The state of North Carolina has conducted a recount, and Vance County is the big winner by losing its grip on an unwanted top ranking.
Vance County no longer is the unemployment capital of North Carolina. According to the January county-by-county figures released by the state’s Employment Security Commission on Thursday, it’s not even close.
The preliminary numbers show Vance’s jobless rate at 9.2 percent, ninth-highest among the 100 North Carolina counties and well below new No. 1 Hyde County at 10.7 percent.
While those numbers might not be cause for dancing in the streets, local ESC director Sara Wester called them a “dramatic change.”
Consider that Vance:
* Had reported an unemployment rate above 10 percent in every month since April 2002.
* Had been in the top two for unemployment since July 2002.
* Had been stuck with the state’s highest jobless rate for 10 consecutive months and 19 out of 21 months.
But now all of those numbers are in doubt, thanks to a major overhaul of how the state’s Labor Area Unemployment Statistics unit measures the work force and estimates who’s working and who isn’t.
“They’re just counting stuff differently. Nothing significant has changed,” Wester said.
The revised statistical methods are difficult to explain — it took four state employees to clarify what had happened in Vance County — and the changes affected each county differently.
Although the new methods were first used for the January numbers, the ESC also applied the changes to the revised data for December to provide a basis for comparison. Those December numbers provide the best glimpse at how much things have changed and how difficult it will be to compare future jobless figures with the past.
Population totals increased across the state, but some of those newly reported people were working, and others were not. Four counties, including Mecklenburg, the state’s largest, had no change in unemployment rate between the preliminary and revised December reports. Fifty-six suddenly found their unemployment problems were worse than they had thought, and 40, including Vance, improved.
But most of the changes were no more than half a percentage point either way. A few were as great as 1.3 percentage points. Only Vance and Yancey changed more 2 percentage points — the unemployment rate in each fell 2.1 points.
If that’s not dramatic enough for you, consider this: In the preliminary December statistics, when Vance was the worst in the state at 11.2 percent, the ESC set the county’s labor force at 17,320; when the ESC revised the December statistics using the new methods and benchmarks, Vance had 17,479 employed residents. The revisions counted more people working than there were available to work before the changes.
The revised December numbers showed a Vance work force of 19,237, and most of the additional potential workers were, in fact, working.
Wester, who has tracked the local job picture for many years as the head of the ESC office in Henderson, couldn’t explain the statistical changes, which the ESC addressed in a two-page news release two weeks ago.
The first thing to understand is that the state doesn’t go around counting every worker and potential worker every month. Instead, like a political poll or television’s Nielsen ratings, the state measures a statistically significant sample and extrapolates the totals with a modeling procedure developed by the U.S. Bureau of Labor Statistics.
The new labor procedures use a revised model based on the monthly Current Population Survey and alter the benchmarking process by introducing reliable national numbers as a control group. The monthly county-by-county labor counts now should add up to match the overall figures found by the federal government.
All of which is a statistician’s way of saying what ESC Chairman Harry Payne Jr. said in a news release Thursday: “We hope a clearer picture of our state’s constantly changing work force will be provided each month.”
But the development of that clearer picture doesn’t immediately explain what happened in Vance.
Larry Parker, a spokesman for the ESC, said he was surprised when he was preparing Thursday’s news release and didn’t see Vance among the unemployment leaders. He couldn’t explain the change.
Pam Davenport could. She’s the supervisor of the ESC’s Local Area Unemployment Statistics unit in Raleigh.
She said several factors played a role.
For example, the state began reporting the jobless statistics for “micropolitan statistical areas” in January. Those micropolises are a Census Bureau innovation to explain the economic realities of rural cities. Henderson is the central city of a micropolitan area that consists entirely of Vance County, so now Vance gets a more thorough look at its job situation. Before, Vance and Warren counties were combined in a local statistical area, and some county numbers were calculated by figuring that Vance had a certain percentage of the overall population of the two-county area and thus had the same percentage of, say, new people in the work force. Now the state measures Vance directly.
The change in approach shows up, Davenport said, when the state estimates the number of 16- to 19-year-olds entering the job market for the first time and the number of people 20 and older re-entering the work force. With the monthly population surveys and the use of more small labor areas, the ESC can detect a rise in the work force that previously might have gone undetected.
But the biggest factor, the one you don’t have to be a statistician to understand, is that the state is doing a better job of measuring people who live in Vance County but commute to work elsewhere.
The federal Bureau of Labor Statistics counts people where they work, Davenport said, and the ESC must convert the data to reflect where those workers live.
Davenport said her unit at the ESC was estimating out-of-county commuters based on fixed ratios that were at least a decade ago. Now the ratios have been updated using the 2000 census, and the ESC forms estimates based not only on Vance’s ratios, but on those of neighboring counties.
In other words, if employment rises in Franklin, Wake and Durham counties, the ESC can make a good guess at how many Vance residents are taking those new jobs. So instead of picturing Vance as an island of unemployment so close to the high-employment Triangle region, the ESC model now embraces an image of Vance residents going to the jobs — not by moving their homes, but by driving their cars each day.
So the dramatic improvement in the Vance County employment picture doesn’t represent a jump in jobs available within the county. It does, however, indicate that the county has far more people who work than the ESC and the rest of us thought. Those monthly jobless rates of 13 percent to 15.5 percent the past few years probably weren’t right.
For now, you can’t compare new ESC figures with historical numbers.
Vance County’s jobless rate for January, 9.2 percent, was up slightly from a revised 9.1 percent in December, which Wester said is typical as companies cut back on workers hired for the holidays. The ESC said 86 counties had higher jobless rates in January than in December because the figures are not seasonally adjusted to account for such predictable changes.
The more meaningful comparison usually is to the previous year. Vance’s official jobless rate in January 2004 was 13.3 percent, and it was 13 percent in 2003.
But those figures were based on a less accurate system that likely undercounted Vance’s partial shift from manufacturing center to bedroom community. We have to wait for retroactive revisions of old data to see how bad things truly were the past four years.
The 2004 revisions should be done by April 28, the ESC said in an announcement this month, and revised numbers through 2000 should arrive in the coming months.
The old numbers seemed to show people drifting away from Vance County. The labor force fell steadily the past few years, and the belief was that people either moved away in search of work or became discouraged and stopped looking for jobs as their unemployment benefits ran out.
The new numbers indicate that most of those people are still here, and many are working again. They’re just doing it outside the county.
That finding fits with what Wester is seeing in the Henderson ESC office, which produced its own statistical oddity in January: The number of new unemployment claims, 249, was exactly the same as in January 2004.
“We’re certainly having job openings to be listed,” she said, but many of them are with companies in Vance‘s neighbors. Those businesses recognize that Vance has a surplus of labor, so “they feel like they can recruit from here.”
But within Vance, companies are shedding as many jobs as they’re adding. “It’s not bleak, but we certainly don’t have the manufacturing jobs people would like,” Wester said. “Companies are not coming to this area.”