Reading and Misreading Polls

The latest CNN/USA Today/Gallup poll finds the Democratic Party leading the Republican Party 55% to 39% among registered voters in the generic congressional ballot. Gallup asks this question to get a sense of how people will vote in this year's elections for the U.S. House of Representatives.

This is the largest lead Democrats have held over Republicans in the 2006 campaign thus far, and the largest lead Democrats have enjoyed among registered voters in a midterm election since 1982. In the 1992 presidential campaign, Democrats led Republicans by 20 percentage points following the Democratic national convention that year. Once likely voter turnout models are applied -- something usually done later in the campaign -- the Democrats' advantage on the generic ballot is reduced given higher turnout rates among Republicans than Democrats. Still, a lead of this size would suggest a solid Democratic advantage among likely voters and the likelihood of Democratic seat gains.

Just what exactly are these polls actually asking? We've got one poll that claims that 55% of Americans would think that Democrats in charge of Congress would be better. 39% think that Republicans would be better. Despite the gloss and hype, this poll, like most all other polls, doesn't mean a whole lot.

Why?

Simple.

Registered voters do not vote for Congress. They vote for their local Congressman or Congresswoman or Senator. They do not vote on the entire makeup of Congress. And if you were to poll on the individual candidates, and not the Congress as a whole, you'd find that most registered voters are mostly satisfied with their representatives.

Try reframing the questions as to whether they'd want to change their own representative so that Congress reflects a Democratic majority, and you'd not only get a different answer, but it would undermine the whole rationale of these kinds of polls.

Republicans currently hold a 231 to 201+1 advantage in the House. 55 to 45 in the Senate. How many of those seats are truly seats with competitive elections with a good chance of the seats shifting. In the House, you need at least 15 seats to change hands to give the Democrats an advantage. In the Senate, 6 must change hands to give Democrats the edge.

Are there simply enough people retiring in the Senate to give the Democrats the edge? Larry Sabato has a few ideas, and while he thinks that Democrats may gain seats, it doesn't look like it would be enough to win in the Senate.