It’s time to cut the GDP some slack. Overhauling the GDP — as some critics would — threatens to politicize one of our most useful economic indicators. It could be twisted to advance or retard political agendas. This is a bad idea.
First, some background.
“GDP” stands for “gross domestic product,” and it’s our standard measure of economic growth. When we hear that the economy grew 2.4 percent in 2015, it really means that GDP grew 2.4 percent. GDP has become a punching bag for pundits, economists, environmentalists and others who claim that it is incomplete and misleading. The respected Economist magazine recently joined the slugfest. It editorialized:
“This one number [GDP] has become shorthand for material well-being, even though it is a deeply flawed gauge of prosperity, and getting worse all the time.”
The trouble with this indictment is that GDP was never intended to be — and shouldn’t be — a measure of general prosperity, which depends on many conditions (a few: job stability, income distribution, health). Prosperity can’t be compressed into a single number. The GDP’s mission is more modest, though still difficult: to measure the economy’s production.
By this standard, GDP is a wonder of the age. Just think: one number that describes something as byzantine as the $18 trillion U.S. economy. The GDP divides into useful segments: consumer spending, business investment, housing construction, government spending and foreign trade.
We take all this for granted, but it wasn’t always so.
In the 1920s, the GDP didn’t exist. To divine the economy, businessmen, investors and government officials relied on many indicators, including railroad boxcar loadings, pig iron output, bank deposits and crop yields. Some entrepreneurs — Roger Babson was the most famous — combined individual indicators into a single index that supposedly predicted the economy. The index “was a hodge-podge of numbers,” says historian Walter Friedman, author of “Fortune Tellers : The Story of America’s First Economic Forecasters.”
The impetus for something better came from the Great Depression and, especially, World War II. To produce more war goods, how much did civilian industry need to be cut? Being able to measure the economy made deciding easier. In 1947, the government began publishing regular GDP estimates. (The initial emphasis was on the closely related “gross national product,” GNP.)
It remains a complex process. To estimate GDP, the Bureau of Economic Analysis (BEA) collects data from more than 300 sources, says Associate Director Brent R. Moulton. The Census Bureau compiles restaurant sales based on regular surveys; the airlines’ trade association supplies travel data; auto sales come from private data providers.
The GDP now poses three major problems.
First is the shift from a manufacturing to a service economy. “It’s relatively straightforward to measure economic output when you can count the number of cars or refrigerators . . . being shipped from factories,” writes Diane Coyle in her book “GDP: A Brief but Affectionate History.” “But how do you measure the output of nurses, accountants, garden designers . . . ?”
The answer: not easily. Particularly difficult are health care (18 percent of GDP) and education (7 percent of GDP). Together, they constitute a quarter of the economy, but we don’t have good measures of whether (or how much) their spending makes people healthier or better educated. The government essentially equates their output with their inflation-adjusted spending. It can’t judge whether the spending is effective.
The second problem is that GDP measures only market output. If you do your own housework, the effort doesn’t count in the GDP; but if you hire someone, the costs add to the GDP. This is acceptable on both theoretical grounds (the GDP aims to measure the market economy) and practical considerations (what home activities wouldn’t be counted?).
Still, it’s confusing, particularly because the boundary between “the market” and everything else is increasingly blurred in a digital world. Some critics argue that GDP is wildly understated, because it omits “free” Internet services such as Google and Facebook that are used everywhere. Google and other Internet services aren’t free, counters the BEA. They’re paid for indirectly through advertising.
The final problem is the most dangerous: the possibility that GDP will be hijacked by those who think it should be modified or replaced to promote other goals — more equality, environmentalism — besides economic growth.
Granted, GDP is not a good measure of economic well-being. It does not subtract many of prosperity’s costs: pollution, crime, congestion, industrial dislocation and inequality, to name a few. Arguably, it also undercounts the benefits of major upheavals. But trying to squeeze these and other factors into one statistic is undesirable. The more we try to measure with one number, the less that number will tell us about anything.
By all means, let’s improve GDP data. But debates over national priorities should be in the open, not hidden in statistical engineering. Let the GDP be the GDP.
Read more from Robert Samuelson’s archive.
Read more on this topic: