In Public Service, Dishonesty Is Not the Best Policy


Public Management

In Public Service, Dishonesty Is Not the Best Policy

Rema Hanna Faculty ResearcherRema Hanna, Associate Professor of Public Policy, Harvard Kennedy SchoolPaper Title
Dishonesty and Selection into Public ServiceCoauthor
Shi-Yi Wang, The Wharton School, University of Pennsylvania

Corruption, in some form or another, can most likely be found in all the governments of the world, though it may be especially pronounced in developing countries. Given the pervasiveness of this problem, economists and social scientists have made an extensive study of lapses in the system, along with financial incentives, which may lead civil servants to improprieties. But what about the people who choose to enter public service? Might those with a predisposition toward misconduct be disproportionately attracted to government jobs?

That question was taken up by Rema Hanna, Associate Professor of Public Policy at Harvard Kennedy School, and Shi-Yi Wang, of the Wharton School. To examine corrupt behavior among potential public service employees, Hanna and Wang devised a series of experiments, which they administered to volunteers—669 college seniors in India and 165 Indian civil servants. The students were drawn from seven large universities in Bangalore and were majoring in disciplines that could lead to jobs in either government or the private sector. All the civil servants were government nurses who worked in five districts in Karnataka, the state in southern India that includes Bangalore.

The crux of the study involved tests designed to gauge the subjects’ honesty and social tendencies. During the so-called dice task, students were instructed to roll a pair of ordinary dice 42 times and record the numbers obtained with each roll. The higher the numbers they wrote down, the more money they would receive in the end. A 6, for example, would earn them three times more money than a 2. Subjects had complete privacy during the exercise and were thus on the honor system; their conduct, however, generally fell far short of honorable.

 “Dishonesty, as measured by the dice task, is rampant,” Hanna and Wang observe. Thirty-four percent of the students reported scores so high that the odds of getting them were one in a hundred, “very far away from what they would have gotten purely by chance,” notes Hanna. Surveys showed that students who cheated on the dice game were more likely (by 6.3 percent) to want a government job.

The “pro-social preferences game” measured willingness to contribute to the public good. Participants were allotted 50 rupees, which they could keep for themselves or give to well-known charities such as UNICEF, the Red Cross, and Save the Children. For every rupee they donated, the amount given to charity was doubled. All told, the students chose to keep most of the money—about 59 percent—for themselves. A majority of students who scored low in pro-social behavior (like those who cheated on the dice task) expressed a preference for government jobs. As Hanna puts it, “Those who kept more money for themselves were more likely to cheat on the dice game, and they were also more likely to want a government job, which is kind of scary.”

The participation of the government nurses gave Hanna and Wang a means of tying their experimental findings to corruption in the real world. The researchers looked at a specific form of corruption—absenteeism, or billing for hours not worked, which squanders public funds and in this case can also have adverse public health consequences. The nurses tended to cheat on the dice task (although to a lesser degree than the students), and those who claimed scores above the median value, the authors wrote, “were 7.1 percent more likely to be fraudulently absent than those below it.”

Hanna believes that their findings could point toward more-effective methods for screening out dishonest job applicants. “If we believe that corruption is a key constraint to service delivery, it may be important to look at characteristics other than ability when screening for bureaucrats,” she says. At the same time, government agencies should make it harder for their employees to collect ill-gotten gains so that those prone to corruption will think twice about going into government work.

- by Steve Nadis

Surveys showed that students who cheated on the dice game were more likely to want a government job.

John F. Kennedy School of Government 79 John F. Kennedy Street
Cambridge, MA 02138
617-495-1100 Get Directions Visit Contact Page