As Magellan continues our daily updates of ballots returned statewide in Colorado, it is necessary for us to make the following clarification: The purpose of these updates is not in any way an attempt to influence the outcome of the vote on Amendment 66. Any inferences drawn from this data, unless specifically mentioned here on this blog, are beyond our control, but suffice to say that the daily updates are not designed on our part to pronounce either triumphantly or despondently on the outcome of any particular campaign or issue on this year’s ballot.  

Rather, as the picture of Colorado’s electorate begins to take shape with each day closer to November 5th, Magellan will be displaying on a daily basis our grasp and understanding of how to use data as a resource in forecasting the outcome of elections. Each day, that picture will become more and more complete, and so each day, the question will become more and more interesting: How does this year’s election compare to other similar elections, and what do any differences in the electorate tell us about what is likely to happen in the future?

In this case, we are interested in comparing this year’s election to November 2011, because that is the last time Colorado’s voters considered an issue on a statewide basis, without the presence of federal or state-level candidates on the ballot. In November 2011, just over 1 million Colorado voters cast a ballot. For comparison’s sake, nearly 2.6 million Colorado voters cast their ballot in the November 2012 Election, while over 1.8 million cast their ballot in the November 2010 Election. Of course the amalgamation of local campaigns and issue around the state, all with varying levels of voter interest, will play some role in determining the electorate, but there is no reason to expect turnout this year to approach November 2012 levels, or likely even November 2010 levels. November 2011, however, provides a solid point of comparison. That is why it has been chosen as our benchmark.

The question then becomes: How will this year’s election differ from November 2011? An obvious place to start is to look for new voters, meaning those people who have already cast their ballot this year, but did not vote in November 2011. For the purposes of taking a deeper look at the data, we’ll also include the November 2009 Election in this analysis, so that a “new” voter will be anyone who has voted this year, but did not do so in either November 2009 or November 2011. There are 44,145 Colorado voters who meet this criteria, as of this morning. They comprise 23% of the electorate thus far.  We can safely assume that these 44,145 voters are people who have never shown any interest in voting in an off-year election, and yet for whatever reason that changed this year. One obvious reason why someone would be included in this category is if they simply weren’t eligible to vote in either November 2009 or November 2011, because they registered to vote after Election Day 2011. Here is the breakdown of what % of the “new” voters are new because of their registration date:

So three out of every four “new” voters decided to vote this year despite choosing not to do so in either November 2009 or November 2011. These 33,018 voters are interesting as a subgroup, and we’ll take a deeper look at them in a blog post tomorrow. But there is more we can learn about the new voters as a whole. What do they look like demographically, in terms of age, gender, and party registration? The tables below provide the answer:
 

Using these data points, we can go back and take a look at November 2011, and see how the addition of these new voters might affect the electorate as a whole when all is said and done. One thing to note is that they are significantly younger than the November 2011 electorate, and so if current trends continue, this year’s turnout will be younger than November 2011’s. Currently, this change has yet to take form – you’ll notice that the % of the electorate age 65 and older is actually 10% higher right now than it was in November 2011. That being said, older voters are more likely to have already returned their ballots. 14.28% (81,803 of 572,690) of active voters 65 and older have already returned their ballot, compared to just 2% (16,667 of 833,966) of active voters ages 18-34 who have returned theirs. And so the age of the electorate moving forward is clearly something to keep an eye on.

Getting back to the “new” voters, it is also worth tracking their party registration. Nearly 37% of them are Unaffiliated, whereas in November 2011, the Unaffiliated portion of the final electorate was around 25%. That’s an almost 12% difference, and the corresponding drop is equally distributed among Republicans (41% to 35%) and Democrats (33% to 27%). Clearly, the “new” voters are less partisan than the rest of the electorate, and whether additional “new” Unaffiliated voters continue to join the electorate will be interesting to watch.

To make a final point on the importance of these “new” voters, let’s assume that final turnout will be around what it was in November 2011. If 1,100,000 voters cast ballots this year, and around 25% of them are new to the electorate, that would be a total of 275,000 “new” voters. That is certainly a large enough number to make a difference in the final results. Keep visiting our blog for our daily updates and posts to track whether that turns out to be a realistic total of “new” voters to expect, and if so, how they will affect the electorate as a whole.

By: Ryan Winger
Magellan Strategies