>>216602970
I think they took the racial statistics from the first which was data from 1994 and then scaled the numbers up or down using population projections from the second. For example the data in the first has non hispanic whites making up ~59.4% of child molestation charges with whites making up 75.7% of the population in 1994. they make up only ~56.3% nowadays so they could have kept the 1994 ratio and scaled the numbers so that a proportionally smaller number would be non hispanic whites today. this is all just a guess though
even using the 1994 numbers the per capita child molestation numbers for non hispanic whites are pretty close to -20% compared to the "average" including all races so the first column seems plausible from my napkin math.