Saturday, September 11, 2010

A Response from Bob Higgs


Whenever I make a comment about some luminary directly, and they respond, I like to bring people's attention to it. The other day in a post about how government output and employment is counted in GDP, Bob Higgs wrote this:

"Just for the record, there is no "Robert Higgs school of GDP." Indeed, to be even more reassuring, I affirm that there is no Robert Higgs school of anything. Which is in no way a bad thing, of course. As Schumpeter said, fish go in schools.

I have criticized the way government is handled in the national income and product accounts, especially in regard to the so-called wartime prosperity during World War II, but my criticism springs from basic theoretical principles that were debated by such fanatics as Simon Kuznets, Paul Samuelson, Moses Abramovitz, James Tobin, and other obscure economists who analyzed the foundations of national accounting back in the 1940s and 1950s -- before the economics profession threw up its collective hands and capitulated to the arbitrary conventions of the Commerce Department. Because the government component has no grounding in market-choice behavior and the price system, as the private components do, it is simply impossible, with a clear conscience, to treat it in the way it is treated in the national accounts. Much as I would like to take credit for having been the first one to arrive at this conclusion, I cannot do so honestly.

Those who suppose that the government component of GDP means something might do well to ponder that if government only chooses to pay more for the final goods and services it purchases, it may simply do so, either by creating money itself or by employing its coercive power directly to tax or indirectly to borrow (putting its coercion in the shadow of its promise to repay). The element of sheer arbitrariness in the price and quantity data that result from this process looms very large."


First, I hope it's perfectly clear that I was using "Robert Higgs school of GDP" casually! I don't think that needs clarification, but just in case! I recognize the problem he associates with it very clearly. We discuss the socialist calculation debate and the price mechanism all the time on this blog, so I'm no stranger to the valuation problems posed by pricing goods that the public sector produces or purchases. I simply don't think Higgs is really being fair when he calls my (and the BEA's) perspective "arbitrary" (and I think he's being a little too sensitive when he facetiously calls his position "fanatical" - to be clear, I never suggest it was fanatical, simply that it was wrong).

I'm not quite sure what to make of the response. Should we go back and revise down all the GDP figures associated with housing between 2000 and 2007? Should we revise down all the carbon-related figures because the price mechanism isn't capturing all costs and benefits there? The price mechanism doesn't always operate as clean theory says it should - this is not news, and as Higgs implies, it's not exactly obscure or fanatical either. So we have a valuation problem. What is government production actually worth? We have a valuation problem for many market goods too, though, unless you plug your ears and close your eyes to much of what we know about the market process. But what is closer to being a biased figure? Not including government production which clearly has a real value simply because we can't price it exactly? Calling it zero instead of some biased positive figure? That seems odd to me. What are malinvestments if not overpaid for goods and services? And we count these.

I think Higgs's fundamental problem is that he's trying to make the argument that GDP should be GDVP, "gross domestic value production". If that's what he wants to measure, that's fine - but I don't think we can measure that. True, the prices government pay don't necessarily reflect true value. But the prices paid in the market don't reflect true value either - they simply reflect the place where marginal revenue is equal to marginal cost for producers and marginal utility is set equal to price for consumers. The value of those goods or services is likely to exceed the total output measured by GDP anyway, even for market goods. My point is simply that anyone who knows how the market process works would have a problem with GDP numbers, period - private or public (and many people do go this far).

So why do we have GDP?

To answer macro questions, not micro questions. GDP measures can't answer value questions and they're misleading because of so many arenas where we depart from the canonical price mechanism - not just in government, but in the market as well.

They can tell you (with a few exceptions that I'll discuss below) how much money people paid for goods and services. They can tell you P*Y, and we can do a pretty decent job at isolating Y. Needless to say, that's an incredibly important thing to know when you're doing macroeconomics. It doesn't matter how P or Y were determined - those aggregates are necessary for certain types of analyses. This is what Keynes was getting at when he said that his economics could be better applied to analyzing a totalitarian system than classical economics. If you read the German preface it was abundantly clear that what he was saying was that neoclassical economics had quickly become useless to German economists addressing the command economy. And for all Keynes knew in 1936 there would be more command economies coming. Macroeconomics asks questions that are fundamentally different from the kind of questions that Higgs is trying to shoe-horn into GDP figures here.

So what are those exceptions? The biggest, of course, is the black market. Illegal drugs stand out the most in U.S. GDP figures, for example. Billions are spent each year on this product, people earn a living from it, people purchase it, money is used to purchase it. It belongs in any sort of macroeconomic calculations. And talk about liquidity preference! I would guess drug dealers have just about the highest liquidity preference of anyone else in the country! Other black market production in the money economy applies too, of course - paying workers under the table, etc. The figures are wrong insofar as they exclude this stuff.

People have mentioned home production in the past as well, but this one doesn't bother me as much, personally. So it would be nice to know how much output was produced in a home because that is real, valuable, production, right? A home-cooked meal, my wife's cross-stiching that we have hanging on the wall now, the beer I brew at home, etc.. All this stuff is valuable, and we make decisions about time allocation when we make these things, but I don't necessarily think it belongs in GDP. It's not a part of the monetary economy - it's not exchanged for money or other goods. So it doesn't seem relevant for answering macroeconomic questions.

I've thought a lot about many things I've written on this blog - the 1920-21 depression, externalities, the nature of liberty, the Austrian School, Keynes, etc.. I haven't thought as deeply on this question. So if anyone wants to challenge any of this please do - but I'm still nowhere near being convinced by Higgs.

17 comments:

  1. Why do you have to analyze the macro economy in aggregates? I know you prefer to, but why is it necessary to "isolate p"?

    I can come to a clear understanding of economics without aggregation.

    ReplyDelete
  2. When you are talking about specific markets or actors you of course don't need aggregates. When you want to say anything useful about money or a monetary economy, you need to know what that money is being exchanged for - you need an output figure.

    Think of it this way - let's just say you're thinking about a market for one product. You only need to quantities to think about that transaction - the quantity of the good, and the quantity of money you exchange for it. You don't need to think about other things. We don't barter, we live in a monetary economy, so talking about any given economic decision can be done with relatively little microdata.

    If you want to say something meaningful about a monetary economy, though, you need to know about everything that money is exchanged for - you need to know what the money stock is traded for in any given period.

    ReplyDelete
  3. "I can come to a clear understanding of economics without aggregation."

    To be honest, a lot of your lack of clarity I think inevitably derives from this hang-up of yours.

    ReplyDelete
  4. damn it - lost a long comment.

    You miss displacement if you rely on disaggregated data exclusively too. This was my point a couple weeks ago about the difference between Hayek and Bastiat's methodology.

    ReplyDelete
  5. I took the same Micro 101 course you did. Of course you need to look at the money stock and what's being traded in order to understand what exactly transpired.

    But it's a different beast in macroeconomics.

    You can't really look at employment "as a whole" or output "as a whole," much less reduce them to mathematical constants. All the traditional aggregates mainstream economists use are arbitrary. Why is a housewife cooking a meal for her husband not considered employment, but a maid servant doing the same is? Why is malinvestment (investment in projects that are toxic and unprofitable) considered as "output" when they help no one?

    To be honest, a lot of your lack of clarity I think inevitably derives from this hang-up of your

    What lack of clarity? Because I happen to think looking at demand side is irrelevant in economic depressions?

    ReplyDelete
  6. You miss displacement if you rely on disaggregated data exclusively too. This was my point a couple weeks ago about the difference between Hayek and Bastiat's methodology.

    But I'm not even looking at data. That's the problem. Economic theory is not subject to the collection of data. Understanding purposeful behavior (action) requires a study in logic. Neither aggregate nor disaggregated data can suffice.

    ReplyDelete
  7. But I'm not even looking at data. That's the problem.

    I couldn't agree more.

    I'm not running around the praxeology bush today.

    ReplyDelete
  8. I don't know what Higgs is suggesting we do about GDP figures. Clearly GDP is not measure of wealth per se. At best, it is a reasonable proxy when particular assumptions hold. Perhaps more care should be taken to clarify these matters, but surely that is not controversial. The popular media and politicians will continue to misuse and abuse such concepts regardless of how much economists complain.

    In any case, although GDP cannot measure the wealth created by government activities, our inability to effectively measure something does not make it disappear. It would be foolish to not account for such difficult to measure variables in some manner, rather than acting as though they didn't exist.

    ReplyDelete
  9. Not only is it not such a measure "per se" - it's not such a measure "at all".

    It's a measure of two things, depending on which angle you want to think about it: "sold output" and "income" - two sides of the same coin. It can measure both those things when the government is involved just fine. Government production for sale is measured with equal accuracy as private production for sale.

    ReplyDelete
  10. No, I think "as such" or "per se" is right. With particular assumptions satisfied, it can also be used a good proxy for wealth in most cases. And people have used it like that for many years and often with no problem. Higgs rightly points out that GDP can become a poor proxy when those assumptions are unsatisfied, such during World War II. One should be careful and critical when scrutinising GDP figures for that reason.

    ReplyDelete
  11. How about answering some of my questions, Daniel?

    ReplyDelete
  12. As for your questions:

    Why is a housewife cooking a meal for her husband not considered employment, but a maid servant doing the same is?

    Because a housewife is not paid for her services and her services are not contracted for.

    Why is malinvestment (investment in projects that are toxic and unprofitable) considered as "output" when they help no one?

    Because "output" isn't "things that help people". Otherwise we'd have to add in hugs and kisses and kind words and perhaps take out tanks and missiles. Output is sold production.

    What lack of clarity? Because I happen to think looking at demand side is irrelevant in economic depressions?

    That's just a bizarre objection to the simple facts of supply and demand. If demand shifts left, quantity decreases. If supply shifts left, quantity decreases. Anyone who doesn't think both are important to situations where output decreases either doesn't understand very basic economics or has some ingrained ideological reason to be allergic to any mention of "demand" - say, because there is a British economist that they instinctively dislike and just want to disagree with things he agreed with.


    And I answered your first two questions in earlier comments.

    ReplyDelete
  13. And here's a question for you - you just say this:

    "That's the problem. Economic theory is not subject to the collection of data. Understanding purposeful behavior (action) requires a study in logic. Neither aggregate nor disaggregated data can suffice."

    Yesterday I said I didn't feel like running around the praxeology bush. Today I suppose I'll have one go.

    You just claim this. You don't demonstrate it at all. Everytime I ask you to, you just say "the logic is very complex - go read Human Action". Could you actually explain why you think this? If you can't explain it, I'm not sure how you can believe something you can't explain.

    ReplyDelete
  14. I'm reading this on my phone. I'll come back later and comment for real.

    ReplyDelete
  15. I'm explaining my position on praxeology and economics in general in a post I'm writing right now. I promise I'm not ignoring that question.

    ReplyDelete
  16. Awesome. Your blog just ate a HUGE post of mine.

    ReplyDelete
  17. Not sure what the problem is, Anonymous - it happened to me too. I usually copy my post before posting it just in case.

    ReplyDelete

All anonymous comments will be deleted. Consistent pseudonyms are fine.