Good empirical evidence about a phenomenon can help inform the debate about whether a management approach will improve government or not, Steve Kelman writes.
The great social psychologist Kurt Lewin once said, “There’s nothing so practical as a good theory.” I would add to this aphorism another one: “The plural of anecdote is data.”
I recently attended the biannual Public Management Research Conference at Ohio State University’s John Glenn School of Public Affairs, which featured papers by leading researchers in the United States and abroad who work on government management.
Nonacademics sometimes dismiss the relevance of scholarly research for the issues that practitioners face. As an academic who also cares about real-world problems, I (probably not surprisingly) disagree. Good empirical evidence about a phenomenon in which opinions outweigh evidence can help inform debate about whether a management approach will improve government.
For example, Sergio Fernandez of Indiana University presented a paper using data from the 2006 Federal Human Capital Survey. His research explored the features of an employee’s job that motivated the employee to seek better ways for the individual or the work group to do its job. The features most closely related to the motivation to innovate? Those would be perceptions that job gave the employee lots of discretion — as opposed to being bound mostly by rules — and that the organization took steps to help the employee develop knowledge and skills.
Those findings are not shocking, perhaps, but they provide evidence for the view that more empowered employees are more innovative ones. At the same time, there was no association between motivation to innovate and an employee’s participation in a pay-for-performance scheme.
In a paper based on data from British local governments, Richard Walker of the University of Hong Kong and Gene Brewer of the University of Georgia looked at the impact of different personnel flexibilities on organizational performance. The researchers found that the more employees believed that it was hard to remove poorly performing managers, the more negative the effect on performance. But the perception that it was hard to reward good managerial performance with higher pay did not have the same effect.
It is interesting that both papers saw no relationship between pay for performance and other performance-related variables — findings I take seriously, given my recent recommendation that government look at the evidence when considering pay-for-performance programs (“The case for pay for performance,” Federal Computer Week, Sept. 28).
Finally, a plenary session featured an interesting presentation by Brint Milward of the University of Arizona. Milward is a well-respected scholar on networks. He expressed concern about the systemic risks that contracting for mission-critical services can create for the government during the recent economic crisis. He used the example of a for-profit company that provides social services to local agencies. The company recently came close to going bankrupt, and Wall Street analysts recommended that it withdraw from its California contracts because of risks that the state would pay late and reduce fees.
What would have happened to social-services recipients if the company suddenly withdrew? And what would have happened to recipients around the country had the firm gone bankrupt? Good questions — and very practical ones.