Tuesday, 10 January 2012

The effect of temporary subsidised employment on employability.

Temporary subsidised employment (TSE), like the Job Guarantee, and the training that is often associated with TSE hopefully raises or at least maintains the employability of those concerned. But is this the actual effect?

These two studies done in Switzerland claim that TSE in the private sector produces significant benefits. In contrast, the benefits of public sector TSE are not as good. As to training, the benefits are not as good as private sector TSE.

TSE does not seem to bring benefits for those with a good chance of finding work anyway (not a big surprise, I suppose).

The above two categories “private sector TSE” and “public sector TSE” do not capture the actual nature of the various TSE schemes in Switzerland with 100% accuracy. But they are a more or less correct characterisation. Anyone interested in precise definitions will have to look at the papers.

1. “A Microeconometric Evaluation of Active Labour Market Policy in Switzerland” 

2. “Does Subsidised Temporary Employment Get the Unemployed Back to Work?”

Some other bits of empirical evidence are as follows.

3. Booth, A.L., Francesconi, M. and Frank, J. (2000), ‘Temporary jobs: Who Gets
Them, What Are They Worth, And Do They Lead Anywhere?’ Discussion Paper 00/54, Institute for Labour Research, University of Essex.


This paper showed that those prepared to do temporary jobs (not necessarily subsidised jobs) fared better in subsequent employment histories than those not prepared to do temporary jobs. This effect was more marked for women than men.

4.  Calmfors, L., Forslund, A. and Hemstrom, M. (2002), ‘Does Active Labour Market Policy Work? – Lessons from the Swedish Experience’, Institute for Labour Market Policy Evaluation, Uppsala.

This confirms the Swiss finding that TES – i.e. “learning by doing” – yields better results than formal training.

5.  Bolvig, I., Jensen, P. And Rosholm, M. (2003), ‘The Employment Effect of Active
Social Policy’, Discussion Paper 736, Institute for the Study of Labour (IZA), Bonn.

This pretty well confirms the above studies, with the surprising additional claim that training actually IMPAIRS employability.


Afterthought (11th Jan). Here is another paper by Calmfors & Co with a more pessimistic take on the benefits of TSE and similar labour market programs. Title of paper: “The effects of active labour market policies in Sweden: What is the evidence?”. See:





  1. To crosspost from Mike Norman's...

    Obama's new CEA chairman Alan Krueger is one of the top labor economists in the country. He included a lot of innovative programs in the back end of Obama's ill-fated American Jobs Act worth looking at. The first one, "Bridge to Work" is a temporary subsidized employment scheme akin to what Ralph's taking about (as Morgan Warstler has suggested, the employers really should have to bid on the subsidized labor).

    324. Bridge to Work Program
    325. Wage Insurance
    326. Enhanced Reemployment Strategies
    327. Self-Employment Programs
    328. Additional Innovative Programs
    344. Grants for Short-Time Compensation Programs
    351. Long Term Unemployed Workers Work Opportunity Tax Credits
    Subtitle C – Pathways Back to Work
    364. Subsidized Employment for Unemployed, Low-Income Adults
    365. Summer Employment and Year-Round Employment Opportunities For Low-Income Youth


  2. It would be interesting to compare temp work to volunteer work. If volunteer work does it, then maybe people should just do this for a while. Build a website for a non-profit, etc.

  3. Marris, The 2nd study above (“Does subsidised temporary employment…”) looks at non-profits. The results are not as good as in the case of regular private sector employers.

  4. Employability isn't really a big part of the problem. Also, you have to differentiate between actual empoyability and perceived employability, where the latter is the one that matters during demand-deficient periods and only for the individual, not for unemployment as a whole.

  5. Pavlina R Tcherneva7 March 2012 09:32

    None of what you present above is evidence against the JG. The Swiss EP is workfare, not JG. Worse it is a means-tested allocation of unemployment insurance. What you point out to be ‘the more successful program’ in Switzerland (TEMP) is actually a very punitive arrangement, forcing people to take ‘unsuitable jobs’ that pay less than their unemployment benefits by overcompensating the difference with additional payments from the UI system (p 7).

    Your next ‘evidence’ for the UK is again anything but JG. We have always argued against workfare. See the work of one of our collaborator’s, Nancy Rose, Workfare vs. Fair Work. JG is the latter. The study finds that temporary contracts pay less and are more dissatisfactory than permanent contracts. Genius! The US did the same thing as the UK--reformed welfare, made the benefit conditional on finding some/any kind of work in the private sector usually very low pay involving long commutes, so the program didn’t lift the welfare recipients out of poverty and left them without assistance after 5 years. While the UK at least bothered to find the welfare recipients some temp private sector contracts, the US didn’t. There are so many problems with these reforms that it will take forever to discuss here. Again, not JG.

    The Swedish model. Huge paper, lots of interesting details and lit review, but did you notice that the whole paper from inception is based on a matching problem assumption, not deficient aggregate demand. Starting with section 2, it uses the worst kind of neoclassical modeling to analyze these programs. General-equilibrium effects? And I should take the results seriously? Yes, much of the literature surveyed uses the same methods. But even if we overlook the neoclassical methodology, let me just point out the obvious: The studies of the 90s found negative effects from active labor market policies (ALMP), which is precisely the period of serious neoliberal reforms in Sweden that undermined the Swedish welfare state and the full employment model. No wonder unemployment rose in the 90s and the programs that were vestiges of the past had to deal with a much bigger problem now. PLUS those employment programs themselves were reformed with more focus on training and job search, not actual employment. How does that help us understand the JG? The Swedish corporatist policies prior to the 90s were not exactly a JG, but the reforms and subsequent active labor market reform follow the ‘labor market’ flexibility neoliberal agenda. But you think this study speaks to the effectiveness of our JG model?!

    The last ‘evidence’ against JG is the topper. The Danish case actually indicates that employment programs are more effective, and that training programs alone should be abandoned (p22). We’ve been arguing forever that retraining is just a game of musical chairs. How is this paper evidence against the JG?

    Ralph, you’ve been part of the JG discussions for many years and I do not understand why you continue to mischaracterize the proposal. But I am more perplexed why studies of several ALMPs which have a very clear neoliberal design serve as evidence against the JG.

    Pavlina R. Tcherneva

    1. Pavlina,

      You are the first human being on planet Earth I’ve come across that actually takes an interest in those Swiss studies. Congratulations. I’ll take your points in turn.

      If by JG you mean temporary jobs in the public sector only, then it strikes me that the fact that PRIVATE SECTOR temporary subsidised jobs produce better restults than public sector temporary jobs IS a valid criticism of JG.

      Re the punative or workfare element in any JG/WPA type scheme, I just view that as a variable which can be incorporated to any degree one wants in any scheme. Plus if private sector subsidised work produces better results than the public sector equivalent under a relatively coercive scheme, I ASSUME the same holds for a less coercive regime. But it’s possible that assumption is wrong.

      Re your opposition to workfare (by which I mean a relatively coercive regime), it’s always possible to go for an “uncoercive” regime. But there is a problem: Calmfor’s “iron law of active labour market policy”. This is the fact that if JG type work is relatively attractive, that ipso facto reduces the relative attractions of regular jobs for those concerned. That reduces aggregate labour supply to the regular jobs market, which is inflatinary. And that in turn necessitates a reduction in aggregate demand. The net result is that JG jobs are at least to some extent AT THE EXPENSE OF regular jobs. Indeed, Calmfors & Co claim in the first Swedish study I cited above that this has been the ACTUAL EFFECT of Swedish ALMP schemes.

      Creating JG/WPA type jobs which are partially at the expense of regular jobs might be the best way to go, but we need to be honest about the latter unfortunate side effect of being “generous” towards JG people.

      Of course the latter unfortunate side effect does not occur given much higher than normal unemployment: in this scenario simply raising demand is the best cure for unemployment, not JG/WPA type schemes. I.e. JG/WPA schemes really comes into their own at relatively low unemployment levels (or at NAIRU, if you like), and it’s in this scenario that the above “unfortunate” side effect arises.

      Re lifting people out of poverty, I think that is largely a separate issue to the question as to what the best JG/WPA type scheme is. I.e. one can have a relatively generous social security and JG/WPA system as per Scandinavia, and that ameliorates poverty. Or one can have a less generous system, as per the U.S., in which case poverty is not amelerioated so much, or is not ameliorated at all.

      Re your objections to section 2 of that Swedish paper, I’m also baffled by it. I was baffled by it years a go when I first saw it. Their figure 4 looks very much like a supply / demand chart much the same as a supply and demand chart for apples, but applied to the macroeconomic arena.

      My preferred chart / graph is the one I set out here:


      I’d be interested in your reactions.

      Finally, I’m not diametrically opposed to JG. I favour SOME SORT OF JG/WPA system. Where I disagree with traditional JG (which seems to be confined to the public sector) is that I favour extending the system to the private sector. I also don’t favour what might be called “specially set up employers” a la WPA. That is, I favour having JG people taken on by EXISTING employers (public and private).

      Re the Danish study and the apparently poor benefits derived from training, I’m glad we both have a similarly jaundiced view of the sort of training often offered as an alternative to JG type work.

  6. Pavlina R Tcherneva7 March 2012 14:35

    Ralph, very quickly, yes i suspect our policy disagreement stems from the different theoretical frameworks we use about the causes of unemployment. i read your 'creative destruction paper' and used these comments above in a new NEP blog to point out in the end that i fundamentally disagree with the use of MPL and NAIRU as theoretical tools. As I am on the road now, I will try to address more of your questions next week. But in the mean time, keep an eye out for my new policy report at Levy: JG through social entrepreneurship. We really are talking about completely different types of policies.


Post a comment.