Something has been bothering me for a while about the jobless claims reports, so I went back and checked the data. For once, I’m not going to opine. I’m just going to present the numbers, and let you draw your own conclusion. I report, you decide.
UPDATE 5/18/12: I found the last week that claims were not revised upwards. 6/19/11
|Week||Jobless Claims||Headline||Revision after 1 week|
|7/14/12||386,000||Weekly Claims Post Rebound; Jobs Market Still in Doldrums|
|7/7/12||350,000||Weekly jobless claims drop sharply to lowest level in four years||352,000|
|6/30/12||374,000||New unemployment claims drop; key report projects stronger job growth||376,000|
|6/23/12||386,000||Jobless Claims in U.S. Hovered Last Week Near 2012 High||388,000|
|6/16/12||387,000||US jobless claims fall, but 4-wk avg at 6-month high||392,000|
|6/9/12||386,000||Jobless Claims in U.S. Unexpectedly Rose Last Week||389,000|
|6/2/12||377,000||U.S. jobless claims fall by 12,000||380,000|
|5/26/12||383,000||Weekly unemployment claims rise modestly||389,000|
|5/19/12||370,000||U.S. weekly jobless claims dip slightly to 370,000||373,000|
|5/12/12||370,000||Jobless claims steady; Mid-Atlantic factories sag||372,000|
|5/5/12||367,000||Jobless claims fall marginally last week||370,000|
|4/28/12||365,000||Weekly Jobless Claims Ease; Productivity Drops 0.5%||368,000|
|4/21/12||388,000||Jobless Claims Fall 1K to 388K||392,000|
|4/14/12||386,000||Jobless claims –2K to 386K||389,000|
|4/7/12||380,000||Jobless claims unexpectedly rise||388,000|
|3/31/12||357,000||Jobless claims fall to lowest since 2008||367,000|
|3/24/12||359,000||US jobless claims fall to 4 year low in latest week||363,000|
|3/17/12||348,000||US jobless claims at 4 year low, lift recovery hopes||364,000|
|3/10/12||351,000||Jobless claims back at 4 year lows||353,000|
|3/3/12||362,000||US Jobless claims rise, but labor market healing||365,000|
|2/25/12||351,000||Jobless Claims fall by 2,000 to 351,000||354,000|
|2/18/12||351,000||Jobless claims hold steady at 4-year low||353,000|
|2/11/12||348,000||Jobs, factory data strengthen growth outlook||351,000|
|2/4/12||358,000||Jobless claims drop brightens labor market picture||361,000|
|1/28/12||367,000||Jobless claims fall, jobs market slowly healing||373,000|
|1/21/12||377,000||Jobless claims rise in latest week||379,000|
|1/14/12||352,000||US jobless claims near 4 year low||356,000|
|1/7/12||399,000||US jobless claims rise||402,000|
|12/31/11||372,000||Private hiring soars||375,000|
Getting too hard to find the articles this far back. But you get the point. Every single jobless claims report this year has been revised upwards the following week. And in just about every case, Reuters headline was either negated by the revision, or proven optimistic the very next week.
Here are the 4 week moving averages, as initially reported, and after revised.
|Week||Initially reported 4 week average||After revision|
The short? Jobless claims are lower than they were at the beginning of the year, but the rolling average has been trending the wrong way for the last several weeks. Bet you wouldn’t have guessed that from the Reuters reports, would you?
UPDATE: Added new data from week of 4/14 on 4/19
UPDATE: Added new data from week of 4/21 on 4/26. Also, you may notice that my 4 week rolling averages are slightly different than the released averages. Why? Because I actually take a 4 week average of the released numbers. Imagine that. BLS seasonally adjusts the averages. Of course, the numbers themselves are seasonally adjusted, so doing a second seasonal adjustment after averaging seems like fudging the data to me, but YMMV.
UPDATE 5/3: Added new data from week of 4/28. Big drop this week. But rolling average still going up. And last week’s revised upwards. Again.
I’ve been meaning to include this for the last couple of weeks. This is from ZeroHedge, who noticed not only the upward revisions, but their size:
Indeed, the most recent revision was the fourth largest on record. If one only read the headlines this year (at such blogs that specialize in architectural billings, deliveries by canoe, and what not), one would have the impression that Initial Claims have fallen nearly 25,000 this year. The revised reality? Barely over 10,000. And that’s a 10,000 improvement against a number that has been revised upward by 50,000. So, yeah.
Stay tuned for next week’s update.
UPDATE 5/10: SSDD. At least the 4 week rolling average went down this week. That’s the only silver lining in this week’s report. Well, I suppose there’s one more. This is two weeks in a row we’ve been in the 365k-370k range, a decent improvement from 380k-390k, which is where we were the previous month.
UPDATE 5/17: Repeat of last week. Again, the 4 week average dropped a bit. But it’s still a good 15-20K higher than it was two months ago. And that’s still not good.
UPDATE 5/18: I added a lot more rows to the tables. Last week claims were not revised upwards was 6/19/11. Initial report for that week was 429,000. After revision, 429,000. You’ll notice that it still wasn’t revised downwards. Average revision in the 46 weeks since then: ~+4,000 (3,978.26).
UPDATE 5/24: Ditto on my update of 5/17.
UPDATE 5/31: 48 weeks in a row. Also, a rather large jump over the last month. The last report in April is 365,000. Last report in May is 383,000. And the 4 week rolling average ended it’s three week run of dips.
UPDATE 6/7: 49 weeks in a row. Pretty big upward revision too. Starting next week, the Department of Labor is making some disturbing changes to the way this information is disseminated. But, I’m sure it’s entirely innocent.
Update 6/14: 50 weeks in a row.
Update 6/21: 51 weeks in a row. Oh, and the DoL has decided not to change their rules on publishing their information. For the time being.