Wednesday, October 27, 2010

On Polls and Early Voting

Lots of back and forth debate on websites, blogs, etc. over the meaning of the early voting statistics and their relationship to polling data.

One fallacious approach encountered in the discussion is comparing the early voting data to the 2008 or even the 2006 election. In both those elections, and particularly in 2008, Democrats won by landslides. In 2008, President Barack Obama won in states that many thought had been permanently lost to Republicans. Large numbers of Republican elected officials, many of them holding office for long periods, were unseated in both the House and Senate.

GOP usually leads GOTV
Normally, Democrats do not perform as well in early voting or in voter turnout (GOTV) as the Republicans. In fact, this is one reason that the GOP is able to stay competitive despite overall lower numbers of voters registered as Republican as compared to Democrat.

However, in 2008, Democrats crushed Republicans in early voting and GOTV allowing them to beat out many incumbents. Defeating sitting elected officials is generally very difficult. They have natural advantages in name recognition, press coverage, government-sponsored outreach, etc. However, the Dems led in practically every statistic related to winning a campaign including leading in the polls.

Now, this time around pollsters are suggesting that the opposite will occur -- that the GOP will unseat large numbers of Democratic incumbents in both the House and Senate -- and particularly in the former.

If that were the case, then it would be fair to expect early voting stats to greatly favor Republicans as they favored Democrats in 2008.

Democrats lead in 2010 early voting
The data, so far, though shows Democrats leading overall in those states that conduct early voting. They are not leading in every state, but they are more than holding their own when everything is taken together.

That really does not bode well for unseating large numbers of Democratic incumbents. However, it is not only early voting that shows a disconnect with the polling data.

For example, in 2008 Democrats were much more active online using "netroots" campaigning. Barack Obama's website was far more active than that of John McCain's, for instance.

Just a few months ago, the GOP and Tea Party did indeed appear to be taking over the internet based on a wide variety of statistics. However, lately that has turned around entirely. Formerly, for example, a YouTube video from OFA would be bombarded by negative comments from Tea Party sympathizers and rated lowly because of much larger numbers of "dislikes." But now the situation has flopped totally in favor of the Democrats.

Dems also hold fundraising lead
In fundraising also, the Democrats wiped the floor with the GOP in 2008 although this is traditionally an area of strength for Republicans. If we are to expect a Dem bloodbath this time around, then shouldn't GOP treasure chests be much larger in the same way that Democratic coffers were in 2008? In fact, the official Democratic campaign committees are raising more money than their GOP counterparts.

The Republicans are benefiting from independent groups receiving donations from large corporations, but its questionable whether that is giving them a significant edge. The conservative groups are at least partially offset by labor unions and progressive spending from groups like and Emily's List.

Sen. Barbara Boxer's supporters in California have been very active showing up not only for Boxer events, but also establishing a strong presence at the events of her opponent Carly Fiorina. (Source:

Weak points in surveys
The discrepancy between the surveys and all the rest of the evidence may lie in weaknesses that polls have in predicting minority and youth turnout. These two demographics traditionally favor Democrats, but both are underrepresented in surveys for a variety of reasons.

When we look at the large turnout at Obama rallies, including about 37,000 recently in Southern California, it is hard to imagine that there is a substantial "enthusiasm gap."   Indeed, the Democrats appear to be turning out in larger numbers overall for these types of public events than Republicans.

The election, of course, will show us whether the survey data was correct.  There is also a possibility that polls will show races tightening dramatically in the last few days before the election.

No comments:

Popular Posts