Governance
Vol. 29, Issue 3
July 2016
Abstract
In addition to difficulties gathering and evaluating complete information, cognitive limitations and biases preclude individuals from making fully value-maximizing choices when making decisions. It has been suggested that, done properly, involving advisors can compensate for individual-level limitations. However, the “groupthink” tradition has highlighted ways group-aided decision making can fail to live up to its potential. Out of this literature has emerged a paradigm Janis calls “vigilant problem-solving.” For this article, we interviewed 20 heads of subcabinet-level organizations in the U.S. federal government, asking questions about how they made important decisions. Ten were nominated by “good-government” experts, 10 chosen at random. We wanted to see whether there were differences in how members of those two groups made decisions, specifically, to what extent executives in the two categories used a “vigilant” process. We found, however, that similarities between the two groups overwhelmed differences: As best as we were able to measure, decision making by U.S. subcabinet executives tracks vigilant decision making recommendations fairly closely. The similarity reflects a common style of senior-level decision making, which we theorize grows out of government bureaucracy's methodical culture. We did, however, develop evidence for a difference between outstanding executives and others on another dimension of decision making style. Outstanding executives valued decision making decisiveness—“bias for action”—more than the comparison group. Perhaps, then, what distinguishes outstanding executives from others is not vigilance but decisiveness. Contrary to the implications of the groupthink literature, the danger in government may be “paralysis by analysis” as much or more than groupthink.
Citation
Kelman, Steven, Ronald Sanders, and Gayatri Pandit. "“Tell It Like It Is”: Decision Making, Groupthink, and Decisiveness among U.S. Federal Subcabinet Executives." Governance 29.3 (July 2016).