As I wrote this morning, Democrats and Big Labor have reason to worry about Gov. Scott Walker’s recall election. Two developments today confirm this.

First, the Walker camp is touting this latest poll: “A new Marquette Law School Poll shows that with three weeks to go until the recall election Governor Scott Walker has taken a six-percentage point lead over Milwaukee Mayor Tom Barrett, 50-44 percent, among likely voters. Just three percent say they are undecided. In the previous poll, taken April 26-29, Walker held a one-percentage point lead among likely voters, 48-47.” You can understand why the DNC might not want to dump money into this race.

In addition, there has been a statistical battle over Walker’s job creation record. Today, the Wisconsin state Workforce Development Department put forth the rationale for a nice juicy number for Walker to crow about:

The year-over-year change in actual job counts as reported by some 160,000 Wisconsin employers shows a net gain of 23,321 jobs from December 2010 through December 2011. Meanwhile, the BLS’ Current Employment Statistics data series estimated a net loss of 33,900 jobs over the year. That puts the CES monthly estimates off from actual, employer-reported job counts by 57,221 jobs.

“It looks like 160,000 Wisconsin employers helped show us the thousands of new jobs that BLS estimates missed last year,” Secretary Newson said. “The bottom line is Wisconsin added jobs in 2011.”

The QCEW is a comprehensive count of jobs, as it includes reports from nearly all Wisconsin employers. This data series is distinct from the Current Employment Survey (CES), which estimates jobs based on a survey of roughly 5,500 employers (3.5 percent of Wisconsin businesses) and is released in preliminary form each month.

In essence, Walker is saying that an actual survey as opposed to a sampling of employers bolsters his case that more than 23,000 jobs have been added under his governorship. The Milwaukee Journal-Sentinel cast the job numbers in a somewhat favorable light:

Economists generally laud the validity of the quarterly census figures, which are time-consuming to compile. One reason they don’t figure as prominently in the political debate over jobs is the six-month reporting time lag.

The Bureau of Labor Statistics uses the quarterly census data to make revisions to the monthly survey data in order to retroactively clean up the inaccuracies that stem from extrapolation.

“The quarterly (census) data is much more reliable,” said Brian Jacobsen, an economist in Menomonee Falls with Wells Fargo Funds Management. “If that one’s showing job gains, that’s going to be tough to argue with. It’s a census as opposed to just a sample. That’s a reason why that survey is used for benchmarking purposes.”

Unlike most employment data, which is the property of the federal government, each state’s quarterly census data belongs to the states, according to the bureau.

That gave the Walker administration leeway with the quarterly data, which the governor seized. The early release amounts to a rare action that breaches tradition but doesn’t violate any agreements between the bureau and the state, officials at the U.S. agency said Tuesday.

Whether this accounting will change any minds remains to be seen, but the this coupled with the latest polling suggests that Democrats are and will have a tough time convincing Wisconsin voters that Walker’s conduct has been so egregious as to justify a recall. In fact, he’s done quite well.