Because income limits are used to place restrictions on the rents that LIHTC property owners can charge, they can be particularly problematic in times of rising operating costs. Because of this, HUD has adopted a series of “hold harmless” policies. Hold harmless policies freeze income limits at their current levels rather than let them decline from one year to the next. NAHB has generally supported hold harmless policies, recommending various versions of them in comment letters and personal communication with HUD over the years.

However, the recent history of “very low” income limits—the limits used in the LIHTC program—combined with rising utility allowances are creating dire situations for affordable housing projects and their owners in a significant number of areas. In fact, the situation is severe enough to cause the stakeholders to seek changes in the way income limits and rents are set in the LIHTC program.  As of this writing, NAHB is scheduled to meet with Congressional staff and the Treasury Department over the next few weeks to discuss the issue.

Background of an Emerging Crisis

Very low income limits are based on 50% of area median income. However, it’s not true that all—or even most—of these limits are precisely half of median family income as estimated by HUD. In addition to hold harmless policies which hold very low income limits above 50 percent of the median, there is a floor determined by median family income in the non-metropolitan portion of each state, and an adjustment either up or down for areas with high or low housing costs (as measured by Fair Market Rents). The charts below show how few metropolitan areas (Figure 1) and nonmetropolitan areas (Figure 2) actually had their very low income limits set at precisely 50% of median family income in 2007.

An Unusual Situation in 2007

In a typical year, HUD’s hold harmless policy should freeze income limits and tax credit rents in a handful of places. In 2002, for example, the hold harmless policy froze income limits rather than let them decline in three metropolitan areas and four nonmetropolitan counties (Figure 3). 

But we haven’t seen a “typical” year like this for some time. In 2003, HUD benchmarked its median family income estimates to the newly available 2000 Census data, making up for ten years of accumulated error in one shot and holding income limits constant in several hundred cases where downward adjustments  would have otherwise caused then to decline. Some of the downward adjustments were so large that the hold harmless policy remained in effect, and tax credit rents remained frozen in many areas throughout 2004 and 2005. Then, in 2006, HUD adopted a totally new set of metropolitan area definitions worked out by the Office of Management and Budget. This resulted in massive changes and another surge of areas held harmless (as when, for instance, outlying counties with relatively low incomes were tacked onto previously existing metro areas), although nothing like the surge seen in 2007.

What made 2007 even more extreme? Why did so many counties fall under the hold harmless policy, even compared to years like 2003 and 2006 when major shocks hit the system of median income estimates?

The answer has to do with the use of a new data source to estimate median family income. The new data come from the American Community Survey (ACS), which is annual survey designed to replace the decennial Census long form. Collecting data annually rather than once every ten years seems like a good idea in general, but, in a way no one anticipated, the ACS wound up producing median family income numbers that were substantially lower than those implied by 2000 Census with any reasonable inflation adjustment. We won’t go into possible explanations here (for those interested, the Census Bureau has produced a paper on the subject http://www.census.gov/acs/www/Downloads/ACS/ASA_nelson.pdf), but we’ll simply note the impact these lower incomes have on income limits and rents in tax credit properties.

Some Rents Have Been Frozen for Five Years

HUD’s hold harmless policies work very well when smoothing over a temporary, anomalous downturn in income data. However, the spike in the number of areas with frozen income limits in 2007 follows several years during which the limits had already been frozen in many places. Thus, tax credit properties have seen little or no increase in for the past five years in some areas. As costs keep increasing, it threatens the financial viability of these properties.

One case where this has happened is Washoe County, which is part of the Reno-Sparks metropolitan area in Nevada. In the LIHTC program, income limits are used to establish restrictions on gross rent, which includes payments for utilities. In the typical case where residents pay for utilities, a “utility allowance” set by local the Public Housing Authority (PHA) is subtracted from the rent the property owner receives. There is no central data base where these allowances are available to the public, but a developer and owner of LIHTC properties in the Reno-Sparks area  provided the recent history of utility allowances for a particular project. This enabled us to analyze both gross and net (i.e., minus the utility allowance) rent for a specific two-bedroom tax credit apartment with a rent restriction based on HUD’s very-low income limit in Washoe County.

In 2002, the limit on gross rent for this apartment was $720 and the utility allowance was $61, so net rent going to the property owner was $639 per month.  Over the next five years, income limits increased by very small amounts in two of the years and were completely frozen in the other three.  Cumulatively, between 2002 and 2007 the following happened (see also Figure 4):

• HUD’s estimate of median family income increased by 2.7% ($1,700)
• The very-low income limit increases by 2.7% ($850)
• The gross rent limit on the tax credit increased by 2.7% ($20)
• The utility allowance increased by 49.2% ($30)
• Net rent declined by 1.6% ($10)

While this was going on, data from the Bureau of Labor Statistics indicate that the average wage in Washoe County increased by 17.4% and Operating Cost Adjustment Factors published by HUD indicate that operating costs for rental properties in Nevada generally increased by 19.6%.
 
Washoe County is far from the only example. In Racine County, Wisconsin, problems caused by flat rents were exacerbated by utility allowance spikes in both 2006 and 2007. In that county, the gross rent on a similar two-bedroom tax credit apartment with a rent restriction based on the very-low income limit was $731 in 2002, and the utility allowance was $75, so that net rent worked out to $656 in 2002. (Again, the utility allowance information was provided by the property owner and developer.) By 2007, gross rent increased by only $10 (being completely frozen in four of the five years) while utility allowances increased by $65, so that net rent fell to only $601—a $55 drop (Figure 5).

The LIHTC properties in Washoe and Racine counties got some (albeit a very small) increase in gross rents in at least one of the years between 2002 and 2007. In nearly 200 other counties, tax credit rents were totally frozen over the entire five-year period. This includes counties in major metropolitan areas such as Atlanta, Chicago, Dallas, Detroit, and Seattle.   

Problems in the Future

The flat income limit problems in the LIHTC program are not likely to end in 2007.  Because the reduction in median family income estimates was so large in some cases, it will take years for the income estimates to “catch up” to the current income limits through ordinary inflation. During that stretch, these places will see no increases at all in income limits or gross rents in tax credit properties, unless the current system changes.

To measure the severity of the problem, we calculate a “critical gap”—how far HUD's median family income estimate in an area must increase before the income limit also starts to increase. For counties that received a high housing-cost adjustment in 2007, the critical gap is set to zero (under the assumption that the chances of costs increasing and driving up income limits further is great enough to make any gap irrelevant). For similar reasons, the critical gap is reduced for counties in the New York and San Francisco metro areas. Although housing costs in those two metros are not high enough to raise income limits above the hold-harmless threshold at present, the costs are nevertheless so high that they are likely to drive up income limits before the critical gap is fully closed.

There are 1,748 counties for which the critical gap is at least $1,000 (Figure 6).  Under the present system for calculating income limits, most of these counties will see little or no increases in gross rents in LIHTC properties in 2008.

In 759 counties, the critical gap is at least $2,000; in 175 it’s at least $4,000; and in 48 it’s at least $6,000. In the most extreme cases, some counties are looking at no increases in income limits or tax credit rents for a decade. 

This has strong implications for the LIHTC program in the affected areas.  Lenders who provide financing for LIHTC projects need to take the level of HUD’s median income estimate into account and, in situations where the critical gap is substantial, reduce or eliminate projected revenue growth when underwriting projects.  This, of course, will make projects more difficult to underwrite. And for projects that already tend to be underwritten on thin margins, such as those intended for extremely low income families, the disabled, or other special needs groups, underwriting issues may be quite serious.

The Housing Finance Agencies (HFAs) who administer the LIHTC program in each state will have to address this issue through some combination of ensuring that underwriting is realistic, and changing the way they allocate tax credits. In particular, they will need to steer credits away from projects underwritten on thin margins in areas where median income estimates indicate these projects are unlikely to experience revenue growth in the near term. 

For existing LIHTC projects that are under financial distress due to a prolonged period of stagnant income limits and rents, the property owners, syndicators, or HFAs will need to find additional subsidies for the projects, or risk losing them from the affordable housing stock. Possible sources of additional subsidies are finite. [ return to top ]