See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/324514039

The Impact of Recreational Marijuana Legislation in Washington, DC on

Marijuana Use Cognitions

Article  in  Substance Use & Misuse · April 2018

DOI: 10.1080/10826084.2018.1461226

CITATIONS READS

8 212

3 authors, including:

Tonya Dodge Michelle L Stock

George Washington University George Washington University
34 PUBLICATIONS   1,764 CITATIONS    50 PUBLICATIONS   2,275 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

The Impact of African American Parents’ Racial Discrimination Experiences and Perceived Neighborhood Cohesion on their Racial Socialization Practices View project

All content following this page was uploaded by Michelle L Stock on 16 January 2019.

The user has requested enhancement of the downloaded file.

 Substance Use & Misuse

ISSN: 1082-6084 (Print) 1532-2491 (Online) Journal homepage: http://www.tandfonline.com/loi/isum20

The Impact of Recreational Marijuana Legislation

in Washington, DC on Marijuana Use Cognitions

Paige Clarke, Tonya Dodge & Michelle L. Stock

To cite this article: Paige Clarke, Tonya Dodge & Michelle L. Stock (2018) The Impact of
Recreational Marijuana Legislation in Washington, DC on Marijuana Use Cognitions, Substance
Use & Misuse, 53:13, 2165-2173, DOI: 10.1080/10826084.2018.1461226

To link to this article: https://doi.org/10.1080/10826084.2018.1461226

Published online: 13 Apr 2018.

Submit your article to this journal

Article views: 212

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=isum20
SUBSTANCE USE & MISUSE
, VOL. , NO. , –
https://doi.org/./..

ORIGINAL ARTICLE

The Impact of Recreational Marijuana Legislation in Washington, DC on Marijuana

Use Cognitions
Paige Clarkea , Tonya Dodgeb , and Michelle L. Stockb
a
Department of Applied Social Psychology, George Washington University, Washington, District of Columbia, USA; b Department of Psychology,
George Washington University, Washington, District of Columbia, USA

ABSTRACT KEYWORDS
Background: There is little published research that tests the effect of recreational marijuana legislation Marijuana legalization;
on risk-related cognitions and how individuals respond immediately after legislative approval. cognitions; willingness;
Objectives: The objective was to test whether learning about the passage of Initiative 71, a voter refer- prototypes; cannabis
endum that legalized recreational use of marijuana in the District of Columbia, would lead individuals
to adopt more favorable marijuana cognitions than they had before the Initiative was passed. Methods:
Undergraduate students (N = 402) completed two web-based questionnaires in 2014. The first ques-
tionnaire was completed prior to the referendum vote and the follow-up questionnaire was completed
after voters approved Initiative 71. Attitudes, perceived norms, intentions, prototypes, and willingness
were measured at time 1 and time 2. Study hypotheses were tested using repeated-measures analysis
of covariance. Results: Results showed that attitudes, intentions, perceived norms, and willingness to
use marijuana were more favorable after Initiative 71 was passed. However, the increase in attitudes
and willingness was moderated by past experience with marijuana whereby the increases were
statistically significant only among those with the least experience. The increase in perceived norms
was also moderated by past experience whereby increases were statistically significant among those
who were moderate or heavy users. The passage of Initiative 71 had no effect on favorable prototypes.
Conclusion/Importance: Legalization may have the unintended outcome of leading to more favorable
intentions to use marijuana and might lead abstainers or experimental users to become more
frequent users of marijuana via more positive attitudes and willingness towards marijuana use.

Marijuana is the most commonly used illicit drug in states in the United States (Alaska, California, Colorado,
the United States (Center of Behavioral Health Statis- Massachusetts, Maine, Nevada, Oregon, and Washing-
tics and Quality, 2015). Approximately 7.5% of people ton). The 2016 election led to the most recent additions to
aged 12 years or older have used marijuana in the past the list (California, Maine, and Nevada). Despite changes
month, with rates highest among those aged 18–25 years in legislation there have been few published studies
(National Institute on Drug Abuse, 2015). College is a documenting how these changes in legislation regarding
particularly common time for marijuana initiation and recreational marijuana use affect marijuana use attitudes
use (Suerken et al., 2014), where 21% of college students and tendencies of young adults.
report having used marijuana in the past 30 days and
more than one third report use within the past year
Medical marijuana legislation and marijuana
(Johnston, O’Malley, Bachman, Schulenberg, & Miech,
cognitions
2015). High rates of marijuana use by college students
is of concern because marijuana use is associated with a The majority of published literature to date has focused
number of problematic health consequences including on the impact of the legalization of medical marijuana
impaired cognitive functioning, respiratory complica- (Anderson, Hansen, & Rees, 2015; Choo et al., 2014;
tions, and mental health problems (e.g., Buckner, Ecker, Harper, Strumpf, & Kaufman, 2012; Hasin et al., 2015;
& Cohen, 2010; Shrivastava, Johnston, & Tsuang, 2011; Lynne-Landsman, Livingston, & Wagenaar, 2013; Pacula,
Taylor, Poulton, Moffitt, Ramankutty, & Sears, 2000). Powell, Heaton, & Sevigny, 2015). For example, Hasin
Several states have recently legalized recreational et al. (2015) analyzed Monitoring the Future (MTF) data
marijuana use. Recreational use of marijuana for adults from years 1991 to 2014. Results showed no significant
is legal in the District of Columbia, plus eight other differences in marijuana use among adolescents in states

CONTACT Paige Clarke [email protected] Department of Applied Social Psychology, George Washington University,  G St., NW, Washington, DC
, USA.
©  Taylor & Francis Group, LLC
2166 P. CLARKE ET AL.

before and after medical marijuana laws were passed, norms. Descriptive norms are an individual’s perception
leading some to conclude that medical marijuana laws of how common the behavior is (i.e., how many people
do not lead to increases in marijuana use among adoles- engage in the behavior).
cents (Choo et al., 2014). Therefore, one may speculate The social reaction pathway of the PWM involves
that because legislation legalizing marijuana use for med- more heuristic processing and willingness is the prox-
ical purposes has not lead to substantial increases in use, imal predictor of behavior. Willingness is predicted by
the effect of legalization of recreational marijuana use will attitudes, norms, and prototypes. Prototypes refer to the
have a negligible effect on marijuana use as well. images that individuals have of people who engage in the
However, some researchers posit that legalizing mar- behavior under consideration. The PWM helps predict
ijuana sends a message that using marijuana is not dan- both intended and unintended risk behaviors including
gerous, which would lead to increased acceptance of use smoking, drug (marijuana), and alcohol use (Andrews,
and thus increases in use (DuPont & Voth, 1995; Miech Hampson, & Barckley, 2008; Gibbons et al., 2015; Todd,
et al., 2015). Indeed, there is some evidence suggesting Kothe, Mullan, & Monds, 2016). Studies have demon-
that laws that legalize or decriminalize (i.e., lessening of strated that cognitions within the PWM are malleable
criminal penalties) marijuana use can lead to greater use and influenced by social and health information (e.g.,
via favorable marijuana cognitions (e.g., beliefs, attitudes, Gerrard et al., 2006; Gibbons et al., 2015; Stock, Gibbons,
perceived norms). For example, Miech et al. (2015) ana- Beekman, & Gerrard, 2015).
lyzed MTF data to test whether intentions to use and rates
of marijuana use differed before and after legalization of
This study
medical marijuana in California in 2010. Results demon-
strated that youth marijuana use and intentions to use Legislation is one aspect of the social environment likely
marijuana in the next five years increased. Furthermore, to impact cognitions within the PWM. While the PWM
perceived risk of harm for regular marijuana use and has been used to predict a wide range of health cognitions
disapproval of regular marijuana use decreased at a sig- and behaviors, no published studies to our knowledge
nificantly greater rate in California following legalization have applied the model to understand how marijuana
of medical marijuana when compared with other states legislation impacts marijuana cognitions. Furthermore,
in the United States. Thus, there is evidence showing that there have been few published studies to date that have
state legalization of medical marijuana use leads to more isolated the cognitions that are impacted by legalization
favorable marijuana cognitions of recreational marijuana use. This study addresses these
gaps in the literature by testing whether marijuana cog-
nitions of undergraduate students attending a university
Using the prototype-willingness model to link
in Washington, DC were affected when informed of the
legislation to risk cognitions
passage of Initiative 71, a voter-approved ballot initiative
Low disapproval of regular marijuana use, greater inten- that legalized the recreational use of marijuana.
tions to use, and lower risk perceptions are concerning Prior to the passage of Initiative 71, it was only legal
because such cognitions are often precursors to mari- to obtain marijuana with a recommendation from a doc-
juana use behavior (e.g., Ito, Henry, Cordova, & Bryan, tor. On November 4, 2014, the Initiative 71 ballot was
2015; Lopez-Quintero & Neumark, 2010). One theoretical approved by voters, and on February 26, 2015 Initiative 71
model that explains how cognitions impact behavior is the became law. Initiative 71 permits adults aged 21 years and
prototype-willingness model (PWM). The PWM posits older to possess up to two ounces of marijuana, grow up
that there are two paths to risk behavior: a reasoned path to six marijuana plants (only three may be mature plants)
and a social reaction path (Gerrard, Gibbons, Stock, Lune, within their primary residences, give up to one ounce of
& Cleveland, 2005; Gibbons, Gerrard, Stock, & Finneran, marijuana to another person of age if no money, goods,
2015). The reasoned path involves more analytic process- or services are exchanged, and to use marijuana on pri-
ing and encompasses constructs from the theory of rea- vate property. Thus, Initiative 71 permits possession and
soned action (TRA; Ajzen & Fishbein, 1980; Fishbein & use, but unlike other states that have legalized recreational
Ajzen, 1975) whereby intentions are the proximal predic- marijuana use (e.g., California, Washington), sale or pur-
tor of behavior. Consistent with the TRA, intentions are chase of marijuana is not legal.
predicted by attitudes and norms. In this study attitudes The objective of this study was to test whether learning
are based within the TRA and reflect how positively or about the passage of Initiative 71 would lead students to
negatively one feels about a behavior. Norms reflect per- adopt more favorable marijuana cognitions than they had
ceptions of social pressure. One specific type of norm that before the Initiative was passed. Furthermore, we antici-
is particularly relevant in the present context is descriptive pated the effect of learning about the passage of Initiative
 SUBSTANCE USE & MISUSE 2167

71 on marijuana use cognitions would be moderated by vote to approve the Marijuana Legalization Initiative
experience with marijuana. That is, we anticipated that 71. Initiative 71 was approved by Washington, DC
learning about the passage of Initiative 71 would lead to voters on November 4, 2014. The time 2 (T2) ques-
more positive cognitions but that this might be greater for tionnaire was completed approximately 30 days after
abstainers and experimental users (i.e., those who have completing T1 from November 16, 2014 to December 4,
used marijuana once a year or less) than for more experi- 2014. Immediately prior to completing the T2 question-
enced users. This is because abstainers and experimental naire participants were reminded that Initiative 71 was
users have little or no direct experience with marijuana approved by voters and they were provided with a sum-
and, as a result, may be more likely to draw inferences mary of Initiative 71. The average amount of time that
about marijuana indirectly from the outcome of the leg- elapsed between T1 and T2 was approximately 4 weeks
islation vote. For undergraduates with limited experience (M = 29.96 days, SD = 3.17).
with marijuana, legalization sends an indirect message
that marijuana use is relatively acceptable and safe.
Measures
Experience with marijuana as a moderator of the effect
of learning about the passage of Initiative 71 on mari- Attitudes
juana cognitions is consistent with existing theorizing Attitudes towards marijuana use was measured with an
and research on the PWM as past experiences with a adapted scale from (Ito et al., 2015) and following Ajzen
behavior have been shown to moderate the relationship (1991). Participants completed a 6-item semantic differ-
between willingness and behavior. When an individ- ential scale with the stem, “When you think about your
ual is inexperienced with a behavior, the relationship using marijuana, what comes to mind?” Responses ranged
between willingness and behavior is stronger than when on a 7-point scale anchored with bad/good, unsafe/safe,
an individual is experienced with a behavior (Pomery, unhealthy/healthy, wrong/right, unacceptable/acceptable,
Reis-Bergan, & Gerrard, 2009). This study extends the and disadvantageous/advantageous. A mean was com-
literature by testing whether experience with marijuana puted, with higher scores reflecting more positive atti-
moderates the effect of the social environment (i.e., tudes (T1α = 0.94; T2α = 0.95).
legalization of recreational marijuana use) on attitudes,
perceived norms, prototypes, intentions, and willingness. Norms
We hypothesize that marijuana cognitions (attitudes, Two items assessed descriptive norms of occasional mari-
perceived norms, intentions, prototypes, willingness) juana use. Participants reported the percentage of college
will be more positive after learning about the passage of students at their university and of their close friends that
Initiative 71 than before the initiative was passed and this use marijuana occasionally (i.e., once a month or less).
effect will be greater for abstainers/experimental users A mean of the two items was calculated (T2r = 0.35,
than for those who use marijuana more regularly. p < 0.01; T2r = 0.36, p < 0.01).

Intentions to use
Methods
Intentions were measured by asking participants their
plans to use marijuana at least once and to use marijuana
Participants and procedures
regularly (i.e., several times a month or more) in the next
Participants (N = 402; male = 121, female = 279, not four weeks. Responses ranged from 1 (definitely no) to
reported = 2) were recruited from introductory-level psy- 7 (definitely yes). A mean of the items was calculated
chology courses and chose to participate in this study. (T1α = 0.83; T2α = 0.82).
Students enrolled in the introductory-level psychology
courses were required to fulfill a research experience- Prototype
learning requirement. This requirement could be met a Images of occasional marijuana users were assessed by
variety of ways such as by participating in a number of asking participants to “ … think about the type of person
different research studies or by writing a brief summary your age and gender who occasionally (once a month or
of a research article. Participation in this study provided less) uses marijuana for recreational/social purposes …
partial fulfillment of this requirement. The study was We are interested in what traits you think this type of
approved by the University’s Institutional Review Board. person is likely to have.” Participants rated this image
Two questionnaires were completed on-line. The time on eleven positive and negative adjectives (e.g., smart,
1 (T1) questionnaire was completed between October unattractive; scored 1 = not at all to 7 = extremely;
15, 2014 and November 4, 2014, prior to the referendum coded such that higher scores reflect greater favorability).
2168 P. CLARKE ET AL.

Participants also rated their similarity to this prototype 5-point scale ranging from “0 occasions” to “10 or more
(scored 1 = not at all to 7 = very similar). The descriptor occasions.”
items were averaged and then multiplied by similar-
ity (Litt & Stock, 2011; Stock, Litt, Arlt, Peterson, & Background characteristics
Sommerville, 2013; T1α = 0.86; T2α = 0.82). Participants reported their gender as male, female, or
transgender. No participants selected transgender so the
Willingness to use responses were coded 0 = male, 1 = female. In addition,
Willingness was assessed by adapting measures used in participants reported age, racial group, and whether they
previous research (e.g., Gibbons et al., 2010; Stock, Gib- had ever used marijuana to treat a medical condition. At
bons, Peterson, & Gerrard, 2013). Participants reported T2, immediately after being reminded of the outcome of
how willing they would be to do three marijuana-related Initiative 71, participants reported whether or not they
behaviors (e.g., use enough to get high) in a hypothetical agreed with the outcome 0 (no) or 1 (yes).
scenario in which they were at a party and offered mari-
juana. Two additional items asked how willing one would
be to use marijuana in the next four weeks and to use Data analytic plan
marijuana with friends this week. Responses ranged from
1 (not at all willing) to 7 (very willing). All items loaded Study hypotheses were tested using repeated-measures
onto a single factor at each time wave, and were averaged analysis of covariance (ANCOVA) in SPSS version 24.
(T1α = 0.95; T2α = 0.95). This allowed us to test for differences on the dependent
variables as a function of time (T1 vs. T2) and type of mar-
Marijuana use ijuana user, while controlling for gender. Approval with
Marijuana use was measured by asking participants how voters’ decision to pass Initiative 71 was also included
often they used marijuana in the past year. Participants as a covariate for the attitudes analyses because a change
responded on a 9-point scale ranging from “never” to in attitudes towards marijuana could be driven, in part,
“every day.” Participants were categorized into four user by one’s attitude towards the law. Statistical significance
levels based on their responses. The categories were: was established when p < 0.05. When a statistically
abstainer/experimental user (if scored 0 (using marijuana significant Time X User interaction emerged, pairwise
never) or 1 (once a year or less)), light user (if scored 2 differences among the adjusted means for marijuana
(using marijuana several times a year)), moderate user user levels were evaluated. The Bonferroni method was
(if scored 3 (using marijuana once a month) or 4 (two to used for correcting for multiple comparisons. Race had
three times a month)), and heavy user (if scored 5–8 (using no effect, so it is not discussed further. Marijuana use
marijuana weekly or more)). These categories reflected during the past 30 days for T1 and T2 are reported but
natural cut points in the data (e.g., 15.7% reported using are not tested because for the majority of participants, the
several times a year, 5.2% and 8.2% reported using once a responses to the marijuana use variable at T2 included
month or two to three times per month, respectively) as their use both before and after Initiative 71. As a result,
well as psychologically meaningful groups. For example, the T2 marijuana use variable provides an estimate of
participants who had never used marijuana or used once use that may be affected by the ballot Initiative, but the
a year or less reflect a group of individuals that are likely effect would differ across participants because of the time
experimenting or may at this age experiment if they they completed the questionnaire. Thus, the extent to
have not yet done so, while those who are using a few which marijuana use in the past 30 days may have been
times a year reflect light users. Additionally t-tests were impacted by legislation is not clear from analyses.
conducted to test whether the cognitions (i.e., either at
T1 or T2) of abstainers (i.e., never tried marijuana) and
Results
experimental users (i.e., used once a year or less or used
several times a year) were statistically significantly dif-
Descriptive statistics
ferent from one another. Results showed that abstainers
and experimental users were statistically equivalent on all Participant characteristics
cognitions, except for prototypes at time 1 (t(73) = 2.06, Of the students who participated at T1, 95% (N =
p = 0.04). Thus, we felt confident categorizing abstainers 382; male = 117, female = 264, not reported = 1)
with experimental users. Participants also reported on also completed T2. Participants self-identified as White
how many occasions (if any) they have used marijuana (66.9%), Asian (14.4%), African American (6.5%), His-
during the past 30 days. Participants responded on a panic (3.7%), Other (2.0%), and Biracial (5.7%) (not
 SUBSTANCE USE & MISUSE 2169

Table . Inter-item correlations of study variables.

Att Att Nrm Nrm Int Int Pro Pro Wil Wil Gen Age

Att —
Att .∗∗ —
Nrm .∗∗ .∗∗ —
Nrm .∗∗ .∗∗ .∗∗ —
Int .∗∗ .∗∗ .∗∗ .∗∗ —
Int .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ —
Pro .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ —
Pro .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ —
Wil .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ —
Wil .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ .∗∗ —
Gen − .∗ − . . − . − . − . − . − . − . − . —
Age − . − . . − . − . − . − .∗ − . − .∗ − . . —

Note.  = time ;  = time ; Att = attitudes; Nrm = norms; Int = intentions; Pro = prototypes; Wil = willingness; Gen = gender; ∗ p < .; ∗∗ p < ..

Table . Marijuana use in past  days for time  and time . and T2 who reported marijuana use across different num-
Marijuana use in past  days Time  Time  ber of occasions within the past 30-days.

 occasions  (.%)  (.%)

– occasions  (.%)  (.%)
– occasions  (%)  (%)
Time and Time X User ANCOVA results
– occasions  (.%)  (%)
 or more occasions  (.%)  (.%) Results of ANCOVAs are presented below. Pairwise dif-
Missing  (.%)  (.%) ferences among means adjusted for covariates across mar-
ijuana use levels are shown in Table 3.

reported = 0.7%). Participants were between 18–35 years Attitudes

(M = 19.14, SD = 1.70). The repeated-measures ANCOVA for attitudes towards
marijuana revealed no statistically significant main effect
Marijuana use and attitudes towards Initiative  for time (F(1, 369) = 0.44, p = 0.51) or for gender (F(1,
Approximately 58% of the sample reported having ever 369) = 0.55, p = 0.46). There was, however, a significant
tried marijuana before, and the average age of first main effect for user (F(1, 369) = 230.79, p < 0.001), and
time trying marijuana was 16.33 years (SD = 1.75). this effect was qualified by the anticipated Time X User
The majority of the sample (59%) were classified as interaction (F(1, 369) = 9.92, p < 0.01). Tests of pairwise
abstainer/experimental users, 16% as light users, 13% as differences revealed that among abstainer/experimental
moderate users, and 10% as heavy users. Almost three- users, attitudes towards marijuana at T1 (M = 2.82, SE =
quarters of the sample (73%) reported that they agreed 0.07) became more favorable at T2 (M = 3.06, SE = 0.07;
with the outcome of the Initiative 71 vote. Not surpris- F(1, 369) = 14.88, p < 0.001). There were no significant
ingly more frequent users were more likely to agree with differences in attitudes towards marijuana at T1 and T2
the outcome of the Initiative 71 vote compared to less fre- for all other levels of marijuana use.
quent users (χ 2 (3) = 52.50, p < 0.001). Only one partic-
ipant reported having used marijuana to treat a medical Norms
condition. The repeated-measures ANCOVA for descriptive norms
Inter-item correlations of study variables are shown in of an occasional user revealed no main effect for time (F(1,
Table 1. Overall scales correlate in ways that one would 368) = 0.42, p = 0.52), or for gender (F(1, 368) = 1.12,
expect. Table 2 shows the percentage of participants at T1 p = 0.29). There was a statistically significant main effect
Table . Pairwise differences among adjusted means for marijuana user levels.
Attitudes Des. Norms Intentions Prototype Willingness
Time  Time  Time  Time  Time  Time  Time  Time  Time  Time 
User M (SE) M (SE) M (SE) M (SE) M (SE) M (SE) M (SE) M (SE) M (SE) M (SE)

Abs/Exp .∗∗ (.) .∗∗ (.) . (.) . (.) .∗∗ (.) .∗∗ (.) . (.) . (.) .∗∗ (.) .∗∗ (.)
Light . (.) . (.) . (.) . (.) .∗∗ (.) .∗∗ (.) . (.) . (.) . (.) . (.)
Moderate . (.) . (.) .∗ (.) .∗ (.) . (.) . (.) . (.) . (.) . (.) . (.)
Heavy . (.) . (.) .∗ (.) .∗ (.) . (.) . (.) . (.) . (.) . (.) . (.)

Note. Abst/Exp = abstainer/experimental user; Des. Norms = descriptive norms; ∗ = significant differences.
(p < .) between T and T; ∗∗ = significant differences (p < .) between T and T.
2170 P. CLARKE ET AL.

for user (F(1, 368) = 75.55, p < 0.001), which was qual- to marijuana use. Findings from this study showed that
ified by a statistically significant Time X User interaction attitudes, intentions, perceived norms, and willingness
(F(1, 368) = 7.34, p < 0.01). Tests of pairwise differences toward marijuana use were more favorable after learn-
revealed that for heavy users, perceived norms increased ing that Initiative 71 was passed than they were before
from T1 (M = 45.45, SE = 2.33) to T2 (M = 50.65, the Initiative had passed. Consistent with our predictions
SE = 2.27; F(1, 368) = 6.78, p < 0.05). For moderate for attitudes and willingness, the increase was statistically
users, perceived norms increased from T1 (M = 39.32, significant only among those with the least experience,
SE = 1.53) to T2 (M = 42.35, SE = 1.49; F(1, 368) = 5.37, p i.e., abstainer/experimental users. One might hypothe-
< 0.05). There were no significant differences in perceived size, based on the PWM model, that legalization of recre-
norms at T1 and T2 for all other levels of marijuana use. ational marijuana use may have the unintended outcome
of leading abstainers or experimental users to develop
Intentions more positive attitudes and willingness towards occa-
The repeated-measures ANCOVA for intentions to use sional use of marijuana. Given the extensive body of
marijuana showed a statistically significant main effect research documenting that positive attitudes and willing-
for time, such that intentions to use marijuana increased ness increase the likelihood of behavioral enactment (e.g.,
from T1 (M = 2.33, SD = 1.84) to T2 (M = 2.50, Todd et al., 2016; Kraus, 1995), this finding is particu-
SD = 1.93; F(1, 370) = 4.62, p = 0.03). There was no main larly important to keep in mind. Findings also imply that
effect for gender (F(1, 370) = 0.28, p = 0.60), but there legalization may have the additional unintended outcome
was a main effect for user whereby intentions increased of leading to more favorable intentions regardless of past
as experience with marijuana increased (F(1, 370) = experiences with marijuana. These more favorable inten-
1010.01, p < 0.001). The Time X User interaction was not tions may lead to increases in marijuana use as past work
statistically significant (F(1, 370) = 0.11, p = 0.74). has shown that intentions to use predict more marijuana
use over the course of one year (Ito et al., 2015).
Prototypes However, the increase in perceived norms was statis-
The repeated-measures ANCOVA for prototypes of an tically significant only among those with more marijuana
occasional marijuana user revealed no statistically signif- use experience (i.e., heavy and moderate users). Although
icant main effect for time (F(1, 370) = 0.59, p = 0.44), this is not what we hypothesized at the outset, in hind-
or for gender (F(1, 370) = 0.46, p = 0.50). There was a sight this is not particularly surprising because of research
significant main effect for user (F(1, 370) = 195.25, p < on motivated reasoning (e.g., Blanton & Gerrard, 1997;
0.001). There was no statistically significant Time X User Klein & Kunda, 1992; Kunda, 1990; Bastardi, Uhlmann,
interaction (F(1, 370) = 1.33, p = 0.25). & Lee, 2011). According to research on motivated reason-
ing, individuals are motivated to see themselves as ratio-
Willingness nal, well-reasoned decision makers. This leads individuals
The repeated-measures ANCOVA for willingness indi- to have the overarching goal of seeing decisions they make
cated a statistically significant main effect for time, such as reasonable or justified (Kunda, 1990). This motivation
that willingness increased from T1 (M = 3.02, SD = 2.03) for justification can influence the way that people evaluate
to T2 (M = 3.11, SD = 2.03; F(1, 369) = 7.34, p = 0.01). relevant information. For example, this perspective has
There was a significant main effect for user (F(1, 369) = shown that individuals will use non-diagnostic informa-
576.39, p < 0.001), but this effect was qualified by the tion to help justify potentially risky decisions (Blanton &
anticipated statistically significant Time X User interac- Gerrard, 1997). In the present context, moderate or heavy
tion (F(1, 369) = 7.09, p < 0.01). Similar to the results for users would be more motivated to identify evidence that
attitudes, tests of pairwise differences revealed that for would support their decision to use marijuana than those
abstainer/experimental users, willingness increased from who are experimental or light users. One way to justify the
T1 (M = 1.89, SE = 0.08) to T2 (M = 2.07, SE = 0.09; decision to use is to evaluate information in a way that is
F(1, 369) = 8.85, p < 0.01). There were no significant consistent with this decision. Therefore, when moderate
differences in willingness at T1 and T2 for all other levels or heavy users learn that Initiative 71 has passed, it pro-
of marijuana use. vides an opportunity to re-visit or re-evaluate perceived
norms in a way that supports their decision to use mari-
juana. As a result, we would expect to see that perceived
Discussion norms would increase more from time 1 to time 2 among
This study tested the effect of learning about the legaliza- moderate and heavy users than among experimental and
tion of recreational marijuana use on cognitions related light users.
 SUBSTANCE USE & MISUSE 2171

In contrast to findings for attitudes, intentions, per- the future (e.g., Cerdá et al., 2017; Miech et al., 2015; Pala-
ceived norms, and willingness, and our hypotheses, learn- mar, Ompad, & Petkova, 2014; Schuermeyer et al., 2014),
ing about the passage of Initiative 71 had no effect on there is noticeably less research reporting how legalization
prototypes. Not surprisingly, however, as past experience affects prototypes. Results of this study imply that legal-
with marijuana increased, so too did prototypes. Stated ization has little immediate effects on prototypes.
another way, the more experienced students in our sam- Results of this study lend additional support to the
ple were with marijuana the more positive were their pro- PWM and contribute to research on the PWM by extend-
totypes of the occasional user of marijuana. ing the role that past experience plays in the model.
Results from this study suggest that legislation legal- While previous research has documented that past experi-
izing recreational marijuana use may have an immediate ences moderate the willingness-behavior relationship, in
impact on personal cognitions (i.e., attitudes, intentions, this study we found that changes in attitudes, perceived
and willingness), but that it does not have an immedi- norms, intentions, and willingness were moderated by
ate impact on more socially constructed beliefs (e.g., pro- past experience with marijuana use. As individuals are
totypes). There are several possibilities that can explain exposed to new information over time, our findings imply
these different outcomes. One is time. That is, it may sim- that changes in attitudes, perceived norms, intentions, and
ply take time for legalization to impact prototypes. This willingness as a result of such information may be more or
means we might not see immediate effects on prototypes, less pronounced based on one’s past experiences with the
but we might see a shift in how favorable individuals feel behavior.
towards use. Another possibility is simply that legaliza- Results of this study provide an important first step in
tion has no impact on prototypes. Future research should showing that marijuana legislation can impact marijuana
explore these two possibilities. cognitions. A next step would be to study whether dif-
This study makes several unique contributions to ferences in legislation have similar effects on marijuana
the literature on outcomes of marijuana legislation. The cognitions. For example, although Initiative 71 permits
majority of the published research has established that adults aged 21 and older to possess and grow marijuana,
legalization of medical and recreational marijuana can buying and selling marijuana in Washington, DC is still
impact cognitions (e.g., perceptions of harm) several illegal. Thus, changes in cognitions after legalization
years after legislation has been enacted (Cerdá et al., of recreational marijuana could be even greater after
2017; Khatapoush & Hallfors, 2004; Miech et al., 2015; legalization for states (e.g., Washington, Colorado) in
Okaneku, Vearrier, McKeever, LaSala, & Greenberg, 2015; which recreational marijuana is more accessible and
Schuermeyer et al., 2014). For example, Cerdá et al. (2017) perhaps acceptable. Future research should test whether
compared MTF data across three years prior to legisla- other states with more lenient recreational marijuana
tion to three years following the legalization of recre- legislation have similar or different cognitions regarding
ational marijuana in Washington and Colorado and found use when compared to Washington, DC. Another avenue
that in Washington, perceived harm of regular marijuana for future research would be to assess how advertisements
use decreased and use increased as a result of legislation. and media coverage relating to marijuana legalization can
While the results show how exposure to legislation can, impact marijuana use cognitions. It might be useful to
over time, trickle down to change perceptions of harm and study whether different sources of communication about
use, our data show that legalization of recreational use can marijuana legalization (e.g., government official, social
have immediate effects on marijuana use cognitions. The media, professor) have different or similar impacts on
present findings, when viewed in conjunction with exist- marijuana cognitions. While we feel results of this study
ing literature, paint a picture whereby legalization can lead provide novel insights regarding the impact of legislation
to more favorable attitudes towards use, greater likelihood regarding use of recreational marijuana on marijuana
of using and a reduction in perceived harm associated use cognitions, there are several limitations that must
with occasional marijuana use both immediately, as doc- be acknowledged. One is that the sample includes only
umented in this study, and several years post-legislative undergraduate students willing to participate in research,
changes, as documented in previous research (Cerdá et al., which restricts the generalizability of study findings.
2017, Khatapoush & Hallfors, 2004; Miech et al., 2015; Future research should replicate the study using a more
Okaneku et al., 2015; Schuermeyer et al., 2014). racially, educationally, and socio-economically diverse
A second contribution of this study is finding that pro- sample. Furthermore, future research should study how
totypes remained relatively unaffected by legalization, at cognitions change as a result of marijuana legislation
least in the short term. While there is evidence focusing on among different age groups to see whether there are
how legalization affects attitudes, personal beliefs about particular age groups that are more or less impacted by
harm/danger of marijuana use, and intentions to use in legislative change. Another limitation is that some of the
2172 P. CLARKE ET AL.

undergraduates might have been seniors and planning References

to leave DC rendering Initiative 71 irrelevant to them.
Ajzen, I. (1991). The theory of planned behavior. Organiza-
However, this likely characterizes a very small percentage tional Behavior and Human Decision Processes, 50, 179–211.
of the sample as only about 12% of the sample were aged doi:10.1016/0749-5978(91)90020-T
21 years and older (thus were likely seniors). Therefore, Ajzen, I., & Fishbein, M. (1980). Understanding attitudes and
Initiative 71 would be relevant for the majority of the predicting social behavior. Englewood Cliffs, NJ: Prentice
sample. Hall.
Anderson, D. M., Hansen, B., & Rees, D. I. (2015). Medical mar-
Another limitation is the relatively short time frame
ijuana laws and teen marijuana use. American Law and Eco-
that elapsed between the passage of Initiative 71 and nomics Review, 17, 495–528.
the T2 questionnaire. This prevents us from knowing Andrews, J., Hampson, S., & Barckley, M. (2008). The effect of
whether legislation leads to longer lasting or more per- subjective normative social images of smokers on children’s
manent shifts in attitudes, intentions, or willingness. intentions to smoke. Nicotine & Tobacco Research, 10, 589–
However, this is an important first step in establishing 597. doi:10.1080/14622200801975819
Bastardi, A., Uhlmann, E. L., & Ross, L. (2011). Wishful think-
that legislation can have immediate effects on marijuana ing: Belief, desire, and the motivated evaluation of scientific
use cognitions. Another limitation is that we reminded evidence. Psychological Science, 22, 731–732.
participants of the outcome of the Initiative 71 vote just Blanton, H., & Gerrard, M. (1997). Effect of sexual motivation
before they completed questionnaires. We did this so on men’s risk perception for sexually transmitted disease:
that all participants could answer the items based on There must be 50 ways to justify a lover. Health Psychology,
16, 374–379.
accurate information about the outcome of the election.
Buckner, J. D., Ecker, A. H., & Cohen, A. S. (2010). Mental
However, informing participants of the outcome might health problems and interest in marijuana treatment among
have increased the salience of the vote and therefore cog- marijuana-using college students. Addictive Behaviors, 35,
nitions reported at T2 may reflect those that were made 826–833.
temporarily salient. Finally, the questionnaires were Center of Behavioral Health Statistics and Quality. (2015).
confidential, but not anonymous, so it is possible that Behavioral health trends in the United States: Results from
the 2014 National Survey on Drug Use and Health. Retrieved
participants underreported their intentions to use and from http://www.samhsa.giv/data/
use of marijuana at T1. This would work in favor of study Cerdá, M., Wall, M., Feng, T., Keyes, K. M., Sarvet, A., Schulen-
hypotheses. While it is impossible to rule out this entirely, berg, J., … Hasin, D. S. (2017). Association of state recre-
a number of precautions were taken to minimize con- ational marijuana laws with adolescent marijuana use. The
cerns about privacy and socially desirable responding. For Journal of the American Medical Association Pediatrics, 171,
142–149.
example, participants were assured that their professors
Choo, E. K., Benz, M., Zaller, M., Warren, O., Rising, K. L., &
would remain unaware of their participation in the survey. McConnell, K. J. (2014). The impact of state medical mar-
Participants were also informed that their data would be ijuana legislation on adolescent marijuana use. Journal of
linked with a codelink so that no identifying information Adolescent Health, 55, 160–166.
was present in the datafile and all identifying information Dupont, R. L., & Voth, E. A. (1995). Drug legalization, harm
would be destroyed prior to datafile download. reduction, and drug policy. Annals of Internal Medicine, 123,
461–465.
Despite these limitations we feel the study makes a Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention and
number of contributions to the literature that advance our behavior: An introduction to theory and research. Reading,
understanding of the effect of legislation on marijuana MA: Addison, Wesley.
use cognitions. This is one of the only published stud- Gerrard, M., Gibbons, F. X., Brody, G. H., Murry, V. M., Cleve-
ies to our knowledge that examines the effect of a spe- land, M. J., & Wills, T. A. (2006). A theory-based dual-focus
alcohol intervention for preadolescents: The Strong African
cific piece of marijuana legislation immediately pre and
American Families program. Psychology of Addictive Behav-
post voter approval. Results showed that when individ- iors, 20, 185–195.
uals were reminded of the outcome of Initiative 71 the Gerrard, M., Gibbons, F. X., Stock, M. L., Lune, L. S.,
attitudes and willingness of the most vulnerable users & Cleveland, M. J. (2005). Images of smokers and
became more positive. Furthermore, intentions for all willingness to smoke among African American pre-
participants, regardless of past experience with mari- adolescents: An application of the prototype/willingness
model of adolescent health risk behavior to smoking
juana, became more positive after learning the outcome initiation. Journal of Pediatric Psychology, 30, 305–318.
of the Initiative 71 vote. doi:10.1093/jpepsy/jsi026
Gibbons, F. X., Etcheverry, P. E., Stock, M. L., Gerrard, M., Weng,
C., Kiviniemi, M., … O’Hara, R. E. (2010). Exploring the
Declaration of interest link between racial discrimination and substance use: What
The authors report no conflicts of interest. The authors alone mediates? What buffers? Journal of Personality and Social
are responsible for the content and writing of the article. Psychology, 99, 785–801.
 SUBSTANCE USE & MISUSE 2173

Gibbons, F. X., Gerrard, M., Stock, M. L., & Finneran, S. (2015). National Institute on Drug Abuse. (2015). Drug facts: Nation-
The Prototype/Willingness Model. In M. Connor & P. Nor- wide trends. Retrieved from https://www.drugabuse.gov/
man (Eds.), Predicting health behavior: Research and practice publications/drugfacts/nationwide-trends
with social cognition models (3rd ed., pp. 189–224), Cam- Okaneku, J., Vearrier, D., McKeever, R. G., LaSala, G. S., &
bridge, UK: Cambridge University Press. Greenberg, M. I. (2015). Change in perceived risk associ-
Harper, S., Strumpf, E. C., & Kaufman, J. S. (2012). Do medical ated with marijuana use in the United States from 2002 to
marijuana laws increase marijuana use? Replication study 2012. Clinical Toxicology, 53, 151–155.
and extension. Annals of Epidemiology, 22, 207–212. Pacula, R. L., Powell, D., Heaton, P., & Sevigny, E. L. (2015).
Hasin, D. S., Wall, M., Keyes, K. M., Cerda, M., Schulenberg, Assessing the effects of medical marijuana laws on mari-
J., O’Malley, P. M., … Feng, T. (2015). Medical marijuana juana use: The devil is in the details. Journal of Policy Anal-
laws and adolescent marijuana use in the USA from 1991 ysis and Management, 34, 7–31.
to 2014: Results from annual, repeated, cross-sectional sur- Palamar, J. J., Ompad, D. C., & Petkova, E. (2014). Correlates
veys. Lancet Psychiatry, 2, 601–608. of intentions to use cannabis among US high school seniors
Ito, T. A., Henry, E. A., Cordova, K. A., & Bryan, A. D. (2015). in the case of cannabis legalization. International Journal of
Testing an expanded theory of planned behavior model to Drug Policy, 25, 424–435.
explain marijuana use among emerging adults in a promar- Pomery, E. A., Gibbons, F. X., Reis-Bergan, M., & Gerrard, M.
ijuana community. Psychology of Addictive Behaviors, 29, (2009). From willingness to intention: Experience moder-
576–589. ates the shift from reactive to reasoned behavior. Personality
Johnston, L. D., O’Malley, P. M., Bachman, J. G., Schulenberg, and Social Psychology Bulletin, 35, 894–908.
J. E., & Miech, R. A. (2015). Monitoring the Future national Schuermeyer, J., Salomonsen-Sautel, S., Price, R. K., Balan,
survey results on drug use, 1975–2014: Volume II, college stu- S., Thurstone, C., Min, S., … Sakai, J. T. (2014). Tem-
dents and adults ages 19–55. Ann Arbor: Institute for Social poral Trends in marijuana attitudes, availability and use
Research, The University of Michigan. Retrieved from in Colorado compared to non-medical marijuana states:
http://www.monitoringthefuture.org/pubs/monographs/ 2003–11. Drug and Alcohol Dependence, 140, 145–155.
mtf-vol2_2014.pdf doi:10.1016/j.drugalcdep.2014.04.016
Khatapoush, S., & Hallfors, D. (2004). “Sending the wrong Shrivastava, A., Johnston, M., & Tsuang, M. (2011). Cannabis
message”: Did medical marijuana legalization in Califor- use and cognitive dysfunction. Indian Journal of Psychiatry,
nia change attitudes about and use of marijuana? Journal of 53, 187–191.
Drug Issues, 34, 751–770. Stock, M. L., Gibbons, F. X., Beekman, J. B., & Gerrard, M.
Klein, W. M., & Kunda, Z. (1992). Motivated person percep- (2015). It only takes once: The absent-exempt heuristic
tion: Constructing justifications for desired beliefs. Journal and reactions to comparison-based sexual risk information.
of Experimental Social Psychology, 28, 145–168. Journal of Personality and Social Psychology, 109, 35–52.
Kraus, S. J. (1995). Attitudes and the prediction of behavior: doi:10.1037/a0039277
A meta-analysis of the empirical literature. Personality and Stock, M. L., Gibbons, F. X., Peterson, L. M., & Ger-
Social Psychology Bulletin, 21, 58–75. rard, M. (2013). The effects of racial discrimination on
Kunda, Z. (1990). The case for motivated reasoning. Psycholog- the HIV-risk cognitions and behaviors of Black adoles-
ical Bulletin, 108, 480–498. cents and young adults. Health Psychology, 32, 543–550.
Litt, D. M., & Stock, M. L. (2011). Adolescent alcohol-related doi:10.1037/a0028815
risk cognitions: The roles of social norms and social Stock, M. L., Litt, D. M., Arlt, V., Peterson, L. M., & Som-
networking sites. Psychology of Addictive Behaviors, 25, merville, J. (2013). The prototype/willingness model, aca-
708–713. demic versus health-risk information, and risk cognitions
Lopez-Quintero, C., & Neumark, Y. (2010). Effects of associated with nonmedical prescription stimulant use
risk perception of marijuana use on marijuana use among college students. British Journal of Health Psychology,
and intentions to use among adolescents in Bogotá, 18, 490–507.
Colombia. Drug and Alcohol Dependence, 109, 65–72. Suerken, C. K., Reboussin, B. A., Sutfin, E. L., Wagoner, K. G.,
doi:10.1016/j.drugalcdep.2009.12.011 Spangler, J., & Wolfson, M. (2014). Prevalence of marijuana
Lynne-Landsman, S. D., Livingston, M. D., & Wagenaar, A. C. use at college entry and risk factors for initiation during
(2013). Effects of state medical marijuana laws on adoles- freshman year. Addictive Behavior, 39, 302–307.
cent marijuana use. Research and Practice, 103, 1500–1506. Taylor, D. R., Poulton, R., Moffitt, E. T., Ramankutty, R. P.,
Miech, R. A., Johnston, L., O’Malley, P. M., Bachman, J. G., & Sears, R. M. (2000). The respiratory effects of cannabis
Schulenberg, J., & Patrick, M. E. (2015). Trends in use of dependence in young adults. Addiction, 95, 1669–1677.
marijuana and attitudes toward marijuana among youth Todd, J., Kothe, E., Mullan, B., & Monds, L. (2016). Reasoned
before and after decriminalization: The case of Califor- versus reactive prediction of behaviour: A meta-analysis of
nia 2007–2013. International Journal of Drug Policy, 26, the prototype willingness model. Health Psychology Review,
336–344. 10, 1–24.

View publication stats