In the weeks ahead we’ll be delving more into research about how the City is establishing what it calls “housing need.” The foundation of almost all the discussion over the last several years has been the assumption that housing is affordable when it consumes 30 percent or less of a household’s income, discounted for some percentage of Area Median Income (AMI). Therefore, housing need is expressed as a ratio of housing cost to income: “My household earns $88,000 annually, which means we can afford $2,200 in rent per month.” Anything more than $2,200 per month means I am cost burdened and any less means I am potentially “downrenting” or “overhoused,” terms that express that I am essentially underpaying for my housing.

There is something satisfyingly simple about the way the City and other government entities like the Washington State Commerce Department define what housing need is at any given time. Just take census data for household income then subtract the number of units priced at 30 percent of those household incomes. Voila! That’s how many units we need. But Michael Stone a professor from the University of Massachusetts Boston has challenged this way of defining affordability.  In his paper, “What Is Housing Affordability? The Case for the Residual Income Approach,” Stone points out,

Affordability is not a characteristic of housing—it is a relationship between housing and people. For some people, all housing is affordable, no matter how expensive it is; for others, no housing is affordable unless it is free. “Affordable” housing can have meaning (and utility) only if three essential questions are answered:

1. Affordable to whom?
2. On what standard of affordability?
3. For how long?

Intuitively we know what this means. In the past, when I have considered household income and what I can “afford” I realized that I didn’t want to pay that much. At other times, if I heard what I could “afford” I would have been stunned, realizing that the standard of 30 percent of my income, even when discounted based on my relationship to median income, was still too much.

Quantitatively, the problems with the Housing Cost Income Ratio (HCIR) are myriad. First, nobody in the housing market, even poor people, make decisions about housing based on the standard used by government to assess their need. So almost immediately, the housing market is out of sync with the normative standard being used to measure the problem. Normative means “you should be paying X for housing.” Some of us are paying too much, for sure, but that also means some of us aren’t paying enough.

But for people struggling to make ends meet, this takes on an outsize importance. Government is determining the “problem” based on income to rents at an arbitrary level of 30 percent of gross (before taxes and other expenses) income. Stone goes on:

Because ratios are pure numbers, they can be compared across time and space and thus are susceptible to being reified as universal and lawful. Such “laws” then become legitimated as appropriate indicators and as the basis for normative standards . . . Once the ratio measure is accepted as the appropriate indicator, ipso facto, the standard must be a ratio or a set of ratios. Yet the notion that a household can adequately meet its nonshelter needs if it has at least a certain percentage of income left after paying for housing implies either that (1) the lower the income of a household, the lower the amount it requires for nonshelter needs, with no minimum whatsoever, or (2) that the normative ratio must diminish with income, all the way to zero below certain incomes. Further, since an affordability standard is intended to measure whether housing costs make an undue claim on household income in relation to other needs, basing such a standard on what people actually pay provides no way of assessing whether they are in fact able to achieve some minimum standard for non-shelter necessities.

Just because I pay 30 percent or less doesn’t mean that my housing costs aren’t consuming so much of my income that I can’t meet other expenses; and conversely, if I am paying more than 30 percent, I may still be able to cover other expenses because they are fewer or non-existant. Take the case of a single mother with two kids who is paying 30 percent of her income compared to a single person who is paying 40 percent. The parent probably has many more household expenses than the single person. While the current statistical regime tries to account for this by adjusting the standard for household size, it probably doesn’t much help the practical problem: the single-parent has many more expenses that draw down limited earnings.

Finally, the methodology of calculating need is deeply suspect. Taking one set of census data and subtracting it from another data set assumes that people should sort themselves into housing according to the HCIR. In a way, this implies that people who aren’t paying enough rent should pay more, while those who pay more than 30 percent should pay less. It’s as if local, state, and federal government could declare housing problems over when we achieve a perfect state of equilibrium, with everyone paying exactly 30 percent of their income for housing.

In the end, the problem with all of this is pressure to tax the building of new housing to build new units priced at 30 percent of monthly income with the appropriate discounts for AMI; i.e. everyone who earns $44,750 in Seattle should pay no more than $1,198 per month. Nevermind that many households could pay more and some might not even be able to afford $900 per month. Clearly if Seattle is going to solve the housing problem it needs to figure out what that problem is first, and the current standard hasn’t done a very good job of doing that. Why would we continue using it as the basis for determining need and imposing taxes to eliminate that need? It’s time for new thinking and methods.