PSD Blog - The World Bank Group

Who Creates Jobs?

Editor's note: Dorsati Madani is a Senior Economist at the World Bank's Strategy and Analysis Unit (CICSA) of the Investment Climate Advisory

There is widespread perception that small businesses are the primary creators of jobs in most countries. A presentation from the World Bank's recent Conference on Entrepreneurship and Growth entitled, "Who Creates Jobs? Small vs. Large vs. Young" used US census data to add an interesting twist to the debate regarding the role of firm size in employment creation.

The paper finds that business start-ups contribute substantially to both gross and net job creation, but that young firms also have very high rates job destruction due to frequent failures. Conditional on survival, young firms grow more rapidly (in terms of employment) than their more mature counterparts. A related finding is that once the authors account for firm age, it becomes the definitive factor in job creation. Firms employing 5-499 employees are found to have lower net growth rates than the largest firms (10000 or more workers) in the economy. 

These results raise an interesting policy issue. Assuming a similar pattern of job creation for developing countries, how do we reconcile the emphasis placed by development institutions - including the WBG - on small and medium enterprise (SME) development? Is this sub-sector the right focus of our PSD efforts? If so, how can we ensure not only SME creation, but survival to support net job creation over the medium to long term? On the other hand, if the SME approach does not create the net medium to long term jobs so desperately needed in high unemployment developing countries, what should our policy recommendations be?

Easterly on the limits of RCTs

Bill Easterly has a very nice post summarizing many of the concerns around randomized control trials (RCTs). The summary overlaps with some of the points I tried to make here, here, and here. Money quote from Easterly's post:

RCTs are infeasible for many of the big questions in development, like the economy-wide effects of good institutions or good macroeconomic policies. Some RCT proponents have (rather naively) claimed RCTs could revolutionize social policy, making it dramatically more effective – this claim itself can ironically not be tested with RCTs. Otherwise, embracing RCTs has led development researchers to lower their ambitions. This is probably a GOOD thing in foreign aid, where outsiders cannot hope to induce social transformation anyway and just finding some things that work for poor people is a reasonable outcome. But RCTs are usually less relevant for understanding overall economic development.

Evaluating the OLPC pilots

Michael Trucano of the World Bank's EduTech blog has posted a valuable round-up of various evaluations of One Laptop Per Child (OLPC) pilots around the world. Michael warns that many of the evaluations are of short-term and small-scale pilots, which limits our ability to extrapolate.

However, the Inter-American Development Bank and others have started evaluations of much larger OLPC implementations, most notably in Peru. But here's another question for the evaluators out there (and one that Michael asks, albeit very nicely, in his post) - can you really get an accurate evaluation of a high-profile project when so many parties have a vested interest in seeing it succeed? Not that there's anything wrong with wanting a project to succeed, but what happens when the next development fad comes around, and the OLPC is no longer the cool new kid on the block? The same question would seem to apply to a whole range of interventions (the Millennium Villages immediately come to mind).    

Evaluating the evaluators

My recent post discussing the unflattering findings from a randomized control trial of microfinance prompted some heated discussion of just what evaluation can tell us in this context. It turns out that a new Policy Research Working Paper takes up exactly this issue (Impact Assessments in Finance and Private Sector Development: What Have We Learned and What Should We Learn?).

If you got worked up about the microfinance findings, then this paper is definitely worth a read. For those who are more time-strapped, here is the CliffsNotes version:

• Randomized control trials can't tell us everything, but that's OK because there are other useful ways to measure impact.
• Evaluations should focus as much on the question of why there was an impact as on whether there was an impact.

Update: David McKenzie, the author of the working paper, has a question for those of you involved in finance and private sector development projects: If you haven't incorporated an impact assessment into your project, why not? And (this one is my own) what would it take to make it possible for you to include this kind of assessment?

The verdict is in on microfinance

And it's not pretty. The results from the first large-scale randomized trial of access to microfinance indicate that it comes up short in many areas of human development. 52 of 104 slums in Hyderabad were randomly selected to receive new branches of a microfinance outfit called Spandana. Abhijit Banerjee and the other randomistas from the Poverty Action Lab describe the results in The Miracle of Microfinance? Evidence from a Randomized Evaluation:

...microcredit does have important effects on business outcomes and the composition of household expenditure. Moreover, these effects differ for different households, in a way consistent with the fact that a household wishing to start a new business must pay a fixed cost to do so. Existing business owners appear to use microcredit to expand their businesses: durables spending (i.e. investment) and business profits increase...

Placing evaluation at the center of U.S. foreign aid

I just saw the news that a piece of legislation was introduced last month in the U.S. House of Representatives to overhaul American foreign aid. The mainstream media seems oddly quiet about it (I checked Factiva but came up with very little coverage.) I was pleased to see that one section of the bill (called the Initiating Foreign Assistance Reform Act of 2009) lays out what appears to be a serious evaluation agenda as part of the reform (starting on page 7):

The President shall develop and implement a rigorous system to monitor and evaluate the effectiveness and efficiency of United States foreign assistance. The system shall include a method of coordinating the evaluation activities of each Federal department or agency primarily responsible for carrying out United States foreign assistance programs with evaluation activities carried out by other such Federal departments and agencies, and when possible with other international bilateral and multilateral agencies and entities.

What are financial literacy programs good for?

Could they help us save the world from future financial crises? If we are to believe the work of economist Robert Shiller and the larger school of behavioral economics, then the answer is "yes", or at least a partial "yes." Households in OECD countries - particularly the U.S. - overleveraged themselves, placing themselves in a precarious position that turned the popping of the housing bubble in the U.S. into a systemic crisis. In theory financial literacy programs will keep households from getting themselves in over their heads, which should help prevent downturns from turning into crises.

More support for good jobs

Some months ago I discussed a paper on the myth of the entrepreneurial middle class. Abhijit Banerjee and Esther Duflo, two of the rockstars in the field of evaluation, argued in the paper that what development requires is "good jobs":

A good job is a steady, well-paid job—a job that allows one the mental space needed to do all those things the middle class does well. This is an idea that economists have often resisted, on the grounds that good jobs may be expensive jobs, and expensive jobs might mean fewer jobs. But if good jobs mean that children grow up in an environment where they are able to make the most of their talents, one might start to think that it may all be worth it.

The idea of a "good job" gets more support from a recent article in Seed Magazine by Sendhil Mullainathan, a professor at Harvard. In Poor Decision Making, he argues:

Smackdown of the OLPC smackdown

Last month I wrote a post called OLPC smackdown, for which I received a number of critical comments. The item pointed to an article by Jon Evans in The Walrus Magazine criticizing the One Laptop Per Child program. Commentor Osimod thought the article wasn't even worth discussing: "I was never really convinced by OLPC, but that article is so *superficial* that I don't think it's worth publicizing it." I'll grant that the article was a superficial analysis, although I still think it brought up some valid points. (And I just pointed to the article - you should see the reaction that Jon Evans got!)

But if you're looking for rigorous analysis, I'm glad to oblige. A new working paper on The Use and Misuse of Computers in Education looks at the results of a randomized evaluation of the use of computers in classrooms in Colombia. (I should note that the computers in this particular evaluation were not provided by OLPC, but the evaluation should still tell us something about the utility of computers in the classroom.) The results were underwhelming, to say the least:

Evaluating Doing Business

Theory and much empirical evidence have suggested that Doing Business reforms are good for the business environment, and, consequently, good for employment and poverty reduction. But Doing Business reforms had never been subjected to the strictest kind of empirical test - until now. Recent evidence from Mexico based on a matched difference-in-difference evaluation - not too different from a randomized evaluation - has found that reducing the number of days to register a business does make a significant difference in the business environment. The second in the FPD Impact series of newsletters has the details: