ABSTRACT
Are business cycles driven by expectations? Models that posit that they do are hard to test empirically due to difficulties associated with measuring expectations and observing random coordination devices (sunspots). I propose an experiment to capture the salient features of Howitt and McAfee’s model of expectationally driven business cycles and investigate whether such cycles arise in an experimental setting. I introduce externalities, uncertainty, and an extrinsic random variable to mimic the crucial features of the model. The results from a pilot session provide preliminary support for the hypothesis that given the right incentives for coordination, expectations-driven cycles do indeed occur.
