## Making it count

8 August, 2012 at 19:53 | Posted in Statistics & Econometrics, Theory of Science & Methodology | 3 CommentsModern econometrics is fundamentally based on assuming – usually without any explicit justification – that we can gain causal knowledge by considering independent variables that may have an impact on the *variation* of a dependent variable. This is however, far from self-evident. Often the *fundamental* causes are *constant* forces that are not amenable to the kind of analysis econometrics supplies us with. As **Stanley Lieberson** has it in his modern classic *Making It Count*:

One can always say whether, in a given empirical context, a given variable or theory accounts for more variation than another. But it is almost certain that the variation observed is not universal over time and place. Hence the use of such a criterion first requires a conclusion about the variation over time and place in the dependent variable. If such an analysis is not forthcoming, the theoretical conclusion is undermined by the absence of information.

…

Moreover, it is questionable whether one can draw much of a conclusion about causal forces from simple analysis of the observed variation … To wit, it is vital that one have an understanding, or at least a working hypothesis, about what is causing the event per se; variation in the magnitude of the event will not provide the answer to that question.

## 3 Comments »

RSS feed for comments on this post. TrackBack URI

### Leave a Reply

Blog at WordPress.com. | The Pool Theme.

Entries and comments feeds.

Indeed, the inability to prove causality is a big problem in all statistics. But I think modern econometricians are well aware of the issue–for any given question in economics, there are reams and reams of research trying to assess whether or not the relationship is causal.

I would not characterize econometrics as “fundamentally based on assuming” that correlation is causal. At a fundamental level, what the theory behind econometrics does is say that if a certain theoretical relationship holds, then it will show up as correlation in the data. So what we do, pedagogically speaking, is start with a theoretical model of the economy in question, and manipulate the predicted result into a form that can be consistently estimated without all the simultaneity and other biases. The economic theory will tell us which correlations are causal and which are spurious, and econometric theory provides us with the basis for identifying each in the data, once we know what we are looking for.

I agree, however, that blindly regressing variables without a strong theoretical motivation about what is and is not causal can lead us astray.

Comment by Matthew Martin— 10 August, 2012 #

I’ve found your recent posts on statistics most refreshing. You might be interested in this paper, Regression and Causation: ‘A Critical Examination of Econometric Textbooks’ Chen and Pearl 2012 at: http://ftp.cs.ucla.edu/pub/stat_ser/r395.pdf

Perl is the 2011 winner of the ACM Turing Award, the highest distinction in computer science, “for fundamental contributions to artificial intelligence through the development of a calculus for probabilistic and causal reasoning. (from Wikipedia)

JRHulls

Comment by JRHulls— 12 August, 2012 #

Thanks for the link. I’ve actually already read the article, and although I concur with Pearl on many things, I’m not sure his alternative causality-concept, in general, is the right one. Maybe I’ll come back to this later, but in the meantime I refer to Nancy Cartwright’s

Hunting Causes and Using Them, where some good arguments are given why Bayes nets may not be the right general causality-concept.Comment by Lars P Syll— 12 August, 2012 #