It's not quite as bad as homeopathy. DNA evidence is worth something in these cases, but it's not conclusive -- the main problem is that government bureaucrats

It's not quite as bad as homeopathy. DNA evidence is worth something in these cases, but it's not conclusive -- the main problem is that government bureaucrats are too dumb to understand this and would treat it as conclusive. If we knew the number of Somali citizens with "non-Somali" DNA and the number of non-Somali citizens with "Somali" DNA, then... well, there's another application for Bayesian statistics, I guess.