Recently I complained on a weblog about the failure of Cleveland media to write about the city's unemployment rate, which according to figures from the Ohio Department of Jobs and Family Services had exceeded 11% for sixteen straight months. This prompted the following e-mail exchange with George Zeller of CEOGC, who is widely regarded as the foremost expert on poverty in Cleveland.

George wrote:

First, the unemployment rates that you mention are phony, and they therefore do not deserve the coverage that you are complaining about...

Second, we do have some honest figures up on the CEOGC web site. One of them shows that Cuyahoga County has lost more than 5% of all its jobs during the past two years. Another way to look at that is that Cuyahoga County has 14.7% of Ohio's jobs, but it has suffered 24.3% of Ohio's job losses during the past two years.

www.ceogc.org/research/index.htm

Further, also on the web site is a comparison between local job growth and cash welfare cuts in all 88 Ohio counties. For every quarter during the last two years, the biggest discrepancy between job growth and welfare cuts in Ohio has been in Cuyahoga County. The gap between the two figures has exceeded 18,000 during every quarter of the last two years. The welfare cuts are continuing every quarter, despite large local job losses in those same quarters. I'll attach the figures here.

I wrote back:

George,

Thanks for all the stuff. That should be good for another entry or two.

But listen... you keep saying the state's figures are "phony", which I can well believe. But they are the official numbers, and I don't see a more accurate alternative among your research pieces.

How would you go about getting real numbers for workforce, employed and unemployed in the city?

Bill

To which George replied:

You ask some very good questions, Bill. So I will take it seriously and give you some answers.

You want to know how you get accurate unemployment figures for the city. The fact is, such figures DO NOT EXIST, ANYWHERE. It's a first rate scandal, but nobody will do anything about it.

My response when the real figures do not exist is to go to some proxies. If you slighly redefine your quesiton, from "what is the unemployment rate" to "how many jobs have we lost," then I can answer your question precisely. I attach the figures here, which are also posted up on the CEOGC web site. These figures are exact counts of jobs, and are not subject to the problem of being "phony." As you see, during the last two years Cuyahoga County has lost more than 5% of its jobs, meaning that we are in an ECONOMIC EMERGENCY here in Cleveland.

If you are looking for a number of people who are in the labor force, that ODJFS figure may be about as close as we get to a reasonable figure. But, if you want to know how many of them are unemployed, that is where the trouble starts.

The problem is not mainly at ODJFS in Columbus, where Dr. Keith Ewald is the director of the Labor Market Information Division that puts out the phony figure. I went to graduate school with Keith, and he is a highly competent fellow who is not normally crooked. But, the federal government tells Keith what to do, and they doom him to having phony figures. In that one WSJ article that I sent to you, they quote both Keith and I in the story, and we were not attacking each other.

The way that it works is as follows:

The statewide Ohio unemployment rate comes from a monthly survey done by the census bureau, called the Current Population Survey. This survey is designed to produce figures for the entire USA, so its sample is of the USA, not of residents of any state. Further, in some budget cuts that go all the way back to Reagan, they cut the sample size of the CPS. Further, it is a panel survey done by interviewers with laptops, who ask the same people month after month whether or not they are employed. Low income and unemployed people are much more likely than employed high income people to drop out of a panel survey like this. They take just the Ohio cases in the CPS sample and calculate a figure for the state of Ohio. They send that figure to Keith, who has a formula to break it down among the counties and large cities. If the statewide figure is wrong, which it is, then Keith's formula has to be wrong as well, which it is.

The figure is wrong because it is not a sample of Ohio in the first place, and because of the bias in the sample that I already mentioned. It was so wrong last year that the Wall Street Journal printed that article on their front page.

The only other alternative for getting a figure on unemployment comes out once every ten years: the decennial census. These are the only figures that are EVER available for neighborhoods. But, you don't want to get me going on the decennial census. Its figures are no good for two reasons. First, there is a large undercount of poor and jobless minorities in the census, which the current administration and supreme court refuse to do anything about. Second, the census was done in April of 1990, which was about four months before the recession started. So, all of their figures on income and unemployment are long since obsolete, even though they are brand new. But, if you do want an alternative underestimate of the labor force, the census would be the place to go. But, you can't use their unemployment numbers, which are three years old, and which do not measure all the lost jobs around here from the recession at all.

Having outlined the problem, I am now going to reprint your question, and then answer it precisely.

Q- How would you go about getting real numbers for workforce, employed and unemployed in the city?

A- ...I recommend that you use my attached proxies, which I guarantee to be both accurate and precise. You could get the same figures from the front page a couple of months ago in Crain's Cleveland Business, or from the front page a month later in the Cleveland Plain Dealer, right underneath the photo of LeBron James. (I am not making any of this up).

In addition, if you look at the most recent unemployment release on the ODJFS Labor Market Information web site, you will see that despite the very high level of new unemployment claims in Cuyahoga County, they show that unemployment went down since last year in both the city of Cleveland and in Cuyahoga County. This is, of course, preposterous.

If you do want an honest count of something, I'll give you the four week moving average of new claims for unemployment insurance in Cuyahoga County for the fourth week of May, the most recent figure available. In May 1999 before the recession started, this figure was 785. In May 2003 a couple of weeks ago, this figure was 1,546. Obviously we are much worse off now than we were four years ago. Plus, you can say with accuracy that right now we are seeing 1,546 Cuyahoga County workers laid off EVERY WEEK, a figure that is twice as high as it was four years ago. You won't find any unemployment figures anywhere that will say that the Cuyahoga County unemployment rate has doubled over what it was four years ago, but it has actually done so. That's about as close as anybody can get to a precise figure on this issue.

George (not Bush)