Neil Irwin at the New York Times (The Upshot) writes that “no one cares about economic data anymore” which he says is “good news.” He points to some indisputable facts, such as the steady growth in both employment and GDP over the past few years. His focus is mostly on short-run monthly or quarterly movements in economic data-movements that perhaps we can safely ignore in the near future.
The GDP numbers for the first quarter of 2014 were released yesterday, and GDP growth was way below forecast. Bad winter weather was the likely culprit, and we agree with Irwin that we are no longer seeing the short-run gyrations in GDP numbers that we saw in the 2006 to 2010 period.
But over a longer horizon, there is something deeply puzzling about the GDP numbers, and economists everywhere should be staring at them and scratching their heads. The chart below shows why. It plots real GDP for every U.S. post-World War 2 recession for 26 quarters after the recession. Each line is indexed to 100 in the quarter before the official NBER start of the recession. The steeper the line for the particular recession, the stronger the recovery.
It is true that the recovery in GDP has been steady over the past couple of years – but it’s been steadily disappointing. The recovery out of the Great Recession looks dismal compared to earlier recoveries- we aren’t even close to the recoveries we’ve seen before. The short-run gyrations are gone, but the longer run issue of dismal growth is as important as ever.
Why has the recovery been so dismal? This has to be one of the central research questions for macroeconomics going forward. We have several candidates: (1) secular stagnation, (2) structural changes in the economy such as demographics (3) a Gordonesque pessimism on productivity growth (4) weak government spending, (5) heightened policy uncertainty, and (6) excessively tight monetary policy. These aren’t mutually exclusive.
But whatever your views, this is not a boring chart. Quite the opposite. This chart demands attention – and further research.