The only reason for the decline in the unemployment rate to 9.5% was yet another decline in the labor force participation rate, which according to the BLS dropped another 652k people in the month of June. This resulted in a labor force to the civilian non-institutional population ratio of 64.7%: the second lowest number in decades of data, and only better than December 2009, when this number was 64.6%. The problem with this is that it badly underestimates the split between those who are marginally attached and those 14,623 who were formally unemployed in June. As the chart below shows, the double dip in the labor force participation is now very much pronounced.
What this chart implies is that if there was a mean reversion to the last 10 year labor force participation average rate of 66.2%, there should be another 3.5 million jobless added to the 14.6 million tally. And as this differential is the easiest thing in the world for the BLS to fudge, adding the two and dividing by the labor force of 153,74, we get an unemployment rate of 11.8%, leaving aside all other such fudge factors are government hiring, temporary workers, birth death, etc. 9.5% or 11.8% - which one is more realistic for an economy finally realizing it never left the second great depression, you decide.