This paper will test three models of underdevelopment against the experience of Thailand from 1850 to 1940, when Thailand became a major rice exporter. The models are a dependency model, a staples model, and a supply and demand model of technical and institutional change. The technical and institutional change model provides the best explanation. Divergences between the goals of national security and economic development as well as those between the private interests of the elite decision makers and the social interest explain Thai underinvestment in increasing agricultural productivity and account in large part for the persistence of underdevelopment.

Students of agricultural history, of pre-industrial labor productivity, of New England town histories, and, especially, of pre-1840 rates of economic growth have long needed a price index of Massachusetts farm prices for this period. The article presents two indexes, covering 15 farm products, weighted by their relative importance, first in 1800, then in 1855. Among the several issues raised by the behavior of farm prices between 1750 and 1855, one of the most interesting may be what this behavior reveals about the market imbeddedness of these farmers, particularly in view of the fact that the prices come, not from market quotations, but from farm account books.

More than half the variance in length of school session in a cross section of 329 localities in Massachusetts in 1855 can be explained by the share of Irish in.the town's population, the family per dwelling ratio, and a proxy for the share of male merchants over 15 in the population, all of which enter regression equations with strong positive coefficients. This paper considers what these results may tell us about a number of hypotheses that link industrialization and educational revitalization in antebellum Massachusetts, discusses independent confirmation of these basic relationships, and concludes with a more general discussion of the implications of this Massachusetts evidence.

This paper examines the behavior of the postal savings system in the Depression. It is shown that because the system was created with an inflexible structure it was unable to carry out its prescribed function when economic conditions changed in the Depression. Instead, the system evolved from a small-scale government program designed for the low-income saver into a financial force capable of causing problems for the savings and loan industry, the housing market, and even the banking system.

In campaigning for the establishment of the National Forests in the late nineteenth century, conservationists pointed to fraud and timber theft in the Pacific Northwest. In this paper we argue that the conservationists were misdirected; that it was a costly Federal land policy that encouraged fraud and theft. In the face of restrictive land laws, fraud was necessary if lumber companies were to acquire large tracts of land to take advantage of economies of scale in logging. Since fraud used real resources, it raised the actual cost of acquiring land and thus delayed the establishment of property rights. Such delays led to theft. The paper examines the public land laws, explains their selection by claimants, and calculates the added transaction costs or rent dissipation that resulted from circumventing the law.

“Government control of cutting on all timberland, private as well as public, is still today, as it was then, the one most indispensable step toward assuring a supply of forest products for the future in the United States.”

In an earlier issue of this Journal, M. E. Sushka criticized the methodology of such “revisionists” as Peter Temin in their evaluation of the impact of the Bank War on the Jacksonian economy. Sushka characterizes the work of the revisionists as the “use of descriptive statistical tools in an attempt to analyze developments empirically.” Sushka, on the other hand, applies “modern econometric techniques” to “an economic framework drawn from contemporary monetary and macroeconomic theory.”

In their comment Curran and Johnston raise two general points. They suggest (1) that my main conclusions are not substantiated by statistical tests of stability, and (2) that the specifications used in my analysis are inappropriate.

Let me summarize my reply. With regard to Curran and Johnston's statistical tests for the structural stability of my money model, there are three major points. First, in using the Chow test for structural stability the authors do not use the relevant statistical information to carry out the test, so the results they derive are not the appropriate test of the economic hypotheses. I on the other hand report and interpret the results of the relevant Chow test that uses the appropriate statistical information and lends support to my hypothesis that the Bank War altered the public's liquidity preference. Second, Curran and Johnston do not correctly analyze and interpret their own results. Actually, the implications of their own evidence support my original hypothesis that the Bank War caused a shift in the liquidity preference of the public. Third, although the Chow test is a statistical test for stability, in this case it is not a direct test of the economic question at hand; that is, it is a weak test. I therefore provide a more direct statistical F test that provides evidence to support my hypothesis that the Bank War caused the Panic of 1837 and that refutes the Temin hypothesis that the Bank War left monetary behavior unaffected.1 Furthermore, the empirical evidence demonstrates that banks were quite interest sensitive during the years prior to the Bank War and that the Bank War altered commercial bank behavior before 1837 in a manner that, coupled with the shift in the public's liquidity preference, culminated in the Panic of 1837.

This article asks: why were crop failures (dearths) so frequent in western India between 1870 and 1920, and why did dearth decreasingly result in famine, that is, in severe elevation of mortality and in major disruptions of the agricultural sector. It is concluded that deficient rainfall caused the crop failures and that after 1900 increasedtrade in agricultural products and increased demand for labor, coupled with increased credit and relief services from government, significantly reduced the probability that a crop failure could result in widespread famine.

In the June 1977 issue of this Journal, Lewis R. Jones criticized my analysis of the diffusion of mechanical reaping and mowing machinery. Jones attempts to revive the threshold argument popularized by Paul David. In doing so Jones either confuses or ignores the main points I raised while concentrating on a peripheral issue of little significance.

I raised three major issues. First, and of least importance, I revised the values of some of the parameters David used to estimate the threshold acreage. Jones correctly notes that I erred by reporting that the revised threshold was more than twice as large as that previously estimated; the correct calculation is about 93 percent, as Jones notes. If I were to write the paper again, I would correct this mistake. Nonetheless, my major conclusions on this issue are unaltered. Revising two of the parameters used to calculate the threshold acreage so that they better conform with the actual historical situation generates a threshold about twice as large as that previously estimated. I felt this discrepancy was sufficient to cast reasonable doubt on previous estimates, and I noted that similar revisions could be made of most of the parameters and variables used for such estimations (p. 333). Anyone who doubts this assertion should see Robert Ankli's excellent work on the subject.

A low rate of growth of world demand for cotton figures prominently in recent attempts to understand the post-bellum retardation of the southern economy. Gavin Wright, especially, stresses this factor in several articles and a recent book.1 Using-sophisticated regression techniques to estimate the rate of growth of demand for American cotton during both the ante- and post-bellum eras and the magnitude of the change in the rate between them, Wright finds a decline of more than two thirds. Such an occurrence could hardly have helped the South make a prompt recovery from the Civil War.

In this paper we stratify a sample of 24 countries by the role of agriculture in industrial growth in the nineteenth and early twentieth centuries, and focus on systematic differences in the nature and strength of the interactions of institutional with agricultural and industrial change. A novel technique, disjoint principal components models, is applied to categorized data representing 35 facets of social, economic, and political structure and institutions. The results stress the systematic variations in the impact of land institutions among countries with different levels and structures of development. They also underline the critical role of social and political forces in differentiating among paths of economic change. Finally, they highlight the need for cross-section studies using institutional data to understand better the complexity of institutional constraints and influences which often vary little in a given national environment.

Since writing “Cotton Competition,“ I have many times wished that I had never included the paragraph joining pages 632 and 633, since figures that were intended as no more than back-of-the-envelope confirmation of the econometric results have been cited as the central conclusions of the study. To make matters worse, the demand-shift formula was incorrectly applied, most seriously in the exaggerated decline in demand for American cotton between 1860 and 1870. For these reasons I am grateful to John Hanson for making these corrections and for giving me an opportunity to disown this particular halfdozen numbers.