Money Mischief_ Episodes in Monetary History - Milton Friedman [114]
The four equations that use U.S. income give a negative coefficient for the real price of silver, though only one out of four comes close to statistical significance. On the other hand, the four others (all for the shorter period) are all positive, in line with theoretical expectations, though none comes close to statistical significance.
This evidence clearly does not contribute to resolving the puzzle.
[back]
***
* Data for 1896 from Encyclopaedia Britannica, 11th ed. (1910), s.v. "Gold"; first-quarter-century figure from Warren and Pearson (1935, p. 122).
[back]
***
* As noted in the footnote on pages 64–65, chapter 3, Hugh Rockoff (1990) argues that The Wonderful Wizard of Oz is a fictional retelling of the silver agitation and this campaign.
[back]
***
* Strictly speaking, this statement is correct only for unanticipated deflation. If deflation is anticipated, the interest rate can be adjusted to allow for the expected deflation. However, there is ample evidence that, certainly during the nineteenth century, deflation and inflation were only imperfectly anticipated and then only after a considerable time lag (see chapter 2 and Fisher 1896).
[back]
***
* Chapter 4 presents estimates of the hypothetical price level under these assumptions.
[back]
***
* The reasons for the overestimate are different for the two cases. My estimates for the continuation throughout of the free coinage of silver necessarily use data affected by monetary disturbances that would not have occurred if free coinage had been continued. However, that does not make those estimates valid for the actual circumstances. The reason is that in making my estimates I have assumed a very different prior pattern of accumulation of both gold and silver than actually occurred. Had free coinage at 16 to 1 been enacted in, say, 1897, there would have been an immediate change in the conditions of demand and supply of silver and gold much larger than the gradual change I have postulated. I conjecture that that would have produced a market ratio lower than my estimate of 24 to 1.
[back]
***
* I am indebted for helpful comments on earlier drafts to Angela Redish, Hugh Rockoff, and Anna J. Schwartz. In addition, I have benefited greatly from detailed comments by the editors of the Journal of Economic Perspectives.
[back]
***
* These are the rounded prices. There are 480 grains in a fine ounce of gold, so the exact legal price of gold was 480/23.22, or $20.6711835..., and of silver, 480/371.25, or $1.2929 ....
[back]
***
* It is not easy to document that this remains the conventional view, since few contemporary textbooks on money or macroeconomics even mention bimetallism. They almost all have some reference to the gold standard, but typically take it for granted that a gold standard is the only kind of commodity standard that needs to be mentioned. I have examined seven popular monetary and macroeconomics texts, dated from 1968 to 1986. Only two mention a bimetallic standard; only the earliest has any reasoned discussion of its advantages and disadvantages, and that in a footnote, which notes that "criticism of the system [bimetallism] has doubtless been overdone" (Culbertson 1968, p. 133n.). I have also examined seven texts on American economic history, dated from 1964 to 1987. All discuss the use of different commodities as monetary standards, bimetallism, and the shift to a gold standard. However, the general approach is strictly factual and, with one exception, conventional. For example, the most recent (and, I understand, the most widely used) text states flatly: "Bimetallism is a poor metallic system to use because the two metals fluctuate constantly against each other in price with strange results" and "Silver had been driven from circulation by the rise in gold supplies in the 1840s and 1850s.... Therefore, in 1873 the Coinage Act omitted any provision