This insightful and carefully researched work by James Raymond Vreeland should end once and for all the debate on whether participation in an International Monetary Fund program tends to improve or depress economic growth. The answer is a resounding no in that, on average, economic growth is roughly 1.5% slower when countries are under an IMF program than otherwise, according to Vreeland. This finding is consistent with several studies—mostly written by authors on the Left—although he takes pains to leave ideology at the door. Rather, he puts together the most systematic and exhaustive study of the subject to date in order to show how the same finding can be arrived at with methods that are qualitative and quantitative, empirical and abstract. Although the result tends toward repetition, there is something here for everyone.

The basic argument of the book is that previous studies of the effects of IMF programs on economic growth have been hampered by poor methods. Vreeland takes aim at conventional approaches used in the past: pre-test/post-test, with/without controlling variables, comparative case studies, and so forth. This said, he utilizes some of these methods himself, but only as a means to an end. His goal is to construct a story of why governments and the IMF enter into contractual arrangements, especially since the result is often disappointing to both in purely economic terms.

Ultimately, he argues that governments see the IMF as a tool for dealing with domestic opposition to what are seen to be necessary—though painful—reforms. This insight is nothing especially new, since it has been noted by IMF scholars going back to the 1970s. He furthermore argues that entering into IMF agreements tends to weaken the incomes of the working class—again an insight already found in work by Manuel Pastor in the 1980s but demonstrated more convincingly by Vreeland. The implication of these two observations is that one must conclude that governments enter into agreements with the IMF without the interests of the working poor at heart. Rather, they seek the IMF's political support (even protection) for what will be painful, regressive policies (elites are sheltered from the programs' worst effects, thereby increasing the likelihood of obtaining their support).

The radical implications of the finding are not seriously pursued, since it is clear that the author seeks only reform rather than revolution. In fact, most of the policy reforms he calls for in the concluding chapter are already being implemented, robbing the work of its political potential.

Be that as it may, this is a work that not only answers important questions but also provides a model for multimethod analysis. Vreeland begins the text with a colorful description of interesting cases (interesting in that they move forward the search for theory): Tanzania, Uruguay, and Nigeria. Already with these three cases we can begin to see some of the patterns that will make up the core of the theoretical model. Following this presentation, he describes an abstract model that factors explicitly the desire of a government to be subjected to conditionality. In the next chapters, he carries out a multivariate regression analysis involving the “usual suspects” in order to explain the propensity to enter into agreements, but with an intriguing twist: He applies a probability-of-agreement model borrowed from his mentor Adam Przeworski that is called “dynamic bivariate probit with partial observability” (p. 84). This allows one to estimate the chance that any two actors will enter into an agreement.

I found that a few paradoxical findings are not given sufficient attention. For example, a positive correlation was found between the length of time a country is under IMF agreements and the country's willingness to enter into a new agreement. This contradicts Vreeland's earlier argument that entering into agreements with the Fund carries “sovereignty costs.” In fact, since governments are generally able to shift the burden of adjustment to the poor and weak, it seems that the sovereignty-costs issue is overdone (this may stem from the discussion of the Tanzania case where the ideologically driven regime had its own reasons to shun the IMF as a Western, capitalist institution).

While the methods are carefully crafted, explained and applied, two observations emerge. First, the gradual unwrapping of the theoretical model over the course of several chapters generates considerable repetition and a few contradictions. A few variables and correlations hinted at in the second and third chapters are discarded by the fifth. It would have been better to begin at the end to maintain focus. Secondly, given the complexity and uncertainty in the model—particularly since uncertainty (“hazard,” “unobserved variables,” and error factors) are made endogenous to the model—it seems that it would be a strong candidate for Bayesian analysis since this would allow for a more inductive approach. That said, my hunch is that it would merely serve to confirm what has already been demonstrated. Overall, The IMF and Economic Development is a very solid piece of work that is successful at many levels. It will be left to the radicals to seize on the full political implications of Vreeland's study.