Measured and found wanting: Difficulties in quantifying poverty

What is the best guide to a household’s resources? The household’s income – or how much it spends?

The answer to this question determines our understanding of people’s living standards and who is in poverty. And perhaps surprisingly, the households who spend the least are often not those with the lowest incomes.

In particular, poverty in the UK is typically measured in terms of having a low income. But if those with the lowest incomes are not the lowest spenders, where exactly should policymakers be focusing their efforts?

Many with low incomes don’t have low expenditures (and vice versa)

Recent analysis by the Office for National Statistics (ONS) looked at how many people fall below a poverty threshold based on low expenditure, compared to a threshold based on low income.

A standard measure of poverty is the proportion of people in ‘relative low income’, meaning people in households with disposable income less than 60% of median income. So let’s define a similar expenditure-based measure, to count people in households spending less than 60% of median expenditure.

The ONS found that a similar proportion of people were in the low income group (23%) as in the low expenditure group (22%) in 2016/17. But between the groups, there was a surprising lack of overlap: only around 12% of people had both a low income and a low expenditure.

In constructing these measures, both income and expenditure are adjusted for the number of people in the household, to enable a fair comparison across different-sized families. (A couple with children will probably need to spend more than a single person, say, to achieve the same standard of living.) For the purposes of this analysis, income and expenditure are measured after deducting housing costs.

Why is there not more overlap?

Part of the reason why quite so many people are counted in low income but not in the low expenditure group (and vice versa) may be that some are only experiencing a low or high income temporarily.

People experiencing a temporary drop in income might keep up previous spending patterns during the survey period, perhaps by drawing on savings or incurring debt, in the expectation that their income will soon return to normal. So we might expect to see some of these people in the low income but not in the low expenditure group. Conversely, some people experiencing a temporary income boost may keep their spending at a low level.

On the other hand, looking at current expenditure does not take into account benefits from goods already bought – for example, housing or cars.

A more prosaic reason is that getting accurate data is difficult. This information is typically collected via surveys and survey respondents may under-report their income or their expenditure (particularly their income). We know that income from various benefits tends to be under-reported in surveys, compared to the total amount paid out by the Government.

You can have a higher income and still be deprived

A more subjective approach might help shed some light on the disparity – we can ask people if they are able to afford certain key goods or services. The Department for Work and Pension’s (DWP) Households below average income dataset (the official source for statistics on low income) provides just such a measure of ‘material deprivation’, although for children and pensioners only. Keen readers may recall we discussed material deprivation in a previous post, Not poor but in a JAM.

Based on the DWP data, around 18% of children were counted as materially deprived in 2016/17 compared to around 30% in low income. But only 11% were both materially deprived and in low income.

By comparison, the ONS analysis suggests around 27% of children were in the low expenditure group and 32% were in low income, but only 18% had both low expenditure and low income.

The ONS and DWP figures are quite different in some respects: they come from different source surveys, and inability to afford goods or services is a different concept to low expenditure. Still, we can see that any one of these approaches to poverty measurement – income, expenditure, something more subjective – is unlikely to give a fully consistent picture of what constitutes a low standard of living in the UK.

So why the focus on income?

Income is arguably the most intuitive way to think about the resources available to households and it is easier to think about policy impacts in terms of household income than household spending. That helps explain why official statistics have in the past generally measured poverty in terms of low income.

The four targets in the Child Poverty Act 2010, which would represent the ‘eradication’ of child poverty, were mostly based on household income. However, one target used a combined low income and material deprivation measure, recognising that income may not always reflect the extent to which families can afford necessities.

Nevertheless, income measures came in for heavy criticism from ministers after 2010, partly on the basis that they failed to address the causes of poverty and led to skewed policy responses. The Welfare Reform and Work Act 2016 removed the targets in the Child Poverty Act and instead introduced two ‘Life Chances’ indicators relating to children in workless households and educational attainment at age 16. These are joined by other non-statutory measures of childhood disadvantage within the current government’s Improving Lives: Helping Workless Families suite of indicators.

Don’t forgo income yet

Although the Improving Lives indicators can tell us about particular types of disadvantage, we still need a yardstick by which to measure a household’s access to resources. So there certainly remains a role for income measures (or expenditure, or a subjective measure of material deprivation, or a combination of these).

Researchers at the London School of Economics analysed the responses to an earlier 2012 government consultation on Measuring Child Poverty and found ‘near universal support for keeping income at the heart of poverty measurement’. But just as the Child Poverty Act used several different income-based measures, keeping the group of low expenditure households in view may give us a more rounded view of who’s actually in need.

The danger, of course, is that debates about poverty degenerate into statistical arguments and we lose sight of the people who actually experience it. Measures are only useful in so far as they help us respond to problems – and are little use at all if we don’t understand what living in poverty is actually like.

Feargal is a senior library clerk at the House of Commons Library, specialising in incomes and poverty.