What are the real job numbers?. Yesterday, I posted a note from Stephen Roach, the Wall Street Economist from Morgan Stanley (he gets paid millions for being right, not for being political). He states that we are 8 m jobs behind where we should be right now with an average recovery (which started in late 2001). How can that number be true? The easiest way to see what is really going on is to look at the absolute numbers of jobs or economy has created rather than fudgy stats about who is looking for work.
Brad DeLong has an excellent graph that shows in absolute terms how many jobs our economy supported over the last four years. Note that the peak was at the end of 2001. Since then, we have lost 2.2 m jobs. In that same time period, the number of new entrants to the job market has grown, due to normal population growth, at nearly 200,000 per month (or 7.2 m workers over the last three years). Combined, that's a net deficit of 9.4 m jobs to stay even with our employment rate at the end of 2001.
So, given these numbers, 8 m new jobs to isn't a typo. So, why the difference between these numbers (generated by an establishment survey that covers 1/3 of all workers) and the unemployment numbers? Unemployment statistics are generated through a household survey that asks people about their job status and whether they are looking for work. The household survey only asks this question of 0.006% of all US households. This sounds about as useful as presidential poll.
FWIW, our economy has even lost jobs since the start of the recovery. This isn't a usual pattern. Something more is at work. Global job arbitrage, enabled in large part by the Internet, may be the culprit. [John Robb's Weblog]