Rent affordability in the affordable housing sector: literature review

3. Measurements of rent affordability

This chapter outlines existing measures of rent affordability suggested by several reports and academic papers in the UK and internationally^[9].

3.1 Rent to income ratio

The oldest and most commonly used measure of rent affordability internationally, because of its simple formula, is based on the ratio of house prices (in terms of rents) relative to income/earnings (affordability ratio = rent / income), which measures the proportion of a household’s income that is spent on rent (Fenton et al. 2011; Young et al. 2017; Wilson and Barton 2018). Fenton et al. (2011) argued that the problem of affordability lies in the combination of a high rent-to-income ratio and a low income.

Traditionally, ‘one week’s pay for one month’s rent’ (a U.S. expression from the 19^th century) was used to define rent affordability as roughly the 25% of the household earnings (Meen 2018, p. 7). The affordability ratio can vary between 25-50% of household income (Bramley 2012, p. 134). In Canada and the U.S., a ratio of 30% is considered acceptable, while in Australia it is 25% (Fenton et al. 2011). In the UK, there is no official benchmark for this ratio, although, according to the National Housing Federation, housing is affordable if the ratio is up to 25%, and according to CIH, if the ratio ranges between 20-30% (CIH 2013). Based on research, a rent can be considered affordable when housing costs do not consume more than 30-40% of households’ incomes (Stephens et al. 2015).

Among the main advantages of this measure are the data availability, which allows for comparisons across time and countries, and the limited need for pre-defined assumptions. It is easy to use for predictions and leads to rather simple and easy-to-understand stories about rent affordability (Meen 2018). However, according to some researchers, this measure is too simple, arbitrary and can be misleading (amongst others Chaplin and Freeman 1999; Stephens et al. 2015; Meen 2018). The ratio provides us with an aggregated overview of housing affordability, but gives no insights on particular groups of tenants or on the quality of the accommodation (Fenton et al. 2011; Meen 2018). Affordability is more complex and ought to take into consideration the needs of different households, as well as the type and location of households (Chaplin and Freeman 1999; Young et al. 2017; Meen 2018). For instance, “two households, one with a rent of £20/week and an income of £100/week, and the other with a rent of £200/week and an income of £1000/week would share the same 20 per cent ratio” (Chaplin and Freeman 1999, p. 1950). Another ambiguity of the ratio measure is whether Housing Benefit^[10]/Universal Credit payments are included in household income or not. Although there are plenty of criticisms of the traditional ratio affordability measure, Bramley (2012), with the use of the British Household Panel Survey (BHPS) in England during 1997-2003, argued that it is the best objective measure, followed by the residual income ratio (discussed later in this chapter).

Together with the rent-to-income ratio, two other descriptive statistics can be used to measure rent affordability: the headcount and the mean (Chaplin and Freeman 1999). The headcount is based on the number of households with a rent/income ratio that exceeds a predefined benchmark. This statistic provides insights on the number of poor households, but not on the depth of their poverty (Chaplin and Freeman 1999). Some European countries, such as Germany, the Netherlands and Austria, use the mean of the affordability ratio calculated by household types to determine the level of housing allowance (Chaplin and Freeman 1999).

As an alternative to the ratio, headcount and mean ratio statistics, Chaplin and Freeman (1999) suggested the use of the Foster, Greer and Thorbecke (FGT) statistic^[11]. This statistic takes into account the depth of poverty, the income gap, the number of poor households, the household income, the poverty line below which a household is considered poor and the total headcounts. In this way, it calculates the distance between each household’s rent-to-income ratio and the poverty line (Chaplin and Freeman 1999).

3.2 Housing expenditure to income ratio

An alternative to the rent-to-income ratio is the housing expenditure-to-income ratio. This measure better reflects the affordability changes across time because it includes the changes in interest rates (Meen 2018). The measure is based on the budget constraint formula:

Consumers’ expenditure (excl. housing) + housing costs + saving from current income = post-tax household earned income + post-tax income from net financial assets

Table 3.1 shows the median ratio of housing costs to (unequivalised^[12]) income by housing tenure and time. Owners (outright or with mortgage) presented the lowest ratios, while ratios for social housing were just under 25%, with very slight differences based on the type of provider. High ratios were also registered for private renters.

Table 3.1 – Median ratio of housing costs to net unequivalised income, by tenure and year, Scotland

	2006/07 to 2008/09		2009/10 to 2011/12		2012/13 to 2014/15		2015/16 to 2017/18
	Ratio	Base	Ratio	Base	Ratio	Base	Ratio	Base
Local Authority	22%	1994	22%	1613	22%	1318	24%	1152
Housing Associations	25%	1310	24%	1129	25%	831	25%	792
Social sector	23%	3304	23%	2742	24%	2149	24%	1944
Private rented	25%	1090	28%	1333	25%	1211	27%	1035
Owned with mortgage	13%	4367	10%	3555	9%	2537	8%	2088
Owned outright	3%	3710	3%	3746	3%	8856	3%	3076

Source: Reproduced by CAD 2019, p. 92; Family Resources Survey

For the base figures see the “Supporting files” https://www.gov.scot/publications/social-tenants-scotland-2017/.

The above measure reflects better the differences across household types. However, it might result in a more optimistic image than reality, especially regarding the private market, because it does not take into account credit market constraints, such as loan and deposit restrictions, as well as the quality and location of the property.

Low-income households might lower their housing standards in order to increase their non-housing consumption or the opposite, with possible dangerous effects on social outcomes, such as health (Bramley 2012; Meen 2018). However, we need to keep in mind that often social tenants do not have a choice on the quality, size and location of the property offered to them by social housing providers, and thus do not always have control on some of their housing costs.

3.3 Residual income measure

Another approach on housing affordability identifies “housing as being unaffordable when the cost of housing of an adequate size and standard reduces income to a level whereby essential non-housing consumption cannot be met” (Stephens et al. 2015, p. 9). The residual income is another measure of rent affordability, which is based on the above formula but also includes non-housing costs and shows the difference between income (net or gross of Housing Benefit/Universal Credit) and housing costs. This approach is based on housing and non-housing consumption standards (Bramley 2012). “This residual income approach subtracts from disposable income the monetary value of a pre-defined standard of non-housing needs; this, therefore, determines how much is affordable for housing” (Meen 2018, p. 15). For example, the Joseph Rowntree Foundation (JRF) claimed that a household of a single working age person needs £10,192 per year after housing costs^[13] (Green et al. 2016). In other words, if a household over-spends on housing, then its non-housing consumption will be inadequate, and vice versa. However, defining a standard for non-housing needs is not straightforward as it depends on the household type and size.

Lydia Marshall from NatCen presented a new residual income approach based on the Minimum Income Standard (MIS) at the 2016 European Network for Housing Research conference. She defined a home as unaffordable when the housing costs are high and the residual income is not sufficient to reach the MIS. JRF’s MIS was introduced in the UK in 2008 and is used to calculate how much income households need in order to afford an acceptable standard of living. The MIS is regularly updated by JRF to reflect changes in the cost of living and social norms. It is crucial that it is often updated based on the required goods and services, as well as on price changes and inflation^[14]. It has often been used by stakeholders and policy makers with the aim of defining a Living Wage. The main UK MIS has been calculated based on people’s needs in urban areas (outside London), but the research has also been extended in rural England, London, remote rural Scotland and Guernsey (Davis et al. 2018). Hirsch et al. (2013) studied the MIS in remote rural Scotland and concluded that households in this part of Scotland required significantly higher incomes to achieve the level of MIS as those living elsewhere in the UK. The reasons for this lie mostly in the higher prices of buying the same goods as elsewhere, the extra cost of heating (often no gas heating is available) and the extra costs of transportation.

According to Fenton et al. (2011), there are two indicators that define the affordable residual income standard: the poverty line standard and a budget standard. In the UK the poverty line threshold is set at 60% of the median household income, while the budget standard defines the residual income remaining after housing costs are covered (there is no specific threshold for the budget standard) (Fenton et al. 2011). Low-income households often receive Housing Benefit or the housing element of the Universal Credit, which covers either all or part of their rent. As mentioned above, when measuring rent affordability in the affordable housing sector, housing benefits can be incorporated into the calculation as an addition to the household income or as a rent reduction (Meen 2018, p. 8). In recent publications^[15] the Scottish Government has used the housing costs-to-(net) income ratio. Housing Benefit payments are included in the net household income: “net income is the total income received by the households excluding taxes such as income tax and council tax. Net income has not been adjusted (“equivalised”) for family size. Housing costs include rent gross of Housing Benefit, as well as water rates and service charges where applicable” (CAD^[16] 2019, p. 91).

As seen above, the most common ways of measuring affordability are as a proportion of the income spent on housing (ratios) or as the income left after paying for housing (residual income measure). According to SFHA, “Recent research suggests the benefit of combining a traditional, low income-based threshold (25%) with information on financial hardship or looking at the joint incidence of a high rent-to-income ratio and low residual income”^[17].

3.4 Rent affordability measure and household composition

A more comprehensive, yet more complex, measure of rent affordability would include household characteristics, such as household type, size and location. Bramley argued that although there was strong rent affordability variation across different age groups and household types, there was not such strong regional variation (Bramley 2012). His research stressed the importance of demographic characteristics, such as type of household (lone parent, single adult, number of children, etc.), when studying affordability (Bramley 2012, p. 146). An example of different ways to measure affordability of a rent of £82/week for different types of households earning a minimum wage of £232/week is presented below (CIH 2013, p. 8).

Table 3.2 emphasizes the differences of rent affordability based on household composition. The same rent represents 32% of the income of a single adult and 18% of the income of a family including two adults and two children. However, when income is adjusted according to the number of household members (equivalised), to capture the average living costs, the same rent represents 21.5% of the single person’s income and one quarter (25.4%) of the family’s income (CIH 2013, p. 8).

Table 3.2 – How affordability can be measured in different ways

	Income after tax/NI/tax credits (Gross of £232.11) (£week)	HB payable at rent level of above: (£week)	Affordability of rent at £82 Min rent HB payable (Proportion of Income)	Rent at 25% of income (£week)	Equivalisation Factor	Equivalised Income (Based on 2 adult household=1)	Affordability of rent at £82 (Proportion of Income)
Gross Minimum Wage	£232.11	N/A	35%	£58.02	N/A	N/A	N/A
Pat	£254.00	£118.95	32%	£63.50	1	£381.00	21.5%
Pat and Chris	£302.00	£112.00	27%	£75.50	1.5	£302.00	27%
Pat, Chris and 1 child	£385.00	£124.00	21%	£96.25	1.8	£320.00	25.6%
Pat, Chris and 2 children	£451.00	£83.00	18%	£112.75	2.1	£322.00	25.4%

Source: Reproduced by CIH 2013, p. 8

*HB = Housing Benefit; NI = National Insurance

3.5 Rent affordability from the tenants’ perspective

Most of the above measures of rent affordability consider affordability as merely an economic term and measure it from a purely economic perspective. An interesting way of measuring rent affordability from the perspective of the tenants has been used by the Centre for Regional Economic and Social Research (CRESR) at Sheffield Hallam University, which carried out a survey for the Flagship Group housing association in order to study how affordable were the housing options of the association in the East of England (Green et al. 2016). They conducted a survey of Flagship tenants (including social tenants, affordable rent and shared ownership customers) using paper and online questionnaires. The questionnaire included information on key household and property characteristics, rents, income and the economic condition of households. The sample of this survey was robust and representative of tenants’ age, type of property, tenure, income and benefits, and household composition^[18]. They defined affordability as “the ability of a household to pay their rent” (Green et al. 2016, p. 7). They adopted a measure of affordability, which is sensitive to tenants characteristics, namely the tenant’s perception of rent affordability, the tenant’s assessment of their overall financial condition and whether the tenant was responsible for paying the rent or whether she/he received housing benefits paid directly to the landlord (Green et al. 2016, p. 7). Based on this measure, they divided the respondent households into three categories:

• affordable rent included those who thought their rent was affordable and had enough money to cover their living costs or were on full housing benefits paid directly to their landlord (59% of interviewed tenants);
• at risk of unaffordable rent included those who thought their rent was affordable, but were not able to cover their living costs (32% of tenants);
• unaffordable rent (6% of tenants) referred to those who found their rent unaffordable and were not able to pay for their housing costs (Green et al. 2016, p. 8)

They then compared the above measure against the conservative rent-to-income ratio. The ratio over-estimated the percentage of unaffordable rent. According to the ratio, 20% of the tenants paid rent above the threshold of 35% of their total household income and were therefore classified as part of the unaffordable rent category. However, a large proportion of tenants who claimed to have an unaffordable rent, were paying less than 35% of their household income for their rent. This highlights a possible problem when using ratio measures that do not take into account personal characteristics of the tenants, especially those belonging to low-income groups.

Similarly, Bramley studied the subjective self-reported housing payment problem indicators, in combination with the objective ratio measure (Bramley 2012). He used the BHPS (1997-2003) question on housing payment difficulties “Many people these days are finding it difficult to keep up with their housing payments. In the last 12 months would you say you have had any difficulties paying for your accommodation?”, together with questions on borrowing money, being behind with payments and the persistence of these problems across time. In an attempt to identify the most appropriate ratio threshold, he tested different threshold values using two approaches: giving different values to different household types and assigning higher values to high-income households (Bramley 2012, p. 138). Based on the results of his research, he claimed that the traditional ratio – set at a threshold of 25% - can predict better self-reported payment problems compared to the residual income approach. The high individual variance that remained unexplained in the regression model, highlighted that other factors (individual and contextual) played an important role in explaining affordability. Bramley then validated the measure by analysing a set of items derived from studies on poverty^[19] in the UK and used in the UK Family Resources Survey and the European Union Statistics on Income and Living Conditions (EU-SILC): households were asked which of the following things they could not afford: a) keep home adequately warm; b) a week’s annual holiday away from home; c) replace worn furniture; d) buy new, rather than second hand, clothes; e) eat meat, chicken or fish at least every second day; f) have friends or family for a drink or meal at least once a month (Bramley 2012, p. 140). Households with a high rent-to-income ratio, as well as households with self-reported payment problems, were more likely to face one or more of the above hardship. Overall, he concluded that the best affordability measure consists of a combination of ratio and self-reported payment problems.

Finally, Meen (2018) suggested two new measures of affordability that focus on the needs of the groups that are more affected by housing unaffordability; namely the low-income renters and the first-time buyers (out of scope in this study). The first measure is based on the relationship between financial stress and affordability and primarily concerns low-income renters. Among other reasons, affordability has become the centre of attention because of its potential impact on household stress and wellbeing (Meen 2018). Although there is little evidence on how stress levels change with affordability, affordability is highly linked with wellbeing (Meen 2018, p. 18). Meen’s study (2018) assumed that any household with difficulties in paying rent faced some level of household stress. Household stress was explained by the variable “whether the household is spending more than a threshold percentage of income (25%) on housing after the subtraction of housing benefits” (Meen 2018, p. 19). Meen (2018) argued that in England in 2016 the impact of affordability on household stress level sharply declined with income. Households with high net housing costs were more likely to face stress, especially if they were low-income households. However, in low-income households, housing costs are reduced by housing benefits. Meen (2018) then tested the same measure including and excluding housing benefits. The results, as expected, showed that low-income households were almost 20% more likely to be under stress, while top income households were not affected at all, since they usually are not in receipt of benefits (Meen 2018, p. 20).

3.6 Summary

In summary, some key points arose from the literature on the various ways to measure rent affordability. The first is conceptual and concerns the measure’s perspective: some measures are economic-centred (based on incomes and rents) and others are from the tenants’ perspective (based on self-reported financial problems and household levels of stress and wellbeing). A combination of the two approaches leads to more comprehensive results from which we are able to draw a wider picture. Affordability measures strongly depend on household characteristics, namely household composition and income, as well as property location and size.