Skip to main content
NCBI home page
Health & Justice logoLink to Health & Justice
. 2025 Jul 25;13:46. doi: 10.1186/s40352-025-00356-2

An experimental investigation of federal messaging on public support for enforcement- and treatment-based approaches for opioid overdose prevention in South Carolina

Lídia Gual-Gonzalez 1, Hunter M Boehme 1,✉, Peter Leasure 2, Pieter A Baker 1, Melissa S Nolan 1
PMCID: PMC12291324  PMID: 40711698

Abstract

Background

As the opioid overdose crisis continues to produce excessive morbidity and mortality in the United States, government agencies have applied various approaches to prevent overdoses, including law-enforcement efforts (e.g., arresting people who use drugs, interrupting drug traffickers, etc.) and treatment-based approaches (e.g., naloxone, medications for opioid use disorder, etc.). Public perception and support of these approaches are relevant for informing policy, allocating resources, and effectively implementing community interventions to prevent drug-related harms.

Methods

Using an embedded informational survey design, we experimentally assessed whether public support for strategies to prevent overdose in South Carolina is influenced by language from federal agencies describing treatment- or enforcement-based approaches. Respondents were randomly assigned to one of three groups: (1) enforcement -based approach, (2) treatment-based approach, or (3) the control condition. Those assigned to experimental groups were presented with statistics on drug overdose deaths, followed by an informational prompt with language about overdose prevention approaches from either DEA (enforcement) or NIH (treatment), while the control group received no informational prompt.

Results

Findings from a sample of 4,675 respondents indicated that those assigned the DEA prompt were significantly more likely to support enforcement-based approaches in arresting drug traffickers and people who use drugs (AME = 0.060, p < 0.001). On the other hand, those assigned to the NIH prompt were significantly more likely to agree that both law enforcement (AME = 0.065, p < 0.0001) and clinicians (AME = 0.044, p < 0.05) are capable of preventing drug overdose deaths.

Conclusions

These findings shed light on public perceptions of approaches to addressing the opioid epidemic and limited modifiability when presented with language from federal agencies. This may inform future research, practice, and/or policy aiming to maintain public safety while also providing treatment options to people who use drugs in order to reduce overdose deaths.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40352-025-00356-2.

Keywords: Drug overdose, Law enforcement, Opioid epidemic, Population survey

Introduction

The opioid crisis continues to pose a serious public health concern in the United States of America (US), with overdose death rates increasing by over 22% between 2020 and 2021 (Centers for Disease Control and Prevention, 2023). Fentanyl, a synthetic opioid approved to treat severe pain, has played an outsized role in driving overdose deaths as clandestine manufacturing and penetration of fentanyl into the illicit drug market has resulted in a serious uptick in overdose and accounting for 88% of opioid-related deaths in 2021 (Center for Disease Control and Prevention, 2023; Han et al., 2019). Opioid-related overdose mortality in the US has followed four separate waves since the beginning of the 21st century: prescription opioid overdose deaths rising in the 1990s, heroin overdose deaths increasing in the early 2010s, the rise in synthetic opioid overdose deaths beginning in 2013 (Center for Disease Control and Prevention, 2023), followed by a recent fourth wave characterized by polysubstance use (opioids and stimulants [i.e. cocaine, methamphetamines])(Friedman & Shover, 2023; Jenkins, 2021).

Amidst this backdrop of rising overdose mortality and drug-related harms, the US government declared an “opioid epidemic” in 2017 and, through the work of federal agencies, has sought to increase public awareness regarding opioid-related risks (Hargan, 2017) and limit the public health impacts of illicit opioid use. While the specific strategies have evolved, the primary approach to addressing drug epidemics in the US has been prohibitionist policy and government-backed enforcement (Kerr et al., 2005). Following the establishment of the Drug Enforcement Administration (DEA) in the 1970s and the escalation of the “War on Drugs” throughout the 1980s, enforcement-based initiatives have spread and persist throughout local, state, and federal agencies. A main focus of enforcement-based approaches is the interdiction and interruption of drug trafficking organizations, often resulting in serious punishments for both drug traffickers and people who use drugs (Cooper, 2015; Lynch, 2012; Meeks, 2006).

As the War on Drugs has expanded and evolved, bipartisan political policies have prioritized expansive drug enforcement, enhanced penalties for drug trafficking, and harsher penalties for crack-cocaine; resulting in disproportionate harm to communities of color (Hinton, 2016; Pyra et al., 2022; Roberts, 2022). These federal decisions may also have affected public perceptions of those who use drugs and/or traffic narcotics (Bobo & Johnson, 2004).

While some have noted increases in incarceration rates driven primarily by drug law enforcement and prosecution during this era, changes in crime trends may best be explained by a complex confluence of social factors including drug law enforcement but also other elements of policing and carceral settings, drug use, economics, and demography (Blumstein & Beck, 1999; Blumstein & Wallman, 2006). Further, other scholars have noted that, in the aftermath of the War on Drugs, there may have been no real reduction in actual drug use or drug overdoses, but rather an increase in the consumption of unregulated drugs, a rise in overdose deaths, and a swelling of the structural impact of policing and incarceration on public health (Beletsky & Davis, 2017; Jalal et al., 2018; Singer, 2018; Werb et al., 2011).

While enforcement-based approaches remain a primary strategy in the US, other efforts, such as treatment-based approaches (and harm reduction approaches), have emerged across the nation. Treatment-based approaches may also include a variety of evidence-based harm reduction interventions like safe consumption sites and syringe service programs that have been implemented successfully with some public support (Kulesza et al., 2015; Levengood et al., 2021).1 More recently, increased distribution of Narcan (Naloxone), a drug that helps reverse drug overdoses, has spread nationwide (Beletsky & Davis, 2017; Britch & Walsh, 2022; Langendam et al., 2001). Recently, the federal government approved and allocated funding for over-the-counter distribution of Narcan as a key strategy to prevent drug overdose deaths (The White House, 2024). Treatment-based approaches also include medications for opioid use disorder (i.e. methadone, buprenorphine) that have been deployed to prevent overdose and also prevent other drug-related harms (e.g., HIV/HCV infection, vascular injuries, etc.). Medications for opioid use disorder (MOUD) have shown long-term effectiveness, potentially contributing to (Nadelmann & LaSalle, 2017; Prendergast et al., 2017). Increasing the access to treatment-based approaches as primary prevention efforts have shown to be effective measures to reduce drug-related harms, but these have remained underutilized, are often left as a final recourse to reduce or reverse opioid overdoses, and are especially lacking in prisons (Hawk et al., 2015). Despite this lack of implementation in communities and carceral settings, evidence-based harm reduction and treatment-based approaches to prevent overdose deaths are key components of federal strategies to combat the opioid crisis, including priorities outlined by the US Department of Health and Human Services (HHS), the US Centers for Disease Control and Prevention, and the National Institute of Health (NIH) Helping End Addiction Long-term (HEAL) initiative. While the US has tested various approaches to combating emergent drug epidemics, less is known about public sentiment towards these approaches or the modifiability of public perceptions when presented with informational messaging from key federal agencies. Public perception of strategies to prevent overdose deaths is relevant as public opinion can apply pressure to shape public health policy, priorities, and resource allocation (Enns, 2014; Manza et al., 2002). Federal agency activities and public messaging may influence how the public views drug-related policy and related enforcement practices. For example, federal agencies (e.g. DEA) charged with enforcement-based approaches to the drug supply may prioritize enforcement-based metrics, publicly highlight law enforcement practices, and potentially influence public support of enforcement-based approaches. Drug law enforcement stories, in particular those involving police misconduct and are covered by mainstream news media, may negatively affect public perceptions towards police. However, less is known about how such narratives may modify or reinforce public perceptions on enforcement-based approaches to address drug use(Succar et al., 2024). Alternatively, healthcare-oriented federal agencies (e.g. NIH) with the mission of providing healthcare access, prevention resources, and treatment to people who use drugs may contribute to public support of treatment-based strategies and/or harm reduction efforts.

Public messaging from leading federal agencies with a role in addressing drug addiction and overdose (i.e. DEA, NIH) may be relevant in shaping public support for either enforcement-based or treatment-based strategies to reduce overdose. Some studies have uncovered correlates to public support of treatment-based (and harm reduction) approaches to drug use including age, gender, political ideology, stigma towards PWUD, crime-related anxiety, and lived experience with addiction (Baker et al., 2020; Kulesza et al., 2015; Mancini & Boehme, 2024; Ruiz Flores López et al., 2025). However, no known study has experimentally contrasted messaging from prominent federal agencies about their respective drug prevention efforts and whether they modify public perceptions of overdose prevention strategies. We aimed to assess the modifiability of public support for approaches to prevent overdose after exposure to informational prompt from federal agencies. We utilized an embedded survey experiment to test whether public opinion may be impacted via exposure to statements about enforcement-related (DEA) and/or treatment-related (NIH) practices in preventing overdose deaths. Based on previous literature suggesting that the public tends to be more punitive (Costelloe et al., 2009), particularly toward drug-related crimes (Neill, 2014), we hypothesized that respondents assigned to the DEA treatment will be more supportive of enforcement-based approaches in addressing the fentanyl/opioid crisis than those assigned to the NIH treatment, and vice versa.

Methods and materials

Study setting and sample

South Carolina, a state heavily affected by drug-related harms, experienced the greatest increase in heroin-related deaths among all US states between 2010 and 2015, holds an age-adjusted mortality rate for drug overdose of 44.7 deaths per 100,00 (ranking 11th nationally) and recorded more than 100,000 overdose deaths in 2023 (Arnold et al., 2019; Florence et al., 2021; Sahebi-Fakhrabad et al., 2024; Tori & Galardi, 2024).

South Carolina is represented by race/ethnicity as 61.6% White alone, 12.4% Black alone, 18.7% Hispanic, 6% Asian alone, and 1.1% American Indian or Native American alone, and is considered a historically conservative-leaning state (U.S Census Bureau, 2023). Syringes used for injecting drugs are not considered “drug paraphernalia” in the state (Fernández-Viña et al., 2020), yet SSPs must currently operate in the absence of full legalization, presenting major barriers to scale up(Herscowitz, 2024). The state has a good Samaritan law with provisions protecting individuals, civilly and criminally, who provide care to those experiencing an overdose(“South Carolina Overdose Prevention Act,” 2015). Additionally, there is no specific drug induced homicide statute in the state, although a bill has recently been introduced to create such a statute (“S. 183: Drug-Induced Homicide,” 2025). The state’s Medicaid program also allow coverage for treatment of opioid use disorder (South Carolina Department of Health and Human Services, 2020).

South Carolina is also heavily impacted by drug enforcement-related incarceration as the drug-related arrest rate increased from 40.7 offenses per 10,000 people in 1993 to more than 94.5 per 10,000 in 2022, 82% of which were for drug possession, purchase, or personal use (South Carolina State Law Enforcement Division, 2022). However, during the opioid epidemic, a governor-led taskforce across the state focused on both prevention, response, and enforcement-based efforts, as well as treatment-based/recovery approaches (South Carolina Emergency Management Division, 2018).

We obtained a distribution list of 904,531 heads of households in the state of South Carolina with an associated email address from Mailers Haven, a third-party listserv. Mailers Haven cross-validates addresses and associated email addresses through multiple sources, including the United States Post Office, county-level public records, and publicly available data (Mailers Heaven, 2023). Mailers Haven ensures high-quality and accurate lists, with deliverability rates in the 94–97% range, offering a high degree of representation.

Data collection

The survey was developed, piloted, and administered by an interdisciplinary research team of faculty and graduate students spanning three departments across two universities. For pilot testing, the survey was sent out to about 30 individuals within our networks, in which 10 replied with comments. The comments consisted of clarity in the statements and comprehension (based on the clarity of questions). There were no perceived issues with the treatments that needed revision. The overall purpose of the survey was to capture public opinion towards various drug-related topics, including governmental and law enforcement responses to the drug crises and perceptions of drug addiction. Data collection was facilitated via e-mail with links to the Qualtrics survey between October 13 and December 13, 2023, with up to five periodic reminder emails. After data cleaning and survey dropout, a final sample of 4,675 unique respondents was obtained for analysis. Cases were excluded for the following reasons: indicating “No” to being a South Carolina resident, indicating “No” to consenting to participate in the survey, duplicate IP addresses, and/or failure to respond to the experimental portion of the survey. Sample size estimates were calculated according to the American Association of Public Opinion Research (AAPOR, n.d.) RR2 estimates, considering the number of complete interviews and partial interviews, suggested a 0.05% response rate. Certainly, the low response rate may raise concerns; however, studies indicate that low response rates may not be an indicator of nonresponse bias (Boehme et al., 2024; Pickett et al., 2018; Pickett, 2017). And although online samples like ours may still exert generalizable findings (Keeter & McGeeney, 2015), we are still cautious in making generalizable claims to the rest of the United States. Even so, with a large sample size, the sample offers enhanced confidence in the internal validity.

Research design

Using the Qualtrics randomization function, one of three experimental conditions were randomly assigned to survey respondents: (1) DEA Treatment, (2) NIH Treatment, or (3) the control condition. In the experimental treatment arms (either DEA or NIH), respondents were presented with a brief informational statement with statistics about drug/fentanyl overdoses in South Carolina and agency (either DEA or NIH) approaches to combating the drug overdoses on a page by itself. The experimental treatments were synthesized from the DEA’s and National Institute of Drug Abuse (NIH) websites and were validated through pilot-testing. The respondent would then have to click the “next” button within Qualtrics to move to the next page. Upon reaching the next page, the respondent would again see the same information from the previous page, followed by a series of Likert-scale questions asking about their perceptions of approaches to drug overdose prevention efforts (dependent variables).

Respondents in a treatment arm (either the DEA or NIH) were purposefully exposed to the informational statement twice to ensure viewing of and engagement with the treatment. Providing South Carolina-specific overdose statistics before the language about DEA or NIH approaches to preventing drug overdoses was intended to contextualize the information provided to the respondent. If respondents were randomly assigned to the control condition, they simply saw the same questions (dependent variables) as those who received treatments, with no information about drug/fentanyl overdose statistics or language about overdose prevention efforts from a federal agency. Below is what the respondents would see if assigned to one of the two treatments (either DEA or NIH):

DEA Treatment: “From 2020 to 2021, drug overdose deaths involving fentanyl increased more than 35% in South Carolina, from 1,100 to 1,494 deaths. The Drug Enforcement Administration (DEA) states that taking an ENFORCEMENT-BASED APPROACH by disrupting drug trafficking organizations and arresting drug traffickers and those for possessing drugs leads to FEWER OVERDOSE DEATHS.”

NIH Treatment: “From 2020 to 2021, drug overdose deaths involving fentanyl increased more than 35% in South Carolina, from 1,100 to 1,494 deaths. The National Institutes of Health (NIH) states that taking a TREATMENT-BASED APPROACH by providing drug users and those who have overdosed with Narcan (a drug that can help reverse overdoses) and other evidence-based treatments for substance use leads to FEWER OVERDOSE DEATHS.”

Dependent variables

Five dependent variables (DV) were analyzed assessing how much respondents agree/disagree with each statement in a Likert-scale format (1 = strongly disagree, 2 = disagree, 3 = neither disagree nor agree, 4 = agree, 5 = strongly agree). The statements are presented below:

DV1 [Label: Treatment-Based Approaches: Treatment instead of Arrest]: “The fentanyl crisis is best addressed by using a treatment-based approach where drug users are given substance abuse treatment rather than arrested and prosecuted.”

DV2 [Label: Get Tough Approaches]: “The fentanyl crisis is best addressed by using a get-tough approach where law enforcement increases their efforts to identify and arrest drug dealers and those who possess drugs.”

DV3 [Label: Clinicians Providing Treatment]: “To combat drug overdose deaths, medical professionals should prioritize providing substance abuse treatment for drug users.”

DV4 [Label: LEO Capable of Preventing ODs]: “Law enforcement officers are capable of preventing overdose deaths.”

DV5 [Label: Clinicians Capable of Preventing ODs]: “Medical professionals are capable of preventing overdose deaths.”

For the primary analysis, the five-point Likert-scale responses were collapsed into a dichotomous variable to assess respondents’ agreement with the above statements (Strongly agree/agree responses = 1, neither/disagree/strongly disagree responses = 0) (Jeong & Lee, 2016). Additionally, we implemented robustness checks to examine the raw nature of the outcomes using multinomial logistic and OLS regressions.

Independent variables

Our independent variable of interest was exposure to the experimental treatment conditions (DEA, NIH, control). For the primary analysis, the uncontrolled models assessed our set of dependent variables and exposure to either DEA or NIH treatment arms as binary variables (yes, otherwise), with the control condition variable left out as the reference category for comparison. For robustness checks, adjusted models were estimated by including relevant post-randomization socio-demographic covariates. These included age (years), race/ethnicity, sex, political affiliation (liberal, moderate, republican), education, marital status, employment status, victim of a crime in the past 12 months (Yes, No), self-reported opioid use in the past 12 months (Yes, No), a previous witness of a drug overdose (Yes, No) [See Table AI in Appendix]. Inclusion of these variables are important for assessing baseline perceptions and experiences relating to overdose prevention, while randomization to the prompt will help ensure balance among observable covariates across treatment groups.

Modelling

We conducted binary logistic regression to assess respondents’ agreement with various statements regarding drug overdose prevention strategies. For robustness checks, we built models controlling for socio-demographic covariates to sharpen model fit, adjust for any covariate imbalance, and assess the correlation between socio-demographic variables and the dependent variables (Long & Hudgens, 2013; Rochon, 1999). Multinomial logistic and OLS regressions were employed for the raw Likert-scale outcomes to ensure that findings did not change based on model selection. All analyses were conducted in Stata 18.0 (StataCorp, 2024).

Results

The final analytical sample of 4,675 respondents was somewhat older (age range 45–64), 76% White, 15% Black, 65% female, and the majority leaning from politically moderate to conservative. In our full sample (including treatment and control arms), the mean score for support of treatment-based approaches (3.31) was slightly lower than support for enforcement-based approaches (3.91). The mean score for the agreement that clinicians should prioritize substance use treatment for drug users was high (3.89) and our full sample was slightly more confident that medical professionals are capable of preventing overdose deaths (2.95) than law enforcement officers (2.65). (Table 1)

Table 1.

Summary statistics of dependent variables and by experimental condition on a sample of 4,661 South Carolinians

Raw Mean (S.D.) / Binary Mean (S.D.)
Treatment-Based Approaches 3.31 (1.26) / 0.53 (0.50)
Get Tough Approaches 3.91 (1.20) / 0.73 (0.44)
Clinicians Providing Treatment 3.89 (1.09) / 0.72 (0.45)
LEO Capable of Preventing ODs 2.65 (1.26) / 0.30 (0.46)
Clinicians Capable of Preventing ODs 2.95 (1.28) / 0.40 (0.49)
DEA Treatment NIH Treatment Control
Mean (S.D.) / Binary Mean (S.D.) Mean (S.D.) / Binary Mean (S.D.) Mean (S.D.) / Binary Mean (S.D.)
Treatment-Based Approaches 3.28 (1.27) / 0.51 (0.50) 3.30 (1.27) / 0.53 (0.50) 3.34 (1.23) / 0.55 (0.50)
Get Tough Approaches 4.03 (1.14) / 0.77 (0.42) 3.88 (1.22) / 0.72 (0.45) 3.85 (1.22) / 0.71 (0.45)
Clinicians Providing Treatment 3.87 (1.10) / 0.71 (0.46) 3.91 (1.07) / 0.72 (0.45) 3.90 (1.06) / 0.74 (0.44)
LEO Capable of Preventing ODs 2.57 (1.26) / 0.28 (0.45) 2.81 (1.25) / 0.35 (0.48) 2.57 (1.26) / 0.28 (0.45)
Clinicians Capable of Preventing ODs 2.87 (1.29) / 0.37 (0.48) 3.08 (1.26) / 0.44 (0.50) 2.90 (1.29) / 0.39 (0.49)
Open in a new tab

Notes: Raw mean is the average score of responses on a 5-point Likert-scale from 1 = strongly disagree, 2 = disagree, 3 = neither disagree nor agree, 4 = agree, and 5 = strongly agree. Binary mean = the collapsed scale into a binary scale where strongly agree and agree = 1 and neither disagree nor agree, disagree, and strongly disagree = 0. S.D. = standard deviation, DEA = Drug Enforcement Administration, NIH = National Institute of Health, ODs = overdose deaths, LEO = law enforcement officers.

Table 2 shows results from logistic regression models on the five dependent variables assessing public perceptions of various approaches to the drug/fentanyl overdose crisis. Those assigned the DEA treatment were significantly more likely to support get-tough approaches to address the fentanyl epidemic than those assigned to the NIH (Average Marginal Effects [AME] = 0.056, p = 0.001) or control (AME = 0.060, p < 0.001) arms. Those who received the DEA treatment were significantly (P = 0.046, CI -0.071, -0.001) less likely to support providing substance use treatment instead of the arrests/prosecution of drug users (AME = -0.036, p < 0.001) or agree that clinicians should prioritize substance use treatment for drug users (AME = -0.034, p < 0.001) than those in the control condition.

Table 2.

Uncontrolled experimental results on a sample of 4,661 South Carolinians

DEA Treatment NIH Treatment
AME (S.E.) 95% C.I. AME (S.E.) 95% C.I.
Treatment-Based Approaches -0.036 (0.018)* -0.071, -0.001 -0.023 (0.018) -0.057, 0.012
Get Tough Approaches 0.060 (0.016)*** 0.028, 0.092 0.004 (0.015) -0.026, 0.035
Clinicians Providing Treatment -0.034 (0.016)* -0.065, -0.002 -0.020 (0.016) -0.051, 0.012
LEO Capable of Preventing ODs -0.001 (0.017) -0.034, 0.032 0.065 (0.016)*** 0.033, 0.096
Clinicians Capable of Preventing ODs -0.030 (0.018) -0.064, 0.005 0.044 (0.017)* 0.010, 0.078
Open in a new tab

Notes: AME = average marginal effects, S.E. = standard errors, C.I. = 95% confidence intervals. ODs = overdose deaths, LEO = law enforcement officers, DEA = Drug Enforcement Administration, NIH = National Institute of Health. *** = P < 0.001, ** = P < 0.010, * P < 0.05. DEA = Drug Enforcement Administration, NIH = National Institutes of Health. Control condition serves as reference category.

As for who is capable of preventing overdoses, respondents assigned to the NIH treatment were significantly more likely to agree that both law enforcement (AME = 0.065, p < 0.001) and medical professionals (AME = 0.044, p < 0.05) are capable of preventing overdose deaths, when compared to those assigned to the control arm. This remained true in comparison to those in the DEA arm as respondents assigned the NIH treatment were significantly more likely to agree that both law enforcement (AME = 0.073, p < 0.001) and medical professionals (AME = 0.044, p < 0.05) are capable of preventing overdose deaths, when compared to the DEA arm.

Robustness checks

Several robustness checks were executed by estimating various regression models, including controlled regression analyses with socio-demographic covariates, multinomial logistic regressions examining the raw nature of the dependent variables, and OLS regressions on the raw nature of the dependent variables. Findings from the controlled model supported the above main findings after controlling for socio-demographic variables that were associated with the dependent variables. Multinomial logistic regression and OLS regression analyses also confirmed the main findings in both significance and direction. In sum, the main findings were confirmed by the robustness checks. All findings from the robustness checks can be found in the Appendix (Tables A2, A3, and A4).

Discussion and conclusion

The present study utilized a survey experiment that randomly assigned respondents one to one of three conditions: (1) DEA approach to combatting the opioid crisis, (2) NIH approach to combatting opioid crisis, or (3) the control condition. Consistent with our hypothesis, we found that those assigned the DEA treatment were significantly more likely to support get-tough approaches (i.e. identify and arrest drug dealers and people who use drugs), but less likely to support any treatment-based approaches (i.e. treatment instead of arrest/prosecution or prioritization of medical professionals providing treatment). Interestingly and contrary to our hypothesis, those assigned to the NIH treatment were no more likely to support treatment-based approaches; however, were in agreement of both law enforcement and clinicians’ capabilities in preventing drug overdose deaths.

It may come to no surprise that our overall sample had a slight preference for enforcement-based strategies or that those assigned to the DEA treatment were significantly more supportive of enforcement-based approaches to the drug epidemic. This aligns with previous criminological research that has found the public tends to have punitive attitudes towards drug users and dealers (Costelloe et al., 2009; Labor & Gastardo-Conaco, 2017; Neill, 2014), as well as research that finds support for enforcement-based approaches in combatting drug epidemics (Enns, 2014), particularly as public-health approaches are sidelined in favor for law-and-order policies (Neill, 2014). While this finding may simply align historical punitive public attitudes found in the USA, support for enforcement-focused approaches to the opioid epidemic may also emerge from the public’s fear of crime and personal safety (Reinarman & Levine, 1997). Those assigned to the DEA treatment may have supported the get-tough approach because they support this strategy due to its perceived protective effects. Additionally, respondents may have supported these efforts because the DEA approach has the potential for a deterrent effect, by ensuring dangerous drugs are removed from the community, and may also consider the potential violence that may occur from the illicit drug trade (Fellner, 2009). It is important to note that our measures did not specify who would be punished via enforcement-based approaches (e.g. people who use drugs or drug dealers). Previous research indicates that messages that either appeal to a respondent’s self-interest or emphasize the unfairness of punishment (i.e. who is punished) may be more effective in gaining public support for reform of enforcement-based approaches (Gottlieb, 2017). It is possible that our findings may have been slightly different if the DEA prompt had more specific details on the intended nature and target of enforcement-based activities.

Respondents assigned the NIH treatment, on the other hand, did not demonstrate differences in support for enforcement or treatment-based approaches. Following the COVID-19 pandemic, there was a lack of trust in health institutions (Ojikutu et al., 2022; Souvatzi et al., 2024), which may have explained, among other factors, the weaker effects of the NIH treatment on our outcomes. If the public loses trust in certain institutions, they may also lose trust that such institutions can adequately respond to pertinent issues, such as drug epidemics. Additionally, our findings suggest that in our sample, the DEA informational primer was more successful in moving responses toward enforcement-based approaches than the NIH information primer in moving responses toward treatment-based approaches. The historical investment in the U.S. to policing, incarceration, and prosecution could be influencing the lower effect of the NIH treatment to our outcome. Those assigned to the NIH arm were significantly more likely to agree that both police and medical clinicians were capable of preventing drug overdose deaths. Bachhuber and colleagues (2015) found that combining factual information and sympathetic messaging about Naloxone administration improved public support for naloxone distribution and intervening laws (Bachhuber et al., 2015). In line with the present study’s findings, increased public awareness and availability of Narcan, both in South Carolina and nationwide, may be evidence of large-scale public. Naloxone distribution and availability has increased in South Carolina over the past decade while several policies supportive of this approach have also been passed over this time (i.e. Overdose Prevention Act in 2015, Senate Bill 571 [H.5193] in 2021, and House Bill 4122 in 2023). Together, these served to expand naloxone distribution by addressing mandated prescribing of naloxone, expanding civil/criminal protections, and providing resources and guidelines for community distributors. Further, this suggests that harm reduction strategies may be gaining broader acceptance, even in politically conservative states like South Carolina where support for enforcement-based approaches may also be high.

It appears the public view both law enforcement (as first responders) and medical clinicians as likely to be either the first on-scene to administer Narcan, or, have the skills to properly administer it to those experiencing a drug overdose. Information from federal agencies like the NIH, with language about Narcan or Naloxone in particular, may influence the public’s perception of harm reduction strategies to prevent overdose. We suggest that community-based programs should be also be supported to educate citizens on its use and provide low-barrier Narcan so that people who use drugs and those nearest to them are capable of administering it to prevent overdose deaths (Shearer et al., 2018). We therefore conclude that harm reduction may be a promising avenue to address drug epidemics, which may be supported by the public(Bachhuber et al., 2015; Brennan et al., 2020; Mancini & Boehme, 2024; White et al., 2023).

Limitations & conclusion

While using an experimental design with a large sample size provides a high degree of internal validity, this study is not without its limitations. First, a sample of heads of households in South Carolina was used in the survey, indicating that the findings cannot extrapolate to those in other social roles or geographic contexts. However, concerns about the representativeness of our sample (head of households with accompanying email) are likely minimal given that previous research suggests that a sample such as ours can still elicit externally valid results (Keeter & McGeeney, 2015; Patten & Perrin, 2015). Furthermore, South Carolina is a historically conservative state with a different treatment and prevention landscape than other states, which may limit the generalizability of the findings. While it appears that the informational prompt may have nudged respondents towards prioritizing enforcement-based approaches, it could also be that this population was primed to support such approaches, given the relevant historical and political context in South Carolina as it relates to prohibitionist drug policy and the role of law enforcement. Additionally, the DEA experiment discussed both drug traffickers and those who possess drugs, which may be two distinct groups and could have swayed public support. This may explain, at least in part, why the NIH prompt was not able to significantly modify support for enforcement-based approaches. Another limitation is that the true number of respondents who received the survey in their inbox remains unknown (e.g., not filtered into junk/spam), indicating that the true response rate is likely higher than our estimate. Finally, this survey provided brief information and immediately asked respondents perceptions about policy. However, the long-term impact of the treatments remains unknown, along with the true effect of more intensive messaging by federal agencies on public perception. In other words, if given additional information about drug overdose response, it is uncertain if these sentiments would still hold or whether more robust targeted exposures to messaging may produce different results.

Notwithstanding these limitations, the current study sheds light on public perceptions of adequate responses to the opioid epidemic and align with recent studies showing that Americans have a multi-faceted view towards the opioid epidemic that in some instances support enforcement/punitive efforts but in other instances, treatment-based approaches (Desilver, 2014; Sylvester et al., 2022). We did not expect to see massive effect sizes as the experiment tested whether the treatments would update respondent priors; however, the value of this experiment was to test whether federal government agency messaging temporarily impact American’s prior beliefs. It is for future research to examine the long-term impact of federal agency messaging. Here we see that for certain policies, there are temporary changes based on federal agency messaging. Future research should continue to investigate public preferences in addressing the drug opioid crises and perceived support for various approaches. This work should inform future efforts to understand and implement effective public health messages surrounding evidence-based interventions to prevent overdose. Policymakers should engage with both scientific evidence and public perceptions in seeking a balance in policies that ensures public safety while also supporting those who use drugs in order to reduce drug overdose deaths.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1 (30.2KB, docx)

Abbreviations

AME

Average marginal effect

CI

Confidence interval

DEA

Drug Enforcement Administration

DV

Dependent variable

HCV

Hepatitis C Virus

HIV

Human Immunodeficiency Virus

LEO

Law enforcement officer

NIH

National Institutes of Health

OD

Overdose death

OLS

Ordinary least squares

SD

Standard deviation

USA

United States of America

Author contributions

L.G.G., H.B., P.L., and M.N. performed survey distribution for data collection. H.B., P.B., and P.L. performed the analysis. M.N. secured research funding. L.G.G and H.B. wrote the initial manuscript draft. All authors reviewed the manuscript.

Funding

This project was funded by the Centers for Disease Control and Prevention grant (CDC-RFA-OT21-2103): National Initiative to Address COVID-19 Health Disparities Among Populations at High-Risk and Underserved, Including Racial and Ethnic Minority Populations and Rural Communities.

Data availability

The datasets generated and/or analyzed during the current study are available in the Harvard Dataverse repository, 10.7910/DVN/KQWYLI.

Declarations

Ethics approval and consent to participate

This study was approved by the University of South Carolina Institutional Review Board (Protocol #Pro00131209). The study was performed in accordance with the Declaration of Helsinki. All participants provided informed consent prior to participating and were informed about the publication of results maintaining confidentiality of their answers.

Competing interests

The authors declare no competing interests.

Footnotes

1

Although harm reduction approaches may be seen as distinct from treatment-based approaches, we use the term “treatment-based” approaches throughout the manuscript for clarity. Both treatment-based and harm reduction efforts are in contrast to enforcement-based approaches. In sum, we used treatment-based approaches as a stark contrast to enforcement-based approaches.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

References

  1. AAPOR. (n.d.). Response rates - An overview. American Association for Public Opinion Research. Retrieved January 30 (2024). from https://www-archive.aapor.org/Education-Resources/For-Researchers/Poll-Survey-FAQ/Response-Rates-An-Overview.aspx
  2. Arnold, J. F., Arshonsky, J. H., Bloch, K. A., Holzman, E., & Sade, R. M. (2019). Opioid abuse prevention and treatment: Lessons from South Carolina. Journal of Public Health Management and Practice, 25(3), 221–228. [DOI] [PubMed] [Google Scholar]
  3. Bachhuber, M. A., McGinty, E. E., Kennedy-Hendricks, A., Niederdeppe, J., & Barry, C. L. (2015). Messaging to increase public support for Naloxone distribution policies in the United States: Results from a randomized survey experiment. PLoS One, 10(7), e0130050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Baker, L. S., Smith, W., Gulley, T., & Tomann, M. M. (2020). Community perceptions of comprehensive harm reduction programs and stigma towards people who inject drugs in rural Virginia. Journal of Community Health, 45, 239–244. [DOI] [PubMed] [Google Scholar]
  5. Beletsky, L., & Davis, C. S. (2017). Today’s Fentanyl crisis: Prohibition’s Iron law, revisited. International Journal of Drug Policy, 46, 156–159. [DOI] [PubMed] [Google Scholar]
  6. Blumstein, A., & Beck, A. J. (1999). Population growth in US prisons, 1980–1996. Crime and Justice, 26, 17–61. [Google Scholar]
  7. Blumstein, A., & Wallman, J. (2006). The crime drop and beyond. Annu Rev Law Soc Sci, 2(1), 125–146. [Google Scholar]
  8. Bobo, L. D., & Johnson, D. (2004). A taste for punishment: Black and white Americans’ views on the death penalty and the war on drugs. Du Bois Review: Social Science Research on Race, 1(1), 151–180. [Google Scholar]
  9. Boehme, H. M., Adams, I. T., Metcalfe, C., Leasure, P., & Nolan, M. S. (2024). Does scientific research change minds? Linking criminology and public perceptions of policing. Criminology & Public Policy, 23(1), 201–224. 10.1111/1745-9133.12644
  10. Brennan, E., Schoenaker, D. A., Durkin, S. J., Dunstone, K., Dixon, H. G., Slater, M. D., Pettigrew, S., & Wakefield, M. A. (2020). Comparing responses to public health and industry-funded alcohol harm reduction advertisements: An experimental study. BMJ Open, 10(9), e035569. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Britch, S. C., & Walsh, S. L. (2022). Treatment of opioid overdose: Current approaches and recent advances. Psychopharmacology (Berl), 239(7), 2063–2081. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. U.S Census Bureau (2023). Total population datahttps://data.census.gov/
  13. Center for Disease Control and Prevention (2023). Understanding the opioid overdose epidemic. Retrieved January 4 2024 from https://www.cdc.gov/opioids/basics/epidemic.html
  14. Centers for Disease Control and Prevention (2023). Fentanyl. Retrieved January 4 2024 from https://www.cdc.gov/opioids/basics/fentanyl.html
  15. Cooper, H. L. (2015). War on drugs policing and police brutality. Substance Use & Misuse, 50(8–9), 1188–1194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Costelloe, M. T., Chiricos, T., & Gertz, M. (2009). Punitive attitudes toward criminals: Exploring the relevance of crime salience and economic insecurity. Punishment & Society, 11(1), 25–49. [Google Scholar]
  17. Desilver, D. (2014). Feds may be rethinking the drug war, but states have been leading the way. Pew Research Center, April, 2, 3014.
  18. Enns, P. K. (2014). The public’s increasing punitiveness and its influence on mass incarceration in the United States. American Journal of Political Science, 58(4), 857–872. [Google Scholar]
  19. Fellner, J. (2009). Race, drugs, and law enforcement in the United States. Stan L & Pol’y Rev, 20, 257. [Google Scholar]
  20. Fernández-Viña, M. H., Prood, N. E., Herpolsheimer, A., Waimberg, J., & Burris, S. (2020). State laws governing syringe services programs and participant syringe possession, 2014–2019. Public Health Reports, 135(1_suppl), 128S–137S. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Florence, C., Luo, F., & Rice, K. (2021). The economic burden of opioid use disorder and fatal opioid overdose in the United States, 2017. Drug and alcohol dependence, 218, 108350. [DOI] [PMC free article] [PubMed]
  22. Friedman, J., & Shover, C. L. (2023). Charting the fourth wave: Geographic, temporal, race/ethnicity and demographic trends in polysubstance fentanyl overdose deaths in the United States, 2010–2021. Addiction, 118(12), 2477–2485. [DOI] [PubMed] [Google Scholar]
  23. Gottlieb, A. (2017). The effect of message frames on public attitudes toward criminal justice reform for nonviolent offenses. Crime & Delinquency, 63(5), 636–656. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Han, Y., Yan, W., Zheng, Y., Khan, M. Z., Yuan, K., & Lu, L. (2019). The rising crisis of illicit fentanyl use, overdose, and potential therapeutic strategies. Translational Psychiatry, 9(1), 282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Hargan, E. D. (2017). Determination that a public health emergency exists. Department of Health and Human Services.
  26. Hawk, K. F., Vaca, F. E., & D’Onofrio, G. (2015). Reducing fatal opioid overdose: Prevention, treatment and harm reduction strategies. The Yale Journal of Biology and Medicine, 88(3), 235–245. [PMC free article] [PubMed] [Google Scholar]
  27. Herscowitz, E. (2024). Are syringe service programs legal in SC? It’s complicated. Post and Courier. https://www.postandcourier.com/health/harm-reduction-clean-syringes-legal-south-carolina/article_d907424c-0702-11ef-acb5-87ea1b7e4caa.html
  28. Hinton, E. (2016). From the war on poverty to the war on crime: The making of mass incarceration in America. Harvard University Press.
  29. Jalal, H., Buchanich, J. M., Roberts, M. S., Balmert, L. C., Zhang, K., & Burke, D. S. (2018). Changing dynamics of the drug overdose epidemic in the United States from 1979 through 2016. Science, 361(6408), eaau1184. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Jenkins, R. A. (2021). The fourth wave of the US opioid epidemic and its implications for the rural US: A federal perspective. Preventive Medicine, 152, 106541. [DOI] [PubMed] [Google Scholar]
  31. Jeong, H. J., & Lee, W. C. (2016). The level of collapse we are allowed: Comparison of different response scales in safety attitudes questionnaire. Biom Biostat Int J, 4(4), 00100. [Google Scholar]
  32. Keeter, S., & McGeeney, K. (2015). Coverage error in internet surveys.
  33. Kerr, T., Small, W., & Wood, E. (2005). The public health and social impacts of drug market enforcement: A review of the evidence. International Journal of Drug Policy, 16(4), 210–220. [Google Scholar]
  34. Kulesza, M., Teachman, B. A., Werntz, A. J., Gasser, M. L., & Lindgren, K. P. (2015). Correlates of public support toward federal funding for harm reduction strategies. Substance Abuse Treatment, Prevention, and Policy, 10, 25. 10.1186/s13011-015-0022-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Labor, P. D. P., & Gastardo-Conaco, M. C. C. (2017). The role of social perceptions, beliefs, and emotions on support for punitive action toward drug dealers and users. Philippine Journal of Psychology, 50(2), 67–96. [Google Scholar]
  36. Langendam, M. W., van Brussel, G. H., Coutinho, R. A., & Van Ameijden, E. (2001). The impact of harm-reduction-based methadone treatment on mortality among heroin users. American Journal of Public Health, 91(5), 774. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Levengood, T. W., Yoon, G. H., Davoust, M. J., Ogden, S. N., Marshall, B. D. L., Cahill, S. R., & Bazzi, A. R. (2021). Supervised injection facilities as harm reduction: A systematic review. American Journal of Preventive Medicine, 61(5), 738–749. 10.1016/j.amepre.2021.04.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Long, D. M., & Hudgens, M. G. (2013). Sharpening bounds on principal effects with covariates. Biometrics, 69(4), 812–819. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Lynch, M. (2012). Theorizing the role of the ‘war on drugs’ in US punishment. Theoretical Criminology, 16(2), 175–199. [Google Scholar]
  40. Mailers Heaven (2023). The highest quality: Most responsive mailing lists.https://www.mailershaven.com/
  41. Mancini, C., & Boehme, H. M. (2024). Public preferences for opioid reform in a post-COVID era: Consensus and divides. Journal of Criminal Justice, 93, 102210. 10.1016/j.jcrimjus.2024.102210
  42. Manza, J., & Cook, F. L. (2002). The impact of public opinion on public policy: The state of the debate. In F. L. Cook (Ed.), Navigating public opinion: Polls, policy, and the future of American democracy. Oxford University Press.
  43. Meeks, D. (2006). Police militarization in urban areas: The obscure war against the underclass. The Black Scholar, 35(4), 33–41. [Google Scholar]
  44. Nadelmann, E., & LaSalle, L. (2017). Two steps forward, one step back: Current harm reduction policy and politics in the United States. Harm Reduction Journal, 14, 1–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Neill, K. A. (2014). Tough on drugs: Law and order dominance and the neglect of public health in US drug policy. World Medical & Health Policy, 6(4), 375–394. [Google Scholar]
  46. Ojikutu, B. O., Bogart, L. M., & Dong, L. (2022). Mistrust, empowerment, and structural change: Lessons we should be learning from COVID-19 (Vol. 112, pp. 401–404). American Public Health Association. [DOI] [PMC free article] [PubMed]
  47. Patten, E., & Perrin, A. (2015). Who’s left out in a web-only survey and how it affects results.
  48. Pickett, J. T. (2017). Methodological myths and the role of appeals in criminal justice journals: The case of response rates. ACJS Today, 41(3), 61–69. [Google Scholar]
  49. Pickett, J., Cullen, F., Bushway, S. D., Chiricos, T., & Alpert, G. (2018). The response rate test: Nonresponse bias and the future of survey research in criminology and criminal justice. Available at SSRN 3103018.
  50. Prendergast, M. L., Podus, D., Chang, E., & Urada, D. (2017). The effectiveness of drug abuse treatment: A meta-analysis of comparison group studies⋆. In M. Natarajan (Ed.), Drug abuse: Prevention and treatment (pp. 279–298). Routledge. [DOI] [PubMed]
  51. Pyra, M., Taylor, B., Flanagan, E., Hotton, A., Johnson, O. D., Lamuda, P., Schneider, J., & Pollack, H. A. (2022). Support for evidence-informed opioid policies and interventions: The role of Racial attitudes, political affiliation, and opioid stigma. Preventive Medicine, 158, 107034. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Reinarman, C., & Levine, H. G. (1997). Crack in america: Demon drugs and social justice. Univ of California.
  53. Roberts, S. K. (2022). The impact of the US drug war on people of color. In P. Gootenberg (Ed.), Oxford handbook of global drug history (pp. 474–494). Oxford University Press.
  54. Rochon, J. (1999). Issues in adjusting for covariates arising postrandomization in clinical trials. Drug Information Journal, 33(4), 1219–1228. [Google Scholar]
  55. Ruiz Flores López, P., Sánchez Hiza, Ó., Zizumbo-Colunga, D., Lathrop, C., R., & Vásquez Salazar, J. C. (2025). What sways public opinion? An experimental study on moral vs. evidence-based arguments for drug harm reduction. Drugs: Education Prevention and Policy, 1–11. 10.1080/09687637.2024.2440413
  56. S. 183: Drug-Induced Homicide, S. 183, South Carolina General Assembly (2025). https://www.scstatehouse.gov/sess126_2025-2026/bills/183.htm#:~:text=(A) A person who knowingly,other person is the injection%2 C
  57. Sahebi-Fakhrabad, A., Sadeghi, A. H., Kemahlioglu-Ziya, E., & Handfield, R. (2024). Exploring opioid prescription patterns and overdose rates in South Carolina (2017–2021): Insights into rising deaths in High-Risk areas. Healthcare. [DOI] [PMC free article] [PubMed]
  58. Shearer, D., Fleming, T., Boyd, J., & McNeil, R. (2018). Naloxone distribution, trauma, and supporting community-based overdose responders. The International Journal on Drug Policy, 74, 255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Singer, J. A. (2018). Harm reduction: Shifting from a war on drugs to a war on drug-related deaths. Cato Institute.
  60. South Carolina State Law Enforcement Division (2022). Crime in South Carolina Annual Report. https://www.sled.sc.gov/forms/statistics/2022 Crime in South Carolina.pdf
  61. South Carolina Department of Health and Human Services (2020). May 1 2020). Medicaid Coverage for Treatment of Opioid Use Disorder. https://www.scdhhs.gov/communications/medicaid-coverage-treatment-opioid-use-disorder
  62. South Carolina Emergency Management Division (2018). Gov. Henry McMaster Announces Opioid Emergency Response Plan. https://www.scemd.org/news/gov-henry-mcmaster-announces-opioid-emergency-response-plan/
  63. South Carolina Overdose Prevention Act (2015). Title 44, chap. 130 § §§ 44-130-10 to 44-130-75. South Carolina Legislature. https://www.scstatehouse.gov/code/t44c130.php
  64. Souvatzi, E., Katsikidou, M., Arvaniti, A., Plakias, S., Tsiakiri, A., & Samakouri, M. (2024). Trust in healthcare, medical mistrust, and health outcomes in times of health crisis: A narrative review. Societies, 14(12), 269. [Google Scholar]
  65. StataCorp, L. (2024). Stata Statistical Software: Release 18.
  66. Succar, R., Ramallo, S., Das, R., Ventura, R. B., & Porfiri, M. (2024). Understanding the role of media in the formation of public sentiment towards the police. Communications Psychology, 2(1), 11. 10.1038/s44271-024-00059-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Sylvester, S. M., Haeder, S. F., & Callaghan, T. (2022). Just say no? Public attitudes about supportive and punitive policies to combat the opioid epidemic. Journal of Public Policy, 42(2), 270–297. [Google Scholar]
  68. The White House. (2024). FACT SHEET: Biden-Harris administration launches the white house challenge to save lives from overdose.:text=This%20includes%3A,and%20pharmacies%20across%20the%20country. https://www.whitehouse.gov/briefing-room/statements-releases/2024/03/13/fact-sheet-biden-harris-administration-launches-the-white-house-challenge-to-save-lives-from-overdose/#:~
  69. Tori, M., & Galardi, C. (2024). Trends in suspected opioid drug overdose events in emergency medical service encounters in South Carolina, 2019–2022. Public Health Reports®, 139(6), 684–690. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Werb, D., Rowell, G., Guyatt, G., Kerr, T., Montaner, J., & Wood, E. (2011). Effect of drug law enforcement on drug market violence: A systematic review. International Journal of Drug Policy, 22(2), 87–94. [DOI] [PubMed] [Google Scholar]
  71. White, S. A., Lee, R., Kennedy-Hendricks, A., Sherman, S. G., & McGinty, E. E. (2023). Perspectives of U.S. harm reduction advocates on persuasive message strategies. Harm Reduction Journal, 20(1), 112. 10.1186/s12954-023-00849-z [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (30.2KB, docx)

Data Availability Statement

The datasets generated and/or analyzed during the current study are available in the Harvard Dataverse repository, 10.7910/DVN/KQWYLI.

Articles from Health & Justice are provided here courtesy of BMC

RESOURCES