«reNtal HousiNg Policy iN tHe uNited states Volume 13, Number 2 • 2011 U.S. Department of Housing and Urban Development | Office of Policy ...»
Renter Income Trends at the National Level Before exploring rent burdens and other measures of rental housing affordability, it is useful to examine how the economic profile of renters in the United States has changed since 1990. The median renter income tended to track the performance of the broader economy. Real renter incomes declined in the first half of the 1990s, but increased as the economy picked up steam in the second half of the 1990s; the recession of the early 2000s drove renter incomes down, although some recovery existed in the mid-2000s, but real renter incomes ended the 2000s nominally below 1990 national levels (exhibit 1).
A more interesting comparison is renter incomes to all household incomes. In the 1990s the median renter earned nearly 70 percent of the median household income, and in the ensuing two decades, the median renter income fell to 62 percent of the median household income (exhibit 2).
From 1990 through the present, renters have become poorer on a relative basis, and as the homeownership rate climbed, the higher income renter households became first-time homeowners.
The flow of higher income renters into homeowners is likely one of the main contributors to the apparent increased stresses in rental housing affordability experienced by median-income renter households. Increased income inequality during this two-decade stretch also contributes to this phenomenon (exhibit 3).
Renter Income Trends at the Metropolitan Level Although national statistics may reveal interesting changes in the aggregate, housing, particularly rental housing, is inherently local. America’s metropolitan areas are incredibly heterogeneous and have undergone profoundly distinct economic and demographic shifts during the past two decades. This article uses a time series of median renter incomes by metropolitan area derived from the 1990 and 2000 Census 5-percent Public Use Microdata Sample (PUMS), American Community Survey (ACS) PUMS 2005, and 2009 microdata from the Minnesota Population Center’s Integrated Public Use Microdata Series (IPUMS). These values were converted to 2009 dollars by deflating them using the local Consumer Price Index (CPI). Accompanying the ACS-based time series are estimates of the median renter income between 1990 and 2009 from the Current Population Survey (CPS) at the census region level.
Exhibit 4 presents the real renter income by census region from the CPS. Only the South experienced an increase in real renter incomes. The other regions ended 2009 at or slightly below 1990 levels. Exhibits 5 through 8 display inflation adjusted median renter incomes between 1990 and 2009 from the metropolitan area estimates derived from the ACS PUMS. The economic boom in the second half of the 1990s lifted renter incomes in 17 of 24 metropolitan areas.2 The 2000s were far less friendly to renters. Stalled wage growth for lower income Americans and the movement of higher income renters into the ownership space, led to falling renter incomes. By 2007 real median renter incomes had fallen below 1990 levels in 22 of 24 metropolitan areas. Only the median renters in San Francisco and San Diego had higher real income than their 1990 counterparts.
1.10 1.00 0.90
Washington, DC, and Phoenix, Arizona, needed to be excluded because the Bureau of Labor Statistics did not publish local CPIs for these two metropolitan areas going back to 1990.
1.10 1.00 0.90
1.00 0.95 0.90
1.05 1.00 0.95 0.90 0.85
* Washington, DC needed to be excluded prior to 2000 because the Bureau of Labor Statistics did not publish local Consumer Price Indexes for this metropolitan area going back to 1990.
Renter Income Trends Since 2007 Beginning officially in the fourth quarter of 2007, the United States entered into what by many measures was the deepest recession since the Great Depression. Although the primary causes of the economic downturn are heavily debated, it is clear that the foreclosure crisis and tumult in the housing market were key contributors to the broader economic collapse. Much attention has been paid to the owner-occupied housing market throughout the crisis, but very little research has examined the effect of the economic recession on the rental housing market.
The most obvious effect of the recession on the rental housing market is lower renter incomes. The inflation adjusted median renter income fell by almost $1,000 from 2007 to 2009 in the ACS. The number of extremely low-income and very low-income renter households—those with incomes 0 to 30 percent and 30 to 50 percent of their local Area Median Income (AMI)—increased from
15.9 to 17.1 million households according to the AHS. These income reductions were not shared equally across U.S. metropolitan areas. The real median renter income fell in 16 of 26 metropolitan areas. Among these areas, Detroit, Michigan; Cleveland, Ohio; Minneapolis, Minnesota; and San Diego, California experienced the largest declines. Houston renters fared the best in the 2-year stretch from 2007 to 2009 with a 6-percent increase in median renter incomes.
Rent Trends The following section explores trends in rents from 1990 through 2009 across a number of large metropolitan areas. These trends help explain the patterns of affordability levels described in subsequent sections.
Rent Trends at the Metropolitan Level Renter incomes have largely remained flat or declined in real terms during the past two decades;
understanding the change in real rents will provide a richer understanding of rental housing affordability trends. One significant limitation to any analysis of rent trends is the lack of frequent data on rents representative of the entire rental housing market. Several private housing market research firms provide rich, frequent data for a rental housing market segment. These data typically sample from exclusively larger, professionally managed properties, however, which represent no more than about one-third of the rental housing market.3 Because a large rental housing stock segment is located in small buildings with mom and pop ownership structures that might differ in their rent-setting methods than larger professionally managed properties, these proprietary sources are not completely representative. For this reason, this analysis uses the local CPI residential rent indices, which are derived from the CPI Housing Survey and capture rental housing statistics across all structure types— mobile homes to multifamily projects. The CPI Housing Survey features six panels that are sampled biannually on a continuous cycle, so that the rent index can be refreshed monthly.
Rental properties with 50 or more units make up 31 percent of the rental stock, according to the 2001 Residential Finance Survey.
The CPI publishes the residential rent index for 27 metropolitan areas and 4 census regions. Exhibit 9 shows the change in real rents by region. The Northeast and the West experienced the largest increases of nearly 15 percent from 1990 through 2009. Exhibits 10 through 13 show the movement of real rents—the rent index deflated by the less shelter index—in 24 of these metropolitan areas by region from 1990 through 2009.4 The real rent index is normalized, so that base year 1990 has a value of zero (0), and changes can be interpreted as percentage changes relative to all other goods. With a few exceptions, the first half of 1990s were marked by downward movement in real rents, although the 10 years from 1996 to 2006 were characterized by upward movement in rents. Very apparent in these exhibits is the incredible heterogeneity in trends across different metropolitan areas. Real rents in the metropolitan areas of Chicago, Illinois; New York, New York; San Francisco, California;
Los Angeles, California; San Diego, California; Miami, Florida; and Washington, District of Columbia increased by more than 15 percent during the time period. Whereas real rents fell in the metropolitan areas of St. Louis, Missouri; Cincinnati, Ohio; Phoenix, Arizona; and struggling Detroit, Michigan, which led with a 4-percent decrease in real rents. Although the statistics are inflation adjusted, median renter incomes declined relative to their 1990 level in 22 of 24 metropolitan areas, 18 of 24 metropolitan areas experienced an increase in real rents during this same period. The metropolitan areas with the largest increases in real rents are primarily major immigrant hubs or supply-constrained rental housing markets. Several studies have estimated the effect of immigration on rents, Saiz (2006), Susin (2001), and Greulich, Quigley, and Raphael (2004) all found evidence that immigrant inflows increases rents. Greulich, Quigley, and Raphael (2004) found that natives experience a commensurate increase in incomes, however, it is such that their rent burden is relatively unchanged.
1.15 1.10 1.05 1.00
Anchorage, Alaska, is excluded because it is nearly twice as small as the next smallest metropolitan area in the analysis.
Local CPIs for Phoenix, Arizona, and Washington, DC, were not published in 1990, so they are excluded from this part of the analysis.
* Washington, DC needed to be excluded prior to 1996 because the Bureau of Labor Statistics did not publish local Consumer Price Indexes (CPIs) for this metropolitan area going back to 1990.
Source: CPI (Rent Index/Less Shelter Index), 1990–2009
Glaeser and Gyourko (2008) have demonstrated the important role that housing supply elasticity plays in the volatility of rental housing cost. In housing markets with inelastic housing supply, the housing stock is unable to accommodate growth in demand without significant upward pressure on rents. This is particularly evident in coastal metropolitan areas such as San Francisco, California, and New York, New York, and in both San Diego and Los Angeles, California, in which regulatory constraints and limited developable land hinder rental housing stock growth (through new construction or additional density), and contribute to rising rents (Glaeser and Gyourko, 2008).
Rents Since 2007 In general, trends in the rental housing market tend to track broader trends in the macroeconomy.
Rental housing demand is very fungible due to the significant young household segment that has more housing flexibility than the established family household segment. With significant job losses beginning in 2008 and rising unemployment rates among the young, rental housing demand contracted, which caused vacancy rates to rise. For the most part, this rising vacancy rate was a function of plummeting demand, and not a function of rental housing overbuilding. Painter (2010) finds evidence that new household formation reached recent historic lows during the recession.
Although widespread job losses created downward pressure on rents, different rental housing markets experienced very different stresses. A closer look at rents from 2007 through 2010 is provided in appendix A. In 14 of 26 metropolitan areas real rents fell from the 2007 to through 2010.
In housing markets hard hit by foreclosures (such as Phoenix, Arizona, and Tampa, Florida), rents fell by a significant 6- to 8-percent range. Even with the prevailing softness in the rental housing market, real rents were up modestly through 2010 in 12 of 26 metropolitan areas—with New York, New York; Washington, District of Columbia; and Seattle, Washington experiencing the largest rent increases. Unfortunately, rent levels reported in the CPI Housing Vacancy Survey are not effective rent (rent net of concessions) levels, so discounts provided to tenants are likely not reflected in these data. High vacancy rates and stalled rent growth tend to ease rental housing affordability stresses (if they are not offset by significant income declines), but create challenges for property owners, particularly those operating relatively affordable unsubsidized rental units. Weakening housing market fundamentals, falling home values, and increasing numbers of foreclosures could potentially reduce the supply of affordable rental housing. The next section explores the dynamics of affordable rental housing stock segments.
Affordable Rental Housing Stock Trends This section examines the dynamics of the rental housing stock during the recession in the context of affordability. The central question is; what happened to the affordable rental housing stock segment from 2007 and 2009? The analysis in this section relies on the 2007 and 2009 American Housing Survey (AHS) and the 2007 and 2009 ACS to explore dynamics in the affordable rental housing market segments during a period of downturn. Although the AHS National Sample does not allow for the granularity to conduct metropolitan level analysis consistently, it has the benefit of being longitudinal and nationally representative, so it is possible to track the how the affordability of a given sample of rental housing units changes over time and estimate the magnitude of these changes for the rental housing market at large.
For this article, rental housing affordability is considered a function of gross rents relative to incomes.5 In addition, rental housing affordability is examined in the context of HUD published AMIs. This approach is sensitive to heterogeneity in wages across housing markets, it allows for greater ease in historical comparisons, and it aligns with HUD program rules. Using the 2007 and 2009 AHSs, rental housing units were categorized according to affordability based on the rent level relative to AMI levels that are consistent with HUD’s Housing Affordability Data Systems (HADS).
Rental housing units are categorized at different percentages of AMI by the ratio of gross rent to the monthly AMI, housing units renting for less than 30 percent of the particular percentage of AMI
threshold are considered affordable. For example: