Mankiw and Postmodernism

When Greg Mankiw argues that we can't know what would have happened in the absence of X policy (for example, Obama stimulus policy) because "we only have one economy," and it's impossible to know what would have happened in the absence of any other series of events, as he does here, he is essentially arguing against counterfactual analysis in a rather postmodern vein. That is, he's arguing that we can never make predictions about what would have happened in the presence of another series of events, since we can't actually observe these events.

I think it's pretty clear that he's also arguing against all social science, which is founded on the notion that if some cause doesn't happen (or happens differently), some effect won't happen (or will happen differently.

Interesting that he's never made this claim about his own economic predictions before, to my knowledge. And interesting that he only seems to be bringing up this (questionable) claim about a policy he doesn't support.

Admittedly, at the bottom of the post, he specifically targets the absence of macroeconomic indicators that can measure the effects of the stimulus. This is a legitimate point. But it's a long step from saying, "we don't yet know how to measure the effects of fiscal stimulus" to saying "the effects of fiscal stimulus are unknowable." Isn't this the kind of stuff that economists do for a living?

Eichengreen on the crisis

Barry Eichengreen's essay is here. Effectively, he argues that the primary cause of the crisis was an ideology (a socially-constructed mental model?) that underestimated the risks of business self-regulation, etc...

He also argues that the age of deductive economics based on abstract theoretical models is over, and the age of inductive economics is beginning.

This is an important point--it's a shame it's buried at the end of the article. With deduction, you've got theory. Maybe, you've got elegant, compelling theory, but theory is only theory, not reality. Theory only should be accepted as something resembling reality when it includes evidence. This comes from inductive analysis. Looking at the real world, through data or otherwise.

And, although I'm getting a little simplistic, I think we in the social sciences sometimes forget that theory is only one part of the scientific method. You've got to have theory, hypothesis, and observations that confirm those hypotheses. Models are only models--and where many economists (and political scientists!) go wrong is that they assume that their models are reality, even when they're based on unrealistic assumptions about human behavior.