Brooks Bashing

I hate to jump on the Brooks bashing bandwagon, but in today's column he makes this claim:

"Republicans still have an advantage the higher you go up the income scale, but the correlation between income and voting patterns is weaker."

Well, not quite. Here are the correlations between income and Democratic vote for president from 1948 to 2000 (strongly negative numbers mean that higher income groups were less likely to vote Democratic than lower income groups):

1948: -.21
1952: -.06
1956: -.06
1960: -.02
1964: -.13
1968: -.03
1972: -.11
1976: -.19
1980: -.19
1984: -.23
1988: -.16
1992: -.17
1996: -.19
2000: -.14

First, it should be noted that the even the strongest correlation (-.23) is pretty weak, given that correlations can run from -1.0 to +1.0. Furthermore, the correlation has weakened since 1984 (Reagan v. Mondale), but the results for 2000 are still stronger than for any election from 1952 to 1972. These numbers, however, are a bit skewed by the fact that blacks (who disproportionately reside at the lower income levels) vote so heavily Democratic. Here are the numbers for whites:

1948: -.22
1952: -.05
1956: -.06
1960: -.01
1964: -.11
1968: +.01
1972: -.05
1976: -.14
1980: -.13
1984: -.17
1988: -.07
1992: -.13
1996: -.13
2000: -.09

As this shows, among whites, class had more influence on white voting in 2000 than in 1988.

Note: All data are from the National Election Study.