Last week I complained about the dearth of practical, policy-relevant literature available to help governments oversee contracts for the construction of civil works, the development of complex software programs, and other products which take months if not years to complete. This is but one of many examples where governments must navigate the procurement process without rigorous, empirically grounded work on what procedures to employ when and how. Absent such guidance, the procurement community falls back on rules of thumbs, old saws, and folk wisdom — the accuracy of which is always suspect.
One of the more suspicious sounding old saws in the procurement practice is the notion that contract execution and contract selection are independent activities — the belief, in other words, is that that how one selects a contractor is of little or no import for how well the contract is performed. But economic theory and recent empirical work both cast doubt on the accuracy of this bit of folk wisdom.
First of all, while belief that the contract selection process doesn’t affect how well the contract is performed might be true if performance monitoring were perfect, in practice monitoring is never perfect — both because hiring an army of monitors would be prohibitively expensive (and impractical), and because the monitors themselves might be corrupted (as I discussed in my last post). For that reason, the quality of contract performance is likely to depend not only on the terms of the contract, but also on the incentives of the contractor.
And that leads to the second point: when contractors are selected through a corrupted process, these contractors might not have the right incentives to do the job well. Consider, for example, bid rigging cartels, which the OECD reports are common in the construction industry. When such cartels operate, a government tender for a highway, bridge, or dam does not lead to vigorous competition among firms for the job. Rather, the colluders have already agreed who will win what job at what price. Why would a firm that knows it has a guaranteed share of the business work hard to meet the contract terms? What incentive does it have to perform? Indeed, the incentives are just the opposite – to lay back and do as little as possible.
Some recent empirical work substantiates the notion that when collusion or bribery impedes true competition in the contract selection process, contract performance also suffers. In an August 2013 paper, Lewis-Faupel, Neggers, Olken, and Pande report the effect of using e-procurement to award public works contracts in India and Indonesia. In both countries its introduction stimulated more competition for the projects, and greater competition in turn improved contract performance. In India the quality of the roads built after e-procurement was introduced was higher, in Indonesia more projects were completed on time post e-procurement.
Decarolis offers even stronger evidence that the method of contract selection cannot be divorced from that of contract monitoring. Using Italian data on the introduction of competitive bidding for public works at different times in two provinces, he shows that where monitoring institutions are weak, the benefit in terms of lower price from the introduction of competition is lost thanks to a large deterioration in contract performance. (Parenthetically, these results, as he notes, calls into question the procurement policy reforms the World Bank and other donor agencies have urged developing countries to embrace.)
Two papers alone may not slay the claim that the quality of contract execution is independent of the process of contract selection. But perhaps now the burden of proof should lie with those who believe they can assure the performance of a contract irrespective of how the contractor was selected.