Unemployment in Various Industries and Speed of Recovery

There have already been a lot of great topics touched on here so far. One thing that I found interesting, though it was not heavily emphasized in the reading, was Keynes' discussion of increases in unemployment in response to changes in production. Unemployment has been traditionally known as a "lagging indicator" in that it peaks after GDP contraction peaks. As Keynes argues, though, "the time which elapses before production ceases and unemployment reaches its maximum is, for several reasons, much longer in the case of the primary products than in the case of manufacture" (Keynes 127). Because agriculture has defined seasons, greater costs of a temporary shutdown, a larger number of self-employed workers, and so on, contractions in employment occur long after contractions in production. In manufacturing, however, unemployment occurs much more rapidly. As Keynes asserts, these differing "speeds" of unemployment can serve to deepen the contraction of growth, as once the agricultural workers become unemployed, they will not purchase as many manufactured goods and so on.

I find this point interesting because at our current point in history, our economy is becoming more and more service and information based. If, like between the agricultural and manufacturing sectors, there are different rates of change in unemployment for the service and information sectors, this will necessarily impact the amount of contraction and the recovery period. If these jobs are fairly easy to eliminate (and fairly easy to fill), economic processes like contraction and recovery will happen at a much faster pace and vice versa. Furthermore, if we understand how key sectors restart employment and what barriers there may be to entry of new labor in those sectors, it will be easier to create appropriate policies (if any) to encourage a stable, efficient, and ideally speedy recovery.

The Bureau of Labor Statistics has extensive data for about the last 10 years on unemployment, including unemployment by industry. Given more time, a thorough analysis of this data would provide a fairly solid idea  of which industries hire and/or fire quickly and which hire and/or fire more slowly. From merely a cursory glance, however, there seem to be very different rates in growth among differing industries. From just 2010-2011, for example, some industries, such as construction, decreased their unemployment percentages by as much as 4 percentage points, while others, such as utilities, actually increased their unemployment rates. Of course, there are many factors involved in changes in unemployment, such as demand for the product of that industry, government involvement, and so on. Unemployment in the public sector, for example, will not necessarily follow a pattern related to the business cycle unless one or more of the factors that form public policy (such as public opinion) also track the business cycle in some way. Nonetheless, there may be some differences in changes in unemployment rates by industry that are related specifically to the kind of work that industry entails.

As this graph shows, the Leisure industry in Virginia dropped off fairly quickly from 2008 to 2009, but not much from 2009 to 2010. On the other hand, the Financial Services and Information industries see a larger increase in unemployment from 2009 to 2010 than from 2008 to 2009. This difference could reasonably be explained by the difference in the type of training and expertise required for these industries. Many, if not most jobs in the "leisure" industry do not require a college degree or extensive training. Jobs in financial services and information (which I'm presuming to mean IT and computer related jobs) often do require technical knowledge and a college education, if not some form of a graduate degree. Because workers in these industries likely require relatively more investment (such as in the form of training) before they can be economically worthwhile to the company, companies are more reluctant to let them go. Furthermore, when such a company is capable of hiring again, they will be much more cautious in doing so because of the required investment in hiring a new person.

Of course, more data to substantiate this theory would be very helpful. However, it seems intuitively rational to assert that industries heavy on jobs that require intense and/or extensive training will necessarily contract and recover more slowly. If this is the case, as specialization increases and more and more careers require advanced technical knowledge and/or higher education, the business cycle will become more and more prolonged. During boom periods this may be a very good thing, but it can make recessions particularly painful, especially for those just entering the job market when it starts. Keynes does not seem to offer an explicit solution to this and I am not sure if there is one good answer either. Nonetheless, if these structural shifts in our economy and our labor force do cause prolonged booms and recessions, that may present some unexpected and complicated policy issues that would be difficult to resolve purely by traditional means such as stimulus spending. Perhaps this could even explain our current situation and why recent attempts at stimulus have failed to produce a robust and rapid recovery.