Economic Prosperity and Housing Affordability in the United States: Lessons from the Booming 1990s

Abstract

The United States is facing an acute shortage of reasonably priced housing with over 35% of households paying more than 30% of their income for housing costs in 2015. As the U.S. economy recovers from the Great Recession, will housing become less unaffordable as incomes rise and households could potentially pay a lower share of their income for housing costs? To see if this is likely, I examined the change in housing affordability in the 100 largest metropolitan statistical areas (MSAs) in the United States between 1990 and 2000, a period of exceptional economic prosperity. I used the percentage of housing cost-burdened households (those that pay more than 30% of their gross income on ownership or rental costs) as a measure of the availability of reasonably priced housing. I used discriminant analysis techniques to detect statistically significant differences in the percentage of cost-burdened households in the 100 MSAs based on a variety of factors. I found that despite the phenomenal economic prosperity of the 1990s, about 30% of households were cost-burdened both in 1990 and 2000. High MSA median income was correlated with a greater shortage of reasonably priced housing. Neither economic growth rate nor poverty rate nor population growth rate distinguished high-shortage MSAs from low-shortage ones. Large MSAs and MSAs in the West had greater shortages than other MSAs. Economic prosperity did not alleviate the problem of lack of reasonably priced housing in the past, and is not likely to do so in the near future. Planners and policy-makers need to enact new policies at local, regional, state, and federal levels to effectively address America’s chronic affordable housing shortage.

Keywords:

• Housing affordability
• U.S. metros
• Cost-burdened households

Notes

1. The homeownership rate reached its peak of 69.2% in the first quarter of 2005. Since then it has steadily fallen and was at 65.0% in the first quarter of 2014.

2. For example, Green and Malpezzi (2003) show that a household at the 25th percentile income level would have to spend much more than 30% of its income on a house priced at the 25th percentile even when the median household did not have to pay more than 30% of its income for the median priced home.

3. Households that spend over 30% of their gross income on housing costs are deemed housing-cost burdened by federal housing programs and by most state and local housing programs. This standard is based on the notion that households that spend a higher proportion than this are compromising on other essential expenditure items, such as childcare, food, healthcare, transportation, specialized training and education. The standard was first formulated in the mid-1960s to set rents in public housing. It became more widely used with the creation and expansion of the Section 8 program (first as a certificate program, then as a voucher program, and now as the Housing Choice Voucher program); this benchmark was used to set the rents paid by renters participating in the Section 8 program. The 1969 Brooke Amendment to the 1968 Housing and Urban Development Act set this ratio at 25% of household income, but the ratio was raised to 30% in 1981, and has not been changed since then. Also, this benchmark was commonly used by financial institutions in the United States in loan underwriting criteria for home-mortgage loans, where this ratio is often known as the front-end ratio. Prior to 2002, most financial institutions set a 30% maximum limit for the front-end ratio. From 2002 to 2008, many banks started originating home mortgage loans with higher front-end ratios—as high as 70%, a practice that, in some measure, precipitated the housing market collapse of 2007–2009. The 2010 Dodd–Frank Act (otherwise known as the Wall Street Reform and Consumer Financial Protection Act) accords privileged status to qualified mortgages and qualified residential mortgages and has thus revived the use of the 30% benchmark in the mortgage industry.

4. In 2010, metropolitan areas accounted for 258 million people (or 83.7% of the total U.S. population).

5. Whereas larger metropolitan areas are often classified as consolidated metropolitan statistical areas (CMSA) that consist of one primary metropolitan statistical area (PMSA) and several MSA, in this study CMSA were regarded as MSA.

6. However, among low-income and renter households the percentage of cases that did not report housing expenses was higher.

7. Households that live in properties supported by the Low Income Housing Tax Credit program may pay over 30% of their income toward rent. These households are cost burdened and are included in this analysis.

8. Consider two households A and B, each with a household income of $100,000 in Year 1. Household A buys a house in Year 1 with a fixed-rate, self-amortizing mortgage, paying $20,000 in mortgage payments, property taxes, and utilities over 12 months. Household B starts renting a house in Year 1, paying $20,000 in rent and utilities for it over 12 months. Both households have a housing-cost burden of 20%. By Year 11, assume the income of each household increases at the rate of inflation (say about 2% per year), to $122,000. Household A, the owner household, has annual housing payments of $20,000 so its cost burden now is 16.4%. Assuming that rent rose at the rate of inflation, Household B would be paying $24,400 in housing costs in Year 11, for a housing-cost burden of 20%. But rents usually rise faster than inflation; therefore, Household B’s housing-cost burden in Year 11 would be greater than 20%.

9. In 1990, there were 11 MSA with more than 3 million people—Miami; Houston, Texas; Dallas, Texas; Washington DC; Boston; Detroit, Michigan; Philadelphia, Pennsylvania; San Francisco; Chicago, Illinois; Los Angeles; and New York, New York—and collectively they had 76.4 million people and 7 million cost-burdened households. In 2000, there were 14 mega MSA—Phoenix, Arizona; Seattle; and Atlanta, Georgia, joined the mega category—with 103 million people and 10.5 million cost-burdened households.

10. Provided the property was owner occupied for at least 2 years in a 5-year period prior to the sale date.

11. Changes to the tax code from the mid-1980s onward have raised standard deductions and lowered benefits of itemizing deductions, thereby negating the advantages from these deductions for most households below 120% of area median income (Capone, 1994; U.S. Department of Housing & Urban Development, 1997). Thus, these deductions now benefit the rich more clearly than they did, say, in the 1980s.