Real Gross Domestic Private Product, 2000–2012

In a new paper, Robert Higgs reports:

Until World War II and the postwar years, when the federal bureaucracy institutionalized the government’s preferred method for calculating national income, economists offered sound arguments for excluding government spending from estimates of gross domestic product. Using their general approach reveals that the private economy’s performance for the past thirteen years has been only somewhat better than complete stagnation.

I don’t think that zero counting of government consumption is the correct approach here, but this is nonetheless an interesting exercise.  Keep in mind that even if government outputs are highly useful, many of them are closer to intermediate than final products.  In other cases the output may be useful, and a final product, but not valued at actual market prices.  There is then still something to be learned by considering and segregating, if only temporarily, those parts of gdp which are sold at true market prices.

For the pointer I thank Daniel B. Klein.


How should we treat government purchases of intermediate goods -- a local government buying police cars or firetrucks or the Pentagon purchasing fighter jets --- that are equivalent to roughly 20% of government output?

For example if ford builds a truck and sell it to an individual it is private consumption. But if it sells it to a business it counts as investment and if it is sold to the government it counts as government o.utput

This creates a problem when trying to measure private GDP that I'm not sure anyone handles very well.

The flipside of the argument that government purchases are not valued at actual market prices is that G understates the value of output. Yet the tacit assumption in these conversations always seems to be the contrary (which may well be the case, granted).

Problem with segregating is C makes no sense without G, and this whole idea suggests G/GDP is a meaningless concept. Interesting, nonetheless, though I think we need a better framework to make sense of the number, and the paper seems simplistic on that front.

For example "However that may be, because so much of this spending may have had little or no value—or even negative value—in itself, the question remains as to whether, despite what the official GDP figures show, the population’s true economic well-being might have suffered a greater contraction than mainstream economists, journalists, policymakers, and others for the most part believe."

It's an exercise in "I believe government spending is useless therefore here's an other measure". And stagnation in this measure is clearly because of onerous government statues whose value I already subtracted.

Quote "when the measures taken under Herbert Hoover and Franklin D. Roosevelt turned
what probably would have been an ordinary, short-lived recession into the Great

Do you really believe this?

It's a really bad sentence because "what would have been" [the counterfactual] is not well-defined at all. Not only does he take some form of free market possibility as a foregone conclusion, the counterfactual here is impossible.

How does one define what would have been? If Herbert Hoover and FDR has passed no laws whatsoever? And what about the tight money Fed. Did FDRs unexpected devaluation make recession worse? Does the Fed exist in this counterfactual? Or would it have been repealed.

But you see spelling all that out would remove from the rhetorical flourish.

Is there any more useless intellectual exercise than speculating on counterfactuals? First, it's hard enough figuring out and describing how things actually happened. Second, once you move beyond the first step in the counterfactual - e.g., if the Pacific fleet carriers had been at Pearl Harbor, the battle would have gone differently - all you are doing is guessing, because changing one event will change a multitude of others in ways that cannot be predicted. (If they could be predicted, then it wouldn't be hard to figure out and describe how things happened in the first place.)

(A) Made it better (B) Had no effect (C) Made it worse.

On the other side, it's the worst recession ever. So, it's funny to me when people claim that anything about that recession or this recession anything that was done intentionally made it better. Pure dumb luck, I'd almost buy.

Of course, because it's true. I mean one can't prove a counterfactual, but all the evidence points that way.

I'm curious, did the $60 parking ticket I paid this morning add to national income or did it net out?

Sorry to go sideways, but I'm interested.

Probably depends on whether the plot is private or public, and in the latter case, whether it is run by a publicly owned company or the municipality.

If it is privately owned, they will probably book it as revenue and thus it is production.

If it is publicly owned (and not in a public company), it probably is a fee and seen as a tax.

The $60 kept you from spending $60 in "C," so that lowered GDP by $60. Meanwhile, your municipality has either $60 extra to spend or needs to borrow or tax $60 less to support the same spending (which would leave the lender with $60 extra to spend on "C." A Keynesian would say that GDP would expand, since the government has a higher MPC. In classical macro, it is a wash.

What if he borrowed $60 dollars to pay the ticket?

Should buying securities that were issued decades in the past count toward GDP?

After all, investing wages in securities is always claimed to boost job creation by driving consumption in capital which requires labor, so wages invested in Wall Street add to GDP by creating demand for production.

And is buying US Treasuries which are paid to private corporations for security state equipment and labor which are for security reasons almost always required to be US citizens, or paid to welfare recipients who are almost all so poor they spend it within two weeks on good and services, additive to GDP or not?

Of course, I guess one can simply blame the radical leftist Obama's socialist policies he rammed through Congress when he took office January 20, 2001, starting with his failed Keynesian stimulus bill (EGTRRA) plus two wars, all paid for with massive debt and printing money which subtracted from GDP. Everything bad is Obama's fault.

The only part of this that shows up in GDP is the brokerage commission.

Capital gains and/or loses are not included in GDP.

When an individual receives a social security check and uses it to buy groceries and pay other living expenses it is recorded in GDP as personal consumption expenditures, ie. it is private not government.

In the linked article it does not say how Higgs handles this..

Do you know how he handled this data problem in his full paper?

Of course one government provided service that's not included in GDP – but has huge financial and supply-side benefits – is issuance of liquid Treasury markets. Not for the sake of financing bridges and education or whatever (that too, but libertarians understandably can doubt this) but for maintaining an extremely efficient and stable financial market.

How can we calculate the financial value provided by Treasuries?

Not to mention the value of all the foreign invasions prevented by the military, or of the crime prevented by police, or of the deaths prevented by health inspectors, or of the Tea Party suppression by the IRS. But for purposes of GDP, the value of all that is zero, just as it is (nearly) for Wikipedia and the cleanest air and water civilized mankind has ever enjoyed. GDP measures expenditures, not value.

Karl Denninger has been making this argument for years. He approaches it a bit differently, but the result is the same. His argument is that government credit emissions have inflated GDP, disguising the stagnation of the private economy. If your growing at 3% and borrowing at 6%, it is not hard to make the argument. Even assuming a fair amount of the debt is domestically issued, some non-trivial amount is foreign or through central banks.

That's one of the central problems with GDP as a tool - its way too easy for governments to boost it artificially - at least for a while. For example, southern Europe.

Or Qatar or the UAE. Economists have become slaves to their tools. Reliance on GDP as a key measure is an excellent example.

For a correct analysis of these issues, see Murray N. Rothbard, "The Fallacy of the Public Sector," in which he points out that G is waste C.

To put it in Keynesian terms,

Y = C + I - (G + T),

where C + I = (private) production and G + T = (state) depredation.

There are two kinds of government "workers," New Class parasites, who do stuff that would be done in the private sector absent state intervention, such as deliver mail, administer law and courts, or teach in public schools (those that would not be shut down); and pure public parasites, who do stuff that would not be done in a free society, such as being a politician, president of the United States, a DEA agent, etc.

If it would be done by the private sector, then it is definitely part of GDP--- it has real value. Otherwise, the mere act of privatizing the post office would raise national output, which is silly since nothing real would have changed.

Good discussion, very clear, nice, I enjoy the comments here. One caveat is that there are some things that cannot be done by the private sector, such as public goods like national defense, or clean air that cannot be transacted (Coase th. stuff), or patents (or Big Science, where you need a certain incentive from government to reach a critical mass--IMO this is why we have a Great Stagnation now. If indeed science is getting harder, then you need more patent protection, not less like Alex T wants, to get society to do something other than create a more cool looking consumer product or trendy looking mobile phone [i.e., industrial design])

My rule of thumb is 10% of what government does adds to GDP, while the other 90% is either robbing Peter to pay Paul (neutral), or simply waste.

I suspect the point Rothbard was trying to make (and with which I agree) is that the tasks currently being performed by the government that would *not* be performed by the private sector are those that have no useful purpose. National defense fits this bill. If US military spending was $0, it would not increase the risk of foreign invasion by even 0.01%

@Ray Lopez,

In a free society there is no nation-state, so there is no national defense. There is also no such thing as public goods. Rothbard showed how air pollution can be handled privately in a Cato Journal article. Patents, which are state-granted monopolies (i.e. crookopolies) impede scientific progress. See Boldrin and Levine, Against Intellectual Monopoly. Government action only subtracts from "GDP."

Question: In a system of private courts, how does one compel non-parties to respond to subpoenas?

I read the screed "Boldrin and Levine, Against Intellectual Monopoly" and, as somebody familiar with patents, it struck me the same way as a modern astronomer reading a tract by the Flat Earth Society. It's nonsense, as is a lot of the Austrian propaganda. As for the "if US military spending was $0, it would not increase the risk of foreign invasion by even 0.01%" comment by MS, it's also fantasy by the Rothbard fanatics.

I was amused by the use of "screed" to dismiss Boldrin and Levine's article cited by Bill Stepp. I found a PDF of it and after reading about 3-4 pages, I conclude that it is not a "screed", but rather an even-tempered critique of the patent system. Without reading the whole thing, I do not know how much I would buy their arguments, but the tone and content seemed reasonable and interesting, based on the excerpt.

What about micro effects of robbing Peter to pay Paul? Not neutral.

Tom Palmer and Stephan Kinsella have written extensively on the patent-crookopoly system. A patent is a government-granted monopoly/crookopoly.

Let's imagine two countries. One is a country that has minimal gov't.

Y = C+I-(G+T) = big number

The other is a socialist gov't where nearly all production happens in gov't plants. Yet amazingly let's just say that this country produces the exact same bundle of goods and services produced by the free market. In this case,

Y = 0

Yet the people in country 2 have the exact same standard of living as country 1.

The libertarian response to this would be such a thing is impossible. No country that is completely socialized could possibly match the production of a free market one. That's fine but isn't it interest the above scheme stacks the deck? Even if a socialist country did match a free market one, the equation by definition makes it a loser.

The correct equation, Y=C+I+G (ignoring exports/imports for now) makes sense. If socialist economies are always bad, then their higher level of G will be offset by equal and greater reductions in C and I. But what's important is that is a position that is actually falesfiable. A free market advocate who uses the correct equation is willing to stand by his position. He is willing to risk someone coming forth with data that proves him wrong because he is confident no such data could ever be validly produced.

By interesting coincidence, the top marginal tax rate has been at or below 35% during three periods of time since 1917. In each case, by the fifth year, the economy turned pretty sour. The fifth year of the first such low-tax rate stretch was 1929, the start of the Great Depression. The fifth year of the second such stretch was 1992, the last year of the GHW Bush administration, a period of pretty mediocre growth. The fifth year of the third such stretch was 2007, the start of the Great Recession. A cynic might wonder why, if low tax rates are such a good thing, do the lowest tax rates tend to be accompanied by an economic mess. A cynic might also point out one big difference between 1929, 1992, and 2007, which is that tax rates rose in 1993.

You might want to look at .

The economy was rather bad during Carter, when the highest MTR was high enough to make Mike happy, while during Clinton the highest rate was only slightly higher than the magic threshold.

One difference between the high rates now vrs. then is the ratio of the income level at which the high rates engage and the average income. When the rates were high they were almost meaningless because they [by design] applied to very few people and because they were in effect compensated for with myriad loopholes.


Mr. King,

You might want to revisit the BEA's real GDP growth figures ( I am certainly no Carter apologist, but it bears noting - real GDP growth rates under Carter (1977 - 1980) were 4.6%, 5.6%, 3.1%, and -0.3%.

Now, for grins and giggles, let's compare that to growth rates during the period from, say, 1987 (when the top tax rate was reduced from 50% to 38.5%) to (let's cherry pick and leave out the Great Recession) 2006. The highest tax rates during that period occurred during the Clinton era, in which they were still just below 40%, and the lowest were 28%. The growth in real GDP during that happens to be 3.1%. If you think growth was bad during the Carter years, I assume you think was bad from 1987 to 2006 as well.

As to loopholes... I keep hearing about those, but having worked at a Big 4 accounting firm, I understand how such data can be manipulated. A cynic might suggest you look at real GDP growth rates and tax burdens and see if the low taxes = faster growth story fares better that way.

"There is then still something to be learned by considering and segregating, if only temporarily, those parts of gdp which are sold at true market prices."

Agree. Actually there are many cuts of GDP like final sales (ex inventories), domestic purchases (ex NX), etc. that are worth considering depending on the question. While GDP gets the spotlight, the national income and product accounts are full of amazing detail. And make no mistake, G is not the only place using non-market prices and imputed values. Large chunks of PCE services (health care, education, and financial services are not valued simply at the money that leaves consumers hands and goes to the provider). Services are hard to measure and some like education are probably as much an intermediate good as defense tech spending. The accounts continue to change as the economy changes. At the end of the month, the BEA is going to bring estimates of intellectual property into GDP.

I applaud the exercise in the paper to think about what we mean by real activity ... what can tell us about health (or sickness) of the economy. I spent all day staring at GDP and many of its sub-aggregates, not all estimates are created equal (or have the same signal) in the accounts and anyone who uses the data should know that. The best measure depends on the question ... and not political affinities.

Indeed, the 1.7% average growth rate of Gross Dometic Private Expenditure for 2000-2012 is stunning, especially when you compare it to the clearly inflated GDP average growth rate for the same period of... 1.6%. Wait, uhm...

But then, you might argue the crux of his argument is about how this measure shows the stagnation of the US economy much more clearly. Calculating the average growth rate for 1973-2000 (the closest I got to Higgs' 3% long-run trend) gives GDPE a 3.4% trend, and a 3.1% GDP trend. So, GDPE shows a 1.7% decrease in trend and GDP a 1.5% decrease. Once again, I fail to see the difference.

Now, I'm all for analyzing the components of GDP (be they private absorption, investment and GDP ex-changes in inventories etc.), but I don't think this is particularly helpful if you don't bother to, you know, actually look at the data.

Well obviously completely ignoring the public sector is stupid, as it ignores the opportunity cost of resources. Imagine the private sector continually gaining in productivity, whilst inputs are continually reallocated politically to the public sector; you'd see stagnation in "private RGDP".

Interesting analysis. Capitalism would be dead if not for massive government spending. Every morning, noon and night the monetarists on Wall Street should get on their knees and pray toward Keynes' place of burial.

I enjoyed reading the comments of people who find this exercise senseless. Now you know how the Austrians feel when they are compelled to make sense of GDP. Are any of these approaches any better than any other? Does anyone actually believe that "Y=C+I+G+NX" means anything?

yes it means the total goods and services produced in an economy during a period of time. Should we care how Austrians 'feel'? Why?

You're free to declare you don't like some types of goods and services. You can, if you wish, calculate GDP produced only by left handed people, or blue eyed people. But here's the thing, GDP works with the concept that there's a finite amount that an economy can produce in a year. That goes whether you're a socialized economy producing tanks and secret police reports or a free enterprise one making hamburgers and iphone apps.

Easy there, tiger. GDP attempts to measure the value of all goods and services produced in an economy during a period of time. Whether GDP actually means that is a completely different question.

It's an interesting point. Zero probably isn't representative, but then no one really knows what is. Of course "real" has a lot of problems too.

Comments for this post are closed