Wisconsin is all abuzz about governor Scott Walker's claim that he added 33,000 in the state since he took office, putting him well on the way to the 250,000 he promised for his first term in office. He now leads challenger Tom Barrett by 6 percentage points in polling, so he and his wealthy and/or corporate backers are displaying great confidence.
I wouldn't be so confident if I were Walker. The ongoing John Doe investigation in Milwaukee is likely to result in an indictment, and if it happens before the recall election, he is finished.
The Milwaukee Journal Sentinel, among other corporate media, supports Walker's job claims, stating "Economists generally laud the validity of the quarterly census figures." The census is compiled by the Wisconsin Department of Workforce Development, and is known as the Quarterly Census of Employment and Wages. The Journal Sentinel doesn't say who any of these economists are, which aroused my curiosity. Then on Friday evening a Wisconsin Public Television show aired an interview with Edgewood College Economics professor Bill Duddleston, in which he said he was withholding judgement until the numbers are verified by the Bureau of Labor Statistics in late June, well after the recall election.
He also said that in his experience the increases in employment in Wisconsin have been in the agricultural sector, and he wondered why the sectoral numbers haven't been released. That got me wondering, and I remembered looking up BLS figures for an Economics class I taught over the winter months, which showed the same trend. I took another look, and came up with the following analysis:
In the figures below, farm employment can be found by subtracting total nonfarm employment from total employment. For example, for April 2012 total employment was 2,863,600. Total nonfarm employment was 2,732,000. Subtracting nonfarm employment from total employment yields 131,600 for farm employment.
The same can be done for March 2012. Total employment was 2,856,800, and nonfarm employment was 2,737,900. Farm employment was 118,900. By subtracting farm employment in March, 118,900, from farm employment for April, 131,600, we see an increase of farm employment of 12,700. This is easily explained by the onset of the planting season.
What also can be seen from the figures below is that nonfarm employment decreased by 5900 - 2,737,000 minus 2,732,000. Though total employment increased by 6800 from March to April, nonfarm employment was lower by 5900 workers.
In January 2011 total employment was 2,833,068. Nonfarm employment was 2,744,800. By subtracting nonfarm employment from total employment, we get a total of 88,268 for farm employment. In April 2012, with farm employment of 131,600, we have an increase over January 2011 of 43,332. With nonfarm employment in January 2011 at 2,744,800, and at 2,732,000 in April 2012, we have a decline of 12,800 over this period. The increase in farm employment, 43,332, less the decrease in nonfarm employment, 12,800, yields an overall increase in employment of 30,532. Using a different sampling method might result in higher totals, but would nonetheless show a decline in nonfarm employment offset by a much larger increase in farm employment.
Whatever sampling error exists in the measurement, it is what social scientists call systematic error, where trends over time would be accurate because the error would be consistent in each sampling. For example, in the present consideration of Wisconsin employment, while the totals may not be 100% reliable, the changes over time are. If an increase in employment can be accounted for by seasonal increases in farm employment, then a politician claiming responsibility for creating jobs is similar to taking credit for the sun rising.
In the case of Governor Walker taking credit for creating jobs, another question is raised by using an inflated sampling method, but comparing the data with totals from the deflated Bureau of Labor Statistics method. If he were to compare his totals from the new method to previous totals from the same new method, he would be comparing apples to apples instead of apples to oranges, and would almost certainly find the same or nearly the same results as the BLS found using the deflated non-Walker method.
Another factor that no one has considered is that, due to the recent crackdown on undocumented workers, many jobs that had been held by “illegal” immigrants off the books are now held by regular employees, who would be counted in the labor force, and in employment statistics. There is no way of telling how many jobs have been added to the employment totals, both farm and nonfarm, but it is likely in the thousands.
So Walker is lying in three ways. He says he is creating jobs when farm employment is following seasonal increases. He is comparing one method of measuring employment for the most recent period with a deflated method for the previous period, and he is very likely attributing a substitution of documented employment for undocumented employment as an increase in jobs, when the very opposite is true.
Walker is also a liar in a fourth way. Even if you give him all his other lies, there is no established connection between his mythical job growth and anything he has done as governor, other than the passage of time with him in office. In logic this assumed connection is known as Post hoc ergo propter hoc - after this therefore because of this.
He actually can be blamed for real job losses, though, because employment is for real growing in neighboring states and nationwide. The only thing discernible that is different in Wisconsin from other states that could cause job losses is the residence of one Scott Walker in the governor's mansion. In an improving economy, Wisconsin is experiencing a decline in nonfarm employment. He should be tossed out of office by acclamation.
Economy at a Glance
Wisconsin
___________________________________________
Little did I know that someone else would be taking the title I used before I did. When it's Rachel Maddow I can live with it.
For a history of the term "lies, damned lies and statistics," click here.
Update, May 23: Walker may be finding himself in more trouble with his lying. Read about it
here.
Update, May 24: In today's news, Walker sneaked in another
tax break for corporations in his new budget. There's more. He is
claiming that opponents of his voter ID law, like the League of Women Voters, are trying to promote fraud. Given Walker's clearly established habit of lying about everything under the sun, it is safe to say that he not only is lying about this, but is himself attempting vote fraud with his voter ID law. This alone is reason enough to recall him, but he really should be impeached, prosecuted as a criminal, and if convicted, sent to prison, where he should be breaking rocks in the hot sun.
This post can also be seen here, here, here and here. It gets more reads elsewhere, and more comments on a couple of them.
Update, May 27: A letter I wrote about the recall appeared in the Wisconsin State Journal and Capital Times this past week. They did similar editing jobs.
Update, May 30:
Here's the Wisconsin Department of Workforce Development's official job numbers. This is a press release, so the numbers are presented to put Scott Walker in the most favorable light, but the results are consistent with what I have presented. Similar results can be seen
here.
Update, May 31:
The plot thickens. All the king's scribes and all the king's spinmeisters won't be able to put this Humpty Dumpty together again. It's getting like whack a mole, but I keep
responding.
Update, June 3: The plot thickens
even more. And
still more. We'll be getting a new governor, one way or another. I prefer one way, but will settle for another.
Here's another something that bodes not-so-well for our criminal governor.
If you didn't make it to the Get Out The Vote rally on Friday, there are a number of videos of it on YouTube.
This is the best one.
Here's another version. I'm not much into Hip Hop, but Brother Ali really rose to the occasion with his smokin' lyrics to this Woody Guthrie classic.
Here's a news story about the rally.