Do Normative Issues Outperform Deterrence for Compliance With Drug Laws?

Abstract

This article tests the impact of deterrence and normative variables in the estimation of drug use among both detected drug users and university students. Consistent with perceptual deterrence research, we expect: (a) a stronger effect of normative variables on estimation of drug use compared to deterrence variables and (b) that any deterrent effect would be restricted to detected drug users. In Portugal, we conducted a cross-sectional survey with detected drug users (N = 147) and with a sample of university students (N = 247). The main results show that normative variables outperform deterrent variables as predictors of intention to use drugs. Descriptive norms have a greater influence on the estimation of drug use in both groups, and proximal social norms have robust effects among university students. Core elements of the deterrence model are not related to the intention to use drugs. Personal certainty of being caught has even an unexpected positive correlation with this intention. A small deterrent effect is limited to the general certainty variable and only among more law-abiding people. The overall findings are consistent with a normative perspective on law compliance and highlight the importance of including all different normative and deterrent variables in models estimating drug use.

Keywords

perceptual deterrence drug use compliance norms propensity

Introduction

In the last century, according to international conventions ratified by the vast majority of countries, the use of illicit drugs or the possession of drugs for personal use has typically been prosecuted by the criminal justice system. Only a few European countries—such as Italy, Spain, Portugal, and the Czech Republic—as well as some American and Australian states have decriminalized drug use. Although the UN endorsing decriminalization as a viable reform strategy, prohibitionist laws and their enforcement continue to be the status quo approach for preventing drug use around the world.¹ However, recently a number of jurisdictions, including Canada, Uruguay, and some US states, have approved cannabis legalization laws, allowing recreational use (and production and retail sale) by adults (EMCDDA, 2020; Kamin, 2017; Pardo, 2014).² In other jurisdictions, and for all other illicit drugs in these jurisdictions, deterrence remains the desired effect of drug prohibition.

Due to the decriminalization of all drugs in small quantities, Portuguese drug policy is generally perceived as less restrictive. On July 1, 2001, the Decriminalization Law (30/2000) entered into force, specifically mentioning “the health and social protection of people who use drugs without a medical prescription.” Possession of small amounts of drugs for use is now considered an administrative offense. Detected drug users are referred to the Commission for the Dissuasion of Drug Addiction (Comissão para a Dissuasão da Toxicodependência [CDT]). The CDT should prefer to choose one of the suspension options that replaced sanctions enforcement (fines, non-financial sanctions, or warnings). Although the decriminalization law explicitly revoked the crime of consumption, the Supreme Court of Justice later defended (Judgment 8/2008) the maintenance of drug possession for use as a crime “when the amount detected is higher than the average individual consumption during the 10-day period.” As drug use remains illegal and even criminal in some circumstances, deterrence remains a goal of Portuguese drug laws (Quintas, 2011).

The classical doctrine of deterrence (Andenaes, 1974; Beccaria, [1766] 1998; Gibbs, 1975, 1977; Zimring & Hawkins, 1973) postulates the threat of legal sanctions as a way to deter crime or other offenses, because it views human beings as hedonistic and motivated by self-interest. As Paternoster (2010) argues “the concept of deterrence is quite simple—it is the omission of a criminal act because of the fear of sanctions or punishment” (p. 766). Deterrence theory is a perceptual theory that “requires individual-level data on how punishments play out in people's minds” (Pratt & Turanovic, 2018, p. 189). The threat of a sanction is a subjective perception that results from an individual's assessment of three classic variables: certainty (e.g., the perception of being caught), celerity (e.g., the swiftness of the sanction), and severity (e.g., the amount of the sanction). Typically, perceptual deterrence research uses surveys to assess the relationship between one or more of these variables and the estimated likelihood of lawbreaking decisions.

The instrumental effect of the famous three deterrence variables applies to virtually all types of illegal behavior (e.g., property and violent crime, tax evasion, drunk driving, environmental violations, copyright infringement, music piracy, illegal fishing, or drug use). The consensus in research reviews is that sanction threat perceptions are weakly related to offenses. At most, criminological research only reports a modest “certainty effect” on unlawful decisions (Apel & Nagin, 2011; Beyleveld, 1980; Dölling et al., 2009; Loughran et al., 2012; Nagin, 1998; Nagin & Pogarsky, 2003; Paternoster, 1987, 2010, 2018; Pratt et al., 2006; Yu & Liska, 1993). Severity is typically not correlated with offenses (Kennedy, 2009; Nagin, 2018; Paternoster, 1987) and celerity is a forgotten variable (for exceptions see Nagin & Pogarsky, 2001 or Pratt & Turanovic, 2018). Despite this consensus, large differences are expected in the weight of deterrent variables when applied to specific offenses, unique individuals, and different sanction threats. Despite this consensus, large differences in the weight of deterrence variables can be expected when applied to specific crimes, unique individuals, and different sanction threats. For example, a comprehensive meta-analytic review (Dolling et al., 2011) confirms that deterrent effects are more likely to be found for petty crimes than for violent crimes. What deserves scientific analysis are the perceived changes in the likelihood of sanctions and whether these perceptions trigger actual changes in behavior under certain conditions (Jacobs & Piquero, 2013).

Perceptual deterrence research also includes some so-called “extra-legal” variables to control for the relationship between sanction risk and lawbreaking decisions or self-reported behavior. The most important of these variables are normative, including personal norms and social norms. Personal norms are typically thought of as an individual's internalized ethics or set of moral standards. Social norms are shared values within a social group and are derived from external standards. Furthermore, social norms can be proximal (such as those observed among close friends or family members) or distant (such as the general rules of a society). Finally, social norms can be divided into descriptive norms (perceptions of how people actually behave) and injunctive norms (perceptions of how people think others should behave) (Cialdini et al., 1991). All of these different normative variables tend to be stronger predictors of behavior than certainty estimates and typically render the ability of any deterrent variable to explain compliance at the multivariate level insignificant or at most negligible (Foglia, 1997; Gallupe & Baron, 2014; MacCoun, 1993; Nagin, 1998; Nagin & Pogarsky, 2001; Paternoster, 1987; Piquero et al., 2016).

Moreover, deterrence and normative variables are not independent sources of behavioral influence. For example, Wenzel (2004) found a complex interaction between norms and deterrence in explaining tax-paying behavior. Personal and social norms are both moderators of deterrence and constrain the relevance of deterrence. Regarding personal norms, deterrence effects are only important for a weak internalized norm condition, since people with internalized norms against tax evasion would be unaffected by sanction estimates. Regarding social norms, they had a moderating effect on the severity of sanctions. Only when social norms were strongly perceived to support tax honesty was sanction severity inversely correlated with tax evasion. These results support the great relevance of normative variables and, also, their differential interaction with deterrence variables.

Finally, it is important to understand how the threat of punishment varies according to a person's motivation or propensity to commit a crime or engage in other deviant behavior. Wright et al. (2004) examined this issue and found the highest deterrent effect “among study participants low in self-control and high in self-perceived criminality” (p. 206). This finding contradicts self-control theory (Gottfredson & Hirschi, 1990) and the traditional perceptual deterrence assumption that all people are equally affected by deterrence. However, the result is consistent with the large body of research showing the irrelevance of deterrence variables when other inhibitions are strong. For example, situational action theory (SAT) research has consistently found that deterrence is more likely to reduce crime among individuals with low morality. Wikstrom et al. (2011) found that people who rarely thought about committing a crime (or other illegal behavior) reported that sanction risk was not based on personal experience, and “regardless of what views they hold about the risk of punishment, these are irrelevant for their compliance with the law” (p. 402). Perceived sanction risk only affects individuals with a high individual propensity to offend (Pauwels et al., 2011), and law-abiding individuals are unaffected by arrest risk assessments. The restraint of moral inhibitions and, in particular, proximal social norms against drug use may be so strong that they preclude consideration of instrumental concerns such as the risk of being caught and suffering formal sanctions (Weatherburn et al., 2000). Conversely, when moral inhibitions are less pronounced, it is the interaction between the setting and the person (as defined by the SAT) that is likely to lead the individual to choose particular action alternatives, allowing for an expected increased power of deterrent variables to influence behavior.

Thus, we expect that people will be differentially affected by perceptions of deterrence in their lawbreaking decisions. Moreover, these deterrent effects may be conditioned by different normative backgrounds.

Current Focus

The first aim is to test the relative importance of deterrence and normative variables in expected future drug use. Directly following the traditional deterrence model, we expect a negative, albeit small, relationship between classical deterrence variables (especially certainty) and expected future drug use. However, in line with contemporary perceptual deterrence research, we also hypothesize a stronger relationship between normative variables and expected future drug use. Thus, we expect the normative variables to outperform the deterrence variables in the regression analysis.

The second aim is to determine the conditions under which deterrence variables influence behavioral intentions. We examine the effects of deterrence on the expected future drug use as a function of an individual's experience with drugs. The sample consists of two groups: detected drug users and university students, most of whom are abstainers. We hypothesize that any deterrent effect is limited to detected drug users. Conversely, we do not expect any deterrent effect among university students, whose behavior should be mostly influenced by norms against drug use.

Method

Participants

Participants were recruited from two contexts in order to select contrasting groups in terms of their propensity to use drugs. The first group, composed of detected drug users (DDU), was recruited at the Commission for the Dissuasion of Drug Dependence (Comissão para a Dissuasão da Toxicodependência—CDT), a Portuguese administrative agency that takes legal action against drug users caught using or possessing small amounts of drugs. They are mainly young adults (N = 145, M age = 23.5 years, SD = 5.52). The second group consists of slightly younger students (N = 247, M age = 20.4 years, SD = 2.98) from universities in the North of Portugal, enrolled in criminology (34.0%), psychology (30.4%), and law (35.6%) courses. The vast majority had abstained from drug use in the past year (81.3%).

Procedure

The data used for this study was part of a larger study of the Portuguese law on the decriminalization of drug use.³ After obtaining the necessary permissions, participants completed a paper-and-pencil questionnaire in private CDT facilities or university classrooms. Participation in the study was anonymous and confidentiality was assured. Only occasionally did participants seek help from the interviewers if they had difficulty understanding.

Materials and Measures

Participants were asked about behavioral, deterrent, and normative variables, inspired in part by perceptual deterrence research, particularly the questions in the Paternoster and Piquero (1995) appendix. Drug use behavior is measured by past self-report and expected future. The deterrence construct is decomposed into several elements: personal certainty, general certainty, severity, and celerity. The normative construct is measured by personal, social (proximal and distant), and descriptive norms. Other variables measured are sanction legitimacy, informal sanctions, and perceptions of the health risks of drug use.

Dependent variable. The dependent variable in this study is the expected future drug use. We asked, on a seven-point Likert-type scale (1 = low to 7 = high), the likelihood of using drugs in the next year.

Deterrence variables. Perceived risk was assessed with questions that also asked about likelihood or probability. To measure certainty, we asked participants to think about their personal likelihood of being caught and about the general likelihood of a drug user being caught. Thus, we asked participants separately how likely or unlikely (1 = very unlikely, 7 = very likely) it was that (a) they would be caught for using drugs (personal certainty) and (b) other people who use drugs would be caught (general certainty). As a measure of the severity of the sanction for drug use, an estimate of the possible fine is requested. Finally, as a measure of celerity, we used another seven-point Likert scale (1 = disagree, 7 = agree) to ask whether the fine would be paid in a short time. Higher scores on these variables indicate a greater perception of deterrence.

Normative variables. To examine respondents’ personal moral beliefs, participants were asked to indicate on a seven-point Likert scale (1 = disagree to 7 = agree) their moral acceptance of drug use (personal norm). Social norms were assessed using the same scale, and participants were asked separately about (a) others’ approval of drug use (distant social norms) and (b) relatives’ approval of drug use (proximal social norms). A seven-point Likert-type scale (1 = low to 7 = high) is used to measure descriptive norms through a question about “others’ estimation of next year's drug use.” Higher scores on each of these variables would indicate greater normative support for drug use.

Control variables. The seven-point Likert scale (1 = disagree, 7 = agree) was used to ask respondents about their opinions on the legitimacy of sanctions, the health risks of drug use, and informal sanctions. Respondents were asked if “punishment for drug use is fair,” if “drug use is a health risk,” and if “I lose respect from others if I use drugs.”

Analytical Strategy

First, we examined descriptive statistics for all variables. Then we have tried to construct deterrence and normative indices. However, the reliability analysis shows that it is inadequate to create a normative index with personal, social, and descriptive norms (α Cronbach = .44) and a deterrence index with certainty, severity, and celerity variables (α Cronbach = .00). Preliminary analyses supported keeping all constructs separate rather than combining them into deterrence or normative scales.

Descriptive data are presented separately for the DDU and the university students, along with the respective difference tests. Zero-order Spearman correlations between expected future drug use and deterrent, normative, and control variables are also examined.

Linear regression analyses were conducted to determine the predictors of expected future drug use. To reduce multicollinearity, all Likert scale variables were z-standardized prior to regression analysis (Field, 2009; Weisburd & Britt, 2007). In addition, all variables have a variance inflation factor (VIF) less than 2, suggesting that there is no multicollinearity problem with these data. Three models are tested: the first shows the main effects of the deterrent variables, the second model adds the normative variables, and the third model adds all control variables.

Results

Descriptive statistics and tests for differences between the DDU and student samples are presented in Table 1. Students report a very low intention to use drugs in the next year (M = 1.87, SD = 1.63), which contrasts significantly with a higher expected future drug use (although below the midpoint of the scale) among the DDU sample (M = 3.31, SD = 1.97).

Table 1.

Descriptive Data, Differences Between Detected Drug Users and University Students and Correlations with Next Year Drug Use Estimation.

	Descriptive data and differences tests							Correlations with drug use estimation
	Detected drug users			University students				Detected drug users		University students
Variable	n	M	DP	n	M	DP	P	r	P	r	P
Drug use estimation	147	3.31	1.97	246	1.87	1.63	<.001	−		−
Certainty (personal)	147	1.90	1.34	246	1.22	.77	<.001	.43	<.001	.46	<.001
Certainty (general)	147	4.26	1.89	245	3.69	1.71	.002	.26	.002	−.12	.065
Severity (fine in €)	109	66	128	190	349	414	<.001	.07	.460	−.10	.178
Celerity	138	3.95	1.73	242	3.52	1.33	.008	−.07	.425	−.09	.166
Personal norms	144	3.60	1.74	247	2.71	1.86	<.001	.27	.001	.40	<.001
Proximal social norms	146	2.53	1.84	243	1.68	1.20	<.001	.20	.016	.55	<.001
Distant social norms	147	3.64	1.87	246	3.07	1.69	.002	.03	.767	.13	.042
Descriptive norms	145	4.46	1.86	245	4.83	1.94	.070	.53	<.001	.27	<.001
Informal sanctions	145	2.83	2.03	244	3.59	2.00	<.001	−.31	<.001	−.20	.002
Sanction legitimacy	144	3.56	2.18	246	4.61	2.01	<.001	−.31	<.001	−.30	<.001
Health risks of drug use	146	6.10	4.32	246	6.25	1.32	.681	.11	.183	−.30	<.001

Scales 1–7, unless otherwise indicated.

The personal risk of arrest is perceived to be very low among DDUs and almost zero among students. However, the general estimation of certainty and the perception of the speed of the authorities’ action tend to be higher, with the mean values closer to the middle of the scale in both samples. For these three deterrence variables, the DDU averages are significantly higher than those of the students. Otherwise, the severity of the sanction is perceived as higher by the students. On average, they estimate the fine to be €349, an amount significantly higher than the more accurate mean estimate of €66 by the DDU.

The descriptive statistics of the normative variables show that the participants and their backgrounds are generally not in agreement with drug use. Personal norms (M = 2.71, SD = 1.86), distant social norms (M = 3.07; SD = 1.69), and especially proximal social norms (M = 1.68; SD = 1.40) are in opposition to drug use among students. The mean values of these three normative variables are significantly higher in the DDU sample, although they remain below the midpoint of the scale. However, the descriptive norm has much higher means in both samples, indicating that participants often perceive other people's drug use as widespread.

We also found among students a moderate disagreement with the possibility of receiving informal sanctions for drug use (M = 3.59; SD = 1.98), a moderate agreement with sanction legitimacy (M = 4.61; SD = 2.01), and a massive perception of drug use as a health risk behavior (M = 6.23; SD = 1.32). DDU tends to express significantly more disagreement with the perception of receiving informal sanctions for drug use (M = 2.83; SD = 2.03), disagreement with the legitimacy of sanctions (M = 3.56; SD = 2.18), and less perception of drug use as a health risk behavior (M = 6.10; SD = 4.32), although the mean for the latter variable is still very high.

Table 1 also reports zero-order correlations between each of the independent variables included in the study and the expected future drug use. For the deterrent variables, most correlation coefficients with prospective drug use are consistent with contemporary tests of deterrence theory and inconsistent with intuitive claims of deterrence doctrine. Severity and celerity, classical deterrence variables, are not significantly associated with expected future drug use in either sample. Intention to use drugs appears to be independent of assessing the amount and swiftness of sanctions for drug use.

Nevertheless, we found one crucial exception: personal certainty estimation is positively associated with expected future drug use among DDU (r = .43) and students (r = .46). This anomalous finding in perceptual deterrence research is largely explained by low estimates of both variables. A low personal risk of arrest is a consequence of low reported expected future drug use for most participants. However, for both samples, an increased risk of arrest does not have a deterrent effect. When participants estimate a higher likelihood of using drugs in the next year, they also believe that drug use will be more easily detected.

Conversely, general certainty estimation is significantly associated with the likelihood of future drug use, positively among DDU (r = .26). DDU does not seem to be deterred by an increased perception of an arrest risk for any drug user. Inversely, general certainty estimation is not significantly correlated with expected future drug use among students (r = −.12).

Future expected drug use is, as hypothesized, positively correlated with normative variables, especially with the descriptive norms (r = .53) and the personal norms (r = .27) among the DDU, and with the proximal social norms (r = .55) and the personal norms (r = .40) in the student sample. DDU shows a greater willingness to use drugs according to the descriptive norms and their moral attitudes toward consumption. Students with more favorable moral attitudes and proximal social background toward drug use tend to report intention to use drugs. Distant social norms are, in both groups, the normative variable less related to behavioral expectation estimation, and it is even irrelevant for DDU.

Furthermore, informal sanctions, sanction legitimacy, and perceived risks of drug use are negatively correlated with prospective drug use, except for the latter among DDU.

At the multivariate level, we performed two separate multiple linear regression analyses, for the DDU and student samples, to explain the estimation of drug use variance (see Table 2). For each group, the first model included the deterrent variables. Model 2 adds all the normative variables, and model 3 tests the effect of the control variables.

Table 2.

Predictors of Drug Use Estimation (Regression Analysis Summary).

Detected drug users (n = 100)
	Model 1 R² = .239; Adj R² = .207 F(4,95) = 7.442; P < .001)			Model 2 R² = .510; Adj R² = .467 F(8,91) = 11.844;P < .001)			Model 3 R² = .540; Adj R² = .482 F(11,88) = 9.388; P < .001)
Variable	B(SE)	β	P	B(SE)	β	P	B(SE)	β	P
(constant)	.296 (.172)		.089	.503 (.146)		<.001	.515 (.145)		<.001
Certainty (personal)	.305 (.077)	.383	<.001	.206 (.066)	.259	.002	.160 (.068)	.201	.021
Certainty (general)	.203 (.088)	.214	.024	−.128 (.096)	−.135	.184	−.050 (.100)	−.053	.618
Severity (fine in €)	−.073 (.268)	−.025	.786	.153 (.234)	.053	.514	.277 (.236)	.096	.245
Celerity	−.099 (.081)	−.110	.225	−.026 (.069)	−.029	.709	−.048 (.069)	−.054	.483
Personal norms				.181 (.100)	.158	.075	.134 (.101)	.117	.187
Proximal social norms				−.010 (.083)	−.010.	.905	−.020 (.084)	−.021	.808
Distant social norms				.033 (.072)	.034	.651	.043 (.072)	.045	.550
Descriptive norms				.616 (.115)	.588	<.001	.539 (.118)	.514	<.001
Informal sanctions							−.120 (.088)	−.116	.175
Sanction legitimacy							−.144 (.083)	−.142	.087
Health risks of drug use							.018 (.043)	.031	.676
University students(n = 183)
	Model 1 R² = .248; Adj R² = .231 F(4,178) = 14.664; P < .001)			Model 2 R² = .528; Adj R² = .506 F(8,174) = 24.343;P < .001)			Model 3 R² = .539; Adj R² = .509 F(11,171) = 18.145; P < .001)
Variable	B(SE)	β	P	B(SE)	β	P	B(SE)	β	P
(constant)	−.118 (.063)		.061	−.122 (.051)		.019	−.117 (.052)		.026
Certainty (personal)	.606 (.086)	.465	<.001	.473 (.072)	.363	<.001	.459 (.073)	.352	<.001
Certainty (general)	−209 (.060)	−.228	<.001	−.264 (.053)	−.288	<.001	−.235 (.057)	−.257	<.001
Severity (fine in €)	−.088 (.049)	−.117	.073	−.012 (.040)	−.016	.762	−.005 (.040)	−.007	.903
Celerity	.040 (.060)	.044	.505	.015 (.050)	.016	.769	.018 (.050)	.020	.715
Personal norms				.120 (.050)	.138	.016	.080 (.055)	.092	.148
Proximal social norms				.340 (.055)	.356	<.001	.331 (.055)	.347	<.001
Distant social norms				−.010 (.049)	−.011	.842	−.011 (.049)	−.012	.820
Descriptive norms				.268 (.051)	.305	<.001	.273 (.051)	.311	<.001
Informal sanctions							−.017 (.050)	−.019	.736
Sanction legitimacy							−.051 (.058)	−.055	.382
Health risks of drug use							−.149 (.103)	−.084	.149

In the DDU sample, Model 1 accounts for 24% of the variance. Personal certainty has the highest standardized coefficients that positively predict drug use, while general certainty has the smallest impact. According to the correlation patterns, celerity and severity are irrelevant variables for estimating drug use. The addition of variables designed to measure normative standards (Model 2) increased the total amount of variance explained to 51%. According to Model 2, descriptive norms strongly predicted drug use, while the influence of personal certainty remained positive and the effect of general certainty was reduced to non-significance. When control variables were introduced into the equation (Model 3), the total amount of variance increased slightly to 54%. Model 3 confirms that descriptive norms and personal certainty are the two strongest predictors of drug use when all variables are entered.

Among students, the deterrence variables explained 25% of the variance in the behavior expectation. As can be seen in Model 1, personal certainty of punishment has a significant positive effect on expected future drug use. In addition, general certainty has a significant negative effect on expected future drug use, although this effect is comparatively smaller. Severity and celerity variables are not important variables for predicting drug use in this group either. After the introduction of normative variables in model 2, the explanation of the variance is remarkably higher (53%). Descriptive norms, proximal social norms, and personal norms have a strong and positive effect on subsequent drug use. However, personal certainty remains a positive predictor, and vicarious certainty retains a significant negative influence on estimated drug use. The addition of control variables in Model 3 almost did not increase the total amount of variance explained (54%), and none of them turns out to be significantly relevant. With some slight changes in magnitude, all of the predictors identified in Model 2, except personal norms, continue to have a significant association with predicting drug use.

Personal certainty estimation is strongly associated with expected future drug use in both samples, and it is also an important predictor of expected drug use. Besides zero-order correlation results, the personal risk of arrest is an even stronger predictor of drug use in the regression analysis, and the direction of its influence remains positive. A higher personal certainty estimation is persistently associated with an increased likelihood of future drug use.

Notably, general certainty, a variable with a low correlation with expected future drug use, is an important predictor in the regression analyses among students. Although previous research has found that certainty usually ceases to be a relevant predictor at the multivariate level (e.g., Paternoster, 1987), we found that this assumption did not apply to our student data. The likelihood that a student will use drugs is reduced the more he or she estimates that other users can be caught. The persistence of this result suggests a deterrent effect, albeit a small one. Nevertheless, general certainty estimation remains irrelevant at the multivariate level among DDUs, and consequently the effect does not occur for this selected group.

The weight of each normative variable found in the regressions is relatively different from the zero-order correlation analysis. Descriptive norms are strong predictors, and distant social norms are not relevant for expected future drug use for either DDU or students. However, personal norms and proximal social norms are predictors only for students. Finally, the control variables are not relevant for the expected future drug use.

In summary, personal certainty and descriptive norms are generally positively correlated with expected future drug use and are also its strongest predictors. Taken together, however, the results also suggest that students and DDUs share the same pattern of influences in only a few characteristics. The effects of personal norms and proximal social norms are most relevant for students. Furthermore, the main group difference is in the general certainty effect. Among DDU, this variable is positively correlated with behavioral intention and is not a relevant predictor. Among students, general certainty is a negative predictor of expected future drug use, despite its non-significant low correlation with behavioral intention.

Discussion

Our primary concern is to compare the relevance of normative and deterrent variables in estimating future drug use. Overall, the results of this study were consistent with a normative perspective on human behavior. Consistent with most previous research on perceptual deterrence (e.g., Foglia, 1997), normative variables are strongly correlated with prospective drug use, and they also tend to outperform deterrence variables as predictors in regression analysis.

On the normative side, we found two main results. First, participants in both groups are more likely to use drugs in the future when they have the perception that others are using drugs. Descriptive normative environments (Cialdini et al., 1991), which are perceptions of how common a behavior is within a population, are found to be an important predictor of future drug use. This finding is consistent with Elek et al. (2006) review of the literature, which shows that norms have significant influences on substance use and that descriptive norms appear to have the strongest influence. Second, other normative variables are only relevant predictors among students, although they are always positively correlated with behavioral intention. Proximal social norms are one of the strongest predictors of student drug use behavior. In this sample, the power of personal norms to influence behavioral intentions is comparatively smaller. Contrary to Wenzel (2004), we find a stronger influence of social norms on behavior than of internalized ethics. Proximal social norms are still crucial even after controlling for personal ethics and deterrence. Interestingly, distant social norms are only weakly correlated with behavioral intentions and are also outside the range of behavioral predictors. At sufficient distance, social norms appear to be irrelevant. These findings highlight the power of the immediate social context in predicting drug use and underestimate the power of general normative standards. Students tend to abstain from drug use primarily because they believe that significant others do not use drugs and will disapprove. They do not necessarily act in accordance with what the community as a whole thinks about the use of drugs. Social bond theory (Hirschi, [1969] 2009) fits well with these findings. Given that the behavior of significant others and proximal normative norms both influence drug use intentions, any prevention effort must address the social context of individuals, including families, peers, and neighborhoods.

The present study produced four main findings related to deterrence variables. First, the deterrence variables were not related to each other. These findings are consistent with several previous studies that also found evidence for largely independent perceptions of risk probability, sanction severity, and celerity. The classical variables appear to be empirically distinct components of deterrence and must be considered interactively rather than additively (see Tittle, 2000). Deterrence was not an empirically reliable concept. Second, even at the bivariate level, behavioral intention was not significantly associated with the severity or celerity of sanctions. These core elements of the deterrence model were not even related to intention to use drugs. This simple finding is consistent with some perceptual deterrence research (see Foglia, 1997) and suggests that people may not refrain from using drugs because they fear the legal consequences of their actions. This pattern of results was obtained in the Portuguese context of mild deterrent threats, but the applicability of a harsh deterrent, such as imprisonment for drug possession for use (available in several jurisdictions), should be studied to test whether severity remains unrelated to behavioral intentions. Third, on a strictly personal level, certainty and behavioral intention have an unexpected positive correlation. We thoroughly analyzed this anomalous result. Participants tend to estimate a lower risk of arrest based on their intention from abstain to use drugs. A clear intention to abstain is congruent with an estimate of never getting caught for a predictably nonexistent behavior. This rather counterintuitive pattern of findings should be at least partially explained by our personal estimation question. Certainty questions in perceptual deterrence research typically ask participants about the probability of being caught if they committed a deviant behavior (cf. Paternoster & Piquero, 1995). However, our personal certainty question placed participants in a real-life scenario, which leads participants with lower intentions to use drugs to lower their certainty estimates. Yet, an increase in the likelihood of using drugs is directly related to an increase in the likelihood of being caught. Taken together, these findings help to establish personal certainty estimation as an important positive predictor of behavioral intentions in both groups. Fourth, general perceived safety is only a relevant predictor of drug use among students. This variable has a predictive negative influence on behavioral intention. Under the same conditions of deterrence variables and personal and social normative standards, the risk of arresting someone else has a deterrent effect on behavioral intention. The deterrent effect will be a function of their estimate of the vicarious experience of being caught. However, this effect does not occur for DDU. The perception of an increased likelihood that a drug user will be arrested does not hold back the intention to use drugs in the future among DDU. We expected a moderate certainty effect, consistent with the theory, only among individuals with behavioral experience, mainly because norms against drug use may render any instrumental reasoning irrelevant for abstainers (Wikstrom et al., 2011; Wright et al., 2004). However, our pattern of findings is inconsistent with this hypothesis.

Finally, and not surprisingly, we found higher estimates of future drug use among DDU. This well-established finding is consistent with virtually all criminological research that identifies history as a major risk factor for all types of deviant behavior (e.g., Andrews & Bonta, 2010).

Conclusion

This study highlights the importance of normative standards comparatively to deterrence in drug law compliance. Descriptive norms are an important normative predictor of drug use in both groups, and proximal social norms also exert a strong influence, especially among university students. The positive influence of normative variables was greater than all the rare negative signs on the deterrence side, including the most promising certainty effect.

We expected the highest deterrent effect among detected drug users, but our results contradict this hypothesis. A small deterrent effect, limited to the general certainty variable, is found only among more law-abiding people. Contrary to our expectation, the estimation of drug use among detected drug users is not influenced by any deterrent variable. It is possible that the minimal or null impact of changes in drug policy and, in particular, drug laws in Portugal (and in other countries) can be explained, at least in part, by the lower capacity of deterrence variables to change intentions and behaviors.

An innovative split of the certainty variable allowed us to examine different patterns of results. The certainty variable evaluated at a strictly personal level shows an anomalous result in perceptual deterrence research. Individuals who do not intend to use drugs estimate the absence of any possibility of being caught. In addition, an increase in the probability of using drugs in the future is associated with a higher estimate of the probability of being caught. These findings strongly eroded the presumed deterrent effects of drug laws, especially the role of the threat of sanctions for compliance.

In general, instrumental concerns are not crucial for drug use intentions among university students and especially among detected drug users. However, since this study largely supports the influence of norms, we cannot ignore the possible normative force of the law (Williams & Hawkins, 1986). The simple fact that drug use is illegal may have a symbolic threshold effect (MacCoun, 1993) that influences behavior by increasing support for social and personal norms against drug use. In addition, a decline in drug use may also limit the strong influence of descriptive norms. Nevertheless, the contribution of drug prohibition to personal, social, and descriptive norms remains unclear. Future studies would benefit from more extensive exploration of this line of research.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Jorge Quintas

Notes

References

Andenaes

(1974). Punhishment and deterrence. The University of Michigan Press.

Andrews

Bonta

(2010). The psychology of criminal conduct (5th ed.). LexisNexis Matthew Bender.

Apel

Nagin

(2011). General deterrence: A review of recent evidence. In Wilson

Petersilia

(Eds.), Crime and public policy (pp. 411–436). Oxford University Press.

Beccaria

([1766]1998). Dos Delitos e das Penas. Edições da Fundação Gulbenkian.

Beyleveld

(1980). A bibliography on general deterrence. Saxon House.

Cialdini

Kallgreen

Reno

(1991). A focus theory of normative conduct: A theoretical refinement and reevaluation of the role of norms. Advances in Experimental Social Psychology, 24, 201–234. https://doi.org/10.1016/S0065-2601(08)60330-5

Dölling

Entorf

Hermann

Rupp

(2009). Is deterrence effective? Results of a meta-analysis of punishment. European Journal on Criminal Policy and Research, 15(1–2), 201–224. https://doi.org/10.1007/s10610-008-9097-0

Dölling

Entorf

Hermann

Rupp

(2011). Meta-analysis of empirical studies on deterrence. In H. Kury & E. Shea (Eds.), Punitivity - International developments. Vol. 3: Punitiveness and punishment, (pp. 315–378). Bochum: Universitätsverlag Dr. Brockmeyer.

Elek

Miller-Day

Hecht

(2006). Influences of personal, injunctive, and descriptive norms on early adolescent substance use. Journal of Drug Issues, 36(1), 147–172. https://doi.org/10.1177/002204260603600107

10.

European Monitoring Centre for Drugs and Drug Addiction. (2020). Monitoring and evaluating changes in cannabis policies: Insights from the Americas (Technical Report). Publications Office of the European Union.

11.

Field

(2009). Discovering statistics using SPSS (3rd ed.). Sage Publications.

12.

Foglia

(1997). Perceptual deterrence and the mediating effect of internalized norms among inner-city teenagers. Journal of Research in Crime and Delinquency, 34(4), 414–442. https://doi.org/10.1177/0022427897034004002

13.

Gallupe

Baron

(2014). Morality, self-control, deterrence, and drug use: Street youths and situational action theory. Crime & Delinquency, 60(2), 284–305. https://doi.org/10.1177/0011128709359661

14.

Gibbs

(1975). Crime, punhishment and deterrence. Elsevier.

15.

Gibbs

(1977). Social control, deterrence and perspectives on social order. Social Forces, 56(2), 408–423. https://doi.org/10.2307/2577733

16.

Gottfredson

Hirschi

(1990). A general theory of crime. Stanford University Press.

17.

Greenwald

(2009). Drug decriminalization in Portugal: Lessons for creating fair and successful drug policies. CATO Institute.

18.

Hirschi

([1969]2009). Causes of delinquency (9th ed.). Transaction Publishers.

19.

Hughes

Stevens

(2010). What can we learn from the Portuguese decriminalization of illicit drugs? British Journal of Criminology, 50(6), 999–1022. https://doi.org/10.1093/bjc/azq038

20.

Jacobs

Piquero

(2013). Boundary-crossing in perceptual deterrence: Investigating the linkages between sanction severity, sanction certainty, and offending. International Journal of Offender Therapy and Comparative Criminology, 57(7), 792–812. https://doi.org/10.1177/0306624X12443944

21.

Kamin

(2017). Marijuana regulation in the United States. In Savona

Kleiman

Calderoni

(Eds.), Dual markets. Comparative approaches to regulation (pp. 105–120). Springer.

22.

Kennedy

(2009). Deterrence and crime prevention: Reconsidering the prospect of sanction. Routledge.

23.

Kilmer

(2002). Do cannabis possession laws influence cannabis use? In Spruit (Ed.). Cannabis 2002 report. Technical report of the international scientific conference (Cap. 8, pp. 119–141). Ministry of Public Health of Belgium.

24.

Loughran

Pogarsky

Piquero

Paternoster

(2012). Re-examining the functional form of the certainty effect in deterrence theory. Justice Quarterly, 29(5), 712–741. https://doi.org/10.1080/07418825.2011.583931

25.

MacCoun

(1993). Drugs and the law: A psychological analysis of drug prohibition. Psychological Bulletin, 113(3), 497–512. https://doi.org/10.1037/0033-2909.113.3.497

26.

Nagin

(1998). Criminal deterrence research at the outset of the twenty-first century. In Tonry

(Ed.), Crime and justice: A review of research (Vol. 23, pp. 1–42). University of Chicago Press.

27.

Nagin

(2018). Deterrent effects of the certainty and severity of punishment. In Nagin

Cullen

Jonson

(Eds.), Deterrence, choice, and crime. Contemporary perspectives (pp. 157–185). Routledge.

28.

Nagin

Pogarsky

(2001). Integrating celerity, impulsivity and extralegal sanction threats into a model of general deterrence. Criminology; An Interdisciplinary Journal, 39(4), 865–892. https://doi.org/10.1111/j.1745-9125.2001.tb00943.x

29.

Nagin

Pogarsky

(2003). An experimental investigation of deterrence: Cheating, self-serving & impulsivity. Criminology; An Interdisciplinary Journal, 41(1), 167–194. https://doi.org/10.1111/j.1745-9125.2003.tb00985.x

30.

Quintas

(2011). Regulação Legal do Consumo de Drogas: Impactos da Experiência Portuguesa da Descriminalização, Porto: Fronteira do Caos.

31.

Pacula

Smart

(2017). Medical marijuana and marijuana legalization. Annual Review of Clinical Psychology, 13(1), 397–419. https://doi.org/10.1146/annurev-clinpsy-032816-045128

32.

Pardo

(2014). Cannabis policy reforms in the Americas: A comparative analysis of Colorado, Washington, and Uruguay. International Journal of Drug Policy, 25(4), 727–735. https://doi.org/10.1016/j.drugpo.2014.05.010

33.

Paternoster

(1987). The deterrent effect of the perceived certainty and severity of punishement: A review of the evidence and issues. Justice Quarterly, 4(2), 173–217. https://doi.org/10.1080/07418828700089271

34.

Paternoster

(2010). How much do we really know about criminal deterrence. Journal of Criminal Law & Criminology, 100(3), 765–823. https://doi.org/0091-4169/10/10003-0765

35.

Paternoster

(2018). Perceptual deterrence theory. In Nagin

Cullen

Jonson

(Eds.), Deterrence, choice, and crime. Contemporary perspectives (pp. 81–106). Routledge.

36.

Paternoster

Piquero

(1995). Reconceptualizing deterrence: An empirical test of personal and vicarious experiences. Journal of Research in Crime and Delinquency, 32(3), 251–286. https://doi.org/10.1177/0022427895032003001

37.

Pauwels

Weerman

Bruinsma

Bernasco

(2011). Perceived sanction risk, individual propensity and adolescent offending: Assessing key findings from the deterrence literature in a Dutch sample. European Journal of Criminology, 8(5), 386–400. https://doi.org/10.1177/1477370811415762

38.

Piquero

Bouffard

Leeper-Piquero

Craig

(2016). Does morality condition the deterrent effect of perceived certainty among incarcerated felons? Crime & Delinquency, 62(1), 3–25. https://doi.org/10.1177/0011128713505484

39.

Pratt

Cullen

Blevins

Daigle

Madensen

(2006). The empirical status of deterrence theory: A meta-analysis. In Cullen

Wright

Blevins

(Eds.), Taking stock: The status of criminological theory (pp. 367–395). Transaction Publishers/Rutgers University Press.

40.

Pratt

Turanovic

(2018). Celerity and deterrence. In Nagin

Cullen

Jonson

(Eds.), Deterrence, choice, and crime: Contemporary perspectives (pp. 187–210). Routledge.

41.

Scheim

Maghsoudi

Marshall

Churchill

Ziegler

Werb

(2020). Impact evaluations of drug decriminalisation and legal regulation on drug use, health and social harms: A systematic review. BMJ Open, 10(9), e035148. https://doi.org/10.1136/bmjopen-2019-035148

42.

Single

Christie

Ali

(2000). The impact of Cannabis decriminalisation in Australia and the United States. Journal of Public Health Policy, 21(2), 157–186. https://doi.org/10.2307/3343342

43.

Tittle

(2000). Theoretical developments in criminology. In LaFree

(Ed.), The nature of crime: Continuity and change (Vol. 1, pp. 51–101). National Institute of Justice.

44.

Weatherburn

Topp

Midford

Allsopp

(2000). Drug crime prevention and mitigation: A literature review and research agenda. New South Wales Bureau of Crime Statistics and Research.

45.

Weisburd

Britt

(2007). Statistics in criminal justice. Springer.

46.

Wenzel

(2004). The social side of sanctions: Personal and social norms as moderators of deterrence. Law and Human Behavior, 28(5), 547–567. https://doi.org/10.1023/B:LAHU.0000046433.57588.71

47.

Wikstrom

Tseloni

Karlis

(2011). Do people comply with the law because they fear getting caught? European Journal of Criminology, 8(5), 401–420. https://doi.org/10.1177/1477370811416415

48.

Williams

Hawkins

(1986). Perceptual research on general deterrence: A critical review. Law & Society Review, 20(4), 545–572. https://doi.org/10.2307/3053466

49.

Wright

Caspi

Moffitt

Paternoster

(2004). Does the perceived risk of punishment deter criminally prone individuals? Rational choice, self-control, and crime. Journal of Research in Crime and Delinquency, 41(2), 180–213. https://doi.org/10.1177/0022427803260263

50.

Liska

(1993). The certainty of punishment: A reference group effect and its functional form. Criminology; An Interdisciplinary Journal, 31(3), 447–464. https://doi.org/10.1111/j.1745-9125.1993.tb01137.x

51.

Zimring

Hawkins

(1973). Deterrence: The legal threat in crime control. University of Chicago Press.