Bayes' theorem states that P(A|B) = P(B|A) * P(A) / P(B), so in this case it says P(PhD|jobless) = P(jobless|PhD) * P(PhD) / P(jobless). It seems that you state

Bayes' theorem states that P(A|B) = P(B|A) * P(A) / P(B), so in this case it says P(PhD|jobless) = P(jobless|PhD) * P(PhD) / P(jobless). It seems that you stated it incorrectly but used the correct version in the computation (the incorrect gives 0.2244).