Saturday, May 3, 2014

Nice graphic on revised employment reports from the St Louis Fed

This graphic comes from the St Louis branch of the US Federal Reserve Bank.

It is a nice illustration of (1) why we should pay only mild attention to INITIAL employment reports and (2) the importance of looking for longer terms trends in economic data reporting.

However, both might be of scant use because the employment report is a "lagging indicator".  It reports on what has already happened and may or may not be useful in assessing the current state of the economy.  But we have to dance with the partner we came with.

As the Bureau of Labor Statistics gathers more data about a particular time period they can revisit and update the number of jobs created

According to the numbers below, there was a +23% difference in the average number of new jobs reported initially and what an updated revised average number 3 or so months later.  That seems significant to me.

Does that mean you can just go ahead and add 23% more jobs to the next jobs report to extrapolate that out to a longer term trend? Nope. Only a good politician or a bad statistician would do that...

The revision could be biasd upward OR downward:
The pattern of the revisions shown in the chart is consistent with the evidence showing that revisions tend to be procyclical. That is, during an expansion, revisions to payroll employment tend to be positive; and, during periods of slow economic growth and recessions, the revisions tend to be negative.--St Louis Fed
Employment Revisions
View My Stats