Does the new poverty measure fully capture disability poverty?

In recent years, we have seen fierce political battles over what poverty is, and the best way of measuring it. The Social Metrics Commission (SMC) is therefore a brave venture – to get a politically diverse group of people to agree how poverty should be measured in the UK, led by one of Iain Duncan Smith’s former special advisers, but well-represented by a host of the-great-and-the-good from the Labour years. Amazingly this seems to have succeeded, with this week’s SMC report being been well-publicised and well-regarded (see the IFS response in The Times, or the BBC or the Guardian, and a more lukewarm Telegraph piece), but in this post I want to put some scrutiny on a central claim of the new measure: that it better captures poverty among disabled people.

Why we need to consider the extra costs of disability

There is a clear problem with the current main measures of poverty (known as ‘Households Below Average Income’, or HBAI): they look only at how much money people get. This assumes that everyone with the same amount of money has the same ability to buy the essentials of life, but it completely ignores the fact that some people – including many disabled people – have higher costs of living. (There’s a huge amount of research showing this empirically in different ways: for a list of references see my previous Inequalities post and the link of publications at the bottom).

One of the main claimed benefits of the SMC poverty measure is that it tries to tackle this by accounting for ‘inescapable costs’, such as ‘the impact that disability has on people’s needs’ (in their own words). It does this in a simple way: (i) Some people are given benefits (PIP, DLA, or AA) to help them meet the extra costs of disability. (ii) The value of these benefits that disabled people receive are treated as an estimate of the costs that they face. In other words, if you receive £80/week of PIP to meet the extra costs you need for daily life and getting around, then you are no better off than someone who doesn’t have these extra needs and doesn’t receive PIP.

They also do a host of other tweaks to measuring poverty, which overwhelmingly seem very sensible (These are summarised on p132 of the report). This includes taking people’s assets into account (if you’ve got savings, you can use these to buy essentials); doing a better job of accounting for childcare spending and mortgage payments as an inescapable cost; and some further tweaks around which groups of people are assumed to share income with one another, assuming that people in overcrowded housing need bigger houses to escape poverty, and including those living on the streets.

The impact on poverty statistics in the UK

The headline is that families that include a disabled person are much more likely to be in poverty than other families – the report summarises this as follows (the figures for disabled people themselves are available from their website’s graphing tool):

The report tells us that nearly half of people in poverty are in a household that contains a disabled person (48% in their measure vs. 44% in HBAI; see p135). They don’t actually compare the poverty rates, but this is easy enough to pick up from the official HBAI release though (using the After Housing Costs HBAI measure):

• Among households with a disabled person, 26% are in poverty in HBAI, vs. 28% in the new SMC measure.
• The gap in poverty rates between households with (vs. without) a disabled person is 6% (percentage points) in HBAI, but 11% in the new SMC measure.

In other words – when we try and account for the extra costs of disability, then the additional risk of poverty among households containing a disabled person (vs. those that don’t) is much higher.

Better – but good enough?

There’s a lot to be said for this approach. Indeed, I myself (with colleagues at the New Policy Institute) have used exactly the same approach to argue that there was a ‘missing million’ of disabled people who should have appeared in the poverty statistics. It was also recommended by Prof John Hills in his review of fuel poverty for the Government, and was used just last month by statisticians at the House of Commons Library. And as the SMC themselves say, the overwhelming majority of people who give an answer think that poverty should take account of the extra costs of disability (according to their commissioned YouGov poll).

However, there’s a problem. It assumes that disability extra cost benefits (DLA/PIP/AA) are a good measure of the extra costs of disability, when we know that they’re only meant as a contribution towards these costs, particularly for those with higher costs (see p17 of our NPI report & the review before it). In our NPI report we use a second crude measure of adjusting for the extra costs of disability, and find that it makes even more of a difference.

Moreover, if our poverty measure assumes that disability extra cost benefits = extra costs per se, then cutting these benefits has no impact on poverty by definition! That is: if people are receiving less money, then the poverty measure assumes their needs are also lower – which clearly isn’t the case. So to the extent that changes to PIP/DLA mean that fewer people claim benefits who might have done, this won’t be captured in the figures. (Note however that on my back-of-the-envelope calculation from official DLA/PIP/AA data, the proportion of the population who receives these benefits hasn’t changed much 2010-17 – let me know if you know of a more a systematic comparison than this).

As the SMC themselves conclude, we need more research on this to enable us to get a robust, policy-independent measure of poverty that captures the extra costs of disability. I completely agree with what they say here – we need both:

• “Significant work to understand and measure the extra costs of disability that different disabled families face”; and
• “Improvements to data captured on disability in both the FRS and Understanding Society to allow for clearer identification of disability types and severity so that extra costs can be understood within the survey data”

The SMC report is a useful step forward, but we need to keep walking…