What is the recent increase in U.S. jobless claims telling us?

The jobs picture is looking grim. Jobless claims for the week ending 14 Feb 2009 were 627,000, pushing the average to 619,000. Additionally, nearly 5 million people are staying on the unemployment roles, the highest figure ever. But, it should be a given that jobless claims have increased in this cycle — the U.S. population has increased. What do these figures really mean?

The number itself is largely meaningless except as a talking point for journalists comparing this recession to previous downturns. How many times have you heard the phrase: “this is the largest number since 1982?” Dozens of times, I bet. Unemployment is a lagging indicator and doesn’t tell us anything about whether the economy is going to rebound. What we need to divine from the number is how it predicts future employment, and therefore future spending and economic growth. That’s going to be important whether you are looking for a job investing money or are looking to strategize for your business.

Here’s my brief look into how one can do that (for a more in-depth analysis, see my October post “The Economy’s Four Horsemen“). When the economy starts to slow, it is usually starts with consumers pulling back somewhat, followed by business realizing they need to pull back in turn. This leads to job losses and a further pull back. The cycle of retrenchment eventually ends and consumers begin spending again. But, what you should notice is that consumers lead and companies follow. That means employment falls only after it is apparent that the economy is slowing, and employment continues to fall even after the economy starts to pick up. So, employment is a lagging indicator.

What is not a lagging indicator is the increase in unemployment. In business cycles I have studied, the 6-month or 12-month change in unemployment is a good gauge of the economy. Translation: if you can compare today to six months ago and the jobs picture looks worse (on a relative basis) than it did last month when you did the same exercise, that means the economy is deteriorating.

Now, there are any number of ways to represent this comparison. You can look at the 12-month change in the unemployment rate. You can use the look at the 12-month change in the number of unemployed, and so on. One of my best metrics is jobless claims. More precisiely, I look at the “12-month change in the 4-week average unadjusted number of initial unemployment claims.” That’s a mouthful, so let’s call this stat the “jobless claims change (JCC).”

What the JCC tells us is the rate of deterioration or improvement in the employment market, which very much more of a coincident or leading indicator than unemployment (see my post “Another Perfect Recession Indicator‘ for more). If the JCC increases a lot and is still increasing, that means the jobs picture is getting worse at a more rapid rate, which means that the economy is also deteriorating.

Below are two charts for jobless claims.



The above two charts indicate that the JCC reached a peak of 326,000 two weeks ago and has since fallen to 292,000. If this trend continues, it would indicate an improvement in the economy. However, two weeks, do not a trend make. Moreover, the continuing claims versions of the same charts show further deterioration i.e. an acceleration in the worsening of unemployment. And, all of these charts show a jobs picture on par with the deep recession of 1974.



So, the long and short of this is: the jury is still out on how quickly things are deteriorating. I fully expect an ugly unemployment number for February and double-digit unemployment before this is over. Because of inventory de-stocking, I also expect a poor GDP number for this quarter as well. But, let’s not get all doom & gloom here. The numbers are plateauing right now.  By the end of March, things could look very different.

Unemployment Insurance Weekly Claims Report – U.S. Department of Labor

  1. David Pearson says

    Great analysis, but I would normalize the numbers for growth in the employed population. I believe they will look, on the one hand, more benign, and on the other, scarier (because we have further to go before topping out).

    This is true of normalized claims data presented by Calculated Risk. It appears from that data that weekly claims should rise to one million to reach ’74/’82 levels. Other data supports the conclusion that this recession will be at least as deep as those two, so one should expect weekly claims to spike in the coming months. The counterargument is that, perhaps, the increase in joblessness this time will be shallower and longer, so that we’ll never reach those claims highs. I disagree with this view because we’ve experienced essentially a housing recession (from 12/07 to 9/08) followed by a deeper, steeper economy wide recession.

  2. Edward Harrison says


    You make a great point about normalizing the numbers. I have a note to self on one of the greatest apps invented Evernote saying:

    Not apples to apples due to manufacturing concentration/leaner inventory blah blah and size of economy.

    Basically, I wanted to work in some themes I didn’t have space for. The important one is the theme you broach on the population growth since the severe downturns in 1974 and 1982. These recessions were worse on a normalized basis.

    But, one must caveat that statement for a couple of reasons. The U.S. economy was much more manufacturing based back in the 1970’s and early 1980s so you were guaranteed to see more layoffs because hiring patterns in the manufacturing sector are more cyclical. Plus, we have a much leaner manufacturing cycle now, with Just-In-Time and all of that. Even in the energy sector, you can see stockpiles of crude, blending components or final output are much lower now than they were in the 1980s. All of which is to say it’s hard to make a apples to apples comparison.

    That said, jobless claims are a much better metric than the unemployment rate which has been manipulated.

    The scary part, as you say, is that there is no clear indication we have reached the zenith. Even when one looks at year-over-year comparisons, they are not getting substantially better, which means the unemployment rate increase is accelerating to the upside.

    Thanks for your input.


Comments are closed.

This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Accept Read More