Bimetallism

960

Journal of Economic Literature, Vol. XL (September 2002)
whereas sunspots cannot, and sunspots can occur in models without bubbles. Garber does not discuss either of these lines of work. Here it will be objected that I am indulging in the reviewer's practice of criticizing Garber for not writing a book different from that which he set out to write. Fair enough. To repeat, the task Garberset for himself was to try to determine from the historical record what actually happened in the famous first bubbles. At this task he succeeded admirably. This book is great reading.
STEPHEN F LEROY

Garber also argues that sharp drops in the prices of newly introduced and rare bulbs, similar to that of 1634-1637, occurred at most other dates for which data are available. Further, given therates at which bulbs can be propagated, such drops are to be expected and do not imply losses to tlie owners of these bulbs. Thus even steep price drops do not indicate the collapse of bubbles. This careful interpretation of the historical record is altogether persuasive, and it is very impressive. However, the theoretical side of Garber's discussion of bubbles is less satisfactory. Garber'sargument is that, as economists, we should accept explanations relying on irrationality only as a last resort. Most economists would go farther, and would argue that apparent evidence of irrationality should always be interpreted as indicating hidden costs or other model misspecification. Along these lines, rationality is interpreted as an organizing principle for interpreting behavior, not as afalsifiable hypothesis about behavior. Does Garber agree with this? If so, it is not clear why he bothers to look at the historical record at all, since he is planning on rejecting irrationality regardless of what the historical record shows. If not, does Garber believe that there are any historical episodes in which economic behavior was genuinely irrational? How are these different from the famous firstbubbles which he discusses? In general, what, if anything, would justify a conclusion in favor of irrationality? Garber does not address these fundamental questions. He does observe that "[The] bubble is one of the most beautiful concepts in economics and finance in that it is a fuzzy word filled with import but lacking a solid operational definition" (p. 4). This is flat-out wrong. There existsa large and growing literature on rational bubbles that offers a precise definition of bubbles and derives conditions under which bubbles—defined as the difference between the price of an asset and the fundamental value of its payoff—can occur. Further, there exists a parallel literature on sunspots that determines conditions under which equilibria can depend on non-fundamental exogenousvariables. Many analysts, Garber included, confuse bubbles and sunspots. This is understandable, given the common use of the term "fundamental" in both cases, but the meaning of fundamental is entirely different in the two contexts. Bubbles can occur in deterministic models.

University of California, Santa Barbara Bimetallism: An Economic and Historical Analysis. By Angela Redish. Gambridge, New York,and Melbourne: Gambridge University Press, 2000. Pp. xii, 275. $54.95. ISBN 0-521-57091-3. JEL 2001-0760 In recent years the gold standard has been subjected to intense scrutiny by a number of able economists and economic historians. The reason for looking so closely at the gold standard, of course, is the hope that it may hold important lessons about the costs and benefits of fixed exchange ratesand monetary integration. But what came before the gold standard? The answer, Angela Redish shows in this detailed yet readable account, is that for several centuries before the gold standard was adopted, European monetary systems were bimetallic: both gold and silver served as the building blocks of the monetary system. One of the main advantages of bimetallism was that full-bodied coins (the...