Cross-pressured voting and the failure of Ohio’s marijuana oligopoly

While some have claimed that last Tuesday’s defeat for Ohio’s Issue 3, which would have legalized recreational marijuana in the state, is an example of liberals losing the culture wars, the actual dynamics of this proposed amendment’s failure are far more nuanced, and may even tell the opposite story. Instead of serving as a defeat for liberals in the culture wars, the failure of this referendum shows what happens when voters are faced with multiple, competing concerns as part of the same issue.

Before getting into the particulars of the amendment, let me offer two considerations that serve as a backdrop for Issue 3. First, recent polling has found that the majority of Ohioans support the legalization of marijuana. An October 8, 2015 Quinnipiac Poll found that 53% of Ohioans supported allowing people to possess small amount of marijuana for personal use, while 44% were opposed. Even among Republicans, 32% were in favor (65% were opposed). Among Democrats, 70% were in favor of legalization, while 27% were opposed legalization; the breakdown among Independents was 51% in favor of legalization and 45% opposed to legalization. Previous Quinnipiac Polls have shown similar results.

Second, the Ohio proposal differed from the 2012 Colorado marijuana referendum that simply legalized recreational marijuana for those over the age of 21 and allowed those who wanted to open stores selling marijuana products to do so. As in some other states, Ohioans would have been allowed to grow up to four marijuana plants for personal use and medical marijuana would have been legalized under the proposed amendment. However, commercial growth and sale of marijuana would have been limited under the amendment to ten pre-designated facilities. Among the owners of these facilities were members of the Taft family and singer Nick Lachey, along with members of the campaign that put the referendum on the ballot in the first place. Opponents of the amendment referred to this arrangement as a marijuana monopoly. Technically, oligopoly is a more accurate description, but the concern stands regardless.

Indeed, fearing that the marijuana issue might pass in Ohio, lawmakers placed a competing anti-monopoly amendment (marijuana or otherwise) on the ballot in Ohio at the same time. Had both issues passed, the competing constitutional considerations would have resulted in court battles. The fact that Issue 2 did pass— even in the face of claims from some liberals that Issue 2 was a “scummy trick” designed by Republicans to subvert the will of the people— suggests that Ohioans do not hold monopolies in high regard.

As a whole, the evidence suggests that 1) marijuana legalization is relatively popular among Ohioans (especially Democrats and independents) and 2) that monopolies are relatively unpopular among Ohioans. Thus many Ohioans who voted in this election, mostly Democrats and Independents (but also some Republicans) faced the cross-pressure of liking the idea of legal recreational marijuana, but not liking the idea of a legal monopoly/oligopoly for the production and sale of legal recreational marijuana. In the end, the second of the two considerations about monopolies won out, dooming Issue 3.

Election results at the county level bear out my explanation for why Issue 3 failed. Issue 3 failed statewide by about 28 percentage points (64.1% no to 35.9% yes), varying from a low of 17.69% yes in Putnam County in the west to a high of 42.36% yes in Jefferson County, which is on the Ohio-West Virginia border. (Interestingly, the second strongest performance was in Cincinnati’s Hamilton County. I wonder what our two oft-opposed founders would think about these results in their namesake counties?)

Initially, one might think that marijuana might do better in counties where President Obama did well, given that Democrats were so much more likely than Republicans to support legalization when polled. Such a relationship does exist, but it is not as strong as one would expect. The correlation between President Obama’s 2012 two-party percentage in a county and the percentage of voters who supported Issue 3 in 2015 is a surprisingly modest 0.59 (R-squared  for bivariate regression= 0.347). In fact, 6 of the 10 counties where Issue 3 performed strongest voted for Mitt Romney in 2012. As a point of comparison, the correlation between President Obama’s 2012 two-party percentage in a county and the percentage of voters who opposed the 2004 referendum on same-sex marriage (a clear culture war vote) is a stronger 0.724 (R-squared for bivariate regression= 0.525).

View post on imgur.com

The relationship we’d expect to see — more Democrats means more support for marijuana legalization — is still there, but it’s weaker than we’d expect to see if the “liberals losing the culture war” hypothesis is true. If Issue 3 failed due to a lost culture war battle, we would expect the relationship to be strong, but the level of support to be lower. In other words, we might see a consistent drop is support of, say, 10 percent across the state as compared with the two-party vote percentage received by President Obama. (Another possibility would be that turnout would fall more among liberals than conservatives, as was the case in Ohio in 2014 after the Ed FitzGerald campaign imploded.)

A consistent drop in support across counties was not what took place in Ohio, meaning that an alternate explanation is needed. If, as I proposed earlier, Issue 3 serves as an example of cross-pressure voting, one would expect to find the biggest drop-offs in support relative to President Obama’s 2012 performance in those counties that supported him most strongly in 2012. This is the case because, since Democrats are most prevalent in these counties, division among them would have the greatest effect in terms of the overall results for that county.

As expected, if one subtracts the vote percentages for Issue 3 in 2015 from President Obama’s 2012 percentage and then correlates that measure (Obama 2012 %- Issue 3 2015%) with President Obama’s percentage in 2012, one finds a more robust correlation of 0.866 (R-squared for a bivariate regression=0.75). In other words, the better a county performed for Obama in 2012, the larger amount of fall-off there is that takes place on Issue 3. The counties where Issue 3 ran behind President Obama by the largest margin are also those where President Obama ran the strongest in 2012. For example, Issue 3 ran more than 30 percentage points behind President Obama’s 2012 vote percentage in Cleveland’s Cuyahoga County and Athens County, home to Ohio University. In fact, all ten of the counties with the biggest fall-off gave President Obama 50% of the vote or more in 2012.

View post on imgur.com

Note the much stronger relationship here. The better President Obama did in a county in 2012, the larger the fall-off between his vote share and the percentage of vote received by Issue 3. Far from being crushed by a rising tide of conservative backlash, Issue 3 was defeated on Democratic and Republican turf alike — even some precincts containing or including college campuses, such as Oberlin College and Kenyon College, wound up voting against the referendum (other precincts at or near campuses provided narrower than expected margins of support).

What’s more, all seven of the counties in which Issue 3 ran ahead of President Obama’s 2012 two-party vote share gave Mitt Romney more than 62 percent of the vote in 2012. At first the idea that Issue 3 would actually run ahead of President Obama’s vote share anywhere seems odd, given the explanation of cross-pressure voting.

Marijuana, via Brett Levin / Flickr

Marijuana, via Brett Levin / Flickr

However, I suspect that even though Republicans were also cross-pressured by the monopoly consideration, Issue 3 in 2015 was still probably more likely to get a Republican vote than was President Obama in 2012. Let me explain: according to the Ohio exit poll from 2012, approximately 5 percent of Buckeye State Republican voters supported President Obama in 2012. Even if the cross-pressure consideration of the monopoly caused support for Issue 3 to be cut in half from the 32 percent who indicated support for marijuana legalization in that Quinnipiac Poll, there would still be about three times the percentage of Republicans voting for marijuana legalization than there were be who voted for President Obama in 2012.

As a result, in heavily Republican counties, Democratic fall-off was offset or (in some cases) even overcome by Republican support for Issue 3 because the number of Republican voters in these counties far exceeds the number of Democratic voters. In contrast, in the most heavily Democratic counties, cross-pressure considerations caused support to drop heavily compared with President Obama’s vote share in 2012. Due to the relative lack of Republican voters in the most Democratic counties, Democratic fall-off was not offset by the rise in Republican support relative to President Obama’s vote percentage in 2012 simply because there were not a whole lot of Republicans to offset this fall-off (and, while greater than support for President Obama in 2012, support for Issue 3 among Republicans still was likely fairly low, requiring quite an imbalance in favor of Republicans for fall-off to be completely offset).

In conclusion, the failure of Issue 3 in Ohio was not a defeat for liberals in the culture wars. Rather the failure of this issue serves as an example of what happens when cross-pressure concerns result in voters from the same party ending up on different sides of an issue, as different concerns are more salient for different voters. While such circumstances are rare in our times of partisan polarization, it seems that internal party division played an important role in the failure of this issue, meaning that it would in fact be more accurate to say a coalition of liberals and conservatives  defeated Issue 3 than it would be to say that liberals were defeated by conservatives alone. Finally, those who support marijuana legalization might take heart in the fact that the proposal did as well as it did in a number of conservative counties. If another proposal without the monopoly aspect is put on the ballot, voters in Democratic-leaning counties are likely to provide the high levels of support that may allow such a proposal to pass and become law if support in these conservative counties can be maintained.

Legalization advocates plan to do exactly that next year.


Jacob Smith is Political Science Ph.D. student at UNC-Chapel Hill. He has a B.A. in Political Science from Kenyon College, where he graduated Phi Beta Kappa and with highest honors in Political Science, and a M.A. in Political Science from UNC-Chapel Hill. His interests relate to Congress, congressional elections, and political parties. All opinions shared here are his own; posting on AMERICAblog does not imply agreement with any posts, current or past, by other authors.

Share This Post

© 2018 AMERICAblog Media, LLC. All rights reserved. · Entries RSS
CLOSE
CLOSE