According to the initial unemployment insurance (UI)claims data for September 2010 there were 1,478,361 claims. According to the Bureau of Labor Statistics (BLS) website this represents an undestimate of jobs actually lost since it doesn't count voluntary quits which aren't eligible for unemployment insurance. Also it doesn't count those lost jobs which aren't covered by unemployment insurance in the first place. For example, self-employed workers, unpaid family workers, workers in certain not-for-profit organizations, and several other small (primarily seasonal) worker categories are not covered. In addition, the insured unemployed exclude the following:
- Unemployed workers who have exhausted their benefits
- Unemployed workers who have not yet earned benefit rights (such as new entrants or reentrants to the labor force)
- Disqualified workers whose unemployment is considered to have resulted from their own actions rather than from economic conditions; for example, a worker discharged for misconduct on the job
- Otherwise eligible unemployed persons who do not file for benefits
So we might safely conclude that at least a million and a half workers lost their jobs in September 2010. But, according to the Employment Situation for September 2010 put out by the BLS, there were 95,000 lost non-farm jobs in September. This is based on the Current Employment Statistics (CES) survey which surveys 140,000 businesses and government agencies each month. For these two figures to compute there would have to be 1.5 million minus 95,000 jobs created in September 2010 and 1.5 million jobs lost in order to jibe with the initial unemployment claims data. This would compute to a figure of 95,000 jobs lost because the BLS is computing net jobs created or destroyed not absolute numbers of jobs created or destroyed. We infer the absolute numbers that would be necessary according to the initial unemployment claims data.
While the number of initial unemployment claims is a hard and fast number, the number of jobs created or destroyed is based on a survey and this survey is inherently inaccurate. Not only is it based on a survey but part of the number is based on an econometric model! And with so many (actual not net) jobs lost in September (or in any other recent reference month) according to initial UI claims data, the net jobs created or lost according to the CES amounts to little more than a margin of error. Note that the BLS doesn't include data about the absolute number of jobs created and the absolute number of jobs lost, only the net. So one must assume based on the initial unemployment claims that the net involves at least 1.5 million actual jobs lost in September. The net of 95,000 jobs lost is so much smaller than the two numbers (jobs created minus jobs destroyed) that must be subtracted that it is inherently inaccurate.
This throws the whole measurement scheme of the BLS into doubt. Since their figures are inherently inaccurate, they can't be trusted. What we know for sure is that at least 1.5 million jobs were lost. This is a hard and fast figure. What we don't know is how many jobs were created (BLS only gives us the net). Therefore, my contention is that nowhere near the 1.4 million jobs that the BLS implies were created in September 2010 were actually created. This leaves us with the conclusion that huge numbers of jobs are being destroyed in the US economy and are not being replaced. The rest is statistical sleight of hand.
The fact of the matter is that the BLS would not even have to use a survey to estimate the number of jobs lost or created in any given month. It has access to hard and fast numbers instead. Those numbers can be derived from the payroll taxes that each employer pays. Employers pay a payroll tax for each employee. The amount of taxes paid divided by the tax per employee gives the number of employees employed per business. This can be summed up over all businesses and government agencies. The IRS has the data to come up with a hard and fast number for the number of employees and hence the number of jobs for the entire economy. The change month to month represents the jobs lost or gained for the current month.
In addition to statistical inaccuracies, the survey presents other problems. It relies on the honesty of businesses and government agencies to report the number of jobs created or destroyed. Private employers and government agencies have a disincentive to provide honest data about job destruction because they have a vested interest in making the economy look as good as possible and to cover up the fact that they are outsourcing jobs. They want consumers to be upbeat about the economy so that they will go out and buy stuff. Consumers are less likely to be upbeat if hundreds of thousands of jobs are leaving the economy each month. And I don't believe there is any penalty for underreporting the number of lost jobs so this in and of itself results in inaccurate data. As we all know, businesses will do everything within their power to profit themselves as the money spent on TV ads and lobbyists attests to.
The American public has a right to expect more accurate figures concerning job creation or destruction. They have a right to know if their jobs are being shipped overseas and to what extent. The BLS needs to revamp its techniques in order to provide the American public with more accurate figures.
