The "Politics of Enforcement" model model as estimated for 1983-96 requires the following parameters entered as discussed below.

• The first variable inputs an index that summarizes the analyses presented by the Bureau of Competition (BC) to the Commission. The value ranges from 0-3, depending on how many of the following three questions are answered in the affirmative by BC. The three questions are: Is the market concentrated (HHI above 2000 with change of over 100 points)?; Do barriers to entry exist?; and Does some theory of competitive concern (either unilateral or collusive) appear supported in the market under review? Abstracting from leaks, this information will not be known to the analyst, thus the application of the model should take a broad view of the potential analyses under the Merger Guidelines, undertake a binary (yes or no) analysis of concentration, barriers (impediments) to entry and competitive theories of concern and then sum the implicit 0-1 (0-No, 1-Yes) variables to compute an index that ranges from 0 to 3.

• The second variable inputs an index that summarizes the analyses presented by the Bureau of Economics (BE) to the Commission. The value ranges from 0-3, depending on how many of the following three questions are answered in the affirmative by BE. The three questions are: Is the market concentrated (HHI above 2000 with change of over 100 points)?; Do barriers to entry exist?; and Does some theory of competitive concern (either unilateral or collusive) appear supported in the market under review? Abstracting from leaks, this information will not be known to the analyst, thus the application of the model should take a narrow view of the potential analyses under the Merger Guidelines, undertake a binary (yes or no) analysis of concentration, barriers (impediments) to entry and competitive theories of concern and then sum the implicit 0-1 (0-No, 1-Yes) variables to compute an index that ranges from 0 to 3.

• The third variable identifies an efficiency defense. Two values are possible, although the actual FTC analysis on the issue should be unknown. Use 1 when the facts suggest the merger will generate merger-specific efficiencies in the area(s) of competitive concern and 0 when the no efficiencies exist or the efficiencies are not linked to merger. 0 should be used if the efficiencies exist only in markets unrelated to the competitive concern, unless these savings follow directly from the merger.

• The fourth variable defines the value of the transaction in nominal millions of dollars (the computer corrects for inflation). Thus a 1 billion dollar deal would be entered as 1000.

• The fifth variable is a dummy for the Clinton administration. Use 1 if Clinton is President. It would appear appropriate to use 1 if a democrat follows Clinton. The index should have the value 0 if a Republican is President. Note one of the final two political variables MUST be zero.

• The final variable defines the average percentage of Democrats in House and Senate under a Republican administration. A simple average is used so a House percentage of 51 percent and a Senate percentage of 49 percent would generate a value of 50. Independents are counted as members of the party they caucus with. NOTE use of this model when a Republican President oversees a Republican Congress (percentage less than 50) involves a prediction out of the historical sample and may generate misleading results. EXTREME CAUTION SHOULD BE USED IN THIS ANAYLSIS. Note again, both the President and Congress Percentage cannot BOTH take on values above 0.

The user can compute approximate probabilities with and without political considerations. (The software numerically integrates the normal function and thus the probabilities are not exact)