Lying with Statistics - A short primer on government stats

The unemployment rate, one of the most useless government stats out there, has come down to 8.5%.

Why is this number useless? Because, while the primary concern should be with the total percent of the population employed, the unemployment rate excludes people who 'leave the labor force' which the BLS gets from surveys asking whether people are currently looking for a job. So if you've been unemployed for 30 weeks and have stopped looking for a job, you are dropped from the stats.

To see how this affects the unemployment rate, assume the current labor force (the number of people who are currently employed + the number of people who the BLS estimates are seeking jobs) is 150m people and the number of people seeking jobs is 15m. The unemployment rate is 10% (15m/150m). If next month, no one gets a new job but 1m people get discouraged and stop looking then there will be a new labor force of 149m (150m - 1m) and new job seekers of 14m (15m - 1m) giving an unemployment rate of 14m/149m = 9.3%. Hurray! Unemployment is down even though there are no more people working this month than last.

Here's the current total civilian labor force:

Notice that the total labor force has declined since 2007. At the end of 2007 it stood at 153705k and at the end of November 2011 it was at 153683k, this despite a population increase from 303287k to 311041k over the same period. So while the population has increased by 8 million people the number of people employed has remained virtually unchanged. Yet since the beginning of 2010 the unemployment rate has been dropping. This is almost entirely due to people leaving the labor force and is not a positive thing.