COMMENTARY

How to prime feds for better performance

One hot trend in corporate America is to use quick, real-time experiments in grocery stores, restaurants and hospitals to measure the impact of various interventions on customer behavior. A common example is testing a variety of website layouts and messages. Those accessing a site will randomly be shown different versions, and the site owner compares reactions (say, click-throughs) by the visitors to see which configurations get the best response.

It is time to introduce the idea of quick experiments into government. As with the example above, agencies could use that approach to test alternative website configurations. Or they could use it to test the impact of different versions of public service announcements, perhaps randomly shown in different TV markets, or other government initiatives designed to influence behavior.

I am especially interested, however, in a different use of experiments: to check the impact of various management interventions on the performance of employees or teams inside agencies.

Recently, I came across some fascinating academic papers by Gary Latham, a professor at the University of Toronto’s Rotman School of Management and the world’s leading researcher on the impact of goal-setting on human performance, and his colleague Amanda Shantz. They tested the impact of “priming” an unobtrusive performance-enhancing message on the performance of call-center employees.

Priming is a technique commonly used to set up social psychology experiments. Say the researcher wants to test the impact of power on behavior. It turns out to be possible to prime feelings of power or powerlessness among experimental subjects — in this case, by asking members of the experimental group to think about a situation in which they felt powerful, while asking members of the control group to think about a situation in which they felt powerless. Amazingly, for most of us, priming succeeds in evoking the feelings whose impact can then be tested in the lab.

Rather than using priming to set up the experiment, Shantz and Latham made it the experiment itself. They gave a group of call-center employees information about the cause and the organization for which they were being asked to raise money. For half the group, the background of the information sheets had a photo of an Olympic race winner crossing the finish line, arms raised toward the sky and face beaming. The control group’s information sheets had a plain background.

The result: The group that received the information sheets with the Olympic photo raised more money than the control group during a four-hour shift.

I cite this research partly because it raises some intriguing questions. For example, could we improve the performance of U.S. Patent and Trademark Office staff by putting pictures of great American inventions as their computer screen savers? More broadly, are there management interventions that could significantly enhance the public service motivation of civil servants? (Adam Grant, an associate professor at the Wharton School at the University of Pennsylvania, has done some fascinating experiments along those lines.) Furthermore, are there management interventions that could improve information sharing among employees or encourage cost-saving ideas?

I also raise this example to illustrate a broader point: A quick-experiment revolution is spreading fast in companies. That revolution needs to spread to government, too.

About the Author

Kelman is professor of public management at Harvard University’s Kennedy School of Government and former administrator of the Office of Federal Procurement Policy. Connect with him on Twitter: @kelmansteve

The 2014 Federal 100

Get to know the 100 women and men honored this year for going above and beyond in federal IT.

Reader comments

Thu, Oct 13, 2011

The best way to get performance from a Fed is to treat them like a scum-of-the-earth contractor.

Wed, Oct 12, 2011

One problem with experiments is that you never know how much is hype and how much is reality. Also, you need to weigh the benefit against the cost of the experiment. Yeah, these experimenters are quick to mention the supposed benefits but neglect to mention how much they are going to charge you for the experiment that may or may not work and may even backfire. Also, you need to include the continuing cost of whatever action you need to implement to get your improved performance. The only group guaranteed to come out ahead are the people hired to run the experiment.

Tue, Oct 11, 2011

Subliminal messages on my PC? OK - I still get paid right?

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above