By just about any measure, this has been the worst recession since the Great Depression. While I think we are coming out of it, the employment market is the last thing to turn, particularly if we follow the pattern of the last two recessions and their aftermath.

We have seen a steady pattern of lower job losses since January, when we hemorrhaged over 700,000 jobs in that single month. August’s loss of 216,000 is sure an improvement over that, and is even a big improvement over the 276,000 lost in July, which in turn was a huge improvement over the 463,000 lost in June.

It is, however, not good enough — the economy needs to add jobs, not just avoid losing them. Every year, the workforce grows by a little over a million, so just to stay even we should be adding about 100,000 a month. To recoup the 6.9 million jobs lost since December 2007, it will have to be much, much higher than that. Historically, though, the participation rate falls as we go into a recession, but then starts to increase as the recovery begins.

Just how bad has this recession been versus other recessions? I think the following graph (from shows it very clearly. In raw numbers of jobs lost, it simply blows away anything since the end of WWII — but then again, the country is a lot larger now than it was then.

However, even when measured as a percent of jobs at the peak of the cycle, only one recession, that in 1948, even rivals it. Then we were winding down from WWII, no longer putting people to work building battleships and tanks and the GI’s were coming home — so in some respects, those job losses were a good thing, or mostly represented friction in the transition to a peacetime economy. Further, by this point, that recession had already recovered all the jobs lost.

By contrast, this time we have yet to bottom out. Also, look at the historical pattern of the amount of time it has taken to get back to square one:

It took a little over two years to recover from the deep 1981 downturn. In the 2001 downturn, it took us almost four years from the start of the recession to get back to the number of jobs at the start of the recession. In the 1990 recession it took almost three years. This is despite both of those more recent recessions being extremely short and mild in terms of the length and depth of the decline in GDP.

It is very possible that it could take until 2015 before we pass the previous peak of 138.2 million jobs in the economy set in December 2007.

We now have 14.9 million unemployed, an increase of 466,000 in just the month of August and of 7.4 million since the recession started. Note the difference between the increase in the number of unemployed and the number of jobs lost.

Part of that is due to the numbers coming from two separate surveys. However, it does have a lot to do with the behavior of the unemployment rate. In July, the unemployment rate fell to 9.4% from 9.5% despite the economy losing 276,000 jobs as people were feeling too discouraged to even look for work.

The drop in the unemployment rate was clearly a bit of an anomaly, but it was an extreme representation of that change in the participation rate I was talking about. If the labor force is growing (in terms of people in normal employment ages) by a million a year, and the recession has now been going on for 20 months, then why is the difference between the number of jobs lost and the number of unemployed only 500,000?

Some of it may have to do with multiple job holders. If you are holding down three jobs, you are only counted as employed once, and if you lose one of them you are still counted as employed, but your three jobs still count as three jobs in the establishment survey.

However, part of it is due to the “there are no jobs out there, so why waste gas driving around trying to find them” effect. In that respect, the fact that the number of unemployed rose by 250,000 more than the number of jobs lost in August is actually good news (well, at least a silver lining in a very dark cloud) since it indicates that people are thinking it is time to get off the couch and start looking for work again.

Don’t get to hung up on the precise numbers, since coming from two different surveys there is plenty of room for statistical discrepancies, but a quarter of a million differential in a single month is too big to be just from such discrepancies.

While the 9.7% (U-3) rate is the one that gets all the headlines, it counts you as employed even if you used to work 40 hours a week, but due to the recession your employer has cut you back to only 10 hours a week. Yeah you still have a job, but you aren’t working much.

A broader measure of unemployment, which includes people working part-time for economic reasons (as opposed to, say, a teen working part time after school) and people marginally attached to the labor force (have not looked in the last 4 weeks because of say family responsibilities, but would be happy to take a job if one were available) showed an even bigger jump, rising from 16.3% in July to 16.8% in August. There were 9.1 million involuntary part-time workers and 2.3 million marginally attached workers in August.

There was a little bit of good news on the duration of unemployment front — the average length of time people were out of work dropped to 24.9 weeks in August from 25.1 weeks in July. That is still a seriously ugly number, though; a year ago it was at 17.6 weeks.

Similarly, the median duration of unemployment dropped to 15.4 weeks from 15.7 weeks in July, but is still well above the 9.3 week level a year ago. Keep in mind that a year ago the recession had already been going on for as long as each of the previous two recessions lasted.

Those changes, however, were driven by changes in the middle of the unemployment duration distribution. The number of people out of work between 5 and 14 weeks rose sharply, to 4.120 million from 3.557 million in July, while the number of people out of work between 15 and 26 weeks dipped to 2.828 million from 2.916 million.

Of particular concern are those that have been out of work for more than 26 weeks, since that is when regular state unemployment benefits run out. That number is still rising, but more slowly than it has been, up to 4.988 million from 4.965 million in July.

One measure I like to look at is the ratio of the long-term unemployed to the number of short-term unemployed (less than five weeks). The history of that ratio is shown in the graph below.

This recession is simply in a whole different league than any other recession we have had. In August, the ratio was 1.65, up from 1.54 in July. A year ago it stood at 0.58, which was already elevated by historical standards. Prior to this downturn, the highest it had ever reached was 0.78 in February of 1983.  If we exclude the last year, the average since 1960 is 0.35.

Most of the almost five million long-term unemployed are getting emergency extended benefits paid for by the Federal Government as part of the stimulus package. However, those do not last forever, and by the end of the year they are expected to run out for almost 1.5 million people. Those are people who are going to have to get their groceries from the food bank rather than Kroger’s (KR) and their clothes from the Salvation Army rather than from Wal-Mart (WMT).

The recession has not hit all demographic groups equally. As usual, it has hit the more underprivileged people hardest. The unemployment rate rose for all major demographic segments in August, but there are vast differences in the levels of unemployment.

Teens have the highest rate at 25.5%, up 1.7 points on the month. Blacks have a 15.1% unemployment rate (up 0.6 points) while Hispanics have a 13.0% rate (up 0.7 points), while the unemployment rate for Whites is 8.9% (up 0.3 points).

The one exception to the normal pattern of under-privilege and unemployment is that this downturn has been much harder on men than on women. The adult male unemployment rate now stands at 10.1% (up 0.3 points) while for women it is “only” 7.6% (up 0.1 points). A year ago, the rate for men stood at 5.8%, while for women it was at 5.3%.   

The recession also points out the age-old wisdom of staying in school. It holds in good times and in bad, but especially in bad times. The unemployment rate for HS dropouts is now at 15.6% up from 9.7% a year ago. For high school grads, it is now 9.7% vs. 5.8% a year ago. For those who have some college or got a 2 year degree, it is at 8.2%, up from 5.0% a year ago, and for those with a bachelors or better the unemployment rate is now 4.7%, up from 2.7% last year.
Read the full analyst report on “KR”
Read the full analyst report on “WMT”
Zacks Investment Research