Here we go again….  I just posted about the broken window fallacy on Saturday, and here comes Eric Zencey writing in the NY Times trying to claim that “If you get into a fender-bender and have your car fixed, G.D.P. goes up” and that “Hurricane Katrina produced something like $82 billion in damages in New Orleans, and as the destruction there is remedied, G.D.P. goes up.”

According to the blurb at the end of the article, Eric Zencey is “a professor of historical and political studies at Empire State College”.   Now I don’t know where  Empire State College is (I’m guessing New York) or what it may have as a claim to fame, but basic economics – or common sense – evidently isn’t a prerequisite in order to be a professor at the institution.

According to Mr. Zencey “If you get into a fender-bender and have your car fixed, G.D.P. goes up.” The only way that makes sense is to assume that the person who got the bent fender just happened to have the money to repair that fender lying around.  And by “lying around” I mean that literally.

If it happens to be invested in company stock, it’s already in the GDP.  If it happens to be in the bank in a savings account, it’s already part of the GDP.  Only if it’s in a mason jar buried in your backyard (or the equivalent) could the money you spent repairing a fender be included in the GDP.  Here’s why.

From Wikipedia – The GDP is calculated as: GDP = private consumption + gross investment + government spending + (exports − imports)

The money to fix the fender HAS to come from somewhere.  Let’s say it cost $1000 to fix.  If that $1000 came from your savings account, it decreases the investment amount by $1000 and the net change to the GDP is zero.

Does this change if you have insurance and the insurance company pays to repair your fender?  No.   Where did they get the money?  It was invested – insurance companies call the money you’ve paid them (but that they haven’t yet had to pay out) “float” and Warren Buffett got to be a billionaire by investing that float in places other than your fender.  All his billions are invested somewhere, so when Geico pays to fix your fender, it’s still subtracting from the overall investment total.  The NET EFFECT IS ZERO.

The same is true if the government spends more money with a “stimulus package”.   The only way GDP can increase via government spending is if the money didn’t come from another category.  But government money comes from higher taxes – which reduces the “private consumption” part of the equation – or it can come from borrowed money, which reduces the “gross investment” and/or increases the “import” (depending on where you borrow the money from) part of the equation.

Government spending only appears to increase GDP if it was previously collected but not spent.  In that case it transferred the spending from one time frame to another.  Government spending can increase the current GDP only at the expense of decreasing the GDP at another time.  If the government spent money it had saved, it decreased the earlier GDP.  It the government borrowed money and spent it, it’s decreasing future GDP.  This ain’t rocket surgery.

Money was simply moved from an investment where (it hopefully) could be used to create something valuable – to somewhere where it HAD to be used in order to repair an asset which had already been purchased.

It’s a bit crass, but think of it this way – if destroying an asset (via fender bender or hurricane or earthquake or nuclear war or whatever) CREATED wealth, here’s a solution to all our problems.


Why wouldn’t that work?  Got an answer for that Professor Zencey?

Cash for clunkers is the same concept.  Taking money from one part of the economy and spending it in a different part has a net effect of ZERO.  It may appear that you’ve got a bump in GDP for awhile, but that’s simply because you haven’t had time to see the effect of the loss in the other area.   Every government dollar spent buying a clunker is a dollar that isn’t spent providing health care, or building bridges or highways, or defending the country, or building a subway, or whatever you think the government should be spending money on.

The net effect is still zero, because every dollar the government spends has to come from somewhere – and “somewhere” means taxpayers.  And that’s why government spending (unless it’s from unspent tax surpluses of which the US has none) doesn’t actually do anything.



2 Responses

  1. [...] of ranting about the “cash for clunkers”-insanity, I will link to this great post from Effor on why you can’t stimulate the economy or [...]

  2. You are right that fiscal stimulus cannot increase the economy’s long-term growth trajectory, or cause the economy to expand beyond full capacity. But your conclusion about the value of fiscal stimulus is based on the static assumption that the economy is at full capacity.

    The broken windows fallacy does not apply to fiscal stimulus in recession, or to put it another way, it is not a fallacy in recessions. By definition, when the economy is below capacity, the supply of savings outstrips investment demand. This is what causes the economy to contract. In essence, there is excess money held in cash, whether in the bank or under the mattress.

    When you stimulate the economy by fixing stuff, then the money comes out of the mattress and into goods and services. If the value of the stimulated economic activity is zero — i.e. if you pay people to dig ditches and fill them back in — then you will have increased economic activity but done nothing to improve national wealth. But if you pay people to do useful things, like fix bridges, then national wealth is increased.

    Another way to look at it is this. Suppose that, during a boom when the economy is at full employment, the government takes some money and pays people to build a building. Well, that’s great, we have one more building . . . but one less of whatever it was those guys would have been doing. But when unemployment is 10%, we would get one more building . . . and we would lose nothing, because nothing is what those unemployed people are sitting around doing.

    As for Cars for Clunkers, obviously it doesn’t increase economic growth over the long term. But it does employ people to do things when those people otherwise would not be doing anything. So in the short run it increases national wealth. . . except for the part about destroying the old cars. Whether you think this is a good program depends on what you think about the value of those old cars. If you think the gas mileage improvements are significant (and while they don’t have to be by law, evidence suggests that they are), then the value gain from a new car exceeds the loss of the old car.

%d bloggers like this: