‘Below Average Resources’ (BAR) is the name of a new poverty measure currently being developed by the UK Department for Work and Pensions (DWP). A methodology document and a first set of BAR statistics were published on 18 January 2024, alongside a request for feedback.
BAR closely reproduces a measure developed by the Social Metrics Commission (SMC) in 2018, which aimed to address some of the issues with existing poverty measures. The SMC’s “ambition was to develop metrics […] which could be used to build a consensus around poverty measurement and action in the UK“. While DWP intends for BAR, once fully developed, to sit alongside their existing measures of low income, it is unclear which measure will take precedence when reporting on poverty.
Below, we summarise the key differences between BAR and the relative poverty measure.
Relative poverty (or ‘relative low income’ as DWP calls it) is currently the most commonly used measure of poverty in the UK. It compares household incomes, which are adjusted for household size, with a typical household income (the median) in the same year. A household is in relative poverty if their combined net income from all income sources is below the poverty line, which is defined as 60% of the UK median household income.
Income-based poverty can be measured before or after housing costs. In Scotland, poverty is usually considered using household incomes after housing costs, which means incomes are net of housing costs, such as rent.
BAR is based on relative poverty after housing costs, and in the remainder of this blog, the term ‘relative poverty’ refers to the existing relative poverty after housing costs measure.
What does the new measure show?
BAR uses the same main data source as the relative poverty measure, the DWP’s Family Resources Survey. But BAR’s definition of income is intended to be more refined: additional assets such as savings are added to the available resources; other inescapable costs (not just for housing), such as childcare costs, are deducted. These and other changes to the methodology, are explained in the section ‘What issues is the new measure trying to address?’.
Together, and when comparing to the relative poverty measure, these refinements increase poverty rates for those more likely to be in households with extra inescapable costs, such as children, and reduce poverty rates of those more likely to have extra assets, such as pensioners.
Poverty in the UK
Compared with the definition of income used for the relative poverty measure, BAR incomes are most affected by including assets, and deducting disability benefits, childcare costs and mortgage capital repayments. In addition, stabilising the poverty line affects how poverty changes over time. These differences are discussed below.
DWP’s analysis goes into some detail on the impact of these changes on the overall UK poverty trends, as well as the poverty trends for children, working-age adults, pensioners, and disabled people. The BAR poverty threshold has been set so that BAR and the relative overall UK poverty rates are the same in 2016-17.
Overall poverty trends don’t change much using BAR compared to relative poverty. DWP produced BAR estimates for 21 years from 2000-01 to 2021-22. During this period, both BAR and relative poverty were broadly stable, with small fluctuations between 20% and 23%. The main difference between BAR and relative poverty is driven by the stabilised poverty line (explained below), which resulted in slightly higher BAR rates between 2007-08 and 2013-14, following the financial crisis.
Chart 1: UK poverty under the two measures
Working-age adults are the largest age group in the population, and poverty trends for working-age adults look very similar to those for the overall population. But children’s BAR rates are consistently higher than relative child poverty rates, and pensioners’ BAR rates are consistently lower than relative pensioner poverty rates. This is because in BAR, childcare costs and mortgage capital repayments reduce the available incomes of families with children, whereas pensioners rarely have these costs and are more likely to have savings that increase their available income under the BAR methodology.
Disabled people who receive disability benefits have these benefit amounts removed from their available income under the BAR measure. This means that BAR rates for families with disabled family members are higher than relative poverty rates, and conversely, BAR rates for families without disabled family member are lower than their relative poverty rates.
DWP also shows that the BAR measure has a stronger association with measures of material deprivation and household food insecurity. This means that people who cannot afford certain essential items including enough food are more likely to be identified as poor under the BAR measure than the relative poverty measure.
Poverty in Scotland
The analysis above summarises the main impacts of the BAR methodology on a UK-level. Additionally, it would have been interesting to see how the components of the BAR methodology affect UK countries differently. For example, the chart below shows poverty rates for Scotland using BAR and relative poverty (using three-year averages data), but we don’t have breakdowns for age groups and disabled people. We would expect different impacts in Scotland because Scotland’s age profile, tenure mix and housing market, and disability benefit policies differ from the rest of the UK.
In recent years, poverty levels and trends in Scotland appear the same on both measures. But prior to 2013-16 and all the way back to 2000-03, relative poverty rates were lower than BAR rates. This means that in Scotland, BAR and relative poverty diverged more consistently than in the UK overall. Without further breakdowns, we cannot be certain about the reasons for this.
Chart 2: Scotland poverty under the two measures
What issues is the new measure trying to address?
The following sections give more detail on the BAR methodology and how it differs from relative poverty.
BAR uses the same main data source as the relative poverty measure, the DWP’s Family Resources Survey. But BAR’s definition of income is intended to be more refined. For example, additional assets such as savings are added to the available resources; other inescapable costs (not just for housing), such as childcare costs, are deducted. These and other methodology decisions are explained below.
Reflecting people’s financial assets
The relative poverty measure accounts for income from investments, but BAR goes much further. BAR treats people’s liquid savings, such as money held in savings accounts, as additional income. This treatment increases the incomes of many pensioners, reducing pensioner poverty rates under the BAR measure.
The additional income from liquid savings is calculated as follows. All liquid savings are divided by 52 and added to weekly incomes. Essentially, this approach assumes that people top up their income with a share of their savings so that all savings are depleted after one year. Clearly, this assumption is not realistic as very few people would treat liquid assets in this way. DWP acknowledges this, and this is one of the many areas where they request feedback.
Capturing other inescapable costs
The relative poverty measure accounts for housing costs (rent and mortgage interest, water charges) as well as direct taxes such as income tax, national insurance contributions and council tax.
In addition to these costs, BAR also deducts other costs from income: mortgage capital repayments, childcare costs and (some) extra costs incurred by disability. The exact costs incurred by disabilities vary from person to person and are unknown. As a proxy, DWP excludes any income from disability benefits (Personal Independence Payment, Disability Living Allowance, and Attendance Allowance), reasoning that these payments are intended to cover the extra costs and should therefore not count towards available resources.
Debt repayments are also intended to be deducted once debt data becomes available. Further inescapable costs are acknowledged but have not yet been considered in detail: social care costs; more accurate costs related to disability; travel to work costs; household energy costs.
Stabilising the poverty line
Traditionally, the poverty line for relative poverty measures is recalculated every year as 60% of median income. When incomes change rapidly, this annual recalculation can lead to surprising results. For example, following the UK recession in 2008-2009, incomes fell sharply, and so did the median and, therefore, the poverty line. The lower poverty line meant that many households were now suddenly sitting above the poverty line even though their income didn’t change. As a result, poverty estimates showed a decline in poverty even though the living standards of many were unchanged or fell.
BAR uses a poverty line that is an average over three years of income data. This means that it changes more slowly, and sudden changes in the economy are only reflected in people’s incomes but not the poverty line itself. Stabilising the poverty line in this way affects the poverty trend.
BAR uses a poverty threshold of 54% of the three-year averaged median, not 60%. This percentage was set by the SMC so that the BAR rate and the relative poverty rate would be the same in the reference year 2016-17. The threshold was chosen so that the measure would not substantially change the understanding of the number of people in poverty but would improve the understanding of the composition of the types of people in poverty.
Other, more minor adjustments
Reflecting unsuitable housing – overcrowding
BAR income is adjusted for families living in overcrowded households. A household is considered overcrowded if the number of bedrooms fails to meet the bedroom standard established in the Housing (Overcrowding) Bill of 2003. The cost of renting an additional room is deducted from household income to reflect the housing costs for the household if their housing was adequate.
Identifying who in the household shares resources
In relative poverty, everyone at the same address with shared cooking space and living room or dining space is assumed to pool and share their resources. However, this also includes house shares and lodging arrangements, where in reality people are likely to only share some costs. In BAR, people are only considered to pool and share if they live together and are related, such as couples and their children, grandparents, grown-up siblings living at home and similar.
The impact of this change in who in the household shares resources is negligible.
Potential future adjustments
The following limitations of relative poverty are acknowledged, but not currently implemented in BAR.
Including the whole population
The relative poverty measure used in the UK (and in Scotland) is based on data from the Family Resources Survey, which surveys people in private households. There are several groups of people who are excluded from the survey and are therefore not reflected in the poverty statistics. These are (some) homeless people, for example those who are sleeping rough, and people living in institutions such as student halls, care homes, or prisons.
DWP considered including rough sleepers in BAR but found that doing so would not affect overall poverty levels meaningfully. Furthermore, “DWP acknowledges the important issue that measures based on household surveys will not be fully representative of the UK population. However, the DWP propose to focus on development of the BAR measure based on the FRS data currently and to acknowledge that this will not include those not represented in household surveys. Given the complexity of meaningfully capturing groups not represented by household surveys, it is agreed that separate substantial work would be required to investigate this issue.”
Improving adjustment for household size
Income-based poverty measures adjust for household or family size via ‘equivalisation’. This process reflects the assumption that all income in the family is pooled and pays for everyone’s shared and individual expenses, such as housing, bills, groceries, activities. Under this assumption, a two-person family requires more money than a one-person family, but not twice as much, as many costs are shared. This means that we cannot directly compare incomes of large and small families, we need to adjust for the number of people. Equivalisation applies a factor to each family’s income so that it is comparable to a 2-adult family. In the established methodology (using the ‘modified OECD scale’), the factor depends on the number of people in the family and how many of them are under 14 years old.
The problem with this adjustment is that it is an oversimplification of how family members share resources, and individual circumstances may vary widely. While most agree that the adjustment isn’t great, it has so far been very difficult to come up with something less arbitrary.
Taking poverty measurement further
DWP provide poverty depth measures in their BAR report, breaking down the households in poverty into five bands, from those just below the poverty line to those in deep destitution. Looking at this data more closely using the published spreadsheets, it shows that over time, a larger share of people in poverty are in deep poverty. The deepening poverty in the UK is something the Joseph Rowntree Foundation have been pointing out over the last few years.
Persistent poverty means someone has been in relative poverty for several years. DWP publish persistent poverty statistics in a separate publication, using data from the Understanding Society survey. The Scottish Government also publish these statistics for Scotland.
The SMC’s “lived experience indicators analysis looks at the association between a range of factors (under domains of: family, relationships and community; education; health; family finances; and labour market opportunity) with the experience of being in poverty or not according to the measure. This analysis seeks to deepen understanding of the experience of poverty and broaden understanding of the potential tools available to address it. Further detail on the lived experience indicators proposed by the SMC can be found in the SMC publications.”
DWP has decided to initially focus on developing BAR further, using Family Resources Survey data only. The development of a persistent poverty measure using BAR, as well as lived experience indicators remains “for future consideration”.
CONCLUSION
BAR, the new poverty measure addresses some limitations of the existing poverty measure by accounting for additional sources of income, and additional inescapable costs, and some other changes. As a result, the new measure better identifies people who are struggling to afford food and other essential items.
On balance, BAR, as it is implemented in DWP’s first iteration, doesn’t require much more data to be collected, and its results also aren’t dramatically different from relative poverty estimates. BAR makes children and disabled people look a little bit poorer, and pensioners a little bit richer.
BAR is currently under development and undergoing an analytical consultation (until 11 April 2024). This is likely to lead to changes in the methodology and resulting estimates.
We look forward to further developments of the new measure.
Maike Waldmann, SPICe statistician