Josh Elliott-Traficante

October 2012

The official unemployment rate has come under fire as of late. The past year has given rise to the charge by many that the unemployment rate under represents the true state of unemployment. Many have cited the U-6 rate which is currently around 14.0%. (For a description of the different rates, click here) The recent drop in the unemployment rate to 7.8% in September brought charges of manipulation for political motives.

While it seems highly unlikely that the Bureau of Labor and Statics is deliberately ‘cooking the books’, there is something to be said about the under representation of the true state of unemployment. This is not to say that the Bureau is doing this on purpose, but rather it is a symptom of the sluggish recovery that is not captured in the data. (For an explanation of the disconnect click here.)

As the recovery of the past two years has only been tepid, at best, a number of individuals have dropped out of the workforce entirely, because they think that they cannot find work, so they do not bother trying. Under the current method of calculating unemployment, these people do not exist. Despite not being counted, they are still unemployed.

However, there is a way to capture these group and project what the unemployment rate would be, were they counted as unemployed. Using historical baselines, thereby only counting those not in the workforce due to the recession, it is possible to calculate the unemployment rate if they were included. When this is done, the unemployment rate for September would be 9.0%.

Below is a chart comparing the revised rate to the official unemployment rate:

The difference between the two rates grows slowly, albeit steadily as the recession began, only to widen drastically by early 2009. As the recession bottomed out and the slow recovery began, this spread has grown only wider to the current 1.2% point spread in the September data.

Though there is some wiggle room in the revised rate (see Methodology below), this is decently accurate way of incorporating those not in the work force, for reasons realistically attributable to the recession, into the unemployment rate.




In addition to counting those no longer in the workforce, the BLS also counts those no longer in the workforce, but want a job. Currently this subset stands at roughly 6.7 million. Looking back at the historical data, we find that even in good times there are people who fall into this category, for whatever reason.

The running average for this group of people between 2000 and the end of 2007 was just over 4.7 million. This subset remained within a fairly narrow band of the average; 96% of the monthly data points were within one standard deviation of it. In other words, during both the recession of the early 2000s as well as the recovery that followed, the total number of people who were not in the work force, but wanted a job, remained fairly consistent. This consistency over 8 years is key, as it gives the perfect baseline.

Using that 8 years’ worth of monthly average as the pre-recession base line, that number is then subtracted from each following month’s reported number of those ‘not in the workforce, but would like a job.’ The resulting difference is (attributable to the recession), is then added back into the labor force data for the month, as well as the total number of unemployed. These revised numbers now include those unemployed persons who have left the workforce due to the recession. The numbers are then divided, resulting in the revised unemployment rate.


x = Number of persons reported as “Not in the Workforce, but would like a job”

y = Revised Number of persons reported as “Not in the Workforce, but would like a job”


x – 4.714 million = y

                y + Total Number of Persons in Labor Force

——————————————————— = Revised Unemployment Rate

               y + Total Number of Persons Unemployed


Accounting for one standard deviation in either direction from the average of 4.7 million, yields only minor changes, a testament to the lack of movement in the data. Taking this into account, there is a band +/- .2% points above and below the revised rate.

1 reply
  1. Ray Pinard says:

    Excellent article. The next step to this discussion: what is the long-term impact the under/unemployment rate has on building individual wealth, personal assets, and the psychological impact on the long-term under/unemployed.

Comments are closed.