"...we have to realize that the statistical economist is an historian and not an experimenter. For the social sciences, statistics constitutes a method of historical research"
--Social Science and Natural Science
Read until you have something to write...Write until you have nothing to write...when you have nothing to write, read...read until you have something to write...Jeremiah
I assume he's talking about how it's impossible to validate or invalidate economic claims based upon history. I've commented before on how I think that he's slightly wrong because at some point empirical evidence must count, but Mises went so far to consider that it was indeed nothing more than history and couldn't realistically be integrated with economics... I think that the german historical school had a bit too much of a negative effect on Mises.
The context is quite revealing here. I reproduce it below:
What the statistician establishes in studying the relations between prices and supply or between supply and demand is of historical importance only. If he determines that a rise of 10 per cent in the supply of potatoes in Atlantis in the years between 1920 and 1930 was followed by a fall in the price of potatoes by 8 per cent, he does not say anything about what happened or may happen with a change in the supply of potatoes in another country or at another time. Such measurements as that of elasticity of demand cannot be compared with the physicist's measurement, e.g., specific density or weight of atoms. Of course everybody realizes that the behavior of men concerning potatoes and every other commodity is variable. Different individuals value the same things in a different way, and the valuation changes even with the same individual with changing conditions. We cannot categorize individuals in classes which react in the same way, and we cannot determine the conditions which evoke the same reaction. Under these circumstances we have to realize that the statistical economist is an historian and not an experimenter. For the social sciences, statistics constitutes a method of historical research.
The point is, I think, that examining, managing, and arranging data does not tell us anything about the abstract, generalizable, relationships between actors and goods in exchange. To look at certain prices that were formed from a certain exchange is simply to look at historical data, the byproduct (if you will), of that particular exchange. These data cannot reveal to us anything about economic theory.
He's absolutely correct with respect to economic science, however, the term "social science" is too broad.
Is psychology part of the social sciences? because it is possible to conduct scientific experiments in behavioral psychology for example.
If only people did realize this...
"Is psychology part of the social sciences? because it is possible to conduct scientific experiments in behavioral psychology for example."
it's not possible to conduct good scientific experiment in behavioral psychology.
I'm not sure what is hard to understand. Statistics concerns aggregations of data, data that naturally occured in the past (and in the case of economics is inherently variable.) So yes, an economist preoccupied primarily with economic statistics is an historian of a sort. I think Gabriel.syne's selected quote explains it very well.
Freedom of markets is positively correlated with the degree of evolution in any society...
"it's not possible to conduct good scientific experiment in behavioral psychology."
"good" is a relative term. But it is not true that it is totally futile.
Also, some of psychology is also starting to integrate itself within the more hard sciences such as biology, genetics, neuroscience, etc... The concept of "social science" is getting more and more blurry.
"statistics concerns aggregations of data, data that naturally occured in the past"
It cannot be statistics per se that makes the economist occupied with statistics an historian, but the mere fact that he cannot conduct controlled experiments. It is impossible to infer causation from statistics without accounting for all the variables, and more importantly, controlling for them.
Hence my parenthetic remark. :p