Tweaking the Fed's unemployment target
Last week I had a look at the US labour market over at Variant Perception's blog based on this excellent piece by Ellen Terry, an economist at the Atlanta Fed, which discusses the drivers of the decline in the US labour force participation rate. The issue should be well known for US economy watchers. The unemployment rate has declined noticeably but if we factor in the declining labour force participation rate the picture looks largely unchanged.
The following sums up the main point from Terry's study and my own comments.
The most interesting aspect of Terry’s analysis however is the finding that the bulk of the decline in the US labour force participation rate (80%) is due to one of three reasons.
- Wants a job, but can’t find one
- Disabled/ill
- Retired
The disabled/ill category is interesting because of the increasing evidence that receiving disability aid is better paid than having a low wage job. Several academic and journalist sources have been pointing to the unsustainable rise in the prevalence of social disability entitlements in the US. Finally, it is interesting to ponder retirement as one of the major causes of the decline in the labour force participation rate. This is significant for two reasons in our view. Firstly, because retired workers are unlikely to re-enter even if economic conditions improve (and if they do it will most likely be in part-time and/or low wage occupations). Secondly, it casts a pessimistic empirical light on the prospect (in all OECD economies) to increase the labour supply by inducing later retirement to correct for a rise in life expectancy. This may work in theory, but it seems much more difficult to implement in practice.
In the context of the official unemployment rate moving rapidly closer to the Fed's target as per the Evans rule the recent labour market trends have obviously put the central bank in a bind. The Fed has already assured markets that the Fed funds rate will be kept low well past the point at which the unemployment rate declines below 6.5%, but for how long and should the rate to which forward guidance is attached be amended (e.g. from 6.5% to 5.5%)? Continuingly massaging forward guidance to reflect a complicated interplay between structural and cyclical drivers of the US labour force participation rate could turn into a big communication challenge for the Fed.
The question surrounding the dwindling labour force participation rate has been the center of much debate and analysis. A recent contribution comes from Morgan Stanley's US economics team also citing Terry's piece. The piece largely come to the same conclusions as above (in terms of the drivers of the labour force participation rate decline), but MS appears optimistic on the prospect of the participation rate to recover as the economy improves. In other words, MS argues that there is scope for significant cyclical improvement.
The gist of the suggestion is that there is more slack in the US labour market than meets the eye and that the Fed could conceivably change the labour market measure it looks at to anchor forward guidance.
Should the Fed, then, shift focus to some alternative measure of labor market slack? A broader measure of unemployment, the U-6, may help. It includes marginally attached workers such as those that are working part-time, but would prefer full-time if it were available (“part-time for economic reasons”), and workers who would like a job but don't search because they don't believe any jobs are available (“discouraged” workers). The U-6 unemployment rate was 13.1% in December compared with the standard (U-3) unemployment rate of 6.7% (Exhibit 7).
The spread between U-6 and U-3 reflects what we consider to be shadow labor - deserters ofthe labor force that could come back if job prospects were to improve enough.
I am skeptical that this shadow labour will be as sensitive to the business cycle as MS suggests. This is especially the case since we have already seen a noticeable improvement in the US economy without any reaction from the shadow labour component (on the contary). The most interesting proposal however is obviously the suggestion that the Fed should change its unemployment target from the U-3 to the U-6 measure. Needless to say this would immediately alleviate a lot of the pressure on forward guidance, but it would also mean that the Fed would need to buy the idea that the difference between these two measures is cyclical. In coming posts I will have a look at whether this can be argued to be true.