Friday, September 27, 2013

White New England Protestants, a liberal lot they are not

As a myopic provincial, the figure I conceptualize in my mind as politically representative of New Englanders is a theistically ambivalent liberal white Protestant. A SWPL, basically. N/A, proprietor of R/H/E Notes, who has set me straight on misconceptions in the past, may have done so again by showing that Romney enjoyed a comfortable edge among white Protestants in New England in last November's presidential election.

The GSS allows us to look back almost as far as LBJ and the Great Society to see how white New England Protestants have voted over the last forty years. To avoid regional skews (self-identified New England Republicans tending to be to the left of Republicans on the whole, etc), only presidential elections are considered. The following graph traces the percentages of white New England Protestants and the national popular electorate as a whole who voted for the Republican candidate over the last four decades of US presidential elections as though there were only two tickets to choose from:

Sample sizes are small, in the 50-100 range for each election, so some sampling error is surely present, but the trend is clear--most Protestant white New Englanders find their seats on the right side of the national assembly hall. Only in 1996 do they appear more Democratically-inclined than does the public, and then only marginally so. If they built the Cathedral, they haven't been its most pious votaries.

GSS variables used: PRES68, PRES72, PRES76, PRES80, PRES84, PRES88, PRES92, PRES96, PRES00, PRES04, PRES08, RELIG(1), RACE(1), REGION(1)


IHTG said...

To avoid regional skews (self-identified New England Republicans tending to be to the left of Republicans on the whole, etc)

Why would you want to avoid that? It is meaningful.

Audacious Epigone said...


It would make white New England Protestants look even more conservative than this method does, to an extent that it might be perceived as intentionally misleading on my part.

JayMan said...

As I explained to n/a over there, religion is a poor proxy for ethnicity (because who considers themself a "Protestant"? What is their genetic heritage?). Maine, for example, contains a fair amount of people of Scottish and Scotch-Irish ancestry. They would be technically "White Protestants", even WASPs, but it wouldn't be surprising that they vote right-of-center. We don't know how many of them are Puritans, which is what we're interested in knowing.

N/a has ran some interesting estimates using surname counts to imagine the share of the New England population that are descended from the original Puritans as well as the Catholic Irish immigrants. His numbers are interesting, but, really, nothing can finally settle the argument except actual genetic analysis of living New Englanders.

Audacious Epigone said...


There are several denominations that are transcribed as Protestant for the GSS question--everything included in the DENOM variable. I'm not making the same argument n/a is, as the confidence I have in assertions on ethnicity are as cautious as your own, though they do clearly cluster as census maps by county show, so self-reports--while surely not precise--are more than just randomness. I suspect they're more accurate than even Cochran expects (but don't let him know I wrote that!).

Anonymous said...

errr...corporeal guy...that chart shows the exact opposite of what n/a wants it to show.

i know n/a personally, and he has sex with animals on his farm and pimps them.

---jorge videla (the only commenter on hbd blogs who's actually in the bgi study)