John J. McCusker, How Much Is That in Real Money? A Historical Commodity Price Index for Use as a Deflator of Money Values in the Economy of the United States. Second Edition, Revised and Enlarged. Worcester, MA: American Antiquarian Society, 2001. ix + 142 pp. $15 (paperback), ISBN 1-929545-01-0. (Available from Oak Knoll Press, 310 Delaware Street, New Castle DE 19720. )

Reviewed by Robert. A. Margo, Department of Economics, Vanderbilt University.

The construction of price indices qualifies as one of the central activities of economic history. Such indices are essential inputs into the measurement of changes in the standard of living over time. When hard data on outputs are absent, price data can provide important clues about the ups and downs of the business cycle. When price series are available for different locations they inform about the degree, or lack thereof, of economic integration.

How Much is the revised version of an earlier "small" (the author's term) book of the same title. As before, the goal is to provide a single, comprehensive price index for the entire sweep of "American" history -- in this revision, 1665-2001. The revisions are of three sorts: extensions back and forward in time, more examples, and additional information on colonial exchange rates. Although economic historians in economics departments will find much that is useful, I suspect the real audience consists of students, and historians other than economic, who, for various reasons, need a specific answer to the question posed in the title.

Most books, even small ones, have several chapters. How Much has a single chapter -- really, an interpretive essay -- five appendices with tables, a bibliography, and a brief Introduction. The essay touches on the history and uses of price indices, and problems therein. Appendix A presents the overall index. As previously, the bulk of the index is derived from the work of David and Solar, whose index itself is mostly derivative of Brady, Bezanson, Cole, and others. McCusker updates his previous updating of David and Solar to 2001 (estimated) using the Bureau of Labor Statistics' CPI-U (urban) cost of living index, available at the click of a button from the BLS website. He also backdates to 1665 using recent work by P.M.G. Harris and Stephen G. Hardy, who have constructed indices from probate records for colonial Maryland and Virginia. These are all spliced together, and the base period set to 1860, allowing easy comparisons over time.

Appendices B and C draw on McCusker's bailiwick, colonial monetary history. The colonies all had their own currency. To convert prices in, say, Maryland and New York into common units requires exchange rates, which are provided in Appendix B. Appendix C does the same for paper money during and just after the Revolutionary War. The tables in these appendices are based on a lot of hard work in archives over many years. Colonial monetary history is an arcane field with more than its fair share of arcane debates. Not being a member of this crowd, I can't really evaluate the issues but, for the rest of us, Appendices B and C certainly seem worth the $15.00 the publishers are asking for the book.

Appendix D provides commodity price indices for Great Britain from 1600 to 2001. Table D-1, like Table A-1, has blank rows extending to 2009, suggesting that another revision may be forthcoming in a few years. Appendix E uses the price index in Appendix A to provide some conjectural dating of business cycles in the United States back to the mid-seventeenth century.

How Much is easy to criticize. The interpretive essay is far from comprehensive. There is barely a mention (except, obliquely, in a footnote) of quality bias and none that I could find of substitution bias. Readers seeking "nuts-and-bolts" advice in constructing a price index will not find much of immediate use in the essay. Although McCusker is correct that some of the criticism directed at historical price indices is off the mark he is too dismissive, in my opinion, of the criticism that many historical indices derive from too limited a range of locations. The essay would have benefited from more (there are some) of McCusker's reasoned judgments on what should be done in future research. The footnotes and bibliography are extensive, but still rather selective -- no mention, for example, is made of this reviewer's rental price index for ante-bellum New York City. Then, there is footnote 22 on pp. 26-27 which claims that "[e]conomic historians are in fact guilty of some of the most simplistic, even misleading use of price indexes" but provides no specific examples.

One can also question the prototypical use -- figuring out what specific items are "worth" in different years -- for which the index is intended. Consider McCusker's example of George Washington's false teeth (p. 37), purchased in 1795 for $60.00. According to McCusker's index, this was "the equivalent of roughly $820 in year 2000 terms." "Roughly" is the operative word in this sentence. In 1984 I purchased my first personal computer, a Kaypro, for $1,500, worth "roughly" $2,485 in year 2000 terms, according to McCusker's index. No one in their right mind would pay $2,485 for such a computer in today's dollars ($51, I note, is the current price on Ebay). However, if such a machine had been available, in say, 1964, I am quite sure someone would have paid many times its alleged 1964 value ($448). Were we able to transport one back in time to 1864, I am also quite sure its value would have been close to zero, since there would have been no way to turn it on.

The point is that, given a suitable estimate of aggregate quality bias, indices like McCusker's can give us a reasonable sense of the average rate of change of the price level -- inflation -- over long periods of time. They do even better over shorter periods of time. And quality bias matters less when we use such a price index to deflate indices of nominal wages, since wage indices are also afflicted by quality bias -- today's workers are healthier, better educated, and so on, than in the distant past. But, to use such an index to predict changes in individual prices is more than a little perilous, because the variance around the average rate of price change -- particularly over long periods of time -- is enormous. Unfortunately, naive users are likely to forget this caveat -- embedded as it is in the word "rough" -- if they are aware of it at all. This caveat notwithstanding, I certainly value having a copy of McCusker at hand, and expect to turn to it often.

(McCusker is Ewing Halsell Distinguished Professor of History, and professor of economics at Trinity University in San Antonio, Texas.)

(Editors Note: The two series: RPI from Great Britian, 1600 to present, and the CPI from the United States, 1650 to present; from this book are available online on the EH.Net server at http://eh.net/hmit )

Robert A. Margo is Professor of Economics and History at Vanderbilt University, Nashville TN. He is the author of Wages and Labor Markets in the United States, 1820-1860 (University of Chicago Press, 2000); and, with Joel Perlmann, Women's Work? American School Teachers, 1650-1920 (University of Chicago Press, 2001).