The Economist is hosting a 10-day debate of the proposition:

This house believes that GDP growth is a poor measure of improving living standards.

Opening statements:


Featured guest: Enrico Giovanni, President, Italian Statistical Office

Featured guest: Michael Boskin, T.M. Friedman Professor of Economics; Senior Fellow, Hoover Institution, Stanford

Featured guest: Me

Closing statements:

Thanks to the Economist for including me in the debate.

On day 1 of the debate, 69% of voters agreed with the motion, 31% disagreed. Agree is now up to 72%. I don’t know whether that 70% range is primarily a function of the quality of the arguments or instead of the initial views of those who chose to vote.

Here is my statement.

Gross domestic product is of course an imperfect measure of improving living standards, primarily because it is incomplete. By excluding changes in non-market goods like clean air and water, GDP measures the market value of goods and services produced within a nation, but excludes many important outputs that are not owned, traded, or easily and objectively valued. The “P” stands for product, so GDP is an output measure, whereas living standards are in part a function of the goods and services we consume and of the income generated by those we produce.

GDP does not measure happiness, or well-being, or what economists call utility. As a gross measure, it aggregates data for a geographic area, ignoring important distributional questions and individual preferences. As a flow measure, it does not account for the value of a nation’s stock of assets and liabilities.

Andrew Oswald argues: “GDP is too narrow a measure of the things that truly matter to humans to be viewed as a valuable indicator in developed nations like ours in 2010.” I shudder to imagine who might assign themselves the role of determining what truly matters to all mankind. Yet the proposition is not whether GDP is a valuable indicator of “things that truly matter to humans”. The proposition is not limited to rich nations or to the present. The proposition is that GDP growth is a poor measure of improving living standards. I oppose that proposition.

Mr Oswald argues that we should measure well-being and include both subjective surveys and sustainability as components of that measurement. I have no quarrel with measuring well-being, but there is no reason to foul up a useful statistic in doing so. GDP is but one indicator that policymakers can and should use to analyse the economic health of a nation, and it is foolish either to use it for a purpose for which it was not intended, or to attempt to change it to suit one’s policy goals. A doctor who monitors only a patient’s pulse is not doing his job, but one who argues we should ask the patient how he feels and call that his “pulse” is outright dangerous.

A patient’s pulse is useful in part because it is an easily measurable and objective metric that is comparable over time and across patients. GDP is quantifiable – it is simply an accounting measure. GDP is objective – we can rely on the data even when personnel in the statistics office change or the party in power flips. GDP is, within limits, roughly comparable across nations and over time, allowing us to make imperfect but still useful policy comparisons and judgments. And since wealthier societies generally devote some of their increased resources to improving non-market attributes like clean air and water, GDP is only partially incomplete as a measure of the non-tradable aspects of improved living standards.

I would rather live in the Turks and Caicos Islands than in Iran, even though the latter has a higher per person GDP. But if you ask me in which of two unlabelled countries I want to live, and if I know only their per person GDP, I will choose the higher one because it probably has a higher standard of living. Similarly, if you tell me only that a country’s GDP has grown 10% and ask me if the standard of living has improved, I will almost always be right if I guess that it has. Higher GDP means more tradable resources for individuals and governments with which to improve standards of living. Economic growth is good, and more economic growth is better. Ask a family in a poor African nation whether they agree with Mr Oswald’s conclusion that man needs “to make fewer things rather than more”, and whether they need to value tranquil beauty more and a car less.

Money cannot buy happiness, and GDP cannot measure it. But as a measure of improving living standards, it is both adequate and superior to subjectively defined, internationally incomparable and time-inconsistent measures of happiness based on someone’s subjective decision about how you should measure your happiness.

(photo credit: JoA in an argument by Anders V)