By Charles Thibault on August 6, 2009 in Unemployment Insurance Claims.

The Department of Labor reported a drop of 38,000 new Unemployment Insurance claims on a seasonally adjusted basis. The 4-week moving average dropped by 4,750 claims, also an improvement.

There had been 30,000 new additional claimants in each of the two preceding releases, which had put the 4-week moving average at risk of hitting an inflection point and moving back in a negative direction. The 4-week moving average is important because the NBER usually identifies the start of the recovery as occurring 8 weeks after the peak in the 4-week moving average of new claims.

On a seasonally unadjusted basis, there were 48,300 fewer claimants.  This corresponds to an increase of 120,000 new job ads two weeks ago. Last week, new ads dipped slightly by 31,000 ads. However, the DOL expects an increase of claims on a seasonally unadjusted basis for next week's release, as the seasonal correction parameter will move from 84.2 to 86.0 for the week ending August 8th.

The graph below shows the historical relationship between the number of new UI claims and the the number of new job ads, on a seasonally unadjusted basis. The axes for UI claims has been inverted to facilitate visual interpretation.

Click chart to view full size

Over the past three years, the correlation between new job ads and new UI claims is -0.60.  As job ads increase the economy is able to absorb more workers, and new UI claims drop. This is why the correlation has a negative sign. Interestingly, lagging the job ads data by one week improves the historical correlation to -0.65.

Even better, using the 4-week moving average of new job ads and the 4-week moving average of new UI claims, on a seasonally unadjusted basis, pushes the historical correlation to -0.69. Lagging the 4-week summation of job ads by one week improves the correlation to -.70. In other words, not only does the number of new job ads influence UI claims, but this is further evidence that Hiring Demand is a leading economic indicator. Lagging job ads by two weeks drops the correlation back down to -0.69.

Comments are closed.