Tuesday, January 12, 2010

More on Employment

The initial unemployment claims continues to drop. Also note the rate of the drop is on pace with the post 1982 recession.

Those unemployed less than 5 weeks has been dropping since the beginning of the year. But the current level is still above the level of 2005-2007 indicating we're still above the historical norm.

The number of unemployed 5-14 weeks has topped out, but is still about 1.2 million above the norm established in 2005-2007.

The number of unemployed 14-25 weeks also appears to have topped out, but is about 1.6 million over the norm of 2005-2008.

The number of people unemployed over 27 weeks continues to increase. This number is over 4 million above the norm of 2005-2008.

The above series of charts should be read as a time progression -- first people are laid off then the move through the various time periods until (regrettably) they are unemployed for over 27 weeks.

The good news in these series is that the shorter numbers (initial unemployment claims and under 5 weeks unemployed) have been decreasing for some time. The less than 5 week number has been decreasing since the beginning of the year and the initial unemployment claims number has been decreasing since the mid/late spring. Unfortunately, because of the rate of job loss during the height of the recession there are still a ton of people unemployed. Let's revisit some numbers to illustrate that point.

Note that at the end of 2009/early 2009 the US lost at least 600,000 jobs/month. We have also learned that the BLS added an additional ~800,000 of job losses to the total amount of jobs lost. While I am sure this will ignite another round of ill-informed conspiracy theories about employment, the reality is that measuring the economy is an incredibly difficult task -- especially during a massive financial shock. As a result, revisions are to be expected.

Let's add two more charts:

Note that construction jobs have dropped by over 1.5 million and

Manufacturing jobs have dropped by over 4 million. That means the vast majority of job losses are attributable to two areas of the economy.

What makes both of these areas of jobs losses concerning is this: the vast majority of these jobs probably aren't coming back. Construction benefited from the housing bubble. Considering the high rate of vacancies and existing home inventory there is no reason to start building en mass again. And note that after the 2001 recession manufacturing didn't come back either. The primary reason is technological advances -- the country saw an increase in overall production and productivity on a declining labor force.

So -- let's sum up.

1.) The employment numbers continue to move in the right direction. Initial unemployment claims continue to drop, leading to a continuing drop in the 5 week and under category. The other categories of time unemployed have topped out save the 27 weeks plus category.

2.) A large percentage of job losses come from construction and goods production industries. Construction jobs aren't coming back because they were caused by the housing bubble. Manufacturing jobs will probably suffer from the same circumstances that hit them in the 2001 expansion -- increased productivity leading to increased production at the expense of a declining workforce.


David Altig at the Atlanta Fed offers this chart:

And these points:

After growing during the 1980s and 1990s, the aggregate labor force participation rate (the percentage of the working-age population active in the labor market employed or looking for work) peaked in the late 1990s and is currently at levels last seen in the 1980s. But this change pales in comparison to changes in labor force participation among America's youth (those folks in the 16- to 24-year-old age range).

During the 1980s participation in the labor market for youth averaged around 68 percent, a rate noticeably higher than for older individuals. The youth participation rate declined sharply to a level at or below the level for older individuals prior to the 1990–91 recession and then remained relatively stable during the 1990s. However, over the past decade youth labor market participation has been on a steep downward trend and currently stands at a little over 55 percent, compared with about 67 percent for older individuals. Moreover, the most recent recession has seen youth participation rates decline at a rate similar to that seen in the early 2000s. In contrast, the labor force participation by individuals over 24 years of age has varied much less, implying that the decline in youth labor force participation has been a major contributor to the reduction in the overall rate of labor force participation (see the above chart).


The big change appears to be that those in school have become increasingly less attached to the labor market. The percentage of school enrollees aged between 16 and 24 who are also participating in the labor market was relatively stable between 1989 and 1998 at around 51 percent. However, labor market participation by those in school declined between 1999 and 2008 from 50 percent to 42 percent. In contrast, labor force participation by those aged between 16 and 24 not enrolled in school has declined only modestly—from 82 percent to 80 percent between 1989 and 2008.

There are economic returns (benefits less costs) to both labor market experience and education. The decreased attachment to the labor market of school enrollees likely reflects, at least in part, factors such as the increased lifetime economic returns to education relative to alternative uses of time. As such, a widening wage premium on education is probably an important influence on youths' schooling choices, including schooling intensity. An example would be enrolling in educational programs during the summer instead of looking for summer employment.