================= HES POSTING ================= [NOTE: Responses to the opening statement for this conference, or the questions it raises are also welcome on the list. If there are responses, I will start a new thread with the subject line "DISC -- Age of Measurement". Conversation about some of these themes might help prompt some more contributions to the conference! -- RBE] THE AGE OF ECONOMIC MEASUREMENT HOPE Conference, 2000 Guest Editors: Judy L. Klein and Mary S. Morgan Since the late nineteenth century, measurement has been a integral part of economics. The metrics of the discipline lend it credence as a science while contemporary debates illustrate the political significance of economic measurement (for examples, the 1996-97 US Congressional hearings on the Consumer Price Index, the symposium on measurement in the winter 1998 issue of The Journal of Economic Perspectives, and the recent attempts by the United Nations to construct a "Human Development Index"). The History of Political Economy conference in the year 2000 will explore the origins of economic measurement and the historical links between such empirical observation and issues of public policy, changes in economic theory and academic practice. There are no geographical boundaries to our proposed exploration, but we imagine that most of the papers will concentrate on the time period from 1850 to 1950, which could well be labelled the "age of measurement in economics." While historians of statistics such as Hacking (1983) have commented on the "avalanche of numbers" which appeared in the 1820- 40 period, that sudden desire to collect statistics seems to have focused more on the social elements of the state and of individual deviancy than on economic magnitudes. The census of population, of course, is the most obvious manifestation, but the fascination with suicide, murder, prostitution and ill health (all noted by Hacking) seemed to form the bulk of social measurement of the period. Despite the interests of economists in joining emerging statistical societies in the 19th century, economic measurement seems to have only reached the same kind of avalanche proportions during the last quarter of the century, when policy makers and economists began to think that economic measurement was a pre-requisite to both scientific knowledge and to reform of the body-economic. During the last part of the nineteenth century, we see economists seriously engaged with questions of how to provide measurements to match their concepts, (eg the price level); with the development of appropriate measurement formulae (eg index numbers); as well as with questions of collection and categorization of data in ways which accurately describe the economy and reveal its problems (eg business cycles). In the early 20th century, we begin to see monographs of measurements and about measurement techniques. Thus, way before econometricians' attempts to measure relationships became widespread, it had become commonplace for economists to describe the economy in numbers by taking and using measurements. But economists weren't the only ones to become obsessed with measurement in the later 19th century. Kelvin's dictum: "If you cannot measure, your knowledge is meagre and unsatisfactory" seems to be widely indicative of the move towards exact and standardized measurements in the physical sciences during that period (see Wise, 1995). It was not that quantification was new, either in the sciences, as evidenced by historians work on 18th century quantification (Frangsmyr et al, 1990), or in the socio-economic realm (as discussed by Kula, 1986). Rather, it seems that, only in the last quarter of the century, did accurate description of the economy and measurement of the objects/elements in economic theory become an essential part of the economic scientist's toolbox. Despite the importance of this topic, the age of economic measurement has received remarkably little attention from historians of economics. The aim of the conference, and the subsequent special issue of History of Political Economy, is to account for the emergence and establishment of economic measurement as a critical element of modern economics. Our research conference therefore invites contributions that analyze the history of attempts to provide a measurement basis to economics as well as papers offering explanations for the general drive to measure that we have noted above. We need to lay out the main path of this age of economic measurement, both in general and in detail, just as much as we need to delve into the reasons for this movement towards measurement, its implications for economics, and the changes it wrought. We would like to understand what questions and problems created the drive to measurement and how economists reacted to the changing status of numbers in their discipline. Conference questions and subthemes There are a number of questions that we suggest might be addressed in papers for the workshop: 1) What has constituted measurement in economics? Do numbers have to be involved or have representations (eg maps or graphs) also counted as measurements? For example, does William Playfair's 1804 chart of Universal Commercial History constitute measurement of national economic strength? Are Griffith Taylor's early twentieth century time-space economic geography diagrams measurements? Do the results of late-twentieth century computer simulations that test economic models overcome accusations of theory without measurement any more than Ricardo's hypothetical calculations? How has the notion of measurement and what counts as measurement changed over time? How have methods of representing measurements altered over time? One of the trends that has been well-documented in the history of statistics and econometrics literature is the gradual trend in economics from measuring things, without analysis or attribution of causes, towards measuring relations and establishing explanations. Indeed, we could reasonably ask the historical question: Has econometrics always been about observation not about testing (as Kevin Hoover (1994) claims for modern econometrics)? Although our conference is not primarily concerned with the history of econometrics, we are very much interested in accounts of this growth in ambition in the measurement movement. 2) How do the histories of the other standard elements of economics: models, theories, evidence etc relate to the history of economic measurement? What is the role of measurement in the relationships between abstract and empirical reasoning? In 1929, at an AEA workshop on quantitative economics, Eveline Burns asserted that "the proper attitude for those of us who are primarily theorists is to ask in what possible ways our view of the economic world and the theories we have built up to account for it are influenced by the methods and results of the quantitative workers." What were the links between the theorists and the quantitative workers? For example, how did contemporary theoretical discussions about national wealth influence attempts to measure that wealth from Playfair's "linear arithmetic"in the late eighteenth century to Michael Mulhall's and Ezra Seaman's measurements of national progress and productivity in the nineteenth century to Kuznet's national income accounting in the twentieth century? Did the "demands of the speculative theorist far outrun the lagging pace of observation" as Frederick Mills argued? 3) How does the history of statistics relate to the history of economic measurement? While there has been work on economic measurements within the history of statistics, and considerable historical discussion of social measurement at the end of the 19th century, more work needs to be done to illuminate the relation of such measurement strategies to the measurement drive in the history of economics. This subtheme can also be interpreted as a question of scientific interaction: What are the relations between economic and social measurement strategies, between the examples provided by physical measurement and those in the economic realm and vice versa? Jevons and Edgeworth, for example saw economists' attempts to codify and measure utility as comparable to the probabilists' attempts to measure doubt and belief. To what extent, and in what ways, did concerns with measurement in the late nineteenth and twentieth centuries function to forge a link between disciplines, such as the biometric link in the adoption of statistical methods? 4) What are the cultural, social, and political contexts within which measurement takes place? Why, for example, was the measurement of the distribution of income and wealth so important to Italian economists in the late nineteenth century? What economic statistics were conceptualized by wartime states or central planning authorities? Are there useful comparisons to be made between, for example, RAND economists equating pilots and planes in the measurement of military worth for resource allocation in the Cold War and William Petty's equation of land and people for the political arithmetic of colonial fiscal policy? Why, for example, were Nikolai Kondratiev's construction of peasant indices or the measurement of the terms of trade for agriculture so threatening to the Soviet government in the 1920s? 5) Context and content are interdependent - measurements are used in policy purposes and policy purposes encourage measurement. Does this explain why some things get measured and others not? This subtheme therefore invites us to consider: What are the feedback loops between policy, measurement, and theory? How have the loops changed in form over time or subject? Is measurement a first step in policy control? This applies most obviously to war-time developments in economic measurement. But the question may also apply more widely. J. R. Hicks argued that monetary theory was stimulated by events, particularly major disturbances. Thomas Humphrey (1973) of the Richmond Federal Reserve argues that Hicks's conjecture also applies to empirical research on money and credit flows. For example much of the research effort of the American monetarists in the 1920's was devoted to the search for the perfect price index with which to fit the quantity theory for stabilization purposes. 6) The questions of context can also be interpreted narrowly to cover the nature of social circles within which measurement aims and practices are constituted and carry across scientific and social borders. On the one hand this invites us to consider the role of learned societies (e.g. Manchester Statistical Society and their periodic analysis of credit flows) as the centre in a measurement network. We might also consider seriously the role of measurement posts or non-state data recorders: the institutions (eg banks, financial journals, business cycle institutes, government agencies etc.) which record information for their own reasons and whose data then become our economic measurements for scientific purposes. What are the historical interactions between economic measurements used in commercial, political and scientific arithmetic? On the other hand this sub-theme also suggests we pose detailed questions about the nature of training in measurement within economics; about the history of practices of measurement and their assimilation; about the differentiation of measurements and their place and role within economic disputes. This subtheme also invites us to consider the role of borrowing of measurement structures within and between sub-fields in economics (for example, the gradual expansion of index number formulae to the measurement of different economic objects; or questions regarding when and why measurement became a necessary condition for the construction of economic history?) 7) How is the movement towards measurement placed within the economic ideas of the period? This internalist question relates both to the changes over time within economics and the relative effects of those, as well as the differences between national preferences and "schools of thought" within economics. For example, the commitment to measurement within both historical and (certain parts of the) institutionalist schools entails a different kind of view of measurement than is supposed to be associated with the utility thinking of the marginalist approach. Some might argue that the latter is a query about measurability not measurement. To what extent did divisions over measurement projects lie behind the split between economic historians and neoclassical economists? How do these varied notions of measurement and its role fit with changing and contested notions of what constituted good economic science? In general: How have economists understood scientific objectivity in relation to measurement? 8) There are clearly some critical methodological questions in economic measurement not least because creating a good correspondence between measurements and economic concepts is often problematic (for example the problem of utility, or the problem of measuring non-market aspects in GNP). At the same time there are problems with getting good accurate measurements, even where the concept is one which is measurable: this leads to standard measuring devices which are often regarded as best proxies rather than good measures (the example of index numbers of cost and standard of living). How do such questions of measurement get resolved and such practices of measurement get standardised in the history of economics? What has been involved in establishing acceptable measurements for things and processes that have tended to be only accessible by indirect means (eg business cycles)? How do some of the philosophies of science of the period - eg pragmatism, operationalism, positivism etc. relate to the age of measurement in its various manifestations? 9) The rhetoric of measurement has been considered by McCloskey (1985) in a well-known attack on econometrics, and by Porter in a discussion of the role of economic numbers in relation to expertise (1995); but a history of the ways that economic measurement has been used in economic discourse over the period, of its rise and establishment, its ability to persuade and to structure our ways of talking in economics remains an interesting project. 10) What account can we give of the changing technologies of measurement? In what ways were the technologies for collecting and analyzing descriptive statistics used in the past different from those now used to provide data for simulation and computational experiments? How are our statistics related to our computational hardware and our algorithmic software? To what extent have measurement technologies and computational resources been a limit to measurement or a stimulus to innovation in measurement and to model construction as a substitute? References: Fransmyr, T. J.L. Heilbron and R.E. Rider (1990) The Quantifying Spirit in the Eighteenth Century (University of California Press). Hacking, I. (1983) The Taming of Chance (Cambridge University Press). Hoover, K. (1994) "Econometrics as Observation: The Lucas Critique and the Nature of Econometric Inference" Journal of Economic Methodology 1(1), 65-80. Humphrey, T. (1973) "Empirical Tests of the Quantity Theory of Money in the United States, 1900-1930" History of Political Economy, 5, 285-316. Kula, W. (1986) Measures and Men (translated by R. Szreter) (Princeton University Press). McCloskey, D.N. (1985) "The Loss Function has been Mislaid: The Rhetoric of Significance Tests" American Economic Review, 75(2), 201-205. Porter, T.H. (1995) Trust in Numbers (Princeton University Press). Wise, N.M. (1995) The Values of Precision (Princeton University Press). THE AGE OF ECONOMIC MEASUREMENT: CALL FOR PAPERS All proposals for contributions to the workshop (in March/April 2000) and subsequent consideration for publication in the special issue of History of Political Economy (in year 2001) are most welcome. The general topic to be addressed: "The Age of Measurement in Economics" is outlined above, along with suggested sub-themes. Of course, proposals which extend our themes are also welcomed. No doubt the history of measurement of specific concepts (e.g. wealth, inequality of income distribution, concentration ratios) or specific processes (cycles) would throw light on many of the above questions, but we expect that proposals of this type would not be entirely descriptive in intent, rather that they will relate to some of the questions outlined above. Please send immediate expressions of interest, followed by one page outline proposals by 1st May 1999, to both the organisers: Judy L. Klein (at Department of Economics, Mary Baldwin College, Staunton, VA 24401, USA. Phone: 1-540-887-7053 Fax: 1-540-887-7137 email: [log in to unmask]) Mary Morgan (at Department of Economic History, London School of Economics, Houghton St., London WC2A 2AE, UK. Phone 44-171-955- 7081; Fax 44-171-955-7730; email: [log in to unmask]). ============ FOOTER TO HES POSTING ============ For information, send the message "info HES" to [log in to unmask]