The formation of expectations: Competing theories and new evidence

•We horserace three simple micro-models of expectations, effort, and utility.•The standard rational expectations model of expectations is rejected.•Our data strongly support a direct utility contribution of expectations.•Self-esteem may play a role in the link between effort and expectations.

Agents form expectations about the future in many markets, and these expectations drive investment and consumption behavior, inform entry and exit choices, and can even provide direct satisfaction or distress. How agents form expectations is therefore of central interest to economists. This paper reviews three competing theories and then provides empirical evidence about what elements make up an agent's expectation about an outcome that is person-specific, susceptible to influence by the agent's own actions, and fairly well predictable on the basis of historical precedent. We examine repeated cross-sections of hundreds of undergraduate students’ expectations at mid-semester of their own final course grades at two Australian universities. Data on actual and expected grades are exploited, as well as demographic and psychological information, course information, and data on students’ academic background, effort levels, happiness, and historical progress at university prior to expectation formation. Results indicate that a simple rational expectations model of expectation formation provides a poorer fit to the data than a utility model involving either direct utility benefits from expectations, or a psychological need to hold high expectations.