Labor and economy watchers may have noticed different takes on June’s slight bump in the state unemployment rate. Does it reflect the IBM layoff, announced June 12? Or not?
The state’s economists came down on opposite sides of the question at their meetings with the Legislature’s Joint Fiscal Committee and Emergency Board. They met July 23 to review the past fiscal year’s budget and this year’s forecast.
It was just a few days after the state released its updated unemployment report for June, and almost two weeks after the last day of work for more than 400 employees let go from the IBM manufacturing facility in Essex Junction.
In Tom Kavet’s brief to the Joint Fiscal Committee, the consulting economist for the Legislature said yes: “Even with this last uptick in June — the unemployment rate went from 4.1 to 4.4 percent, some of the IBM layoffs are counted there — but it’s still the fourth lowest in the country.”
In Jeff Carr’s brief to Gov. Peter Shumlin and the Emergency Board, the governor’s consulting economist said the opposite: “IBM isn’t in any of the numbers yet. … They didn’t go off the payroll until 30 days after the (layoff) announcement. So we haven’t seen any impact on our employment numbers yet from the IBM employment action.”
The different interpretations, also reflected in some news reports about the monthly labor update, highlight the ambiguity of unemployment numbers.
Unemployment rate
A lot depends on the unemployment rate: federal monetary policy, economic forecasting, political campaigns. It’s a (hopefully) small number that means a lot.
There are some things, however, the unemployment rate doesn’t mean. It doesn’t mean that X percentage of people are actually unemployed; it means they feel that way. The unemployment rate is an estimate of the number of people who feel unemployed.
State unemployment numbers come from the federal Bureau of Labor Statistics. Part of a bigger dataset about the population, the unemployment rate is derived from monthly surveys conducted for the Current Population Survey.
The CPS is like a cousin of the census: Instead of literally asking every person a set of questions, the survey asks a representative sample of the country. The responses are then used as a proxy for the whole. There’s a lot to it, even disagreement about its accuracy.
The government has settled, therefore, on the sample survey as the most reliable method for approximating the unemployment rate. But that’s not to say the CPS is without limitations of its own.
But in the context of Vermont and the IBM layoff, one thing is clear — sort of.
The unemployment rate is not a tally of how many people are filing claims for unemployment insurance (UI) benefits. As the BLS points out and as the long-term unemployed can attest, some people remain unemployed after their benefits run out. Others may lose their jobs but not apply for benefits, and still more may not even be eligible to collect unemployment.
“So, quite clearly, UI information cannot be used as a source for complete information on the number of unemployed,” the bureau explains.
The government has settled, therefore, on the sample survey as the most reliable method for approximating the unemployment rate. But that’s not to say the CPS is without limitations of its own.
Vermont’s Economic & Labor Market Information Chief Mat Barewicz attributes some of its constraints to a “human” element.
“Humans will answer surveys any way they feel that day,” he said.
This means that June’s IBM layoff may — or may not — have been reflected in that month’s unemployment rate. Most laid-off workers were given a 30-day notice, meaning they technically were employed until their last day on July 12, as Carr pointed out. But if any of those employees were in June’s CPS sample, or if survey takers had just learned about the IBM layoff, it potentially could have colored their responses.
There’s no way to say definitively that the IBM layoff contributed to the slight bump in Vermont’s unemployment rate in June. But it’s also impossible to prove it didn’t.
