The Anomaly of the Surprising September Unemployment Rate

The unemployment rate has dropped two months in a row. The Department of Labor announced Friday that it declined to 7.8% from 8.1%, the first time it dipped below 8% in the 43 months of the Obama Administration. The unemployment rate was 8.3% in July.

The economy must be improving. President Obama clearly let out a sigh of relief.

Conversely, Jack Welch, the retired Chair of General Electric, claimed the Administration had engaged in a conspiracy to fix the rate.

If it doesn’t make sense, then it’s probably wrong. If the numbers don’t add up, something’s wrong. If it doesn’t look right, then it’s probably wrong.

However, it’s not a conspiracy, but an anomaly due to small samples.

Let's look at the underlying statistics and methodology to understand the anomaly.

The reports on new filings for unemployment compensation are based on actual state reports.

However, the monthly unemployment rate is based on surveys, and thus has to be revised as more information comes in. For example, the initial August report of 96,000 new jobs was increased to 142,000 and July was revised upwards to 181,000 from 141,000.

The Bureau of Labor Statistics surveys 141,000 of the 6 million private employers and 345,000 federal, state and local government agencies in the United States to estimate the unemployment rate. This survey estimated 114,000 new jobs were created with 104,000 in the private sector and 10,000 in the public sector. (How can we be adding jobs in the public sector in this economy?)

However, a different survey of 60,000 households out of 117 million American households reported 873,000 new jobs.

Whether you use a calculator, adding machine, slide rule, abacus, math tables or your fingers, these numbers don’t add up.

They also tell an incomplete tale. 12.1 million Americans are counted as unemployed. The number though of Americans who are unemployed, underemployed, or no longer looking for work remained constant at 23 million. 8.6 million Americans of the 23 million were working at part time jobs in September, up from 7.7 million in March.

The number of Americans unemployed or working at part time jobs remained constant at 14.7%.

A study from the National Employment Law Project reported in late August that most of the jobs added in the recovery were low paying jobs; lower wage occupations with median wages of $7.69 to $13.83, accounted for 21% of the job losses during the economic retrenchment, but 58% of the job gains in the recovery.

We also know that Federal Reserve Chairman Ben Bernanke announced on September 13 that the economy remained weak. He said the “weak job market should concern every American.” He announced a third round of “quantitative easing” with the Fed buying $40 billion of mortgage backed securities in the open market. Thus, the Fed is continually to pump money into the economy.

We also know that the economy grew only 1.7% in the second quarter, down from 4.1% in the last quarter of 2011.

Employment is not going to rise substantially as the economy is slipping in an already weak recovery. The problem lies with the methodology by which the BLS computes the numbers.

If the October report is an anomaly, then the November numbers, released 5 days before the election, may show a spike upwards. That will not be good news for the President.