Measuring Housing Affordability
The HAI Affordable Housing Index
Introduction
Given the recent run-up in house prices in many parts of the United States, the issue of housing affordability has received increasing attention; particularly in places like Boston where house prices nearly doubled during the period from 1998 to 2005. The most common notion of affordable housing implies that households that spend more than 30 percent of their income on housing have an affordability problem. According to a U.S. Department of Housing and Urban Development report to Congress, approximately 12.5% of households allocated more than 50 percent of their income to housing in 2003.
In order to address this issue, the MIT/CRE Housing Affordability Initiative (HAI) has developed an innovative measure of the affordability of towns within a metropolitan area. By focusing on area affordability, we recognize that the price of a house is not only affected by its structural characteristics but also by amenities such as adequate access to jobs, school quality and environmental amenities associated with a house’s location. In this sense, the HAI Index is far more powerful than existing measures of affordability such as the National Association of Realtors’ Housing Affordability Index.
Since the policies of a metropolitan area affect outcomes in the broader economic region, the MIT/CRE HAI seeks to create housing affordability indices throughout New England. In this way, the HAI index will assist in the formulation of effective policy responses to current challenges in regional planning. As a result, the HAI indices provides lenders, planning agencies, advocacy groups, and consumers with a more nuanced and accurate picture of the factors affecting the affordability of housing in their area.
2007 Research Findings: Maps of the HAI Affordability Index
Housing Affordability: A Regional Map (pdf)
Rhode Island Affordability 2006 (pdf)
Greater Boston Area Affordability 2006 (pdf)
Change in Greater Boston Area Affordability 2000-2006 (pdf)
Springfield Metro Area Affordability 2006 (pdf)
How is Affordability Measured in the HAI Index?
In determining whether individual households can afford adequate housing, there are three dimensions of affordability that are considered in the current literature: adequate structural quality, the specific needs of varying types of households (especially differences in size), and commuting costs.
Many of the existing measures of housing affordability are indicators of so-called “rent burdens”—a ratio of monthly housing costs to incomes—computed using a simple ratio of median rents or house prices to median incomes. This approach, while mathematically simple and based on readily available data, obscures the true questions of affordability.
Reports using medians fail to identify an acceptable level of housing quality or town level public amenities (such as school systems). In addition, the median house price in many well-to-do areas is likely to reveal little about the distribution of acceptable housing units that cost less than the median in that town. Furthermore, households eligible for affordable units neither earn the median nor seek the median housing unit.
Some of the median-ratio statistics are also created by using an individual town’s median house price and the same town’s median income. Those who already live in the town can most likely afford to live there. It is more appropriate to consider a metropolitan-wide distribution of households who need to live somewhere. Finally, because the median measures fail to incorporate structural characteristics they provide little useable information to households of different sizes.
The HAI Housing Affordability Index is a powerful alternative to previous measures of affordability. One of its greatest strengths is its ability to capture the Total Cost of Ownership of individuals’ housing choices. In computing the index the obvious cost of rents and mortgage payments are modified by the hidden costs of those choices. The Housing Affordability Index adjusts the nominal Total Cost of Ownership by factoring in the opportunity costs associated with access to employment, the quality of public schools, and environmental conditions. Inexpensive housing built with little or no access to employment imposes high commuting costs on households with less capacity to absorb those costs. The Index’s innovation in affordability measurement is its incorporation of job accessibility as an important aspect of the economic health of the region. Likewise, poor schools and environmentally fraught neighborhoods also impose costs on residents that are not entirely reflected in the price of housing. The Index modifies each community’s rents and housing values to reflect these costs using a sophisticated multivariable regression to generate an adjusted distribution of the housing stock.
In addition to more accurately describing the total cost of housing in a town within its local metropolitan region, the Index allows for the varying needs of different sized households and imposes a minimum standard for structural quality. Different indexes can be computed for specific household incomes and household sizes, taking into account the variety of groups who have a stake in area affordability. The flexibility and rigorous approach of the HAI Index make it a powerful tool for helping to address the current challenges to providing affordable housing.
Index Construction and Methodology
Computation of the Index relies on information for rental and owner-occupied stock from the 2000 Decennial Census and updated stock information for rental and owner-occupied units obtained from the Warren Group and several state agencies. In order to compare rental and owner-occupied units, imputed rents are calculated for owner-occupied housing taking into account the prices of recently sold homes, interest rates, and tax liabilities. The imputed rents are then used to compute the marginal values of three town level amenities (job accessibility, school quality, and open spaces) on housing prices.
While very high quality schooling and lots of open space may be a luxury, an adequate level of school and environmental quality is assumed to be a necessity for households in the region. If towns have worse than average schooling or open space the housing units in that town are assessed an additional cost associated with the inadequacy. The same method is used to adjust for job accessibility, except that housing units in towns with better than average accessibility are given a cost rebate while units in low accessibility places are assessed an additional cost.
From these individual adjustments, a total adjustment due to amenity factors is computed for each town, and used to adjust the rents/prices of the entire stock of rental and owner-occupied units.
The standard used for affordability is based upon 30% of the income of the target household. This standard takes into account the size of the target household as well as its income relative to the area median income for households of the same size. For example, in an index for 2-person households there is no restriction on the number of bedrooms in a unit for that unit to be eligible in the housing stock count, while a 4-person household index only considers units with 2 or more bedrooms. Inadequate units (based on structural quality and, if applicable, bedroom characteristics) are removed from both the rental and owner-occupied stock, and a total count of the remaining units costing less than or equal to the affordable rent standard is taken. The ratio of affordable units to total units in a town represents the town’s final score in the index.
For a detailed account of the methodology, please refer to Fisher, Pollakowski and Zabel (2007).
