Misleading Indicator
Will the Great Recession finally end our misguided obsession with gross domestic product?
Here’s a question no one ever asks about the Great Recession: How do we even know it’s happening? Daily, almost, someone releases employment figures, production estimates, consumer-confidence indexes—almost all of them bad. There’s no need to wonder what’s happening when information about everyone’s problems constantly streams across your TV screen.
And in this dolorous statistical parade, no number is quite so central to public life as the gross domestic product. Political scientists build formulas around it to predict who will win the presidency. The stock market trembles at the approach of new quarterly figures. Other economic statistics, like budget deficits or health-care spending, are quoted as percentages of GDP. It has become the common denominator of economic well-being.
But GDP’s broad dominion has long had its critics. It was never meant to be the measure of our well-being, they say, only the measure of our production—literally, the total value of the goods and services produced within the national borders in a given year. While the quest for some broader measure of progress has been going on for a while (more than a decade ago, for example, The Atlantic was running articles like “If the GDP Is Up, Why Is America Down?”), it may finally be gaining traction at a time when people understand, as never before, how easily GDP and well-being can diverge.
One of the leaders of a huge global effort to build a better statistical yardstick has been Enrico Giovannini, until recently the chief statistician of the Organization for Economic Cooperation and Development. Though the OECD is the global coordinator of the project, its partners represent a who’s who of economic development: the World Bank, various UN programs, the African Development Bank, and the European Commission. They are looking to create more-reliable metrics for measuring change in our health, education, the environment—the many ways that human beings make themselves better off or worse off. This fall, in the wake of the OECD’s third World Forum on Statistics, Knowledge, and Policy, these groups are set to move ahead on a broader, better set of indexes.
Crisis seems to be the mother of statistics. The germ of the idea that eventually became GDP emerged after World War I, when American economists who had been frustrated by the lack of reliable statistics to guide war production founded the National Bureau of Economic Research (despite its name, the NBER is a private organization). Still, not until the Great Depression did we finally get our first national income accounts, which measured the annual income of people, companies, and the government.
It is rare for people to write about that era without taking a swipe at Herbert Hoover, but we might be kinder if we remembered just how little information he had. With no national income accounts, Hoover had to rely on fragmentary indicators such as freight-car loadings, steel production, and the gyrations of the New York Stock Exchange. There weren’t even comprehensive national unemployment statistics, because Congress didn’t authorize the Department of Labor to collect them until the middle of 1930. Forget Hoover’s economic theory, most of which was pretty bad; the man barely had any data.
Beyond newspaper anecdotes and a bunch of unrelated industrial indexes, he had little way of knowing just how awful things were or, more important, exactly where intervention might be needed. It’s as if someone hired you to cater a lavish formal dinner—then gave you no head count, a partial list of available ingredients, and the July 1953 issue of Gourmet magazine to work with.
That started to change in 1932. The legendary progressive Senator Robert La Follette introduced a resolution that directed the secretary of commerce to estimate national income for the prior three years. Unable to find anyone in the department who was up to the task, Commerce hired Simon Kuznets, who had already begun working on the problem at the NBER. He spent the next 14 years creating a system of national accounts that, with constant refinements and additions, we still use today.
It’s almost impossible to overstate what a titanic achievement this was. But here’s something that hints at the magnitude of the difficulties: despite his earlier work, Kuznets didn’t make the first tentative presentation of his figures to Congress until 1934. By 1941, he had expanded the series backward to 1919, and all the way to 1938. Measures of output—the true forerunners of GDP—didn’t follow until 1942, as war production ramped up. (Total mobilization depended on extensive statistics, which is why the Nobel Prize winner Paul Samuelson called World WarII “an economist’s war.”) The Great Depression and World WarII created the modern world in a lot of ways. They also created one of the primary lenses through which we view it.
Unfortunately, that lens is a trifle distorted. It counts the dollar value of our output, but not the actual improvement in our lives, or even in our economic condition.
Think about a house, any of the millions that were constructed during the bubble that burst in 2008. Let’s make it a nice house: four bedrooms, 3.5 baths, with an attached garage and a quarter-acre lot. During its construction, that house did its own little bit to boost GDP. Lumber was purchased and swathed in fluttering robes of Tyvek. Tiles were pressed out of clay and nailed to the roof. Pipe was laid, glass was sealed into the window slots, granite was hewn from a Vermont mountain and shipped all the way to its kitchen counters. All of this output, which swelled GDP (at least to the tune of its purchase price), has ended up in … nothing. The house, in an exurban cul-de-sac, sits empty while bankers, borrowers, and regulators squabble. One of the estimated 2.4million excess homes on the market, its only function right now is to bankrupt its owner.
GDP does not, and cannot, reflect the waste of enormous effort, and precious natural resources, that went into building something that suddenly no one wants. Moreover, it misses many other aspects of our existence. Strip-mining a picturesque mountaintop, or clear-cutting a primeval forest, shows up in GDP only as a boost to output. Meanwhile, in India’s national accounts, all of Mother Teresa’slabors among the poor would have had only the most minimal possible impact. GDP can record how much money we spend on health care or education; it cannot tell us whether the services we are buying are any good.
Consider the housewife, or rather, her modern successor, the stay-at-home mom. There’sevidence that the recession may have forced her back onto the job market. Since the downturn, the labor-force participation rate for working-age men has fallen significantly. But the participation rate of women has actually ticked up. The most likely explanation is that women whose husbands have been laid off or had their income cut are going back into the workplace (at least temporarily) to help make ends meet.
Some feminists might rejoice, but to an economist, or a social conservative, this development is almost certainly a bad thing. Assuming that she and her family both wanted her to stay home, each woman who leaves for the office out of economic necessity represents a loss to the country, a loss of what economists call utility and what we may think of as net national happiness.
As she heads back to the workplace, that mother will be boosting GDP. If her husband has lost not his job but merely some income from sales commissions or a business, she will probably have to pay for child care. She may need to buy new work clothes. Money will be spent on commuting, and the family will probably shift away from homemade meals to costlier prepared foods that save time. All of these transactions further swell the national income accounts. Yet all of them also represent a decrease in life satisfaction.
These are but some of the reasons why analysts on the left and right think we need a better measure of national welfare than GDP. But if Kuznets had a hard time figuring out how to measure all the transactions in our marketplace, how much harder will it be for his multilateral successors to put a number on the things we often call priceless?
There are, broadly speaking, three ways you can try to build a “well-being” index. You can use what economists call “shadow prices,” imputing dollar values to the various things that contribute to our quality of life. But the index usually ends up being incredibly sensitive to your starting assumptions. As the economist Tyler Cowen says, “How much is a fish worth? .00000000000001 cents per fish or .0000000000000000000000000001 cents per fish? It makes a big difference!” He wryly adds, “The point of a GDP statistic is to drain away those sorts of problems. The point is not to commit you to an anti-fish position.”
The second approach is to attach weights to various indicators and use them to build a composite gauge like the Human Development Index. Unfortunately, the weights will always be somewhat arbitrary. That inherent subjectivity makes the accuracy of your index somewhat suspect, and leaves the people who created it open to accusations of bias.
Take the World Health Organization’s infamous ranking of national health systems, which in 2000 put the United States below such health-care luminaries as Oman, Colombia, and Morocco. Would you really rather get sick in Bogotá than in Berkeley? The WHO analysts heavily weighted somewhat murky estimates like equality of access, knocking us down to two places above Cuba, where antibiotics are scarce for everybody. This is so bizarre that conservatives don’t sound entirely crazy when they voice suspicions that the WHO chose its weights to produce just this result. The WHO has since stopped publishing the index.
When all else fails, of course, you can just ask people: Are you better off now than you were three years ago? Even this approach poses problems. As any marketing expert can tell you, the difference between what people say they want, and what they actually want, can be large. Their memories of years past are quite poor, and often highly selective. Survey respondents are vulnerable to all sorts of influences, from how the question is framed to a desire to impress the interviewers. That’s why polls often find substantial majorities against raising taxes and against cutting any major programs—and for reducing the deficit. Surveys also show that heterosexual men on average have many more sexual partners than heterosexual women do, which is mathematically impossible.
Besides, much of the progress in important areas of life is invisible to most people. You are indisputably better off having the option of getting a liver transplant if you should need one. But unless your skin actually starts turning yellow, you probably never think about it.
The OECD project’s daunting task is to find better ways to handle these kinds of obstacles. One possible approach is to focus on hard indicators that we can measure in a fairly standard way. But these are scarce for some aspects of life, and even when they exist, can be tricky to interpret. Life expectancy, for example, seems pretty objective. It’s a metric on which the United States does relatively poorly, causing us endless consternation. A few years ago, however, the health-care economists Robert Ohsfeldt and John Schneider recalculated the numbers after controlling for deaths from homicides and traffic accidents. Because these things tend to strike very young people, they can have an outsize impact on mortality statistics. Those deaths reflect America’s crime policy and its driving habits more than the effectiveness of its health-care system. And if you remove them from the picture, say Ohsfeldt and Schneider, America jumps to the top of the life-expectancy tables. Assuming they’re correct, does America have good health, or bad? Neither answer is the obviously right one. Such conundrums will vex analysts long into the future.
But of course, measurement problems also bedevil GDP, which is why the U.S. Bureau of Economic Analysis employs economists and statisticians rather than bookkeepers. Whatever the difficulties, the OECD project’s leaders, like their forebears during the Great Depression, are convinced of the urgent need to get a better handle on the progress our societies are making.
That sense of urgency doesn’t mean we should look for speedy results. Even though his staff had to tabulate the results without the aid of computers, Kuznets and his team produced the first national accounts more rapidly than his successors will create the newer, better set of statistics they aspire to. For all the burdens under which Kuznets labored, he had one advantage: the absence of an existing economic establishment. He did not have to convene working groups and seminars to make sure that academics around the world and the staffs of dozens of institutions felt included. Nor, once he was finished, did he have to convince governments that they should care. Giovannini describes today’s next step as developing “a framework that at least lists the domains that outline what this project should look like.” That’s several nouns away from any sort of working model.
It doesn’t help that Giovannini has left the OECD to head Italy’s statistics authority. But efforts like these have always been larger than one person. When our grandchildren face their financial Waterloo, they may have Giovannini’s brainchildren to help guide them through it. But we will have to muddle through with the legacy of Kuznets and the generations of economists who expanded and improved on his remarkable achievement.