When the counterfactual *really* matters

Jonathan Portes, the Director of Britain's National Institute of Economic and Social Research has a great post up discussing the slightly controversial mandatory work experience placements in supermarkets for unemployed youngsters. Whether the scheme constitutes slave labour or not, it would be interesting to think for a second about its effectiveness.

The Minister in charge of the scheme has proudly trumpeted:

The fact is that 13 weeks after starting their placements, around 50 per cent of those taking part have either taken up permanent posts or have stopped claiming benefits.

Mr. Portes, formally a Chief Economist at the same government department, considers the counterfactual: claimants of job-seekers allowance (JSA) who do not participate in a work experience placement:

Off-flows from JSA remain high - almost 60% of claimants leave within three months

So you have a 60% chance of leaving benefits in 3 months unless you take part in this scheme, after which you only have a 50% chance. Awk-ward. Of course:

Now this is not definitive - without a proper control group and a counterfactual, we do not know what would have happened to the participants without the programme. Maybe I am wrong, and in fact those who go on the programme have very poor characteristics, and would have done even worse without it. Without proper evaluation, we just don't know. But certainly the evidence and analysis so far published by DWP does not make a good case.

Bottom line #1: if you're interested in smart and well-presented UK economic policy analysis you should really be reading Jonathan Portes.

Bottom line #2: There is probably a very good business case to be made for sending Mr. Iain Duncan Smith on the J-PAL Exec Ed course on evaluating social programs. Training budgets must be tight with all the cuts going on and that though, so - genuine offer - if you're interested Mr. Smith I'll pay your tuition fees out of my own pocket :)