Why contests are a smart procurement tool

Steve Kelman is professor of public management at Harvard University’s Kennedy School of Government and former administrator of the Office of Federal Procurement Policy.

Compared to many of my colleagues, I am more inclined to read academic journals from cover to cover rather than just the one or two papers that are directly related to my areas of research. This has the virtue of exposing me to ideas off my beaten track, which is good for creativity, but at the cost of my often being way behind on the latest issues.

I say this as an excuse for reporting only now on a paper in the September/October 2010 issue of Organization Science. Fortunately, academic papers have a longer shelf life than, say, a daily newspaper. And this article was worth the wait.

It has the forbidding title “Marginality and Problem-Solving Effectiveness in Broadcast Search.” The authors are Lars Bo Jeppesen of Copenhagen Business School and Karim Lakhani of Harvard Business School. By “broadcast search,” the authors mean what in government jargon is called “contests” — that is, situations in which an agency announces a problem and offers a prize for the first or best solution.

The paper’s starting point is an observation from what was probably the world’s first government-sponsored contest and perhaps the best-known one, due to its popularization in Dava Sobel’s 1996 book “Longitude”: the prize offered by the British government in the 1700s to anyone who could solve the vexing problem of determining longitude at sea. The paper notes that Sir Isaac Newton, who served on the board of scientists that reviewed the entries, predicted that the solution would need to be based on astronomical science. However, the eventual winner was a largely self-taught carpenter and clockmaker, John Harrison, who developed a chronometer suitable for the task by coming up with a design that differed from the clockmaking establishment’s typical approach.

The paper also cites research showing that the scientists responsible for major innovations in medicine and molecular biology have tended to be marginal players.

With that in mind, Jeppesen and Lakhani reviewed the winners of a sample of contests launched on the website InnoCentive.com and found two results. The first is that the more the solution submitters characterized the problem they were trying to solve as at the boundary of or outside their field of expertise, the more likely they were to win the contest. The second was that women were more likely to win than men.

The researchers suggest a common explanation for those findings: Outsiders have “a useful ignorance of prevailing assumptions and theories.” Furthermore, the greater the number of different perspectives and toolkits that are applied to a problem, the greater the likelihood that one of them will work. Indeed, in many cases, organizations conduct a contest only after trying to solve the problem internally using conventional toolkits.

The greater success rate for women, the researchers say, relates to the enforced marginality of many women in science, which produces the positive side effect that women scientists might be less bound by conventional wisdom.

The findings suggest another reason why contests might be a good procurement tool for government. Because contests do not require an understanding of the government procurement system, they are more likely to bring in players other than the usual suspects, and such players might be better able to solve the government’s problems. Therefore, the advantage of using contests might be even greater for government than for the private firms studied in this research.