In the past few months, following statements by Eric Pickles and then the release of the Casey Report, there has been considerable debate on so-called 'troubled families'. But the focus in much of the media has been less on the government’s multi-agency approach to helping these families, and more on the way in which 'troubled families' have been defined and quantified. In particular, much has been made of the Prime Minister’s description of these families as 'neighbours from hell' and his statement that they are:

'The source of a large proportion of the problems in society. Drug addiction. Alcohol abuse. Crime. A culture of disruption and irresponsibility that cascades through generations.'

In fact, as many point out, the analysis that identified these families did not take crime, alcoholism or anti-social behaviour into account at all. Instead, the analysis isolated what the previous Labour government might have called 'multiple disadvantage' – i.e. poor health, poor housing, poor education, unemployment and low income.

Several excellent contributions to this debate (here and here) have subsequently criticised the government’s somewhat Victorian approach of conflating poverty with immorality or criminality. These families are simply poor – not some sort of morally corrupt sub-class – and they pointed out that Casey’s report, which painted a squalid picture of child abuse, violence and teen pregnancies, is based on interviews with 16 families, and hardly representative of the 120,000.

What this debate has shown (apart from the Government’s fast and loose treatment of statistics) is that our understanding of those living in poverty is critically limited. So much so that our attempts to lift the lid on the blunt descriptor of 'the poverty line' has very quickly fallen into stereotype and guesswork. We rely on statistical analysis from 2004 (as this is where the 120,000 figure arose) and a handful of anecdotes to inform a policy with a £448 million budget. It’s hardly the way to go about things in such straitened times.

This is perhaps why the Government announced last month that it would consult on a new measurement of child poverty, one which considered a range of factors alongside low income. Whilst many accused the Government of moving the goal posts in order to cover their failure to meet the child poverty target , the concept of looking at a range of drivers of poverty (such as poor housing, education, disability and so on) is inherently sensible, as Jo explains here.

The critical point, however, when widening the poverty measure, is to ensure that low income remains central to the definition and subsequent measurement. If this is achieved, then we are not moving the goal posts – rather, deepening the net. We are giving ourselves a deeper understanding of how people in poverty live, what put them there, and crucially, what is most likely to get them out.

For years, our focus on income has meant we have exclusively looked to employment as the sole solution to poverty. But in-work poverty keeps rising. At some point, therefore, we must stop tackling the symptoms of poverty and start addressing the causes – and this means identifying what they are, and who experiences them.

The Multi-dimensional Poverty project is trying to do just that. We have, with the help of a range of experts and practitioners, selected 20 indicators associated with poverty, and are analysing the population living below the poverty line to identify which of these indicators cluster together. We are generating 'types' of poverty in three separate cohorts – working age families with and without children, and older people. We have found each cohort has five or six different common types of poverty, characterised by a different combination of indicators.

Overall, this means we have 16 individual types of poverty in Modern Britain, each with low income as a common denominator. What does this mean for tackling poverty on the ground? Well, in an ideal world, it means having a range of different strategies, characterised by a different combination of agencies working together to fight poverty in a nuanced, multi-dimensional way. One type of poverty might require adult education and social housing to work together, whilst another might require mental health services and Sure Start to work together.

Of course, we are aware that all analysis can be used inappropriately, and some may try to give our types of poverty moral or cultural characteristics which cannot be proven from our data. However, these sorts of leaps of logic are symptomatic of limited evidence, or a limited grasp of the evidence.

Demos and NatCen are working hard to create an evidence base to understand modern poverty which is both robust and accessible. If it is easily understood and transparent it will tackle- rather than feed- the stereotypes and assumptions about poverty borne from a limited grasp of relevant data. It is only through a greater understanding of the lived experience of modern poverty that we can put to bed the anecdotal and morally loaded approaches to tackling it.

New Comment