Employment Flows

Data on unemployment for January were released today. I think that the process that generates changes in the unemployment rate is poorly understood. The overall change in employment and unemployment is a small net amount relative to massive movements in the labor market.

Federal Reserve economists Bruce C. Fallick and Charles A. Fleischman, in a paper written last April, point out that

about 5.2 million workers per month in 1999, moved from employment to nonemployment (that is, unemployment plus not in the labor force) in an average month, and a slightly higher number moved from unemployment to employment.

Moreover, Fallick and Fleischman point out, about 4 million workers per month change employers. Overall, therefore, about nine million employment relationships end each month, and about nine million employment relationships start each month. The difference between those two numbers is the change in aggregate employment, where a movement of 300,000 in a month is considered large.

In other words, we have a strong economic month when 9 million jobs are terminated and 9.3 million new jobs are started. We have a weak economic month when 9 million jobs are terminated and 8.7 million new jobs are started.

Discussion Question. In view of this data, can one view reports of layoffs as a reliable indicator of changes in overall employment? If you read about a layoff, can you assume that the unemployment rate will rise?