimates is adequate to fully fund the operating subsidies for which agencies are eligible; this has not happened since 2002.
 MTW requires housing agencies to serve “substantially the same” number of families as they did at the start of the demonstration. But in at least some cases, HUD has interpreted this requirement to allow declines as great as 10 percent, and it is not clear that HUD has enforced even that loose requirement.
 New HIP agencies would be permitted to blend voucher and public housing funds only after they have achieved utilization of 95 percent of their authorized vouchers or spent 95 percent of their voucher funds. An agency would only have to meet this standard in one year, however, and would then be permitted to divert funds for the remainder of the demonstration.