35. Cumulative history or a debate without an end? (28-4-2020)
Cumulative history or a debate without an end? (28-4-2020)
Written by: Jan Luiten van Zanden
My supervisor Ad van der Woude saw the writing of history as a ‘gezelschapsspel voor fijne luiden’, an intellectual pastime without much pretentions. He had studied history in Utrecht in the 1960s, and was influenced by the Utrecht school, in particular Pieter Geijl, who considered history to be a ‘debate without an end’, in which, almost at random, interpretations of the past changed, disappeared and came back again. Why intellectuals would want to invest their time in a such a game, is not entirely clear to me. This view of course contrasted sharply with positivistic and sometimes left-wing ideas that the study of the past was serving a purpose, which was to understand the laws of societal evolution, making it possible to ‘assist’ these laws and steer society in the right direction .
A few blogs ago I raised the question if we understand the economy of the 16th century better than, for example, Charles V, who had access to almost all information that was available at the time. It is probably impossible to answer this question, but we perhaps can try to address a more simple one: do we know more than economic historians in the 1960s, or the 1980s? Have we made ‘progress’, or has our paradigm simply shifted from the kind of Braudelian cum Marxian perspective that was fashionable in the 1970s, to the Douglass Northian – Daron Acemogluan Institutional approach that we have adopted in the last 20 years or so? It is difficult to argue that one theory is superior to another, and for sure, at some point we will conclude that North is passé, so it is not entirely clear how to defend the case for progress at the level of theory.
However, there is at least one argument in favour of the idea that we know more than 40 years ago, and that is the development of quantitative economic historical research. When I present the results of the Maddison project to an audience, I show how much we knew about economic growth in for example the UK the 1960s (a few observations between 1870 and 1960), the 1980s (a time series from 1700 to 1980) and now (a time series that covers the better part of the past millennium, with a first observation in 1086, and annual observations since 1270). And this thousand-years-graph can be enriched with similar estimates for Japan, China, Spain, Italy, Holland and half a dozen other countries. I do not think anybody in the 1960s or 1970s, when Maddison started this kind of international comparative research in a systematic way, could have anticipated such an amazing result. This success story was made possible by the fact that he was able to organize some kind of cooperation of the scholars who were separately working on their country and their period, and that he at the end of his career handed this collaboratory over to the next generation. This is, I think, the key to making such a project really cumulative: the research of all scholars has to be included, and at the same time the consistency of the results has to be maintained. Underlying ‘success factors’ were of course that conceptual problems had to a large extent been solved, thanks to the international standardization of the national accounts, on which the historical research could build. And that GDP was seen as the key variable of economic development. This is not to say that all estimates are perfect and reflecting the historical ‘truth’. There are large differences in the quality of the estimates and margins of error obviously become larger when going back in time. Maddison sometimes made impressionistic ‘guestimates’ to challenge scholars to disagree with him and to make better estimates that were based on large-scale empirical research. This often worked, and resulted in much improved (or real) estimates. So the Maddison dataset is work in constant progress, gradually covering more countries, longer time periods (also because each year another year can be added), and replacing low quality estimates by higher quality data. Truly cumulative.
The Maddison dataset is, for us as economic historians, probably the mother of all collabs, but there are similar international comparative projects that have produced very valuable datasets. The work by Bob Allen and many others on real wages springs to mind first, but another good example is the database on slave trade in the Atlantic. The oldest initiative in the field dates already from the 1930s, when the International Scientific Committee for the Study of Prices and Wages was set up; Posthumus was the Dutch member of this ‘global’ initiative. The historians of heights have done similar, international comparative work on the biological standard of living, and international datasets of educational attainment, demographic and gender related variables, and even the activity index of Parliaments have been constructed.
Quantification – the mystery of the number, which makes things comparable in time and space – is making it possible to write the economic and social and demographic history of countries and the world in a much more detailed international comparative way. Perhaps from the Geijl perspective this is just another story, another interpretation that will some day be replaced by another. And perhaps it is possible to agree that what really matters are not the numbers – which have no meaning in themselves – but the stories we tell on that basis. Quantification is a tool, like many others, but the last 60 years or so have shown that it is a very powerful one indeed.
Continue reading: The solitude of the organ (29-4-2020)
The website of the Maddison-project: