The antitrust agencies regularly forecast whether prospective mergers are likely to be anticompetitive. Assessment of their success in this mission is complicated by two selection problems. First, it is difficult to analyze whether a blocked merger would have benefited consumers, since no post-merger activity is ever observed. Similarly, since the antitrust authorities try to prevent (or modify) the most egregious transactions, relatively few plausibly anticompetitive mergers are actually completed. For these reasons, evidence regarding the competitive effects of consummated mergers is relatively scarce (Vita and Sacher, 2001). This shortcoming limits evaluation of whether prospective merger analysis provides a reliable estimate of a transaction’s future competitive impact.