The bank employs almost 10,000 people. But only 77 now work full-time as researchers in its economics department, and the headcount has fallen in recent years. (About the same amount of research manpower is also scattered throughout the banks other divisions and regions.) They produce hundreds of studies a yearfrom panoramic surveys of global inequality to close examinations of pesticide poisoning. A number of these papers make it into the top academic journals; some are put between glossy covers in the banks armada of flagship reports; more collect dust on the shelf.
Some of their work is wonkishly heroic. While others emote about the worlds poor, the banks researchers try to count them, despite all the formidable conceptual challenges that poses. Its project to gauge purchasing power around the globe is now reckoned to be the worlds biggest statistical initiative. Economists could say little authoritatively about world growth or poverty without it.
The banks inspectors were suitably impressed by this invaluable spadework. They would, indeed, like to see a lot more of it. Too much of the banks work, they complained, tried to do what they, as academics, could do better. It showed just enough technical aplomb to make it into a second- or third-tier journal, but no other discernible purpose.
But if they were withering about this long tail of inconsequential work, they were much ruder about the way more striking research findings were used by the banks top brass. They accuse the leadership of taking new and untested results as hard evidence that its preferred policies work; and of then using this research to proselytise on behalf of bank policy. This has a cost, the inspectors point out. Placing fragile selected new research results on a pedestal invites later recrimination that undermines the credibility and usefulness of all bank research.
In the main, the academics levelled this charge at the banks big-think studies of globalisation. Comparing scores of countries in one statistical sweep, these studies concluded that the poor do not get left behind by growth; that poverty falls when tariffs are cut; and that aid works, but only if the governments that receive it pursue sound policies. These results, especially the last, have been enthusiastically promoted by the banks leaders. The underlying propositions are probably true.
It is inconvenient, then, that the results tend to fall off their pedestals when prodded or poked a little. In 1997, for example, Craig Burnside and David
Dollar published a paper showing that aid results in faster growth, but only in countries with stable prices, lots of trade and prudent budgets. The paper eventually appeared in the American Economic Review, as well as in a bigger research report the following year on Assessing Aid. This newspaper welcomed the work, and its message helped to fire the new enthusiasm for aid shown at the UNs Monterrey Conference in 2002.
Unfortunately, the study, like many in the genre, proved fragile. Critics have shown that adding four more years to the sample (and plugging some other gaps) undermined the result. Using more plausible measures of aid and policies had the same effect. And the paper did not convincingly disentangle cause and effect: perhaps faster growth attracts more aid money from donors keen to back a winner.
According to Franois Bourguignon, the banks chief economist since 2003, Assessing Aid was championed by the banks leaders because it was convenient. The work confirmed a plausible intuition they already held. It was a very good piece of research the banks management could use for its advocacy and it was difficult to resist.
This temptation was easy to understand. The bank spent much of the 1990s under attack from critics on the left and right, claiming that trade was akin to imperialism, growth bypassed the poor, or that aid was money down a rat hole. Unsurprisingly, it sent in its best economists to lance its detractors.
These battles, and the economic weapons used in them, have since moved on. None of the bright young things in development economics today, several of whom contributed to this evaluation, do big cross-country studies. They show more interest in absenteeism among doctors, intestinal worms in children or flip charts in schools than in the grand questions of trade, aid and growth. And the advice economists give to countries aspiring to prosperity no longer comes off-the-rack. It is more tailored to the idiosyncrasies of time and place.
True to this intellectual fashion, the evaluators urge the banks economists to immerse themselves in the banks work in the field. They also call on them to unbutton their lips, crying foul if their superiors over-egg (or ignore) their results. The best way for the bank to disarm its dafter critics is to replace them with more sophisticated ones. That is a role its shrinking band of economists should be encouraged to play.
* The evaluation and responses are available at tinyurl.com/yck7wc.
The Economist Newspaper Limited 2007