New Metro Definitions: Potential Winners and Losers
The U.S. Office of Management and Budget (OMB) has substantially changed its official list of U.S. metropolitan areas, using new criteria for defining metro areas and incorporating 2000 Census data on population and commuting patterns into the definitions for the first time.

Because these metro area definitions play an important role in government housing programs, the changes have the potential to have a significant impact on the multifamily industry. Multifamily developers may find that some of these changes open up new market opportunities for them over the next couple of years in those counties that have benefited from the changes. At the same time, multifamily owners in counties that have been negatively affected by the changes might find their rental allowances even further restricted and the number of eligible households reduced because of changes to income limits.

Both scenarios are certainly possible because metro area definitions play an important role in government housing programs, including Section 8, the Low Income Housing Tax Credit, the HOME program, tax exempt bond financing, and programs operated by the Department of Agriculture’s Rural Housing Service. Virtually all of these programs operate under restrictions (maximum incomes and rents, for example) that vary from place to place. Those restrictions are tied to the “place” for which they’re defined. When the geographic boundaries of the places change, areas in which drastically different incomes exist may be combined together or separated from each other. Such changes could result in increases or decreases to rental allowances as well as changes in renter eligibility guidelines.

We discussed, in general terms, some of the likely impacts of these changes in the June issue of Multifamily Outlook. Now that OMB has published the list of new areas, we can look at potential changes in specific counties.

New Definitions Could Affect FMRs, Income Limits

Once a decade, OMB undertakes a comprehensive review of metropolitan area definitions using data that becomes available from the decennial census. Its latest review resulted in dramatic changes. The commuting threshold for including a particular county in a metro area was lowered and made more consistent. Old terms (such as Primary Metropolitan Statistical Area) disappeared and new ones (Micropolitan Area, Combined Statistical Area) were introduced. We will not go into the technical definitions here, but will simply use the general term “metro area” to refer to whatever definitions are relevant for housing programs.  [Anyone interested in more technical details can find them on OMB’s web site:  http://www.whitehouse.gov/omb/fedreg/metroareas122700.pdf]

OMB claims to define metropolitan areas only for statistical purposes, in order to provide a consistent set of definitions for collecting and publishing data. It considers housing programs to be “nonstatistical,” and therefore deliberately avoids considering  potential impacts on them when establishing the definitions. Agencies can choose to use the data for nonstatistical purposes, as is done with housing programs’ use of this data, but they are directed to exercise great care in doing so. ¹

Mid-2004 is the earliest that new metro definitions could begin having an impact on housing programs.  The initial version of the 2004 Section 8 income limits (which are used in a wide variety of programs beyond Section 8) will not use the new definitions. HUD’s current plan is to issue revised 2004 income limits, based on the new definitions, at the same time final 2005 Fair Market Rents are published. Thus, the first time HUD would use the new metro definitions is when it publishes the notice of proposed 2005 Fair Market Rents (the proposed 2004 FMRs were published in May 2003).

In private conversation, HUD economists have told NAHB that they will seek to err on the side of generosity, and try to avoid reducing income limits and disrupting housing programs. In practice, however, that will be quite complicated to carry out and will require a number of policy decisions stretching beyond the authority of the Office of Economic Affairs at HUD.

Negative Impact Expected for Many Counties

When shifting the definitions of income limits alters the way programs measure housing costs and incomes, those programs can experience a negative impact. There are three ways this could happen:

1. A low-income county can become detached from a high-income metro
2. A high-income county can be added to a low-income metro
3. A county can be moved from a high-income metro into a low-income metro

NAHB has conducted a preliminary analysis of counties that fall into these categories, and examples are shown in Tables 1, 2, and 3. Table 1 shows counties that belonged to metro areas before the 2003 change, but not afterwards.  Because HUD has not been computing annual estimates of median family incomes for those counties, the income comparisons are based on 1999 incomes reported in the 2000 Census. 

Based on 1999 incomes, the counties in Table 1 have median family incomes at least $10,000 lower than the metro area from which they’ve become detached. In the case of the Culpeper County and King George County in Virginia at the top of the table, HUD had already diverged from the old OMB definitions, and was publishing separate, lower Fair Market Rents and income limits for these counties.  So, among the counties in Table 1, Henderson in Texas looks like the one subject to the largest housing program shocks, with median family income more than $17,000 lower than the Dallas metro area from which it has become detached.

When outlying counties are added to metro areas, we would typically expect the outlying counties to have lower incomes, on average, than the metro areas to which they are added.  There are notable counterexamples, however, and several of these are shown in Table 2. The most extreme case is Summit County in Utah, which was just added to the Salt Lake City metro area, where HUD’s estimate of median family income was about $22,000 lower.

Analyzing counties that have switched from one metro area to another is difficult, because it often involves a new metro area for which no close analogue existed prior to the 2003. We were, however, able to identify a few counties that moved from one metro to another where a reduction in median income is clearly a possible outcome. These are shown in Table 3.

Berkeley County in West Virginia is at the top of the table, but it is another county in the DC metro area for which HUD was already publishing separate, lower Fair Market Rents and income limits. The aggregation of the old Bergen-Passaic entity into the New York metro area appears to be one of the most problematic of the changes for housing programs, involving a very large population center, two different states, and a difference in median family incomes of almost $27,000.

How Will HUD Handle This Data?

The intent of the HUD economists who calculate the numbers is to incorporate the new metro area definitions into their calculations in a way that does as little damage as possible to housing programs. But precisely what that would mean in this context is not entirely clear.

When income data became available from the 2000 Census last year, HUD decided to grandfather in the higher income limits by freezing them at the previous year’s levels in places where they otherwise would have fallen. NAHB applauded HUD's decision to do this, and recommended the same approach when dealing with the new metro area definitions.

But a freeze is trickier to implement when geography is shifting and new “places” that didn’t exist before are being created. Consider what happens when a high-income county is added to a low-income metro area. One approach to freezing in that case is to raise the income limit for the entire metro area up to what it had previously been in the high-income county. In fact, HUD currently seems somewhat attracted to that approach—but only in cases where the county-metro income difference is small. It’s  unlikely HUD would be willing to raise income limits for the entire New York area up to what they had been in Bergen and Passaic, for instance.

The other approach to freezing limits would be to carve out the relatively high-income counties from OMB’s new metro areas and continue to publish separate Fair Market Rents and income limits for those counties. (As noted, HUD already is publishing separate numbers for some counties in the Washington DC area and a few other places around the country.) That would involve very large departures from the OMB definitions, and constitute substantial government policy decisions, considerable complicating the procedure of calculating and releasing the numbers.

A related issue is whether revising income limits in the middle of 2004, as HUD is tentatively planning to do, is a good idea.

And the implications of the new metro area definitions for housing programs don’t end with income limits and Fair Market Rents. At a minimum, the published lists of Qualified Census Tracts and Difficult Development Areas, the allocation formulas for Community Development Block Grants, and the definitions of areas eligible for Rural Housing Service programs can be impacted. Those impacts, however, are less immediate. HUD, for example, has pretty much decided not to revise the lists of QCTs and DDAs until 2005, after the policy decisions involving income limits and Fair Market Rents have been resolved.

New Metro Areas Will Benefit Some Counties

So far, we have discussed the potentially negative impacts that call for some type of policy response.  But there also is a potential for positive impacts — cases in which rent and income limits rise substantially due to the new metro area definitions. An increase, in fact, is quite likely in cases where outlying counties are being added to metro areas. Preliminary counts have found only 25 cases where a county has a lower median income than the metro area to which its being added, compared to almost 200 where incomes are higher in the receiving metro area.

Some of the more extreme examples are shown in Table 4. All the counties in that table have been added to metro areas in which median family income is at least $24,000 higher. In the case of San Benito County in California being added to the San Jose Metro Area, the spread is more than $38,000.

It’s less common for a high-income county to become detached from a lower-income metro area, but there also are cases where that has occurred. Examples are shown in Table 5. Los Alamos County in New Mexico is in a class by itself, with a 1999 median family income that is $35,000 higher than the Santa Fe metro area it was just separated from.

Additional Information

NAHB has undertaken a systematic evaluation of the new metro area definitions.  More complete results than those reported here can be obtained by calling 1-800-368-5242 x8449.  The NAHB analysis is based on entering the definitions and comparing the new and old lists essentially by hand, however, so we can not guarantee that the results are completely comprehensive and free of error. Moreover, New England is excluded from the analysis, because metro areas do not follow county boundaries in that part of the country.

Therefore, we recommend that those who build, own, or manage residential property and use (or are considering using) a government housing program check out in detail what’s happening in the relevant parts of the country directly from government sources.

To check out which metro area, if any, a particular county was part of before the latest change in definitions, consult HUD’s current county-level spreadsheets for income limits http://www.huduser.org/datasets/il/fmr03/Section8.xls  or Fair Market Rents http://www.huduser.org/datasets/FMR/FMR2004F/FMR2004P_County.xls.   For the status after the change, check the revised list of metro areas on the Census Bureau’s web site http://www.census.gov/population/estimates/metro-city/03msa.txt. Going through the list of new metro area will be somewhat tedious, because it’s organized by metro area name rather than by state or county (so you’d probably want to open the file in word processing software that will let you do a text search for a county name), but it will be worthwhile for anyone seriously interested in what’s going to happen to government housing programs in a particular part of the country over the next couple of years.

-----------

1. The general advice OMB gives to federal agencies on the subject is, “In cases where there is no statutory requirement and an agency elects to use the Metropolitan, Micropolitan, or Combined Statistical Area definitions in nonstatistical programs, it is the sponsoring agency's responsibility to ensure that the definitions are appropriate for such use. When an agency is publishing for comment a proposed regulation that would use the definitions for a nonstatistical purpose, the agency should seek public comment on the proposed use. An agency using the statistical definitions in a nonstatistical program may modify the definitions, but only for the purposes of that program. In such cases, any modifications should be clearly identified as deviations from the OMB statistical area definitions in order to avoid confusion with OMB’s official definitions.”