Thursday, June 25, 2009
initial claims jump
My view is the most useful number in the weekly claims report is the number of seasonally adjusted initial claims (with a 4-week moving average because it is so noisy). This has declined from the peak of 10 weeks ago, but is still very high. This suggests that the peak of job losses might be behind us, but also that there are still significant job losses occurring. We will probably see monthly job losses reported by the BLS until the weekly initial claims numbers declines close to 400 thousand.
ed harrison has noted that the seasonal adjustment may obscure the turn in claims, however, so looking at the unadjusted number provided by the department of labor
The advance number of actual initial claims under state programs, unadjusted, totaled 566,586 in the week ending June 20, an increase of 8,548 from the previous week. There were 358,159 initial claims in the comparable week in 2008.
here's the fed chart (link is updated; snapshot at left is through last week only). the trend in both adjusted and unadjusted is still clearly indicating a slow dissipation of the unemployment cycle -- but a pessimist will want to keep an eye on the unadjusted claims low point in the may 30 report of 500,383. this june 20 report is some 13% higher that that reading. obviously there is considerable volatility in the series and no real conclusion can be drawn -- but the low reading in a normal year comes in august. (2008 was clearly not a normal year.) if unadjusted claims continue to rise -- or even stay level -- it implies a strengthening employment problem.
more from harrison today. note that both he and CR are discounting possible improvements in continuing claims, noting that even extended benefits are running out for many former workers and thereby reducing the count without reflecting any improvement in hiring.