The Misuse and Meaning of GDP
Friday’s report on America’s Gross Domestic Product (GDP) held a few surprises. But the real story lies behind the data. The economy grew at a 2.5 percent annual rate in the first quarter of 2013. That was a little less than what most economists had guessed, even though business invested more and consumers spent more than predicted. The key takeaway: Americans were clearly feeling better about their own prospects, since consumer spending accelerated. The bad news came mainly from misguided austerity politics, at home and abroad. Under our current nitwit sequester policy, government cutbacks continued to drag down growth. In addition, trade detracted from growth for the first time in a year, in large part because exports to austerity-stricken Europe are falling.
It’s no mystery why markets and politicians track the GDP figures so zealously. Compiled by 2,000 economists and statisticians at the Bureau of Economic Analysis (BEA), GDP pulls together everything they can measure about how much American households and industries earn, consume and invest, and for what purposes. But earlier in the week, the BEA also tacitly acknowledged that its current GDP measure lags pretty badly behind the actual economy. Earlier in the week, the bureau released a set of changes in how it will calculate GDP, which will take effect in the second-quarter report to be released July 31. The main point of those changes is to take better account of the economic value of ideas and intangible assets. Few among us today question the notion that new ideas can have great economic value. And some 15 years ago—long before smartphones, tablets, and protein-based medications—the BEA started to study how to revise the GDP measure to better measure that value. Finally, this week, the bureau announced that starting soon, and for the first time, when a company undertakes research and development or creates a new book, music, or movie, BEA will count those costs as investments that add to GDP, rather than as ordinary business expenses, which do not.
BEA also will apply these changes to its GDP numbers going back more than a half-century. In the present, the revisions, in an instant, will add some $400 billion to the official accounting of the economy’s current total product. Voilà! Business profits also will look larger—now and for the past half century—because ordinary business expenses reduce reported profits, while investments do not. Most important, the revisions tell us that American businesses and government, together, now invest just 2.1 percent of GDP in R&D—that’s less investment than in the 1990s, especially by businesses.
While this week’s BEA changes bring us closer to an accurate picture of GDP, last week we also learned how naive we can be about blatant misuses and distortions of GDP. This story began four years ago, when two well-respected economists, Carmen Reinhart and Kenneth Rogoff, published an economic history of financial crises. R&R’s timing (2009) was impeccable, and their book was a bestseller. More important, it gave its authors wide public credibility when they issued a paper the following year, “Growth in a Time of Debt,” that claimed to have found a deep and strong connection between high levels of government debt and a country’s economic growth. The data, they reported, showed that when a country’s government debt reaches and exceeds the equivalent of 90 percent of GDP, its growth slumps very sharply.
With the big run-up in government debt spurred by the financial crisis and subsequent deep recession, conservatives who had waited a long time for a plausible economic reason to slash government found it in the new R&R analysis. And based on its authors’ newly elevated reputations, conventional wisdom-mongers from think tanks to editorial boards echoed the new line on austerity. Even the most liberal administration since Lyndon B. Johnson’s couldn’t resist the new meme. Despite a palpably weak economy, the president and congressional Democrats grudgingly accepted large budget cuts and then pumped the economy’s brakes some more by insisting on higher taxes. And we were not the only ones so economically addled. As government debt in Germany, France, Britain, and most other advanced countries rose sharply, conservatives there argued that less government was a necessity for average Europeans as well.
Just last week, however, we learned that the R&R 2010 analysis was so riddled with technical mistakes that its “findings” about what moves GDP are meaningless. When three young economists from the University of Massachusetts found they couldn’t replicate the results—the standard test for scientific findings—they took R&R’s model apart, piece by piece, to figure out why. It turns out that R&R—or more likely, their graduate assistants—left out several years of data for some countries, miscoded other data, and then applied the wrong statistical technique to aggregate their flawed data. And as bad luck would have it, all of their disparate mistakes biased their results in the same direction, amplifying the errors. In the end, instead of advanced countries experiencing recessionary slumps averaging -0.1 percent declines once their government debt exceeded 90 percent of their GDP, the correct result was average growth of 2.2 percent carrying that debt burden.
Utterly wrong as R&R’s analysis was, the austerity advocates proceeded to badly misuse it. The authors had merely reported a correlation between high debt and negative growth—or, as we now know, between high debt and moderate growth—without saying what that correlation might mean. Hard-line conservatives and their think-tank supporters, here and abroad, quickly insisted it could only mean that high debt drives down growth. That can happen, but only rarely—when high inflationary expectations drive up interest rates, which at once slows growth and increases government interest payments. In the much more common case, Keynes still rules: slow or negative growth leads to higher debt, not the other way around. In those more typical instances, cutting government only depresses growth more, further expanding government debt. Occasionally, the correlation of negative growth and high government debt reflects some independent third cause. The tsunami and nuclear meltdown that struck Japan in 2012, for example, simultaneously drove down growth and drove up government debt. And sometimes, there is no correlation: Britain carried government debt burdens of 100 percent to 250 percent of GDP from the early to mid-19th century, while it was giving birth to the Industrial Revolution.
The R&R analysis did not distinguish between these various scenarios. Yet, the conservative interpretation became the received public wisdom. The International Monetary Fund, the World Bank and most politically unaffiliated economists insisted that slashing government on top of weak business and household spending would only make matters worse. No matter. The inevitable result was not the stronger growth as promised, but persistently high unemployment and slow growth here, and double-dip recessions for much for Europe and Japan. In the end, R&R deserve less criticism for their mistakes than for their failure to correct the damaging distortions of their deeply flawed work.
The broader takeaway from this week’s wonky discussion of economic growth? We shouldn’t put too much stock based on a single GDP report. The data are likely to be revised, reexamined, and reinterpreted over time.