One problem with which economic historians always contend, to which they usually give acknowledgment and with which they occasionally think they have dealt successfully, is that of the representativeness of their empirical answers. As Clapham said,

Every economic historian should, however, have acquired what might be called the statistical sense, the habit of asking in relation to any institution, policy, group or movement the questions: how large? how long? how often? how representative?

The major purpose of this paper is to call attention to an article by Keynes (1937) published in Eugenics Review the year following the publication of the General Theory, and which has been neglected in the literature on Keynes. The emphasis in the General Theory is on the short-run with given resources and a given state of technology. In his Eugenics Review article, Keynes examined the effects of a changing population upon savings and investment and the level of income and employment. He was thereby led on to an approach nearer to what might be regarded as growth theory than that of the General Theory. In this sense, Keynes' 1937 article may be viewed as supplementing the General Theory.

Recently an article by Samuel Pao-San Ho appeared in this Journal, which, while using some of the data from my previous study of Taiwan's agriculture, criticizes the input series I compiled. Samuel Ho offers competing estimates of productivity change and its sources, and provides estimates of the transferred agricultural surplus and an interpretation of the transfer process. His estimates of the productivity change in Taiwan's agriculture during the colonial period are considerably lower than my own.

Between the years 1860 and 1913 approximately twelve million people took passage from the United Kingdom to extra-European countries. The bulk of the migration stream (about 125,000 people per year) was directed toward the United States; it is this movement of population that is the subject of our article. The flow of individuals from the United Kingdom to the United States in this period ranged from 38,000 in 1861 to 202,000 in 1887 with marked cyclical fluctuations. For example, in 1873 the flow was 167,000 and by 1877 it was only 45,000. Variations of this magnitude pose the interesting intellectual question of whether or not they can be explained. This is not a new question; there are frequent references in the literature to the possible causes of this movement and the emigration from the United Kingdom that it implies. Studies focus on various economic influences on emigration. There is little in this period in the socio-political environment of the United Kingdom that would prompt individuals to emigrate in order to flee intolerable religious or political persecution.

The production of economic history, like that in many fields of scholarly endeavor, increased sharply in the past quarter-century, compared to the rate of output in earlier eras. While the “new” economic history, with its emphasis on economic theory and measurement, has attracted considerable attention during the last decade, “traditional” economic history, written along institutional lines, has continued to be significant, both quantitatively (in terms of numbers of books and articles) and qualitatively (as assessed by contributions to our understanding of economic processes.)

Forecasting developments in firm and industry studies during the next decade is a risky business. Students of firms and industries ask widely different questions about the causes, processes, and/or results of change, and they employ an equally wide variety of methodologies in seeking answers. As a result, objectives are more surrounded and enveloped than taken in accordance with a widely accepted and nicely articulated battleplan. This paper, then, is not offered as a battleplan for the next decade but as a survey of the fronts where the possibilities of advance seem most promising, with some effort to suggest targets and weaponry that are or might become relevant.

In his comment on my 1968 article, Professor Y. M. Ho questions: (1) my selection of input measurements, (2) my explanation of the residual, and (3) my discussion of the agricultural surplus. I shall comment on each of these issues in turn.

The scarcity of quantitative evidence on economic growth in the nineteenth century is a strong incentive for the use of representative indicators, instead of compound indexes or supplemented by them. Consumption series of coal, iron, and cotton are prominent candidates to serve this purpose, due to the central roles played by these materials in European industrialization. A further point in their favor is the relatively high availability and quality of mining and foreign trade statistics. The specific functions and link-ages of those three staples, however, were quite different, and thus their growth paths represent different aspects and processes. For the overall modernizing transformation of an economy, coal consumption evidently suggests itself as the best single indicator, at least until the end of the century. Its growth reflects the increasing output of industry, transportation, and even agriculture, inasmuch as they relied on coal for heat and/or for motive power. This means that coal stands for a broad cross-section of the economy, and furthermore, for a growing one. The substitution of mineral fuel for wood or charcoal and of steam for water power was not only the most general component of industrialization, in the wider sense of the term. It also was a prerequisite for the growth potential of firms and industries in many instances, among them such strategic sectors as iron and steel or overland transportation. Similarly, the introduction of steam engines was a characteristic of the innovating firm in agriculture and in land reclamation. On these grounds one would expect the growth of coal consumption to represent the pace of transition to modern economic growth, rather than the growth of industrial product only.

Much of the federally owned public lands in the American South to the west of the Appalachians was sold to private interests between 1820 and 1860. Land sales' policy remained a great political issue during the nineteenth century because of its perceived effects on the distribution of wealth, sectoral economic growth, and the geographic location of political power. In this essay we consider the marginal impact of Southern land sales on national income. Like all models, our model is only as good as its underlying assumptions. Our aim is less to provide the last word on an important historical issue than to place the problem in a context in which analytical tools can be employed.

The meteoric rise of the new economic history reflects the fact that historians abhor a vacuum and sometimes it seems to reflect an even baser instinct that any theory is better than no theory. Of course, any explanation always uses theory; however, it was usually implicit and frequently a bouillabaisse in which Marx, Veblen and the German historical school floated around in as equally indigestible lumps. But the combination of an internally consistent paradigm based on a few simple assumptions and the mysteries of econometrics simply overwhelmed the older historian, as much as anything else, because he could not understand the rules of the game, much less play it. But even while the new breed was destroying one traditional explanation after another, the traditional historian even in retreat kept muttering over and over, “But you are destroying the existing myths without replacing them. Soon there will be no explanation—no economic history—just an immense heap of numbers.” And sometimes, plaintively from the left flank of the retreating historians, there would come the cry, “But institutions ARE important!”