Scott Sumner says there’s nothing mysterious about “jobless recoveries,” they’re just weak recoveries:
The thing we seem to have forgotten is that the end of a recession used to be associated with really fast GDP growth, not just a return to normal levels of growth. And that makes sense. If you’ve got a full employment economy, increasing output is hard. You need to apply more capital or think up new ideas. But if you’ve got 9-10 percent of your labor force not working then it’s easy to increase output—you need those unemployed people to do, well, something. Anything! So rapid increases in real output from the trough of a recession are both possible (because there are all these idle workers) and necessary (because absent rapid output growth the idle workers won’t get jobs) but we’ve forgotten how to step on the gas hard enough.