Elsevier

Social Science Research

Volume 78, February 2019, Pages 12-27
Social Science Research

How and why have attitudes about cannabis legalization changed so much?

https://doi.org/10.1016/j.ssresearch.2018.12.011Get rights and content

Abstract

Since the late 1990s public opinion about cannabis legalization has become drastically more liberal, and some states have begun to legalize cannabis for recreational use. Why have attitudes changed so much? Prior research has considered a few of the reasons for this change, but this is the first comprehensive and empirically-based study to consider the wide range of potential causes for how and why this happened. We use data from the General Social Survey, National Study of Drug Use and Health, and word searches from the New York Times. We find that attitudes largely liberalized via intracohort changes. Most Americans developed more liberal views, regardless of their race and ethnicity, gender, education, religious or political affiliation, or religious engagement. Changes in cannabis use have had minimal effects on attitudes, and legalization of cannabis has not prompted attitude change in neighboring states. As to root causes, evidence suggests that a decrease in religious affiliation, a decline in punitiveness, and a shift in media framing all contributed to changing attitudes.

Introduction

Americans’ attitudes about cannabis legalization have changed dramatically in the last twenty-five years, and the reasons are largely unknown. Between the late 1960s and early 1990s, pro-legalization attitudes fluctuated within a fifteen-point range from 12 to 27 percent approval. Since the early 1990s, the proportion of the public in favor of legalization has risen at approximately 1.5 percentage points a year (calculations shown below). Trends in attitudes have been reflected in the punitiveness of state laws. During the 1980s concerns about cannabis were part of the escalation in the “War on Drugs” (Harcourt and Ludwig, 2007; Tonry, 2004). When attitudes became more disapproving in the early 1980s, legal punishment for carrying cannabis became harsher, whereby possession of 100 cannabis plants now had the same punishment as 100 g of cocaine (Golub et al., 2007).

Cannabis was also part of the “Three Strikes” sentencing laws, requiring life sentences in many states for repeat drug offenders (Caulkins and Chandler, 2011; McCoy and Krone, 2002). As Americans’ views began to liberalize, so did state-level legislation (Weitzer, 2014). In 1996 California became the first state in the union to legalize medical cannabis. By 2017, 21 states had legalized cannabis for medical purposes, and eight states (plus DC) had legalized the recreational use of cannabis.1

Clearly cannabis-related attitudes and laws have become more liberal since the 1990s. But why did they change? While some studies have considered a few of the factors that have shaped public opinion (Maričić et al., 2013; Meares, 1997; Nielsen, 2010), there has not yet been an empirical study examining the wide range of possible explanations for why attitudes have liberalized. Understanding the factors shaping attitudes would provide insight into why what was once widely perceived as radical is now increasingly the norm.

This is the most thorough study to date examining why attitudes about cannabis have changed. We draw on data from the General Social Survey (GSS), the National Survey on Drug Use and Health (NSDUH), and a content analysis of the New York Times. We also include a new data source – individual and state-level identifiers from the GSS – to test the relationship between state-level cannabis-related legislation and changes in attitudes. We find that attitudes largely liberalized via intracohort changes. Most Americans developed more liberal views, regardless of their race and ethnicity, gender, education, religious or political affiliation, or religious engagement. Changes in cannabis use have had minimal effects on attitudes and legalization of cannabis has not prompted attitude change in neighboring states. As to root causes, evidence suggests that a decrease in religious affiliation, a decline in punitiveness, and a shift in media framing all played contributing roles.

As noted above, support for the legalization of cannabis has increased substantially since the 1980s. Fig. 1 uses repeated cross-sectional data from the General Social Survey and from Gallup polls to show that support increased through the mid-1970s, fell in the 1980s, and increased again after the early 1990s. By the 2010s, over 50% of Americans supported legalization.

While there are a wide range of potential reasons why Americans’ attitudes changed, to date most studies have focused on only one or two factors. Nielsen (2010) and Miech and Koester (2012), for example, were some of the first scholars to notice the liberalization trend.

Nielsen (2010) also found more evidence for period effects than cohort effects, suggesting that people of all ages were changing their minds at roughly the same rate. However, Nielsen's data extended only until 2006.

Some researchers have also found that reading newspapers and watching television were associated with increasingly liberal drug-related views (Nielsen and Bonn, 2008; Stringer and Maggard, 2016). A lot of attention has also been given to research within individual states, some of which focuses on how cannabis-related legislative changes have affected attitudes and use (Friese and Grube, 2013; Khatapoush and Hallfors, 2004; Miech et al., 2012; Schuermeyer et al., 2014).2 But these studies cannot address why laws and attitudes changed initially or what factors may have contributed to changing views across the nation because they focus largely on legislative changes within a single state.

We know of only two studies that have tried to unravel the multiplicity of factors that could explain more liberal views across the nation. The first is a report published by Brookings that examined Pew public opinion data from 2010 to 2013 (Galston and Dionne, 2013). The other is a book chapter that provides a largely narrative account of the possible reasons for attitude change (Musgrave and Wilcox, 2013). While both are insightful, neither one assesses the wide range of factors that could have shaped attitudes since the 1980s across the country.

Our study proceeds by first establishing the extent to which attitudes have changed through period effects versus cohort replacement. We then look at whether some groups of people (e.g., more religious, Republicans, women, whites, people living in closer proximity to places where marijuana was legalized in some way) have been more likely to change than others. Next, we consider a variety of causes. We consider distal causes, such as recent declines in religious affiliation, increases in the proportion of the population with a college-education, and changes in attitudes about punitiveness of the criminal justice system. We also consider proximate causes, such as increases in cannabis use, perhaps by aging Baby Boomers looking for more effective pain relief, changes in the perceived risk of using cannabis, as well as changes in the way cannabis is framed by the media.

One of the key factors for understanding the forces shaping attitudes is determining whether changes have been largely driven by cohort succession or intracohort changes. Cohort succession corresponds to the concept of “generations” whereby older cohorts are gradually replaced by newer ones that share similar social experiences and historical events (Alwin and Krosnick, 1991; Brooks and Bolzendahl, 2004). As one generation is replaced by the next, the population shift can result in attitudinal changes across the entire country. The alternative process is intracohort change whereby both older and younger people develop more liberal views during a similar time period.

Some prior research has examined cohort effects on changes in drug-related attitudes. Musto (1999) suggested that during the Reagan era of the 1980s anti-drug rhetoric may have influenced the cohort coming of age to be less supportive of drug legalization than preceding more liberal cohorts. In contrast to Musto's work, Nielsen (2010) was unable to find such an effect for this cohort. Rather, she only found consistent differences in support for legalization for people born before World War II compared to those born after and no significant differences for the many cohorts born after 1945. Similarly, Musgrave and Wilcox (2013) identified eight distinctive cohorts (from “Flappers” to “Clintons”), finding that cohort effects had a strong influence on legalization support only until the 1980s, after which the support levels of Baby Boomers and subsequent cohorts moved in tandem. This previous work suggests that earlier in the 20th century there may have been cohort succession effects, whereby younger cohorts replaced older ones leading to more liberal views. However, later in the century both younger and older people may have developed more liberal views about cannabis. These ideas lead to our first hypothesis:

Hypothesis 1

Changes in cannabis legalization support since the 1980s are largely the result of intra-cohort changes in attitudes as opposed to cohort succession.

If cohort effects are not driving change, which intra-cohort changes might be responsible? As we explain below there are good reasons why the attitudes of some groups of people may have changed. Nevertheless, we suspect that changes between men and women and differences across various racial, political and religious groups are not the key for understanding American's relatively recent liberalization in attitudes.

Over the past four decades, researchers have consistently found that men are more likely to support cannabis legalization than women (Meares, 1997; Saieva, 2008; Toch and Maguire, 2014). Compared to men women tend to have lower prevalence and frequency of cannabis use, and less personal use is associated with higher risk estimates and lower levels of legalization support (Maričić et al., 2013; Schepis et al., 2011; Trevino and Richard, 2002). Some research has also suggested that mothers may be more likely to frame legalization as a family rather than an individual issue (Cubbins and Klepinger, 2007; Musgrave and Wilcox, 2013; Rienzi et al., 1996).

As gender expectations and roles have changed, there is reason to think that women's attitudes on this issue might have begun to align more with men's views, explaining friendlier attitudes over the years. At the same time, there may be a small net effect if both sets of gender expectations have evolved in tandem.

Significant but less persistent gaps have also been found between political parties in their support for cannabis legalization. Since the early 1990s, independents have generally been most likely to support legalization, with Democrats a close second, and Republicans a distant third (Musgrave and Wilcox, 2013; Timberlake et al., 2001; Toch and Maguire, 2014). Although cannabis legalization is a policy position on which Democrats tend to disagree with Republicans (Geiger, 2016; Motel, 2015; Newport, 2011), until recently this issue has not been central to most party platforms (Schwartz, 2014).

Prior research and polls have also found modest racial differences in attitudes. African Americans tend to support legalization at slightly higher levels than other ethnic groups, with the lowest rates of support among Latinos (Chen and Killeya-Jones, 2006; Galston and Dionne, 2013; Geiger, 2016). African-Americans have been found to think about issues like cannabis legalization in group terms using what Dawson (1995) called a “racial utility heuristic” – that is, evaluating whether the policy is good or bad for their community as a whole (Musgrave and Wilcox, 2013). While we suspect that there will be a racial gap in attitudes, it is unlikely that this gap has widened or that change within one racial group is disproportionately responsible for overall changes.

People with different levels of education are also likely to differ in their views. In the 1970s, the college-educated were more likely to favor legalization than were others, but in the early 1990s the gap narrowed (Musto, 1999; Nielsen, 2010; Ours and Williams, 2007). Some research has suggested that the relationship between education and legalization support may be curvilinear, with more education associated with less support until college, after which education is associated with greater support (Saieva, 2008). While we suspect that more educated individuals will be more supportive, it's unlikely that the gap has widened since the 1980s.

Finally, levels of support for cannabis legalization differ by religious identity, with conservative Protestants least likely to support it, Jews being most supportive, and Catholics in the middle (Merrill et al., 2005; Rothwell and Hawdon, 2008). These differences can be partially explained by differences in the interpretation of doctrine concerning substance use and ideological differences among religious denominations (Hoffmann and Miller, 1997; Saieva, 2008). People who are more engaged in their religion are also likely to be more disapproving of cannabis use (Adamczyk and Palmer, 2008). We expect that these differences have remained largely the same since the 1980s.

Based on the above considerations, we expect attitudinal differences across groups which are reflected in our second hypothesis. We also hypothesize that over time the pace of change will be similar.

Hypothesis 2

Support for cannabis will differ for men and women, African Americans and whites, Democrats and Republicans, different religious groups, and for people with higher and lower levels of religious engagement.

Hypothesis 3

Changes in attitudes have occurred roughly to the same extent across genders, political parties, racial groups, religious groups, educational attainment and religious belief.

Pro-legalization sentiment could have increased in part because more liberal groups have become more prevalent in the population. Coinciding with the pro-legalization trend since the early 1990s was a threefold increase in the percentage of people claiming no religious affiliation. And as Rothwell and Hawdon (2008) found, a secular worldview is associated with greater tolerance of deviance, in particular cannabis use.

Given the positive relationship between education and pro-legalization attitudes, rising education levels could have contributed to changing attitudes as less educated cohorts died out and were replaced with more educated people. However, this effect is likely to account for only a few percentage points of change at most, since the percentage of the population that is college-educated has increased less than 8% since 1992 (our calculation).

Hypothesis 4

Changes in attitudes have resulted partly from increases in the percentage of the population that is religiously unaffiliated and/or an increase in the proportion that has graduated from college.

Studies have found that people who used cannabis regularly at some point are much more likely to support legalization and decriminalization than people who used only occasionally or less(Cerdá et al., 2012; Friese and Grube, 2013; Maričić et al., 2013). One reason may be that regular users perceive lower risks from cannabis (Andersson et al., 2009; Okaneku et al., 2015; Wall et al., 2011). If cannabis use has increased since the 1980s, this could explain attitude liberalization. These ideas lead to our next hypothesis.

Hypothesis 5

Changes in attitudes are partially the result of increases in cannabis use.

As noted above, social scientists have looked at whether legislative changes (e.g., a state making medical cannabis legal) affect attitudes or vice versa (Caulkins and Chandler, 2011; Marion and Hill, 2016; Page and Shapiro, 2010). The findings show that attitudes typically drive legislative changes. However, as states have become more liberal and legalized cannabis, residents in neighboring states may have become more tolerant. Likewise, access to cannabis from states that allow for medical cannabis could affect the attitudes of residents living nearby, helping them see cannabis as something that is not as problematic as expected. These ideas lead to the following hypothesis:

Hypothesis 6

Changes in attitudes have diffused across states, whereby states that have legalized medical cannabis and have more supportive residents have influenced the attitudes of people in neighboring states.

One commonly cited explanation for changing attitudes about cannabis is the increasing extent to which it is framed in medical terms (Musto, 1999; Pallone and Hennessy, 2003; Sznitman and Lewis, 2015). Media frames refer to how journalists simplify complexity by invoking the preexisting cognitive schemas of their audience (Borah, 2011; Entman, 1993; Scheufele and Tewksbury, 2006). Such intentional or unintentional media reframing of social issues influences how individual readers interpret information in news stories, which can then influence collective public opinion (Matthes, 2009; McCombs, 2018; Scheufele, 1999).

Media framing research has shown that even minor changes in how the news presents a social issue can shape public perceptions (Barry et al., 2013; Forsyth, 2012; Lancaster et al., 2011). Research has also found that the negative media framing of cannabis and other drugs peaked in the “War on Drugs” of the mid-1980s, then decreased both in terms of the number of articles as well as media framing of cannabis as a negative, criminal issue in the late 1980s (Gonzenbach, 1996; Johnson et al., 2004). This reduction in anti-cannabis media coverage was followed by an increase in the framing of cannabis as a medical issue starting in the mid-1990s (Schwartz, 2002; Stryker, 2003; Vickovic and Fradella, 2011).

Studies by Nielsen and Bonn (2008) and by Stringer and Maggard (2016) found that support for legalization was correlated with exposure to media that framed cannabis in a positive way.

Although there has been some increase in tolerant attitudes toward other illegal drugs, since the mid-1990s this change has not been as dramatic as the increase in tolerance for cannabis (Millhorn et al., 2009; Nielsen, 2010; Trevino and Richard, 2002). Based on these ideas we expect that since the 1990s cannabis would be more likely than other drugs to be framed as a medical issue. Additionally, increases over time in the medical framing of cannabis should correspond with increasingly positive attitudes about cannabis, thus suggesting that the medical framing may have, in part, led to more positive views. As the medical frame has become more popular, we would expect media discussions about cannabis to be decoupled from those related to other illegal drugs such as cocaine and heroin.

Hypothesis 7

As attitudes about cannabis changed, there was also a significant increase in the number of newspaper articles framing cannabis as a medical issue.

Hypothesis 8

Medical framing of other illegal substances has increased to a much lesser extent than the medical framing of cannabis.

Hypothesis 9

The media has become less likely to discuss illegal drugs such as cocaine and heroin in reports about cannabis.

Given the increasing extent to which cannabis is framed in medical terms, over time Americans should be less likely to see cannabis as harmful. A lot of research has found that legal substances, like cigarettes, may pose major health risks (Cummings and Proctor, 2014; Pacheco, 2011), possibly more than moderate cannabis use, which may lead some Americans to question the extent to which cannabis is problematic. Additionally, seeing cannabis as a drug for the medical treatment of pain is likely to affect views. If Americans are less likely to see cannabis as harmful, we would also expect them to be less likely to disapprove of it, which leads to our next hypothesis.

Hypothesis 10

As Americans increasingly see cannabis as less risky, they are also less likely to disapprove of it.

In the 1990s people who used or sold illicit drugs came under heightened scrutiny from law enforcement. Possession of small amounts of drugs could lead to a jail sentence (Harcourt and Ludwig, 2007). In 1994 several states began adopting a version of the “Three Strikes” law whereby people who were convicted of three felonies could be sent to prison for life (Lock et al., 2002). At this time California implemented a particularly harsh version whereby even misdemeanors could qualify someone for a lifetime prison sentence (Kovandzic et al., 2002). Over time this tough-on-crime approach led to the mass incarceration of people convicted of drug-related crimes, most notably cannabis (Tonry, 2001). Over the last twenty years the number of arrests for cannabis possession has been greater than for any other drug violation (Borden et al., 2016). By removing a high proportion of people (mostly men) who could provide important financial, social, and family support, a number of studies have found that cannabis-related mass incarceration has had a detrimental effect on families and local communities (Rose and Clear, 2003; Sampson and Loeffler, 2010; Wakefield and Wildeman, 2014).

Both the media and social science research have given a lot of attention to the problems with mass incarceration (Ramirez, 2013), especially as it relates to cannabis possession. Legislatures have begun to respond. Hence, in a 2012 referendum, Californians voted for a more lenient version of the “Three Strikes” law (Albonetti, 2016). It is possible that as more Americans began to view the criminal justice system as unreasonably harsh, they also tempered their views about cannabis, leading to greater tolerance (Tonry, 2004). These ideas lead to our final hypothesis:

Hypothesis 11

As more Americans have come to believe that the criminal justice system is too harsh, they have become more supportive of legalizing cannabis.

Section snippets

Data

To test our hypotheses, we draw on three datasets: the General Social Survey (GSS), the National Survey on Drug Use and Health (NSDUH), and articles from the New York Times. We discuss each of these datasets below. Descriptive statistics for GSS and NSDUH are in Table 1.

The General Social Survey (GSS) is a nationally representative cross-sectional survey that has been conducted 30 times between 1972 and 2016. We use a subset of the GSS in which respondents were asked: “Do you think the use of

Methods

We test our hypotheses in the order they were presented in the literature review. We test Hypothesis 1 using a decomposition technique demonstrated by Firebaugh (1997, p. 22; Table 4.1) on data from the General Social Survey. This technique estimates the extent to which change happens through people changing their minds and the extent to which it happens through the replacement of older cohorts with differently-minded young people.

To examine Hypotheses 2 and 3, we ran six logistic regressions

Results

We examine evidence for Hypothesis 1 in Table 2. We parse total change within consecutive four-year periods into average intra-cohort change and cohort replacement. Average intra-cohort change is a mean of change within four-year birth cohorts weighted by cohort size. The table shows that cohort replacement accounts for relatively little of the changes in attitudes about cannabis legalization between the early 1990s and 2016 and plays a very minor role during the most recent period, offering

Conclusion

Over the last three decades public opinion swung from three-to-one against to two-to-one in favor of legalizing cannabis. Since the 1990s people across different sociodemographic subgroups largely changed their attitudes to a similar extent. We found only two exceptions – since 2006, Hispanics have changed less than whites, and Democrats started liberalizing earlier than Republicans did. Differences by religion, gender and education remained the same. Our analysis shows that all birth cohorts

References (89)

  • Barbro Andersson et al.

    The prevalences of and perceived risks from drug use among teenagers in 33 european countries

    J. Subst. Use

    (2009)
  • Colleen L. Barry et al.

    Framing childhood obesity: how individualizing the problem affects public support for prevention

    Polit. Psychol.

    (2013)
  • Kristen Bialik et al.

    Key Trends in Social and Digital News Media

    (2017)
  • Porismita Borah

    Conceptual issues in framing theory: a systematic examination of a decade's literature

    J. Commun.

    (2011)
  • Tess Borden et al.

    Every 25 Seconds the Human Toll of Criminalizing Drug Use in the United States

    (2016)
  • Paul R. Brewer

    Values, political knowledge, and public opinion about gay rights: a framing-based account

    Publ. Opin. Q.

    (2003)
  • Jonathan Caulkins et al.

    Long-run trends in incarceration of drug offenders in the US

    QNRS Repository

    (2011)
  • Kevin W. Chen et al.

    Understanding differences in marijuana use among urban black and suburban white high school students from two US community samples

    J. Ethn. Subst. Abuse

    (2006)
  • Renita Coleman et al.

    The young and agenda-less? Exploring age-related differences in agenda setting on the youngest generation, Baby Boomers, and the civic generation

    Journal. Mass Commun. Q.

    (2007)
  • Lisa A. Cubbins et al.

    Childhood family, ethnicity, and drug use over the life course

    J. Marriage Fam.

    (2007)
  • K. Michael Cummings et al.

    The changing public image of smoking in the United States: 1964–2014

    Cancer Epidemiology and Prevention Biomarkers

    (2014)
  • Michael C. Dawson

    Behind the Mule: Race and Class in African-American Politics

    (1995)
  • Fedor A. Dokshin

    Whose backyard and what's at issue? Spatial and ideological dynamics of local opposition to fracking in New York state, 2010 to 2013

    Am. Sociol. Rev.

    (2016)
  • Robert M. Entman

    Framing: toward clarification of a fractured paradigm

    J. Commun.

    (1993)
  • Lauren Feldman et al.

    Polarizing news? Representations of threat and efficacy in leading US newspapers' coverage of climate change

    Publ. Understand. Sci.

    (2017)
  • Glenn Firebaugh

    Analyzing Repeated Surveys

    (1997)
  • Bettina Friese et al.

    Legalization of medical marijuana and marijuana use among youths

    Drugs Educ. Prev. Policy

    (2013)
  • William A. Galston et al.

    The New Politics of Marijuana Legalization: Why Opinion Is Changing

    (2013)
  • A. Geiger

    Support for Marijuana Legalization Continues to Rise

    (2016)
  • Andrew Golub et al.

    The race/ethnicity disparity in misdemeanor marijuana arrests in New York city

    Criminol. Public Policy

    (2007)
  • W.J. Gonzenbach

    The Media, the President, and Public Opinion: a Longitudinal Analysis of the Drug Issue, 1984-1991

    (1996)
  • Bernard E. Harcourt et al.

    “Reefer madness: broken windows policing and misdemeanor marijuana arrests in New York city, 1989–2000

    Criminol. Public Policy

    (2007)
  • John P. Hoffmann et al.

    Social and political attitudes among religious groups: convergence and divergence over time

    J. Sci. Stud. Relig.

    (1997)
  • Michael Hout

    Religious ambivalence, liminality, and the increase of No religious preference in the United States, 2006–2014

    J. Sci. Stud. Relig.

    (2017)
  • Michael Hout et al.

    Explaining why more Americans have No religious preference: political backlash and generational succession, 1987–2012

    Sociological Science

    (2014)
  • Thomas J. Johnson et al.

    Drug peddlers: how four presidents attempted to influence media and public concern on the drug issue

    Atl. J. Commun.

    (2004)
  • S. Khatapoush et al.

    ‘Sending the wrong message’: did medical marijuana legalization in California change attitudes about and use of marijuana?

    J. Drug Issues

    (2004)
  • T.V. Kovandzic et al.

    “Unintended consequences of politically popular sentencing policy: the homicide promoting effects of ‘three Strikes’ in US cities

    Criminol. Public Policy

    (2002)
  • Kari Lancaster et al.

    Illicit drugs and the media: models of media effects for use in drug policy research

    Drug Alcohol Rev.

    (2011)
  • E.D. Lock et al.

    Battle fatigue: is public support waning for ‘War’-Centered drug control strategies?

    Crime Delinq.

    (2002)
  • Jelena Maričić et al.

    Risk perception related to (il) licit substance use and attitudes towards its' use and legalization–the role of age, gender and substance use

    Društvena Istraživanja: Časopis Za Opća Društvena Pitanja

    (2013)
  • Nancy E. Marion et al.

    Legalizing Marijuana: a Shift in Policies across America

    (2016)
  • Jörg Matthes

    What's in a frame? A content analysis of media framing studies in the world's leading communication journals, 1990-2005

    Journal. Mass Commun. Q.

    (2009)
  • Maxwell McCombs

    Setting the Agenda: Mass Media and Public Opinion

    (2018)
  • Cited by (71)

    • Post-retirement enlightenment syndrome: Worthy of investigation

      2023, International Journal of Drug Policy
    View all citing articles on Scopus
    View full text