The Results of Some Tests of Mechanism Designs for the Allocation and Pricing of Collections of Heterogeneous Items
During the discussion and evaluation of proposals for the design of the Federal Communications Commission (FCC) mechanism to sell the spectrum, over 130 auctions were run under controlled conditions at Caltech for the National Telecommunications and Information Administration (NTIA), the FCC and others. In this paper we look at these data and try to extract some useful findings for those who may be involved in creating future designs of similar auctions. For those whose experience with experimental economics methodology is limited, we begin with a section on the general framework within which experimental work underlying applied mechanism design is conducted. Next we cover, in section 2, the various technical pieces needed to understand the data: performance measures, economic environments, mechanisms tested, and the major issues considered. The experimental data are presented and our observations are summarized in section 3. We end, in section 4, with some thoughts for future work and with the observation that there is a huge gap between, theory, scientific evidence, and practice in the design of complex auctions. Much needed research remains to be done.
Some of the trials and the data generated are described in a report to the FCC. See Ledyard, Plott, and Porter (1994). For a discussion of the role of experimentation in the FCC design process, see Plott (1996). We would like to thank Robin Hanson for his design of the spatial environments. Published as Ledyard, J.O., Porter, D., & Rangel, A. (1997). Experiments testing multiobject allocation mechanisms. Journal of Economics & Management Strategy, 6(3), 639-675.
Submitted - sswp978.pdf