As reported in a story by Mark Williams in today’s Columbus Dispatch, Ohio suffered a very large net loss of 21,000 jobs in September. The largest contributors to this loss were leisure and hospitality (down 6,300 jobs) and two other sectors that had been doing well: private education and healthcare (down 5,200) and manufacturing (down 4,700).
This is an impact of the weak national economy. But as I point out in the article, I tend to be suspicious of a very large change in state or local employment (up or down) that is not echoed in the national numbers. The survey generating these estimates draws from a fairly small sample (about 140,000 employers nationally representing about 440,000 worksites) so there is a fairly large sampling error in the Ohio estimates — and even more in the Columbus estimates. So if we don’t see this decline whittled down in the revision next month, we may when the estimates for all of 2011 are revised next March.
The fact remains that even with this large decline, Ohio’s employment recovery is still outpacing the national average: a gain of 1.9 percent vs. 1.6 percent nationwide. Since February 2010, Ohio has gained 92,800 jobs, with manufacturing, wholesale, business services, healthcare, and government all outperforming the national average.
A separate survey of households generates the unemployment rate, which held steady at 9.1 percent, equal to the U.S. average. This paints a very different picture: it shows 3,600 more Ohioans employed in September than in August. Almost exactly the same number of people entered the labor force, so the unemployment rate remained constant. This is based on a much smaller sample than the survey of businesses, so it is less reliable. It is also a different question: the number of people working rather than the number of jobs. When the two are in conflict, as they were this month, I tend to side with the survey of businesses. So in this case, the truth probably lies somewhere in between — probably closer to the decline than the increase.