Matthew O’Keefe

We develop an empirical model of auctions in which bidders have non-expected utility preferences described by Yaari (1987)’s dual utility theory. Yaari’s dual utility admits risk-aversion but remains linear in wealth, offering an appealing framework for auctions in which bidders are profit-maximizing firms. We show that identification is analogous to identification in models of expected utility maximization (Guerre et al., 2009), while estimation and equilibrium computation are comparatively simpler. Applying our estimation method to bidding data from United States Forest Service (USFS) timber sales, we find evidence of bidder risk aversion that we confirm by testing a shape restriction. We use the resulting estimates to quantify the expected increase in USFS revenue from adopting an optimal reserve price.

Wastewater reuse in the shale gas industry reduces firms' private costs and mitigates many of the local environmental harms associated with fracking. Most reuse occurs within the firm boundary, but rival operators often exchange (or “share”) wastewater prior to reuse. I study the effect of wastewater sharing on reuse rates, transportation efficiency, and other outcomes. I find substantial private benefits from wastewater sharing, and modest external benefits. However, these benefits are diminished by the presence of large transaction costs. I explore the sources of these transaction costs and consider potential policy interventions to improve sharing markets and incentivize reuse.