TCS Daily

First-Time Jobless Claims Gauge the Economy's Health Well

By Kevin Hassett - August 7, 2001 12:00 AM

Just about everyone is looking around for signs that we have hit bottom. The Federal Reserve is undoubtedly doing that as well. One of the most frustrating problems facing the Fed is the tardiness of the economic data. Here we are in early August, and the most recent data on factory orders, for example, cover June. Fed policy can't affect June, what matters is how we are doing now. What's the timeliest indicator of that?

There have been numerous studies that have attempted to identify the best contemporaneous economic indicator, and there is a clear champion: initial claims for unemployment insurance. This data release provides a simple count of the number of new claims for unemployment insurance that occur during a given week. It is a very precise measure of the number of people who have lost their jobs, which makes it a pretty important number. Timeliness is what makes this indicator receive such a heavyweight in the thinking of economy watchers. On Thursday (August 9) we will receive data for the first week of August.

The data on claims over the past few months tell a very interesting story. When the economy is healthy we usually see 200,000 to 300,000 new claims for unemployment insurance in a given week. During a recession (and adjusting for the growth of the economy that has occurred since the last one) we might expect to see 500,000 to 600,000 new claims. Up until about a year ago, claims stayed fairly comfortably in the "boom" range, but they began a steady march northward last fall. In early July, they peaked at about 450,000. Since then, they have steadily declined, with the most recent reading showing 346,000 new claims in the last week of July.

If that trend holds, then we will be able to safely call last June the bottom. Another month of that trend and the series will be back in "boom" territory. I choose June instead of July because layoffs generally lag bad news a little bit. Employers hate laying off workers, and often have to pay significant costs when they do. Layoffs tend to happen only after signals about bad times are convincing.

If the downtrend continues, the latest episode will have provided strong evidence in support of arguments made by New Economy students. Claims at their peak did not approach a recessionary level. This means that firms were indeed able to use information technology to manage their inventories and production in a way that smoothed significantly the stresses that the business cycle put on their production.

So the key question is, will that pattern hold? Here are the arguments for and against.

Let's look at the dark side first. Data over the summer on layoffs are very noisy. This is because many manufacturers temporarily lay off workers for a few weeks while they retool, but they do this in a haphazard fashion. Since the seasonal filters correct for such short-term issues, claims could look like they are declining just because temporary layoffs are not as high as usual. Why might that be? Maybe manufacturers released workers who normally face temporary layoffs earlier in the year, giving us less short-term pain now. If that happened, then the claims decline in July will just be a blip.

On the brighter side, claims have declined steadily for three straight reports, and have not fluctuated wildly as one might see if seasonal adjustments were causing data chaos. Moreover, many of the other indicators we have discussed in this space have signaled that the bottom might have been near in June. The claims data look exactly like we predicted they would look now. The reduction in layoffs makes sense given the other things we have seen.

All this means is that this Thursday's release is going to have an unusually large effect on the outlook. Three good reports might have been luck, or seasonal weirdness. A few more, and we will have to take the new trend seriously.

Postscript: Well, the frustration continues. The claims data last Thursday could have sent us a strong signal that we were out of the woods, but they did not. The number of unemployment claims jumped up to 385,000, reversing the recent downtrend. On the brighter side, claims stayed well below the recession danger zone, but a number in the neighborhood of 350,000 or below would have signaled that we are headed back toward a vibrant economy. All this suggests that the frustratingly low positive growth experienced so far this year will continue in the third quarter.

TCS Daily Archives