The Labour Party has created a stir in advance of today’s vote on restricting benefit increases by proposing a ‘jobs guarantee’ for 130,000 long-term unemployed. This would be funded from cutting tax relief on pensions for better-off employees.

While there is possibly some political benefit to Labour, and also the beginning of a welcome recognition that benefits cannot be unconditional (Ed Balls says the scheme will be compulsory for those receiving Jobseeker’s Allowance), I can’t say that this proposal looks a sensible use of public money.

The proposed scheme is one in a very long line of job subsidy programmes both in the UK and abroad. The coalition government already has one for young people, offering a payment to firms taking on 18-24 year olds as part of the Work Programme. The Labour proposal is to subsidise private sector firms who take on those aged 25 plus who have been on JSA for more than two years. These subsidies will only last for six months, so the ‘job guarantee’ is of strictly limited duration.

There are some practical issues around this scheme. One is that the 130,000 on JSA accounts for less than a third of all those who have been unemployed for two years or more.  The most recent Labour Force Survey figures show 444,000 people aged 25-64 who meet the ILO definition of unemployed and have been without work for two years. Many of these will be on other benefits. Would the scheme ignore them?

Another is the willingness of private firms to offer jobs of this kind: past experience suggests that the numbers required would be difficult to achieve in the short run, leading those running the scheme to seek expedients with charities, not-for-profits and local authorities.

It also seems likely that these short-term jobs will be low-level roles: firms will not want to take on people in responsible positions on this basis. Given that a fair proportion of the long-term unemployed will be qualified and experienced, many will be less than enthusiastic. It is rather different from the position of young people with no previous employment record, where entry-level jobs can be useful.

No selection process has been specified. If firms have a choice about whom they take on, it seems very unlikely indeed that all those to whom the offer applies will be accepted by employers, meaning that the residue of largely unemployable benefit recipients will have to be found ‘make work’ jobs – arranged no doubt by the sort of dodgy agencies and fly-by-night operators which have plagued other schemes over the last thirty years.

Looking more generally at job subsidies, there is now a large international theoretical and empirical literature – and it makes depressing reading. Headline figures tend to focus on payments made as if they indicated outcomes. But even if Labour’s scheme was successful in the sense that all 130,000 (or whatever) benefit recipients were taken on by employers, this does not mean that employment rises by 130,000.

There are deadweight effects, where the scheme pays for workers who would have found work anyway within the duration of the scheme. There are substitution effects, where the targeted group is taken on in preference to people who would have found jobs:  those unemployed for two years are taken on in preference to those unemployed for a shorter period, who would otherwise have found work and now remain jobless longer. And there are displacement effects: firms which have their employment subsidised (typically larger firms which can handle the bureaucracy inevitably involved) gain market share at the expense of other firms, which reduce employment as a consequence.

All this means that targeted job subsidies largely redistribute unemployment rather than reduce it overall. Net job ‘creation’ may be less than 10% of the nominal coverage of the scheme, even during the short period it operates. Moreover few of the target group taken on subsequently obtain a permanent post. There is even some evidence from matched studies in other countries (where only some individuals are randomly eligible, and thus it is possible to control for individual characteristics) that participation in job creation schemes can actually harm future job prospects. This may be because the experience of a bad job placement can demotivate job-seekers, or because participants are stigmatised in the eyes of future potential employers.

So Labour’s scheme may be headline-grabbing, but it would be unlikely to have a significant favourable impact on the long-term unemployed, let alone on overall unemployment. Schemes like this are no substitute for fundamental welfare reform, significant deregulation of the labour market, and reductions in the cost of employing labour (by for example making permanent reductions in employer and employee national insurance contributions for lower-paid workers).