Economics and the Historian

Issues in the Study of Economic Trends

Issues in the Study of Economic Trends THOMAS G. RAWSKI Investigation of economic trends occupies an important position in historical research. Did prices or output of particular commodities orservices or broad aggregates rise, fall, or remain constant over various periods? Can we detect substantial change in levels of economic welfare, distribution of income and wealth, degree of commercialization, patterns of cropping , organization of economic activity, or significance and functioning of various economic institutions? Answers to these questions and many others involve systematic study of economic trends. Attempts to ascertain the presence and determine the direction and magnitude of economic trends call for certain procedures and encounter characteristic difficulties that display strong similaritiesregardless of the trends being studied. The objective of this essay is to discuss the methods and problems that can properly be regarded as typical ingredients in the study of economic trends. THE MEANING OF ECONOMIC TRENDS What do we mean by economic trends? Nobel laureate Sirnon Kuznets, whose work shaped the way that all economists think about long-term economic change, emphasized that any effort to define economic trends requires a suitably homogeneous economy and time period within which we seek to identify unidirectional changes that are sufficiently persistent and sustained to allow us to identify them as long-term trends, as opposed to short-term fluctuations or cycles (1961, chap. 2). M 16 ThomasG. Rawski Several essential points emerge immediately. The study of economic trends is an art or craft rather than an exact science. While technical procedures unfamiliarto most historians may play a role, for instance in determining whether long-term changes "produce movements that dwarf the short-term variations" in some particular context (Kuznets 1961,45), there are crucial issues of judgment and evidence that every historian is trained to comprehend. What isthe appropriate period of analysis? How do we determine if a change is "unidirectional" or "persistent"? How can we satisfy ourselves that we have sorted out the causes of long-term change and short-termfluctuations? Kuznets's criteria also help to explain why we can rarely expect finality in controversies about economic trends. If one's view of certain trends is shaped by judgments concerning the scope of inquiry; if it is possible to reinterpret trend X as the outcome of forces A, B, E, and F twenty years after another scholar has received general acclaim for an explanation focusing on forces A, B, C, and D, we can imagine successive rounds of inquiry into broad issues such as the British industrial revolution or American slavery even in the absence of new sources, flawed data, or disagreement about the validity or interpretation of certain evidence, all of which are, of course, commonplace. Of particular importance here is the passage of time. Access to long series of quantitative or qualitative evidence enormously simplifies the determination of trends. When the data span is short, ambiguity is unavoidable . Scholars concerned with measuring the growth rate of China's post-1949 economy engaged in sharp controversy during the 19608 and early 19703. While political viewpoints no doubt contributed to these disputes , the main problem was insufficient evidence. At a 1972 conference, Simon Kuznets attacked China specialistsfor their inabilityto determine whether output per head had risen or fallen since 1949.' Five years later, these disputes ceased, not because of any change in the sympathies of the combatants or in Peking's publication policies, but simply because the passage of time and accumulation of (scattered and fragmentary) evidence left no room for doubt that substantial increases in per capita output had occurred in the People's Republic. Unfortunately, historians cannot always rely on the passage of time to generate new raw material; only diligent search for unused evidence or for new ways of using well-known sources can add to our stock of data. When we focus on a short period, it can never be easy to separate i. Kuznets's comments were initially directed at the draft version of Perkins 19753. Issues in the Study of Economic Trends 17 trend and cycle. For example, we find one scholar of early twentiethcentury China who predicts that "as we undertake more studies of local conditions . . . brief moments of prosperity will continue to reveal themselves , to be sure. . . . But the long-term pattern" will be quite different (Bell 1985, 2.6). There are others who acknowledge instances and periods of decline but see an overall pattern of economic growth...