Why is the recovery from this recession different from recoveries from past recessions? In the previous two recessions, it took 32 months for nonfarm employment to reattain its June 1990 peak, and 48 months for it to reattain its January 2001 peak. Assuming the economy keeps adding nonfarm employment at the current rate, it will have taken 88 months to reattain its January 2008 peak. The explanation most often heard is that “financial crises are different”: after a debt crisis, shaken consumers are reluctant to spend and shaken firms are reluctant to hire, slowing private-sector job growth even after the recession has bottomed out.

There is some truth in this, but it is not the whole story. In fact, while the latest recession was particularly deep, the recovery in private-sector employment, once it finally started, has not been particularly slow by recent historical standards. In the 27 months since the start of the current employment recovery, the private sector has added 4.3 million jobs, fewer than the 5.0 million it added in the 27 months after February 1992 but not many fewer than the 4.5 million it added in the equivalent period after August 2003.

But there is something historically different about this recession and its aftermath: in the past, local government employment has been almost recession-proof. This time it’s not. Going back as long as the data have been collected (1955), with the one exception of the 1981 recession, local government employment continued to grow almost every month regardless of what the economy threw at it. But since the latest recession began, local government employment has fallen by 3 percent, and is still falling. In the equivalent period following the 1990 and 2001 recessions, local government employment grew 7.7 and 5.2 percent. Even following the 1981 recession, by this stage local government employment was up by 1.4 percent.

Who is losing these local government jobs? In 1981 it was mostly teachers. Now, the losses are shared by teachers and other local government workers alike.

State government is much less important than local because it is a much smaller share of total nonfarm employment: 4 percent versus 10 percent. Nevertheless, a similar story can be told there. This far into each recession since 1955, state government employment had grown. Since the start of the latest recession, state government employment is still down 1.2 percent.

Without this hidden austerity program, the economy would look very different. If state and local governments had followed the pattern of the previous two recessions, they would have added 1.4 million to 1.9 million jobs and overall unemployment would be 7.0 to 7.3 percent instead of 8.2 percent.

Why is this happening? One possibility is that we are witnessing a secular change in state and local politics, with voters no longer willing to pay for an ever-larger work force. An alternative explanation is that even though many state and local governments are constrained not to run deficits, they can muddle through a standard recession without cutting jobs. But when hit by a huge recession like that of 1981 or the latest one, the usual mix of creative accounting and shifting in capital expenditures cannot absorb the shock, and jobs have to go.

It has become commonplace to contrast the American and European responses to the Great Recession, with stimulus in the former and austerity in the latter. European austerity has been at the level of member states and local governments — there is no meaningful federal government of Europe to provide either stimulus or austerity. But the United States has also seen unprecedented austerity at the level of state and local governments, and this austerity has slowed the job recovery.