Housing affordability - short life working group: final report 2022 to 2024

Analysis: Assembling a Shared Understanding

Up till now this report has set out the range of views expressed about what affordability means and how it might be redefined and understood. From here on we try to build a case for a particular perspective, one that achieves sufficient consensus to move things forward. We have already indicated decisions about the shared understanding in earlier sections. First, we view the shared understanding affordability policy assumption as something that should be progressively realised (between now and 2040). Second, the fundamental heterogeneity and different experiences of different types of households, especially through interaction with the social security system, is likely to generate very different affordability outcomes, and these need to be part of any analysis of how affordable Scottish housing works in practice.

What are the remaining critical questions that still need to be determined to achieve our shared understanding objectives? These are summarised in the box below and examined in the rest of this section.

Questions to Address in Advancing a Specific Shared Understanding

1. Do we look just at renters or also include owner-occupiers?

2. Should we include housing conditions and minimum standards?

3. Where do we situate the analysis in terms of the spectrum of narrow housing to broad home costs? Can we do more to consensually define costs and income measures?

4. To what extent is this definition related to immediate short-term considerations or longer term more chronic housing unaffordability?

5. Do we require clear definitions around the level of social rents, affordable (e.g. mid-market) rents and the basis by which they change over time so that they remain affordable?

6. Do we propose a single or combined measure of housing affordability and should it distinguish between an overall policy assumption or guideline regarding the meaning of housing affordability and a more specific application to be used and defended in housing practice for individuals and specific providers?

7. What due diligence approaches are required to support the introduction of the shared understanding proposed here? What are the concerns over transition to a feasible operationalisation of the shared understanding?

The Shared Understanding, the right to adequate housing and the National Performance Framework

We think that the centrality of the progressive realisation of housing affordability on Scotland’s overarching child poverty goals, and as a cornerstone of the human right to adequate housing, means that the shared understanding of housing affordability should be enshrined in the National Performance Framework within key national outcome indicators, and their performance measured and reported. This means that the commitment to shared affordability goals should be reflected on, and considered in, all policy decisions that directly or indirectly affect housing affordability. This also recognises the centrality and pervasiveness of housing outcomes across a wide range of prioritised social and economic domains and their desired policy outcomes (Marsh, et al, 2025). This will enhance the preventative profile of housing within public policy and make explicit policy trade-offs that must be addressed (e.g. the impact of net zero retrofit funding on rents).

Recommendation 1: As a key plank of fighting child poverty and constructing the human right to adequate housing, the new shared understanding definition ought to be a component part of the National Performance Framework national outcome indicators in line with the Government's proposal for a new National Outcome for Housing.

All tenures or just renting?

Affordability, potentially, is an issue for all households regardless of tenure. However, there is a qualitative difference between owners as purchasers of assets and renters as tenants. A further fundamental difference is that mortgage regulation – the binding deposit constraint on first time buyers alongside affordability stress tests – is a UK function, and not devolved. However, existing owners, having overcome these accessibility barriers, have in recent times faced rising interest rates and vulnerability to large increases in monthly bills.

Our conclusion for the purpose of the present task is to focus on renter affordability only – owner-occupier concerns do need to be considered by the Scottish Government, but the issue must primarily be about accessibility given deposit size, mortgage stress tests and underlying excessively high price to income ratios in most parts of Scotland. Among existing borrowers, affordability and financial hardship clearly remain an issue in a context of rising interest rates and potential income shocks. Nonetheless, we observe the long-term exponential increase in affordability cost to income ratios faced by the bottom quintile of private renters across the UK that sets them apart from the rest of the distribution and which we think, along with child poverty and residual income concerns, makes them the key focus of the shared understanding (see Tomlinson, 2019). We also note that our SLWG members identified that both RSLs and mortgage lenders tend to currently adopt the same cost to income threshold ratio of 30% (see further discussion under recommendation 6).

A second tenure issue relates to private renting where market rents function and, as one tenant in the qualitative research noted, establishing an affordability ratio as a maximum payment may simply mean many properties are out of reach. Clearly, there are other housing policies that are concerned with rent regulation and control. Developing a coherent set of policies for the PRS, including lobbying to reform the local housing allowance, is informed by better evidence and this in turn will be supported by the gap between what people pay in the PRS and what the shared understanding of affordability suggests is the upper limit of affordability. This is an important additional function that such an indicator can provide and a further benefit of a shared understanding.

Recommendation 2: the shared understanding focuses on social, affordable and private rental affordability, but this is on the basis that other aspects of housing policy are supporting access to home ownership, also working and lobbying to support policies to help owners in financial difficulty.

Include conditions or physical standards?

We have seen that there are wide variations in views regarding whether minimum standards and property conditions should be part of the shared understanding of affordability. On balance, while seeing merit in both perspectives, we are persuaded on grounds of both simplicity and the existence of other complementary policy interventions aimed at raising conditions and standards - not to include it in the working definition. This is on the basis that any landlord charging rent ought to be based on meeting minimum legal and regulatory standards.

Recommendation 3: on grounds of simplicity and the existence of other complementary policies, we propose not to include minimum standards regarding housing conditions/standards in the working definition other than to be clear that the rent charged by a landlord should be based on meeting minimum legal and regulatory standards.

Narrow or broad housing costs?

Our primary research suggested very different views among our working group members but evidenced a broader sense of housing costs within the qualitative research findings. We again are struck by the case for simplicity (which is also in part a data issue). We also recognise the merits of the argument that distinguishes between direct housing costs over which landlords have control over and residents have none, relative to those elements of wider home costs that also operate in other policy spheres (e.g. the council tax or indeed energy costs which have seen massive interventions during the recent crisis), and for which there are other mitigations or choices that can be made by the resident. None of this is to in any way discount the impacts that these additional property-related costs have on struggling residents, but we are mindful of the need to draw limits. For these reasons we recommend keeping the housing cost measure narrow and restricting it to rent and service charges. We also note that across the evidence there is for once a degree of consensus around measuring income as net or take-home monthly income.

Recommendation 4: on grounds of simplicity, locus of decision-making and alternative policy domains relevant to certain property related costs, we conclude that a narrow measure of cost, rent plus service charge, should be used.

Immediate or more long-term focus?

There are different views expressed as to whether affordability information and hence arguments for intervention to support meeting housing costs should be aimed at achieving long term stable housing cost affordability or more immediate pressing concerns expressing themselves now e.g. very rapid rent increases while incomes are stagnant. The issue is that on the one hand long term stable affordability relationships may tell us more about trends and sustained need for intervention (and should be routinely monitored). If, however, the view is that housing costs and affordability are more dynamic and that trends are not stable – there is also a second reason to monitor outcomes (and to disaggregate across key different household types).

The shared understanding task arose out of Housing to 2040’s specific requirement. There is a compelling case to view the fact that moving the housing system on to the desired path set out in that strategy will take a long time (the existing stock dominates the new flow of housing even over the long term) – implies the need for a long-term target. This allows for a progressive continuous realisation of the achievement of the target and would indicate a target to be met by the end of the strategy’s life in 2040 but monitored throughout to gauge performance.

A second timing issue concerns the shelf life of the shared understanding definition. Does the evidence suggest that it needs to be revised in the light of the evidence? If there is agreement on the approach, this then becomes a data issue which can be monitored and regularly updated on an agreed basis (we note in particular the impact of annual decisions about uprating, freezing or changing eligibility rules with the social security system, in turn having impacts on net incomes and affordability). A specific requirement to assess the fitness for purpose of the indicators used should be carried out every 5 -10 years.

Third, we recognise that both initial rents and ongoing rent increases must both remain affordable i.e. the shared understanding has a direct interest in the role of regular rent increases and whether they remain affordable.

Recommendation 5: agreeing a framework for the shared understanding should allow regular monitoring and periodic review to consider updating of the indicators used in the light of evidence collected about affordability. In this way it should be possible to adopt the shared understanding to both immediate and longer-term questions re housing policy.

Clarifying what is a social rent

We note that one perspective from working group members was to simply define affordability as social rents. This begs questions about how this could be achieved in practice in the PRS and raises wider questions about the consistency of existing social rents and guaranteeing them in the future and for new developments. It also raises important challenges regarding the level and future path of affordable rents in the case of mid-market rents (and their relation to both the LHA and to social rents). We also must recognise the affordability consequences associated with the private sector rent control regime envisaged in the current housing Bill going through the Scottish Parliament.

Ideally, social rents would be differentiated on a consistent basis though some form of harmonious rent policy across Scottish social housing. However, the rent convergence experience in England in the 2000s suggests this is a risky and complex policy commitment (and there has never been appetite in Scotland for a national rent policy). The social sector housing system is dominated by the stock of existing homes, and it would be more feasible to focus on rent increases within the stock and maintaining their position in the non-market hierarchy, as well as ensuring that new stock, social or mid-market, are initially, and remain, genuinely below market depending on the subsector they inhabit. The affordability outcomes to avoid are that in the future rent increase shift rents out of a recognisably social level if in the social sector, or that mid-market rents grow beyond what can reasonably be defined as affordable. A further important concern is that the shared understanding definition can play a role in informing how future higher rent, non-grant-aided new build is considered, especially where such provision is proposed to be the subject of Scottish Secure Tenancies and, therefore, is claimed to be social housing.

Recommendation 6: it is essential that we have a shared understanding of what social rent levels are and the extent to which future rent increases in social housing can rise to over time but still meet the accepted sense that rents remain at social levels. The Scottish Government should undertake further work to better understand what that social level is across the existing housing stock and how it compares for instance to assumptions within published guidance regarding RSL social rent benchmarks. Affordability scrutiny is also required for the mid-market rent sector, specifically, regarding future MMR rent increases.

Single or combined measures? Towards Measures of a Shared Understanding

There is we believe a case that can be made for a combined approach but one that is also simple to understand and deploy. This takes its lead directly from the right to adequate housing and its sense of what constitutes housing affordability, but also incorporates the cross-government focus on reducing child poverty.

In doing this we need to make an important distinction between the overall policy measure or guideline that the Scottish Government would apply as a standing policy assumption, and the application of the proposal to disaggregated groups of households facing very different affordability outcomes. Thus, in the former case, the policy assumption would guide questions such as the affordability central measure in Housing Need and Demand Assessments, in local affordable housing policies, in starting rents in new social and affordable housing supply, to use as evidence to argue for policy reform, and in the monitoring of household survey data to analyse affordability across Scotland. The latter measure would facilitate a more considered analysis of how different groups are impacted by their interaction with benefit systems and the types of housing they occupy.

Below we propose three elements to a shared understanding: a cost: income ratio, a residual income measure and an after-housing cost poverty measure. The issue that raised the most contention was the form of the residual income measure (not that there should be a residual income element).

The essential reasoning behind incorporating a residual income into the shared understanding reflects both the shortcomings associated with the housing costs to income ratio but also the logic of the right to adequate housing. The residual income focuses on post-shelter net income and it is standard practice to compare the actual income of a household, or all renter households in this case, with a benchmark or standard to be achieved. One area of concern raised by housing providers concerning a benchmark such as the after-housing costs (and income equivalised) Joseph Rowntree Foundation Minimum Income Standard (MIS), which was described by some as challenging and aspirational, and not in the real world of providers setting rents. However, the same MIS (90%) is the basis of the legislated fuel poverty definition in Scotland, something supported by housing providers. It is also a direct source of intelligence used in the SFHA rental affordability toolkit. Further exploration of the MIS can be found in Annex 3.

A perhaps plausible alternative can also be found directly in the SFHA rental affordability tool in the form of the Annual Survey of Hours and Earnings 30^th percentile (ASHE30) benchmark. This is a measure of a specific distributional point in the distribution of earnings from the ASHE survey. The 30^th percentile is of such interest because the developers of the tool argued that the equivalent level of household income implied by ASHE30 tended to be around the level where means-tested benefits run out – it is a proxy for low to moderate income households on the margins of benefit reliance. This is also a clear echo of the original late 1980s definition of housing affordability in Scotland of housing cost to net income for low to moderate working age working households i.e. affordability for those not significantly supported by means-tested housing benefits. Bear in mind that through the third proposed shared understanding indicator, we are also tracking after housing cost renter poverty.

More detail on the way ASHE30 is used in the rent affordability tool is provided in guidance published by SFHA, March 2021 for users of the tool. On pg. 11 it says:

“The figures shown in the tool are net , reflecting income tax and national insurance thresholds and bands”.

“Moderate incomes are calculated based on the 30th percentile for each local authority, which is above housing benefit eligibility. The justification for using these moderate incomes is that incomes converge around this point once housing benefit and tax credits are included in lower incomes. Following the Joseph Rowntree Foundation’s Living Rent methodology, the household incomes are weighted to account for additional income for adults and child benefit entitlements”.

“The method assumes that a single person’s earnings have a value of 1.0 while each additional adult or child aged over 14 years ‘earning’ an extra 0.5 score for the household, with younger children scoring 0.3.”

In other words, the utilisation of ASHE30 in the tool has important assumptions inherent to it:

• The earnings of the individual (the output data from ASHE) are reweighted based on a JRF equivalence scale to generate a household income (also incorporating child benefit entitlements).
• Gross income (the survey measure) is netted off by applying relevant tax and national insurance bands and thresholds (presumably also personal allowances).
• The 30^th percentile of the income distribution generated is considered to be a convergence point of moderate income, just above Housing Benefit eligibility.

There are pros and cons with both MIS and ASHE30 measures.

The MIS is a widely accepted, carefully developed standard with which to base a long-term tracker of affordability on the country as a whole and for specific disaggregated groups. It does include broader measures of standards of minimum consumption based on survey evidence. It is also an evolving annually updatable model. It is a UK national model but does have weights for rurality. Focusing on this level of income also helps to counter the point we heard more than once that net rents are much more affordable for tenants in receipt of benefit. MIS plays an analogous role in the fuel poverty definition to the multi-strand shared understanding of affordability we propose here, establishing a precedent: 90% of MIS is used as a residual income measure alongside a ratio measure of 10% of income spent on fuel in a suitable heating regime. A range of 75-90% of MIS might be used as a long term goal with the sense that falling below 75% should represent a warning about low residual income.

On the other hand, ASHE30 is favoured by some, not just on the argument that housing policy needs to focus on the aspiration of residents moving off benefits, it allows the focus to move on to those without such support but still on incomes well below the median. ASHE30 is a potential proxy for the low in work income level for working age households, but it is essential to confirm regularly if this specific percentile point remains the actual tipping point as described when the rent tool was established using ASHE30’s percentile point specifically. Changes to benefits and their interactions with means tested benefits alter over time as does the distribution captured by ASHE. It would be surprising if this key policy percentile did not shift over time.

The ASHE30 indicator, built on the assumptions noted earlier, leads to reflection on the SFHA rent affordability tool itself. This is a web-based model that allows users to plug in their own rents and potential rent increases for different property archetypes and sizes. It is important to recognise that the SFHA make it clear that the tool is only that – an illustrative guide under different scenarios to help providers consider the implications of different rents. Nonetheless, it is important to recognise that the income measures are hypothetical points in different income or earnings distributions (or an application of the living wage) – they do not operate in the real world, as it were.

Following from this debate, what is proposed as a recommendation regarding the shared understanding definition and, specifically regarding this choice of which measure to use for our residual income component?

The clear statement that this is a long-term measure covering the period to the end of the Housing to 2040 span, and therefore about the gradual and progressive realisation of the implied targets of the three-part definition, should lessen immediate worries regarding the rent-setting implications of the residual income choice. This is reinforced by the widespread use and acceptance of the JRF MIS for fuel poverty, and its use as a comparator in the SFHA rent affordability tool.

It is apparent why providers might prefer the ASHE30 definition, but there are three important caveats. First, it clearly needs to be periodically refreshed and reappraised to confirm that it remains correct to say that the 30^th percentile is the general tipping point between benefit dependence and full exposure to rent levels and rent increases. We need robust contemporary evidence to this point, which we don’t currently have. It may now be too high or indeed too low (and not at the Goldilocks point, as it were). Second, this is a measure at a point in the income i.e. earnings distribution – it is a fairly crude proxy conceptually for low to moderate income. This contrasts with the well-developed, conceptually and empirically robust, regularly updated MIS, which as we saw earlier is widely used across social policy analysis in the UK. Third, there is a chain of assumptions that goes from individual gross earnings to net household income that also need to be regularly assessed and revalued.

On this basis, the proposal is that we adopt the MIS as the primary measure of residual income and chart its progress over time both nationally and with respect to the disaggregated groups identified in the main report. However, provided that the appropriate measure of switchover on/off housing benefit eligibility for low to moderate incomes can be identified empirically, and there is a commitment to regularly revisit this and other key assumptions, the ASHE30 (or other point observed in the distribution with this characteristic) measure could be a supplementary indicator of residual income for the national and disaggregated scales, focused more on shorter term performance of affordability – but should be assessed for fitness of purpose on the grounds above before doing so.

Overall, looking at the evidence and at our guiding principles, we consider that a combined approach could work well and embrace three different aspects of affordability that are considered important. The housing cost to income measure remains a useful summary measure for both existing and new rents but its shortcomings can be reduced by incorporating a residual income element.^[14] Here we propose an adaptation of the Joseph Rowntree Foundation income standard (i.e. after housing costs). Third, and even if the cost to income ratio was less, and the residual income so defined was more, than the stated threshold for unaffordability, if the household’s net after-housing cost (AHC) income remained below the standard UK poverty threshold, then the housing, would be deemed unaffordable^[15]. We propose 30% of net income for the cost to income ratio – beyond the range of views in this broad area, one further argument for 30% rather than 25% is the contention that the latter is the more commonly used ratio for home owner access by lenders - should the desire be to widen the definition to all housing. The ratio incorporates service charges^[16].

Recommendation 7a: (1) national policy guideline or assumption – rent and service charge is no more than 30% of net monthly income; second, there is a minimum residual income relating to the after-housing costs version of the Joseph Rowntree Foundation Minimum Income Standard (with a possible supplementary indicator based on ASHE30, as discussed, conditional on it being successfully refreshed in the light of tax and benefit changes since its introduction and use in the SFHA affordability tool); third, if, after housing cost income is below the UK poverty benchmark, the housing is therefore not affordable. Recommendation 7b: (2) disaggregated version of policy assumption guideline – as (1) but with focus on average, national measures for significant groups of household types where we expect to observe and analyse important variations e.g. working age single adults, single families, couples, couples with children, single and couple retired households, young adults, women, ethnicity and other important protected characteristics. This will require a degree of new data capture, analysis and publication to support work done by providers, trade bodies and wider citizen interests.

What due diligence is required for a working definition to be operationalised?

The national policy assumption has many possible consequences and will require impact assessments, an assessment on public spending implications, as well as policy design work to consider how this may be implemented, phased-in and transitioned towards a more consistent approach to affordability on Scotland. These considerations need to also reflect impacts on providers’ business plans. As has been indicated, the shared understanding measure will need to be regularly monitored and reports published in a timely way, and this will require data and statistical investment, for the most effective utilisation of the performance of recommendations 7a and 7b.

Recommendation 8: Scottish Government, local authorities and providers need to collaborate to undertake a series of economic and equality impact assessments around the new affordability measure. They also need to consider how the new measure might be efficiently but fairly transitioned over time to deliver a progressive realisation of the new shared understanding. There will need to be new data collected to support monitoring the measure. This can draw on recent experience with the new fuel poverty measure.

Recommendation 9: The shared understanding should be viewed as a long-term objective or target of the Housing to 2040 strategy. Targets should be progressively realised by its end date, but with ongoing monitoring. At the same time, the shared understanding target should be used forthwith to influence and shape future policy discussions relevant to affordability e.g. seeking to converge over time the definitions of affordability used in policymaking.