Bernanke rolls the dice on what seems to be a bad bet

In Ben Bernanke’s speech to the National Association of Business Economists, he argues that the main reason for unemployment being stubbornly high was not a structural reason but a cyclical reason. Thus he has made the call for economic policies to spur demand as a solution. So far markets seem to like it. I find myself on the opposite side of this argument. The table below looks at the five categories of unemployment by reason. These data reveal the percentage point contribution to the overall rate of unemployment from each type of unemployment. The table clearly shows that ‘by reason,’ or at least by most reasons, unemployment is very close to or below the averages for each component for recoveries lasting 32 months since 1970. The BIG exception is ‘not-temporary layoffs where 3.9 pct points of the overall rate is created. Compared to 2.5 pct points of contribution on average this is where the main problem resides. And there is a slight contribution from the category, labor market re-entrants, which at 2.2pct points in this cycle compares to an average of 1.8 pct points. There is a small relative elevation here. There is a chart I would like to insert here but cannot. You can find it a Most of the excess in the ‘unemployment rate’ in this cycle comes from the category ‘not-temporary’ layoffs. This category contributed 3.9% points to the overall rate and is higher than the highest previous cycle at this stage (2.7 pct points) and higher than the average for this category by 1.4 pct points. The rate for new entrants is above average too; the two together add 1.8pct points to the overall rate of unemployment in the cycle. Without these two categories in excess the unemployment rate could now be around 6.3%. The category ‘job losers’ is split into two parts. Typically it dominates the unemployment picture. The category ‘temporary job losers’ refers to those former workers on temporary furlough during the recession. They may be unemployed for the moment but are waiting to be called back to work. In contrast ‘not-temporary job losers’ are those who lose jobs but are not waiting to be called back. Either they were fired or their firms went out of business or the plant they worked for was closed, etc.. This category is the one most likely to be affected in job changes of a structural nature. Not surprisingly for temporary job losers unemployment tends to fall rather quickly in a recovery; it tends to be only 65% of its end recession level 32 months later (our current data are for the 32nd month of the expansion). While unemployment among not-temporary job losers tends to be a lot stickier; it averages 92% of its end recession level 32 months later. Not-temp jobs losses when they loom large as a share of unemployment (as they do now) should tend to make the duration of unemployment longer. And that is just what we are seeing. I think Bernanke is barking up the wrong tree. The duration of unemployment is not ‘a problem’ it is ‘a symptom’. It is symptom of large not-temporary job losses and that is a symptom of structural change. One thing to remember is that this is a statistic for people who are STILL unemployed. The duration data possibly are made SMALLER by potential workers getting discouraged and dropping out of the labor force. Discouraged drop outs tend to have been unemployed for a while…And if this happens and after dropping out they come back in, when the job market improves they will appear as unemployed- that’s right- re-entrants. What we find in cycle is that the drop in temp job losers is almost exactly what is normal; levels are 65% of their recession end level. Of the two ‘problem’ categories, the drop in not-temp job losers is larger than average as it is 78% of its end-recession level compared to an average of 91%. Re-entrants, the other category that is relatively high in this cycle, are 101% of their end recession level and that is lower than the 104% that is normal at this time in an expansion. It turns out that reentrants are contributing more to the rate of unemployment despite a slightly better-than-normal drop off because at the recession end the unemployment rate of re-entrants was relatively high. On balance the strongest evidence we have is that the real unemployment problem stems from the classification unemployed for not-temporary reasons. This bulge explains the slow drop in the overall rate and the extended period of unemployment. It suggests more directly than Mr Bernanke’s indirect evidence that the real unemployment problem is from permanently closed businesses causing permanent job loss, the kind that is hard to make up by just driving demand harder. In other words our unemployment problems look to be more structural than cyclical. Now… To defend Bernanke, he does not want to tighten policy right away. He believes he has learned a lesson from the Great Depression. And having that ‘knowledge’ he must express the unemployment problem as malleable for he will lose the option to use discretion if he does not. If he endorsed the argument that the unemployment rate elevation is structural he could not easily call for the Fed to run a continuing easy policy. But if this is the reason for his policy choice then he is aware that with this choice comes a greater danger of inflation. (I have no proof that Bernanke has chosen to say unemployment is more cyclical than structural because he needs to say that that to run the policy he wants to run. If he did such a thing this it would be doing things backwards, in some sense; picking the policy then choosing the rationalization). Still if Bernanke is doing this he knows there is a risk. And I don’t think he would let inflation run out of control. As a policymaker he is in a tough spot. The Fed is under more political pressure than ever before. He does not want the Fed to lead rates higher so he will lag and let the market make the judgment on the economy. He will ‘err on the other side,’ by keeping policy as easy as he can for as long as he can, hoping that the unemployment problem really is more cyclical than structural. In the end he will not let inflation go too far before he switches to fight it. Although Not-Temporary unemployment is slower to fall in a recovery than temporary unemployment, the drop in the not-temp level of unemployment is the second fastest (in percentage terms) among the last six economic recoveries. So it is proving to be somewhat malleable, so far.