How should we measure poverty?
This blog first appeared on Bright Blue.
The Welfare Reform and Work Bill currently passing through Parliament proposes to repeal the four income targets under the Child Poverty Act 2010 (relative poverty, absolute poverty, persistent poverty, and combined low-income and material deprivation), and to introduce ‘life chances’ measures in four areas: children living in workless households in England; children living in long-term workless households in England; the educational attainment of children in England at the end of Key Stage 4; and the educational attainment of disadvantaged children in England at the end of Key Stage 4.
Writing in the Times, Tim Montgomerie makes important arguments on the role of statistics in government. Accurate, timely statistics measuring things that are of interest from a public policy perspective can provide the evidence base that drives intelligent policy. Montgomerie gives examples of occasions where statistics that – with the benefit of hindsight – fell short of those standards prevented good policy. It’s a history lesson that reinforces his key argument that evidence matters. We would agree.
When it comes to poverty statistics and targets, however, we tend not to share Montgomerie’s analysis. No poverty measure can be perfect, and we would not wish to pretend otherwise. Poverty is a concept – one ably defined by the Prime Minister soon after becoming party leader (‘some people lack those things which others in society take for granted’) – not a single number, and it has many facets. The targets under the Child Poverty Act cover different dimensions – relative, fixed income threshold (‘absolute’), persistent and combined low income and material deprivation – for that reason. There is always room to add more measures or to refine existing ones as the evidence dictates. Social science is an evolving discipline.
Looking at the policy change that the targets have helped drive, the evidence is impressive. The number of children in poverty fell by more than a million in the decade after Tony Blair’s 1999 pledge to eliminate child poverty. And the ‘poverty plus a pound’ critique (that policy focused on lifting families from just under the poverty line to just above it) doesn’t really stand up to scrutiny: the IFS found that there would have been significant poverty reductions if the line was drawn anywhere between 43% and 100% of median income. Those are substantial achievements, with the UK outperforming the rest of the OECD, and it is clear that the poverty targets were a key driver behind them.
That’s not to say that everything was perfect. Progress was slowing even before the financial crash, and statistics can help us to understand why. For example, on the latest figures (2013/14), a million fewer children live in poverty than in 1996/97, yet the number from working families (around 1.4 million) is unchanged. This evidence is strongly suggestive that policies to get people into work have been broadly successful, but that more needs to be done on in-work poverty.
There are lessons here about what has worked – both successful policies and the measures and targets driving them – and look at where they can be built upon. So, if progress is stalling because in-work poverty is proving stubbornly persistent, let’s address those measures, and think about how we can tackle low pay, bring down the cost of childcare and housing, and support family incomes. If you’re worried that poverty thresholds focus policy too narrowly (although, as noted, the evidence we have doesn’t support this), then introduce measures on poverty gaps and give them equal weight. As a general rule, and providing it is of sufficient quality, more evidence is generally better than less.
There is plenty of evidence to suggest that social mobility is a problem in the UK today, and so the Government’s proposal to produce measures of life chances is, on the surface, welcome. But this takes us to the other half of our ‘more, high quality’ principle on evidence. We have in the UK a rich social science tradition, and a depth of expertise on measurement with a long history behind it that exists across government, business, and civil society. We also have a democratic process, through which this expertise can be brought to bear to help create the best measures of life chances possible. The measures in the current Child Poverty Act were the result of a lengthy process along these lines, driven by DWP. There is a lot here for the Government to work with.
Instead, however, they have presented a smattering of disconnected measures, introduced into primary legislation without any consultation. Their proposed supplementary measures on the drivers of poverty suffer from similar flaws, and ignore structural determinants like low pay, the cost of childcare and housing, and, indeed, low income itself.
Tim Montgomerie is right to highlight the value of evidence in policy, and the importance of getting that evidence right. We want to see light shed upon the distinct but interlinked issues of child poverty, life chances, and the drivers of poverty. We are deeply worried, however, that the current process that aims to redefine how we measure these concepts is being done in a haphazard, political manner. It risks abandoning what works without having thought properly about what replaces it. The government should be taking a step back, and tapping into the goodwill of all the individuals and organisations who have an interest in developing the best evidence possible. It is vital to our children’s futures that we get this right.