GDP is a flawed measure of economic welfare. So what?

The Economist indulges in a bit of hand-wringing about the exhaustively-documented weaknesses of using GDP to measure economic well-being:

It's high time that economists looked at more than just GDP: Economists spend much time discussing how to boost GDP growth. The OECD itself drew attention this week to the widening gap between America's and Europe's GDP per head. Yet a nation's well-being depends on many factors ignored by GDP, such as leisure time, income inequality and the quality of the environment. GDP was developed primarily as a planning tool to guide the huge production effort of the second world war. It was never intended to be the definitive yardstick of economic welfare. Would another indicator change the ranking of countries or their relative performance over time?

The answer to that last question is 'almost certainly'. What's less clear is why the question is even of interest. Offhand, I can't think of any policy question that would be affected by a mismeasurement of the magnitude - or even the sign - of the difference between economic well-being in Canada and in the US.

And of course economists do look at more than GDP - again, I can't think of a policy matter where discussion revolves entirely around the GDP numbers.

Other data can and should be used to supplement GDP in discussing trends in economic welfare. But any attempt to put together a single index of well-being - why not call it a Social Welfare Function? - is doomed by Arrow's Impossibility Theorem to be an exercise in futility. How would we trade off an increase in (say) a given measure of income inequality against an increase in another indicator of public health or a decrease in air quality? And why would that tradeoff be constant?

Economists should, can and indeed do look at other data when evaluating economic welfare. But trying to come up with a single index for economic welfare is not a high-priority claim on our time.

(See also this post.)

Comments

[Mandos]

Except that well-being is relative.

[thwap]

Can we then down-grade the "goddamm gdp"? After all, if everyone knows it's just one indicator out of several, why should we be so concerned that "the economy" is "growing"?

[Stephen Gordon]

Because it's still a pretty useful indicator: it's broad-based, and it's published on a timely basis. And it's still *generally* the case that that short-run movements in GDP reflect fluctuations in welfare - those other factors generally move much more slowly.

It's not perfect, but it's still pretty useful.

[Declan]

I don't know. It seems to me that the all-pervasive tracking of GDP favours policies which increase GDP (say by pushing non-market activities into the market, or by privatizing government functions) even where these policies aren't optimal for society.

While it may be hard to point to any specific case and say this decision was made wrong due to focus on GDP, over time I think this has a negative impact. For example, how many people have advocated that Canada adopt American policies because they have a higher GDP per capita than we do, while failing to allow for the differences in GDP caused by a more market-based society and a population that works longer hours and takes fewer holidays. If you really think that no policy decisions would have been made differently over the last couple of decades if Canada was consistently believed (measured) to be outperforming the U.S. economically (e.g. if we had reversed the sign of the difference in GDP), then I think you are you are mistaken - hard to prove of course!

I also think back to how many Americans have argued against moving to a more socialist model based on the 'fact' that the GDP/capita (and hence the standard of living in many minds) of Missouri was higher than that of Sweden. Or consider proposals for adding more statutory holidays. I think in the back of some people's minds would be concerns about falling behind in GDP. Generally, I think that what gets measured and what gets treated as our prime objective has an impact.

Economists may look to other measures for evaluating economic welfare but I'm not sure many other people do.

Also, just because there are theoretical issues that prevent coming up with a perfect alternative measure, that doesn't mean that a more useful alternative measure couldn't be constructed. How does one trade off artistic impression against technical prowess in figure skating? No idea, but weighting them equally is likely a better solution than ignoring either one entirely.

Good blog, by the way.

[Stephen Gordon]

Thanks.

But I'm going to stick to my main point: for policy purposes, the measurement issues involved with GDP just don't matter that much, at least, not to economists. It's certainly the case that you'll see GDP numbers being presented in many weird and wacky ways in order to back up some point or other, but the real problem there isn't GDP: it's bad economics reporting. Of the two problems, I'd say GDP measurement issues rank a distant second behind that of the quality - or lack thereof - of popular economics commentary.