Since the comparison between the sets of anecdotes affects the case for bias at PolitiFact, the "PFB Smackdown" feature takes the attempts by liberals to show a right-leaning bias at PolitiFact and exposes the errors therein.
Which brings us to "WisFactCheck":
Another day, another outrageous claim from the right-leaning Politifact.We at PFB admit that PolitiFact's work exhibits problems that harm both conservatives and liberals at times. And, of course, we expect that the legitimate examples of the former exceed those for the latter. So what's the evidence in this case?
The latest eyerolling claim from Politifact is that Sen. Lena Taylor's claim that Walker "has not created at single job" was not accurate:
WisFactCheck:
Let's look at the actual information. According to the DWD, there were 2,318,900 people in private sector jobs in December of 2010. In February there were 2,262,700 people in private sector jobs. Therefore, the state had a NET LOSS of 56,200 in January and February.WFC obtains its 2,318,900 figure for Dec. 2010 from a chart listing non-seasonally adjusted job figures. Seasonally adjusted, the figure was 2,317,200 (a difference of 1,700). Where WFC used 2,262,700 for February, the corresponding seasonally adjusted figure for February was 2,331,500 (a difference of 68,600). In short, WFC completely flubbed the math, perhaps failing to notice that the DWD numbers were reversed from one report to the other. The seasonally adjusted numbers came second on the January report (the origin of the Dec. 2010 figure) but first on the report covering February.
Rather than a net loss of 56,200 the seasonally adjusted private sector employment numbers in the reports show a gain of 12,600.
WFC:
Politifact was looking at January and February since that was the latest information available when Taylor made the remarks on March 29, however, if March's private sector job count (2,275,300) is included Walker's total is a net lost (sic) of 43, 600 in the months that he has been governor.The first paragraph is shown wrong by our preceding analysis (though WFC also misreported the seasonally adjusted employment figure, which is actually 2,341,300 for March). So what about that supposed loss of 3,200 jobs from January to February? A researcher concerned about seasonal distortions ought to use seasonally adjusted figures rather than simply pretending that seasonal work ends by the end of January and remains absent throughout February.
In fairness, there is a jobs-bumb (sic) because of retailers hiring for Christmas, so let's just look at at January numbers, which were 2,265,900. In February, there were 2,262,700 jobs, meaning that from January to February therewas (sic) a net loss of So (sic), when Taylor made her comments Wisconsin had a lost a total of 3,200 jobs.
The seasonally adjusted private employment number for Jan. 2011 was 2,327,300 and the corresponding figure for Feb. 2011 was 2,331,500. Rather than a loss of 3,200 the seasonally adjusted numbers show a gain of 4,200. The difference between the WFC figure and the correct (at least based on our source documents) figure is 7,400.
WFC's struggle with the facts continues despite a belated recognition of seasonal adjustments:
So, where did Politifact get the "more than 13,000" number? They link to a DWD press release that came out a month and half after Taylor made her comments. There is no reference to anything over 13,000, but presumabely (sic) they're using the "seasonally adjusted numbers," which as previously discussed, are when you take the real numbers and adjust them to take into account seasonal employment drivers (such as Christmas-time hiring). So, they're using a starting point of 2,328,500 private sector jobs for January, 2,331,500 for February, and 2,341,300 for March. Even though the March numbers are irrelevent because they weren't released until several weeks after Taylor made here (sic) comments, they're taking March numbers and subtracting the January to get 12,800, which apparenty constitutes "more than 13,000." If they wanted to use the seasonally adjusted numbers for job growth from January to February, the correct number would be 3,000.I detect nothing in the PolitiFact story that supports the WFC reconstruction of the reasoning process. The seasonally adjusted job gains amount to 12,600 from Dec. 2010 to Feb. 2011, but it's important to note that the original story does not limit its job count to private sector jobs. Total seasonally adjusted employment increased from 2,736,300 to 2,752,200 from Dec. 2010 to Feb. 2011, an increase of 15,900.
Unfortunately, the 15,900 figure does little to explain PolitiFact using the words "more than 13,000." Why not "more than 15,000" if PolitiFact used the same numbers I used?
I don't see an easy answer. The increase in total employment from February to March without a seasonal adjustment was 13,200. But it's hard to see why PolitiFact would use that figure based on the URL embedded in the story:
According to Department of Workforce Development statistics, the state had a net gain of more than 13,000 jobs in January and February. To be sure, Walker’s policies hardly had time to take effect in that time frame, but that’s not what Taylor is citing as evidence.The cited source contains no figures for January.
The WFC claim that the "correct number" of jobs created from January to February is 3,000 also fails to match the DWD figures. Seasonally adjusted private employment actually increased by 3,600 according to the DWD, using preliminary figures for February available as of a March 23 press release.
Here's the bottom line: When Taylor made her claim (Mar. 29, 2011), the only numbers available to her via Wisconsin's DWD press releases were preliminary January numbers, with Secretary Manny Perez announcing the addition of 10,100 private sector jobs for that month. That means that Taylor, lacking any other source of information, would have to assume negative or no job growth from February to late March in order to make an accurate claim that Walker created no jobs even if she were to count as lost 10,000 rail jobs lost in Wisconsin with Walker's rejection of federal funds for that purpose.
Is it reasonable to count the total number of rail jobs supposedly lost over the course of the rail program against job gains in the first month of Walker's governorship? Of course not. PolitiFact probably made its comparison to emphasize just how shaky was Taylor's grounding in fact.
PolitiFact's policy calls for grading politicians according to information available at the time. While that procedure seems to cut Taylor a break by allowing her to allude to the January numbers in her late March press release, Taylor should have recognized the importance of other information available to her at the time. To wit: She was making her comparison in March using January numbers. It is therefore reasonable and fair to point out Taylor's rash lack of caution and to point to the contradicting numbers released subsequent to her statement. Taylor should have recognized that numbers from February and March could contradict her claim. And so they did.
WFC:
3,000 versus 13,000 is important, because Taylor qualified her comments with the loss in jobs that Walker cost Wisconsin by destroying the high-speed rail deal, which would have created approximatly (sic) 10,000 jobs in Wisconsin.Taylor did not allude to "rail" jobs by name at all, at least not in her press release. She alluded to a loss of "1200 quality jobs" in her press release (yellow highlights added):
Therefore, even using the seasonally adjusted numbers, there is no validity in Politifact's claim that Taylor's claim was false.
(MADISON) Senator Lena C. Taylor (D-Milwaukee), ranking member of the Joint Committee on Finance, expressed her disbelief and great concern over provisions of Scott Walker’s budget as the committee began its agency briefings today. Taylor noted that Walker’s budget falls short of fiscal responsibility in the criminal justice system, eliminates over 1200 quality jobs, and fails to uphold at least three campaign promises made by the Governor.WFC engaged in a hyper-partisan spin job, failing to produce any valid evidence of a conservative bias at PolitiFact. The attempt stands as a profound embarrassment to "WisFactCheck."
Update 5-10-11: Added link to original source-Jeff
Good catch..
ReplyDelete