Last Updated Aug 24, 2011 6:03 AM EDT
My view, one that is widely shared, is that the problem is predominantly cyclical. However, the point I want to emphasize is that cyclical unemployment can turn into a structural problem over time, and those problems can cause permanent losses for the economy.
The following graph showing the average duration of unemployment over time helps to illustrate the problem:
Many employers shy away from workers who have been unemployed for a considerable period of time. That is, the longer a person is unemployed, the more negatively he or she will be viewed by the job market. In response, many of the long-term unemployed will drop out of the labor force. For example, some workers in their late 50s or early 60s might give up even looking for a job and find a way to hang on until retirement by living with a family member, etc. Others will take any job they can get, perhaps one that is ill-suited to their talents or in the underground labor market, and get stuck in these jobs long-term. Because the jobs do not make the best use of their talents, their output will be less than it might be otherwise.
This is one source of aggregate losses, and it's not confined to older workers. There's evidence that the first job a person takes has a large influence on their lifetime earnings. When young workers have trouble finding employment and settle for a job that doesn't make the best use of their talents, and then get stuck in those jobs as they buy cars, houses, have families to support, etc., they suffer permanent losses of income.
For this reason, it's essential to keep people connected to the labor force through job creation and other means. We are not doing enough to help the unemployed -- particularly the long-term unemployed -- find jobs and the longer we wait, the more permanent the damage. Job creation isn't free, and budgets are tight at all levels of government. But investing in workers now can avoid losses for years or even decades into the future, and the sooner we get started, the better.