The May 2016 jobs report was bad news in every meaningful way. Even ignoring the ongoing CWA strike, job creation was too low to absorb new workers. Nearly a half of a million folks quit looking for work and job losses plagued nearly every sector. A spike in involuntary part-time work erased months of full-time job gains and inflation-adjusted hourly wages declined. In total, this report was too bad to be merely a white noise error or data gathering anomaly. The reason why it was bad is another issue.

Labor markets are lagging economic indicators, and so the only solace in these numbers is that they may be a hangover from the global slowdown that already appears to be stabilizing. Still, this challenges the Federal Reserve to reconsider the expected interest rate hikes later this month. It also begs the question of just how much policymakers can rely on macroeconomic models to explain the world.

Macroeconomic forecasts perform fairly well in every domain except one; the timing and magnitude of a downturn. Given that there have been only a dozen US recessions since the computer was invented, that's not too surprising. Imagine how accurate weather forecasters would be after only a dozen storms. Still, forecasting recessions would offer a huge opportunity for policymakers, businesses and households to mitigate or eliminate the downside of a business cycle. However, forecasting a recession (or its precursors like a financial bubble or global shock to oil prices) is not the same thing as preventing one. For that, we have to evaluate what economic theory offers policymakers.

The current slow growth recovery is largely consistent with predictions from two competing models of the economy. One set of models focuses on failures in financial markets, which lead to the bad investments that trigger a recession, e.g. the overbuilding of homes. In that model, failure to allow the recession to run its course would result in the current economy we now suffer. This is often called the Austrian model.

The second set of models allows for many of the same causes to a recession. But, this view credits the slow recovery to workers and businesses recalibrating their purchasing and investment decisions to lower expectations. In this model, increased government spending, lower taxes and loose monetary policy would all increase the speed of recovery. This is often called the new Keynesian model.

Both of these approaches can be convincingly argued and it is easy to see how someone who isn’t a trained economist might be attracted to one or the other based on their existing world view. For those of us who are trained economists, the performance of these models depends on how well they perform empirically, across more than predicting a slow growth recovery.

For what it’s worth, my money is on the new Keynesian models. Still, after close to a decade of fiscal stimulus and easy money, whatever policy choices we have are limited by the huge Federal deficit and the Fed’s balance sheet, not economic theory.