← Back

How would you test the hypothesis that transparency is a more effective means of control than top-down rules? That’s the question a group of midlevel managers in a global pharma company asked themselves during a workshop led by one of our colleagues.

The first step was to look for a Byzantine policy that was widely regarded as a pain in the ass. As you might suspect, they had plenty of candidates, but the company’s irksome travel policies seemed a particularly juicy target. In an attempt to rein in a corporate travel budget of roughly $500 million per year, the finance function had developed a maze of niggling rules. There were strict guidelines on who could travel, for what purposes, on which airlines, and in which class of service. Hotel and rental-car choices were similarly constrained…

The experiment, modeled on the company’s methodology for drug trials, involved two pairs of treatment and control groups—one pair at head office and the other in an operating unit. The experiment was designed to test the hypothesis that increased autonomy and transparency would (1) simplify travel planning, (2) reduce frustration, and (3) not raise costs. Fifty people were recruited for each group, for a sample of two hundred individuals. The treatment groups were told that for the next ninety days, they’d be able to make their own travel arrangements with no pre-trip authorizations or post-trip audits. The catch: all their travel expenses would be posted online for everyone to see.

At the end of the trial, the team analyzed the results. A large majority of those in the two treatment groups—74 percent and 87 percent— reported that the new process was less time consuming than the old one. What was more surprising was that 45 percent of the participants said the simple rule change had increased their overall job satisfaction. The researchers had expected travel costs to edge up slightly, and were prepared to argue this was a price worth paying for a more time-efficient process, but in the end, travel costs fell for the treatment groups while remaining essentially unchanged for the control groups.