Groups Are More Cautious Decision-Makers Than Individuals

01 Jan 2008

Groups are more reluctant than individuals to take on high-risk projects. That is the central finding of new experimental research by Professors Robert Shupp and Arlington Williams, published in the January 2008 issue of The Economic Journal.

The researchers asked both individuals and small groups to make similar decisions about how much they would be willing to pay to invest in various lotteries each of which had a different chance of winning. Their results indicate that groups are often less willing to take on a risky project than any of the members of the group would have done individually.

As anyone who frequently attends meetings knows, many decisions involving uncertain outcomes are made through group interaction. Chief executives collaborate with boards of directors about future business directions, sales teams meet to develop a sales strategy and couples get together to decide how to save for their children''s university education or to plan for retirement.

It is not unreasonable to expect that the outcomes of these decisions would have been different if the chief executive, the sales team leader, or just one spouse made the decision unilaterally. Professors Shupp and Williams'' study attempts to assess whether this is indeed the case.

In their experiment, individuals were given $20 each and were asked how much of the $20 they would be willing to pay to participate in nine different lotteries. If they won, individuals would win an additional $20, but each lottery had different odds of winning, ranging from a low of 10% to a high of 90%.

Three-person groups were presented with an equivalent situation: they were given $60 and asked how much they, as a group, would be willing to pay to participate in lotteries with a 10% to 90% chance of winning an additional $60.

In each case, one lottery was randomly chosen and then a purchase price was randomly chosen for this lottery between $0 and the payoff of a winning lottery ($20 or $60). Individuals or groups with willingness-to-pay bids greater than or equal to the randomly chosen price purchased the right to play the lottery and the purchase price was deducted from the initial $20 or $60.

Individual participants got to keep any of the initial $20 left plus any winnings from the chosen lottery. Groups split their earnings (what was left of the $60 plus any winnings) evenly.

The experimental results suggest that groups do make different decisions relative to individuals, but that the differences are not consistent across all levels of riskiness:

  • For the higher-risk lotteries, where the chance of winning the prize was 40% or less, groups made decisions that indicated they were much less willing to take a risk than individuals.
  • For the very low-risk lotteries, the result was reversed. When the chance of winning was high (80% or 90%) groups made decisions that indicated they were more willing to take a risk than individuals.

After having found these differences, the researchers ran further experiments with cash incentives in an effort to begin to understand how group decisions compare to the individual decisions of group members. In these experiments participants first made bid decisions as individuals, then as a part of a three-person group.

The researchers found evidence that in the higher-risk lotteries, group discussion not only led to bids that were lower than that of the average individual in the group, but sometimes led to bids that were lower than what would have been the lowest bid for any one individual in the group.

Many current economic models implicitly assume that groups and individuals will behave identically when faced with making a decision involving risk. Gaining insight into behavioural differences between group and individual decision-making processes could help better explain behaviour in everyday situations.

''Risk Preference Differentials of Small Groups of Individuals'' by Robert Shupp and Arlington Williams is published in the January 2008 issue of The Economic Journal.
 

Robert Shupp

Michigan State University | +1 517 432 2754 | shupprob@anr.msu.edu

Arlington Williams

Indiana University | +1 812 855 4564 | williama@indiana.edu