According to a press release from the RI Department of Labor and Training — the last to report this data before the election — Rhode Island’s unemployment rate dropped to 10.5% in September, its lowest level since April 2009. The change was on the strength of the state’s ” largest monthly increase of employed RI residents since the Bureau of Labor Statistics implemented the current methodology in 1976.”
Another measure provided in the press release makes it seem plausible:
Estimated nonfarm payroll in Rhode Island totaled 458,000 in September, reflecting a gain of 2,000 jobs from the revised August employment estimate of 456,000.
It’s important to note that these are two different methods of measuring employment. The first number, above, comes from a survey of Rhode Islanders, asking them a series of questions about their employment status. The second number comes from a survey of Rhode Island businesses, which will later be benchmarked against tax forms, concerning the number of employees that they have.
As I’ve explained before, the numbers don’t have to correlate. First of all, the methodologies could produce different results. Second of all, Rhode Islanders may be employed in other states, or people from other states could be taking jobs in Rhode Island.
But shouldn’t both markers’ strongly pointing in the same direction be a very positive sign? Perhaps, but a few considerations make me wonder.
First, a table in the email version of the DLT’s press release shows 11.3% unemployment (per 1st survey) on the strength 459,700 jobs (per the 2nd survey) in September 2011. Yet, September 2012 has 10.5% unemployment on the strength of 458,000 jobs. That’s a 7.1% drop in the unemployment rate somehow corresponding with a 0.4% drop in actual jobs.
Second, note the revised August number provided in the above quotation: 456,000. That revision reduced last month’s number by 900, or down from 456,900. The DLT reported the initial estimate, that month, as an increase of 1,100 jobs… which turned out to be only 200 jobs. And that increase was based on a revision of July numbers that was 100 jobs lower than the initial estimate.
I understand how these numbers start to make the head swim, but the upshot is that, since April, every revision has been down (except June, which was zero), which has the effect of making each month’s increase seem bigger than it was. If you go through every such press release from this year (not including September) and add up the increases, you’ll think that there are 1,400 more jobs in Rhode Island than there were in December. But if you combine the increases of the revised numbers, you’ll find that 500 have been lost.
Mainly, this is a comment about how wary we should be about these numbers from month to month. For this month, we should be especially cautious, given the peculiar jump in employment nationwide.
The Bureau of Labor Statistics will release all state-level employment data tomorrow. It will be very interesting to see what other states had historic jumps.