The Cost of Rent-Seeking: Evidence from Cellular Telephone License Lotteries

Abstract

The inventors of the theory of rent-seeking used an analogy with competitive markets to conclude that all monopoly rents would be dissipated in the competition to obtain the property rights to them [23; 29]. Subsequent theorists showed that this conclusion is quite sensitive to assumptions about foresight, risk aversion, conjectures, and bias [15; 21; 30]. Since little quantitative information about actual rent-seeking competitions is available, the theory has developed in an empirical vacuum. In this paper, we attempt to fill that vacuum with an estimate of rent-seeking expenditure by applicants in the Federal Communications Commission's (FCC) cellular telephone license lotteries, held between 1986 and 1989. We go on to compare expenditures with estimates of the value of the licenses at stake. These lotteries constitute a unique natural experiment. Any U.S. citizen was eligible to enter each geographic lottery once, and only once, for an expense which we calculate as modest.' The lotteries were widely publicized and heavily entered. There were over 320,000 applications (one per market) filed for 643 cellular licenses, and third-party application preparers entered the market rapidly. Because licenses were awarded sequentially, prospective participants in later lotteries could find out that expected net returns from earlier lotteries were positive because of undersubscription. The value of the prize was known because sales of previously awarded licenses had begun to take place before later lotteries were held, and market analysts were using such sale prices to value licenses not yet awarded. The later lotteries seem almost optimally contrived to maximize the dissipation of rents. Yet comparing the resources invested in entry (the price of a fully-prepared application) with the rents available (the market value of cellular telephone licenses) shows that dissipation was far short of complete.