Art Carden  

Is Collecting GDP Data An Exercise in the Fatal Conceit?

Is the middle class being "squ... Social Security: The Inside St...

While prepping for yesterday's Principles of Macroeconomics discussion of GDP, I listened again to Diane Coyle's EconTalk episode in which she and Russ Roberts discussed the history and development of GDP. Toward the end, they discuss a philosophical question: is it "hubris" to think we can measure and control an economic system?

I tell students that GDP is imperfect, but that imperfect isn't the same thing as useless or misleading. There's a lot that GDP leaves out and a lot it likely measures incorrectly, but it's sufficiently highly correlated with the things we'd really like to measure that it's quite literally good enough for government work.

Russ asked an interesting question at about the 42 minute mark: should we really be collecting and publishing GDP data (he notes that this is a heretical view among economists)? From an Austrian perspective, are we indulging what Hayek called "The Fatal Conceit" by trying to measure and control the aggregate economy? Does it "lead to a false sense of control"?

I think there are two ways to look at this. First, there's a radical change that Deirdre McCloskey emphasizes in her "Bourgeois Era" series. The idea that a society is a project for improvement is one of modernity's major contributions. Schumpeter pointed out that the "capitalist achievement" is not more silk stockings for the Queen of England but more and more stockings for progressively lower levels of effort for factory girls. The achievement of the capitalist era is that we actually care about the factory girls rather than the Queen of England. Hence, the idea of trying to measure national income for something other than our ability to tax and fight wars is pretty marvelous.

Second, there is Russ's provocative invocation of the Austrian approach. By collecting and reporting statistics on aggregate economic activity, are we implicitly endorsing the idea that societies can and should be measured and controlled? In the absence of data on things like GDP, how would we make the kinds of comparisons economists and policymakers want to make? How would we make public policy?

Comments and Sharing


COMMENTS (8 to date)
Thomas writes:

By collecting and reporting statistics on aggregate economic activity, are we implicitly endorsing the idea that societies can and should be measured and controlled? In the absence of data on things like GDP, how would we make the kinds of comparisons economists and policymakers want to make? How would we make public policy?
Your second and third questions answer the first one, in the affirmative.

andy writes:

How do GDP numbers help in making public policy?

vikingvista writes:

Public policy shouldn't be based on aggregates. It should be based on individual behavior.

There is nothing wrong with using aggregates to try to judge societal outcomes, however imperfect that may be, as long as it is done by those (e.g. academics) whose only power to intervene is through voluntary persuasion. The problem comes when those (e.g. politicians) with enormous incentive to lie and distort are able to force their interventions on large swaths of the population.

Andrew_FL writes:

I think the more serious problem is treating accounting identities as predictive models.

That being said, I seem to recall that whoever it was, some official in Hong Kong, who helped ensure it became as economically free as it is now, did not allow the government to collect economic information at all, knowing they would only use it to make mischief. I can't recall the reference for this but I was somewhat taken aback by it at the time. Still, to the extent that the question is an empirical one, this would suggest an affirmative answer.

Daublin writes:

This question comes up with any form of measurement. The answer is generally yes, measure whatever you can.

Measured performance gives us a way to improve how we do things without having to fully understand why the improvement worked. It allows us to drop things that are spectacularly not working, again, without even having to have a clear explanation of why it didn't work.

We should also improve a great many things that are not measurable, but we tend to be very bad at it.

Thucydides writes:

Uh, who is "we"?

MikeP writes:

Andrew_FL, you are thinking of John Cowperthwaite. I thought of him immediately as well.

Here's a quote of his about GNP from 1970:

There was a plea from honourable Members relating to the need for formal Gross National Product figures. Such figures are very inexact even in the most sophisticated countries I think they do not have a great deal of meaning, even as a basis of comparison between economies. That other countries make use of them is not, I think, necessarily a good reason to suppose that we need them. But, although I am not entirely clear what practical purpose they would serve in Hong Kong, I am sure they would be of interest. I suspect myself, however, that the need arises in other countries because high taxation and more or less detailed Government intervention in the economy have made it essential to be able to judge (or to hope to be able to judge) the effect of policies, and of changes in policies, on the economy. One of the honourable Members who spoke on this subject, said outright, as a confirmed planner, that he thought that they were desirable for the planning of our future economic policy. But we are in the happy position, happier at least for the Financial Secretary where the leverage exercised by Government on the economy is so small that it is not necessary, nor even of any particular value, to have these figures available for the formulation of policy. We might indeed be right to be apprehensive lest the availability of such figures might lead, by a reversal of cause and effect, to policies designed to have a direct effect on the economy. I would myself deplore this.
R Richard Schweitzer writes:
How would we make public policy?

As noted by another commentator, therein is the essence of conceit.

First, who shall be the "we" (and why) ?

Next, what is Public Policy (or for that matter what is "policy") other than determination by some, of the conduct of others?

Next, what generates the "need" for policy; the perceptions of some as to how society should be ordered and function (including to ameliorate perceived defects); the need for members of that "some" to attain a sense of significance?

Next, is it that some can determine the particular information from which, by making connections, knowledge can be derived to shape decisions for the conduct of others through Public Policy ?

Is it all a smoke screen in the efforts to entrench positions for some in our increasingly "managed" society?

Comments for this entry have been closed
Return to top