Policy recommendations for the World Bank

• Yahoo! asked to explain alleged perjury
• Giant space monster continues its approach
• Yahoos in suits choose cash flow over human rights
• Junior high textbook publishers act like junior high students
• Giant space monster is watching us
• Murdoch conspiracy uncovered
• Web site induces massive loss in worker productivity
• Reading public fail to surprise.
• Super-smart objectified.
• Porn averaged.

topics covered:

[PDF version]

If you know me personally, you know that I had a number of contracts at the Bank. Some went well, but most were for a rather shifty individual who was somewhat dishonest, both academically and more generally. The contract stipulated pay for eight days, and I cut bait and left after maybe four months of mostly full-time work trying to fit models to his hypotheses--and he still withheld payment.

There you have it, right in front, and you can read the rest of this article knowing my bias. Though, I'd worked out most of the below over a year ago, and you can tell by this header that I'm self-conscious enough not to base any of the below on one personal data point.

The World Bank is huge, and that is not necessarily wrong. If we somehow broke it into ten separate organizations, those organizations would have constant turf squabbles and fights over funding, constantly need each others' assistance and/or replicate each others' work, and would probably wind up setting up some sort of interorganization council that doesn't differ too much from the Bank's top management. There are umpteen thousands of people working in development, and that's an organizational problem no matter what.

Now, there are two eminently sensible ways to divide expertise: you can go by geographic region, because Central Asia is a very different place from Sub-Saharan Africa, or you can go by topic, because poverty reduction is a very different issue from gender issues. The Bank does both: the management chart is a grid, with departments like Central Asia-poverty and Africa-gender (except everything is an acronym, so it'd be ECA PREM). That is, everybody has two bosses, thus giving everybody something to complain about at the water cooler.

The next problem is that new projects come and go at a pretty rapid pace, and always involve subject that are a bit far afield from the expertise of the managers. The solution: hire contractors. Lots and lots of contractors. The Bank doesn't have a dam-building division--they contract. Nor do they have a dam-evaluating division; that's contracted too. Nor do they have sufficient expertise in virtually any of the subjects that your typical Bank report is about.

So when you sit at the coffee shop in the Bank's spacious atrium--and a whole lot of Washington is like this--you find a small core of individuals who are there for life and a constant flux of two-year RAs and contractors. Also, you'll see that only two of the five coffees they serve will be Fair Trade certified.

Now, if you work in a technical field, you know that finding people who are both technically apt and good managers is supremely difficult. Further, even if they were the best of the best twenty years ago, maybe their COBOL skillz aren't so impressive now. I would contend that in the pool that the Bank has to draw from, there are simply not enough people who are experts in country plus subject plus general management to fill all the posts. They have to compromise.

And the compromise the Bank chooses, time and time again, is to go for the people who can best manage the stream of contractors. Subject knowledge can be hired; familiarity with the Bank and its protocols can't. To add to this, managers are rotated every few years, so if somebody really buckles down and learns everything there is to know about labor in Europe, it's down the drain when they're moved to managing technology dissemination in China.

Now, management skill generally goes with political skill, and much of the work of a Bank manager consists of not pissing off politicians. So this is one more reason to bias toward managerial knowledge over subject knowledge. It is also why most Bank reports don't actually say anything of substance.

I suppose the arrangement of hiring subject knowledge could work, but the Bank culture kills it. The Bank's employees are trained to think like business managers. Impoverished countries are referred to as "clients", and I have heard at least one manager give a speech explaining in no uncertain detail about how the Bank must be run like a business. There are clear management goals; growing GDP by 7% per year in an given country is always a popular one--and even if it is achieved via inequality-expanding means, all boats will eventually rise with the tide.

However, the Bank's questions are only partly business on the ground. A huge percentage of the Bank's product is reports, about the business climate, about how [trend of the month] can reduce poverty, about the environmental effects of dams. So the contractors the Bank will hire are a mix of the usual survey-takers and other such laborers, and academic types.

As demonstrated by me and my shifty boss (really, I had some nice ones too!), academics and hard-nosed managers don't mix. Academics want to write something that is correct to the last detail; managers want something that gets the point across. Academics want to hold off judgment until the data is in; managers want to write the introduction to the report before hiring the academics. Academics need time; managers need output by Friday. Further, experienced academics are not cheap. They have day jobs, and if you need original research, that distracts from the navel-gazing they'd rather be doing.

I asked one well-known academic, “what do you think of the World Bank?” and he snapped back, “Did they pay you?” Then there's the graduate student who confided to me that she didn't tell her adviser that she's working at the Bank; due to so many bad past experiences, her department has a policy of not allowing students to work there. After I gave up on the project at the head of this column, shifty manager attempted to contract two experienced individuals working in my field, and both refused. I'll stop with the anecdotes there.

The end result from all of this: the Bank generally runs on the young and inexperienced. If you just got your Master's, they've got work for you to do. If you can't necessarily do original research, but you know how to run Microsoft Excel, then you have what it takes to write the flagship publication, the World Development Report. If you skim through the thing, you'll notice that (1) it is primarily a very long lit review, and (2) none of the original data work goes further than a bivariate graph. Most of those bivariate graphs (e.g., pp 2.14 and 2.20) are Excel charts based on a manually-enterable number of data points. That is, it's the sort of report that you expect from a team of Master's students.

Of course, that report has its value, and if you want a survey of global poverty in 2005, you should definitely click the link above. But does the following statement inspire confidence in you?: "Resources to eradicate world poverty are allocated based upon the best research by the most eager graduates of the last few years."

In short, the Bank's structure consists of a core of managers, many of whom learned their economics in the '70s, and a constant stream of generally inexperienced subject-knowledge labor. The result is well-managed and well put-together graduate-level work, upon which life-changing policies are decided.

Policy recommendations

The Bank is clearly over-managed, in the sense that there are just too darn many layers of management and the corresponding politics and unproductive inoffensiveness is stifling. It is also overmanaged in the sense that it is run like a business with a series of financial goals for its clients, rather than an organization whose goal is to understand poverty and eradicate it.

Coauthoring is efficient. The latest Journal of Political Economy has seven articles, five of which were coauthored. To produce a good report about environmental issues in India, you can hire a full-time India-environment specialist, or you can hire a full time India specialist and a full time environment specialist. The team stands a good chance of producing a better paper and finishing in fewer person-hours than the lone author. Thus, the grid is redundant: a team of the best and the brightest for each country and for each topic, plus a nice coffee stand where then can all interact, may be sufficient, and can run with a fraction of the managerial overhead.

What do the best minds in academia want? [And let me make very clear that I am decidedly not referring to myself.] First, they want to feel that their work will actually go somewhere. Everybody understands that things go slowly and nobody's recommendations are magically implemented a week later. But it is disheartening to know that your careful and interesting work will be reduced to the politically least offensive denominator. For a report to be interesting to one's academic peers, it has to say something.

For Bank contractors, much of the pay is in the form of carrot-dangling about how maybe one day you'll have a real job at the Bank and at the least you'll have a recognized name on the resume, both of which are premia that only a recent graduate could love. The incentives to advance for the full-time employees are primarily about being able to control more money. The World Bank has no advancement path of any sort for a person who has experience and knowledge regarding poverty and economics.

Since there is no means of retaining those who are interested and capable regarding the intellectual challenge of fighting poverty, such people are guaranteed that the Bank will be an unpleasant place for them, and if they want a community of people thinking about poverty, they are better off elsewhere. The Bank could be the focal point where the smartest people in the world meet to tackle the world's hardest problem, but instead it is a house of bureaucracy, hiring lowest-bidder contractors to provide management solutions to the government officials who are its clients.

[link] [4 comments]
[Previous entry: "Invariants"]
[Next entry: "Transparency vs effectiveness"]

Replies: 4 comments