Abstract
Although anecdotal reports suggest that many religious communities in the United States oppose public health policies such as medication assisted treatment and syringe services, the relation between religiosity and drug policy attitudes is currently unclear. A survey of support for protective and punitive drug policies was conducted with 3,096 residents from 14 Appalachian and Midwestern states currently affected by the rural opioid epidemic. Despite high prevalence of drug use in the sample, only 59 and 36 percent of the respondents respectively supported medication assisted treatment and syringe exchange services, and 52 and 50 percent respectively supported punishment and incarceration for people who use drugs. Furthermore, although religious affiliation had no association with personal support for either protective or punitive drug policies, frequency of religious service attendance was positively correlated with support for punitive policies and negatively correlated with support for protective policies. In addition, the perception of punitive norms among religious leaders was positively correlated with personal support for punitive policies, and the perception of protective norms among religious leaders was positively correlated with personal support for protective policies. Thus, religious attendance and religious norms may reduce compassion towards others in the context of the rural drug use epidemic in the United States.
Keywords: Opioid epidemic, religion, altruism, drug policy, attitude, norm
In the United States, the use of drugs including prescription pain relievers, heroin, and synthetic opioids (e.g., fentanyl) threatens public health and the welfare of many rural communities. In the United States, there were 67,367 drug-related overdose deaths in 2018 (NIDA, 2020), and the CDC (Centers for Disease Control and Prevention) estimates the cost of this crisis (i.e., health care, loss of productivity, treatments, and criminal justice) to be $87.5 billion a year. The opioid crisis has also been linked to outbreaks of HIV (Human Immunodeficiency Virus) and HCV (Hepatitis C Virus), mostly concentrated in rural areas of the United States. After initial outbreaks in Scott and Austin in the state of Indiana, and accumulating evidence of alarming overdose rates in other states, Van Handel and her collaborators (2016) modeled regional risk for rapid dissemination of HIV and HCV and found that the majority of the most vulnerable 220 counties in the country were located in Appalachia and the Midwest. Hence, this region has a need for policies to treat SUD (Substance Use Disorder) and reduce injection-related infections, including medication assisted treatment and syringe exchange programs.
Drug policies involve protective ones referred to as “harm reduction policies,” particularly medication assisted treatment and syringe services, which reduce long-term drug use (Deschamps et al., 2019) and communicable diseases (Hagan et al., 2011). For example, SSP (Syringe Service Programs) can reduce the likelihood of transmitting infectious disease via injection drug use (Samoff et al., 2020), and MAT (Medication-Assisted Treatment) can reduce overdoses and transmission of infectious disease (Fullerton et al., 2014). This paper addresses these issues. Other policies are punitive and entail a “war-on-drugs” approach (Coyne & Hall, 2017) typically associated with more overdoses, violence, and disease transmission (Coyne & Hall, 2017; Csete et al., 2016; Rogeberg, 2015), as well as unfair criminalization (Abadie et al., 2018; Vogel & Porter, 2016). We refer to each set of policies as protective and punitive respectively.
In rural areas of the US affected by drug use, religious values are likely correlates of support of or opposition to protective and punitive drug policies. On the one hand, 12-steps programs that meet in churches have gained high levels of popularity and provide positive support for people trying to stop using substances (Gianelli et al. 2019). On the other, more religious people have negative attitudes toward drug users (Jara-Concha & Cumsille, 2019) and lower support for protective policies (Gebelhoff, 2019). According to Szott (2018), this opposition stems from a moral model of addiction, which increases drug use stigma and blocks key initiatives like the provision of sterile syringes for disease prevention, or the use of medication as a part of SUD treatment. Many religious communities have either disapproved or overtly repudiated medication–assisted treatment, retail access to syringes, or syringe exchange programs (Szott, 2020), largely because they interpret substance use as a moral failure rather than a disease and see these science-supported programs as “enabling” drug use (Ibragimov et al., 2017). Although some churches have cooperated with campaigns delivering syringes (Shimron, 2019), many others continue to view such participation as facilitation of and tacit approval of drug use (Sulmasy, 2012). Thus, in the end, even though the Catholic doctrine, for example, supports syringe exchange programs on the grounds of justifiable cooperation and absence of intrinsic evil (Sulmasy, 2012), many Catholic groups and other Christian communities effectively oppose establishing syringe-exchange programs in their area (for a discussion of objections in Albany, NY, see Sulmasy, 2012).
In this paper, we studied the degree to which religious affiliation, frequency of attendance to religious services, and perceived norms of religious leaders correlate with support for protective and punitive policies in the area of substance use and HIV/HCV prevention. Religious affiliation, religious attendance, and religious-leader norms are different facets of religious identification and ongoing social influence within the religious community, and dovetail well with Putnam and Campbell’s (2010) conceptualization of religiosity as involving formal affiliation as well as behaviors and internalized norms. The question in our paper is whether religious affiliation, attendance, and norms have implications for community support of policies that may punish or protect people who use substances.
We were interested in determining which aspects of religiosity correlate with support for punitive and protective drug policies in Appalachia and the Midwest. For example, religious affiliation may be associated with policy support. In addition, attendance of religious services and religious leaders’ norms may be associated with policy support as well. That is, people who attend services regularly and people who perceive that leaders oppose syringe exchange programs may be more likely to oppose syringe exchange programs than may people who do not attend services regularly or who do not perceive policy opposition by religious leaders. This research tested these hypotheses with a large sample representative of states identified as having counties with high vulnerability to HIV and HCV outbreaks related to injection drug use. These counties were located in Appalachia and the Midwest.
Studying Associations between Drug Policy Support and Religiosity
The impact of religious convictions on concern for other people has attracted scholarly interest since the 1990’s. Some authors, mostly affiliated with religious institutions, have reported positive effects of religiosity on altruistic behaviors like donations to charity and serving the poor (Koenig, 2012). Along those lines, people whose faith declines appear to reduce their solidary activities (Krause & Pargament, 2017), people who receive spiritual support from religious organizations are more likely to practice compassion and forgiveness (Krause et al., 2019), and religious involvement correlates positively with volunteering (Gutierrez & Mattis, 2014). Moreover, a meta-analysis of 92 studies of religious priming reported evidence that reminders of religion can heighten prosocial behavior (Shariff et al. 2015; see also Schumann, 2020). From this standpoint, religious affiliation, religious service attendance, and religious-leader norms could be associated with support for protective drug policies.
Research contradicting a positive association between religious variables and concern for others have proposed that compassion and generosity, rather than religiosity per se, underlie the prosocial outcomes of religiosity (Steffen & Masters, 2005), and that generosity correlates with a higher educational level and more financial resources, which are often confounded with religiosity (Wiepking, 2009). Other research contradicting a positive link between religiosity and concern for others has shown that conservative religious beliefs are associated with labeling people who use drugs as violent and criminal, as well as responsible for damaging their families and society (Ibragimov et al., 2017). In Tajikistan, where 95 percent of the population is Sunni, campaigns that distributed syringes in the middle of a serious epidemic of HIV and HVC among people who injected drugs, crashed against religious stereotypes (Ibragimov et al., 2017). In the US, the proposal to create a safe injection place in Philadelphia was labeled a crack house’ by Christian conservatives (Gebelhoff, 2019), leading to a ban of the proposal. Conservative Christian groups also tend to espouse the 12-step method as the ideal and express disapproval of methadone and other medications for the treatment of SUD (Frank, 2011). In this light, religious affiliation, religious service attendance, and religious leader norms could be associated with support for punitive drug use policies.
This research involved a survey of general population within states affected by the rural opioid epidemic. The survey gauged respondents’ attitudes toward protective and punitive policies. Specifically, protective policies included providing clean syringes for people who inject drugs as a way of avoiding infections; paying for treatment for SUD; providing free treatment for infections that result from drug use; supporting the use of medication that reduces addiction; and funding drug treatment programs that help people out of addiction. Punitive policies included penalizing the use of drugs without a prescription and incarcerating people who use drugs illegally. The survey also included questions about religion affiliation if any; frequency of attendance of religious services; and religious leaders’ norms with respect to punitive and protective policies. Analyses relied on structural equation modeling.
Method
To understand the associations among support of protective or punitive policies, religious affiliation, religious-service attendance, and religious leaders’ norms in areas affected by the opioid crisis and vulnerable to HIV/HCV outbreaks, we recruited participants through Qualtrics Panels (Online Panels: Get Responses for Surveys & Research | Qualtrics, n.d.) covering the states of interest. The Qualtrics panel provides an online sample chosen from a pre-arranged pool of respondents who’ve agreed to be contacted by Qualtrics to respond to surveys. The panels are recruited nationwide via various local and national advertising methods according to industry standards.
Participants
The survey included 3,096 who completed the survey online. The sample was drawn from fourteen states in Appalachia and the Midwest, including Alabama, Georgia, Illinois, Indiana, Kansas, Kentucky, Michigan, Missouri, Ohio, Pennsylvania, South Carolina, Tennessee, Virginia, and West Virginia. To ensure good representation of highly vulnerable counties, half of the participants were drawn from state counties that ranked in the top 5% of injection-related vulnerability to HIV and HCV (van Handel et al. 2016); the other half were drawn from other counties in the same states. Participants received payment for their completion through their panel according to industry standards, and the study was approved by the University of Illinois’ Institutional Review Board.
Measures
Participants in this study were asked questions about demographics, alcohol/drug use, attitudes toward alcohol/drug use, social support, mental health, and attitudes and resources within their communities. Of interest in this paper were items concerning attitudes toward protective and punitive drug policies, religious affiliation, religious-service attendance, and religious leaders’ norms. We also use demographics and political ideology as controls.
Religious affiliation.
The item about religious affiliation asked participants to indicate ‘Which of the following, if any, describes your religious beliefs?’ The options were Baptist, Other Protestant, Catholic, Jewish, Muslim, Hindu, Agnostic, and Other.
Attendance to religious services.
Participants were asked by ‘How often do you attend services?’ Respondents provided their answer on a scale with the following points: 1(Never) to 6 (More than once a week).
Religious leaders’ norms concerning drug policies.
The survey included two items that measured religious leaders’ norms concerning punitive drug policies: “Religious leaders in my community want to make people afraid of the consequences of using drugs,” and “Religious leaders in my community believe that drug problems should be dealt with by punishing people for using drugs illegally.” Participants used a scale from 1(strongly disagree) to 5 (strongly agree) to rate support for protective and punitive policies. These two items were used for an index of religious leaders’ norms concerning punitive drug policies.
The survey also included measures of religious leaders’ norms concerning protective drug policies. These included (a) “Religious leaders in my community try to help people who are misusing drugs to recover,” (b) “The top concern of religious leaders in my community regarding drugs is to help people stay safe,” (c) “Religious leaders in my community are supportive or would be supportive of programs that provide ways for people who misuse drugs to stay safe (such as with clean needles to prevent spreading),” and (d) “Religious leaders in my community are supportive or would be supportive of treatment programs that use medication to help reduce drug addiction (e.g., MAT).” Participants used a scale from 1(strongly disagree) to 5 (strongly agree) to rate support for protective and punitive policies. These items were used to form an index of religious leaders’ norms in support of protective drug policies.
Support for drug policies.
The measures of support for protective policies included the following items: (a) “To prevent the spread of disease, the government should provide clean syringes for people who inject drugs as a way of avoiding infections;” (b) “The government should pay for treatment for drug addiction when community members need treatment;” (c) “The government should provide free treatment for infections that result from drug use;” and (d) “The government should deal with drug misuse by funding drug treatment programs that help people out of addiction.” The measures of support for punitive policies included (a) “The government should punish people for using drugs without a prescription;” and (b) “The government should jail people who use drugs illegally.” Participants used a scale from 1(strongly disagree) to 5 (strongly agree) to rate support for protective and punitive policies.
Demographics, lifetime drug use, and other descriptors.
Our survey also included demographic measures and lifetime drug use measures. To measure lifetime drug use, the survey asked whether participants ever used any of the substances in Table 1. It also asked if participants had ever been arrested, and measures about income, living situation, and having a health provider. We also had a measure of political ideology scored from 1 (extremely liberal) to 5 (extremely conservative) with the following points: Extremely liberal, liberal, neither liberal nor conservative, conservative, and extremely conservative.
Table 1.
Sociodemographic, Health, and Political Characteristics (N = 3096)
| Variable | N | % |
|---|---|---|
| Gender | ||
| Male | 1867 | 39.1 |
| Female | 1209 | 60.3 |
| Non-binary | 7 | 0.2 |
| Race/Ethnicity | ||
| White | 2713 | 87.6 |
| Black | 282 | 9.1 |
| American Indian/Alaskan | 59 | 1.9 |
| Asian | 66 | 2.1 |
| Hispanic/Latinx | 123 | 4.0 |
| Native Hawaiian or Pacific | 7 | 0.2 |
| Education | ||
| Less than high school | 149 | 4.8 |
| High School | 900 | 29.1 |
| College degree | 545 | 17.6 |
| Master’s degree | 211 | 6.8 |
| Professional degree | 44 | 1.4 |
| Doctoral degree | 17 | 0.5 |
| Employment status | 545 | 17.6 |
| Not working (student) | 99 | 3.2 |
| Not working (temporary layout) | 58 | 1.9 |
| Not working but looking for work | 238 | 7.7 |
| Not working-disabled | 361 | 11.7 |
| Not working-other | 216 | 7.0 |
| Not working-retired | 574 | 18.5 |
| Working (paid employee) | 1206 | 39.0 |
| Working/self employed | 227 | 7.3 |
| Looking for an employ | 1052 | 47.0 |
| Income less than 50000/year | 1355 | 60.3 |
| Receiving public assistance | 1000 | 44.9 |
| Living situation | ||
| Living alone | 1000 | 32.3 |
| Living with others | 2043 | 66.0 |
| In group home/ recovery home | 32 | 1.0 |
| Living in a hospital | 44 | 1.4 |
| A prison or jail | 13 | 0.4 |
| Relationship status | ||
| Married | 1380 | 44.6 |
| Divorced | 368 | 11.9 |
| Never married | 755 | 24.4 |
| Partnered but never married | 325 | 10.5 |
| Separated | 84 | 5.4 |
| Widowed | 166 | 5.4 |
| Political ideology | ||
| Liberal/extremely liberal | 713 | 23.1 |
| Neither liberal nor conservative | 1236 | 39.9 |
| Conservative/extremely conservative | 1129 | 36.4 |
| Religious affiliation | ||
| Baptist | 834 | 26.9 |
| Another Protestant | 593 | 19.2 |
| Catholic | 427 | 13.8 |
| Jewish | 39 | 1.3 |
| Muslim | 29 | 0.9 |
| Hindu | 4 | 0.1 |
| Agnostic | 251 | 8.2 |
| Religious service attendance | ||
| Attending religious services | 1975 | 64.3 |
| Not attending religious services | 1096 | 35.7 |
| Drug use | ||
| Marijuana | 1380 | 44.8 |
| Amphetamines without prescription or more than prescribed | 423 | 13.7 |
| Methamphetamine | 326 | 10.6 |
| Cocaine, crack | 407 | 13.2 |
| Prescription opioids without prescription or more than prescribed | 708 | 23.0 |
| Heroin | 205 | 6.7 |
| Fentanyl | 193 | 6.3 |
| Non-opioids prescription for pain without prescription or more than prescribed | 471 | 15.3 |
| Hallucinogens | 340 | 11.0 |
| Thinner or other inhalants | 149 | 4.8 |
| GHB | 114 | 3.7 |
| Prescription pills for anxiety without prescription or more than prescribed | 677 | 22.0 |
| Cigarettes | 1487 | 48.3 |
| Chewing tobacco | 425 | 13.8 |
| Nicotine vaping products | 665 | 21.6 |
| THC vaping products | 431 | 14.0 |
| Ever arrested | 159 | 5.1 |
| Having a health provider | 2318 | 51.6 |
The age of the sample was M = 44.96 (SD = 17.08)
Results
Description of the Sample
General characteristics.
A description of the sample appears in Table 1. As shown, we had good representation of participants in all of our target states (i.e., Alabama, Georgia, Illinois, Indiana, Kansas, Kentucky, Michigan, Missouri, North Carolina, Ohio, Pennsylvania, Tennessee, Virginia, and West Virginia). Participants’ average age was 45 years, and 60 percent of them were female. In accordance with the population in this region, participants were mostly White (88%) followed by African-American (9%), Hispanic (4%), Asian-American (2%), Native-American (2%), and Hawaiian or Pacific Islander (0.2%) respondents. Participants with less than a high school level of education comprised 5 percent of the sample; participants with complete high school comprised 29 percent of the sample; and participants with a college degree comprised 18 percent of the sample. For comparison, the US Census for the specific states in our sample involves 51% females, 79% Whites, 15% African Americans, 0.6% Native Americans, 3% Asian Americans, and 0.09% Native Hawaiian or Pacific Islander.
Politically, 23 percent of the sample indicated being Extremely Liberal or Liberal, whereas 36% indicated being Conservative or Extremely Conservative. Forty percent of the respondents chose the option Neither liberal nor conservative. About 50% of the sample was currently working, whereas the other half was not due to temporary unemployment, disability, being a student, or being retired. With respect to living arrangements, 66 percent of our respondents lived with other people, 32% lived alone, 1% lived in a hospital or recovery house, and 1% were homeless. Sixty-six percent of the sample reported having used some kind of substance, including marijuana, amphetamine without a prescription or more than prescribed, methamphetamines, cocaine, crack, prescription opioids, anxiolytics, or amphetamines without a prescription or more than prescribed, heroin, fentanyl, hallucinogens, inhalants, GHB, or tobacco. Lifetime drug use was 23% for opioids without a prescription or more than prescribed, 22% for anxiolytics without a prescription or more than prescribed, and 14% for amphetamines without a prescription or more than prescribed. Lifetime use of illicit substances was13% for cocaine or crack, 11% for hallucinogens, 11% for methamphetamines, 7% for heroin, and 6% for fentanyl, with smaller prevalence of other illicit substances as well. About three fourths of the sample had a health service provider, about a quarter were receiving public assistance, 38% had used alcohol rehabilitation services, and 11% had used drug rehabilitation services. Finally, 1% reported having been in jail or prison.
Religious affiliation, attendance, and religious leaders’ norms.
The sample was generally religious. Ninety-one percent of the participants reported belonging to a religion. Forty-six percent were Protestant, 14% Catholic, 1% Jewish, 1% Muslim, 0.1% Hindi, 9% Agnostics, 29% Other. With respect to attendance to religious services, 64% reported ever attending, with about half of those (29%) reporting attending between one and a few times a year, and the rest (35%) reporting attending monthly to attending more often than weekly. The norms of religious leaders were M = 2.42 (SD =1.15) in support of punitive policies and M = 2.60 (SD = 1.03) in support of protective policies.
Support for protective and punitive drug policies.
The index of attitudes toward protective policies had a M = 3.30 (SD = 1.00) and the index of attitudes toward punitive policies had a M = 3.01 (SD =.90). Considering participants who checked either agreeing or strongly agreeing with these items, support for protective was 67 percent for the government funding drug treatment, 59 percent for the government providing treatment with medication, 48 percent for the government paying for drug treatment of some sort, and 36 percent for the government providing clean syringes for people who inject drugs. Support for punitive proposals included 53 percent agreeing that the government should punish people who use drugs and 49 percent agreeing that the government should put people who use drugs in jail.
Associations between Religious Variables and Policy Support
Using the R package Lavaan (Lavaan Package-R, n.d.), we conducted Structural Equation Modeling with latent variables to test the associations of religious affiliation, religious service attendance, and religious leaders’ norms with support for punitive and protective policies (see correlation matrix in the Appendix). Age, income, sex (1 = male, 2 = female), and indicator variables for race and ethnicity were introduced as controls. Religious affiliation was assessed through a series of indicator variables for Agnostics, Catholics, Protestants, and religious service attendance was treated as a continuous variable. Punitive and protective religious norms and support for punitive and protective policy were treated as latent variables, with punitive religious leaders’ norms, protective religious leaders’ norms, support for punitive policies, and support for protective policies each being a latent factor. The model fit very well (CFI = .95; TLI = 0.93; RMSEA = .039; SRMR = .024) and all factor loadings within the measurement model were high. Table 2 presents all of the model’s coefficient, and Figure 1 presents a summary of the results. For simplicity, this figure does not show nonsignificant paths or correlations among external variables. Please refer to Table 1 for that information.
Table 2.
Effects of Religious Affiliation and Attendance on Relative Support for Protective versus Punitive Drug Policies
| Coefficient | SE | z | p | Standardized Coefficient | |
|---|---|---|---|---|---|
| Latent variables | |||||
| Punitive religious leaders’ norm | |||||
| Punitive religious leaders’ norm item 1 | 1.000 | ||||
| Punitive religious leaders’ norm item 2 | 0.764 | 0.046 | 16.738 | 0.000 | 0.648 |
| Protective religious leaders’ norm | |||||
| Protective religious leaders’ norm item 1 | 1.000 | 0.886 | |||
| Protective religious leaders’ norm item 2 | 0.909 | 0.033 | 27.245 | 0.000 | 0.689 |
| Protective religious leaders’ norm item 3 | 1.006 | 0.033 | 30.857 | 0.000 | 0.776 |
| Protective religious leaders’ norm item 4 | 1.044 | 0.032 | 32.663 | 0.000 | 0.828 |
| Support for punitive policies | |||||
| Support for punitive policies item 1 | 1.000 | 1.054 | |||
| Support for punitive policies item 2 (l2) | 1.052 | 0.035 | 30.170 | 0.000 | 0.879 |
| Support for protective policies | |||||
| Support for protective policies item 1 | 1.000 | 0.944 | |||
| Support for protective policies item 2 (l2) | 1.052 | 0.035 | 30.170 | 0.000 | 0.791 |
| Support for protective policies item 3 (l3) | 1.039 | 0.039 | 26.359 | 0.000 | 0.759 |
| Support for protective policies item 4 (l4) | 0.727 | 0.035 | 21.014 | 0.000 | 0.583 |
| Support for protective policies item 5 (l5) | 0.855 | 0.036 | 24.068 | 0.000 | 0.679 |
| Regressions: | |||||
| Support for punitive policies | |||||
| Punitive religious leaders’ norm | 0.222 | 0.034 | 6.613 | 0.000 | 0.205 |
| Age | −0.004 | 0.002 | −2.052 | 0.040 | −0.055 |
| Education | −0.047 | 0.021 | −2.238 | 0.025 | −0.062 |
| Conservative ideology | 0.230 | 0.028 | 8.065 | 0.000 | 0.223 |
| Income | 0.005 | 0.010 | 0.452 | 0.652 | 0.012 |
| Sex | 0.266 | 0.068 | 3.893 | 0.000 | 0.101 |
| Frequency of service attendance | 0.037 | 0.018 | 2.084 | 0.037 | 0.059 |
| Protestant | 0.080 | 0.064 | 1.244 | 0.214 | 0.038 |
| Catholics | 0.166 | 0.088 | 1.891 | 0.059 | 0.053 |
| Agnostic | 0.011 | 0.395 | 0.028 | 0.978 | 0.001 |
| Hispanic | −0.131 | 0.138 | −0.946 | 0.344 | −0.025 |
| White | 0.065 | 0.139 | 0.464 | 0.643 | 0.020 |
| Black | −0.176 | 0.152 | −1.159 | 0.246 | −0.049 |
| Support for protective policies | |||||
| Protective religious leaders’ norm | 0.174 | 0.032 | 5.491 | 0.000 | 0.163 |
| Age | −0.003 | 0.002 | −1.910 | 0.056 | −0.051 |
| Education | 0.043 | 0.019 | 2.296 | 0.022 | 0.064 |
| Conservative ideology | −0.229 | 0.026 | −8.926 | 0.000 | −0.248 |
| Income | −0.032 | 0.009 | −3.562 | 0.000 | −0.098 |
| Sex | −0.246 | 0.061 | −4.038 | 0.000 | −0.105 |
| Frequency of service attendance | −0.011 | 0.016 | −0.669 | 0.503 | −0.019 |
| Protestant | −0.095 | 0.057 | −1.653 | 0.098 | −0.050 |
| Catholics | −0.060 | 0.078 | −0.773 | 0.440 | −0.022 |
| Agnostic | −0.420 | 0.353 | −1.191 | 0.234 | −0.031 |
| Hispanic | 0.074 | 0.123 | 0.600 | 0.549 | 0.016 |
| White | 0.004 | 0.124 | 0.033 | 0.974 | 0.001 |
| Black | 0.286 | 0.136 | 2.107 | 0.035 | 0.089 |
| Punitive religious leaders’ norm | |||||
| Age | −0.004 | 0.002 | −2.237 | 0.025 | −0.066 |
| Education | −0.041 | 0.021 | −1.896 | 0.058 | −0.058 |
| Conservative ideology | −0.080 | 0.028 | −2.818 | 0.005 | −0.084 |
| Income | 0.004 | 0.010 | 0.422 | 0.????? | 0.013 |
| Sex | 0.009 | 0.069 | 0.127 | 0.899 | 0.004 |
| Frequency of service attendance | 0.078 | 0.018 | 4.369 | 0.000 | 0.136 |
| Protestant | 0.172 | 0.065 | 2.644 | 0.008 | 0.088 |
| Catholics | 0.019 | 0.089 | 0.218 | 0.827 | 0.007 |
| Agnostic | −0.273 | 0.402 | −0.678 | 0.498 | −0.019 |
| Hispanic | 0.105 | 0.140 | 0.747 | 0.455 | 0.022 |
| White | −0.045 | 0.142 | −0.318 | 0.750 | −0.015 |
| Black | −0.129 | 0.154 | −0.838 | 0.402 | −0.039 |
| Protective religious leaders’ norm | |||||
| Age | 0.003 | 0.001 | 2.122 | 0.034 | 0.056 |
| Education | −0.050 | 0.017 | −2.888 | 0.004 | −0.079 |
| Conservative ideology | 0.033 | 0.023 | 1.427 | 0.154 | 0.038 |
| Income | 0.005 | 0.008 | 0.610 | 0.542 | 0.017 |
| Sex | −0.037 | 0.056 | −0.662 | 0.508 | −0.017 |
| Frequency of service attendance | 0.130 | 0.015 | 8.817 | 0.000 | 0.249 |
| Protestant | 0.178 | 0.053 | 3.364 | 0.001 | 0.100 |
| Catholics | 0.083 | 0.073 | 1.142 | 0.254 | 0.032 |
| Agnostic | −0.380 | 0.327 | −1.161 | 0.246 | −0.030 |
| Hispanic | 0.148 | 0.114 | 1.295 | 0.195 | 0.034 |
| White | −0.077 | 0.115 | −0.666 | 0.505 | −0.029 |
| Black | 0.219 | 0.126 | 1.745 | 0.081 | 0.073 |
| Covariances: | |||||
| Punitive religious leaders’ norm | |||||
| Protective religious leaders’ norm | 0.443 | 0.029 | 15.127 | 0.000 | 0.556 |
| Support for punitive policies | |||||
| Support for protective policies | −0.142 | 0.027 | −5.212 | 0.000 | −0.163 |
| Variances: | |||||
| Punitive religious leaders’ norm item 1 | 0.434 | 0.053 | 8.237 | 0.000 | 0.313 |
| Punitive religious leaders’ norm item 2 | 0.767 | 0.040 | 19.245 | 0.000 | 0.580 |
| Protective religious leader’s norm item 1 | 0.553 | 0.025 | 22.263 | 0.000 | 0.413 |
| Protective religious leader’s norm item 2 | 0.719 | 0.029 | 24.725 | 0.000 | 0.526 |
| Protective religious leader’s norm item 3 | 0.523 | 0.024 | 21.794 | 0.000 | 0.397 |
| Protective religious leader’s norm item 4 | 0.394 | 0.021 | 18.763 | 0.000 | 0.315 |
| Support for punitive policies items 1 | 0.622 | 0.046 | 13.436 | 0.000 | 0.359 |
| Support for punitive policies item 2 | 0.361 | 0.047 | 7.682 | 0.000 | 0.227 |
| Support for punitive policies item 1 | 1.208 | 0.049 | 24.890 | 0.000 | 0.575 |
| Support for punitive policies item 2 | 0.590 | 0.030 | 19.578 | 0.000 | 0.374 |
| Support for punitive policies item 3 | 0.707 | 0.033 | 21.120 | 0.000 | 0.423 |
| Support for punitive policies item 4 | 0.913 | 0.035 | 25.987 | 0.000 | 0.660 |
| Support for punitive policies item 5 | 0.761 | 0.032 | 24.058 | 0.000 | 0.539 |
| Punitive religious leaders’ norm | 0.916 | 0.067 | 13.776 | 0.000 | 0.965 |
| Protective religious leaders’ norm | 0.693 | 0.040 | 17.235 | 0.000 | 0.884 |
| Support for punitive policies | 0.969 | 0.054 | 18.044 | 0.000 | 0.871 |
| Support for protective policies | 0.776 | 0.052 | 14.844 | 0.000 | 0.871 |
Figure 1.
Structural equation model.
Figure 1 provides support for the idea that religious variables shaped respondent’s attitudes toward drug policies. First, attending religious services and a punitive religious-leader norm were positively correlated with personally supporting punitive policies. Second, a protective religious-leader norm correlated with personally supporting protective policies.
Several of the control variables correlated with policy support as well. For example, as one might expect, conservative ideology correlated positively with support for punitive policies but negatively with support for protective policies. Older age and higher education were negatively correlated with support for punitive policies, whereas being female was positively correlated with support for punitive policies. Being black and more educated correlated positively with support for protective policies, whereas being female and having a higher income correlated negatively with support for protective policies. Overall, younger, female, less educated, non-black, and higher income participants endorsed harsher drug policies than did older, male, more educated, black, and lower income participants.
Religious leaders’ norms correlated with specific religious affiliations and religious service attendance. A punitive norm correlated positively with service attendance and being protestant, and negatively with being conservative and older in age. A protective norm correlated positively with being protestant, attending religious services, and being older, and negatively with higher education and being female. That is, even though service attendance differentially affected personal support for punitive and protective drug policies, more frequent attendees perceived stronger punitive as well as stronger protective norms among religious leaders. Being protestant had similar associations by which respondents who identified as protestant perceived stronger punitive and protective norms. Finally, similar to the policy support findings, females perceived harsher policy norms, whereas older respondents perceived more benevolent norms among religious leaders. More highly educated respondents perceived weaker protective norms among their leaders, which combined with the more benevolent attitudes of the more highly educated, suggests that this group perceived religious leaders as being harsh.
In summary, attendance and religious leaders’ norms are important correlates of people’s support for punitive and protective policies. However, the model in Figure 1 assumes that the norms precede personal support for the policies although it is also possible that people infer norms that are consistent with their own attitudes (Festinger, 1954; Turner, 1985). For this reason, we tested another model in which norms were the endogenous variables and support for punitive and protective policies predicted the norms. This model, whose details appear in the Appendix, also had a good fit (CFI = .94, TLI = .92, RMSEA = .041, SRMR = .033). Although the two models cannot be directly compared, the fit of our proposed model was slightly better than the fit of this alternate model.1
Discussion
We investigated patterns of association between religiosity and attitudes toward policies intended to reduce harm from drug use as well as punitive policies. The respondents, who were demographically similar to the rural populations affected by the current opioid and methamphetamine epidemic, were mostly religious and had moderate attendance. In general, religious affiliation had no impact on either protective or punitive policy attitudes. However, consistent with past findings that people with religious practice are less compassionate towards people who use drugs than are those without such a practice, we found that religious service attendance correlated with stronger support for punitive policies and weaker support for protective ones. That is, the greater the frequency of religious service attendance, the greater the support for punishment and incarceration for people who use drugs. Thus, despite churches often supporting services for recovery from alcohol and drug use disorder, regions with higher religious service attendance may continue to experience support for punitive than protective drug policies.
Our study contributes to the literature on links between the public stigma surrounding opioid use disorder (Magnus et al., 2013; Neale et al., 2008; Rivera et al., 2014; Van Boekel et al., 2013a) and drug use policies. For example, people who use drugs anticipate the negative judgment of providers if they need harm reduction interventions (Earnshaw et al., 2013; Paquette et al., 2018; Van Boekel et al., 2013), and providers who prescribe MAT feel stigmatized for providing these medical services (Madden, 2019). It is likely that the relation between religious variables and drug policy support is partly due to the stigma of drug use, which is known to correlate with religiosity (Ibragimov et al., 2017; Vigliotti et al., 2020). However, the objections to drug policies are also based on the notion of the lack of efficacy of actions that are framed as “enabling” drug use (Ibragimov et al., 2017).
There are limitations to our study. First, the survey was conducted electronically, may exclude participants who may only be able to respond via phone or in person. Second, our analysis of religious affiliation was limited to the affiliations as reported in the region and led to very low numbers of Jewish and Muslim individuals. Despite these limitations, to the best of our knowledge, the study is the first examine associations between religion and policy attitudes in rural areas that carry a heavy burden of the current substance use disorder epidemic.
These findings may be useful in finding ways of working with religious leaders and communities to become more open to science-based policies that protect people who use drugs and consequently society at large. For example, religious conferences may begin to address these issues and mass media messages may be used to link religiosity to the protection of all vulnerable populations, including those who use drugs. There are two large community projects in Appalachia currently funded by the NIDA (National Institute of Drug Abuse). HEALing Communities (NIDA, n.d.) is a large-scale attempt to integrate prevention, overdose treatment, and medication-based treatment in select communities in the United States. Also with NIDA funding, the Grid for the Reduction of Vulnerability (Grid for the Reduction of Vulnerability, n.d.) has invited agencies from 99 counties to participate in community efforts to support services and protective policies in communities in Appalachia and the Midwest. In this context, our findings suggest that finding ways of incorporating religious leaders and developing an agenda that incorporates religious value in a way that increases compassion may go a long way in reducing the harm of drug use in rural areas on the United States.
Public Significance Statement.
A large survey of rural areas in Appalachia and the Midwest suggested that religious attendance and religious norms reduce compassion towards others in the context of the rural drug use epidemic in the United States. Religious leaders may be mobilized to support protective and efficacious drug policy to curb the opioid epidemic.
APPENDIX
Correlation Matrix
| Characteristics of the sample | 1 | 2 | 3 | 4 | 5 | 6 | 8 | 9 | 10 | 11 | 12 | 13 | 14 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1. Age | 1 | ||||||||||||
| 2. Sex | .152 .000 |
1 | |||||||||||
| 3. Education | .118 .000 |
.130 .000 |
1 | ||||||||||
| 4. Income | −.014 .467 |
.160 .000 |
.399 .000 |
1 | |||||||||
| 5. Conservative political ideology | .163 .000 |
.043 .017 |
−.015 .407 |
.036 .407 |
1 | ||||||||
| 6. Frequency of religious service attendance | .103 .000 |
.022 .218 |
.204 .000 |
.128 .000 |
.199 .000 |
1 | |||||||
| 7. Punish drug use | .026 .155 |
−.030 .097 |
.006 .735 |
.034 .076 |
.179 .000 |
.130 .000 |
.1 | ||||||
| 8. Provide clean syringes | −.148 .000 |
.030 .092 |
.021 .245 |
.007 .708 |
−.279 .000 |
−.065 .000 |
−.187 .000 |
1 | |||||
| 9. Pay for treatment | −.121 .000 |
−.014 .440 |
−.019 .302 |
−.058 .003 |
−.229 .000 |
−.023 .204 |
−.219 .000 |
.521 .000 |
1 | ||||
| 10. Free treatment for infections from drug injection | −.094 .000 |
.052 .004 |
−.029 .110 |
−.030 .121 |
−.239 .000 |
−.046 .011 |
−.134 .000 |
.559 .000 |
.642 .000 |
1 | |||
| 11. Put people who use drugs in jail | .000 .984 |
−.059 .001 |
−.013 .466 |
.025 .189 |
.220 .000. |
.155 .000 |
.711 .000 |
−.195 .000 |
−.153 .000 |
−.133 .000 |
1 | ||
| 12. Support medication-assisted treatment | −.050 .006 |
.053 .003 |
.025 .174 |
.009 .632 |
−.155 .000 |
−.026 .156 |
−.050 .006 |
.397 .000 |
.480 .000 |
.449 .000 |
−.080 .000 |
1 | |
| 13. Fund programs for treatment | −.064 .000 |
.028 .117. |
.008 .651 |
.018 .359 |
−.211 .000 |
−.049 .007 |
−.047 .009 |
.395 .000 |
.592 .000 |
.513 .000 |
−.074 .000 |
.530 .000 |
1 |
Effects of Religious Affiliation and Attendance: Alternate Model
| Unstandardized Coefficient | SE | z | p | Standardized Coefficient | |
|---|---|---|---|---|---|
| Latent variables | |||||
| Punitive religious leaders’ norm | |||||
| Religious leaders’ norm item 1 | 1.000 | 0.988 | 0.844 | ||
| Religious leaders’ norm item 2 | 0.736 | 0.048 | 15.437 | 0.000 | 0.635 |
| Protective religious leaders’ norm | |||||
| Religious leaders’ norm item 1 | 1.000 | 0.883 | 0.765 | ||
| Religious leaders’ norm item 2 | 0.908 | 0.034 | 27.100 | 0.000 | 0.687 |
| Protective religious leaders’ norm item 2 | 1.006 | 0.033 | 30.717 | 0.000 | 0.776 |
| Protective religious leaders’ norm item 2 | 1.044 | 0.032 | 32.462 | 0.000 | 0.826 |
| Support for punitive policies | |||||
| Support for punitive policies item 1 | 1.000 | 1.046 | 0.794 | ||
| Support for punitive policies item 2 (l2) | 1.067 | 0.037 | 28.983 | 0.000 | 0.887 |
| Support for protective policies | |||||
| Support for protective policies item 1 | 1.000 | 0.935 | 0.647 | ||
| Support for protective policies item 2 (l2) | 1.067 | 0.037 | 28.983 | 0.000 | 0.793 |
| Support for protective policies item 3 (l3) | 1.050 | 0.041 | 25.897 | 0.000 | 0.760 |
| Support for protective policies item 4 (l4) | 0.733 | 0.035 | 20.743 | 0.000 | 0.583 |
| Support for protective policies item 5 (l5) | 0.864 | 0.036 | 23.727 | 0.000 | 0.680 |
| Regressions: | |||||
| Punitive religious leaders’ norm | |||||
| Support for punitive policies | 0.103 | 0.027 | 3.765 | 0.000 | 0.109 |
| Age | −0.003 | 0.002 | −1.884 | 0.060 | −0.055 |
| Education | −0.035 | 0.021 | −1.629 | 0.103 | −0.049 |
| Conservative ideology | −0.102 | 0.029 | −3.520 | 0.000 | −0.106 |
| Income | 0.005 | 0.010 | 0.450 | 0.653 | 0.014 |
| Sex | −0.020 | 0.070 | −0.291 | 0.771 | −0.008 |
| Frequency of service attendance | 0.074 | 0.018 | 4.107 | 0.000 | 0.126 |
| Protestant | 0.160 | 0.065 | 2.454 | 0.014 | 0.081 |
| Catholics | 0.006 | 0.089 | 0.063 | 0.950 | 0.002 |
| Agnostic | −0.252 | 0.402 | −0.627 | 0.531 | −0.018 |
| Hispanic | 0.117 | 0.140 | 0.835 | 0.404 | 0.024 |
| White | −0.051 | 0.142 | −0.361 | 0.718 | −0.017 |
| Black | −0.115 | 0.155 | −0.747 | 0.455 | −0.035 |
| Protective religious leaders’ norm | |||||
| Support for protective policies | 0.092 | 0.025 | 3.641 | 0.000 | 0.097 |
| Age | 0.003 | 0.001 | 2.289 | 0.022 | 0.060 |
| Education | −0.054 | 0.017 | −3.082 | 0.002 | −0.084 |
| Conservative ideology | 0.053 | 0.024 | 2.252 | 0.024 | 0.062 |
| Income | 0.008 | 0.008 | 0.954 | 0.340 | 0.026 |
| Sex | −0.014 | 0.056 | −0.253 | 0.800 | −0.006 |
| Frequency of service attendance | 0.129 | 0.015 | 8.799 | 0.000 | 0.248 |
| Protestant | 0.184 | 0.053 | 3.495 | 0.000 | 0.104 |
| Catholics | 0.087 | 0.072 | 1.207 | 0.227 | 0.033 |
| Agnostic | −0.336 | 0.326 | −1.031 | 0.303 | −0.026 |
| Hispanic | 0.139 | 0.114 | 1.221 | 0.222 | 0.032 |
| White | −0.076 | 0.115 | −0.662 | 0.508 | −0.028 |
| Black | 0.190 | 0.125 | 1.518 | 0.129 | 0.064 |
| Support for punitive policies | |||||
| Education | −0.055 | 0.021 | −2.628 | 0.009 | −0.074 |
| Conservative ideology | 0.208 | 0.028 | 7.308 | 0.000 | 0.203 |
| Age | −0.004 | 0.002 | −2.510 | 0.012 | −0.068 |
| Income | 0.005 | 0.010 | 0.540 | 0.589 | 0.015 |
| Sex | 0.264 | 0.068 | 3.862 | 0.000 | 0.101 |
| Frequency of service attendance | 0.053 | 0.018 | 3.027 | 0.002 | 0.086 |
| Protestant | 0.116 | 0.064 | 1.812 | 0.070 | 0.055 |
| Catholics | 0.168 | 0.088 | 1.907 | 0.057 | 0.054 |
| Agnostic | −0.049 | 0.396 | −0.124 | 0.902 | −0.003 |
| Hispanic | −0.105 | 0.138 | −0.759 | 0.448 | −0.020 |
| White | 0.060 | 0.140 | 0.429 | 0.668 | 0.019 |
| Black | −0.198 | 0.152 | −1.300 | 0.194 | −0.056 |
| Support for protective policies | |||||
| Age | −0.002 | 0.002 | −1.555 | 0.120 | −0.042 |
| Education | 0.034 | 0.019 | 1.821 | 0.069 | 0.051 |
| Conservative ideology | −0.221 | 0.026 | −8.631 | 0.000 | −0.242 |
| Income | −0.031 | 0.009 | −3.442 | 0.001 | −0.096 |
| Sex | −0.250 | 0.061 | −4.102 | 0.000 | −0.107 |
| Frequency of service attendance | 0.012 | 0.016 | 0.753 | 0.452 | 0.021 |
| Protestant | −0.063 | 0.057 | −1.108 | 0.268 | −0.034 |
| Catholics | −0.046 | 0.078 | −0.586 | 0.558 | −0.017 |
| Agnostic | −0.483 | 0.352 | −1.371 | 0.170 | −0.036 |
| Hispanic | 0.098 | 0.123 | 0.799 | 0.424 | 0.021 |
| White | −0.009 | 0.124 | −0.075 | 0.940 | −0.003 |
| Black | 0.321 | 0.135 | 2.371 | 0.018 | 0.101 |
| Covariances: | |||||
| Support for protective policies | |||||
| Support for punitive policies | −0.108 | 0.027 | −4.008 | 0.000 | −0.122 |
| Punitive religious leaders’ norm | |||||
| Protective religious leaders’ norm | 0.424 | 0.029 | 14.761 | 0.000 | 0.532 |
| Variances: | |||||
| Punitive religious leaders’ norm | 0.393 | 0.059 | 6.693 | 0.000 | 0.287 |
| Punitive religious leaders’ norm item 2 | 0.785 | 0.041 | 19.041 | 0.000 | 0.597 |
| Protective religious leader’s norm item 1 | 0.552 | 0.025 | 22.204 | 0.000 | 0.415 |
| Protective religious leader’s norm item 2 | 0.719 | 0.029 | 24.704 | 0.000 | 0.528 |
| Protective religious leader’s norm item 3 | 0.522 | 0.024 | 21.716 | 0.000 | 0.398 |
| Protective religious leader’s norm item 4 | 0.394 | 0.021 | 18.714 | 0.000 | 0.317 |
| Support for punitive policies items 1 | 0.642 | 0.048 | 13.384 | 0.000 | 0.370 |
| Support for punitive policies items 2 | 0.339 | 0.050 | 6.816 | 0.000 | 0.214 |
| Support for punitive policies item 1 | 1.213 | 0.049 | 24.945 | 0.000 | 0.581 |
| Support for punitive policies item 2 | 0.587 | 0.030 | 19.384 | 0.000 | 0.371 |
| Support for punitive policies item 3 | 0.706 | 0.034 | 21.065 | 0.000 | 0.423 |
| Support for punitive policies item 4 | 0.914 | 0.035 | 25.989 | 0.000 | 0.661 |
| Support for punitive policies item 5 | 0.760 | 0.032 | 24.019 | 0.000 | 0.538 |
| Support for punitive policies | 0.932 | 0.071 | 13.087 | 0.000 | 0.955 |
| Support for protective policies | 0.681 | 0.040 | 17.156 | 0.000 | 0.874 |
| Punitive religious leaders’ norm | 1.001 | 0.056 | 17.986 | 0.000 | 0.914 |
| Protective religious leaders’ norm | 0.781 | 0.054 | 14.563 | 0.000 | 0.894 |
Footnotes
The results in Figure 1 do not chance when lifetime drug use is added into the model.
References
- Abadie R, Gelpi-Acosta C, Davila C, Rivera A, Welch-Lazoritz M, & Dombrowski K (2018). `Ìt Ruined My Life{‘‘}: The effects of the War on Drugs on people who inject drugs (PWID) in rural Puerto Rico. INTERNATIONAL JOURNAL OF DRUG POLICY, 51, 121–127. 10.1016/j.drugpo.2017.06.011 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Coyne C, & Hall A (2017). Four Decades and Counting: The Continued Failure of the War on Drugs. https://www.cato.org/sites/cato.org/files/pubs/pdf/pa-811-updated.pdf [Google Scholar]
- Csete J, Kamarulzaman A, Kazatchkine M, Altice F, Balicki M, Buxton J, Cepeda J, Comfort M, Goosby E, Goulao J, Hart C, Kerr T, Lajous AM, Lewis S, Martin N, Mejia D, Camacho A, Mathieson D, Obot I, … Beyrer C (2016). Public health and international drug policy. LANCET, 387(10026), 1427–1480. 10.1016/S0140-6736(16)00619-X [DOI] [PMC free article] [PubMed] [Google Scholar]
- Deschamps J, Gilbertson J, Straube S, Dong K, MacMaster FP, Korownyk C, Montgomery L, Mahaffey R, Downar J, Clarke H, Muscedere J, Rittenbach K, Featherstone R, Sebastianski M, Vandermeer B, Lynam D, Magnussen R, Bagshawl SM, & Rewa OG (2019). Association between harm reduction strategies and healthcare utilization in patients on long-term prescribed opioid therapy presenting to acute healthcare settings: a protocol for a systematic review and meta-analysis. SYSTEMATIC REVIEWS, 8. 10.1186/s13643-019-0997-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Earnshaw VA, Bogart LM, Dovidio JF, & Williams DR (2013). Stigma and racial/ethnic HIV disparities: Moving toward resilience. American Psychologist, 68(4), 225–236. 10.1037/a0032705 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Festinger L (1954). A Theory of Social Comparison Processes. Human Relations. 10.1177/001872675400700202 [DOI] [Google Scholar]
- Frank D (2011). The Trouble with Morality: The Effects of 12-Step Discourse on Addicts’ Decision-Making. Journal of Psychoactive Drugs, 43(3). 10.1080/02791072.2011.605706 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fullerton CA, Kim M, Thomas CP, Lyman DR, Montejano LB, Dougherty RH, Daniels AS, Ghose SS, & Delphin-Rittmon ME (2014). Medication-assisted treatment with methadone: Assessing the evidence. Psychiatric Services, 65(2), 146–157. 10.1176/appi.ps.201300235 [DOI] [PubMed] [Google Scholar]
- Gebelhoff R (2019). Safe injections sites aren’t immoral. But attempting to block them sure is. In The Washington Post. [Google Scholar]
- Grid for the Reduction of Vulnerability. (n.d.). Retrieved March 19, 2021, from https://www.thegrov.org/ [Google Scholar]
- Gutierrez IA, & Mattis JS (2014). Factors Predicting Volunteer Engagement Among Urban-Residing African American Women. Journal of Black Studies, 45(7). 10.1177/0021934714543189 [DOI] [Google Scholar]
- Hagan H, Pouget ER, & Des Jarlais DC (2011). A Systematic Review and Meta-Analysis of Interventions to Prevent Hepatitis C Virus Infection in People Who Inject Drugs. JOURNAL OF INFECTIOUS DISEASES, 204(1), 74–83. 10.1093/infdis/jir196 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ibragimov U, Cooper HL, Haardorfer R, Dunkle KL, Zule WA, & Wong FY (2017). Stigmatization of people who inject drugs (PWID) by pharmacists in Tajikistan: sociocultural context and implications for a pharmacy-based prevention approach. HARM REDUCTION JOURNAL, 14. 10.1186/s12954-017-0190-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jara-Concha P, & Cumsille F (2019). Urban Residents Atttitudes Towards People With Drug Abuse in Five Districts of Concepcion Commune, Chile. Texto & Contexto - Enfermagem, 28(spe). 10.1590/1980-265x-tce-cicad-6-20 [DOI] [Google Scholar]
- Koenig HG (2012). Do Religion and Compassion Go Hand-in-Hand? In Catholic Exchange. [Google Scholar]
- Krause N, Hill PC, & Ironson G (2019). Evaluating the relationships among religion, social virtues, and meaning in life. Archive for the Psychology of Religion, 41(1). 10.1177/0084672419839797 [DOI] [Google Scholar]
- Krause N, & Pargament KI (2017). Losing My Religion: Exploring the Relationship Between a Decline in Faith and a Positive Affect. Applied Research in Quality of Life, 12(4). 10.1007/s11482-016-9495-2 [DOI] [Google Scholar]
- Lavaan Package-R. (n.d.). Retrieved March 19, 2021, from https://www.lavaan.ugent.be/about.html [Google Scholar]
- Madden EF (2019). Intervention stigma: How medication-assisted treatment marginalizes patients and providers. Social Science and Medicine, 232, 324–331. 10.1016/j.socscimed.2019.05.027 [DOI] [PubMed] [Google Scholar]
- Magnus M, Herwehe J, Murtaza-Rossini M, Reine P, Cuffie D, Gruber D, & Kaiser M (2013). Linking and Retaining HIV Patients in Care: The Importance of Provider Attitudes and Behaviors. AIDS Patient Care and STDs, 27(5), 297–205. 10.1089/apc.2012.0423 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Neale J, Tompkins C, & Sheard L (2008). Barriers to accessing generic health and social care services: A qualitative study of injecting drug users. Health and Social Care in the Community, 16(2), 147–154. 10.1111/j.1365-2524.2007.00739.x [DOI] [PubMed] [Google Scholar]
- NIDA. (n.d.). HEALing Communities Study. https://heal.nih.gov/research/research-to-practice/healing-communities [Google Scholar]
- Paquette CE, Syvertsen JL, & Pollini RA (2018). Stigma at every turn: Health services experiences among people who inject drugs. International Journal of Drug Policy, 57, 104–110. 10.1016/j.drugpo.2018.04.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rivera AV, DeCuir J, Crawford ND, Amesty S, & Lewis CF (2014). Internalized stigma and sterile syringe use among people who inject drugs in New York City, 2010–2012. Drug and Alcohol Dependence, 144, 259–264. 10.1016/j.drugalcdep.2014.09.778 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Rogeberg O (2015). Drug policy, values and the public health approach - four lessons from drug policy reform movements. NORDIC STUDIES ON ALCOHOL AND DRUGS, 32(4), 347–364. 10.1515/nsad-2015-0034 [DOI] [Google Scholar]
- Samoff E, Mobley V, Hudgins M, Cope AB, Adams ND, Caputo CR, Dennis AM, Billock RM, Crowley CA, Clymore JM, & Foust E (2020). HIV Outbreak Control With Effective Access to Care and Harm Reduction in North Carolina, 2017–2018. American Journal of Public Health, 110(3), 394–400. 10.2105/AJPH.2019.305490 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shimron Y (2019). Church opens door to syringe exchange. In The Christian Century. [Google Scholar]
- Steffen PR, & Masters KS (2005). Does compassion mediate the intrinsic religion-health relationship? Annals of Behavioral Medicine, 30(3). 10.1207/s15324796abm3003_6 [DOI] [PubMed] [Google Scholar]
- Sulmasy DP (2012). Catholic Participation in Needle- and Syringe-Exchange Programs for Injection-Drug Users: An Ethical Analysis. Theological Studies, 73(2). 10.1177/004056391207300207 [DOI] [Google Scholar]
- Szott K (2020). ‘Heroin is the devil’: addiction, religion, and needle exchange in the rural United States. Critical Public Health, 30(1). 10.1080/09581596.2018.1516031 [DOI] [Google Scholar]
- Turner JC (1985). Social Categorization and the self concept: A social cognitive theory of group behavior. Advances in Group Process. [Google Scholar]
- Van Boekel LC, Brouwers EPM, Van Weeghel J, & Garretsen HFL (2013). Stigma among health professionals towards patients with substance use disorders and its consequences for healthcare delivery: Systematic review. In Drug and Alcohol Dependence (Vol. 131, Issues 1–3, pp. 23–35). 10.1016/j.drugalcdep.2013.02.018 [DOI] [PubMed] [Google Scholar]
- Vigliotti V, Taggart T, Walker M, Kusmastuti S, & Ransome Y (2020). Religion, faith, and spirituality influences on HIV prevention activities: A scoping review. PLOS ONE, 15(6). 10.1371/journal.pone.0234720 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vogel M, & Porter LC (2016). Toward a Demographic Understanding of Incarceration Disparities: Race, Ethnicity, and Age Structure. JOURNAL OF QUANTITATIVE CRIMINOLOGY, 32(4), 515–530. 10.1007/s10940-015-9265-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wiepking P (2009). Resource that make you generous: Effects of social and human resources on charitable giving. Social Forces, 87(4), 1973–1995. [Google Scholar]