Behind the job reports.

I quote from "A Lift for the Worst Off" By FLOYD NORRIS on the NYT Economix blog:

The overall unemployment rate fell in September to 7.8 percent from 8.1 percent a month earlier. But the really impressive figures were in the categories of people who have suffered the most. 

The jobless rate among people with college degrees was unchanged at 4.1 percent. But the rate among high school dropouts fell to 11.3 percent, the lowest figure for that group in nearly four years. It has declined by 1.4 percentage points over the last two months. 

Similarly, the number of people who have been out of work for more than six months fell below five million for the first time since mid-2009. It peaked at 6.7 million in the spring of 2010.

Obviously, this is good news for the economy and very good news for the Obama campaign.

I would point out that the television news soundbites have not focused on the fact that unemployment has been primarily a problem for the less educated, not for the college graduate. Nor have the minute-pundits described very well the geography of unemployment (except as it relates to the election in swing states).
I quote further:

(W)e should remember the household survey, on which the unemployment numbers are based, can be volatile. And the establishment report found 114,000 jobs were added during the month, in sharp contrast to the 873,000 additional people with jobs found in the household survey........ 

The two surveys often diverge. One is of employers, who are asked how many workers they have. The other is of households, who are asked which members of their households are working. If both surveys were perfect, there would still be differences, as self-employed people get counted in the household survey but not the other one, and people with two jobs get counted twice in the employer survey.

How many people understand the statistics?

In Chile recently there has been a public debate over the government estimate of the poverty level, with considerable vocal comment as to whether it is 15.1% or 14.8% or even 14.2%. When I lived there, there would be front page newspaper articles on how the market basket should be composed for the measurement of inflation.

Here in the United States the right wing has been accusing the Obama administration as skewing the employment figures. That paranoia might be related to the right wing willingness to draw upon skewed scientific findings to support its own positions; the government statistics bureaus are very, very professional and independent. I have worked with their personnel.

But few Americans really understand government statistics and their real limitations. Norris (the chief economic correspondent for the New York Times and a grad of one of my schools, U.C. Irvine) points out that the household survey gives different employment numbers than the survey of hiring by establishments. A lot more Americans these days are self employed, showing up as working in the household survey but not showing up as joining the work force of government nor a corporation. Are there also a non-formal sector and an illegal sector? I suppose so. The drug dealers don't respond to government surveys, nor do the people who pick up casual laborers to do odd jobs around their homes for pay. (Indeed, even our politicians have kept house maids off the books to avoid paying payroll taxes.)

In another post, Norris reported:

Mitt Romney said on Friday that there were 23 million Americans struggling to find work. It looks as if he got that wrong, by engaging in a little double counting. The real number is around 21 million. 

The just-released Labor Department report for September says there are 12,088,000 people classified as unemployed, meaning they looked for a job during the previous month and did not find one. That is the seasonally adjusted figure. The actual number the department estimated was 11,742,000...... 

There are 6,427,000 people counted as out of the labor force but wanting to work, and 2,517,000 classified as marginally attached. 

Add them together, and use the higher (seasonally adjusted) figure for unemployment, and you get 21,032,000. If you were trying to be fair and compare apples to apples, you’d used all numbers before seasonal adjustment, and get 20,686,000. 

Either way, that is a long way from 23 million.

Virtually no American voters understand these subtleties in the numbers, and even it fact checkers publish articles showing Romney's errors, few will read them and he will get away with misstatements to the general electorate.

Fortunately the much maligned bureaucrats in the federal government do understand these figure, and usually use them properly in policy formulation.