Evidence is the best basis for policy-making, and this summer saw the launch of another fact-filled report. Sponsored by the Tenant Service Authority, the Joseph Rowntree Foundation and the Scottish Government, Growing up in social housing in Britain: a profile of four generations 1946 to the present day is an excellent piece of work. On four occasions since the war, the British government has undertaken big cohort studies. Put simply, you take everyone born in a particular week - around 17,000 people - and follow them through their lives. We have studies of people born in 1946, 1958, 1970 and 2000. So far, the first group have been surveyed 21 times, with the others eight, seven and three times respectively. We know a lot about these people!

The authors, led by the London School of Economics’ Ruth Lupton and Becky Tunstall, are serious people. They respect data and they are no right-wing patsies: it was Ms Lupton who took Michael Gove to task last year for his misleading statistics, when he claimed educational inequality had increased under Labour.

Their report finds that for those born in 1946, growing up in council housing is associated with worse health, well-being, education and employment outcomes. The authors show that this is not a ‘council housing effect’, but rather is caused by differences in family backgrounds. Children who are likely to have poor life outcomes are more likely to live in council housing, and council housing per se does not lead to worse life outcomes. ‘Phew’, I hear you say, ‘thank heavens for that’.

Unfortunately, the authors find that for people born in 1958 and even more for people born in 1970, living even for a short while in social housing as a child is associated with worse adult outcomes even after taking into account the child’s background, with the results strongest for women.

Let us be clear here, when we say that the authors take into account the child’s background, we don’t just mean parental income or some crude measure. These cohort studies allow them to take into account 66 and 56 different background characteristics for the 1958 and 1970 cohorts respectively, including parents’ education, their interest in their children’s education, whether the child wet the bed, whether the child described themselves as happy, the characteristics of their schools, their height, their weight, the number of siblings and birth order and so on.

Equally, the cohort study allowed the authors to look at a wide range of outcomes, including the person’s judgement of their own health status, whether they are depressed, whether they smoke, their declared life satisfaction ratings, and self-efficacy, their employment rates and whether they are on means-tested benefits, their qualifications, and whether they have literacy or numeracy problems.

As the report concludes, even when you take everything and the kitchen sink into account, the statistics are clear: growing up in social housing if you were born in 1958 and even more so if you were born in 1970 is associated with worse outcomes on each and every outcome measure listed above. This is true across all regions of Britain, irrespective of housing types and quality, irrespective of neighbourhood conditions and remains the case even when you restrict the analysis to social classes commonly found in social housing. There is not a single measure on which children growing up in social housing do better. As Ms Lupton said at a seminar to launch the report, ‘There is no positive story here for social housing.’

Of course, and as the authors are quick to point out, statistics assess correlation, not causation. But until British policy makers allocate otherwise identical families to different tenures in sufficient quantities to generate statistically significant randomised samples, correlation is all we have. And as Professor John Hills remarked, once you have taken this many things into account, and once you get results that are this consistent and this robust, you have to think that maybe you have found the answer. And that answer, sadly, is that social housing is bad for our children’s prospects.
Inevitably, readers will write in with examples of children they know whose lives have been transformed by social housing. But these are the exception, not the rule: the report shows that, taken as a whole, social housing makes it more likely that they will have less fulfilling lives than children whose families are identical in every way except that they did not live in social housing.

As ever, more research would be useful. It would be good to have more explicit comparisons with those in private rented accommodation. This is hard, since not that many children born in 1958 or 1970 lived in private rented accommodation, and many who did also lived in the social rented sector. But it remains a research priority since the private rented sector and not owner-occupation is the obvious practical alternative to social housing.

But notwithstanding an academic’s wish to see more research, we need to think of the policy implications. For those in the sector, this report makes for dreadful reading. And if I were a government minister, I would want to know why we have £500 billion worth of state- and quasi-state-owned housing that reduces people’s life chances.

When politicians talk about the environment they often talk of the precautionary principle: the evidence is not conclusive, but strongly suggestive, and that is sufficient to warrant action. The evidence in this report is not conclusive, but it is strongly suggestive. Apply the precautionary principle here and the right policy would be to keep children out of social housing at almost any cost.

Dr Tim Leunig is an academic in the department of economic history at the London School of Economics