|
View the H-German Discussion Logs by month
View the Prior Message in H-German's August 2007 logs by: [date] [author] [thread] View the Next Message in H-German's August 2007 logs by: [date] [author] [thread] Visit the H-German home page.
Germany? Date: Wednesday, May 15, 2007 When I learned that H-German launched a forum on economic history, I was both delighted and skeptical. Delighted because H-German is certainly on the culturalists' side (I am not), skeptical because it is my impression that efforts to communicate between culturalist historians (CHs) and economically-minded economic historians (EMEHs) are honorable but doomed to fail for they live in entirely different worlds. The fact that the partly quite provocative remarks so far do not have sparked a more lively debate seems to underline this. Susan Boettcher's contribution has tried to bridge the ocean between CHs and EMEHs by formulating a number of questions. I will try to give answers, and I will raise new issues. Boettcher asked whether econometrics was a "neutral" methodology. Doing econometrics is applying statistical techniques to economic questions. The usual way EMEHs do it (if they see themselves as "cliometricians") is to formulate testable hypotheses which may emerge from economics or not, and test them statistically. Any theory is welcome--neo-classical, Keynesian, Marxist, or whatsoever--and if you browse through the _Journal of Economic History_ or any other economic history journal you will realize that many seemingly econometric articles do not even rely on economic theories but on ad- hoc assumptions like "if you are unemployed, you will tend to vote for a more radical party, be they communists or Nazis." Which is fine. The point is that the hypothesis is testable, that is open to be falsified. In this sense econometrics is "neutral". Does it include its own biases? Absolutely. But the bias is not political, it is rather that there are many questions around that are not easily formulated in a testable way, especially the big and important ones. In particular, it is often difficult to find historical data for the variables that you think are important. E.g., one could formulate the hypothesis that the more capitalist a country is, the more culturalist are its historians (grin). For the first part of the statement you will easily find sufficiently reliable variables, but for the second you will have difficulties: how can we measure culturalishness? Boettcher wondered further whether econometrics is simply the quantitative version of capitalist theory. Whatever "capitalist theory" may be, the answer is a clear no. Any hypothesis can be subjected to a statististical test if you have sufficiently adequate data. I am not familiar with the East German debate on the Peasants' War, but contributions published in the 1980s or earlier did certainly not rely on the quantitative methods that are available now -- including the access to powerful statistical software packages. As to the cultural studies Boettcher refers to: My dissatisfaction with CHs is that they (well, many) "prove" an asserted causal relationship with a number of citations from sources, and that's it. Efforts to test this empirically are, as far as I can see, rare. Although the CHs I know are not happy to hear this, "cultural" hypotheses may of course be subjected to empirical tests. Boettcher asked: "What is being offered by a 'new' economic history that is not present in the 'old' one?" Very simple: multivariate rather than bivariate analysis. The statistical tables in musty old texts Boettcher refers to do not amount to quantitative history in the modern sense. Modern quantitative methods actually use the information displayed in tables. Boettcher: What is the purpose of the apparent advocacy of moving away from cultural history? If anything, from my perspective, an understanding of the role of cultural factors such as religion in world history and relations has taken on more, not less, urgency since the turn of the century. Spoerer: Good. Why not test the empirical importance of these factors? Boettcher: I am wondering if some of these contributions do not simultaneously underestimate the current level of economic literacy among historians... Spoerer: This I very, very strongly doubt (from a German perspective)... Boettcher: ... and understate the amount of time and commitment necessary to obtain the necessary quantitative proficiency to carry out economic history research. Spoerer: ... and this, given the deficiencies of my own knowledge of statistics, I fully support. Boettcher: My impression is that graduate programs are tending toward eliminating the potential of substituting statistics or economics for a language because so few students are interested in this option. Spoerer: Sure, history students behave rationally: If their supervisors tell them you can get through by quoting Foucault rather than taking Popper seriously, of course they no longer choose statistical methods. Boettcher: My question is thus: shouldn't we (for both undergraduates and graduate students) be emphasizing increased economic _literacy_ either as a first step toward or in preference to, a fully quantitative economic history? Spoerer: In my view, economic and statistical literacy are two completely different things. Economics is a social science, statistics (in so far as we historians deal with it) is a tool. Boettcher: Still, the early modernist wrestles even more necessarily with the problem of incomplete and/or unreliable data series (which is one reason why estimates of the German population before 1800 vary so wildly). What do we do with this source situation? Spoerer: Some cliometricians simply take the numbers they can find, assert that the variables for which they can find data are representative (proxies) for the data they really needed, throw them into a statistical software package and hope that the results are "significant." This is certainly not very convincing (at least not for me). My answer would be: without quantitative data, no statistical tests. But one can still try to use economic reasoning, or reasoning from other social sciences. Boettcher: As a related question, of how much value are economic theories and quantitative methods gauged to the evaluation and explanation of capitalism in analyzing pre-capitalist economies? Spoerer: This is one of the issues where CHs and EMEHs will never agree. The latter are fully convinced (perhaps sometimes a bit too self-confident) that most of human behavior, and certainly that targeted at solving economic problems, follows a, grosso modo, time- invariant behavior pattern which in its essence may be circumscribed with the much contested and much misunderstood metaphor of homo oeconomicus. Boettcher: Don't we simply end up (once again, as the discussion about the emergence of globalization has done) characterizing early modernity as an incomplete version of something that has its full- fledged version later, and thus missing something important? Spoerer: Would that be so wrong -- given that today's world is an incomplete version of something that has its full-fledged version in the future? Mark Spoerer University of Hohenheim
 Humanities & Social Sciences Online 
|