Last night I noted that many polls are showing a huge advantage for Democrats in their samples, sometimes as large nine, fourteen, or sixteen percent. I noted that since 1988, the largest advantage either party has ever had was four percent, in 1996 (39-35).
I was trying to find the partisan breakdown in the electorate for the last horrific era for the GOP, the post-Watergate years of 1974 and 1976.
The Post has the New York Times/CBS exit poll for 1976, putting it at 41 percent Democrat, 34 percent Independent, 25 percent Republican. (It’s really interesting to note that the 16 percent Democratic advantage yielded… a 2 percent win for Jimmy Carter.)
And Campaign Spot reader James writes in:
From Craig Shirley’s excellent book on the 1976 GOP primary, “Reagan’s Revolution”: Shirley says that in 1974 Democrats had a “two to one” party ID advantage, and that “only 22 percent of the American people claimed allegiance” to the GOP that year. Shirley also cites “iconoclastic conservative pollster Arthur Finkelstein” as saying that in 1974 “Democratic turnout did not exceed the normal pattern for an off-year election, but Republican turnout was off by ten percent.”
So ask yourself two questions: (1) Do Democrats have anything close to a 2-1 edge this year in party ID? (2) Will GOP turnout drop 10% as against 2004?
Jay Cost writes in with additional data to chew:
I don’t have the exit poll data, but here’s the National Election Studies breakdown:
1972: 54 D; 13 I; 32 R
1974: 51 D; 15 I; 31 R
1976:52 D; 15 I; 33 R
1978: 53 D; 14 I; 31 R
1980: 52 D; 13 I; 33 R
1982: 55 D; 11 I; 32 R
1984: 48 D; 11 I; 39 R
1986: 50 D; 12 I; 36 R
1988: 47 D; 11 I; 41 R
1990: 51 D; 10 I; 37 R
1992: 50 D; 14 I; 39 R
1994: 47 R; 11 I; 42 R
1996: 51 D; 9 I; 39 R
1998: 51 D; 11 I; 37 R
2000: 49 D; 12 I; 37 R
2002: 48 D; 8 I; 41 R
2004: 49 D; 10 I; 40 R
These numbers are going to be different from the exit poll b/c these respondents are not necessarily voters. In fact, these outsized Dem advantages can profitably be contrasted to the actual exit poll data – for the fact that the electorate tends to be more Republican than pre-election polls might otherwise predict.
Despite this pro-Dem tendency, you might find them of use for the purposes of analyzing pro-GOP years versus anti-GOP years. E.g. the swing from ‘82 to ‘84 was D+13 to D+7. ‘86 was another Dem year and that was Dem +14. There’s a similarity there.
But Jay and I agree that it’s worth noting is that the electorate of the mid-70s is different from today’s electorate. A lot of voters, particularly Southerners, saw themselves as Democrats but were conservative in nature. Today, most of those voters identify themselves as Republicans.
In the past 20 years, the worst cycle for Republicans was 1996, and even that year, Bob Dole took 46 percent of the two-party vote. (Four years earlier, George H.W. Bush took 46.7 percent of the two-party vote.) There won’t be a significant third-party vote this year (cue the screams of protest from the Barr/Nader/McKinney fans). If the undecideds split evenly, McCain loses by the current RCP average of 7 percent – a result that would be frustrating for the GOP, still considerably less than the 10 percent margin in the ABC/Washington Post poll, the 11 percent margin in Newsweek, etc.