Unemployment: Very Mixed

Today’s employment report from the Department of Labor has an odd feel to it. The figures are always adjusted for seasonal factors so whenever we read about the ‘new claims for unemployment assistance’, we are not looking at the raw data, but at a figure manipulated to offset any peculiar and recurring movements. The typical surge in seasonal hiring during the summer is a good example of such an adjustment. The idea being , of course, to get at the underlying trends that such a seasonal ‘bounce’ might obscure.

Ironically today’s report that initial claims fell by 52,000 last month may be an example of the adjustment process producing a perverse result. The data has been scrubbed to exclude the usual summer variations, one of which is the ‘normal’ auto industry round of layoffs prior to the summer time re-tooling for next years model introductions. This year all those layoffs occurred in the spring when Chrysler and GM went through bankruptcy. So the seasonal adjustment downwards of the raw number may not be justified this year and so the reported improvement may not be representative of a long term trend.

There’s a lot of ‘mays’ and ‘ifs’ in that analysis, which only goes to highlight the difficulty we have in getting at what’s really going on. So I repeat my mantra: watch the data accumulate and don’t focus too heavily on any one week’s or month’s numbers.

With that in mind I want to point out one worrisome fact that we ought to be paying attention to closely over the next few months: ‘repeat claims for assistance’. This figure shows us the trend in long term unemployment. While ‘initial claims’ obviously reveal recent layoffs, ‘repeat claims’ give us an idea of how easy or difficult it is to find work. If the ‘repeat claims’ keep on edging up, as they did this week, then we can assume that employers are not re-hiring. So the two figures provide insight into the two parts of the employment cycle: when the initial claims figure starts to drop, we know that the run up in unemployment is ending; and when the repeat claims start to drop we know that the unemployment rate will start to fall also.

Right now the trends seem to show a slowing but still rising trend in unemployment with no shift towards decline in sight.

To put the employment picture in perspective: the number of people claiming unemployment insurance was 3,218,000 at this time last year and is now 6,883,000. That’s the size of our problem. We need to get these folks back to work.

Addendum: Unemployment and the stimulus package …

As an afterthought: unemployment is a lagging indicator. It rises after a recession has begun and improves after a recovery has started. There is no way that the unemployment rate will improve in the near term. So all the current hyped up talk over the ‘failure’ of the stimulus package is just that: hype. There hasn’t been enough time to have an effect. The original proposals laid out very clearly that the earliest expected improvement in employment resulting from the stimulus was in the late third and early fourth quarters this year.

We should therefore ignore the critics.

In any case anyone suggesting that we don’t need a stimulus should be clear about what their alternative is. Reduce interest rates? Can’t do that: they are at zero. Cut taxes? Did that already in 2001 and 2003: nothing happened, the last eight years have been the weakest employment years of any post war expansion. What else? …

Exactly. We have one policy option. Get over it.