The 11 essays, all save one newly commissioned for this work, largely represent extensions of previous work by the authors. As one title notes, cartelization or collusion is “a practice without defenders.” The hostile attitude of the economics profession was established by Adam Smith's observation (itself cited in several of the essays) that collusion is “… a conspiracy against the public.” In Common Law countries historically, and in European countries since the advent of the Treaty of Rome, this hostility has inspired legal sanctions against collusion, with significant implications for the scientific study of collusion.

The first consequence is that most collusive activity will be covert, resulting in a lack of reliable statistical data from which to draw valid inferences. A related problem is that only a subset of collusive activities will be detected. (This provides another answer to the question, “When is collusion successful?” which is a topic in several essays, i.e., “When it is undetected by authorities!”) The question then is how representative are the industries where attempted collusion has been detected? Uncertainty about this limits the lessons that can be drawn from case studies of such identified colluding industries.

Another consequence of legal hostility towards collusion is that “exceptional circumstances” are often discovered to apply to specific industries. They frequently result in conscious exemption from, or at least, nonenforcement of antitrust strictures. Once again, although this may create a larger data “footprint,” it calls into question whether observed behaviors may be generalized. The point is particularly relevant here, because at least half of the papers examine such special cases. Except for the brief introduction and an initial paper reviewing prior empirical work, the balance of the studies consider periods prior to the development of modern legal doctrines regarding collusion.

The essays fit comfortably into the traditional Structure-Conduct-Performance tradition of Industrial Organization. This is largely a consequence of the data limitations discussed previously, which make it impossible to estimate models employing more modern structural specifications. However, the contribution of these papers may be greater than either the authors or the editor of this volume realize.

Start with the third element of the structuralist formula, Performance. Janice Rye Kinghorn and Randall Nielson's “A Practice without Defenders: The Price Effects of Cartelization” examines what, according to economic theory, ought to be the chief effect of collusion, increased prices. Their method is to compare the relative price performance of cartelized German industries with that of their British counterparts. They first demonstrate that prior to the formation of cartels German coal and steel prices closely tracked their British equivalents. After the adoption of a cartel German prices became a fraction (and a decreasing fraction at that) of British prices. Although one may have reservations about the result—perhaps British firms colluded more effectively covertly than German ones did openly or some tariff or nontariff barrier in international trade prevented equilibration—it is striking and thoroughly counter-intuitive for most economists.

In the same performance-related vein, Andrew Dick's, “If Cartels Were Legal, When Would Firms Fix Prices?” notes that the Webb-Pomerene Act created a broad exemption in United States antitrust laws for exporters. Despite the exemption granted from 1919 to 1965, as Dick observes (p. 151), “… the most noteworthy observations is that relatively few industries ever made use of the antitrust exemption. In a typical year, only 30 to 35 cartels were active, and collectively these cartels accounted for between 5 and 10 percent of US exports.” When collusion promises such great rewards and the U.S. government has promised to look the other way, why were there so few cartels?

These puzzles are addressed in Dick's and other contributions to the book. They largely involve questions of the sorts of market structures that facilitate the establishment of cartels. Dick's essay uses sets of logit regressions matching 125 industries that established Webb-Pomerene cartels with 125 noncartelizing export industries. Many of the variables Dick identifies represent the “usual suspects” in investigations of collusion, i.e., high concentration, barriers to entry, and nondurable and undifferentiated goods. The notable extension is that regionally distributed industries are more likely to cartelize, and in this case the cartel will usually take the form of a joint sales agency. Dick speculates that the economies achievable through joint export marketing programs for regionalized industries were responsible for their cartelization.

This speculation finds support in Peter Grossman's, “Why One Cartel Fails and Another Endures,” in which he examines the comparative cases of the Joint Executive Committee and Railway Express. The Joint Executive Committee was an attempt to cartelize freight carriage on the important Chicago-New York rail routes. It failed, besieged by cheating and repeated entry. In contrast, the Railway Express cartel, performing a function similar to FedEx today, proved extremely durable and successful. Grossman explains the performance differences by the fact that, in addition to the obvious price rise achieved by the express cartel, membership also permitted member firms to realize substantial economies. Cheating both lowered revenues and raised costs, making firms reluctant to exit the cartel.

Two other contributions highlight additional reasons for cartelization beyond the traditional achievement of monopoly prices. The first is Christopher Gilbert's, “International Commodity Agreements as Internationally Sanctioned Cartels.” This presents a plausible model of situations where governmental cooperation with a cartel may actually be welfare enhancing. William Sjostrom's “The Stability of Ocean Shipping Cartels” suggests that, given integer problems in the supply of ocean shipping, there may be no sustainable competitive outcome in these markets. Both provide case study evidence supporting their models.

In summation, this is an interesting and useful volume (including other contributions not discussed here due to space constraints) that could have been even better with some additional editorial organization and commentary.