The reason for both the illusion and the counter intuitive point is related to the calculation of unemployment and the dynamics of the components. In the latest report the denominator shrunk by 315,000 people during the most recent measurement. That is, people seeking a job became totally discouraged and left the workforce [actively seeking a job]. Meanwhile the numerator grew by 278,000 jobs [employed]. Hence with the numerator growing and the denominator shrinking you get an illusion.
Why would the unemployment rate rise if the economy was growing? Growth of the economy would cause the discouraged worker (not currently seeking a job) to be attracted back to the workforce [those actively seeking a job]. Hence with growth, the initial response would be an upward movement in the measurement [unemployment rate] as the denominator would be flooded by people now seeking jobs. Hence the new 8.6% unemployment rate down from 9% is really an indicator of the overall weakness in the economy not a robust economy.
Furthermore, the January 2012 unemployment rate will likely fall as well. Why? Historically speaking, the January unemployment rate is below the preceding December unemployment rate due to an arbitrary statistical decision each January: a statistically significant number of workers are automatically retired by labor statisticians each January. Therefore the denominator is reduced and unemployment arbitrarily looks better in January than the preceding December all things being equal.
Historically speaking--that is examining the seasonally-adjusted monthly unemployment data from 1948 to 2011 from the Current Population Survey--there is absolutely no difference in the December and following January unemployment rates. The average seasonally-adjusted unemployment rate for December from 1948 to 2010 is 5.8, and the average seasonally-adjusted unemployment rate for January from 1949 to 2011 is 5.8. BTW, from 1949 to 2011, all of the other months also have average seasonally-adjusted unemployment rates of 5.8.
ReplyDeleteThe analysis looks the same if you examine the last 30 years of monthly data instead of the whole series. Over the last 30 years, the monthly average in all months was 6.3 percent.
David posts: :Historically speaking--that is examining the seasonally-adjusted monthly unemployment data from 1948 to 2011 from the Current Population Survey--there is absolutely no difference in the December and following January unemployment rates. The average seasonally-adjusted unemployment rate for December from 1948 to 2010 is 5.8, and the average seasonally-adjusted unemployment rate for January from 1949 to 2011 is 5.8. BTW, from 1949 to 2011, all of the other months also have average seasonally-adjusted unemployment rates of 5.8.
ReplyDeleteThe analysis looks the same if you examine the last 30 years of monthly data instead of the whole series. Over the last 30 years, the monthly average in all months was 6.3 percent".
In the post I was eluding the “January Effect”.
An excellent explaination [“The January phenomenon is interesting. The birth-death number for January is always a huge negative number”] appears in the link below:
http://www.businessinsider.com/bls-employment-numbers-black-box-2011-2
Other resources are:
CES Net Birth/Death Model
http://www.bls.gov/web/empsit/cesbd.htm
Labor Force and Unemployment Statistical BS
http://globaleconomicanalysis.blogspot.com/2011/02/labor-force-and-unemployment.html
Did Unemployment Really Go Down?
http://www.minyanville.com/businessmarkets/articles/unemployment-report-unemployment-rate-bureau-of/2/7/2011/id/32650