Job creation and the deep gap

In February 2011, 192,000 jobs were added in the US. Most reports say that it's "encouraging" while cautioning that it's going to take several years at this pace to recover the ground lost in the recession.

How long?

Well, in February 2011, 138 million people were employed, whereas three years before there were 144.5 million employed (BLS data, monthly employment level, table A, no seasonal adjustment). The difference is actually 6.457 million.

In the same period (2/2008 to 2/2011), the number of multiple job holders dropped from 7.6 million to 6.9 million. To be exact, 728,000 jobs were lost. Most of these jobs were part-time, though, and counting them as full jobs lost seems excessive. Let's count part-time jobs as half jobs, so 364,000 full job equivalents were lost.

Again over the same period, the number of people working part-time went up from 25 million to 27.4 million. This amounts to 2.4 million people going part-time from full-time. Counting part time jobs for 50%, this amounts to another 1.2 million jobs lost. There is little overlap between these numbers and the numbers of the multiple job holders, I think, as the latter end up being fully employed.

So, my estimate of the number of full-time equivalent jobs lost in the last three years is 8 million.

Moreover, in the same period, the civilian non-institutional population also increased by 6.042 million or 2.6%. Normally a rise in the population should lead to an increase in the labor force. In fact, we should expect the rise in the labor force to be the same, i.e. 2.6%. However, this didn't happen: the labor force rose only by a bit over 100K. The reason for this is that, although there are more people of working age, relatively fewer are actually interested in work at all. If we are to get back to the same employment situation as before the crisis, people will have to be brought back into the labor force. Suppose we should get back to the pre-crisis employment level, the labor force should expand by 2.6% over the 2008 numbers, which amounts to 3.957 million new workers. If we're to keep the same participation rate as in February 2008 (65.5%), this will require the creation of 2.952 million jobs.

So the job gap - the amount of jobs needed to get back to where the work force was in February 2008 - is almost exactly 11 million. This disregards the quality of these jobs, of course, but that is a very different topic.

Now, how long does it take to get 11 million jobs back? Over the past 3 years, keeping up with the growth in the civilian population would have required the creation of 2.952 million jobs. As population growth is slowing, this hurdle will lower, but for the next couple of years an annual addition of 900,000 jobs should probably do. That amounts to monthly job creation of 75,000.

So, only when job creation exceeds 75K we start working on reducing the job gap. That means that a monthly job creation rate of 200K will close the job gap in 88 months - a bit more than 7 years. An average job creation rate of 400K per month can get the gap closed in less than 3 years. The fastest sustained job growth we had in recent history (from 1992 to 2000) saw on average a 200K jobs created per month (all of this is still BLS data).

So there you go: if job creation rivals that of the nineties (200K/month), the job gap will be closed by 2018. It is worth remembering that the growth in the nineties was partly due to the internet bubble, so it's debatable whether this really is a sustainable rate.

In reality, then, it's more likely that some of the changes in the employment situation are permanent: more people will remain part-time employed, more people will stay completely out of the labor force and the participation rate will drop. If we just try to recover the 8 million jobs lost (and keep up with population growth), a 200K/month rate will accomplish this in something like 3 to 5 years. Call it a recovery to a new normal by 2015. Whether this new normal is stable remains to be seen.