By Christopher Shea

From the people who brought you the study that found that legalizing medical marijuana was associated with fewer deaths on the highways — possibly because people substitute marijuana for drinking — comes new contentious research involving pot. Its title is “High on Life? Medical Marijuana Laws and Suicide,” and this is the gist:

Using state-level data for the period 1990 through 2007, we estimate the effect of legalizing medical marijuana on suicide rates. Our results suggest that the passage of a medical marijuana law is associated with an almost 5 percent reduction in the total suicide rate, an 11 percent reduction in the suicide rate of 20- through 29-year-old males, and a 9 percent reduction in the suicide rate of 30- through 39-year-old males. Estimates of the relationship between legalization and female suicides are less precise and are sensitive to functional form.

It’s an IZA Discussion Paper, from January. http://www.iza.org/en/webcontent/publications/papers/viewAbstract?dp_id=6280

D I S C U S S I O N P A P E R S E R I E S
Forschungsinstitut
zur Zukunft der Arbeit
Institute for the Study
of Labor
High on Life? Medical Marijuana Laws and Suicide
IZA DP No. 6280
January 2012
D. Mark Anderson
Daniel I. Rees
Joseph J. Sabia
High on Life?
Medical Marijuana Laws and Suicide
D. Mark Anderson
Montana State University
Daniel I. Rees
University of Colorado Denver
and IZA
Joseph J. Sabia
San Diego State University
Discussion Paper No. 6280
January 2012
IZA
P.O. Box 7240
53072 Bonn
Germany
Phone: +49-228-3894-0
Fax: +49-228-3894-180
E-mail: iza@iza.org
Any opinions expressed here are those of the author(s) and not those of IZA. Research published in
this series may include views on policy, but the institute itself takes no institutional policy positions.
The Institute for the Study of Labor (IZA) in Bonn is a local and virtual international research center
and a place of communication between science, politics and business. IZA is an independent nonprofit
organization supported by Deutsche Post Foundation. The center is associated with the University of
Bonn and offers a stimulating research environment through its international network, workshops and
conferences, data service, project support, research visits and doctoral program. IZA engages in (i)
original and internationally competitive research in all fields of labor economics, (ii) development of
policy concepts, and (iii) dissemination of research results and concepts to the interested public.
IZA Discussion Papers often represent preliminary work and are circulated to encourage discussion.
Citation of such a paper should account for its provisional character. A revised version may be
available directly from the author.
IZA Discussion Paper No. 6280
January 2012
ABSTRACT
High on Life? Medical Marijuana Laws and Suicide
Using state-level data for the period 1990 through 2007, we estimate the effect of legalizing
medical marijuana on suicide rates. Our results suggest that the passage of a medical
marijuana law is associated with an almost 5 percent reduction in the total suicide rate, an 11
percent reduction in the suicide rate of 20- through 29-year-old males, and a 9 percent
reduction in the suicide rate of 30- through 39-year-old males. Estimates of the relationship
between legalization and female suicides are less precise and are sensitive to functional
form.
JEL Classification: I10, I18
Keywords: medical marijuana laws, marijuana, alcohol, suicide
Corresponding author:
Daniel I. Rees
University of Colorado Denver
Department of Economics
Campus Box 181
P.O. Box 173364
Denver, CO 80217-3364
USA
E-mail: Daniel.Rees@ucdenver.edu
1
1. INTRODUCTION
Under the Controlled Substances Act of 1970, marijuana was classified as a
Schedule I drug because it was considered to have no “accepted medical use in treatment
in the United States” (Eddy 2010). Since then, 16 states and the District of Columbia
have legalized the use of medical marijuana; six more state legislatures have recently
considered medical marijuana bills.
The relationship between marijuana and mental health has received a great deal of
attention from both proponents and opponents of legalization. Proponents argue that
marijuana can be an effective treatment for bipolarism, depression, and other mood
disorders (Rosenthal et al. 1996; Grinspoon and Bakalar 1997; Zimmerman 1999).1 They
also argue that medical marijuana patients are able to reduce their use of pain killers,
tranquilizers, and psychiatric medicines because of marijuana (Lucido 2004).2
Opponents, on the other hand, argue that marijuana use increases the likelihood of
depression, anxiety, psychosis, and schizophrenia (Zammit et al. 2002; Henquet et al.
2004; Goldberg 2006; Shulman 2008). They also argue that the negative effects of
marijuana are long-lasting and that users are at risk of suffering from decreased
psychological well-being later in life (Green and Ritter 2000; McGee et al. 2000).
This paper examines the relationship between medical marijuana laws (hereafter
MMLs) and suicides. While the majority of people who suffer from mental illness do not
commit suicide, over 90 percent of those who do commit suicide have a diagnosable
mental or substance use disorder (Moscicki 2001). The relationship between marijuana
1 Tetrahydrocannabinol (THC), the primary psychoactive ingredient in marijuana, produces mild euphoria
and has analgesic effects (Elphick and Egertova 2001).
2 Among some advocates, marijuana is referred to as “green Prozac” (Chaboya-Hembree 2011).
2
use and suicide-related outcomes (e.g., depression, suicidal ideation, and suicide
attempts) has been studied extensively (Petronis et al. 1990; Felts et al. 1992; Borowsky
et al. 2001), but there have been no previous attempts to estimate the effect of medical
marijuana laws on completed suicides, the tenth leading cause of death in the United
States (National Institute of Mental Health 2010).
This study begins by detailing the channels through which MMLs could impact
the risk of suicide. We cite research from epidemiological and medical studies showing
that marijuana use is associated with depression and suicidal ideation, although we note
that any observed association between marijuana use and outcomes such as these could
be driven by difficult-to-observe factors (Van Ours and Williams 2011) or simply reflect
self-medication (Gruber et al. 1996; Rey et al. 2002). We also note that previous studies
have found evidence to support the hypothesis that marijuana and alcohol are substitutes
(Chaloupka and Laixuthai 1997; Saffer and Chaloupka 1999; DiNardo and Lemieux
2001; Anderson et al. 2011; Crost and Guerrero 2011), and cite research linking alcohol
policies and suicides (Birckmayer and Hemenway 1990; Jones et al. 1992; Markowitz et
al. 2003; Carpenter 2004; Carpenter and Dobkin 2009).
The empirical analysis draws on data from the Centers for Disease Control from
1990 through 2007 to estimate the effect of legalization on suicide rates. Our results
suggest that the passage of a MML is associated with an almost 5 percent reduction in the
total suicide rate.
When we examine the relationship between legalization and suicides by gender
and age, we find evidence that MMLs are associated with decreased suicides among 20-
through 29-year-old males and among 30- through 39-year-old males. This result is
3
consistent with registry data from Arizona, Colorado, and Montana showing that most
medical marijuana patients are male, and that roughly half are under the age of 40
(Anderson et al. 2011). Estimates of the relationship between legalization and suicides
among females are less precise, and are sensitive to functional form. We conclude that
the legalization of medical marijuana leads to an improvement in the psychological wellbeing
of young adult males, an improvement that is reflected in fewer suicides.
2. BACKGROUND
2.1. MMLs and substance use
MMLs remove criminal penalties for using, possessing, and cultivating marijuana
for medicinal purposes. Because it is prohibitively expensive for the government to
ensure that all marijuana ostensibly grown for the medicinal market ends up in the hands
of registered patients, diversion to the illegal market likely occurs.3 Only a handful of
MMLs permit individuals with mental health problems to obtain medical marijuana.
California includes anxiety as a qualifying condition, while Delaware and New Mexico
both allow the use of medical marijuana for post traumatic stress disorder (Marijuana
Doctors 2011). However, without exception, MMLs allow patients to register based on
medical conditions that cannot be objectively confirmed (e.g., chronic pain and nausea),
and in some states it is possible to petition for access to medical marijuana based on
ailments not included on the official list of qualifying conditions (Marijuana Doctors
2011).
3 All MMLs enacted during the period 1990-2007 allowed for home cultivation. Since 2007, Delaware,
New Jersey and Washington D.C. have passed MMLs that do not permit home cultivation (Marijuana
Policy Project 2011).
4
Surprisingly little is known about the relationship between MMLs and marijuana
use. Gorman and Huber (2007) examined data on adult arrestees for the period 1995–
2002 from Denver, Los Angeles, Portland, San Diego and San Jose. They found little
evidence that marijuana use by arrestees increased as a result of legalization. In contrast,
Wall et al. (2011) found that rates of marijuana use among 12- through 17-year-olds were
higher in states that had legalized medical marijuana than in states that had not, but noted
that “in the years prior to MML passage, there was already a higher prevalence of use and
lower perceptions of risk” in states that had legalized medical marijuana.
Perhaps the best evidence on the relationship between MMLs and marijuana use
comes from Anderson et al. (2011). These authors used a difference-in-differences
estimation strategy and data from the National Survey on Drug Use and Health (NSDUH)
from three states that legalized medical marijuana in the mid-2000s: Montana, Rhode
Island, and Vermont. They found that marijuana use among 18- through 25-year-olds
living in Montana increased by 19 percent as a result of legalization. Marijuana use
among 18- through 25-year-olds living in Rhode Island increased by 21 percent, and
marijuana use by Rhode Island residents ages 26 and older increased by 44 percent.
However, there was no evidence that legalization in Vermont led to an increase in
consumption.4
Anderson et al. (2011) also examined the relationship between MMLs and alcohol
consumption. Using data from the Fatal Accident Report System (FARS) for the period
1990-2009, they found that legalization of medical marijuana led to a sharp reduction in
4 Vermont passed a MML in 2004, but did not allow for medical marijuana dispensaries until 2011. This
likely explains Anderson et al.’s (2011) null finding for Vermont.
5
fatal crashes involving alcohol. In addition, using data from the Behavioral Risk Factor
Surveillance System (BRFSS) for the same period, they found evidence that legalization
led to a 25 percent reduction in self-reported alcohol consumption among 20- through 29-
year-olds; using data from the Beer Institute, they found that legalization led to a 5.3
percent decrease in per capita beer sales, the most popular alcoholic beverage among
young adults (Jones 2008).5 Anderson et al. (2011) concluded that their results were
“consistent with the hypothesis that marijuana and alcohol are substitutes” (p. 18).6
2.2. Stressful life events and suicide
Hamermesh and Soss (1974) were the first economists to model suicidal behavior.
According to their “economic theory of suicide”, negative shocks to happiness can reduce
expected lifetime utility to the point where an individual will decide to take his or her
own life. Even if a shock is perceived to be temporary, an individual with “timeinconsistent
preferences” may commit suicide at the prospect of having to cope with an
acutely “painful state in the present” (Cutler et al. 2001, p. 235).7
In fact, there is evidence that suicide among adolescents and young adults is often
triggered by “stressful life events”, or SLEs. SLEs include the breakup of a romantic
5 Estimates from Anderson et al. (2011) on the relationship between MMLs and alcohol consumption are
reproduced in Appendix Tables 2 and 3.
6 Anderson et al. (2011) noted that a consensus has not been reached with regard to the relationship
between marijuana and alcohol consumption. A number of studies have found evidence of
complementarity between marijuana and alcohol (Pacula 1998; Farrelly et al. 1999; Williams et al. 2004;
Yörük and Yörük 2011), while others studies lend support to the hypothesis that marijuana and alcohol are
substitutes (Chaloupka and Laixuthai 1997; Saffer and Chaloupka 1999; DiNardo and Lemieux 2001; Crost
and Guerrero 2011).
7 There is evidence to support the hypothesis that preferences are time inconsistent (O’Donoghue and Rabin
2001, p. 41). Non-economists have exerted considerable effort to understand the rationality (or lack
thereof) of suicide. See, for instance, Henry and Short (1954), Brandt (1975) and Graber (1981).
6
relationship (De Wilde et al. 1992; Brent et al. 1993; Beautrais et al. 1997; Johnson et al.
2002), conflict with a parent or sibling (Brent et al. 1993; Johnson et al. 2002), an
abortion (Gissler, Hemminki, and Lonnqvist 1996; Gissler et al. 2005), and
legal/disciplinary problems (Brent et al. 1993).
Among older adults, problems at work, financial difficulties, unemployment and
separation/divorce are common triggers of suicide (Heikkinen et al. 1994; Johansson and
Sundquist 1997; Cavanagh et al. 1999; Preti and Miotto1999; Duberstein et al. 2004;
Kõlves et al. 2006). Among the elderly, suicide is often associated with physical illness
and functional impairment (Whitlock 1986; Mackenzie and Popkin 1987; Harris and
Barraclough 1994; Conwell et al 2000; Waern et al. 2002). A number of researchers
have concluded that the symptoms of depression (or self-reported depression) mediate the
relationship between SLEs and suicidal behavior (Fergusson et al. 2000; Conwell et al.
2002; Hardt and Johnson 2010).
2.3. Marijuana use and mental health
Could marijuana use help individuals cope with the negative shocks to happiness
that, according to Hamermesh and Soss (1974), lead to suicide? Alternatively, could
marijuana use magnify the impact of SLEs?
Animal studies provide evidence that, at low doses, synthetic cannabinoid
injections can have a potent anti-depressant effect (Jiang et al. 2005; Bambico et al.
7
2007). However, higher dosages can reduce serotonin transmission and lead to
“depression-like behavior” (Bambico et al. 2007; Bambico et al. 2010).8
The majority of epidemiological studies have found that marijuana use is
positively associated with both the symptoms of depression and suicidal ideation (Rey et
al. 2002; Moore et al. 2007).9 However, any association between marijuana use and
outcomes such as these could be driven by difficult-to-observe factors (Van Ours and
Williams 2011) or simply reflect self-medication (Gruber et al. 1996; Rey et al. 2002).
Although there have been attempts to account for these sources of endogeneity, none
have been particularly convincing. Moore et al. (2007, p. 325) concluded their review of
the literature by noting that “the majority of studies for affective outcomes [such as
depression and suicidal ideation] did not adequately address the problem of reverse
causation as a possible alternative explanation for any association observed.”
2.4. Alcohol use and mental health
Alcohol consumption represents an alternative route through which legalization of
medical marijuana could impact suicides. Numerous studies have shown that alcoholism
is more common among individuals with major depression (Sullivan et al. 2004).
8 Deficiency of the neurotransmitter serotonin, a chemical produced by the brain, is linked to depression
(Karg et al. 2011).
9 For instance, Bovasso (2001) found that marijuana use was positively associated with these outcomes
among Baltimore residents who participated in the Epidemiologic Catchment Area Study; Fergusson et al.
(2002) found similar results among teenagers and young adults in New Zealand. In contrast, Denson and
Earleywine (2006) found a negative association between marijuana use and the symptoms of depression.
Beautrais et al. (1999) found that “much of the association between cannabis abuse/dependence and suicide
attempt risk arose because: (a) individuals who develop cannabis abuse/dependency tend to come from
disadvantaged socio-demographic and childhood backgrounds…, or (b) because cannabis
abuse/dependence is co-morbid with other mental disorders which are independently associated with
suicidal behaviour.” Moore (2007) described the case of one woman who “without the relief that marijuana
delivered to her…killed herself at home”.
8
Moreover, alcoholism is associated with suicidal ideation as well as attempted and
completed suicides (Cornelious et al. 1995; Kessing 1999; Chatterji et al. 2004;
Rodriguez Andres 2005).
Policy studies provide further evidence of a link between alcohol use and suicide.
For instance, Birckmayer and Hemenway (1990) and Carpenter and Dobkin (2009)
provide evidence that prohibiting young adults from drinking protects against suicide. In
a similar vein, Carpenter (2004) found that the adoption of stricter drunk driving laws
lead to fewer suicides among young males. Markowitz et al. (2003) found a strong
negative relationship between the excise tax on beer and male suicides. 10 Interestingly,
neither Carpenter (2004) nor Markowitz et al. (2003) found consistent evidence that
alcohol policies had an impact on female suicides.
3. SUICIDE, TRENDS AND DATA
Data on suicides come from the National Center for Health Statistics Mortality
Detail Files, produced by the Centers for Disease Control.11 Table 1 shows the number
of suicides per 100,000 population by age and gender for the period under study, 1990-
2007.
Females suffer from depression and attempt suicide at higher rates than males
(Murphy 1998).12 Nevertheless, males are approximately four times more likely to
10 See also Son and Topyan (2011), who examined the relationship between alcoholic beverage excise taxes
and suicides using data from the United States for the period 1995-2004.
11 As pointed out by Birckmayer and Hemenway (1999) and Carpenter (2004), suicides are likely
underreported in these data. However, there is no reason to believe that this underreporting is
systematically related to the legalization of medical marijuana.
12 Murphy (1998) argued that females are less likely to commit suicide than males because they are more
likely to seek treatment for depression.
9
commit suicide. Since 1990, the male suicide rate has decreased slightly, while the
female suicide rate has remained fairly constant (Figure 1).
Fifteen- through 19-year-olds were less likely to commit suicide than older
individuals during the period under study; males over the age of 59 were the most likely
to commit suicide (Figures 2 and 3). Canetto (1992) argued that, as they age, men are
especially prone to suicide because they have unrealistic expectations and an inadequate
range of coping strategies. Clarke (1993, p. 21) hypothesized that “elderly persons who
die by suicide have a lifelong character fault that remains invisible until aging lifechanges
force the issue into the open.”
Figures 4 through 14 illustrate suicide trends by legalization status.13 For states
that legalized medical marijuana, we report the suicide rate for the three years prior to
legalization, the year in which the law changed (year 0), and the three years following
legalization. For states that did not legalize medical marijuana during the period under
study, the year in which the law changed was randomly assigned.
Figure 4 presents pre- and post-legalization trends in the total suicide rate. The
total suicide rate falls smoothly during the pre-legalization period in both MML and non-
MML states. However, beginning in year 0, the trends diverge: the suicide rate in MML
states continues to fall, while the suicide rate in states that never legalized medical
marijuana begins to climb gradually. Comparing Figures 5 and 6, it is clear that the postlegalization
downward trend in male suicides is much more pronounced than the postlegalization
downward trend in female suicides.
13 In each figure, a vertical line separates the years before legalization from the year of the law change and
the post- legalization period.
10
Consistent with the hypothesis that marijuana can be an effective treatment for
depression and other mood disorders, there appears to be a sharp decrease in the suicide
rate of 15- through 19-year olds males in the treatment states as compared to the control
states approximately two years after legalization (Figure 7). A similar decrease can be
seen for 20- through 59-year-old males (Figures 8 and 9), and there is a modest decrease
in the suicide rate of males 60 years of age and older approximately three years after
legalization (Figure 10). Among females under the age of 40, there is little evidence that
legalization lead to a decrease in the suicide rate (Figures 11 and 12). However, among
older females, there appears to be a small decrease in the suicide rate two to three years
after legalization (Figures 13 and 14).
4. THE EMPIRICAL MODEL
The trends discussed above could be driven by a number of factors such as statelevel
economic conditions or alcohol policies. In order to address this shortcoming, we
use a difference-in-differences (DD) regression framework to estimate the relationship
between MMLs and suicide rates. This framework allows us to control for both
unobserved state-level heterogeneity and national trends. Our baseline estimating
equation is:
(1) ln(Suicides Totalst) = β0 + β1MMLst + Xstβ2 + vs + wt + εst,
where s indexes states, t indexes years, and vs and wt represent state and year fixed
effects, respectively. The variable MMLst is an indicator for whether medical marijuana
11
was legal in state s and year t. The coefficient of interest, β1, represents the effect of
medical marijuana legislation. In alternative specifications we replace Suicides Totalst
with the other dependent variables listed in Table 1 and add state-specific linear time
trends to the right-hand side.14 The state-specific trends are intended to control for the
influence of difficult-to-measure factors at the state level that evolve smoothly over time
such as attitudes.15
The vector Xst includes indicators for whether Zero Tolerance drunk driving and
0.08 BAC laws were enforced in state s and year t, and the state beer tax. Previous
studies provide evidence that stricter alcohol policies and higher beer taxes can reduce
suicides (e.g., Jones et al. 1992; Birckmayer and Hemenway 1999; Markowitz et al.
2003; Carpenter 2004).16 In addition, the vector Xst includes the state unemployment rate
and income per capita.17 Previous studies have shown that suicide rates are sensitive to
economic conditions (Gerdtham and Johannesson 2003; Andres 2005; Minoiu and
Rodriguez Andres 2008; Chung 2009). Finally, we control for whether marijuana use
and possession was decriminalized.18
14 In other words, we include an interaction between the state dummies and a variable equal to 1 in 1990, 2
in 1991, 3 in 1993, and so forth.
15 DeCicca et al. (2008) found that attitudes toward cigarette smoking were an important determinant of
youth smoking participation.
16 Data on the Zero Tolerance laws, 0.08 BAC laws, and beer taxes are from the National Highway Traffic
Safety Administration, Freeman (2007), and the Beer Institute, respectively. Jones et al. (1992) and
Birckmayer and Hemenway (1999) both found that higher minimum legal drinking ages (MLDA) were
associated with fewer youth suicides. Carpenter and Dobkin (2009) used a regression discontinuity design
to show that granting legal access to alcohol at age 21 leads to a large and immediate increase in suicides.
But, because all states had a minimum legal drinking age of 21 by 1990, we do not control for the MLDA.
17 The unemployment and income data are from the Bureau of Labor Statistics and the Bureau of Economic
Analysis, respectively.
18 During the period 1990-2007, the decriminalization indicator only captures one policy change: Nevada
decriminalized marijuana in 2001. The majority of these laws were passed prior to 1990.
12
5. THE RESULTS
5.1. The baseline results
Table 2 presents OLS estimates of the relationship between MMLs and suicides.
All regressions are weighted by the relevant population in state s and year t, and the
standard errors are corrected for clustering at the state level (Bertrand et al. 2004).
The estimated relationship between MMLs and the total suicide rate is presented
in the first column of Table 2. Controlling for year and state fixed effects, the
legalization of medical marijuana is associated with an 8.1 percent decrease in the suicide
rate (e-0.084 – 1 = -0.081). When the controls listed in Table 1 are included, legalization is
associated with a 6.9 percent decrease in the suicide rate; when state-specific linear time
trends are included, legalization is associated with a 4.8 percent decrease in the suicide
rate.
In the remaining columns of Table 2, we allow the estimated relationship between
legalization and suicides to differ by gender. MMLs are associated with a 4.6 to 7.0
percent decrease in the male suicide rate. MMLs are negatively associated with the
female suicide rate, but these estimates are not statistically significant at conventional
levels.
5.2. Estimates by gender and age group
Next, we examine the relationship between the legalization of medical marijuana
and suicides by age group (Table 3). Our estimates suggest that legalization leads to a
sharp reduction in suicides among young adult males. For instance, in the fully-specified
model, legalization is associated with a 10.8 percent decrease in the suicide rate of 20-
through 29-year-old males, and a 9.4 percent decrease in the suicide rate of 30- through
13
39-year-old males. These results are consistent with those of Anderson et al. (2011), who
found that MMLs have the greatest impact on the substance use of young adults, and are
also consistent with registry data from Arizona, Colorado, and Montana showing that
most medical marijuana patients are male and that roughly half of medical marijuana
patients are under the age of 40 (Anderson et al. 2011).
Although the estimates of β1 are negative for males between the ages of 15 and
19, they are never significantly different from zero. Likewise, the estimates of β1 for
males ages 40 through 59 are consistently negative, but not precisely estimated. For
males 60 years of age and older, estimates are negative in two of the three specifications,
but none are precisely estimated.
Table 4 presents the results for females by age group. There is little evidence that
legalization of medical marijuana impacted suicides among females under the age of 30.
The estimated relationship between legalization and suicides is stronger among older
females, although it is sensitive to model specification. Controlling for state and year
fixed effects and the covariates listed in Table 1, the legalization of medical marijuana is
associated with a 10.4 percent decrease in the suicide rate of 30- through 39-year-old
females. However, the estimated impact of legalization becomes much smaller in
magnitude and is no longer statistically significant when we add state-specific linear time
trends.
Although consistently negative, the estimated impact of legalization for females
ages 40 through 49 is never statistically significant. However, when we control for statespecific
linear time trends, legalization is associated with large and statistically
14
significant decreases in the suicide rate of 50-through 59-year-old females and the suicide
rate of females ages 60 and older.
5.3. Adding leads and lags to the model
In Table 5 we examine pre- and post-legalization trends in suicides using the
regression framework introduced above. Specifically, we include three lead indicators,
an indicator for the year of the law change, and four lag indictors. The omitted category
is 3+ years before a MML was passed.
The results are consistent with the hypothesis that MMLs have a causal impact on
male suicide rates. Controlling for year and state fixed effects, the covariates listed in
Table 1 and state-specific linear trends, none of the lead dummies are statistically
distinguishable from zero. However, once a MML is passed, there is a strong negative
relationship between legalization and the male suicide rate.
In contrast, the evidence that legalization reduces suicides among females is much
weaker. The estimated coefficients of the three lead indicators are large, negative, and
statistically significant, suggesting that there was a systematic change in the suicidal
behavior of females preceding legalization. Although the estimated coefficients of the
lag indicators are two to three times larger than those of the lead indicators, we cannot
reject the hypothesis that they are equal.19
5.4. A placebo test
19 As noted above, Carpenter (2004) found evidence that Zero Tolerance drunk driving laws were
negatively associated with male suicides. He also found evidence of an implausibly large (and statistically
significant) relationship between Zero Tolerance drunk driving laws and female suicides.
15
Lastly, we performed a robustness check based on Luallen (2006).20 Specifically,
we generated a placebo MML indicator using a random number generator based on the
continuous uniform distribution. Because 12 states passed a MML during our sample
period, we assigned 12 placebo policies for each trail run.21 In total, 50 trials were run.
Table 6 illustrates the average coefficient estimates for the placebo MML on total,
male and female suicide rates. In all three sets of regressions, the average estimate is
quite small in magnitude. Moreover, in 50 trials, only four estimates were negative and
statistically significant at the 5 percent level for the total suicide rate. For male and
female suicide rates, only three and four of the estimates were negative and statistically
significant at the 5 percent level, respectively. This pattern of results provides evidence
that our estimates cannot be readily reproduced through random assignment of
legalization.
6. CONCLUSION
Numerous epidemiological studies have found a positive association between
marijuana and suicide-related outcomes such as depression and suicidal ideation.
However, this association could be driven by difficult-to-observe factors or simply reflect
self-medication. In fact, a recent review of the epidemiological literature concluded that
most studies had not adequately addressed the problem of reverse causation (Moore et al.
2007).
20 Luallen (2006) examined the relationship between teacher strike days and juvenile crime.
21 A year of legalization between 1990 and 2007 was randomly selected. Randomly assigning a
legalization year between 1999 and 2007 produced qualitatively similar results.
16
The current study avoids the problems of reverse causality and unobserved
heterogeneity at the individual level by examining the relationship between the
legalization of medical marijuana and completed suicides. Specifically, we use a
difference-in-differences empirical strategy to estimate the impact of MMLs on statelevel
suicide rates. Our results suggest that the legalization of medical marijuana is
associated with a 5 percent decrease in the total suicide rate, an 11 percent decrease in the
suicide rate of 20- through 29-year-old males, and a 9 percent decrease in the suicide rate
of 30- through 39 year-old-males. Estimates for female suicide rates are generally
measured with less precision and are sensitive to functional form.
In an often-cited article, Hamermesh and Soss (1974) argued that negative shocks
to happiness may reduce expected lifetime utility to the point where an individual will
decide to take his or her own life. The negative relationship between legalization and
suicides among young adult males is consistent with the argument that marijuana can be
used to cope with such shocks. However, estimates provided by Anderson et al. (2011)
provide an alternative explanation. These authors found that the passage of MMLs led to
sharp decreases in alcohol-related traffic fatalities, self-reported alcohol use, and per
capita beer sales. The strong association between alcohol consumption and suiciderelated
outcomes found by previous researchers (Markowitz et al. 2003; Carpenter 2004;
Sullivan et al. 2004; Rodriguez Andres 2005; Carpenter and Dobkin 2009) raises the
possibility that medical marijuana laws reduce the risk of suicide by decreasing alcohol
consumption.
Although 16 states have passed medical marijuana laws, and others are
considering legalization, very little is known about the effects of these laws.
17
Policymakers weighing the pros and cons of legalization should consider the possibility
that medical marijuana laws may lead to fewer suicides among young adult males.
However, the exact mechanism through which suicides are reduced remains a topic for
future study.
REFERENCES
Anderson, D. Mark, Daniel I. Rees, and Benjamin Hansen. 2011. “Medical Marijuana
Laws, Traffic Fatalities, and Alcohol Consumption.” Working Paper, Montana
State University.
Bambico, Francis Rodríguez, Noam Katz, Guy Debonnel, and Gabriella Gobbi. 2007.
“Cannabinoids Elicit Antidepressant-Like Behavior and Activate Serotonergic
Neurons through the Medial Prefrontal Cortex.” Journal of Neuroscience 27:
11700-11711.
Bambico, Francis Rodrigues, Nhu-Tram Nguyen, Noam Katz, and Gabriella Gobbi.
2010. “Chronic Exposure to Cannabinoids during Adolescence but Not during
Adulthood Impairs Emotional Behaviour and Monoaminergic
Neurotransmission.” Neurobiology of Disease 37: 641-655.
Beautrais, Annette, Peter Joyce, and Roger Mulder. 1997. “Precipitating Factors and
Life Events in Serious Suicide Attempts Among Youths Aged 13 Through 24
Years.” Journal of the American Academy of Child & Adolescent Psychiatry 36:
1543-1551.
Beautrais, Annette, Peter Joyce, and Roger Mulder. 1999. “Cannabis Abuse and Serious
Suicide Attempts.” Addiction 94: 1155-1164.
Bertrand, Marianne, Esther Duflo, and Sendhil Mullainathan. 2004. “How Much Should
We Trust Differences-in-Differences Estimates?” Quarterly Journal of
Economics 119: 249-276.
Birckmayer, Johanna, and David Hemenway. 1999. “Minimum-Age Drinking Laws and
Youth Suicide, 1970–1990.” American Journal of Public Health 89: 1365-1368.
Borowsky, Iris, Marjorie Ireland, and Michael Resnick. 2001. “Adolescent Suicide
Attempts: Risks and Protectors.” Pediatrics 107: 485-493.
Bovasso, Gregory. 2001. “Cannabis Abuse as a Risk Factor for Depressive Symptoms.”
American Journal of Psychiatry 158: 2033-2037.
18
Brandt, Richard. 1975. “The Morality and Rationality of Suicide.” In Seymour Perlin,
ed., A Handbook for the Study of Suicide. Oxford: Oxford University Press.
Brent, David, Joshua Perper, Grace Moritz, Marianne Baugher, Claudia Roth, Lisa
Balach, and Joy Schweers. 1993. “Stressful Life Events, Psychopathology, and
Adolescent Suicide: A Case Control Study.” Suicide and Life-Threatening
Behavior 23: 179-187.
Canetto, Silvia. 1992. “Gender and Suicide in the Elderly.” Suicide and Life-
Threatening Behavior 22: 80-97.
Carpenter, Christopher. 2004. “Heavy Alcohol Use and Youth Suicide: Evidence from
Tougher Drunk Driving Laws.” Journal of Policy Analysis and Management 23:
831-842.
Carpenter, Christopher, and Carlos Dobkin. 2009. “The Effect of Alcohol Consumption
on Mortality: Regression Discontinuity Evidence from the Minimum Drinking
Age.” American Economic Journal: Applied Economics 1: 164-182.
Cavanagh, J.T.O., D.G.C. Owens, and E.C. Johnstone. 1999. “Life Events in Suicide and
Undetermined Death in South-East Scotland: A Case-Control Study using the
Method of Psychological Autopsy.” Social Psychiatry and Psychiatric
Epidemiology 34: 645-650.
Chaboya-Hembree, Jan. 2011. “Medical Marijuana as a Treatment for Depression.”
Available at:

http://medicalmarijuana.com/experts/expert/title.cfm?artID=65.

Chatterji, Pinka, Daval Dave, Robert Kaestner, and Sara Markowitz. 2004. “Alcohol
Abuse and Suicide Attempts among Youth.” Economics and Human Biology 2:
159-180.
Chung, Andy. 2009. “Gender Difference in Suicide, Household Production and
Unemployment.” Applied Economics 41: 2495-2504.
Conwell Y., J.M. Lyness, P. Duberstein, C. Cox, L. Seidlitz, A. DiGiorgio, and E.D.
Caine. 2000. “Completed suicide among older Patients in Primary Care Practices:
A Controlled Study.” Journal of the American Geriatrics Society Soc 48: 23–29.
Conwell, Y, P.R. Duberstein, and E.D. Caine. 2002. "Risk Factors for Suicide in Later
Life.” Biological Psychiatry 52:193-204.
Cornelius, Jack, I. Salloum, J. Mezzich, Marie Cornelius, Horacio Fabrega, Jr., J. Ehler,
19
R. Ulrich, M. Thase, and J. Mann. 1995. “Disproportionate Suicidality in Patients
with Comorbid Major Depression and Alcoholism.” American Journal of
Psychiatry 152: 358-364.
Cutler, David, Edward Glaeser, and Karen Norberg. 2001. “Explaining the Rise in
Youth Suicide.” In Jonathan Gruber, ed., Risky Behavior among Youths: An
Economic Analysis. Chicago, IL: University of Chicago Press.
DeCicca, Philip, Donald Kenkel, Alan Mathios, Yoon-Jeong Shin, and Jae-Young Lim.
2008. “Youth Smoking, Cigarette Prices, and Anti-Smoking Sentiment.” Health
Economics 17: 733-749.
Denson, Thomas and Mitchell Earleywine. 2006. “Decreased Depression in Marijuana
Users.” Addictive Behaviors 31: 738-742.
De Wilde, E., I. Kienhorst, R. Diekstra, and W. Wolters 1992. “The Relationship
between Adolescent Suicidal Behavior and Life Events in Childhood and
Adolescence.” American Journal of Psychiatry 149: 45-51.
Duberstein, P. R., Y. Conwell, K.R. Conner. S. Eberly, and E.D. Caine. 2004. “Suicide at
50 Years of Age and Older: Perceived Physical Illness, Family Discord and
Financial Strain.” Psychological Medicine, 34: 137–146.
Eddy, Mark. 2010. “Medical Marijuana: Review and Analysis of Federal and State
Policies.” Congressional Research Service Report for Congress. Available at:

http://www.fas.org/sgp/crs/misc/RL33211.pdf.

Elphick, Maurice and Michaela Egertova. 2001. “The Neurobiology and Evolution of
Cannabinoid Signalling.” Philosophical Transactions of the Royal Society B:
Biological Sciences 356: 381-408.
Felts, W. Michael, Thomas Chenier, and Robert Barnes. 1992. “Drug Use and Suicide
Ideation and Behavior among North Carolina Public School Students.” American
Journal of Public Health 82: 870-872.
Fergusson, David M., Annette Beautrais, and L. John Horwood. 2003. “Vulnerability and
Resiliency to Suicidal Behaviours in Young People.” Psychological Medicine 33:
61-73.
Fergusson, David M., L. John Horwood, and Nicola Swain-Campbell. 2002. “Cannabis
Use and Psychosocial Adjustment in Adolescence and Young Adulthood.”
Addiction 97: 1123-1135.
Fergusson, David M., L. J. Woodward., and L. J. Horwood. 2000. “Risk Factors and Life
Processes Associated with the Onset of Suicidal Behaviour during Adolescence
and Early Adulthood.” Psychological Medicine 30: 23-39.
20
Freeman, Donald. 2007. “Drunk Driving Legislation and Traffic Fatalities: New
Evidence on BAC 08 Laws.” Contemporary Economic Policy 25: 293-308.
Gerdtham, Ulf-G. and Magnus Johannesson. 2003. “A Note on the Effect of
Unemployment on Mortality.” Journal of Health Economics 22: 505-518.
Gissler, Mika, Cynthia Berg, Marie-Helene Bouvier-Colle, and Pierre Buekens. 2005.
“Injury Deaths, Suicides and Homicides Associated with Pregnancy, Finland
1987-2000.” European Journal of Public Health 15: 459-63.
Gissler, Mika, Elina Hemminki, and Jouko Lonnqvist. 1996. “Suicides after Pregnancy
in Finland 1987-94: Register Linkage Study.” British Medical Journal 313:
1431-1434.
Goldberg, Carey. 2006. “Studies Link Psychosis, Teenage Marijuana Use.” Boston
Globe 26 January. Available at:

http://www.boston.com/yourlife/health/children/articles/2006/01/26/studies_link_

psychosis_teenage_marijuana_use/.
Gorman, Dennis and Charles Huber Jr. 2007. “Do Medical Cannabis Laws Encourage
Cannabis Use?” International Journal of Drug Policy 18: 160-167.
Graber, Glenn. 1981. “The Rationality of Suicide.” In Samuel Wallace and Albin Eser,
eds., Suicide and Euthanasia: The Rights of Personhood. Knoxville: U. Tennessee
Press.
Grinspoon, Lester and James Bakalar. 1997. Marihuana: The Forbidden Medicine.
New Haven, CT: Yale University Press.
Gruber, Amanda, Harrison Pope, and Michael Brown. 1996. “Do Patients Use
Marijuana as an Antidepressant?” Depression 4: 77-80.
Hardt, Jochen, and Jeffrey G. Johnson. 2010. “Suicidality, Depression, Major and Minor
Negative Life Events: A Mediator Model.” German Medical Science
Psychosocial Medicine 7: Doc05.
Harris, E.C., and B.M. Barraclough. 1994. “Suicide as an Outcome for Medical
Disorders.” Medicine 73: 281–296.
Hamermesh, Daniel, and Neal Soss. 1974. “An Economic Theory of Suicide.” Journal
of Political Economy 82: 83-98.
Heikkinen, M, H. Aro, and J. Lönnqvist. 1994. “Recent Life Events, Social Support and
Suicide.” Acta Psychiatrica Scandinavica 89: 65–72.
Henry, Andrew, and James Short. 1954. Suicide and Homicide. Glencoe, IL: Free Press.
21
Henquet, Cecile, Lydia Krabbendam, Janneke Spauwen, Charles Kaplan, Roselind Lieb,
Hans-Ulrich Wittchen, and Jim van Os. 2005. “Prospective Cohort Study of
Cannabis Use, Predisposition for Psychosis, and Psychotic Symptoms in Young
People.” British Medical Journal 330:11-15.
Jiang, Wen, Yun Zhang, Lan Xiao, Jamie Van Cleemput; Shao-Ping Ji, Guang Bai,
and Xia Zhang. 2005. “Cannabinoids Promote Embryonic and Adult
Hippocampus Neurogenesis and Produce Anxiolytic- and Antidepressant-Like
Effects.” Journal of Clinical Investigation 115: 3104-3116.
Johansson, S.E. and J. Sundquist. 1997. “Unemployment is an Important Risk Factor for
Suicide in Contemporary Sweden: An 11-year Follow-up Study of a Crosssectional
Sample of 37,789 People.” Public Health 111: 41-45.
Johnson, Jeffrey, Patrick Cohen, Madelyn Gould, Stephanie Kasen, Jocelyn Brown,
and Judith Brook. 2002. “Childhood Adversities, Interpersonal Difficulties, and
Risk for Suicide Attempts during Late Adolescence and Early Adulthood.”
Archives of General Psychiatry 59: 741-749.
Jones, Jeffrey. 2008. “Beer Back to Double-Digit Lead Over Wine as Favored Drink.”
Gallup. 25 July. Available at:

http://www.gallup.com/poll/109066/beer-back-doubledigit-lead-over-wine-f

avored-drink.aspx.
Jones, Nancy, Carl Pieper, and Leon Roberston. 1992. “The Effect of Legal Drinking
Age on Fatal Injuries of Adolescents and Young Adults.” American Journal of
Public Health 82: 112-115.
Karg, Katja, Margit Burmeister, Kerby Shedden, and Srijan Sen. 2011. “The Serotonin
Transporter Promoter Variant (5-HTTLPR), Stress, and Depression Meta-analysis
Revisted.” Archives of General Psychiatry 68: 444-454.
Kessing, Lars Vedel. 1999. “The Effect of Comorbid Alcoholism on Recurrence in
Affective Disorder: A Case Register Study.” Journal of Affective Disorders 53:
49-55.
Kõlves, Kairi, Airi Värnik, Barbara Schneider, Jürgen Fritze, and Jüri Allik. 2006.
“Recent Life Events and Suicide: A Case-Control Study in Tallinn and
Frankfurt.” Social Science and Medicine 62: 2887-2896.
Lewinsohn, Peter, Paul Rohde, and John Seeley. 1994. “Psychosocial Risk Factors for
Future Adolescent Suicide Attempts.” Journal of Consulting and Clinical
Psychology 62: 297-305.
Luallen, Jeremy. 2006. “School’s Out…Forever: A Study of Juvenile Crime, At-Risk
22
Youths and Teacher Strikes.” Journal of Urban Economics 59: 75-103.
Lucido, Frank. 2004. “Implementation of the Compassionate Use Act in a Family
Medicine Practice: Seven Years Clinical Experience.” Available at:

http://www.medboardwatch.com.

Marijuana Doctors. 2011. “Medical Marijuana: Medical Marijuana Laws, Qualification
Guidelines, and ID Card Information.” Available at:

https://www.marijuanadoctors.com/medical-marijuana/index.

Marijuana Policy Project. 2011. “Key Aspects of State and D.C. Medical Marijuana
Laws.” Available at:

http://www.mpp.org/legislation/state-by-state-medical-marijuana-laws.html.

Markowitz, Sara, Pinka Chatterji, and Robert Kaestner. 2003. “Estimating the Impact of
Alcohol Policies on Youth Suicides.” Journal of Mental Health Policy and
Economics 6: 37-46.
Marmor, Jane. 1998. “Medical Marijuana.” Western Journal of Medicine 168: 540-543.
Mackenzie, T.B. and M.K. Popkin. 1987. “Suicide in the Medical Patient.”
International Journal of Psychiatry in Medicine 17: 3–22.
McGee, Rob, Sheila Williams, Richie Poulton, and Terrie Moffitt. 2000. “A
Longitudinal Study of Cannabis Use and Mental Health from Adolescence to
Early Adulthood.” Addiction 95: 491-503.
Minoiu, Camelia and Antonio Rodriguez Andres. 2008. “The Effect of Public Spending
on Suicide: Evidence from U.S. State Data.” Journal of Socio-Economics 37:
237-261.
Model, Karyn. 1993. “The Effect of Marijuana Decriminalization on Hospital
Emergency Room Drug Episodes: 1975-1978.” Journal of the American
Statistical Association 88: 737-747.
Moore, Michael. 2007. “Medical Marijuana User Commits Suicide After Long Being
Denied Its Use.” Missoulian 27 October. Available at:

http://digg.com/news/story/Medical_Marijuana_User_Commits_Suicide_After_L

ong_Being_Denied_Its_Use.
Moore, Theresa, Stanley Zammit, Anne Lingford-Hughes, Thomas Barnes, Peter Jones,
Margaret Burke, and Glyn Lewis. 2007. “Cannabis Use and Risk of
Psychotic or Affective Mental Health Outcomes: A Systematic Review.” Lancet
370: 319-328.
23
Moscicki, Eve. 2001. “Epidemiology of Completed and Attempted Suicide: Toward a
Framework for Prevention.” Clinical Neuroscience Research 1: 310-323.
Murphy, George. 1998. “Why Women Are Less Likely Than Men to Commit Suicide.”
Comprehensive Psychiatry 39: 165-175.
National Institute of Mental Health. 2010. “Suicide in the U.S.: Statistics and
Prevention.” 27 September. Available at:

http://www.nimh.nih.gov/health/publications/suicide-in-the-us-statistics-andprevention/

index.shtml.
O’Donoghue, Ted and Matthew Rabin. 2001.“Risky Behavior among Youths: Some
Issues from Behavioral Economics.” In Jonathan Gruber, ed., Risky Behavior
among Youths: An Economic Analysis. Chicago, IL: University of Chicago
Press.
Pabst Battin, Margaret. 1996. The Death Debate: Ethical Issues in Suicide. Upper Saddle
River, NJ: Prentice-Hall.
Petronis, K., J. Samuels, E. Moscicki, and J. Anthony. 1990. “An Epidemiologic
Investigation of Potential Risk Factors for Suicide Attempts.” Social Psychiatry
and Psychiatric Epidemiology 25: 193-199.
Preti, Antonia and Paola Miotto. 1999. “Suicide and Unemployment in Italy, 1982-1994.
Journal of Epidemiology and Community Health 53: 694-701.
Reading, Anthony. 2004. Hope and Despair: How Perceptions of the Future Shape
Human Behavior. Baltimore, MD: Johns Hopkins University Press.
Rey, Joseph, Michael Sawyer, Beverley Raphael, George Patton, and Michael Lynskey.
2002. “Mental Health of Teenagers Who Use Cannabis: Results of an Australian
Survey.” British Journal of Psychiatry 180: 216-221.
Rodriguez Andres, Antonio. 2005. “Income Inequality, Unemployment, and Suicide: A
Panel Data Analysis of 15 European Countries.” Applied Economics 37: 439-451.
Rosenthal, Ed, Dale Gieringer, and Tod Mikuriya. 1996. Marijuana Medical
Handbook: A Guide to Therapeutic Use. San Francisco, CA: Quick American
Archives.
Scott, Emilee Mooney. 2010. “Marijuana Decriminalization.” Office of Legislative
Research Report. Available at:

http://www.cga.ct.gov/2010/rpt/2010-R-0204.htm.

Ross R.K., L. Bernstein, L. Trent, B.E. Henderson, and A. Paganini-Hill. 1990. “A
24
Prospective Study of Risk Factors for Traumatic Death in the Retirement
Community.” Preventative Med 19: 323–334.
Shaffer, David, Madely Gould, Prudence Fisher, Paul Trautman, Donna Moreau,
Marjorie Kleinman, and Michael Flory. 1996. “Psychiatric Diagnosis in Child and
Adolescent Suicide.” Archives General Psychiatry 53: 339-348.
Shulman, Matthew. 2008. “Marijuana Linked to Heart Disease and Depression.” U.S.
News 14 May. Available at:

http://health.usnews.com/health-news/familyhealth/

heart/articles/2008/05/14/marijuana-linked-to-heart-disease-anddepression.
Son, Chong Hwan and Kudret Topyan. 2011. “The Effect of Alcoholic Beverage Excise
Tax on Alcohol-Attributable Injury Mortalities.” European Journal of Health
Economics 12: 103-113.
Sullivan, Lynn, David Fiellin, and Patrick O’Connor. 2004. “The Prevalence and Impact
of Alcohol Problems in Major Depression: A Systematic Review.” American
Journal of Medicine 118: 330-341.
Vaillant, George E. and Susan J. Blumenthal. 1990. “Introduction: Suicide Over the Life
Cycle - Risk Factors and Life Span Development.” In Susan J. Blumenthal and
David J. Kupfer, eds., Suicide Over the Life Cycle: Risk Factors, Assessment, and
Treatment of Suicidal Patients. Washington DC: American Psychiatric Press Inc.
Van Ours, Jan and Jenny Williams. 2011. “Cannabis Use and Mental Health Problems.”
Journal of Applied Econometrics 26: 1137-1156.
Wall, Melanie, Ernest Poh, Magdalena Cerdá, Katherine Keyes, Sandro Galea, and
Deborah Hasin. 2011. “Adolescent Marijuana Use from 2002 to 2008: Higher in
States with Medical Marijuana Laws, Cause Still Unclear.” Annals of
Epidemiology 21: 714-716.
Waern, Margda, Bo S. Runeson, Peter Allebeck, Jan Beskow, Eva Rubenowitz, Ingmar
Skoog, and Katarina Wilhelmsson. 2002. “Mental Disorder in Elderly Suicides: A
Case-Control Study.” American Journal of Psychiatry 159: 450–455.
Watson, Stanley, John Benson, Jr., and Janet Joy. 2000. “Marijuana and Medicine:
Assessing the Science Base.” Archives of General Psychiatry 57: 547-552.
Whitlock, Francis. 1986. “Suicide and physical illness.” In A. Roy, ed., Suicide.
Baltimore: Williams & Wilkins.
Zammit, Stanley; Peter Allebeck, Sven Andreasson, Ingvar Lundberg, and Glyn Lewis.
2002. “Self Reported Cannabis Use as a Risk Factor for Schizophrenia in
Swedish Conscripts of 1969: Historical Cohort Study.” British Medical Journal
25
325: 1199-1203.
Zimmerman, Bill. 1999. Is Marijuana the Right Medicine for You? New York, NY:
Pocket Books.
0
5
10
15
20
25
30
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Mean Suicides per 100,000
Figure 1. National Suicide Rates
Total Suicide Rate
Male Suicide Rate
Female Suicide Rate
0
5
10
15
20
25
30
35
40
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Mean Suicides per 100,000
Figure 2. National Suicide Rates by Age, Males
Males 15 to 19
Males 20 to 39
Males 40 to 59
Males 60+
0
1
2
3
4
5
6
7
8
9
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Mean Suicides per 100,000
Figure 3. National Suicide Rates by Age, Females
Females 15 to 19
Females 20 to 39
Females 40 to 59
Females 60+
12
13
14
15
16
17
18
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 4. Pre- and Post-Legalization Trends in Suicide Rates,
Total
MML
Non-MML
22
23
24
25
26
27
28
29
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 5. Pre- and Post-Legalization Trends in Suicide Rates,
Males
MML
Non-MML
4
4.5
5
5.5
6
6.5
7
7.5
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 6. Pre- and Post-Legalization Trends in Suicide Rates,
Females
MML
Non-MML
8
10
12
14
16
18
20
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 7. Pre- and Post-Legalization Trends in Suicide Rates,
Males 15 to 19 years-old
MML
Non-MML
19
20
21
22
23
24
25
26
27
28
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 8. Pre- and Post-Legalization Trends in Suicide Rates,
Males 20 to 39 years-old
MML
Non-MML
23
23.5
24
24.5
25
25.5
26
26.5
27
27.5
28
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 9. Pre- and Post-Legalization Trends in Suicide Rates,
Males 40 to 59 years-old
MML
Non-MML
25
27
29
31
33
35
37
39
41
43
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 10. Pre- and Post-Legalization Trends in Suicide Rates,
Males 60+ years-old
MML
Non-MML
2
2.5
3
3.5
4
4.5
5
5.5
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 11. Pre- and Post-Legalization Trends in Suicide Rates,
Females 15 to 19 years-old
MML
Non-MML
3
3.5
4
4.5
5
5.5
6
6.5
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 12. Pre- and Post-Legalization Trends in Suicide Rates,
Females 20 to 39 years-old
MML
Non-MML
5
5.5
6
6.5
7
7.5
8
8.5
9
9.5
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical
marijuana during the period under study (1990-2007).
Figure 13. Pre- and Post-Legalization Trends in Suicide Rates,
Females 40 to 59 years-old
MML
Non-MML
3
4
5
6
7
8
9
-3 -2 -1 Year of
law
change
1 2 3
Mean Suicides per 100,000
On the horizontal axis, 0 represents the year in which medical marijuana was
legalized. It was randomly assigned to states that did not legalize medical marijuana
during the period under study (1990-2007).
Figure 14. Pre- and Post-Legalization Trends in Suicide Rates,
Females 60+ years-old
MML
Non-MML
Table 1. Weighted Means of Dependent and Independent Variables
Dependent Variables Mean (SD)
Suicide rates per 100,000
Suicides Total 14.3 (3.47)
Suicides Male 23.7 (5.72)
Ages 15-19 14.4 (6.13)
Ages 20-29 22.9 (6.92)
Ages 30-39 22.9 (6.27)
Ages 40-49 24.0 (6.19)
Ages 50-59 23.3 (6.10)
Ages 60+ 30.3 (8.81)
Suicides Female 5.51 (1.49)
Ages 15-19 3.09 (1.82)
Ages 20-29 4.13 (1.53)
Ages 30-39 5.79 (2.00)
Ages 40-49 7.35 (2.59)
Ages 50-59 6.97 (2.43)
Ages 60+ 4.89 (1.90)
Independent Variables
Medical Marijuana Law 0.111 (0.311)
Marijuana Decriminalization Law 0.327 (0.469)
Ln (State Income) 10.3 (0.153)
Unemployment Rate 5.43 (1.36)
Blood Alcohol Content 0.08 0.509 (0.491)
Zero Tolerance Law 0.721 (0.440)
Real Beer Tax (2000 dollars) 0.252 (0.214)
N 918
Table 2. Estimates of Relationship between Medical Marijuana Laws and the
Natural Log of the Suicide Rate
Suicides Total
Suicides Male
Suicides Female
Year/State FEs
-0.084*
(0.050)
[0.94]
-0.073*
(0.043)
[0.94]
-0.106
(0.070)
[0.82]
Year/State FEs and
Covariates
-0.072**
(0.036)
[0.94]
-0.065***
(0.031)
[0.94]
-0.080
(0.050)
[0.83]
Year/State FEs,
Covariates, and State
Trends
-0.049*
(0.025)
[0.96]
-0.047**
(0.021)
[0.96]
-0.060
(0.042)
[0.87]
Observations 918 918 918
***Statistically significant at the 1% level; **at the 5% level; *at the 1% level.
Each cell represents the results from a separate regression. The dependent variable is equal to the
natural log of the suicide rate per 100,000 population, and is based on information collected from
the Centers for Disease Control for the period 1990-2007; the covariates are listed in Table 1.
Regressions are weighted using the relevant state-by-sex populations. Standard errors, corrected
for clustering at the state level, are in parentheses and R2s are in brackets.
Table 3. Estimates of Relationship between Medical Marijuana Laws and the
Natural Log of the Male Suicide Rate by Age Group
15-19
yr.-olds
20-29
yr.-olds
30-39
yr.-olds
40-49
yr.-olds
50-59
yr.-olds
60+
yr.-olds
Year/State FEs
-0.077
(0.077)
[0.72]
-0.096***
(0.025)
[0.85]
-0.147***
(0.048)
[0.82]
-0.129
(0.083)
[0.78]
-0.030
(0.062)
[0.72]
-0.008
(0.037)
[0.89]
Year/State FEs
and Covariates
-0.091
(0.080)
[0.72]
-0.096***
(0.022)
[0.85]
-0.129***
(0.032)
[0.83]
-0.113**
(0.054)
[0.80]
-0.010
(0.045)
[0.73]
-0.010
(0.034)
[0.90]
Year/State FEs,
Covariates, and
State Trends
-0.118
(0.114)
[0.75]
-0.114***
(0.077)
[0.87]
-0.099***
(0.037)
[0.86]
-0.060
(0.043)
[0.84]
-0.015
(0.041)
[0.77]
0.036
(0.028)
[0.91]
Observations 918 918 918 918 918 918
***Statistically significant at the 1% level; **at the 5% level; *at the 1% level.
Each cell represents the results from a separate regression. The dependent variable is equal to the natural log
of the suicide rate per 100,000 population, and is based on information collected from the Centers for Disease
Control for the period 1990-2007; the covariates are listed in Table 1. Regressions are weighted using the
relevant state-by-age populations. Standard errors, corrected for clustering at the state level, are in parentheses
and R2s are in brackets.
Table 4. Estimates of Relationship between Medical Marijuana Laws and the
Natural Log of the Female Suicide Rate by Age Group
15-19
yr.-olds
20-29
yr.-olds
30-39
yr.-olds
40-49
yr.-olds
50-59
yr.-olds
60+
yr.-olds
Year/State FEs
-0.069
(0.159)
[0.29]
-0.062
(0.061)
[0.40]
-0.125**
(0.059)
[0.53]
-0.120
(0.106)
[0.64]
-0.068
(0.103)
[0.50]
-0.082
(0.062)
[0.63]
Year/State FEs
and Covariates
-0.105
(0.165)
[0.30]
-0.044
(0.065)
[0.40]
-0.110**
(0.054)
[0.53]
-0.078
(0.066)
[0.65]
-0.019
(0.072)
[0.51]
-0.082
(0.057)
[0.63]
Year/State FEs,
Covariates, and
State Trends
0.083
(0.235)
[0.34]
-0.008
(0.065)
[0.45]
-0.035
(0.088)
[0.59]
-0.041
(0.055)
[0.71]
-0.104*
(0.060)
[0.55]
-0.121**
(0.059)
[0.65]
Observations 918 918 918 918 918 918
***Statistically significant at the 1% level; **at the 5% level; *at the 1% level.
Each cell represents the results from a separate regression. The dependent variable is equal to the natural log of
the suicide rate per 100,000 population, and is based on information collected from the Centers for Disease
Control for the period 1990-2007; the covariates are listed in Table 1. Regressions are weighted using the
relevant state-by-age populations. Standard errors, corrected for clustering at the state level, are in parentheses
and R2s are in brackets.
Table 5. Estimates of Relationship between Medical Marijuana Laws and the
Natural Log of the Suicide Rate: Adding Leads and Lags to the Model
Suicides Total
Suicides Male
Suicides Female
2-3 years before MML
-0.020
(0.015)
-0.006
(0.018)
-0.066*
(0.033)
1-2 years before MML
-0.023
(0.021)
-0.002
(0.010)
-0.106***
(0.036)
0-1 year before MML
-0.045**
(0.017)
-0.018
(0.019)
-0.140**
(0.053)
Year of law change
-0.085***
(0.029)
-0.066**
(0.022)
-0.154**
(0.058)
0-1 year after MML
-0.063*
(0.034)
-0.043
(0.027)
-0.138*
(0.070)
1-2 years after MML
-0.098***
(0.029)
-0.080***
(0.021)
-0.160**
(0.062)
2-3 years after MML
-0.154***
(0.035)
-0.124***
(0.023)
-0.257***
(0.094)
3+ years after MML
-0.174***
(0.052)
-0.129***
(0.032)
-0.327**
(0.124)
Year FE yes yes yes
State FE yes yes yes
Covariates yes yes yes
State Trends yes yes yes
R2 0.97 0.96 0.87
Observations 918 918 918
***Statistically significant at the 1% level; **at the 5% level; *at the 1% level.
Each cell represents the results from a separate regression. The dependent variable is equal to the
natural log of the suicide rate per 100,000 population, and is based on information collected from
the Centers for Disease Control for the period 1990-2007; the covariates are listed in Table 1.
Regressions are weighted using the relevant state-by-sex populations. Standard errors, corrected
for clustering at the state level, are in parentheses and R2s are in brackets.
Table 6: The Average Effect of Randomly Assigned MMLs on Suicide Rates
Suicides Total Suicides Male Suicides Female
Average MML estimate -0.002 -0.001 -0.009
(0.020) (0.017) (0.038)
Number of trials 50 50 50
Number of negative MML 30 29 31
coefficients
Number of MML estimates 4 3 4
that were negative and
significant at 5% level
Year FE Yes Yes Yes
State FE Yes Yes Yes
Covariates Yes Yes Yes
State Trends Yes Yes Yes
*Statistically significant at 10% level; ** at 5% level; *** at 1% level.
Notes: Each column represents the results from a series of regressions. Average standard errors, corrected
for clustering at the state-level, are in parentheses.
Appendix Table 1. Medical Marijuana Laws, 1990-2007
Effective date
Alaska March 4, 1999
California November 6, 1996
Colorado June 1, 2001
Hawaii December 28, 2000
Maine December 22, 1999
Montana November 2, 2004
Nevada October 1, 2001
New Mexico July 1, 2007
Oregon December 3, 1998
Rhode Island January 3, 2006
Vermont July 1, 2004
Washington November 3, 1998
Notes: Arizona, D.C., Delaware, Michigan, and New Jersey have passed medical marijuana laws since
2007. Source: Eddy (2010).
Appendix Table 2. Alcohol Consumption and Medical Marijuana Laws: Summary of Results from Anderson et al. (2011)
(1) (2) (3) (4) (5) (6)
18-19 20-29 30-39 40-49 50-59 60 +
yr.-olds yr.-olds yr.-olds yr.-olds yr.-olds yr.-olds
Number of Drinks (Males)
MML -18.7% -21.6%** -2.53% -4.70% -3.74% -3.07%
Observations 852 852 852 852 852 852
Number of Drinks (Females)
MML 5.26% -21.0%** -10.3%* -12.6%** -5.43% -12.2%
Observations 855 855 855 855 855 855
Year FE Yes Yes Yes Yes Yes Yes
State FE Yes Yes Yes Yes Yes Yes
Covariates Yes Yes Yes Yes Yes Yes
State-specific trends Yes Yes Yes Yes Yes Yes
*Statistically significant at 10% level; **at 5% level; ***at 1% level.
Notes: Each column represents the results from a separate regression. The dependent variable is the mean number of drinks consumed in the past 30 days in
state s and year t, and is based on information collected from the Behavioral Risk Factor Surveillance System (BRFSS) for the period 1990-2009. Controls
include the state unemployment rate, per capita income, the relevant population size (e.g., the number of males or females in state s and year t), the state beer tax,
and indicators for marijuana decriminalization, BAC 0.08, administrative license revocation, and Zero Tolerance. Regressions are weighted using sample sizes
from the BRFSS. Standard errors, corrected for clustering at the state level, are in parentheses.
Appendix Table 3. Alcohol Sales and Medical Marijuana Laws: Summary of
Results from Anderson et al. (2011)
(1) (2) (3)
State-level per State-level per State-level per
capita beer sales capita wine sales capita spirits sales
(1990-2009) (1994-2009) (1994-2009)
MML -5.3%*** -1.2% -0.2%
Observations 1020 816 816
Year FE Yes Yes Yes
State FE Yes Yes Yes
State covariates Yes Yes Yes
State-specific trends Yes Yes Yes
*Statistically significant at 10% level; **at 5% level; ***at 1% level.
Notes: Each column represents the results from a separate regression. The dependent variable is equal to
the natural log of per capita sales (measured in gallons) and is based on data from the Brewers Almanac,
published by the Beer Institute. Controls include the state unemployment rate, per capita income, the state
beer tax, and indicators for marijuana decriminalization, BAC 0.08, administrative license revocation, and
Zero Tolerance. Regressions are weighted using state populations. Standard errors, corrected for
clustering at the state level, are in parentheses.