‘Cannabis Coattails’ and the challenges of polling in 2016

A thriving marijuana plant is seen at a grow operation in Denver, Colorado December 31, 2013. REUTERS/Rick Wilking/File Photo - RTSJ4EC

Much attention has been paid to the fact that several states are voting on marijuana initiatives this November. Five states—Arizona, California, Maine, Massachusetts, and Nevada—will vote on recreational legalization. Four more states—Arkansas, Florida, Montana, and North Dakota—will have medical marijuana on their ballots. In many places, polling is tight and there are no sure bets on passage (not even in the Golden State).

The presence of these initiatives on ballots is important in and of itself, and passage would mean tremendous drug policy changes in any state that votes to approve its initiative, but they could also have significant ripple effects on other elements of this year’s election. How might the ballot initiatives affect voter turnout, the outcomes of other races, or the accuracy of polling in these states?

Let’s break this down into a few important points.

Recreational vs. medical marijuana: Is there a difference?

From a policy perspective, differences between medical marijuana and recreational marijuana are significant. Medical markets are smaller, access to the product is more limited, and overall public support is higher for medical marijuana. That broader support also makes it harder to identify concrete and significant differences of opinion across demographic groups. For recreational marijuana, where support nationwide stands at around 60 percent divisions exist and they are significant. We know that Democrats support legalization at higher rates than do Republicans (and about the same as Independents). Men support legalization at higher rates than women. We also know that the younger you are, the more likely you are to support legalization. For instance, voters under the age of 30 support legalization at around 80 percent, while fewer than four-in-ten voters 65 and older support it.

This is not to say these differences are absolute, of course. You can find plenty of older Americans who support legalization and happily use the product themselves. There are many Republicans and conservatives, especially those identifying as libertarian, who favor marijuana reform. But, the trend lines are clear on recreational.

For medical, it makes sense that differences are muted because overall support is higher. What adds to that trend is the product’s broader consumer base. While there are plenty of young medical marijuana patients in states that have passed reform, it is not necessarily a “young person” issue. Patients who swear by the therapeutic benefits of marijuana span age groups and, in fact, many qualifying conditions disproportionately affect voters aged 30 and over—think MS, ALS, arthritis, cancer, etc. Similarly, you don’t have to be young to be convinced by a relative or friend that medical marijuana is helping them. For voters of all ages, seeing is believing.

Cannabis Coattails and the impact this year

Having recreational marijuana on the ballot matters for more than just these initiatives in and of themselves. Advocates put the initiatives on the ballot in even numbered years—especially presidential election years because turnout is significantly higher. However, there is some evidence from 2012 that marijuana initiatives have coattails, too. Those coattails have meaningful effects up and down the ballot.

To assess how marijuana ballot initiatives affected turnout, young voter behavior in 2008 and 2012 can provide a baseline. Given the honeymoon period for President Obama had ended by 2012, compared to the aura of his candidacy and the historic nature of his run in 2008, one would expect liberals and younger people to vote at lower or roughly similar levels in 2012. Nationally, the data suggest that to be the case. The share of voters under the age of 30 stayed relatively constant from 2008 to 2012—18 percent and 19 percent respectively. The share of the electorate made up of the youngest voters aged 18-24 was similar—10 percent in 2008 and 11 percent in 2012. The percent of the electorate made up of self-described liberals, ticked up slightly between 2008 and 2012 from 22 percent to 25 percent.

Nationally, those data show a fairly stable composition of the electorate. So what happened in states with legalization initiatives? As I have written previously the 2012 initiatives in Colorado and Washington had unique impacts on turnout in those states. Let’s first look at Colorado.

Voter turnout in Colorado changed dramatically between 2008 and 2012. In 2008, self-described liberals composed 17 percent of the electorate and that exploded to 28 percent in 2012. With marijuana legalization on the ballot in the Rocky Mountain State, more young people turned out, too. In 2008, 18-29 year olds made up 14 percent of the electorate, that increased to 20 percent in 2012. The changes were even more significant among 18-24 year olds. With legalization on the ballot, they made up 12 percent of the electorate compared to just 5 percent four years prior.

Washington saw many of the same changes with their legalization measure—Initiative 502—on the ballot. Self-described liberals increased slightly, from 27 to 31 percent of the electorate. But youth turnout changed dramatically. Between 2008 and 2012, while younger voters turned out at similar levels nationally, Washington was a different case. Voters aged 18 to 29 increased their share of the electorate from 10 to 22 percent between 2008 and 2012—more than doubling! For voters 18 to 24, their share grew from five to 13 percent—a 160 percent increase.

Marijuana legalization impacted who turned out in 2012 and we should believe it may have similar effects in 2016.

Marijuana’s bump for Democrats

 It is no secret that when liberals turn out and younger voters turn out, it helps Democrats, but is that necessarily true when legalization initiatives are on the ballot? Support for legalization spans the political spectrum and while there are trend lines by ideology, age, gender, and party, people from all corners of society find themselves getting behind marijuana reform.

From 2012, we have evidence about the relationship between support for marijuana reform and support for a presidential candidate. In Colorado, those supporting the legalization initiative also supported President Obama’s reelection. That year, 68 percent of legalization supporters also cast a ballot for Obama, compared to just 30 percent who voted for Romney. In Washington, Initiative 502 supporters voted for Obama 72 percent of the time, with just 36 percent backing Romney.

So, legalization initiatives have a clear Democratic benefit. Reform supporters and those behind ballot initiatives strategically time measures to happen during presidential years in order to capitalize on the increased turnout. Yet, we know that the initiatives themselves can dramatically transform turnout and can have significant effects on other races as well.

Marijuana legalization and the 2016 race

 With five recreational legalization initiatives on the ballot this year, what might it mean for the election generally? First, there is a clear Democratic benefit to these initiatives. In some states that may not matter, though. California and Massachusetts have ballot measures but they are not competitive in the presidential election and California’s Senate race is not competitive. In Maine, Arizona, and Nevada, however, there are competitive presidential contests. In the latter two, there are also competitive Senate races. Even though none of the candidates have embraced legalization, Hillary Clinton and Democrats across those ballots may see a bump because of changes in turnout. Democratic-leaning voters, who otherwise might have stayed home, could turn out to vote on marijuana reform. Some may leave other parts of the ballot blank, but Democrats could see a meaningful benefit overall. In a race that is close, a few thousand votes here or there could force an incumbent Republican Senator to pack up his office or shift a state’s electoral votes from red to blue.

Marijuana reform’s challenges for polling

In 2016, polling has been a tremendous topic of discussion. Candidates either consider them gospel or junk science, depending on how positive they are to a given campaign. Polling, of course, is a complicated business, and pollsters must make guesses as to what the electorate will look like on Election Day. How many Latinos will turn out? What will the age breakdown of the electorate be? What will the percentage of independents in the electorate be? How many women will go to the polls? These questions and many more must be answered, and answered effectively, in order for a poll to be accurate.

Presidential years are a bit easier for pollsters. We have excellent polling and tremendous amounts of historical data about what an electorate looks like in a presidential year at the national level and at the state level, as well. Unexpected changes to the composition of the electorate pose challenges for pollsters, however.

Marijuana legalization initiatives can generate one of those changes. The challenge for pollsters is that there is scant data about what such an initiative can do to the electorate of a state. The above data from Colorado and Washington tell us something about what may happen in Arizona, California, Maine, Massachusetts, and Nevada. But those are two data points. Because there have been only two recreational legalization initiatives in presidential election years, pollsters are left guessing, and many may not be factoring the effect of these initiatives into their calculus.

If turnout among those under the age of 30 and among self-described liberals explodes in those five states, it could transform the outcome of many other races. It may also mean that the actual effect of Cannabis Coattails could lead pollsters to underestimate both support for Democrats and the support for the initiatives themselves.

This really is not the fault of pollsters; it is a problem that stems from a lack of data. However, 2016 will provide additional data on the effect of marijuana initiatives on the composition of the electorate and the benefit for Democrats, so that the next time we face a similar situation—and 2020 will almost certainly have more legalization initiatives—pollsters will be better informed when designing poll samples and generating results.

None of this is to say that marijuana legalization initiatives will have a disruptive effect on the election. However, if Nevada’s Democratic Senate candidate, Catherine Cortez Masto, narrowly defeats Joe Heck, or if Arizona’s Ann Kirkpatrick outpaces John McCain, or if Hillary Clinton manages to hold the electoral vote awarded from Maine’s Second Congressional District, it may not be a hard-fought campaign that made the difference. Such wins might occur because cannabis has a coattail effect—and even candidates who oppose legalization may find that marijuana was the medicine their campaign needed.

Editor’s note: The source for all exit polling data in this post is CNN.