Political Animal

Blog

May 12, 2014 12:14 AM The Generic Ballot Poll is Useless for Senate Elections

By Martin Longman

Chris Cillizza makes some good points about the differences between House and Senate races and why the generic ballot poll isn’t a very reliable barometer for predicting how the Senate elections will turn out.

But he misses something that people really ought to keep in mind. It’s well known that the Democrats have a hard time turning out their base in midterm elections, but the Republicans see a drop-off in voter participation, too. So, let me put it this way. If you are Mitch McConnell, would you rather run for reelection in a midterm or in a presidential year in which your party’s nominee carries your state 60.5% to 37.8%, with a margin of four hundred thousand votes?

If you are Sen. Mark Pryor of Arkansas, would you rather run in a midterm or in an election in which your party’s presidential nominee gets only 36.9% of the vote and loses by over two hundred and fifty thousand votes?

The truth is, midterm turnout cuts both ways for senate candidates. Sen. Kay Hagan of North Carolina would definitely benefit from Obama’s turnout machine, but Mary Landrieu and Michelle Nunn are glad not to be running with Obama on the same ticket.

Another reason that the generic ballot question is kind of useless for predicting how the Senate will turn out is that only a third of the Senate is even up for reelection. It doesn’t matter what people in Florida or Missouri or Indiana or Utah think about the Senate because they won’t be voting for a senator in November, and neither will anyone in New York, Massachusetts, Arizona or California.

In any case, low turnout elections are bad for Democrats in a general kind of way, but Democratic senators running for reelection in deep red states would find it harder to win in a high turnout election because there are simply more Republicans than Democrats in those states. What they have to do is win over a considerable number of people who vote for the Republicans in presidential elections. That’s easier to do when a presidential election isn’t taking place.

Yet, as Jon Tester and Heidi Heitkamp just demonstrated in the last cycle, Democratic senate candidates can win in deep red states even in a presidential election year. So, basically, look at the races themselves and the candidates, their campaigns, their fundraising, and the polls. The generic ballot number is only useful for predicting House elections.

Martin Longman is the Web Editor for the Washington Monthly and the main blogger at Booman Tribune. He has worked as a community organizer for ACORN/Project Vote and as a political consultant for Democracy for America.

Comments

(You may use HTML tags for style)

comments powered by Disqus