The U.S. Bureau of Labor Statistics will be releasing the national unemployment rate for August on September 6, and the atmosphere is tinged with an extra element of suspense this time around. For most observers, the big question is this: Will the rate continue its downward trend from July, or will it jump back up again, as it did after the dip in April?

Obviously, we'll all have to wait until the release date to find out for certain, but national polling firm Gallup has done some research of its own, and the news isn't good.

An unexpected surge in unemployment?
Gallup and BLS each track national unemployment levels, and the pollster is noting that for August, the rate is probably going to be at least 8.8%. Considering that the month started with a Gallup estimate of 7.9%, this is disturbing, to say the least. With the government's reported jobless rate for July clocking in at a current-year low of 7.4%, I wondered just how reliable Gallup's numbers really are.

Gallup performs its own survey on unemployment, conducted daily and posted on its website. Since the pollster's is a rolling 30-day average, I took the last value for each day of a particular month, comparing it to the BLS statistics published for those same months. Placing these values in a chart, we can see how the two percentages have compared to each other for the current year.

Source: BLS and

Gallup and BLS have never been exactly the same in their estimates of the unemployment rate, but they are not far off, either. For any given month, the two have never been more than 0.4 of a percentage point apart, and BLS has actually reported rates higher than Gallup twice this year, so the latter doesn't just consistently track higher than the government's results.

Different research methods
Granted, the two entities do not use the exact same methods to come up with their results. BLS interviews approximately 60,000 households each month, using specific sampling criteria in order to reflect the entire U.S. Each month, 25% of these households are changed, and none are used as a sample for four months in a row. After the four consecutive interviews, those respondents are set aside for eight months, then used at the same time the next year, four months in a row.

Gallup conducts daily telephone interviews, encompassing a randomly chosen, monthly sample size of 30,000. While the government sample includes people aged 16 years and older, Gallup's respondents must be at least 18 years old. In addition, Gallup acknowledges its numbers are not seasonally adjusted, and that the margin of error is +/-1%. 

Food for thought
Despite the differences in research methods, the similarities in results can't be denied. This lends credibility to Gallup's claim that unemployment may be on its way up again -- an unfortunate set of circumstances for the overall health of the economy.

It seems clear to me that both BLS and Gallup have a high level of expertise when it comes to taking the pulse of working America. For those who want a truer barometer of the jobless rate on any given month, I suggest plotting halfway between the two metrics offered by these two entities. It generally won't be a prettier picture than the official story, but it will likely be a more realistic one.