Pierre Siklos indicates that “The aim [of this book] is to try and isolate how monetary policies have been shaped by economic, political, and institutional forces.” (p. xviii) This is an extraordinarily ambitious task. Normally, a statement about a book being ambitious is followed by a statement that the goal was not reached or was impossible anyway.

Such is not the case here. Pierre Siklos has largely succeeded in providing an interesting and informative analysis of how monetary arrangements have changed since World War II and an explanation of why, using an extraordinary blend of theory, data, and institutional considerations. Siklos summarizes conclusions of much of the literature on monetary policy since the 1950s. He also presents a valuable set of data on institutional arrangements for 21 central banks, as well as evidence on how those institutional arrangements affect inflation outcomes. Coverage includes Western European central banks and those of the United States, Canada, Australia, New Zealand, and Japan.

Siklos has kept the analysis at a fairly general level, with mathematical and econometric details in the background. At least some of these details are available in associated working papers. For a reader used to seeing them, it is sometimes frustrating to be unsure how something was done, although it is not particularly likely that the empirical results presented are sensitive to technical choices. The overall task in this book is sufficiently broad that leaving the details in the background usefully makes the book accessible to a wider readership.

The first part of the book analyzes central banks' institutional arrangements and their implications, including the importance of personal as opposed to institutional reputation. The empirical analysis of the central 90 pages indicates that these institutional factors are important for central banks' behavior. The last part of the book shows how experience with alternatives led to the conclusion that inflation targeting is the best monetary policy for generating low inflation and why this is a correct conclusion.

For postwar European central banks, there is an overarching question: Why was there a big increase in inflation in the 1970s—sometimes called the “Great Inflation”—and much lower inflation in the 1990s? Siklos concludes that the Great Inflation was an episode of bad policy due to poor institutions. Central banks were insufficiently accountable, with vague goals. A lesson was learned: central banks are generating less inflation now because they are more accountable and are targeting inflation instead of mistakenly attempting to stabilize economic activity.

This answer offers an interesting contrast with Thomas Sargent's analysis of the same question in The Conquest of American Inflation (Princeton, NJ: Princeton University Press, 1999), which also relies on learning, but of a different sort. In Sargent's analysis, the central bank learns about the economy instead of politicians and monetary economists learning the best way to organize and operate a central bank. Siklos specifically rejects a principal-agent analysis of central banks (p. 237). Instead, as is typical in much recent literature, he proceeds from an assumption that a central bank is attempting to maximize the welfare of the population of the country. Although this is a nice thought—at least to someone working at the Federal Reserve—there are reasons to think that it misses details, details that are likely to be important for understanding the implications of institutional arrangements. Analyzing the principal-agent problems between the voters, representative government, and the central bank would clarify the roles of goals as well as of disclosure and accountability in the operation of the central bank.

Siklos concludes that we have learned the correct lesson: An accountable central bank with sufficient disclosure and an inflation target will generate low inflation, a state of affairs that can continue for the foreseeable future.

I am not so optimistic. As Siklos points out, failure to achieve the inflation target has had no real consequences for the governor of the central bank, even in New Zealand, the country that probably has the closest connection between unwanted inflation and the governor's position. A couple of alternatives, which are hard to distinguish given current data, are more plausible to me personally.

Sargent's view has been examined in the most depth. In his analysis, central bankers learned in the 1970s that the Phillips curve was steeper than they thought and, therefore, that the gains from inflation were less than originally thought. As a result, inflation fell. As he sees it, the central bank learned something incorrect: while it appears that there is an exploitable Phillips curve tradeoff between inflation and unemployment, there really is not. For the future, Sargent's analysis suggests that a period of low inflation will generate data consistent with a flatter, exploitable Phillips' curve, and central banks will increase inflation again.

Alternative explanations for the Great Inflation can also be built on the ability of unexpected inflation to redistribute wealth across voters for a time.

Even given these alternative explanations, Siklos's analysis of inflation targeting probably reflects the reigning orthodoxy. It may even be correct.

The book is reasonably well edited in general. It may seem churlish to say this after the author has put in so much work, but my understanding of the book would have benefited from more editing of the writing.

Overall, this book is a very valuable addition to the literature. It is worthwhile reading if you are interested in central banking generally or the history of central banking since World War II, or if you want a readable introduction to how central banks ended up targeting inflation.