David R. Henderson  

GDP Fetishism

African Poverty Declines... Reflections on The Baader-M...

Econlib's featured article this month is GDP Fetishism by me.

Opening graf:

When economics professors teach the basics of Gross Domestic Product (GDP), we usually caution our students that it is not a good measure of welfare. Unfortunately, many economists go on to give GDP far more credit than it deserves. They tend to consider fiscal and monetary policy positive if these policies increase GDP, but they often fail to ask, let alone answer, whether those same policies increase or reduce welfare. I have a term for giving GDP such a sacred a place in economists' reasoning: GDP fetishism. If we return to some basic principles of economics, we will avoid GDP fetishism, do better economic analysis, and propose better policies.

An explanation about how I've categorized this post. I've put it under Macroeconomics for obvious reasons but also under Microeconomics because all my criticisms are based on well-accepted microeconomic principles. My main point is that the careless use of GDP ignores some of these principles.

Thanks to Jeff Hummel and Bob Murphy for looking at earlier drafts and providing comments.

Comments and Sharing

COMMENTS (11 to date)
Nathan Smith writes:

I think GDP is a better guide in the short run than in the long run. For example, estimates of GDP greatly overstate the improvement of human welfare over the long run, because so much of human happiness comes from free goods like sunlight and fresh air and the beauties of nature, which modern people sometimes enjoy less of because of living in cities (though it's complicated) and goods that can't be priced, like love, or courage, or telling the truth.

There are certain periods of history where GDP is a bad guide. For example, macroeconomists tend to make the mistake of thinking that US economic performance during WWII is a vindication of Keynesianism. But surely WWII wasn't a good time for human welfare, even if GDP rose fast. The macroeconomic lesson of WWII is that massive government spending that is purely destructive but has a credibly limited time horizon, under conditions of price controls, and with some labor conscripted, can lead to a rise in economic activity as measured by GDP, with little or no corresponding rise in human welfare, or more likely, a setback to human welfare. This isn't a precedent which it is sensible to try to repeat, in any way.

Bill Woolsey writes:


I think GDP figures (per capita) are the best information we have for "well being." Of course, changes in leisure, environmental quality, and the size and scope of government should be considered.

For example, a program for compulsory work on Saturdays increased GDP per capital by 20%! Oh, but everyone has to work 6 days a week.

The draconian environmental protection program reduced real GDP per capita by 50%! But the air and water are pristine.

Our $10,000 lump sum tax to build the giant pyramids raised per capita GDP! Oh, but private consumption and investment dropped and everyone worked one weekend a month to pay taxes. And all we got was a pyramid. Hopefully Obama will be happy in the afterlife.

On the other hand, when these things aren't changing much, the GDP figures do provide useful information about macroeconomic fluctuations.

Real GDP dropped very sharply in late 2008 and early 2009. Perhaps everyone just decided to take time off. Oh, there was presumably slightly less industrial pollution. Hey, maybe the furlough of county employees meant that they were doing a bit less harmful regulation.

But figuring out exactly how recessions impact human well-being isn't really my primary interest. Why do these decreases in production occur? What was going on during that time?

Production began to rise again during the summer of 2009. Why? It remains below its peak value from the summer of 2008, even now. It is way below its trend growth path.


Again, I am sure there are many silver linings from the depressed value of output. It is even possible that there is no problem at all. Counterproductive government programs ended and everyone responded to the reduced tax burden by taking off an hour of work each day.

But I am pretty sure that wasn't it.

I think your microeconomic points about how the unemployed enjoy leisure during recessions is almost offensive. (True, but offensive.)

Having people dig ditches and fill them in is wasteful. Just give the unemployed people the money and let them play. Or, put them to work doing something with positive value.

Finally, your analysis of inflation is very faulty.

If you assume that prices and wages are perfectly flexible, then the real quantity of money always equals the demand for real purchasing power.

While it is always true that an increase in the quantity of money, ceteris paribus, raises the equilibrium price level, if the equilibrium price level is below the current price level, it doesn't necessary raise today's price level compared to today.

Your analysis of inflation either assumed that the price level is always at equilibrium, or else, played on an ambiguity between the price level being higher than it eventually would have been, rather than higher than it already was.

Nominal expenditure in the economy today remains below its peak and nearly 10% below its long run growth path. Real expenditures could have been maintained. It just would have required that all prices and wages drop 10% below their long run trends. They didn't.

mobile writes:

Many metrics of general welfare have been tried and will be tried in this world of sin and woe. No one pretends that Gross Domestic Product is perfect or all-wise.

Indeed, it has been said that GDP is the worst metric there is for measuring general welfare except for all those other metrics that have been tried from time to time.

David R. Henderson writes:

What I notice about the comments so far is that except for Nathan Smith's, they are not very responsive to my particular criticism. I wasn't saying GDP is useless, as a careful reading of my article will show. Rather, I said that one can make mistakes with it and I even gave two examples. The discussion would be furthered if someone said why he/she thinks those two examples are wrong.
I also don't see why it's offensive or almost offensive to suggest that most people who are out of work would willingly work for one cent per day. I'm imputing some value of their time, which, to me, seems complimentary.

Joe Cushing writes:

I just heard a show on NPR last night that refutes your claim that GDP is a poor measure of welfare. They talked about Frances effort to come up with a better welfare measure. Then they pointed out that GDP is highly correlated with anything you might measure to measure people's welfare. They also said that a doubling of a rich country's GDP had the same impact on increasing welfare as did doubling a poor country's GDP according to survey data. It turns out that having money allows people to enjoy all of the things that people measure as welfare --even free time.

David R. Henderson writes:

@Joe Cushing,
Nothing you quoted from the NPR show refutes the claims I made.

Fred Thompson writes:

Spending on goods and services by government is included in Gross Domestic Spending (GDP) as equal to the cost of production, or factor cost. In contrast, spending on consumer spending is added to GDP based on the market value of the goods and services purchased. If government stimulus spending goes down a proverbial rat hole, it nevertheless boosts nominal GDP by the amount of the spending increase. For instance, if a government embarks on the building of a pyramid, which adds absolutely nothing to the well-being of individuals, the GDP framework will regard this as economic growth. In reality the building of the pyramid will divert real funding from productive activities, thereby stifling the production of wealth. The measured increase in GDP due to an increase in government spending overstates the true impact on the economy and the subsequent well-being of its citizens.

I absolutely agree with David.

Bill Woolsey writes:

There is nothing offensive about stating that unemployed workers value their leisure. What is offensive is the implication that the plight of the unemployed should be discounted because they are enjoying leisure.

(And, yes, make work projects are stupid.)

I see nothing wrong with your examples of how GDP fails to measure well-being. They are correct. I think I listed other examples where changes in GDP would fail to measure wellbeing.

Oddly enough, I think that the high per capita GDP of the U.S. does tell us something even if government spending portions are wasted in promoting world empire. While the GDP figures give a poor estimate of the well-being generated by government spending, the "cost" is an estimate of the private goods and services sacrificed. The U.S. economy produces a lot of goods and services measured in terms of valuable private goods and services. There are the ones produced, and then there are the ones sacrificed to provide government services of doubtful value.

If you are interested in comparing the productivity of different countries or changes over time, then the GDP figures tell you something useful about it, even if government is grabbing it and wasting it.

But this value of GDP figures pales relative to the value in trying to understand the coordination failures that generate short run macroeconomic fluctuations. The silver lining aspects to recessions (less pollution, more housework, more rest and relaxation) exist, but what does it have to do with the problem at hand?

If, on the other hand, you are thinking about changes in leisure, or different environmental regimes, or shifts in the private/public mix of goods, using GDP as a measure of well being would be insane. (Economic growth destroyed the enviroment..but GDP went up? Everyone converted to buddism and now live simply and enjoy life. But GDP went down.)

P.S. An expansion in the quantity of money sufficient to return nominal expenditure to its long term growth path might result in somewhat higher inflation for a time, and during this period the liquidity services that people get from holding currency (as opposed to interest bearing deposits) might be depressed. I rather hope it would motivate them to hold less currency. However, thinking about some kind of long run inflationary equilibrium--a permanent 5% inflation or something, and how that would create an excess burden because too little currency would be held, has nothing to do with the problem at hand. In fact, I would say that the problem at hand is that the demand for currency is too high because the nominal yield is too high, and that market efficiency requires a negative yield on perfectly liquid and perfectly safe assets when the real investment projects that generate yields take time and involve risk.

spencer writes:

You claim that the TSA has a value of zero.

But the fact is that there has not been a US airplane hijacked since the TSA took over airline security from the private sector.

Do you really believe that the 100% absence of airplane hijackings has zero value?

David R. Henderson writes:

@Bill Woolsey,
"There is nothing offensive about stating that unemployed workers value their leisure. What is offensive is the implication that the plight of the unemployed should be discounted because they are enjoying leisure."
So, Bill, do you think we should do cost benefit analysis by systematically excluding some benefits?

"But the fact is that there has not been a US airplane hijacked since the TSA took over airline security from the private sector.
Do you really believe that the 100% absence of airplane hijackings has zero value?"
No. I believe that you've made a non sequitur. You've attributed the absence of hijackings to the TSA. I attribute it to the fact that we've virtually all learned that the payoff matrix has changed. We won't sit passively by and let someone hijack or blow up the plane. Notice what I said in the article:
"the examples we have of terrorists on airplanes being defeated are all examples of victory by fellow passengers."

Allen writes:

There are definitely some serious problems with the concept of GDP. The phrase "GDP Fetishism" is a good way to put it considering the dichotomy between how serious the issues are with GDP, and just how much many people continue to rely on GDP anyway. I also wrote about some of these issues in February:


Comments for this entry have been closed
Return to top