Revolutionary High? Exposure to State Violence & Drug Use After

Revolutionary High? Exposure to State Violence & Drug Use after Revolution Alexandra Domike Blackman* Sarah Kammourh† Elizabeth R. Nugent‡

June 17, 2021 Abstract Does exposure to state violence during revolution increase drug use? We leverage a unique panel study, the Population Council’s Survey of Young People in Egypt (SYPE), to test whether rates of drug use are higher among those who witnessed violence during the 2011 revolution. We nd that exposure to state violence signicantly increases drug use; respondents exposed to violence are twice as likely to use drugs than those who were not, an eect size similar to important predictors such as losing employment, experiencing depression, or becoming less religious. Results are robust to specications that control for respondents’ reported exposure to state violence prior to the revolution and drug use among family and friends, factors identied in medical research as key predictors of drug use. Our study sheds light on the downstream eects of state violence and its implications for public health outcomes and drug policy preferences under repressive regimes.

*Assistant Professor, Department of Government, Cornell University. Contact: [email protected]. †Yale College, Class of 2022, Yale University. Contact: [email protected]. ‡Assistant Professor, Department of Political Science, Yale University. Contact: [email protected]. Introduction

The use and abuse of drugs is a global epidemic but disproportionately aects developing coun- tries due to high rates of economic precarity and inequality, rapid urbanization, and overburdened medical systems (United Nations’ 2020 World Drug Report). Egypt is a developing country with particularly high levels of drug use and abuse. In 2018, the Egyptian Ministry of Social Solidarity reported that the drug abuse rate reached 10.4 percent, nearly double that year’s global average of 5.6 percent (United Nations Oce on Drugs and Crime 2018; Menawy 2018). The widespread use of tramadol, a synethetic opioid, and various cannabis products tops public health issues; addiction to these drugs accounts for the vast majority of admissions to addiction treatment centers and calls to national drug addiction hotlines (Egypt Independent 2020). While rates of drug use in Egypt have steadily risen since the 2000s, recent spikes have blamed on the 2011 revolutionary uprising and its chaotic aftermath. In 2012, the managing director of the addiction unit at Egypt’s Ministry of Health said, “there should be little doubt that drug abuse has risen signicantly since 25 January,” shorthand for the 2011 revolution (Viney 2012). Similarly, the director of Egypt’s Nation Fund for Drug Control and Addiction Treatment claimed the revolution created a 300% increase in drugs rehabilitation center admissions in 2013 (Reuters 2013). These claims are vague on exactly how the revolution increased drug use and abuse. In this paper, we seek to understand what, if any, relationship exists between revolution and drug use. We draw on existing medical understandings of drug use to focus on the eect of one specic aspect of revolution: exposure to state violence. During the revolution and its aftermath, the Egyptian state responded to protest mobilization and opposition in all forms with brute force. A July 2013 counterrevolution included a military coup, renewed deadly violence against protesters, and massive waves of arrests. Between the revolution and the counterrevolution, a large number of Egyptians were exposed to state violence, many for the rst time. We argue that increased exposure to state violence during revolution increases drug use as people self-medicate to cope with this traumatic experience. We leverage a unique panel study of young Egyptians to test whether exposure to violence during the revolution increased drug use. The Population Council’s Survey of YoungPeople in Egypt (SYPE)

1 interviewed 10,916 young adults in 2009 and again in 2013 and 2014. Both waves of the survey asked several questions about the respondent’s personal drug use and drug use by family and friends. The 2014 instrument also included a battery capturing the respondent’s experience of the 2011 revolution and asked whether they witnessed violence during that time. The panel structure of the data permits us to control for the individual and social factors that might increase an individual’s propensity to use drugs while iso- lating the eect of exposure to violence during the revolution. We nd that exposure to state violence during the 2011 revolution signicantly increased drug use; respondents exposed to violence are 1 to 2 percentage points more likely to subsequently report using drugs than those not exposed to violence. Substantively, the size of this eect is akin to the eect of losing employment, experiencing depression, or becoming less religious, major life events that are found to increase drug use in medical studies. Our results are robust to specications that control for respondents’ reported exposure to state violence prior to the revolution and drug use among friends and family. The paper proceeds as follows. First, we conceptualize revolution and counterrevolution as an experience that exposes the average citizen to state violence, often for the rst time. Second, we provide a brief overview of the 2011 revolution and the history of drug use in Egypt. We then turn to our research design, data, and results. Next, we discuss the implications of our results for public health outcomes and drug policy preferences in repressive political contexts, as well as how these downstream eects may create two unexpected avenues through which state violence can destabilize repressive regimes. In exploring drug policy preferences, we present an original online survey conducted in Egypt in March 2020, in which we nd that respondents who have used drugs or know others who do are more likely to support the depenalization of drug use in Egypt. We conclude by outlining future research.

Revolution and State Violence

A political revolution is “any and all instances in which a state or political regime is overthrown and thereby transformed by a popular movement in an irregular, extra-constitutional, and/orviolent fash- ion” and necessitates “the mobilization of large numbers of people against the existing state” (Goodwin

2 2001, 11).1 Existing studies of revolutionary success tend to cluster into two groups. Procedural analyses focus on the processes of revolutions – the events, decisions, grievances, and mobilization that create a revolutionary situation (Gurr 1970; Kuran 1991; Tarrow 2011). In contrast, structural analyses pay attention to a revolution’s social, economic, and institutional predecessors, such as the balance of power and the level and nature of socioeconomic development, to explain its outcome (Dix 1984). Scholars have struggled to stipulate a singular measurement of political revolutionary success (Bosi and Uba 2009), simultaneously pointing to a host of policy changes, the replacement of leadership, and shifts in the democratic nature of government. It is much clearer when a revolutionary situation is defeated, its gains reversed and the ancien regime restored or resurrected (Clarke 2021).2 These denitions focus on revolution at the institutional level. Instead, we conceptualize revolution and its aftermath as an experience, with an eye towards how the average citizen might encounter these dramatic events. For participants, a revolution is a whirlwind of emotions. Initially, a revolution is a joyful, even euphoric moment of deep connection with other protesters and innite possibility (Arm- brust 2019). Its defeat, then, creates feelings of hopelessness and inecacy, and increases distrust in others for those invested in its success (Nugent and Homan 2021). However, for the average citizen, a revolution and its defeat are experiences characterized by chaos and increased exposure to state violence. Revolutions are often initiated through nonviolent protests but rarely remain free of violence. Regimes defend themselves by exploiting their monopoly (or at least disproportional ownership) of state violence to repress their challengers. This tends to create interactive escalation, wherein protesters mobilize, states repress protesters in an attempt to demobilize them, and subsequent protests grow by mobilizing both the original targets as well as drawing in additional supporters, in turn increasing the level of force re- quired to tame the masses (Lichbach 1987; Davenport 2007). We hypothesize that it is through exposure to state violence that revolution increases drug use.

1 See also Brinton (1938) and Gross (1958). Mass mobilization stands in contrast to top-down political changes, such as coups d’etat or pacted transitions. 2 How and why the Arab Spring uprisings did not achieve lasting change is beyond the scope of this paper but has been discussed in detail elsewhere (Brownlee, Masoud, and Reynolds 2015; Nugent 2020; Abadeer et al. 2021).

3 Medical and psychological studies nd that exposure to violence correlates with an increase in drug use (Sullivan, Kung, and Farrell 2004; Lofving-Gupta¨ et al. 2018). Violence is broadly dened, encompass- ing violence experienced at school, in the neighborhood, or in the workplace; domestic and parental violence; and mass shootings (North, Smith, and Spitznagel 1997; Sullivan, Kung, and Farrell 2004; Menard, Covey, and Franzese 2015; Lofving-Gupta¨ et al. 2018). In line with these ndings, we expect that individuals exposed to state violence during the 2011 Egyptian revolution will be more likely to report subsequent drug use.

A Brief History of the 2011 Revolution and Drug Use in Egypt

Egypt’s 2011 revolution occurred in conjunction with a wave of uprisings in the Middle East and North Africa known as the Arab Spring. January 25, 2011 marked National Police Day, a government holiday commemorating the role of the police in maintaining Egypt’s security and stability. Activists re- claimed Police Day to protest against police brutality and related infringements on civil liberties, freedom of speech, and economic justice. The rst demonstration saw tens of thousands gathering in Egyptian cities, followed by 18 days of sustained protest around the country by people from all ages, classes, walks of life, and political persuasions (Beissinger, Jamal, and Mazur 2015). On February 11, president Hosni Mubarak stepped down after nearly 30 years in oce. While Mubarak’s removal was swift, it was not without violence. In 2011, nearly 900 protesters were killed in Cairo alone, according to a leaked government report (Kingsley and Doss 2013). February 2 marked the “Battle of the Camel” after pro-Mubarak thugs riding camels and horses attacked unarmed protesters in Tahrir Square, leaving 11 dead and 600 injured. Egyptian security forces killed 24 and injured 212 while forcibly clearing a sit-in outside of Cairo’s Maspero television building in October 2011. In November 2011, the state’s Central Security Forces responded to commemorations of those who died during the revolution on Mohamed Mahmoud with excessive force, snipers and tear gas, killing 51 (Arm- brust 2019). The counterrevolution was similarly violent (Clarke 2021). In June 2012, Mohamed Morsi, a

4 former member of the Muslim Brotherhood’s Guidance Bureau and a parliamentarian under Mubarak, defeated former Prime Minister and Air Force commander Ahmed Shaq with 51.7 percent of the vote in a two-man run-o, and was inaugurated as Egypt’s rst democratically-elected president (Abadeer, Blackman, and Williamson 2018). Polarization between the Muslim Brotherhood and rival political elites quickly resurfaced and deepened (Nugent 2020). On the one-year anniversary of Morsi’s inauguration, the youth group Tamarod (Arabic for “Rebellion”) organized massive anti-Morsi protests with nancial and logistical support from elements of Egypt’s security apparatus. On July 3, 2013, the military an- nounced Morsi’s removal from power, suspended the constitution, and installed an interim government. The coup brought current Egyptian president, retired Field Marshall Abdel Fattah El-Sisi, to power. Under Sisi, Egypt has witnessed state violence on a scale that dwarfs the abuses of his predecessors. Between the July 2013 coup and May 2014, the number of individuals arrested and prosecuted reached 41,153, the majority accused of ties with the now outlawed Muslim Brotherhood.3 In September 2019, small protests broke out in several Egyptian cities in daring opposition against the corruption of the Sisi regime. The regime arrested nearly 4,000 people in response over the following week. In 2019, the total number of political prisoners reached 60,000 and likely tops 100,000 at the time of writing. Under Sisi, at least 22 new prisons have been constructed to house Egypt’s rapidly growing prisoner population (Stork 2019). The revolution and counterrevolution coincided with a dramatic increase in drug use and abuse. Drugs have a long and complex history in Egypt. The use of cannabis products, particularly hashish, has been documented in Egypt for several hundred years and was largely tolerated until the 20th century (Nahas 1985; Robins 2018). The “white drug” epidemic followed the 1919 revolution against the British occupation of Egypt and the Sudan. At that time, drugs such as heroin, cocaine, and opium were smug- gled into Egypt and used at high rates (Nahas 1985). During the 1920s, the Egyptian government passed a series of laws aimed at curtailing drug use. As a result, it is estimated that drug users made up nearly 30

3 As detailed by independent statistical database Wiki Thawra in the “Report on those Detained and Prosecuted during the Sisi/Adly Mansour Period.”

5 percent of the prison population in the 1920s. Following the end of World War II until the 1980s, there was a signicant decline in drug use in Egypt (Nahas 1985; Cowell 1989). With the rise of the “War on Drugs,” the United States exerted pressure on Presidents Sadat and Mubarak to increase eorts to combat the drug trade in Egypt, with only limited success (Robins 2018). Prior to the 2011 revolution, observers noted the increased use of specic cannabis products and a synthetic opioid, tramadol (Robins 2018). Since the early 2000s, the abuse of tramadol has became a major public health issue in Egypt. The crisis has only worsened since 2011. Egypt’s National Fund for Drug Control and Addiction Treatment reported a dramatic rise in patients seeking out drug use rehabilitation, from 12,000 patients per year before the revolution to 40,000 after (Reuters 2013). In 2010, the Anti-Narcotics General Administration (ANGA) seized over 132 million tablets of tramadol (Robins 2018). The number of seized tablets rose to 435 million in 2012, leveling out at high pre-revolutionary rates of approximately 100 million tablets seized in following years (United Nations Oce on Drugs and Crime 2017). Among a 2018 sample of university students, 12.3 percent reported using tramadol, and 60 percent of these tramadol users had drug-related problems (Bassiony et al. 2018).

Research Design and Data

To explore the relationship between exposure to state violence and drug use, we analyze the Population Council’s Survey of Young People (SYPE). SYPE rst surveyed a nationally representative sample of nearly 15,000 Egyptians between the ages of 10 and 29 in 2009. In 2013 and 2014, SYPE re- interviewed 10,916 respondents.4 We highlight two important aspects of the SYPE that are helpful for our analyses here. First, instruments for both waves included extensive and comparable questions about individual drug use and use by family and friends. This allows us to compare increases in usage while controlling for a variety of individual, social, and community factors that might increase an individual’s propensity to use drugs. Second, the 2014 instrument included a battery about experiences of the 2011

4 Further information about SYPE is available at https://www.popcouncil.org/research/ survey-of-young-people-in-egypt-2009-and-2014-datasets.

6 revolution. This allows us to examine whether similar individuals with diering exposure to violence during the revolution report dierent rates of subsequent drug use.

Dependent Variable

Our main dependent variable is drug use, measured by responses to the question, “have you experimented with any drugs before?” Just over 1 percent of respondents in the rst wave reported ex- perimenting with illegal drugs in wave 1 of the survey, while roughly 2 percent responded armatively in wave 2. The vast majority of those who use drugs report using tramadol or cannabis products.

Independent Variable

Our independent variable of interest is exposure to state violence during the 2011 revolution. Specically, the SYPE asks respondents to “think about the period since the start of the January 25th revolution” and then respond whether they saw a person being injured or killed. Since this is only measured in the second wave of the survey, we create a 2009 baseline measure for exposure to violence using the question: “Have you ever witnessed police violence?” While the questions dier slightly in phrasing, we believe they measure exposure to police brutality and state violence in a consistent manner.5

Control Variables

We include a number of controls in our analyses, motivated by the medical literature on drug use. Table A.1 in the Appendix reports the summary statistics for all variables. We control for family drug use, which is a signicant predictor of future use.6 Individual mental health issues related to anxiety

5 While beyond the scope of this paper, we note that exposure to state violence during revolution can be direct or indirect. Individuals may personally witness or experience state violence, directed at themselves or others (LeBas and Young 2020). Individuals may also indirectly experience violence by seeing it or hearing about it on social media or on television. More direct (i.e., personal) exposure to violence tends to have a stronger eect on a variety of political, social, and economic outcomes (Zimmerman and Posick 2016). In additional analyses reported in the article’s appendix, we conrm that direct exposure to state violence matters for the ndings presented later in the piece; witnessing violence second-hand, on television or through social media, does not have a consistent eect on drug use. 6 There is a debate within the medical eld whether the eect of family constitutes a genetic predisposition or environmen- tal inuence. Ahmed et al. (2020) nd that genetics explained up to two-thirds of the variance in vulnerability to cocaine addiction.

7 and depression are highly correlated with drug use, particularly when these conditions are untreated (Ed- lund et al. 2015; Lai et al. 2015) Recent studies conducted with Egyptian populations conrm this robust nding. Researchers from Assiut University found that 72 percent of those with substance use disorder had severe depression compared to 6 percent of the control group (Mohamed et al. 2020). Accordingly, we control for depression among respondents using a module in the SYPE on mental health and social development.7 Weinclude demographic controls for sex, employment status, education, socioeconomic status, and age, which are identied in medical studies as being important predictors of drug use (Biederman et al. 2000; Swendsen and Merikangas 2000; Galea and Vlahov 2002; Hosseinbor et al. 2014; Lai et al. 2015; Avenevoli et al. 2015; Robins 2018).8 Studies of drug use in Egypt suggest the highest rates among young adults, particularly males (El-Sawy, Abdel Hay, and Badawy 2010; Bassiony et al. 2018; Rabie et al. 2020). Finally, we also control for religiosity, which research suggests is negatively correlated with drug use (Chit- wood, Weiss, and Leukefeld 2008). In addition to demographic covariates, an individual’s relationship with the surrounding community is also important in predicting drug use. Those who feel socially dis- connected or are embedded in communities of other users are more likely to use drugs (Bond et al. 2007; Branstetter, Low, and Furman 2011). We include a measurement of generalized trust and the respondents’ reported number of friends to measure social connectedness. To measure connections to other drug users, we control for reported drug use by close friends.

Empirical Strategy

We conduct two sets of analyses to examine the eect of exposure to state violence on drug use. First, we conduct an analysis that highlights the cross-sectional nature of the SYPE data. This permits us to control for variables that may not change between the two waves of the panel data identied in existing studies as potentially important correlates of substance use. We measure the eect of witnessing violence

7 We do this following the diagnosis criteria set out in the Diagnostic and Statistical Manual of Mental Disorders (DSM–5). Respondents are coded as showing signs of depression when they agree with ve or more of nine questions that measure dierent potential symptoms of depression (American Psychiatric Association 2020). 8 Time-varying variables are all included in the panel analysis, including employment status, education, socio-economic status.

8 during the 2011 revolution on reported drug use in the second wave of the survey while controlling for the individual and community characteristics outlined above, measured in the 2009 wave of the survey. Second, we exploit the panel structure of the data, which permits us to account for unobservable bias that confound causal inferences in cross-sectional analyses. We also include controls for variables with values that change between 2009 and 2014.

Results

Our results demonstrate that exposure to state violence during the 2011 revolution led to a subsequent increase in reported drug use. First, in the cross-sectional analysis (Table 1), we nd that those respondents who reported witnessing violence in person during the 2011 revolution are 1.1 percentage points more likely to report that they use drugs in 2014 than those who did not. This result is robust to a model specication including individual demographic and socioeconomic controls, measures of drug use among the respondents’ friends and family, and respondents’ reported use in 2009.

9 Table 1: Cross-Sectional Analysis: Exposure to Violence and Drug Use

Dependent variable: Probability of Reported Use (2014) Drug Use Drug Use (1) (2) Witnessed Violence (2014) 0.011∗∗ 0.013∗∗ (0.005) (0.006)

Drug Use (2009) 0.046 (0.032)

Friend Drug Use (2009) 0.023∗ (0.013)

Family Drug Use (2009) 0.003 (0.014)

Witnessed Violence (2009) 0.004 (0.012)

Constant 0.018∗ 0.076∗∗∗ (0.009) (0.021)

Sex/Age/Religion Controls Y Y Additional 2009 Controls N Y Governorate FE Y Y Observations 10,864 7,659 R2 0.035 0.051

Note: ∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01. OLS models with standard errors clustered by sampling unit. All models control for respondent sex, age, and religion. Model 2 also controls for 2009 measures of: employment, education, SES, depression, generalized trust, number of friends, and religiosity.

We nd similar results in the panel analysis (Table 2). Respondents who reported exposure to state violence in 2011 were between 1 and 2 percentage points more likely to report using drugs in 2014, compared to those who were not exposed to violence. The size of this eect is similar to the that of becoming unemployed, experiencing depression, or becoming less religious, major life events identied in medical studies as contributing to an increase in drug use. Given a baseline usage rate of 1 percent, this change represents a doubling (100 percent increase) in reported drug use. We note that consistent with medical studies, whether a respondent’s family or friends use drugs has the largest eect on drug use. However, above and beyond the eects of other important predictors, exposure to violence during the revolution remains a highly signicant and substantively important variable.

10 Table 2: Panel Analysis: Exposure to Violence and Drug Use

Dependent variable: Probability of Reported Use Drug Use Drug Use (1) (2) Witnessed Violence 0.021∗∗∗ 0.011∗∗ (0.006) (0.005)

Friend Drug Use 0.132∗∗∗ (0.012)

Family Drug Use 0.221∗∗∗ (0.020)

Employed 0.003 (0.005)

Unemployed 0.010∗ (0.005)

Attended School (Completed below secondary) 0.005 (0.007)

Attended School (Completed secondary+) 0.003 (0.008)

Wealth Quintile −0.001 (0.001)

Depression 0.011∗ (0.006)

Religiosity −0.010∗ (0.006)

Generalized Trust 0.008∗ (0.004)

Number of Friends 0.0003 (0.0003)

Observations 15,478 15,104 R2 0.003 0.223

Note: ∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01 OLS models with standard errors clustered by respondent. Model 2 also controls for time-varying measures of: friends’ and family members’ usage, employment, education, SES, depression, religiosity, generalized trust, and number of friends. For employment status, Out of Labor Force is the reference category, and, for education, No Formal Education is the reference category.

11 State Violence, Public Health Outcomes & Drug Policy Prefer- ences

Increased drug use resulting from exposure to state violence during revolution has two important implications for politics in repressive contexts. First, our ndings demonstrate that state violence is bad for public health; exposure has a signicant negative eect on public health outcomes as it increases drug abuse. Rising drug use in Egypt is creating a new crisis for the Sisi regime. As we noted in our introduction, nearly one-tenth of the Egyptian population – close to 10 million people – abuses drugs. If our results are correct, in that exposure to violence increases drug use, then the current regime may need to rethink its current governing strategy. At the core of authoritarian retrenchment in Egypt following the 2011 revolution is a massive and widespread increase in the state’s use of repression against both organized opposition and regular citizens. Since the regime came to power in 2013, over 100,000 people have been imprisoned, the vast majority on politically-motivated charges. Due to the state of the penal system in Egypt, the process of being arrested, held, and jailed involves signicant exposure to state violence and potential post-traumatic stress syndromes both for direct victims, their networks of friends and family, and other Egyptians who witness or learn about these experiences. In addition, prisons are spaces notori- ous for drug use among inmates even under the best conditions, and the appalling environment created in Egyptian prisons likely further promotes drug abuse. While it is unlikely that the Egyptian government will meaningfully democratize in the near future, it is possible for the non-democratic regime like the current one in Egypt to address certain aspects of the current crisis. For example, scaling back the level of police brutality, undoubtedly condoned by the state, in turn decreasing the population’s exposure to state violence, might signicantly reduce drug abuse rates. Reforming the prison sentencing process, especially for those sentenced on low-level drug charges, and improving living conditions for prisoners may have a similarly positive eect on the drug crisis. However, the Sisi regime has taken a dierent approach to drugs in recent years. While the government continues to impose harsh sentences on some drug users, there is an increased focus on

12 treating addiction once it has occurred (Menawy 2018; Middle East Monitor 2019). In March 2018, the Egyptian Drug Control and the Treatment of Addiction started a campaign titled “Anta Aqwa Min Al- Mokhadarat” (“You Are Stronger than Drugs”). Headed by the popular soccer player Mohamed Salah, the campaign sought to encourage people to seek help for drug addiction. The Egyptian Ministry of So- cial Solidarity, which oversaw the campaign, reported that the national drug rehabilitation hotline had a 400 percent increase in calls in the rst three weeks of the campaign (Middle East Monitor 2018; Menawy 2018). Such anti-drug campaigns are a positive development, but they are unlikely to be eective without addressing the systemic issues underpinning increased drug abuse. Second, the growth of the drug crisis, by exposing more of the population to its ills, may potentially change public opinion and create demand for new government policies. Building on existing research that nds that exposure to drug use shapes policy preferences (De Benedictis-Kessner and Han- kinson 2019), we conducted a survey of 1,500 Egyptians focused on attitudes towards penalization of drug use and government funding of drug treatment facilities. We also asked respondents about their level of exposure to recreational drug use, both personal and through friends and family.9 Results demonstrate that exposure to drug use is associated with reduced support for punitive governmental drug policies. In- dividuals who report cannabis and tramadol use themselves or by friends and family are more supportive of drug depenalization policies and of increased funding for drug treatment. Columns 1 and 2 of Table 3 demonstrate that those who are exposed to the recreational use of either cannabis or tramadol are over 12 percentage points more likely to support depenalization.10 This result holds when we control respondent gender, age, religion, education level, and employment status.11 Columns 3 and 4 of Table 3 demonstrate that those with drug use exposure are over 5 percentage points more likely to support increased state funding for drug treatment programs. Both downstream eects create unexpected avenues through which state violence may actually

9 See Supplementary Appendix for question wording. 10 For this analysis, we use a linear probability model that takes 1 if the respondent supports the full decriminalization of cannabis possession or the depenalization of the rst two oenses or if the respondent supports increasing funding. 11 Figures A.1 and A.2 in the Supplementary Appendix display the distribution of policy preferences graphically.

13 destabilize repressive regimes rather than shore them up. First, contributing to an increase in drug abuse, even inadvertently, is bad public health policy and creates a new problem the regime must then address. Moreover, if exposure to state violence creates more drug users, and more drug users increases public exposure to the drug crisis, regimes may contribute to demands for drug policy reforms. With these demands left unmet, the state may create new popular grievances.

Table 3: Qualtrics Survey: Exposure to Drug Use and Policy Attitudes

Dependent variable: Probability of Support for Policy Drug Depenalization Drug Depenalization Drug Treatment Drug Treatment (1) (2) (3) (4) Exposure to Drug Use 0.122∗∗∗ 0.158∗∗∗ 0.059∗∗ 0.077∗∗∗ (0.023) (0.026) (0.026) (0.028)

Employed −0.029 0.013 (0.027) (0.029)

Unemployed 0.003 −0.068 (0.045) (0.048)

Secondary+ −0.050 0.221∗∗∗ (0.053) (0.057)

Female −0.084∗∗∗ −0.064∗∗∗ 0.031 0.034 (0.022) (0.024) (0.024) (0.026)

Age −0.003∗∗∗ −0.003∗∗∗ −0.001 −0.001 (0.001) (0.001) (0.001) (0.001)

Muslim −0.112∗∗ −0.100∗∗ −0.023 −0.011 (0.046) (0.049) (0.050) (0.052)

Constant 0.422∗∗∗ 0.456∗∗∗ 0.782∗∗∗ 0.575∗∗∗ (0.052) (0.074) (0.057) (0.079)

Observations 1,433 1,211 1,426 1,206 R2 0.049 0.056 0.005 0.025

Note: ∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01. OLS models. All models control for respondent sex, age, and religion. Models 2 and 4 also control for employment and education.

14 Conclusion

In this paper, we demonstrate that exposure to state violence during the 2011 revolutionary uprising in Egypt resulted in higher reported drug use among a panel study of 10,916 young adults. The size of this eect was 1-2 percentage point increase, an eect size similar to a number of variables identied as particularly important for shifting an individual’s drug use, such as losing employment, experiencing depression, or becoming less religious. Increased drug use, as caused by increased exposure to state violence, not only creates a major public health issue that requires state resources for drug addiction treatment, it also may increase demand for drug policy reforms by the public. Additional survey evidence from a sample of Egyptians in March 2020 demonstrates that people who have used drugs or know others who do are more likely to support the depenalization of drug use. Our ndings speak to the long and varied nature of the legacies of revolution and counterrevolution, and we hope our work will inspire others to expand the breadth of the outcomes we analyze when thinking about political violence. We focus here on the eects of revolution on drug use, with implications for public health at the national level and mental and physical health at the individual level. We encourage other scholars of politics to theorize and analyze the public health eects and implications of major political events, particularly violent ones such as revolutions and counterrevolution. This research is important not only for understanding how political change aects individuals and public health outcomes but also for grasping the lived experience of revolution and its aftermath, particularly that of the Arab Spring uprisings of 2010-2011. The uprisings largely failed to create meaningful democratization at the institutional level. In every case, protests were countered by signicant state violence as regimes defended themselves (Bellin 2012). Many regimes continue to use signicant repression to try and quell mobilized populations. Should our results from Egypt hold elsewhere in the region, state violence may be contributing to burgeoning drug crisis as populations seek to cope with the trauma of brutal regime responses to the Arab Spring.

15 References

Abadeer, Caroline, Alexandra Domike Blackman, Lisa Blaydes, and Scott Williamson. 2021. “Did Egypt’s Post-Uprising Crime Wave Increase Support for Authoritarian Rule?” Working Paper. Abadeer, Caroline, Alexandra Domike Blackman, and Scott Williamson. 2018. “Votingin Transition:Par- ticipation and Alienation in Egypt’s 2012 Presidential Election.” Middle East Law and Governance 10 (1): 25–58. Ahmed, Serge H., Aldo Badiani, Klaus A. Miczek, and Christian P. Muller.¨ 2020. “Non-pharmacological factors that determine drug use and addiction.” Neuroscience & Biobehavioral Reviews 110:3–27. American Psychiatric Association. 2020. What is Depression?, October. https://www.psychiatry.org/ patients-families/depression/what-is-depression. Armbrust, Walter. 2019. Martyrs and Tricksters: An ethnography of the Egyptian revolution. Princeton University Press. Avenevoli, Shelli, Joel Swendsen, Jian-Ping He, Marcy Burstein, and Kathleen Ries Merikangas. 2015. “Major depression in the national comorbidity survey–adolescent supplement: Prevalence, correlates, and treatment.” Journal of the American Academy of Child & Adolescent Psychiatry 54 (1): 37– 44. Bassiony, Medhat M., Mohamed M. Abdel-Ghani, Ghada M. Salah El-Deen, Mervat S. Hassan, Hayam El-Gohari, and Usama Yousef. 2018. “Opioid Use Disorders Attributed to Tramadol Among Egyp- tian University Students.” Journal of Addiction Medicine 12 (2): 150–155. Beissinger, Mark R., Amaney A. Jamal, and Kevin Mazur. 2015. “Explaining Divergent Revolutionary Coalitions: Regime strategies and the structuring of participation in the Tunisian and Egyptian revolutions.” Comparative Politics 48 (1): 1–24. Bellin, Eva. 2012. “Reconsidering the Robustness of Authoritarianism in the Middle East: Lessons from the Arab Spring.” Comparative Politics 44 (2): 127–149. Biederman, Joseph, Stephen V Faraone, Michael C Monuteaux, and Jennifer A Feighner. 2000. “Pat- terns of alcohol and drug use in adolescents can be predicted by parental substance use disorders.” Pediatrics 106 (4): 792–797. Bond, Lyndal, Helen Butler, Lyndal Thomas, John Carlin, Sara Glover, Glenn Bowes, and George Pat- ton. 2007. “Social and School Connectedness in Early Secondary School as Predictors of Late Teenage Substance Use, Mental Health, and Academic Outcomes.” Journal of Adolescent Health 40 (4): 357– e9. Bosi, Lorenzo, and Katrin Uba. 2009. “Introduction: The Outcomes of Social Movements.” Mobiliza- tion: An International Quarterly 14 (4): 409–415. Branstetter, Steven A, Sabina Low, and Wyndol Furman. 2011. “The Inuence of Parents and Friends on Adolescent Substance Use: A Multidimensional Approach.” Journal of Substance Use 16 (2): 150– 160.

16 Brinton, Crane. 1938. The Anatomy of Revolution. W. W. Norton & Company. Brownlee, Jason, TarekE. Masoud, and Andrew Reynolds. 2015. The Arab Spring: Pathways of Repression and Reform. Oxford University Press. Chitwood, Dale D., Michael L. Weiss, and Carl G. Leukefeld. 2008. “A Systematic Review of Recent Literature on Religiosity and Substance Use.” Journal of Drug Issues 38 (3): 653–688. Clarke, Killian. 2021. “Overthrowing Revolution: The Emergence and Success of Counterrevolution, 1900-2015.” Working Paper. Cowell, Alan. 1989. “A Poor Land Torn by a Rich Craving.” The New York Times (May). https://www. nytimes.com/1989/05/20/world/cairo-journal-a-poor-land-torn-by-a-rich-craving.html. Davenport, Christian. 2007. “State repression and political order.” Annual Review of Political Science 10:1–23. De Benedictis-Kessner, Justin, and Michael Hankinson. 2019. “Concentrated Burdens: How Self-Interest and Partisanship Shape Opinion on Opioid Treatment Policy.” American Political Science Review 113 (4): 1078–1084. Dix, Robert H. 1984. “Why Revolutions Succeed & Fail.” Polity 16 (3): 423–446. Edlund, Mark J., Valerie L. Forman-Homan, Cherie R. Winder, David C. Heller, Larry A. Kroutil, Rachel N. Lipari, and Lisa J. Colpe. 2015. “Opioid abuse and depression in adolescents: Results from the National Survey on Drug Use and Health.” Drug and Alcohol Dependence 152:131–138. Egypt Independent. 2020. “Over 45,000 drug addicts in Egypt nd help through hotline during January- May: Minister.” Egypt Independent (June). https://www.egyptindependent.com/over- 45000- drug-addicts-in-egypt-find-help-through-hotline-during-january-may-minister/. El-Sawy, Hosam, Mohammed Abdel Hay, and Adel Badawy. 2010. “Gender Dierences in Risks and Pattern of Drug Abuse in Egypt.” The Egyptian Journal of Neurology, Psychiatry and Neurosurgery 47 (1): 413–418. Galea, Sandro, and David Vlahov. 2002. “Social determinants and the health of drug users: socioeconomic status, homelessness, and incarceration.” Public Health Reports 117 (Suppl 1): S135. Goodwin, Je. 2001. No Other Way Out: States and Revolutionary Movements, 1945-1991. Cambridge Uni- versity Press. Gross, Feliks. 1958. The Anatomy of Revolution. Philosophical Library. Gurr, Ted R. 1970. Why Men Rebel. Princeton University Press. Hosseinbor, Mohsen, Seyed Mojtaba Yassini Ardekani, Saeed Bakhshani, and Somayeh Bakhshani. 2014. “Emotional and social loneliness in individuals with and without substance dependence disorder.” International Journal of High Risk Behaviors & Addiction 3 (3).

17 Kingsley, Patrick, and Leyla Doss. 2013. “Egyptian police ’killed almost 900 protesters in 2011 in Cairo’.” The Guardian (March). https://www.theguardian.com/world/2013/mar/14/egypt-leaked-report- blames-police-900-deaths-2011. Kuran, Timur. 1991. “Now Out of Never: The Element of Surprise in the East European Revolution of 1989.” World Politics 44 (1): 7–48. Lai, Harry Man Xiong, Michelle Cleary, Thiagarajan Sitharthan, and Glenn E. Hunt. 2015. “Prevalence of comorbid substance use, anxiety and mood disorders in epidemiological surveys, 1990–2014: A systematic review and meta-analysis.” Drug and Alcohol Dependence 154:1–13. LeBas, Adrienne, and Lauren Young. 2020. “Repression and Dissent around a Potential Critical Junc- ture: Panel Data Evidence from Zimbabwe.” Working Paper. Lichbach, Mark Irving. 1987. “Deterrence or escalation? The puzzle of aggregate studies of repression and dissent.” Journal of Conﬂict Resolution 31 (2): 266–297. Lofving-Gupta,¨ Sandra, Mimmie Willebrand, Roman Koposov, Marek Blatny,´ Michal Hrdlicka,ˇ Mary Schwab-Stone, and Vladislav Ruchkin. 2018. “Community violence exposure and substance use: cross-cultural and gender perspectives.” European Child & Adolescent Psychiatry 27 (4): 493–500. Menard, Scott, Herbert C Covey, and Robert J Franzese. 2015. “Adolescent Exposure to Violence and Adult Illicit Drug Use.” Child Abuse & Neglect 42:30–39. Menawy, Marwan. 2018. “Egypt ghts a losing battle against drugs.” Arab News (February). https : / / www.arabnews.com/node/1254306/middle-east. Middle East Monitor. 2018. “400% rise in calls to Egypt’s rehab hotline after Mohamed Salah anti-drugs campaign.” Middle East Monitor (April). https://www.middleeastmonitor.com/20180409-400- rise-in-calls-to-egypts-rehab-hotline-after-mohamed-salah-anti-drugs-campaign/. . 2019. “Egypt approves death penalty for drug dealers.” Middle East Monitor (January). https: //www.middleeastmonitor.com/20190131-egypt-approves-death-penalty-for-drug-dealers/. Mohamed, Ikram I, Hossam Eddin Khalifa Ahmad, Shehab H Hassaan, and Shymaa M Hassan. 2020. “Assessment of anxiety and depression among substance use disorder patients: a case-control study.” Middle East Current Psychiatry 27 (22): 1–8. Nahas, G G. 1985. “Hashish and drug abuse in Egypt during the 19th and 20th centuries.” Bulletin of the New York Academy of Medicine 61 (5): 428–444. North, Carol S, Elizabeth M Smith, and Edward L Spitznagel. 1997. “One-year Follow-up of Survivors of a Mass Shooting.” American journal of psychiatry 154 (12): 1696–1702. Nugent, Elizabeth R. 2020. After Repression: How Polarization Derails Democratic Transition. Vol. 24. Princeton University Press. Nugent, Elizabeth R., and Michael T. Homan. 2021. “Disappointed and Demobilized: Political Atti- tudes and Behavior after Defeated Revolution.” Working Paper.

18 Rabie, Menan, Nermin M Shaker, Eman Gaber, Mahmoud El-Habiby, Dalia Ismail, Maha El-Gaafary, Amina Lotfy, Noha Sabry, Wael Khafagy, and Richard Muscat. 2020. “Prevalence Updates of Sub- stance Use among Egyptian Adolescents.” Middle East current psychiatry 27 (1): 4. Reuters. 2013. “Video: Increase of drug users post Egyptian revolution.” Al-Arabiya (May). https : / / english.alarabiya.net/News/middle- east/2013/05/12/Video- Increase- of - drug - users- post- Egyptian-revolution. Robins, Philip. 2018. “Drugs of choice, drugs of change: Egyptian consumption habits since the 1920s.” Third World Quarterly 39 (2): 248–260. Stork, Joe. 2019. “Egyptian Prisons As Islamic State Recruitment Centers.” Lobe Log (February). https: //lobelog.com/egyptian-prisons-as-islamic-state-recruitment-centers/. Sullivan, Terri N., Eva M. Kung, and Albert D. Farrell. 2004. “Relation Between Witnessing Violence and Drug Use Initiation Among Rural Adolescents: Parental Monitoring and Family Support as Protective Factors.” Journal of Clinical Child & Adolescent Psychology 33 (3): 488–498. Swendsen, Joel D, and Kathleen R Merikangas. 2000. “The comorbidity of depression and substance use disorders.” Clinical Psychology Review 20 (2): 173–189. Tarrow, Sidney G. 2011. Power in Movement: Social Movements and Contentious Politics. Cambridge Uni- versity Press. United Nations Oce on Drugs and Crime. 2017. “Conference room paper submitted by the Arab Republic of Egypt on strengthening international cooperation in addressing the non-medical use and abuse, the illicit manufacture and the illicit domestic and international distribution of tramadol.” In Commission on Narcotic Drugs: Sixtieth session. United Nations publication, March. https://www.unodc.org/documents/commissions/CND/CND Sessions/CND 60/CRPs/ ECN72017 CRP4 V1701497.pdf. . 2018. World Drug Report 2018. United Nations publication, Sales No. E.18.XI.9, June. https: //www.unodc.org/wdr2018/. Viney, Steven. 2012. “As drug addiction rises in Cairo, experts oer recommendations.” Egypt Indepen- dent (May). https://egyptindependent.com/drug-addiction-rise-cairo-exacerbated-revolution- experts-offer-prescriptions/. Wiki Thawra. 2014. “Report on those Detained and Prosecuted during the Sisi/Adly Mansour Period.” Wiki Thawra (May). https://wikithawra.wordpress.com/2014/01/09/sisi-mansour-detainees/. Zimmerman, Gregory M, and Chad Posick. 2016. “Risk Factors for and Behavioral Consequences of Direct versus Indirect Exposure to Violence.” American Journal of Public Health 106 (1): 178–188.

19 A Supplementary Appendix: Revolutionary High? Exposure to Violence and Drug Use after Revolution

A.1 Summary Statistics Table A.1 displays the summary statistics for the SYPE panel and our original March 2020 survey.

Table A.1: Survey Summary Statistics

Variable N Mean St.Dev Min Max Survey of Young People in Egypt (SYPE) Witness Violence (2014) 10890 0.18 0.39 0 1 Witness Violence (2009) 10916 0.03 0.16 0 1 Drug Use (2014) 10903 0.02 0.13 0 1 Drug Use (2009) 7761 0.01 0.12 0 1 Family Drug Use (2014) 10903 0.04 0.19 0 1 Family Drug Use (2009) 7761 0.03 0.16 0 1 Friend Drug Use (2014) 10903 0.07 0.25 0 1 Friend Drug Use (2009) 10916 0.05 0.21 0 1 Female 10916 0.54 0.50 0 1 Muslim 10916 0.97 0.17 0 1 Age (2014) 10891 24.23 5.70 14 36 Age (2009) 10891 19.23 5.70 9 31 Secondary Education+ (2014) 10915 0.67 0.47 0 1 Secondary Education+ (2009) 10916 0.46 0.50 0 1 Employed (2014) 10912 0.30 0.46 0 1 Employed (2009) 10916 0.20 0.40 0 1 Unemployed (2014) 10782 0.08 0.26 0 1 Unemployed (2009) 10824 0.06 0.24 0 1 Wealth Quintile (2014) 10916 3.14 1.41 1 5 Wealth Quintile (2009) 10916 2.94 1.35 1 5 Depression (2014) 10916 0.04 0.20 0 1 Depression (2009) 10916 0.07 0.26 0 1 Religiosity (2014) 10889 1.98 0.35 1 3 Religiosity (2009) 7761 1.92 0.36 1 3 Generalized Trust (2014) 10888 0.22 0.41 0 1 Generalized Trust (2009) 7761 0.10 0.30 0 1 Saw Violence on TV (2014) 10890 0.98 0.12 0 1 Saw Violence on Social Media (2014) 10889 0.30 0.46 0 1 March 2020 Qualtrics Survey Drug Exposure 1500 0.26 0.44 0 1 Tramadol Exposure 1500 0.18 0.38 0 1 Cannabis Exposure 1500 0.19 0.39 0 1 Supports Drug Depenalization 1486 0.20 0.40 0 1 Supports Increased Drug Treatment 1476 0.75 0.43 0 1 Female 1500 0.43 0.50 0 1 Muslim 1500 0.94 0.23 0 1 Age 1446 35.12 13.13 18 88 Secondary Education+ 1267 0.94 0.24 0 1 Employed 1500 0.62 0.49 0 1 Unemployed 1500 0.09 0.29 0 1 Out of Labor Force 1500 0.29 0.45 0 1

1 A.2 Additional Analyses and Robustness Checks A.2.1 Dierent Types of Exposure to Violence

Table 2: Cross-Sectional Analysis: Exposure to Violence and Substance Use (Dierent Types of Expo- sure)

Dependent variable: Probability of Reported Use Drug Use Drug Use Alcohol Use Alcohol Use Tobacco Use Tobacco Use (1) (2) (3) (4) (5) (6) Witnessed Violence (2014) 0.013∗∗∗ 0.015∗∗ 0.021∗∗∗ 0.023∗∗∗ 0.032∗∗∗ 0.030∗∗ (0.005) (0.006) (0.007) (0.007) (0.010) (0.013)

Saw Violence on TV (2014) 0.010 0.020∗∗∗ 0.006 0.009 0.021 0.025 (0.006) (0.004) (0.010) (0.012) (0.021) (0.028)

Saw Violence on Social Media (2014) −0.007∗ −0.005 0.001 0.003 −0.030∗∗∗ −0.022∗∗ (0.004) (0.005) (0.004) (0.005) (0.008) (0.010)

Drug Use (2009) 0.045 (0.032)

Friend Drug Use (2009) 0.023∗ (0.013)

Family Drug Use (2009) 0.003 (0.014)

Alcohol Use (2009) 0.008 (0.026)

Friend Alcohol Use (2009) 0.021 (0.021)

Tobacco Use (2009) 0.235∗∗∗ (0.021)

Friend Tobacco Use (2009) 0.023∗∗ (0.010)

Family Tobacco Use (2009) −0.006 (0.008)

Witnessed Violence (2014) 0.004 −0.002 0.041 (0.012) (0.012) (0.028)

Constant 0.012 0.058∗∗∗ 0.050∗∗ 0.072∗∗ −0.004 0.161∗∗∗ (0.011) (0.022) (0.021) (0.032) (0.030) (0.055)

Sex/Age/Religion Controls Y Y Y Y Y Y Additional 2009 Controls N Y N Y N Y Governorate FE Y Y Y Y Y Y Observations 10,863 7,658 10,863 7,658 10,863 7,241 R2 0.035 0.051 0.025 0.031 0.206 0.293

Note: ∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01 OLS models with standard errors clustered by sampling unit. All models control for respondent sex, age, and religion. Models 2, 4, and 6 also control for 2009 measures of: employment, education, SES, depression, generalized trust, number of friends, and religiosity.

2 A.2.2 Alcohol and Tobacco Use

Table 3: Cross-Sectional Analysis: Exposure to Violence and Substance Use (Alcohol and Tobacco)

Dependent variable: Probability of Reported Use Probability of Reported Use Alcohol Use Alcohol Use Tobacco Use Tobacco Use (1) (2) (3) (4) Witnessed Violence (2014) 0.021∗∗∗ 0.023∗∗∗ 0.021∗∗ 0.022∗ (0.007) (0.007) (0.010) (0.012)

Alcohol Use (2009) 0.008 (0.026)

Friend Alcohol Use (2009) 0.021 (0.021)

Tobacco Use (2009) 0.235∗∗∗ (0.021)

Friend Tobacco Use (2009) 0.024∗∗ (0.010)

Family Tobacco Use (2009) −0.005 (0.008)

Witnessed Violence (2009) −0.002 0.041 (0.012) (0.028)

Constant 0.057∗∗∗ 0.083∗∗∗ −0.004 0.173∗∗∗ (0.017) (0.027) (0.021) (0.045)

Sex/Age/Religion Controls Y Y Y Y Additional 2009 Controls N Y N Y Governorate FE Y Y Y Y Observations 10,864 7,659 10,864 7,242 R2 0.025 0.031 0.205 0.292

Note: ∗p<0.1; ∗∗p<0.05; ∗∗∗p<0.01 OLS models with standard errors clustered by sampling unit. All models control for respondent sex, age, and religion. Models 2 and 4 also control for 2009 measures of: employment, education, SES, depression, generalized trust, number of friends, and religiosity.

3 Table 4: Panel Analysis: Exposure to Violence and Substance Use (Alcohol and Tobacco)

Dependent variable: Probability of Reported Use Probability of Reported Use Alcohol Use Alcohol Use Tobacco Use Tobacco Use (1) (2) (3) (4) Witnessed Violence 0.023∗∗∗ 0.018∗∗∗ 0.026∗∗ 0.026∗∗ (0.006) (0.006) (0.012) (0.012)

Friend Alcohol Use 0.277∗∗∗ (0.025)

Friend Tobacco Use 0.118∗∗∗ (0.009)

Family Tobacco Use 0.094∗∗∗ (0.007)

Employed −0.008 0.053∗∗∗ (0.005) (0.012)

Unemployed 0.009∗ 0.009 (0.005) (0.012)

Attended School (Completed below secondary) 0.0003 0.013 (0.008) (0.018)

Attended School (Completed secondary+) −0.001 0.004 (0.008) (0.020)

Wealth Quintile 0.001 −0.001 (0.001) (0.003)

Depression 0.020∗∗∗ 0.018∗ (0.007) (0.011)

Religiosity −0.008 −0.018∗ (0.005) (0.011)

Generalized Trust −0.003 −0.006 (0.004) (0.009)

Number of Friends 0.001 0.001 (0.0004) (0.001)

Observations 15,478 13,792 15,478 13,792 R2 0.003 0.128 0.001 0.089

4 A.3 March 2020 Qualtrics Survey In an online survey elded by Qualtrics to a panel of 1500 Egyptian respondents, we asked the following four questions related to drug use and drug policy preferences:

• Have you or anyone close to you (family or friend) ever used hashish as a recreational drug?

A. Yes. B. No.

• Have you or anyone close to you (family or friend) ever used tramadol as a recreational drug?

A. Yes. B. No.

• Which is closest to your opinion regarding Egypt’s law concerning the criminalization of the consumption of marijuana and the eorts to decriminalize marijuana in Egypt? [Those selecting A or B are coded as supporting depenalization.]

A. Possession of marijuana should not be a crime punishable by law. B. People arrested for possession of marijuana should not serve mandatory prison sentences for at least the rst two oenses. C. Judges should determine prison sentences on a case by case basis. D. People arrested for possession of marijuana should always serve mandatory prison sentences.

• Do you think that the Egyptian government should provide additional nancial support for drug treatment facilities? [Those selecting A are coded as supporting increased funding.]

A. Yes, the Egyptian government should provide more nancial support for drug treatment facilities. B. Government support should remain the same for drug treatment facilities. C. No, the Egyptian government should reduce its nancial support for drug treatment facilities.

5 A.3.1 Qualtrics Survey Results Figure A.1 displays the distribution of policy preferences regarding drug penalization. We see that respondents who have used drugs or know others who have are more likely to support drug depenalization, dened as supporting either full decriminalization of cannabis possession or the depenalization of the rst two oenses. This result holds when controls for respondent gender, age, religion, education level, and employment status are included.

Figure A.1: Attitudes toward penalization for drug use, Egypt 2020

Figure A.2 displays the distribution of respondent preferences regarding drug treatment resources. Respondents who have used drugs or know others who have are more likely to support increased state funding for drug treatment programs. This result holds with the inclusion of controls.

6 Figure A.2: Attitudes toward funding for drug treatment programs, Egypt 2020