Reinhart/Rogoff-gate isn't the first time that incorrect data have been a thorn in the side of those seeking austerity and deficit reduction. It’s been a remarkable feature of attempts to “pivot” away from dealing with mass unemployment that they’ve relied in large part on government data to make their points -- data that usually turn out to be incorrect.
As Henry Farrell and John Quiggin document, the “Keynesian revival saw its peak in early 2009. By mid-2010, Keynesians ... were lamenting that they had lost the war.”
The counter-Keynesian arguments fell into two broad categories. The first is that the economy has no short-term, demand-driven cyclical problems that the government can address and that the real problems come from the supply constraints of our economy. The second focuses on the discovery of serious limits to how much debt a country can carry, as well as evidence that austerity can create enough growth to offset itself.
Though austerity seekers move effortlessly between the two, those are separate lines of argument, the first having to with the issue of supply, the second with the issue of universal limits. And neither has done particularly well in the past several years.
The interesting part of the first set of arguments is that despite their pessimistic view of government’s effectiveness, they usually depend on government-provided data rather than market information. Worse, those data usually turn out to be wrong.
Take the idea of “structural unemployment.” Here the argument is that there are too many job openings given the level of unemployment in the country, so there must be something keeping the unemployed from taking the jobs available. In 2010, people ranging from David Brooks to Bill Clinton argued that the jump in job openings, as calculated by the Bureau of Labor Statistics, was reason to worry that there were no more short-term solutions.
The numbers turned out to be incorrect. In 2011, the BLS revised its job openings numbers significantly downward, finding some 200,000 fewer jobs open from 2009 to 2010. This was the result of its birth/death model for businesses being calibrated from periods that didn’t include the largest downturn since the Great Depression. This also assumed that a job opening was something that stayed the same through a recession, although newer research shows that the health of the economy affects how actively employers look to fill a job and thus close an opening, which means how government statistics understands openings is limited in weak times.
Or take the idea that unemployment skyrocketed faster and higher than people had expected at the beginning of the recession. This gave many people the impression that somehow the relationship between growth and employment was broken, or that bringing down unemployment was going to be much more difficult than traditional recession-fighting.
Sadly, the data turned out to be wrong. In 2011 the Bureau of Economic Analysis released new estimates finding what many had suspected: that the recession was deeper and longer than conventional measures were telling us at the time. The weak recovery and high unemployment actually made perfect sense in light of the fact that “percent change in real GDP was revised down 0.3 percentage point for 2008, was revised down 0.9 percentage point for 2009.” (Here is what GDP looked like plotted before and after the revisions.)
Or take the idea of “housing lock,” to pick a random but indicative example. The Census Bureau reported a significant decline in interstate mobility numbers after 2006. Some immediately jumped at the idea that underwater mortgages were to blame for mass unemployment, as people couldn’t move to new jobs, which would blunt stimulus. The Washington Post reported in 2010 that “economists worry that decreasing labor mobility is a growing factor in keeping the nation mired in its worst employment situation since the Great Depression.”
Sadly, the data turned out to be wrong. The Minnesota Fed found that those interstate mobility numbers were the result of a data error. The supposed drop never even happened, yet many people spent valuable time in 2010 worried about it.
Contrast that with what the Keynesians were doing. They were looking at unused capacity alongside record-low interest rates and wondering why the government wasn’t acting. The treasuries market is one of the biggest, most liquid and traded markets in all of human history, and it is actually offering negative rates when controlled for inflation.
When they did look at government data, the Keynesians used data that was better. The government does a great job of keeping track of wage incomes (the better to tax them), and they could see that wages were stuck in a downward direction. The government also keeps great track of unemployment insurance, because it actually has to write the checks. Those numbers were skyrocketing in the crisis and stayed high, and their move back up recently has led to worries about premature austerity. (There’s something oddly Soviet about relying on questionably accurate bureaucratic data over market signals to determine government macroeconomic policy; however, it is difficult to imagine Leonid Kantorovich using Excel when really good statistical software packages are available for free.)
As for the second line of argument, anyone who’s read Timothy Mitchell’s excellent book on the creation of the Egyptian economy through colonialism, "Rule of Experts," should know that when economists show up claiming to have found “principles true in every country,” you should watch out. And sure enough, the most notable thing about the second wave of arguments was how they were meant to abstract from the specific situations countries face.
The two big arguments here were that austerity could expand the economy if carried out properly and that there was a 90 percent debt cliff that would cause countries to have significantly less growth once they wandered beyond it. Both claims were severed from the contexts of any specific economies. Big or small, well-functioning or not even competent, having strong growth or suffering through a recession: Whatever the economy, these two lessons were supposed to hold. This is what fueled the idea that market confidence would boost the economy if we started to cut.
As many have found, the examples that were driving the expansionary austerity results were countries in which the economy was already strong and growing. Or they didn’t have low interest rates, so monetary policy could offset the cuts. Or they exported their way out of their austerity. And we are currently cutting the deficit along the lines Alberto Alesina recommended, and no analyst believes private spending is increasing faster to compensate.
And now the last idea, that there’s a debt-to-GDP threshold that is true for every country, falls apart now that researchers are finding the causation is backwards. As Gavyn Davies notes about this week, the “moral of this story is that it is an illusion to expect that the complicated relationship between public debt and GDP growth will always and everywhere be the same.” The idea that there is a stable relationship between debt and growth across time and places, independent of weak economies, is now behind us.
The timing of these developments is interesting. Right now there’s a serious effort to rethink the move to austerity. Between the developments in Japan and the IMF’s efforts in Europe and England, the common wisdom will soon be that austerity as a solution was oversold, with all the toxic side effects hidden. The question next will be how to turn that into political power.
Mike Konczal is a fellow at the Roosevelt Institute , where he focuses on financial regulation, inequality and unemployment. He writes a weekly column for Wonkblog. Follow him on Twitter here.