Sorry for the delay, folks; my power lunch ran longer than expected, and so this post is coming a few hours later than I had intended. I hope it will be worth the wait. One final caveat: This post will be a bit longer and more academic than most, but I really want to get to the heart of this issue.

In this post I want to pinpoint exactly what is wrong with the typical financial reporting--and it's not really the reporters' fault, they're just quoting the mainstream economists!--when it comes to the strength of the economy. We always hear the mantra that consumer spending represents some huge component (70 percent at the end of 2006) of GDP, and therefore is the key determinant of economic growth. Lately, we've heard things like, "The one silver lining has been the fall of the dollar, which has boosted exports and kept the economy out of recession."

This is all palpable nonsense.

What has happened here is a classic confusion of correlation with causation. Ex post, if you want to measure the "total gross output" of an economy, it sort of makes sense to count up how much money people spent on stuff produced within the United States. (Even here there are serious methodological issues, but those aren't relevant for the present post.) You can't simply count up all the physical things being produced, because that would just give you an immense vector of different items: x amount of iPods, y amount of golden delicious apples, etc. There would be no way to just look at the gross physical output from one year to the next, and be able to say, "The economy grew 3.4% in real terms over the year."

So the way economists get around this issue of aggregation is to use the money prices of the goods as a way to compare their relative importance. So if one more unit of something is produced that a buyer is willing to spend $100 on, that counts as twice as much "output" as one more unit of something that a buyer is willing to spend $50 on. Again, there are methodological problems here, but let's ignore them to concentrate on an even more blatant mistake in the conventional approach.

Every transaction involves an expenditure and an equal receipt of income. That is, when you buy a TV for $100, that's $100 more in "consumer spending" but it is also $100 more in "business revenue." For various reasons--including accessibility of the data--macroeconomists focus on the expenditure side of the coin, and sum up all the expenditures in the economy in order to measure gross output for the period.

In an ex post fashion, there is nothing intrinsically wrong with this choice; any philosophical problems you had with it, would (I believe) also apply to measuring income.

However, the fundamental mistake comes in when people stop using the expenditure approach as a backward-looking bookkeeping trick, and instead use it as a forward-looking, causal theory of GDP growth. That is to say, the act of people spending money on consumer goods does not constitute economic growth.

This is so obvious that I feel funny spelling it out, but somebody needs to do this once in a while to remind everyone how ridiculous the standard commentary on GDP and recession is. If you want to boost economic growth, go work more, produce more stuff. Don't go to the mall and buy things; that's consuming, it's not producing.

Although what I just said is a truism, the reason it's a bit subtle is that (as I explained above) you really do need a link to consumer spending, in order to quantify how much your "production" is worth. If you go out and bust your butt for 8 hours "producing," you really might not be very productive in an economic sense, if nobody wants to buy whatever it is that you made. So in a sense, it's true that if nobody spent a penny on anything made within the US, then we could say "nothing" was produced that year (unless we bring in a timing issue, where people produce stuff that they plan on selling in the future). This subtlety notwithstanding, it should still be clear that consumer spending per se does not constitute production, or contribute to economic growth. It is rather a reflection of economic output.

When I tried to get this point across to my macro students, the best example I could think of was the component of depreciation. Here is the standard formula for GDP:

(1) Y = C + I + G + (X-M),

where X is exports and M is imports. However, we can rewrite this equation as

(2) Y = C + (N+D) + G + (X-M),

where N is net investment and D is depreciation. Back in equation (1), "I" stood for gross investment, and net investment is gross minus depreciation, so that's why I = (N+D).

OK now just stare for a minute at equation (1). In the typical press discussions of the economy, they would say things like, "Consumers are getting increasingly worried about winter heating costs and gasoline prices, and so C is falling. That's bad for total output, Y, so we might enter a recession. On the other hand, the weak dollar is boosting exports X, and reducing imports M, so that is tending to keep Y up, keeping us out of recession." (Of course they don't use the letters, but I'm trying to tie the verbal discussion into equation (1) above.)

As anyone with a background in old-school economics knows, this type of talk can't be right; how in the heck is it "good for the economy" if people go spend like sailors, or if the dollar falls? That doesn't make any sense. But what specifically is the mistake?

Here's where equation (2) comes in. I'm going to employ the same reasoning, but with a new variable. I'm hoping that in this new context--where it hasn't been hammered over your head by "expert" PhDs for your whole life--the absurdity will jump right out at you.

Looking at equation (2) above, suppose someone said, "Well Maria, the economy is experiencing tough times. Consumer spending is down, government spending is down, and net investment is down. However, the one silver lining in all this, is that because of a new type of road salt recently adopted, all of the nation's tractor trailers are wearing out much more quickly than anticipated. Trucks that were thought to have 100,000 more miles in their useful life, are now expected only to eke out another 25,000 miles. Because of this sharp increase in depreciation, the economy might just barely skirt recession. Treasury Secretary Paulson is in fact calling on local governments even in southern states, where there is no snow, to begin salting their roads to promote this growth spurt from depreciation."

OK, does everyone see how I used the exact same logic as the talking heads on CNBC, to "prove" that raising depreciation boosts economic output? Something is clearly crazy here.

I'll end the suspense. Specifically, the problem is that people look at that accounting tautology--Y = C + (N+D) + G + (X-M)--and then mistakenly assume that if they increase a variable on the right-hand side, then the way the equation remains true is that the variable Y on the left-hand side goes up accordingly.

But duh, that's not the only way to balance the equation. In the salt example, what happens is that if D goes up, then N goes down, so that the equation is still true. Gross output isn't affected, but the capital stock grows less than otherwise, because more of the gross investment I (from eq. 1) is being sucked up by the wearing away of the tractor trailers.

It's similar with the nonsense over the rebate checks. If people are good little citizens and do what Paulson wants, then yes C goes up. But that doesn't mean Y has to go up, instead it can just push down I (in eq 1) or N (in eq 2). Duh! If you pay down your debts with your stimulus check, then that gives more loanable funds to the financial sector so some businessperson can borrow it and expand his operations.

Finally, what about the falling dollar? Again, just looking at the equation: If X goes up and M goes down, this isn't necessarily balanced by a rise in Y. On the contrary, it could be offset by a fall in C and/or a fall in I. This should be obvious; a falling dollar makes foreign imports more expensive, and so of course Americans can't consume as much (because part of what they consume is imported goods!). A falling dollar also makes inputs more expensive for US businesses, like, oh I don't know, those who rely on petroleum products. (Duh!)

In conclusion, the GDP accounting tautology is true, if we ignore serious methodological problems. But because of Keynesian theory (not a tautology!), the causal relationships between the variables in the tautology are assumed to work in a particular direction. That is, it is Keynesian theory that says if you increase C (or G), the result will be an increase in Y. This is consistent with the tautology, but is not required by it.

Classical (and modern Austrian) economic theory would say that in general, when people save more, this lowers interest rates and leads to higher investment, and so doesn't affect Y in the near-term. With freely floating wage rates, it doesn't matter how much of their income people "spend" versus how much they save. If consumers start saving 25% of their income, the relative prices and wage rates adjust so that teenagers who used to flip burgers and tear movie ticket stubs, are now working in factories cranking out tractors and drill presses. Total gross output (which includes both consumer and capital goods) doesn't need to fall, and in fact will grow at a faster rate year after year, the higher the savings rate and hence the higher net investment is every period.

Alas, when it comes to the financial press--and even free market economists who ought to know better--"we are all Keynesians now."