摘要：We conducted a randomized experiment at the time of the 2004 flu vaccine shortage, providing information about the sharply reduced number of clinics and their schedule, and an appeal on cooperative restraint to a campus population. This strategy was intended to reduce demand for vaccination among non-priority individuals and to free available supplies for the priority population. It failed to achieve its purpose. Information induced a net increase in vaccines distributed and, perversely, the net increase originated entirely in non-priority individuals. The surprising finding is that calls on cooperative restraint induced an uncalled for positive response among priority individuals, while they induced an increase in cheating among non-priority individuals. Age as a qualifying factor was in particular widely abused, with the number of “65 years old” more than twice the predicted value, while about half of the predicted 61-64 years old are missing.
