Building better jobs for feds

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.

I have written a great deal in the past few years about my worry that although the government has hired many talented young people, it is in danger of quickly losing them because it doesn’t give them jobs that are interesting enough.

My conversations with young people in the contracting field have made me especially sensitive to this issue in the context of that particular workforce, but the issue is broader than that. It also applies to other professional groups and to more experienced workers. And it isn’t just about retention, especially because the tough economy might discourage people from leaving the government, at least in the short term, while there aren’t a lot of jobs out there. It is also related to the quality of performance and job satisfaction of career civil servants.

I recently learned from a colleague about a 20-year-old research article (published in 1993 in the Journal of Applied Psychology) that provides some practical advice for federal managers trying to design jobs for greater employee satisfaction and productivity. The article reports on a two-year, real-life experiment among clerical employees at a financial services company who were processing paperwork for different operating units that sold various products.

Two kinds of changes were introduced in the information-coding jobs of employees. For some employees, the task was enlarged; coding and keying, previously two separate jobs, were combined into one. For others, the knowledge needed for the original job was increased; they still only coded, but they did so in a larger number of product areas using different sets of information, cost structures, etc. Both approaches had been recommended in the past as ways to make jobs more meaningful for employees.

For the next two years, researchers took measures of employee productivity, error rates and satisfaction. Their findings were striking. For employees whose task was enlarged, performance got worse. They made more errors and were less efficient than before. Their job satisfaction was lower and mental overload higher. By contrast, the opposite occurred for employees whose knowledge requirements were increased. Errors went down, mental overload declined, and satisfaction went up.

It’s hard to know to what extent those findings would apply to other kinds of jobs. The study used relatively low-skilled clerical jobs, not high-end professional ones. But they probably have a good deal of similarity to many jobs in the government, except those at the most senior levels.

That kind of study represents the best in academic research: a rigorous examination of a practical management question with real-world significance. The answer is not obvious. Indeed, many have argued on theoretical grounds that both approaches should improve productivity and/or satisfaction or that neither should.

I think the results are likely applicable to the government as well as to the private sector and that managers have at least some level of control about introducing changes in the suggested direction. If I were a supervisor or manager, I would think about how I might apply the results to help my employees.

The results don’t support the common view that it’s a good idea to combine several job responsibilities into one, although it might be appropriate when an artificial division of work creates bottlenecks along a sequential production line or makes employees feel less responsible for their work, neither of which was an issue in this experiment. But the results do support the view that, from both a productivity and satisfaction standpoint, we should be seeking to provide employees with challenges that require more knowledge so they have an opportunity to perform their current work better and with greater satisfaction.