Dear governments and aid agencies: Please stop hurting poor people with your skills training programs

Here is an incredible number: From 2002 to 2012 the World Bank and its client governments invested $9 billion dollars across 93 skills training programs for the poor and unemployed. In lay terms, that is a hundred freaking million dollars per program.

Unfortunately, these skills probably did very little to create jobs or reduce poverty.

Virtually every program evaluation tells us the same thing: training only sometimes has a positive impact. Almost never for men. And the programs are so expensive—often $1000 or $2000 per person—that it’s hard to find one that passes a simple cost-benefit test.

You might think to yourself: That’s not so bad. Nobody hurt the poor. Plus the trainers and the firms probably benefited. So it’s not a total loss.

If you think this, I urge you to transfer to an organization where you can no longer affect the world. I can think of a couple UN agencies with excellent benefits.

Because when you take billions of dollars a year (because the World Bank is hardly the only spender on skills programs) and you spend them on vocational bridges to nowhere, you have denied those dollars to programs that actually work: an anti-retroviral treatment, a deworming pill, a cow, a well, or a cash transfer. You have destroyed value in the world.

I know what some are thinking: skills program just have to be more market-driven, or on-the-job, or linked to firms, or targeted to the right people.

Maybe. And these might pass a cost-benefit test if you can make them cost much less. But I want you to ask yourself: do you want to run programs that are hard to get right, or hard to get wrong?

Because if you want to create work for unemployed people, and reduce extreme poverty, there are in fact programs that are hard to get wrong.

It gets better. Currently, about two billion people live in countries that are deemed fragile or have high homicide rates. Jobs and incomes in these countries will probably mean less crime, and maybe even a decrease in other kinds of violence. Especially if they are targeted to the highest-risk men.

If you’re thinking to yourself “hey, I would like to read 20,000 more words on this, preferably in dry prose,” well do I have the paper for you. I have a new review paper with Laura Ralston: Generating employment in poor and fragile states: Evidence from labor market and entrepreneurship programs.

It is a draft for discussion, and comments and criticisms (in emails, blog comments, and prank calls) will be integrated over the coming months.

Fortunately the paper includes a 4-page executive summary. And, even better, an abstract!

The world’s poor—and programs to raise their incomes—are increasingly concentrated in fragile states. We review the evidence on what interventions work, and whether stimulating employment promotes social stability.

Skills training and microfinance have shown little impact on poverty or stability, especially relative to program cost. In contrast, injections of capital—cash, capital goods, or livestock—seem to stimulate self-employment and raise long term earning potential, often when partnered with low-cost complementary interventions. Such capital-centric programs, alongside cash-for-work, may be the most effective tools for putting people to work and boosting incomes in poor and fragile states.

We argue that policymakers should shift the balance of programs in this direction. If targeted to the highest risk men, we should expect such programs to reduce crime and other materially-motivated violence modestly. Policymakers, however, should not expect dramatic effects of employment on crime and violence, in part because some forms of violence do not respond to incomes or employment.

Finally, this review finds that more investigation is needed in several areas. First, are skills training and other interventions cost-effective complements to capital injections? Second, what non-employment strategies reduce crime and violence among the highest risk men, and are they complementary to employment programs?

Third, policymakers can reduce the high failure rate of employment programs by using small-scale pilots before launching large programs; investing in labor market panel data; and investing in multi-country studies to test and fine tune the most promising interventions.

75 thoughts on “Dear governments and aid agencies: Please stop hurting poor people with your skills training programs

  1. Interesting post. Wonder if this applies across sectors. I am thinking that the agricultural sector in particular can greatly benefit from training. Smallholder farmers are amongst the poorest in the world but they own land, their productive asset, and perhaps also livestock.

    They have a chance of applying new skills right away and are not dependent on external organisations to provide jobs. Of course there must be some sort of market access but have seen how training based programmes (sustainable agriculture, animal management and gender) can easily treble or a household’s income.

    It is particularly important for development as farmers often lack access to inputs so need to rely on a knowledge intensive agriculture to be resilient to the challenges they face.

  2. Great post, but I would add several other reasons why training programs are at best expensive and often counter-productive. Training programs that do not lead to jobs are a cruel hoax visited on vulnerable people. Even if some people do get jobs, I would argue that these programs often just reshuffle who gets the few jobs that are out there. This is bad enough, but to the extent that these programs give the appearance of addressing youth unemployment, without actually doing so, they are complicit in governments’ failures to address the real constraints to job creation. There is a long list of these, from bad governance to closed markets to bad roads to weak education systems, and on and on. Ultimately, these constraints add up to calculations in the minds of potential employers that hiring an additional young worker is not a winning financial proposition. The causes of low worker productivity extend far beyond those supposedly addressed by training programs, with many of them having nothing to do with labor, per se. If you can count on electricity just half the time, half the potential labor productivity just flew out the window. If you can’t get credit for new machinery, there went another chunk, etc. Donors want to do the right thing, but wanting and doing are very different.

  3. Hello & thanks a lot for the interesting paper! It quotes a paper on your research project in Ethiopia (Blattman, C. and Dercon, S. (2015). More sweatshops for Africa? A randomized trial of industrial
    jobs and self-employment). I was eagerly waiting to read more about it, is it available somewhere?

  4. Interesting post. It might be interesting to compare what job training programs for households with very low incomes in developed nations have to say that might influence how we think of evaluations of these programs in low-income nations or what dynamics might work.

  5. $ 9bn over 10 years doesn’t seem like a huge amount compared to the figures being discussed in Addis. People know that capital injections, cash-for-work, and public works work. But they take a long time, and unless government led and sustained, may not provide an inter-generational route out of poverty.
    Some ‘skills training’ programs are also viewed as more politically palatable than telling people to get a job in a sweatshop, and that’s why governments often advocate for them.

  6. Based on my in-depth research in Tanzania with Josaphat Kweka and Xinshen Diao, I am happy to report that the good news is that people are not waiting for training programs or any other kind of government or NGO assistance to improve their livelihoods. We show that the majority of new jobs created in Tanzania between 2002 and 2012 were created in small firms. Moreover, the productivity in a great deal of these small firms is higher than economywide productivity in manufacturing! Using one of the only nationally representative surveys of small firms that I knokw of in Africa, we use information from the qualitative section of the survey to show that 99% of these firms never received any kind of assistance from the government or from an NGO.

  7. Very interesting piece. I agree totally that skills training programs belong in the 60s. What are poor communities trained to do other than deepening their engagement on those activities that continue cementing their status as the poor? Door making, bead making, masonry, automotive engineering skills will not transform poverty status magically or fast enough. Bring the capital and innovation to the table….a scholarship program, agribusiness and tech investments….all else is just Messianic posturing