The state Department Division of Employment Security (formerly known as the Employment Security Commission) released its employment report (PDF) for October earlier today. Of interest: in seasonally adjusted terms, government employment in North Carolina increased by 4,800 from September to October. (Government employment not seasonally adjusted — in other words, the raw count of jobs — grew by 13,900 over that same period.)

This paragraph in the DES press release by Larry Parker caught our attention:

While the Government sector experienced the largest over-the-month employment change, it is important to note that Government employment estimates are subject to large seasonal patterns, especially during the summer and fall. This is largely due to the contraction and expansion within the public education sectors. Seasonal adjustment factors are applied to the data. However, it appears that these factors may not be fully capturing the seasonal trend. Therefore, when interpreting the changes in Government employment, it is advisable to focus on over-the-year changes in both the seasonally adjusted series and the not seasonally adjusted series.

Why, this sounds a lot like something Carolina Journal’s Don Carrington (who used to be the deputy director of the ESC’s Labor Market Information Division) wrote about last month. Consider:

ESC serves as the local agent for the federal Bureau of Labor Statistics, which produces the employment estimates. The “seasonal adjustment” factor used by BLS for public education jobs in the summer, for instance, is little more than a guess. It’s an estimate based on observations from several recent years of the number of jobs that should exist if all schools remained open year-round and there were no summer break.

A more reliable way to look at public employment is to consider an actual count of jobs, the so-called not seasonally adjusted or unadjusted figure. Since the fiscal year began July 1, the ESC’s seasonally adjusted employment estimates show a net gain of 3,600 state and local government jobs. By contrast, the unadjusted jobs figure for local education has plummeted — from 229,100 in May to 132,300 in July. Such drops are typical in the summer when schools are out of session. The unadjusted jobs number rebounded to 189,300 in August and should rise even more when the September report is released Oct. 21.

ESC officials defend their approach. Agency spokesman Larry Parker told CJ that seasonally adjusted estimates are “accurate.” In a Sept. 16 interview with Raleigh television station WTVD, Deputy Chairman David Clegg said, “We are using the same methodology that’s been used for decades. We are looking at the same snapshot of North Carolina’s economy each and every month.”

[SNIP]

ESC also emphasizes seasonally adjusted numbers in its monthly release rather the raw estimates.

The BLS raw count of government jobs in North Carolina drops by as much as 100,000 every summer, because most public school teacher jobs are not counted during the summer break.

ESC’s Parker told CJ that most economists would agree a more reliable method of measuring changes in government jobs would be to compare unadjusted employment numbers from September 2011 with those from September 2010.

We report, you decide.