Proving the Opposite

In an article purporting to show that Republican Congressional districts have more people who lack medical insurance, we see:

Among the nation’s 435 congressional districts, 207 have coverage levels below the average of 85.3 percent for the non-institutionalized civilian population, the census data shows. Of those, 105 are held by Republicans, while Democrats in mostly urban areas represent 100 and two are vacant. The rankings are based on the Census Bureau’s annual American Community Survey, which includes margins of error for each district.

Wait a moment … 105/232=0.45 whereas 100/200=0.5. The numbers cited prove the opposite of the claim. On the other hand, I don't think the difference is statistically significant.