volved from wasting time and effort on case-by-case analysis, even at the cost of some small errors It makes sense to have rules, perhaps with waivers, when the outcome of nearly every case can be readily predicted on the basis of easily ascertainable facts. For example, thirty years ago only 2.7 percent of U.S. cities with a daily newspaper had more than one such newspaper. Similarly, most cities had no more than three significant commercial TV sta- tions. A merger between the newspaper and one of the TV stations in these cities would almost certainly have increased concentration significantly in local advertising markets however broadly defined. In those circumstances, a rule banning such cross-ownership was likely more efficient than repetitive case-by-case analysis. Today the relevant facts vary significantly across local markets, and a rule-based approach is no longer approprh When facts differ significantly from one transaction to another, a case-by-case approach is likely to be superior to a rule if it employs analytical tools that are well-defined and easy to understand. If prospective applicants understand these tools, they can model the agencys decision process. This increases predictability and, presumably, reduces the number of applications likely to be rejected. The Merger Guidelines are again a very use- ful model of such a tool. 9 In spite of the Guidelines state-of-the-art analytical framework, there might be pragmatic reasons to reject the use of traditional antitrust enforcement standards in media industries. Imagine, for example, that empirical studies by the Commission demonstrated significant adverse effects on the price of advertising in local media markets when HHI levels exceeded 800. That might justify the Commissions use of 800 rather than 1,000 or 1, 800 as a safe harbor, or it might justify an owner ship cap at 800, depending on the nature of the empirical findings. The Merger Guidelines stan- dards are of general applicability. Their numerical values, frankly, are arbitrary. Certainly they are not necessarily appropriate for every industry. As it happens, there is no evidence that a special standard is required for broadcasting and related industries 88 volved from wasting time and effort on case-by-case analysis, even at the cost of some small errors. It makes sense to have rules, perhaps with waivers, when the outcome of nearly every case can be readily predicted on the basis of easily ascertainable facts. For example, thirty years ago only 2.7 percent of U.S. cities with a daily newspaper had more than one such newspaper. Similarly, most cities had no more than three significant commercial TV sta￾tions. A merger between the newspaper and one of the TV stations in these cities would almost certainly have increased concentration significantly in local advertising markets, however broadly defined. In those circumstances, a rule banning such cross-ownership was likely more efficient than repetitive case-by-case analysis. Today, the relevant facts vary significantly across local markets, and a rule-based approach is no longer appropri￾ate. When facts differ significantly from one transaction to another, a case-by-case approach is likely to be superior to a rule if it employs analytical tools that are well-defined and easy to understand. If prospective applicants understand these tools, they can model the agency’s decision process. This increases predictability and, presumably, reduces the number of applications likely to be rejected. The Merger Guidelines are again a very use￾ful model of such a tool. 9 9 In spite of the Guidelines’ state-of-the-art analytical framework, there might be pragmatic reasons to reject the use of traditional antitrust enforcement standards in media industries. Imagine, for example, that empirical studies by the Commission demonstrated significant adverse effects on the price of advertising in local media markets when HHI levels exceeded 800. That might justify the Commission’s use of 800 rather than 1,000 or 1,800 as a safe harbor, or it might justify an owner￾ship cap at 800, depending on the nature of the empirical findings. The Merger Guidelines stan￾dards are of general applicability. Their numerical values, frankly, are arbitrary. Certainly they are not necessarily appropriate for every industry. As it happens, there is no evidence that a special standard is required for broadcasting and related industries