One of the interesting aspects of this week’s primaries was Colorado’s experiment with allowing unaffiliated voters to participate in the mail-in primary. The unusual aspect of it was that unaffiliated voters could request one party’s ballot in advance, or they would receive both major party ballots in the mail but could only send one back. This was potentially very confusing, so it’s worth examining to see how it affected the state’s primary. My focus here is on three main questions: 1) How many of the unaffiliated actually voted? 2) How did they vote? 3) How many messed up?
How many voted?
Unaffiliated voters, who make up some 36 percent of the state’s active voters, only voted at about half the rate of partisans. Below, I’ve charted the ballot counts tabulated by the secretary of state’s office over the past two weeks broken down by party affiliation. (Note: These figures are not final.) Each figure is reported as a percentage of the state’s registered voters for each party category.
Overall, about 1.2 million of the state’s 3.2 million active voters turned out, for a voter turnout rate of about 36 percent. (That’s up substantially from the 21 percent turnout in 2016 and 22 percent in 2014.) The rate was pretty close among partisans, with 46 percent of Democrats and 41 percent of Republicans sending in ballots. (Democrats had a late rally.)
Yet only 24 percent of unaffiliated voters sent in ballots. Another way of looking at this is that unaffiliated voters make up 36 percent of the state’s active voters but only submitted 24 percent of the ballots.
That’s obviously a lot more unaffiliated ballots than were submitted in the state’s previous closed primaries (zero), but the lower numbers are consistent with political science research showing that independents tend to have less interest in politics and participate less. It’s also possible some were confused by the two-ballot system and didn’t participate as a result. We may see greater participation in future elections as people grow used to this primary system.
It’s also interesting to note the rate of ballot returns. A campaign consultant once told me that voters in a mail-in election vote like consumers pay bills; some pay as soon as the bill shows up, many others wait until the due date, and there’s a low rate of bill-paying in the middle. If we look at the daily rate of ballot returns, we see something like that, with an initial spike, followed by a relatively quiet period, and then a rush of ballots on Election Day.
How did the unaffiliated vote?
Among the unaffiliated who received both major party ballots, 153,095 sent back the Democratic ballot and 93,041 sent back the Republican one. Relatedly, the unaffiliated could signify in advance which party ballot they wanted; Democratic ballots outnumbered Republican ones 2:1 among this group.
Does this mean that the unaffiliated lean Democratic or will vote Democratic in the fall? Well, maybe, but there are other possibilities. These voters might well have found the Democratic gubernatorial race more interesting than the Republican one and wanted to weigh in. They also might have found the Democratic congressional primary in their district more competitive than the Republican one. But it’s certainly plausible that the unaffiliated currently contains a large number of Democratic-leaning voters who choose not to identify with the party.
To get a sense of this, I looked at the share of unaffiliated voters who received both ballots and submitted a Democratic one by county, and then compared that with the county’s vote in 2016’s presidential election. As can be seen in the scatterplot below, county partisan lean is a very strong predictor of the unaffiliated ballot choice. (The relationship has an R-squared of 0.80.)
This suggests that ballot choice is a good indicator of partisan lean, and that this week’s results portend well for the Democrats in the fall. Still, we should be cautious in inferring too much about individual voters’ behavior from county-level results.
How many messed up?
Given the somewhat confusing new primary system, it was widely anticipated that there’d be some ballot spoilage. That is, some unaffiliated voters would send back both major party ballots, invalidating both of them.
The secretary of state’s office is still gathering the data on this. However, the most recent reports suggest that the ballot spoilage rate is surprisingly low — in the range of 3 to 4 percent. That is, of those who received two party ballots, more than 95 percent managed to send back just one ballot and have their vote counted.
This is an impressive rate, suggesting that efforts by the state and various county election directors to educate unaffiliated voters about the new process were effective. I hope to have some more to say on this effort in the near future.