An Experimental Study of the Nature of Consumer Expectations

Although important both theoretically and practically, the nature of consumer economic expectation formation has been little studied, particularly by psychologists. The most relevant previous research suggests that expectations are based on a heuristic that results in them being significantly biased. Further, relevant indicator series are poorly utilized. However, this earlier research used a task lacking in potentially important features of the real world, and this may have impaired performance. In the current experiment, participants received a more ecologically-valid task. Although there was still evidence of heuristic use, leading to suboptimal performance and bias, this performance was significantly better than anticipated from previous research, particularly regarding use of indicator series. However, when a strong trend in the criterion series allowed accurate forecasting without consideration of indicators, they were little used. I conclude that expectations are formed by first extrapolating the criterion series and only if that works poorly is other relevant information considered. Thus consumers appear to trade-off accuracy against effort, such that more effort is expended only when some threshold of acceptable performance fails to be reached.