Are we getting housing affordability wrong?
The Week in Housing 17/02/23
Any form of housing that involves housing costs (rents or mortgage repayments) that exceed 30% of household disposable income is considered ‘unaffordable’. This ‘30% rule’ has become firmly embedded in discourse around Irish housing policy. It is a simple rule of thumb that has become an unquestioned feature of how we think about housing affordability, how we think about ‘housing stress’ and how we evaluate the overall health of our housing system. Taken to an extreme, the 30% rule can be deployed to argue that any form of new housing supply that exceeds the 30% rule, say for households on the average income, makes housing affordability worse.
The problem with all of this is that the 30% rule has no logical or theoretical basis, is arbitrary, and is widely believed by researchers to be a poor measure of housing stress. It is not at all clear (a) what housing affordability is/means; or (b) that it is the most appropriate way to consider housing’s impact on wellbeing; or (c) to assess the impact of policy interventions. Given all this it is concerning that it has become a largely unquestioned feature of public discourse on housing. In today’s issue I look into housing affordability measures and the critiques thereof that can be found in the academic literature, mainly drawing on the work of Michael Stone, one of the main critics of ‘ratio approaches’ to affordability. Next week I look more into the measurement challenges around affordability.
But first, what is housing affordability? Stone defines housing affordability as “an expression of the social and material experiences of people, constituted as households, in relation to their individual housing situations”. This is, in a broad sense, what affordability tries to capture, i.e. a set of relationships between households and their housing. More specifically, Stone goes on to argue that affordability ‘expresses the challenge each household faces in balancing the cost of its actual or potential housing, on the one hand, and its nonhousing expenditures, on the other…”. This is how affordability is most typically understood and measured, i.e. as the relationship between housing costs and other economic aspects of household life, typically their income or other forms of expenditure.
MacLennan and Williams’ definition of affordability draws our attention to the fact that in practice notions of affordability always imply some implicit or explicit standard of housing and housing costs, against which household’s experiences can be measured or benchmarked. They note that affordability is “concerned with securing some given standard of housing (or different standards) at a price or rent which does not impose, in the eyes of some third party (usually government), an unreasonable burden on household incomes”. Here of course it is worth noting that what might constitute a ‘given standard of housing’ and what might constitute and ‘unreasonable burden’ is far from obvious.
There is quite a strong prima facia case for considering the housing-wellbeing relationship in these terms. First, in a market society wellbeing, including housing, is largely or primarily achieved through forms of expenditure. The cost of housing will thus have a large impact on how accessible housing is and the quality of housing which can be accessed. Moreover, because ‘the rent eats first’, i.e. housing costs are typically one of the first (and largest) claims on household income, the proportion of income eaten up by housing costs will often determine how much is left over for other vital goods and services. Finally, measuring the relationship between housing costs and a household’s broader economic situation is often feasible due to the availability of data and, moreover, in many instances allows comparison across time and across jurisdictions. As Stone notes:
“Such indicators and standards make it possible to arrive at conclusions—potentially contentious to be sure—about the overall extent of affordability problems and needs, as well as their distribution socially and geographically. They also provide an important foundation for the at least somewhat rational formulation, implementation, and evaluation of policies and practices that deal with affordability”.
The dominant approach to housing affordability focuses on the proportion of income spent on housing costs – so called ‘ratio measures’. The most widely used approach defines a household as experiencing unaffordable housing, or being in ‘housing stress’, if they spend more than 30% of their disposable income on housing costs. This is often, for example in ESRI research, nuanced by focusing on the bottom two income quintiles, i.e. the bottom 40% of the income distribution. This is known as the ‘30/40’ rule and measures the proportion of households in the bottom 40% of the income distribution who spend more than 30% of their disposable income on housing costs. The rationale for focusing on the bottom 40% of household by income is that some better off households may choose to allocate more than 30% of their income to housing costs, but this does not reflect housing stress per se, but rather a choice to ‘overconsume’ housing.
The criticisms of the ratio approach are many and varied. They relate to two main issues. First, the arbitrary nature of the 30% figure. Second, the lack of precision in capturing the actual incidence of housing stress because of (a) the variation in household characteristics or circumstances; and (b) the ‘trade offs’ involved in housing consumption choices. Here I focus on the first issue, next week I will turn to the second.
The ‘arbitrary’ nature of the 30% figure is particularly interesting. According to Stone, the 30% figure appears to have been derived from various pieces of research on what households typically spend on housing costs, i.e. deducing that if the average household pays X than X constitutes what affordability is. However, as Stone notes, this is meaningless as some households will be paying more than they can afford while others will be paying less, and it doesn’t really tell us anything about the key thing we are trying to capture, i.e. housing stress.
Crucially, Stone argues that ‘as a normative concept, an affordability standard must have some independent logical or theoretical basis against which households’ actual circumstances can be measured’. If this is not the case, any standard, such as the 30% figure, will be tautological or arbitrary. To be clear, the ratio approach is based on the idea that ‘if a household pays more for housing than a certain percentage of its income, then it will not have enough left for other necessities’ but ‘there is no theoretical or logical foundation for the concept or the particular ratio or ratios that are used’.
This AHURI research found that examining the validity of ratio approaches using Australian data, and found that 45% of households deemed to be in housing stress according to the 30/40 rule, in fact regarded themselves as financially ‘reasonably comfortable’ or ‘very comfortable’:
“This supports the proposition that a significant proportion of low–moderate-income households can sustain housing cost burdens exceeding 30 per cent of their income and still be in a financially sustainable position”.
They can to argue that ‘it would be erroneous to assume that traditional housing stress measures such as the 30:40 rule would be broadly indicative of being in financial stress”.
I’ll return to the question of why this may be the case next week, but for now I want to focus on a crucial point made by Strong:
“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”.
What is interesting here is that while a ratio standard (e.g. 30%) should be based on some independent normative standard of housing, it is in fact arbitrary but, paradoxically, has itself become the basis for normative standards. This is very much the case in Ireland, where the 30% rule has become widely accepted as a rule of thumb in policy and political discussion, as if it referred to some kind of agreed normative standard on acceptable levels of housing costs.
Next week I’ll look in more detail at why the 30% measure, which makes so much intuitive sense, is quite poor at capturing the complexity of housing, housing costs, household needs, and household incomes interact.
Events
Today I’ll be speaking at this Housing Agency event on Tenant Unions. This event this evening is part of the final days of the Housing Unlocked exhibition at the Science Gallery, looking at the PRS and the experiences of renters. Simon are organizing what looks to be a very interesting event looking at approaches to addressing housing vacancy across Europe.
What I’m reading
Not really a reading but Renters’ Voice in Northern Ireland are asking PRS tenants in the North to fill out this survey. Michelle Norris and colleagues have just published this report on preventing homelessness among care leavers. The latest report of the Dublin Regional Homeless Executive is now available. Last but not least, NESC have a new report out on the PRS in Ireland.