Collusion can profitably be classified into three distinct types. In our classification, "Type I" collusion is the familiar direct agreement among colluding firms (a cartel) to raise prices or, equivalently, restrict output. Alternatively, firms can collude to disadvantage rivals in ways that causes those rivals to cut output. We term this "Type II" collusion. Its indirect effect is an increase in market prices.
A number of important collusion cases neither direct manipulation of prices or output, nor direct attacks on rivals. Examples include Supreme Court cases such as National Society of Professional Engineers v. US, Bates v. State Bar of Arizona, and FTC v. Indiana Federation of Dentists. In each of these cases, cartel members set prices and output independently. Their collusion shaped the rules under which the independent decisions of the colluding firms were made. Collusion permitted the cartel members to insulate themselves from each another, at least partially. Their newfound isolation provided benefits similar to those attainable from market power acquired in a more traditional fashion. By increasing the insulation of cartel members, each achieved the power and independence to raise its own price-the colluding firms competed on price, but their competition was rendered less vigorous than by collusion over rules.
Archetypal examples of this type of collusion include softening competition by limiting information available to consumers through direct restrictions on advertising. In this Article we explore a number of examples of previously unexplained or uncategorizable cartels that can be explained by this construct. We show that, together, they form a third general category of anticompetitive behavior that we term "Type III" collusion. With considerable enforcement activity directed at collusion of Types I and II, we believe that Type III collusion will prove increasing attractive to firms and, accordingly, a growing source of social welfare loss from collusion.