Unemployment Claims Decline, But Less Than Expected

Here is Goldman's take:
1. The biggest news in this battery of reports that the number of unemployed workers receiving extended or emergency benefits dropped a combined total of 508,000 in the week ending August 28. At this writing we have no clear explanation for this. Among the possible reasons to discount it: (a) Statistical noise; however, this drop comes two weeks after a 320,000 drop, with no significant change in the intervening week. (b) September is a strong month for hiring; payrolls tend to rise by about 500,000 between August and September before seasonal adjustment. Since the data on emergency/extended benefits are not seasonally adjusted they may simply reflect this seasonal rhythm, though we don't see anything similar in the corresponding week of 2009. (c) The declines of recent weeks are coming about two years after the labor market began to deteriorate seriously in the wake of the Lehman bankruptcy; since benefits are paid for 99 weeks, we may be seeing exhaustions of eligibility in the data.

2. Otherwise, the data on claims were mixed relative to expectations but fairly unremarkable. Initial claims fell 3,000 to 450,000, suggesting that last week's drop was not a fluke. Continuing claims fell 84,000 but from a level that was revised up sharply (up 91,000 from the level reported last week). None of the data points covered in this report applies to the survey week for the September employment report, which is being conducted this week.

3. Producer prices behaved as expected, with energy prices accounting for most of the boost in the headline. The core indexes for both finished goods and intermediate product rose only 0.1% with little in the detail on finished goods worth mentioning.

4. The current account deficit was likewise close to expectations, as the widening in the trade deficit already reported drove the increase. Balances on income and unilateral transfers improved by $1bn and $2bn, respectively.