Tuesday, October 28, 2008

This cycle may get to 1980s low; unlikely according to cyclical unemployment

I received some comments regarding different measures of unemployment when I predicted that the unemployment rate would not rise above 10% in this cycle; specifically, how some measures are already reporting 11%. However, these measures do not give any new information. The best measure to identify the depths of the recession is cyclical unemployment. At 10%, the cyclical unemployment rate implies a recession that tops the 1980s, and since growth is not forecasted to reach such lows, 10% is an upper bound.

Admittedly, there are huge risks that the unemployment rate actually reaches this level, and the biggest being the failure of the Fed’s and Treasury’s measures to instill confidence in the banking sector. However, assuming that they do, 10% implies high excess unemployment (cyclical), surpassing that following the 1980-1982 recessions when the unemployment rate reached 10.7% in January of 1983.
The chart illustrates quarterly cyclical unemployment spanning the years 1975, quarter 1 to 2008, quarter 3. Cyclical unemployment is the difference between the current unemployment rate and the non-accelerating inflation rate of unemployment (NAIRU), or the natural rate of unemployment. The current unemployment rate, 6.0% in the third quarter (average over July, Aug, and September), is 1.2% higher than NAIRU, 4.8%. Thus, the cyclical unemployment rate is 1.2%.

The current cyclical unemployment implies that this recession may be more severe than that in 1990-1991. The recession likely started in July 2008 (third quarter) and cyclical unemployment was 1.2%, which is higher than cyclical unemployment entering the recession of 1990, -0.1% (the current rate of unemployment was slightly lower than the NAIRU level).

If the cycle peaks at 10%, the current recession may be just as, if not more, severe compared to the recessions in 1980-1982. During 1980-1982, when economic growth hit a cyclical low of -7.8% in 1980, and the cyclical unemployment rate hit 4.58%. If unemployment hit 10% during this cycle, then the implied cyclical unemployment is, 5.22% = 10% - 4.8%, which exceeds the 1980-1982 cycle high.

I have not seen a forecast where growth approaches the levels implied by the 1980-1982 recession. Thus, 10% is likely too high for this cycle, and a worst-case scenario because growth is just not expected to be -7.8% for any quarter in the near term.

I am sure that some are going to refer to the U-6 measure of unemployment, which is currently 11% and already exceeds my forecasted peak of 10%. This measure estimates the unemployment rate, calculated as unemployed plus part-time for economic reasons plus marginally-attached workers all divided by the labor force plus marginally-attached workers. This level is higher than the headline unemployment rate at any given time and adds no new information to the severity of this cycle.

The best way to assess the severity of a recession is to use cyclical unemployment because NAIRU changes over time. And if we do reach 10% unemployment, the U.S. economy is looking at a nasty recession – worse that the 1980-1982 cycle. The risks that this recession does reach such lows are ubiquitous, but venerable forecasters (like Macro Advisers) do not see the U.S. economy reaching such depths.

Rebecca Wilder

1 comment:

  1. If you look back to forecasts of unemployment in the past number of years, they are notoriously inaccurate. The numbers for say May will change when updated in June, July and later, some even going from negative to positive or vice versa. So, the cyclical approach seems to be more to the point. We also may hear more about the subject as so many "white collar workers" are being handed the dreaded pink slip.