Jon pretty much nailed almost everything you need to know about employement/unemployment statistics. I'll try to add a just little more context. First, there is a methodology which is part of the establishment survey called the "birth/death model". It is not what it sounds like, however. You see, a big part of the jobs created or lost each month comes from the creation of new businesses or the closing of old ones. The survey can only gauge such activity with a considerable lag, because new jobs and old lost jobs are not yet added to the unemployment insurance rolls. To fix this, the Bureau of Labor Statistics (BLS) must estimate how many new business "births" are created and likewise "deaths". They base the calculation on actual birth/death statistics, but naturally these reflect what happened in the past and now what is happening now. so they tend to be way off the mark during turning points in the economy- overstating employment when the economy slows and understating new employment when the economy picks up.
For example, in June the BLS says 289,000 jobs were created, but of those, 177,000 came from the birth/death model. Even more suspicious is construction added 29,000 of those jobs. Huh? Construction of what, empty homes? No doubt these will be revised downward when the real data comes in months from now. If you are dying to read those BLS tables, you can find them here: http://www.bls.gov/web/cesbd.htm
Despite problems with the surveys, they do give us the general temperature of the economy as far as labor goes. If you are interested in predicting where the economy might head next, however, they are both pretty worthless. Why? The main reason is that most employment statistics are a lagging indicator. That is they tell us what already happened. Economists will look at different indicators and classify them as leading, lagging or coincident. Employment/unemployment statistics are considered lagging and therefore of little value in divining the future, but they can be of great value when studying what recent trends have been. For example, if you are trying to determine how many more jobs might be created before igniting upward wage pressures (competition for workers) and thus inflationary pressures, then employment statistics are a good barometer. Here is a fun fact: only about 62% of the population are employed.
Friday, August 22, 2008
Subscribe to:
Post Comments (Atom)
Thanks for shedding some light on this topic!
ReplyDelete