Thursday, November 29, 2012

Initial Jobless Claims: comparing Katrina and Sandy


 - by New Deal democrat

Initial jobless claims fell 23,000 this past week but were well above the trend before Sandy hit at the end of October.  Is this something we should be concerned about, or is this still the effect of the hurricane?  A comparison with jobless claims after Hurricane Katrina suggests that while most of the effect has worn off, the entire effect won't abate for a few more weeks.

Here are initial jobless claims beginning with the first week of August 2005:

2005-08-06 311000
2005-08-13 319000
2005-08-20 315000
2005-08-27 318000 * Katrina hits 8/29
2005-09-03 326000
2005-09-10 422000
2005-09-17 424000
2005-09-24 359000
2005-10-01 384000
2005-10-08 383000
2005-10-15 348000
2005-10-22 324000

As you can see, the effects of Katrina on jobless claims weren't felt until 2 weeks later, when claims shot up by about 100,000.  After two weeks, most of the effect wore off, but claims were still at least 30,000 (and as much as 65,000) above the pre-Katrina baseline for 4 more weeks.

Now here is the current data beginning with the first week of October:

2012-10-06 342000
2012-10-13 392000
2012-10-20 372000
2012-10-27 363000 * Sandy hits 10/29
2012-11-03 361000
2012-11-10 451000
2012-11-17 416000
2012-11-24 393000

 Just as with Katrina, claims didn't spike up for 2 weeks.  Just as with Katrina, most of the spike has disappeared 2 weeks later, but we are still significantly (about 30,000) above the pre-Sandy trend.  If the pattern holds true, we will have 3 more weeks of elevated data due to Sandy before the underlying trend is discovered.