History and statistics serve a common purpose: to understand the causal force of some phenomenon. It seems to me, moreover, that statistics is a simplifying tool to understand causality, whereas history is a more elaborate tool. And by “more elaborate” I mean that history usually attempts to take into account both more variables as well as fundamentally different variables in our quest to understand causality.
To make this point clear, think about what a statistical model is: it is a representation of some dependent variable as a function of one or more independent variables, which we think, perhaps because of some theory, have a causal influence on the dependent variable in question. A historical analysis is a similar type of model. For example, a historian typically starts by acknowledging some development, say a war, and then attempts to describe, in words, the events that led to the particular development. Now, it is true that historians typically delve deeply into the details of the events predating the development – e.g., by examining written correspondence between officials, by reciting historical news clippings to understand the public mood, etc. – but this simply means that the historian is examining more variables than the simplifying statistician. If the statistician added more variables to his regression, he would be on his way producing a historical analysis.
There is, however, one fundamental way in which the historian’s model is different from the statistician’s: namely, the statistician is limited by the fact that he can only consider precisely quantified variables in his model. The historian, in contrast, can add whatever variables he wants to his model. Indeed, the historian’s model is non-numeric.[1]Econolosophy
There are advantages and disadvantages of quantitative, numeric methods and qualitative, non-numeric methods. In the first place, qualitative, non-numeric methods can incorporate qualitative, non-numeric methods, but the reverse does not apply, since quantification is limited to that which can be reduced to quantity without sacrificing anything that is relevant.
This is clear in equating price with value, for example. Value is qualitative and price is quantitive. The key assumptions on which conventional economics rests involve the quantification of economic value independent of other values in terms of price discovered in markets, as well as the quantification of utility. These are simplifications that eliminate from consideration a great deal that is both qualitative and also relevant. While a calculus of utility is an arbitrary construct, there is no moral calculus at all.
On the other hand, history is able to view context holistically in terms of both quantity and quality. Whereas power other than market power is irrelevant in conventional economics, it is central in history. It is also central in economics as Marx and institutionalists recognize.
5 comments:
History is simply the study of the past, and the attempt to gain some knowledge of it. There are many different kinds of knowledge one can possess about the past, and the method one needs to use depends on the type of knowledge one seeks. I think it is a mistake to suppose there are two different fields here with two different methods.
I think it is pretty generally agreed that historical method is different from statistical method, certainly wrt to quantification.
Statistics is a branch of math while history is not and is not even considered to be a social science.
Statistics and social science makes use of the historical method wrt to data selection, however. But the historical method is one tool among many.
Historical context is also relevant in statistics wrt data selection in order to avoid adding apples and oranges. MMT economists — Bill Mitchell comes to mind — initially criticized Reinhart & Rogoff for not paying attention to historical context and mixing fixed and floating rate systems when the type of monetary regime was relevant.
However, the statistical method is of limited application in history, although it is relevant where applicable, of course. A lot of economic history is statistical, for example. But history involves much more than quantifiable data and historical method is nowhere near as precise or as precisely defined as statistics.
They overlap is some areas but are different wrt to both subject matter and method as disciplines. The R&R study is quite relevant, since Reinhart and Rogoff seem to be unaware of the relevance of historical context in data selection.
If a historian wants to answer a question like, "Did the median lifespan increase or decrease under the Tudor dynasty?" or "What was the likelihood of being killed in battle for a Roman legionnaire?" they will have to use statistical methods. And if the historian wants to assemble a more complex causal narrative pertaining to some longer term trend or event, they will need to assemble a number of verifiable answers to empirical questions of this kind as the factual basis of this narrative. There is no one thing that can be referred to as "historical method." The historian has special challenges related to verifying the authenticity and reliability of documentary sources, but beyond that, every branch of science is of potential use to the historian.
Statistics is only as good as the data. Statistics is a mathematical method applied to suitable data (evidence). There is actually very little highly reliable data collected officially even today.
Verifying historical data is often difficult to impossible. Economic data, the most carefully tracked, differs in availability and quality from country to country today, so many statistical comparisons are questionable. For example, the data produced by the US and China is believed to differ widely in quality, but no one actually knows the quality of data reported by China and there are varying estimates.
The best way to use stats for economic and historical arguments/discussions is as descriptive stats. I have written on this at some length here:
Why inferential statistics are inappropriate for development studies and how the same data can be better used
Post a Comment