What Can Contract Data Tell You about World Bank Projects?

Yesterday in a blog about the World Bank and open contracting, I mentioned the bank had put out  more data on contracts that it finances.  The covered contracts are those that were large enough for World Bank procurement procedures to mandate “prior review” by bank staff before they were awarded, a designation that covers the considerable majority of bank-financed contract value. 

You can learn a lot about the World Bank by looking at this data—some of which should be interesting to those attending the spring meetings this weekend. Take three examples of what I take away from the data:

Local firms are winning contracts

Together with Sarah Dykstra, in July last year I looked at the data back to 2007.  Across time, nearly 60 percent of all contracts are won by local firms (based in the borrower country), which is interesting if you believe the model of the World Bank is to provide scarce foreign currency so that countries can afford to bring in the skills and goods from abroad they need for high-return investments.  (Although note: local subsidiaries of international companies are listed as local firms).

There have also been some strong trends regarding which foreign firms are winning contracts.  In 2007, firms based in the US, UK, Germany, France, and Japan accounted for 18 percent of the value of contracts in the database between them.  Excluding contracts awarded in China, Chinese firms won less than 1 percent of the value of World Bank contracts in that year.  By the first half of 2013, those proportions had dramatically changed: US, UK, German, French, and Japanese firms between them accounted for 6 percent of the value of contracts awarded compared to 10 percent for China-based firms operating outside of China.

Procurement in some countries is concentrated among a few firms

From data on contracts in the civil-works category with transport as the major sector, we calculated a measure of market concentration known as the Herfindahl-Hirschman Index (the sum of the square of the percentage market share of firms divided by 10,000) for each country with ten or more such contracts. While an HHI of 1 represents a complete monopoly, anything above 0.25 is considered highly concentrated—when high concentration is associated with a poorly functioning market.

For World Bank–financed civil-works contracts in transport, the following countries have an HHI of above 0.25: Guatemala, Jordan, Bolivia, Burundi, Ghana, Ukraine, Sierra Leone, Madagascar, Macedonia, Serbia, Pakistan, Liberia, Senegal, and Honduras.  To be fair, in a lot of these cases there’s one really large contract that dwarfs all of the others, which guarantees a high concentration score.  And it isn’t clear that “World Bank–financed civil-works contracts in transport” is a well-defined market you’d be worried about concentration in.  Still, HHI might be an interesting filter to focus on where procurement might be working better.

Bank staff spends a lot of time monitoring low-value contracts

In 2012, the bank did prior review on 12,130 contracts.  The average value of those contracts was a little under $1.2 million, and 86 percent of all contracts had a value of under $1 million.  In short, bank staff members are spending a lot of their time monitoring low-value procurements, approving terms of reference and bidding processes and award decisions and contracts.  Wouldn’t it be wonderful if more of that time could be used to focus on designing better projects—and monitoring project results at the other end?

There are lots of other things you could study—especially with access to more data like how many companies bid for the contract and their bid prices.  So there’s a real value in putting this contract data in the public domain, and the World Bank should be applauded for doing it—and encouraged to do more.