If Gallup Says Most Americans Want to Legalize Marijuana, Gallup Must Be Unreliable


Senate Judiciary Committee

At the International Drug Policy Reform Conference in Denver last week, many attendees were excited about the recent Gallup poll finding that 58 percent of Americans support legalizing marijuana. Writing at The Huffington Post, anti-pot activist Kevin Sabet tries to piss on this parade, but his aim is not so good. Sabet cites "at least three major problems with using Gallup as a reliable marker for marijuana attitudes in the U.S." Let's consider them one at a time:

1) "The poll asked about marijuana use, not sales and production." That's true, but surveys that ask about legalizing the marijuana business get similar results. In a Reason-Rupe survey last January, for example, 53 percent of respondents said "the government should treat marijuana the same as alcohol." In a 2010 A.P.-CNBC poll, 56 percent of respondents said regulations for marijuana should be either the same as or less strict than regulations for alcohol. A 2011 YouGov/Economist poll found a similar level of support (58 percent) for treating marijuana like alcohol. That is precisely how the successful legalization measures in Colorado and Washington, which attracted around 55 percent of the vote in both states, were described. Even in Texas, a recent Public Policy Polling survey found that 58 percent of respondents either "somewhat" or "strongly" supported "changing Texas law to regulate and tax marijuana similarly to alcohol." As I noted a few weeks ago, the appeal of the alcohol model is so strong that legalizing the commercial production and distribution of marijuana counterintuitively can attract more support than merely legalizing use. In the Reason-Rupe survey, only 47 percent of respondents favored "legalizing marijuana for recreational use," while 53 percent thought marijuana should be treated like alcohol.

2) "Gallup has always shown more support for legalization than other polls, and there's reason to think it may be an outlier." That first statement is not accurate. In 2010, for example, Gallup found 46 percent support for marijuana legalization; A Newsweek poll and an ABC News/Washington Post poll that year got similar results: 45 percent and 48 percent, respectively. In 2001 and 2003, Gallup found that 34 percent of Americans favored legalization; a 2002 CNN/Time poll got the same result. And as Sabet himself notes (via a quote from The Guardian's Harry Enten), Angus Reid during the last few years has found stronger support for legalization than Gallup has. So it clearly is not true that "Gallup has always shown more support for legalization than other polls." What about this year? At least three 2013 surveys put support for legalization above 50 percent: Gallup (58 percent), Reason-Rupe (53 percent), and Pew (52 percent). Enten mentions two others that did not find majority support: Fox News (46 percent) and the Public Religion Research Institute (45 percent). So according to most of these polls (three out of five), most Americans support legalization.

3) "The poll had a small sample size of only about 1000 respondents." The surveys by Fox News and the Public Religion Research Institute, which Sabet seems to consider more reliable, used samples of 900 and 1,000, respectively. While Gallup's results supposedly are compromised by its "small sample," the same apparently is not true of the surveys that yielded results more congenial to Sabet's cause.  

We obviously are not at the point where polls consistently find majority support for legalizing marijuana. But that seems to be the direction in which we are moving, given the generational divide on this issue. In the Gallup poll, 67 percent of 18-to-29-year-olds favored legalization, compared to 45 percent of respondents 65 and older—the only age group that was mostly opposed to legalization. 

Sabet concedes "there is no doubt that marijuana legalization enjoys more support than it did a few years ago." He attributes that change to the machinations of "billionaire marijuana smoker Peter Lewis," who "has spent millions convincing Americans that marijuana legalization will bring 'money for new schools!' and 'safer roads!' and 'no more drug cartels!' if passed." Sabet seems to think you can persuade people of anything as long as you spend enough money. But if that were true, the federal government, which has vastly greater resources than Peter Lewis, would not be losing this argument.

In any case, pro-legalization philanthropy during the last few years cannot possibly explain the long-term arc of public opinion on this issue. According to Gallup's numbers, "Public support for legalization more than doubled in the 1970s, growing to 28%. It then plateaued during the 1980s and 1990s before inching steadily higher since 2000, reaching 50% in 2011." This trend does not correspond to giving by pot-smoking billionaires, but it does correspond pretty well to rising familiarity with marijuana, which has fatally undermined the government's credibility on this subject.