PITTSBURGH -- Some of the talk about making people and organizations more accountable may amount to just that -- talk with little real payoff, a Carnegie Mellon University study shows.
That's the result of the first comprehensive review of what a new push toward accountability on the job can mean and where it is likely to go, says Jennifer Lerner, an assistant professor of social and decision sciences and psychology at Carnegie Mellon.
In recent years, increased accountability has been proposed as a solution for everything from failing schools, the rising costs of health care and unresponsive politicians. Accountability, or responsibility to others, is widely regarded as a positive thing. Most of us believe that requiring decision-makers to be accountable will lead them to give more thought to the consequences of their actions. And, we hope that giving more thought to the consequences of those decisions will lead to improvements in overall performance.
However, a comprehensive analysis of accountability by scientists at Carnegie Mellon and Ohio State University finds the reality is not that simple. Lerner, writing with Philip Tetlock, the Harold Burtt professor of psychology and political science at Ohio State University, concludes that only in special cases does the feeling of "being accountable" increase a person's attention and effort. Researchers also find that paying more attention and giving more effort to being accountable doesn't necessarily help. In fact, it sometimes makes matters worse. Their article assessing accountability appeared in a recent issue of the Psychological Bulletin, a publication of the American Psychological Association.
In their study, Lerner and Tetlock point out that there are actually many different kinds of accountability. This is important because different kinds of accountability have different effects on judgment and decision making -- some improving, some having no effect and some degrading judgment and decisions. For example, accountability to an audience whose views are known elicits less effort and more biased judgments than accountability to an audience whose views are unknown.
In everyday life, this means that bosses who want their subordinates to engage in even handed and effortful evaluations of alternatives should refrain from revealing their own views until all opinions are heard.
Lerner and Tetlock also discovered that two of the most important factors in accountability are the pressures to make good decisions and the timing of those decisions.
People who are on the spot to make good decisions will often factor in irrelevant information more heavily. Because they feel under pressure to perform, Lerner and Tetlock said decision-makers are more likely to go on "wild goose chases" that use readily available but generally irrelevant information.
"The reason is that a desire to avoid appearing foolish in front of an audience activates a tendency among accountable decision-makers to use all available information," Lerner says. "If some of the available information is biased or irrelevant, but not obviously so, then predictions from accountable decision-makers are even worse."
Whether someone learns before or after making a decision that they will be held accountable makes a big difference on the outcome. In business, for example, managers who knew they would be accountable for decisions from the start felt more comfortable engaging in brainstorming and free thinking. But if they publicly committed to a position and then were held accountable, the managers tended to justify their actions and dig in their heels, even if it meant increasing commitment to a failing course of action.
Lerner and Tetlock say there's a message in empirical research about accountability -- a message that does not support the view of accountability as a social panacea.
"Simply making someone accountable will not necessarily help. Effects depend on the type of accountability, the type of task and the characteristics of the decision-maker," Lerner said. "Perhaps policy planners and mangers can now draw on this comprehensive analysis of accountability research and design accountability systems that help, rather than hinder, performance."