Difficulty affording housing is widely acknowledged as the most common housing problem in the United States. No matter how one chooses to measure the problem, it is clearly widespread and growing worse among the lowest income renters. But how one conceives of and measures housing affordability matters to policy making as well as public perceptions of the scope and nature of the problem.
Defining housing affordability problems is complicated and entails subjective judgments. For example, should households that spend a small fraction of their income on housing but that live in a substandard home or in an unsafe neighborhood or at great distances from their jobs be construed as having affordability problems? If so, then which such households ought to be counted? Should households with moderate incomes who spend so much on housing that they have too little leftover to save and invest be viewed as having an affordability problem? Should a low- or moderate-income household that spends a large share of their income on housing to live in an affluent neighborhood be viewed as having an affordability problem or as having just made a choice to spend more on housing? Indeed, distinguishing between who is allocating large shares of income to housing or taking long commutes out of choice and who is doing so out of necessity is a bedeviling task.
Standard measures of affordability do not engage with these issues. Importantly, standard measures fail to take into account tradeoffs that people make to lower housing costs. These tradeoffs include housing quality, neighborhood quality, and location. Making these tradeoffs can impose other costs on households. These added costs are not now captured by the simple approach of measuring only the share of income households spend on their housing. Counting a portion of those who incur such costs would add to counts of the number of households with housing affordability problems. For example, households in the bottom expenditure quartile that spend 30 percent or less on housing spend on average $100 more on transportation than those that allocate over half their outlays to housing. Should this $100 tradeoff get added back to housing costs when estimating who is spending more than a certain amount on housing? Should the time value of longer commutes get added in as well? Creating measures that capture such tradeoffs is possible but will require considerable research and debate over appropriate methods.
This paper explores the challenges of conceptualizing and measuring rental affordability for the purposes of formulating public policy. The strengths and weaknesses of the standard definitions of affordability are examined, suggestions for improving them are made, and reasons for differences in estimates using apparently the same definitions of affordability are explained. Given differences in estimates and criticisms of how incomes are measured and defined in developing rental affordability measures, stylized conclusions about rental affordability problems are presented that are robust to differences in datasets used, decisions made about how to treat special cases, and decisions made upon how to define income and rents.