Greg Mankiw, the author of the most popular economics textbook, briefly explains why unemployment numbers are often so confusing. Read here. He excerpts several paragraphs from his textbook, ending with:
... In the end, the divergence between the household and establishment surveys from 2001 to 2003 remains a mystery. Some economists believe that the establishment survey is the more accurate one because it has a larger sample. Yet one recent study suggests that the best measure of employment is an average of the two surveys. [George Perry, “Gauging Employment: Is the Professional Wisdom Wrong?,” Brookings Papers on Economic Activity (2005): 2.]
More important than the specifics of these surveys or this particular episode when they diverged is the broader lesson: all economic statistics are imperfect. Although they contain valuable information about what is happening in the economy, each one should be interpreted with a healthy dose of caution and a bit of skepticism.
The Gallup website weighed in with this:
This blog mentioned in mid-September that it was possible -- if unlikely, based on Gallup's survey data that include 30,000 interviews per month -- that September's seasonally adjusted unemployment rate could fall to 7.9%. Still, Friday's BLS report of a drop to 7.8% in the Household survey seemed to surprise everyone, as has been the case on many occasions this year.
The problem is that even though the Household survey tends to be very volatile, this decline seems to lack face-validity, particularly after the prior month's numbers. The consensus estimate was that the government would report that the unemployment rate was unchanged at 8.1% in September. GDP growth was 1.3% in the second quarter and seems to be no better this quarter. The government's Establishment survey shows there were 114,000 new jobs created in September -- very close to the consensus of 113,000 -- and not sufficient to lower the unemployment rate.
The obvious conclusion is that a new employment measure is needed. Gallup has proposed such a measure -- Payroll to Population (P2P) -- the number of Americans employed full-time for an employer as a percentage of the U.S. population. This is a much simpler measure that has none of the numerous adjustments made to the seasonally adjusted unemployment rate. The P2P deteriorated slightly to 45.1% in September from 45.3% in August, suggesting the real jobs situation was essentially unchanged last month. ...
The problem is that even though the Household survey tends to be very volatile, this decline seems to lack face-validity, particularly after the prior month's numbers. The consensus estimate was that the government would report that the unemployment rate was unchanged at 8.1% in September. GDP growth was 1.3% in the second quarter and seems to be no better this quarter. The government's Establishment survey shows there were 114,000 new jobs created in September -- very close to the consensus of 113,000 -- and not sufficient to lower the unemployment rate.
The obvious conclusion is that a new employment measure is needed. Gallup has proposed such a measure -- Payroll to Population (P2P) -- the number of Americans employed full-time for an employer as a percentage of the U.S. population. This is a much simpler measure that has none of the numerous adjustments made to the seasonally adjusted unemployment rate. The P2P deteriorated slightly to 45.1% in September from 45.3% in August, suggesting the real jobs situation was essentially unchanged last month. ...
The Gallup piece goes on at some length to explain their case. Complicating matters with the current system is that the big announcement is made about monthly numbers, but a little later, those numbers are revised, sometimes significantly so, when more complete data is available. All the public ever remembers is the originally announced numbers.
Comments