One has to apply the usual caveats to survey data here. For one, the mean can be skewed upward by outliers, so median is probably a more informative statistic. Two, who is being surveyed? If half of pathologists were unemployed and only employed pathologists were surveyed, you wouldn't be seeing the whole story. Three, are survey answers likely to be truthful? Mikesheree has stated IIRC that he underreports in surveys like this. Others might overreport for some bizarre reason, who knows.
Whatever you think of this survey, please remember that NOT EVERYONE ON THIS FORUM THINKS THE JOB MARKET IS BAD. There are a few very frequent posters who repeat this stuff, and there are numerous others like myself who don't comment much on the job market but who think it is fine.