Understanding the Relationship Between Public Opinion and Experience With the Criminal Justice System

Understanding the Relationship Between Public Opinion and Experience With the Criminal Justice... Abstract Surveys indicate that between 70% and 80% of people consider judges to be too lenient. This current research confirms the inadequacy of this general measure of public opinion by contrasting it with alternative opinion-based measures of court sentencing across specific criminal cases. The article also explores the impact of public’s experiences of the criminal justice system on their opinions about the courts. Having been a victim or attending a criminal case where the punishment was seen as being too lenient seems to cause greater dissatisfaction, whereas contact with people working with victims or who have been in prison improves the level of satisfaction. In September 2011, when he was being questioned about the rationale for a draft law on crime, a Conservative minister in Canada’s House of Parliament proclaimed, “We believe that the best experts in Canada are the voters who give us the mandate to do what we said we would do.”1 Indeed, and, as Garland (2001) once suggested, the dominant voice in matters of penal policy is no longer that of the experts, but that of the public. But what does the public really want in terms of sentencing? The results of surveys on criminal justice are disturbing: between 70% and 80% of the population feel that the sentences imposed by the courts are too lenient. According to Roberts, Stalans, Indermaur, and Hough (2003), this result is constant across both time and place, although rates are slightly lower in Canada (between 60% and 70% dissatisfied) and slightly higher in the United States (up to 85% dissatisfied). Politicians generally rely on such popular outlooks to justify measures that increase the length of sentences or more general tough-on-criminal stances (Sprott, Webster, & Doob, 2013). However, several researchers have demonstrated that the gap between public demand and provision by the courts is not considerable2 (Sprott, 1998; Hough & Roberts, 1998; Roberts & Stalans, 1997; Rossi, Berk, & Campbell, 1997; Zamble & Kalm, 1990; Roberts & Doob, 1989; Doob & Roberts, 1988). In several studies, it has been shown that when people were given a detailed vignette and subsequently asked to determine the suitable sentence, their recommendations were as severe or less severe than the sentences generally imposed by courts (Stalans & Diamond, 1990; Zamble & Kalm, 1990; Hough & Roberts, 1999). The same conclusion is reached when examining opinions about support for repressive policies. Applegate and his colleagues (1996) found that, when asked about their support for the “Three Strikes” policy, citizens may express clear support, but this drops dramatically when people are given a specific scenario and asked, for example, what sentence they would give someone convicted of a serious offense for the third time. Similarly, Durham, Elrod, and Kinkade (1996) found that when respondents were presented with 17 scenarios describing situations in which murder had been committed, the percentage of respondents supporting the death penalty was both higher (91%) and lower (31%) (depending on the scenario) than the percentage obtained with a general question about support for that penalty (80% in favor). Such nuances are rarely presented when presenting or discussing public opinion measures. Thus, there is a more general factor that is often overlooked. An individual’s knowledge base and experience with the criminal justice system will have a direct effect on how that person perceives leniency and severity in court cases and decisions. This article develops this general point. First, by assessing the limits of traditional measures of public opinion on sentencing. Second, by a review of past research that explained the public’s apparent dissatisfaction with courts and the criminal justice system. The remainder of this article then focuses on a series of analytical demonstrations that support alternative outlooks and measures of this phenomenon, with specific attention on the respondent’s knowledge-based and experience thesis. Limits of Public Opinion Measures There are three common criticisms of opinion polls. First, the general questions in surveys are subject to a variety of interpretations (Casey & Mohr, 2005). For instance, studies showed that people think of different crimes (often, but not always, violent crime) when expressing an opinion about court sentences in general3 (Indermaur, 1987; Sprott, 1996; Doob & Roberts, 1983). Second, polls are criticized for capturing the emotional responses of citizens, responses that are often volatile (Rossi, et al., 1997; Casey & Mohr, 2005) and more complex than suggested by the polls (Hutton, 2005; Roberts & Stalans, 1997; Cullen, Fisher, & Applegate, 2000). Third, citizens are required to give an opinion about a reality with which they are unfamiliar. Surveys of public knowledge show unequivocally that people almost systematically overestimate frequency and severity of crimes, as well as underestimating the severity of courts (Gray, 2009; Hough & Roberts, 1998; Mattinson & Mirrlees-Black, 2000; Roberts & Stalans, 1997; St-Amand & Zamble, 2001). Studies on heuristics suggest that in the absence of specific information or details, people fill the gaps by resorting to stereotypes (Stalans, 1993; Stalans & Diamond, 1990). Because most people get the bulk of their knowledge about the courts from the media, we can assume that people will generally take as their reference point cases that are serious, unusual, and not representative of the reality of the courts (Jewkes, 2004). These limitations have prompted several researchers to question the current use of opinion polls on criminal justice. It is frequently suggested that research should offer new ways of gauging public opinion by providing as much detail as possible to citizens so that they have the necessary information to formulate and support their opinions. However, there are only few studies that have ventured down this path. Warner and Davis’s study (2011) is both unique and innovative because it looked at a new way of measuring opinion about court sentences by questioning people who had recently served as jurors in one or more criminal cases. The first result of the study confirms the conclusions put forward by other researchers (see Gwin, 2010, for a comparison of juror recommendations and sentencing guidelines): the jury generally recommended sentences that were less severe than the one handed down by the judge in the same case. A more interesting result is that although the vast majority of jurors expressed a general dissatisfaction with the leniency of court sentences4 for the type of crime they had to judge, 90% found the final sentence imposed by the judge to be appropriate. While some jurors, whom the authors referred to as the “extrapolators,” used their experience and knowledge of the sentences of the courts to more favorably reevaluate their opinion about the broader work of the courts, others remained unaffected. What Warner and Davis’s study shows us is that there is a significant difference between a more general opinion expressed in a survey and an opinion that is more specific and placed in the context of the various contingencies that comprise a decision-making process. However, this study does not provide information about whether it was access to detailed facts, experience as a juror, or knowledge about the judge’s sentence that led to more positive opinions. Furthermore, as the study involved people who had recently served as jurors, we do not know its relevance for the majority of citizens, who typically have no experience with the criminal justice system. Explaining Public Opinion Knowing that incorrect knowledge about criminality and the courts encourages greater dissatisfaction with courts (Gelb, 2006; Roberts & Stalans, 1997; Hough & Roberts, 1998; Mattinson & Mirrlees-Black, 2000; Chapman, Mirrlees-Blaws, & Brawn, 2002), some researchers have suggested that, if the public knew more about the criminal justice system, they would be more satisfied with it (see Mirrlees-Black, 2002; Indermaur & Hough, 2002 for a discussion on that topic). However, empirical studies of this particular hypothesis provide contradicting results. Examining the effect of court experiences on public confidence in the criminal justice system, Van de Walle (2009) concluded that research in this area is “inconclusive” and noted a number of limitations. He and others (Benesh & Howell, 2001; Rottman, Hansen, Mott, & Grimes, 2003) pointed out the importance of taking into account different dimensions of experience with the criminal justice system: the level of implication (direct or indirect experiences), the timing (past or recent), the outcome (positive or negative), the role played; and the motivation to participate in the criminal justice system. Despite their recommendations, few researchers have considered the variability of experience (Van de Walle, 2009; Benesh & Howell, 2001). Most studies interested in opinions about sentencing described the typical “punitive” or “dissatisfied” citizen by presenting sociodemographic characteristics or attitudes. Roberts and Indermaur (2007), for example, tried to predict punitive attitudes (measured by support for death penalty, stiffer sentencing, and sentences reflecting public opinion) with different sets of factors. They concluded that, “It appears to be people’s knowledge and beliefs of crime and the criminal justice system that underpin punitive attitudes to a greater extent than demographics, political orientation, religious attendance or media use.” However, these kinds of studies that explain public opinion about court decisions all rely on general measures of public dissatisfaction. Even if these general measures or questions (e.g., is the sentence imposed by the court too lenient? Would you support stiffer sentencing?) are recognized to be severely biased, they continue to constitute the basis of research on public opinion about sentencing. Current Study In the first part of the article, three different measures of public opinion are proposed to highlight differences between general and specific opinions. According to Kahneman (2011), people think and process information by two different processes. The first process (referred to as System 1) is related to intuition and emotions. It treats information automatically, quickly, and with little or no cognitive effort. This process uses stereotypes and prior information to give a fast and simple answer. The second process (System 2) is more cognitively challenging and is mobilized when the situation requires analytical thinking. It processes information more slowly and involves a level of concentration and reasoning. Based on these ideas, we can hypothesize that when people are asked for a general opinion, they will answer more spontaneously (using System 1) and their answer is more likely to be based on stereotypes (most often inspired by the media representation of crime; Voumvakis & Ericson, 1984). On the other hand, when given detailed information about a specific case, individuals will be forced to use System 2 in dealing with the information, which will lead to a change in their point of reference and a more informed opinion. In the second part of the article, we use hierarchical regression analysis to examine the impact of different experiences with crime and the justice system on individual opinions about the leniency of the courts. In this article, we use two measures of public opinion that are put into context (reducing possible biases) and different types and levels of experience to gain a better understanding of what matters in court experiences. As the presentation of specific information would mobilize System 2, the respondent will make an effort to analyze the information provided to them. In such a context, past experiences or indirect knowledge acquired by contact with criminal justice actors might also be used, affecting an individual’s thought process. Method Data were collected during the summer of 2009 through a survey conducted with 200 French Canadian participants living in Montreal.5 Face-to-face interviews were conducted with each participant and respondent’s responses were noted on a paper questionnaire by the interviewer. The questionnaire consisted in 57 pages and lasted on average 90 min. The survey began with questions on opinions and knowledge about the court, the parole system, the crime, and offenders. Respondents were then provided three detailed criminal cases (see Appendix I for a sample description) and asked to sentence the defendant and give their view about the case (e.g., sentence provided by the court, sentencing goal). Questions about experiences with crime (as a victim or an offender) and the criminal justice or its actors (judge, police, etc.) were included before respondents had to evaluate three other crime scenarios. In these last three scenarios, respondents had to decide whether to grant parole or not and give their view about the case (e.g., opinion on the sentence originally imposed, recidivism risk). The survey ended with a set of sociodemographic questions (age, sex, education, etc.). Measuring Public Opinion At the beginning of the questionnaire, people were asked to answer the classic opinion poll question: “In general, would you say that sentences handed down by the courts are too severe, about right, or too lenient.” In the first three cases, participants had to make a series of decisions on the appropriate sentence to be imposed on three men convicted of various crimes: Peter, guilty of theft and handling stolen goods worth $8,000; Paul, guilty of a domestic homicide; and John, guilty of incest with his daughter. For each criminal case, respondents had to indicate the sentence they would impose, the length of sentence they considered acceptable (the minimum and the maximum sentence they would agree with), and the sentence they believed the courts would have imposed.6 Past research has demonstrated that citizens, as judicial actors, do not have a fixed concept of a just sentence, but are more likely to suggest wide ranges of sentences that are perceived as equally acceptable (Leclerc, 2010). It is therefore imperative to see whether the sentence they think the courts imposed for each case lies within this range. For each of the three criminal cases, we created a new opinion variable, the anticipated sentence of the court, to check whether the anticipated sentence was (1) too severe (more severe than the maximum acceptable sentence); (2) about right (between the minimum and maximum acceptable sentence); or (3) too lenient (less severe than the minimum acceptable sentence). This conception of public opinion is interesting for two reasons. First, it focuses on opinion and not on sentencing expectations. Researchers have often concluded that there is a difference between general and specific opinion by comparing responses with general questions on the work of the courts with the sentences that the public proposed for different types of offense (Stalans & Diamond, 1990; Zamble & Kalm, 1990; Hough & Roberts, 1999). Thus, they merged two different concepts: the opinion (“what I think of the courts”) and the sentence required (“what I hope the courts will impose”). The measure we propose avoids this problematic merge. Second, the measure reflects the willingness of the respondents to negotiate. It is not because an individual wants to impose 5 years in prison and believes that the court gave 3 years that he will necessarily be dissatisfied. This measure takes into account the openness of citizens to various penal options, which they consider as equally satisfactory. In the last three scenarios, participants had to decide whether to grant parole to three individuals who had requested it7: Thran, serving 2 years less a day in prison for a car theft and a robbery; Costa, jailed for 2 years in prison for dealing in hard drugs; and Hervé, sentenced to 18 months in prison for domestic violence. As one of the questions on the assessments made of each of these three cases, participants were asked what they thought of the sentence imposed by the courts. As with the earlier variable, they could say that the sentence seemed to be too severe, about right, or too lenient. This variable was then used as a measure of specific opinion on the actual sentence of the courts. This measure has the advantage of providing a sentencing “tariff” for respondents, making it possible to analyze opinion that is no longer biased by a lack of knowledge about the reality of the court. It also provides the opportunity to test the hypothesis that citizens are dissatisfied with judges’ decisions because they underestimate the severity of the courts. Explaining Public Opinion Tables 1 and 2 present all independent variables8 used to explain why some citizens consider sentences, in particular case, as being too lenient. As few people consider courts to be “too severe,” we decided to merge these people with the ones who consider the court decision to be “about right.” Moreover, people interested in public opinion are generally focusing on the “punitive person,” who considers the court to be too lenient. Models presented in this article thus compare sentences that are considered too lenient with all others (e.g., sentence considered severe enough and too severe). Additional analyses show identical results when excluding sentences considered too severe (7.8% of the sample). Table 1 Descriptive Measures of Independent Variables (Level 1) Type of crime Crime seriousness (1–100) (Median score) Victimization (% of respondents saying that a close family or themselves have been a victim of this type of crime) Theft and Handling stolen goods 40 10 Domestic homicide 90 3 Incest 90 21 Robbery and car theft 70 33 Hard drugs trafficking 60 14 Domestic violence 80 16 Type of crime Crime seriousness (1–100) (Median score) Victimization (% of respondents saying that a close family or themselves have been a victim of this type of crime) Theft and Handling stolen goods 40 10 Domestic homicide 90 3 Incest 90 21 Robbery and car theft 70 33 Hard drugs trafficking 60 14 Domestic violence 80 16 Table 1 Descriptive Measures of Independent Variables (Level 1) Type of crime Crime seriousness (1–100) (Median score) Victimization (% of respondents saying that a close family or themselves have been a victim of this type of crime) Theft and Handling stolen goods 40 10 Domestic homicide 90 3 Incest 90 21 Robbery and car theft 70 33 Hard drugs trafficking 60 14 Domestic violence 80 16 Type of crime Crime seriousness (1–100) (Median score) Victimization (% of respondents saying that a close family or themselves have been a victim of this type of crime) Theft and Handling stolen goods 40 10 Domestic homicide 90 3 Incest 90 21 Robbery and car theft 70 33 Hard drugs trafficking 60 14 Domestic violence 80 16 Table 2 Descriptive Measures of Independents Variables (Level 2) Characteristic % of respondents answering in the affirmative (N = 179a) Involvement in the criminal justice system     Were a witness in a case 12.1     Involved in a case as the accused or relative of the accused 11.7     Involved in a case as a victim or relative of the victim 14.7     Involved in a case where the sentence was considered to be too lenient 14.2 Indirect knowledge of the criminal justice system     Knows a lawyer 55.4     Knows a judge 13.1     Knows someone who works with criminals 39.2     Knows someone who works with victims 27.9     Knows someone who has been in prison 38.4 Personal characteristics     <30 17.3     >65 11.7     Male 48.0     Separated 19.6     Married or Living with 49.0     Children 55.7     Diploma of Collegiate Studiesb 21.6     University 43.3 Characteristic % of respondents answering in the affirmative (N = 179a) Involvement in the criminal justice system     Were a witness in a case 12.1     Involved in a case as the accused or relative of the accused 11.7     Involved in a case as a victim or relative of the victim 14.7     Involved in a case where the sentence was considered to be too lenient 14.2 Indirect knowledge of the criminal justice system     Knows a lawyer 55.4     Knows a judge 13.1     Knows someone who works with criminals 39.2     Knows someone who works with victims 27.9     Knows someone who has been in prison 38.4 Personal characteristics     <30 17.3     >65 11.7     Male 48.0     Separated 19.6     Married or Living with 49.0     Children 55.7     Diploma of Collegiate Studiesb 21.6     University 43.3 Note.aA seriousness scale ranging from 1 to 100 was added during data collection to compensate for the fact that another measure of seriousness (amplitude scale) seemed problematic for respondents. The first 19 respondents did not assess the severity of the different scenarios (n = 114) and therefore had to be excluded from the analyses. Twelve other cases were eliminated because they had one or more missing values on the variables under study. bThe school system in Quebec provides a college education (2 or 3 years) before university-level education. Table 2 Descriptive Measures of Independents Variables (Level 2) Characteristic % of respondents answering in the affirmative (N = 179a) Involvement in the criminal justice system     Were a witness in a case 12.1     Involved in a case as the accused or relative of the accused 11.7     Involved in a case as a victim or relative of the victim 14.7     Involved in a case where the sentence was considered to be too lenient 14.2 Indirect knowledge of the criminal justice system     Knows a lawyer 55.4     Knows a judge 13.1     Knows someone who works with criminals 39.2     Knows someone who works with victims 27.9     Knows someone who has been in prison 38.4 Personal characteristics     <30 17.3     >65 11.7     Male 48.0     Separated 19.6     Married or Living with 49.0     Children 55.7     Diploma of Collegiate Studiesb 21.6     University 43.3 Characteristic % of respondents answering in the affirmative (N = 179a) Involvement in the criminal justice system     Were a witness in a case 12.1     Involved in a case as the accused or relative of the accused 11.7     Involved in a case as a victim or relative of the victim 14.7     Involved in a case where the sentence was considered to be too lenient 14.2 Indirect knowledge of the criminal justice system     Knows a lawyer 55.4     Knows a judge 13.1     Knows someone who works with criminals 39.2     Knows someone who works with victims 27.9     Knows someone who has been in prison 38.4 Personal characteristics     <30 17.3     >65 11.7     Male 48.0     Separated 19.6     Married or Living with 49.0     Children 55.7     Diploma of Collegiate Studiesb 21.6     University 43.3 Note.aA seriousness scale ranging from 1 to 100 was added during data collection to compensate for the fact that another measure of seriousness (amplitude scale) seemed problematic for respondents. The first 19 respondents did not assess the severity of the different scenarios (n = 114) and therefore had to be excluded from the analyses. Twelve other cases were eliminated because they had one or more missing values on the variables under study. bThe school system in Quebec provides a college education (2 or 3 years) before university-level education. Real or perceived knowledge of the sentence A dichotomous variable enabled a distinction to be made between the three cases of sentencing where citizens had to determine the likely sentence of the courts and the three cases for parole where they were asked to comment on the quality of the sentence chosen by the judge. This variable is used to test the hypothesis often cited in the literature that people would be more satisfied if they had a better knowledge of the sentence handed down by judges. Crime seriousness The literature on public opinion suggests that citizens are more dissatisfied with sentences for serious crimes (Frost, 2010). We introduced a variable in order to test whether citizens’ levels of satisfaction is related to their assessment of the offense’s seriousness. The seriousness of the criminal scenarios was evaluated on a scale ranging from 1 to 100. The results in Table 1 show that citizens consider cases of incest and conjugal homicide as similar in severity (a median of 90)9 and more serious than other offenses. Crime experience Although several studies found no direct link between victimization and punitiveness (Tufts, 2000; Hough & Roberts, 1999; Brillon, 1988; Langworthy & Whitehead, 1986), we know little about the impact of victimization on opinions about court sentencing (with the exception of the work of Sprott & Doob, 1997 who discovered complex relationships between victimization and citizen opinions). Furthermore, it is possible that the lack of a correlation between victimization and punitiveness observed in the literature results from the use of a general measure of victimization. This variable tested whether experience of a similar crime,10 either personally or through a close family member, influenced the opinion someone had of sentencing by the courts in these particular cases. Table 1 shows that the rate of victimization varies depending on the crime (3–33%). Involvement in the court system Several researchers suggested that citizen’s opinions about the courts are problematic because they are based only on information obtained from the media—known to be a highly biased source (Jewkes, 2004). We wanted to examine whether having participated in or attended a criminal case would influence public opinion on sentencing by the courts. As shown in Table 2, participation in the court system is measured by having attended a criminal case as a witness (12.1%), as a defendant or a friend of the accused (11.7%), or as a victim or a friend of a victim (14.7%). These categories are not mutually exclusive (8% attended a trial in two different “statuses”), and we found that 27.6% of the respondents had attended a criminal trial in one or more of these three statuses. Among all respondents, 14.2% believed that the sentence in one or more of the cases they attended was too lenient. Indirect knowledge of the criminal justice system Indirect knowledge of the criminal justice system was measured on the basis of knowing someone who works directly in the criminal justice system (Table 2). A number of respondents knew a judge (13.1%), a lawyer (55.4%), an individual who works with victims (27.9%), an offender (39.2%), or a person who has directly experienced the criminal justice system by being in prison (38.4%). Most citizens have an indirect knowledge of the courts and several may have been exposed to different views because 21.1% of them knew two different types of participants, while 30% knew at least three. Personal characteristics Individual characteristics are also included to control for the fact that experiences with the court might not be randomly distributed (Van de Walle, 2009). We also took accounted for age, gender, and education level of respondents; variables often found in studies on opinion and attitudes of citizens. We also included marital status and whether respondents had children because these characteristics could potentially influence their opinion about certain crimes (e.g., domestic violence or incest). Analytical Framework The nested data structure (six cases evaluated per individual) requires the use of hierarchical regression analysis11 (or multilevel analysis, see Hox, 2010; Hox & Kreft, 1994; Snijders & Bosker, 1999). The opinions of an individual on court sentences for the six crime scenarios are not independent and they may be more similar than they would be if they were the responses of two different respondents. Hierarchical regression analysis has the advantage of avoiding the violation of the assumption of independence of the error terms (residuals) required by conventional regression analysis.12 The results of this analysis13 can be assessed in the same ways as those from an ordinary logistic regression. Results Measuring Public Opinion Table 3 presents different measures of public opinion on sentencing by the courts. Findings confirm the hypothesis based on the use of two different processes. Indeed, the percentage of respondents dissatisfied with the sentences of the courts is much higher when the question is asked in a general way without reference to a specific criminal case story. Information in a specific case will activate System 2 and ameliorate satisfaction regarding the work of the court. Table 3 Comparison of Different Measures of Public Opinion on Sentencing by the Courts Model 1 General opinion Model 2 Decision taking Model 3 Evaluation of the decision taken by a court In general Theft and handling stolen goods Domestic homicide Incest Robbery and car theft Hard drugs trafficking Domestic violence Court sentencea (2 years) (10 years) (3 years) 2 years–1 day 2 years 18 months Court too lenient 63% 25% 26% 53% 20% 31% 56% Court about right 36% 61% 61% 42% 74% 67% 42% Court too severe 2% 14% 13% 5% 7% 3% 3% Model 1 General opinion Model 2 Decision taking Model 3 Evaluation of the decision taken by a court In general Theft and handling stolen goods Domestic homicide Incest Robbery and car theft Hard drugs trafficking Domestic violence Court sentencea (2 years) (10 years) (3 years) 2 years–1 day 2 years 18 months Court too lenient 63% 25% 26% 53% 20% 31% 56% Court about right 36% 61% 61% 42% 74% 67% 42% Court too severe 2% 14% 13% 5% 7% 3% 3% Note. aSentence estimated by the public. Adapted from "Explorer et comprendre l'insatisfaction du public face à la clémence des tribunaux" by Leclerc (2010). Adapted with permission. Table 3 Comparison of Different Measures of Public Opinion on Sentencing by the Courts Model 1 General opinion Model 2 Decision taking Model 3 Evaluation of the decision taken by a court In general Theft and handling stolen goods Domestic homicide Incest Robbery and car theft Hard drugs trafficking Domestic violence Court sentencea (2 years) (10 years) (3 years) 2 years–1 day 2 years 18 months Court too lenient 63% 25% 26% 53% 20% 31% 56% Court about right 36% 61% 61% 42% 74% 67% 42% Court too severe 2% 14% 13% 5% 7% 3% 3% Model 1 General opinion Model 2 Decision taking Model 3 Evaluation of the decision taken by a court In general Theft and handling stolen goods Domestic homicide Incest Robbery and car theft Hard drugs trafficking Domestic violence Court sentencea (2 years) (10 years) (3 years) 2 years–1 day 2 years 18 months Court too lenient 63% 25% 26% 53% 20% 31% 56% Court about right 36% 61% 61% 42% 74% 67% 42% Court too severe 2% 14% 13% 5% 7% 3% 3% Note. aSentence estimated by the public. Adapted from "Explorer et comprendre l'insatisfaction du public face à la clémence des tribunaux" by Leclerc (2010). Adapted with permission. The percentage of respondents dissatisfied with the leniency of the courts drops from 63% (first column) to 20–30% in four of the six scenarios.14 In addition, when people are asked to make a decision (Model 2), they are more likely to think that the sentences of the courts were too severe and, therefore, longer than the sentence they deem acceptable in this specific case: the percentage increases from generally 2% to 7% to almost 15% (second and third column). When citizens are asked to put themselves in the place of the judge and make a decision, they seem much more aware of the severity of the decision than when asked to comment on the fairness of the sentences imposed by the courts (both in the abstract—Model 1—and in the concrete—Model 3). This result is similar to that of Warner and Davis (2011). Explaining Public Opinion: the Impact of Experiences and Contacts With the Court System Table 4 presents the results of different logistic regressions (multilevel analysis) in explaining public opinions about the courts’ leniency in each of the six cases. Parameters presented are odds ratios. The first model includes variables related to the criminal cases and the experiences of citizens. The second model adds the characteristics of the individuals. The last model only retains the variables that are significant or relevant from a theoretical point of view. Table 4 Multilevel Logistic Regression on the Dissatisfaction of Citizens With the Leniency of the Courts in Six Criminal Cases (N = 1,074 Cases, 179 Individuals) Model 0 Model 1 Model 2 Model 3 Model 4 Intraclass correlation coefficient (ICC)a 0.194*** 0.180*** 0.149*** 0.135*** 0.156*** Odds ratio Crime seriousness (1–100) 1.02*** 1.02*** 1.02*** 1.02*** Sentence of the judge known 1.02 1.03 1.03 1.03 Victim (direct or indirect) of the crime being assessed 2.05** 2.18*** 2.14*** 2.23*** Participated in a case as a witness 1.10 1.14 Involved in a case as the accused or relative of the accused 0.75 0.71 Involved in a case as the victim or relative of the victim 1.63 1.69 Involved in a case where the sentence was seen as being too lenient 1.69* 1.72* 1.78* Knows a lawyer 0.82 0.85 Knows a judge 1.54 1.70† 1.54 Knows someone who works with criminals 1.15 1.16 Knows someone who works with victims 0.60* 0.61* 0.63* Knows someone who has been to prison 0.66* 0.64* 0.67* <30 1.03 >65 0.73 Male 0.89 Separated 0.68 Married or living with 0.70 Children 1.44 Diploma of collegiate studies 1.50 University 1.03 Model 0 Model 1 Model 2 Model 3 Model 4 Intraclass correlation coefficient (ICC)a 0.194*** 0.180*** 0.149*** 0.135*** 0.156*** Odds ratio Crime seriousness (1–100) 1.02*** 1.02*** 1.02*** 1.02*** Sentence of the judge known 1.02 1.03 1.03 1.03 Victim (direct or indirect) of the crime being assessed 2.05** 2.18*** 2.14*** 2.23*** Participated in a case as a witness 1.10 1.14 Involved in a case as the accused or relative of the accused 0.75 0.71 Involved in a case as the victim or relative of the victim 1.63 1.69 Involved in a case where the sentence was seen as being too lenient 1.69* 1.72* 1.78* Knows a lawyer 0.82 0.85 Knows a judge 1.54 1.70† 1.54 Knows someone who works with criminals 1.15 1.16 Knows someone who works with victims 0.60* 0.61* 0.63* Knows someone who has been to prison 0.66* 0.64* 0.67* <30 1.03 >65 0.73 Male 0.89 Separated 0.68 Married or living with 0.70 Children 1.44 Diploma of collegiate studies 1.50 University 1.03 Note. †p ≤ 0.10; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001. aICC (logistic reg) = (Level2 Var)/(Level2 var + (π2/3)) as suggested by Snijders et al. (2012). Table 4 Multilevel Logistic Regression on the Dissatisfaction of Citizens With the Leniency of the Courts in Six Criminal Cases (N = 1,074 Cases, 179 Individuals) Model 0 Model 1 Model 2 Model 3 Model 4 Intraclass correlation coefficient (ICC)a 0.194*** 0.180*** 0.149*** 0.135*** 0.156*** Odds ratio Crime seriousness (1–100) 1.02*** 1.02*** 1.02*** 1.02*** Sentence of the judge known 1.02 1.03 1.03 1.03 Victim (direct or indirect) of the crime being assessed 2.05** 2.18*** 2.14*** 2.23*** Participated in a case as a witness 1.10 1.14 Involved in a case as the accused or relative of the accused 0.75 0.71 Involved in a case as the victim or relative of the victim 1.63 1.69 Involved in a case where the sentence was seen as being too lenient 1.69* 1.72* 1.78* Knows a lawyer 0.82 0.85 Knows a judge 1.54 1.70† 1.54 Knows someone who works with criminals 1.15 1.16 Knows someone who works with victims 0.60* 0.61* 0.63* Knows someone who has been to prison 0.66* 0.64* 0.67* <30 1.03 >65 0.73 Male 0.89 Separated 0.68 Married or living with 0.70 Children 1.44 Diploma of collegiate studies 1.50 University 1.03 Model 0 Model 1 Model 2 Model 3 Model 4 Intraclass correlation coefficient (ICC)a 0.194*** 0.180*** 0.149*** 0.135*** 0.156*** Odds ratio Crime seriousness (1–100) 1.02*** 1.02*** 1.02*** 1.02*** Sentence of the judge known 1.02 1.03 1.03 1.03 Victim (direct or indirect) of the crime being assessed 2.05** 2.18*** 2.14*** 2.23*** Participated in a case as a witness 1.10 1.14 Involved in a case as the accused or relative of the accused 0.75 0.71 Involved in a case as the victim or relative of the victim 1.63 1.69 Involved in a case where the sentence was seen as being too lenient 1.69* 1.72* 1.78* Knows a lawyer 0.82 0.85 Knows a judge 1.54 1.70† 1.54 Knows someone who works with criminals 1.15 1.16 Knows someone who works with victims 0.60* 0.61* 0.63* Knows someone who has been to prison 0.66* 0.64* 0.67* <30 1.03 >65 0.73 Male 0.89 Separated 0.68 Married or living with 0.70 Children 1.44 Diploma of collegiate studies 1.50 University 1.03 Note. †p ≤ 0.10; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001. aICC (logistic reg) = (Level2 Var)/(Level2 var + (π2/3)) as suggested by Snijders et al. (2012). Crime seriousness The analysis shows that, as expected, the more serious someone judges a crime to be, the more likely they are to think that the individual’s sentence will be too lenient. This suggests that people are aware that their assessment of the seriousness of a crime can differ (e.g., they know they “overestimate” the crime’s seriousness) and that this can partially explain why they believe the sentences will not be sufficiently severe (they know that the judge chose the right sentence resorting on proportionality between the punishment and an assessment of the seriousness of the crime). However, it is important to note that not all violent and serious crimes lead to lenient opinions. As shown in Table 3, dissatisfaction regarding homicide or robbery is pretty low (respectively 26% and 20%). Knowledge of the sentence Research has shown a link between more general dissatisfaction and a lack of public knowledge about the work of the courts (Gelb, 2006; Roberts & Stalans, 1997; Hough & Roberts, 1998; Mattinson & Mirrlees-Black, 2000; Chapman et al., 2002). However, our results do not enable us to observe this relationship when the opinion is more specific and placed into context. Indeed, the variable that differentiates cases where respondents estimated the sentence with those in which they gave their opinion on the judge’s sentence is not significant. This result is contrary to what we expected but is in line with those obtained by St-Amand and Zamble (2001), who concluded that informing the public about the likely sentence of the courts in a detailed question does not significantly influence their acceptance of its severity. Our results suggest that informing the public about the real work of the courts (the solution usually proposed, see for example, the initiative of the British Home Office in the 2000s) would not necessarily have an impact on more specific views of the courts. This can be explained in part by the fact that, once freed from their traditional stereotypes (as expressed in the more general questions), respondents have a good knowledge of the court’s work (as shown in the findings in Table 3) and an opinion that is already more nuanced. Crime experience Our analyses of crime experiences also reveal that the odds of being dissatisfied with sentencing are twice as high if the respondent had been a direct or a vicarious victim of a crime. While this relationship may be expected, it contrasts with empirical literature that suggests that victims are not harsher or more negative in their view of justice or the judicial system (see, for example, Unnever, Cullen, & Fisher, 2007). The difference can probably be explained by our measure of victimization, which takes into account the type of crime assessed and does not restrict victimization experiences to the last year or to personal victimization (as it includes victimization of close family). Additional analyses concluded that a more general measure of victimization, which took into account victimization experiences for 15 crimes, has no effect on these opinions. Involvement in the court system Some authors have suggested that people’s dissatisfaction stems from a lack of knowledge combined, among other things, with an absence of involvement in the criminal justice system (Gelb, 2006; Mattinson & Mirrlees-Black, 2000; Chapman et al., 2002). Although this lack of direct contact can potentially explain the more general opinion of the courts found in some studies (see Van de Walle, 2009), it does not seem to be linked to opinions on more specific cases. Indeed, having been involved in a criminal case as a witness, a victim (or relative of a victim), or a defendant (or relative of a defendant) did not influence respondents’ opinions about the work of the courts in the six cases presented. These experiences have a significant impact only in cases where people considered the sentence imposed by the judge to be too lenient. Table 4 shows that people who had been involved in a case where they considered the sentence to be too lenient were one and a half times more likely to be dissatisfied with the severity of the courts. Indirect knowledge of the criminal justice system While knowing a lawyer or someone who works with offenders has no effect on the opinions of citizens,15 three types of contacts can have an influence. First, a person who knows someone who has been in prison is almost one and half times more likely to be satisfied with the work of the courts. This effect may be explained by the fact that people who have been in prison can break down some stereotypes about the judicial process, such as conditions of prison life and the effect of punishment on their criminal life. Knowing someone who works with victims has the same effect of increasing citizen satisfaction (odd ratio: 1.6). This result may, a priori, seem surprising, as victims often report dissatisfaction with their judicial experience. However, past researches showed that victims are rarely dissatisfied with the sentence and that their dissatisfaction most often concerns the place or the voice they did not have in the judicial process (Wemmers, 1996). Finally, people who know a judge are more likely to be dissatisfied with sentencing, but the effect is marginally significant and only emerges when the education level of citizens is considered (there is some covariance between these variables as people who know a judge are twice as likely to have a university degree—69% vs. 39%). Like the press, judges probably tend to talk about the more unusual or frustrating cases they presided over. In their personal relationships, judges have no obligation to inform and demystify the criminal justice system and, like other citizens, they may tend to focus on the more negative or darker sides of their work and of the criminal justice system. Given its marginal significance, this result should be interpreted with the usual caution. Personal characteristics Finally, Table 4 also shows that, contrary to our hypothesis, the characteristics of respondents are not related to their views on the work of the courts.16 However, this result is not surprising when one takes into account the variability of opinions among individuals. As shown in Table 5, because a respondent is dissatisfied with the work (actual or anticipated) of the courts in a particular case does not mean that that he or she will necessarily be dissatisfied with their work in another case. Most respondents (around 60%) reported being dissatisfied with one to three of the six scenarios, showing that their attitudes are not fixed and need to be explored in multiple contexts. This conclusion is important because it suggests that it may be unproductive to seek personal explanations (age, sex, or socioeconomic status) for public dissatisfaction because this dissatisfaction is nuanced and therefore subject to change depending on the context. Table 5 Distribution of Respondents by Number of Cases They Consider the Sentence Being too Lenient Number of cases % of respondents 0 17.4 1 20.5 2 22.6 3 19.5 4 13.8 5 4.6 6 1.5 Number of cases % of respondents 0 17.4 1 20.5 2 22.6 3 19.5 4 13.8 5 4.6 6 1.5 Table 5 Distribution of Respondents by Number of Cases They Consider the Sentence Being too Lenient Number of cases % of respondents 0 17.4 1 20.5 2 22.6 3 19.5 4 13.8 5 4.6 6 1.5 Number of cases % of respondents 0 17.4 1 20.5 2 22.6 3 19.5 4 13.8 5 4.6 6 1.5 Conclusion Some authors have emphasized the differences between general and specific attitudes of citizens, but they always rely on two different concepts (opinion about court decisions vs. sentencing expectations) to arrive at these conclusions (Zamble & Kalm, 1990, or Applegate et al., 1996). The analyses presented in this article differ in that they are based on alternative measures of the opinions of citizens. Following Warner and Davis (2011), the results confirm that there is a significant difference between the general opinion that citizens express on an abstract question about the work of the courts and the specific opinion that they give when asked to look at the work of the courts in a specific case. While the more general opinion seems to be an emotional response that can reveal perceived injustices and is often based on stereotypes about offenders or the criminal justice system (an automatic response from System 1 while processing information), specific views are more nuanced and result from a more analytic reasoning (System 2). This article suggests that dissatisfaction is not real because it disappears for most citizens when they are put in a decision-making situation (thus, activating System 2) where the use of stereotypes is limited by the provision of information about the details of the criminal case. It is therefore wrong to conclude that people want tougher penalties for all violent or serious crimes—citizens’ views are much more complex and cannot be easily generalized. Experiences with crime or the criminal justice system may also explain people’s opinions about sentencing. The variable with the most impact in this study was having been a victim of a crime (directly or indirectly through close family). As with the factor of having been involved in a case where the sentence handed down was considered too lenient, such negative experiences tarnish the respondent’s view of the court decisions. Only two variables lead respondents to more positive opinions: knowing someone who has been in prison and someone who works with victims. It is possible that direct experience with the court system (such as having attended a trial) does not change the image people have of courts because they do not believe that the case in question was representative of the daily activities of the courts (an argument noted by Warner and Davis, 2011) or because the experience is marginal when compared with the amount of information and experiences generated from the media (Dorfman & Schiraldi, 2001). Finally, studies that examine the degree of punitiveness of individuals should consider how this notion is conceptualized and integrate more flexible measures that vary depending on the context. If the results of previous studies are so contradictory and individual characteristics have such a low explanatory power (Frost, 2010), it is because the distinction between punitive individuals and nonpunitive (or satisfied vs. dissatisfied) individuals is not as simple as we hoped. Some individuals are potentially more punitive in some contexts, while remaining more lenient than their peers in other situations. Of the three cases presented, only 5% of respondents were consistently in the top quartile of the toughest respondents and 4% were consistently in the quartile of the less severe respondents. Measuring the punitiveness of an individual according to response to abstract questions or by adding up responses to several specific situations remain strategies that underestimate the adaptability of individuals. They fail to bring out the rational, analytical, and intelligent aspects of public perceptions and opinions. Funding The research had been funded by the Social Sciences and Humanities Research Council. Chloé Leclerc is an Assistant Professor of Criminology at Université de Montréal and a researcher at the International Centre for Comparative Criminology. Her research focuses on the penal and judicial system. She had worked on criminal justice professionals and citizens’ conceptualization of “just sentencing.” Anta Niang is a PhD candidate of the Department of Psychology in Université Rennes 2 at the Laboratoire de Psychologie: Cognition, Comportement et Communication (LP3C, EA 1285). Her research interests focus on social judgments about crime, criminals, and sentencing. She currently works on the impact of psychological and psychiatric expertise on jurors’ decision making. Marie-Chloé Duval has a Master’s Degree from the School of Criminology, at Université de Montréal. She had worked on media representation of criminal trials. Footnotes 1De grandpré, H. (2011). Les conservateurs prêts pour une loi omnibus. La Presse, 19 September. 2This result is contrary to the findings of De Keijser, Van Koppen, and Elffers (2007) and of Tremblay, Cordeau, and Ouimet (1994) who found that, as the public underestimates the severity of the courts, the gap between the two groups’ understanding of how severe sentences should be is even more important than one might have anticipated. 3A “general” opinion captures the overall views that people possess about a policy issue. “Specific” opinions are the views they express when the policy is applied to a case that has a certain set of attributes. These “specific” opinion/attitudes are especially relevant to criminal justice policy because decisions are made about cases that involve offenders, victims, and acts, which may vary on many dimensions and interact in unique ways (Cullen et al., 2000). 4The percentage varies depending on the type of crime. Just >50% of citizens considered sentences for drug offenses and crimes against property to be too lenient. This percentage increases to >75% for violent crimes or crimes of a sexual nature. 5The sample was randomly chosen from a list of voluntary participants drawn up using a telephone survey of the general public (randomly selected). The sample was stratified by gender and age to reflect the demographic importance of Francophone citizens living in the Montreal area (the survey was only conducted in French). The final sample reflects most of the characteristics of the population of Montreal, albeit with a slightly higher proportion of common-law or married people (49% as compared with 40% in the total population) and people with either a college education (21.6% compared with 15% in the total population) or a university degree (43.3% compared with 31.8%). Since socio-demographic variables are known to have a low impact on people’s opinions and attitudes (Roberts & Indermaur, 2007: 58) and given that the purpose of the article is less to provide a representative view of the population than to propose an alternative way of measuring public opinion and to determine the effect that some experience with the justice system can have on public opinion, these differences do not undermine the main conclusions of the article. 6Since sentences of different nature are not easy to compare, and since all respondent provided prison term without mentioning a problem, we limited our analyses to prison terms in this article. 7Respondent were told, that the individual was presenting his request on his first admissibility to parole, namely, after serving a third of his prison term. 8Multicollinearity was tested among all independent variables (the highest Spearman’s rho was <0.5). 9Although it may seem surprising that many citizens have evaluated the crime of incest as severely as the crime of homicide, it must be emphasized that the attitude of the accused can greatly influence the severity perceived by citizens. In the homicide case, the man regretted his actions and called the emergency services, whereas the father in the incest case did not seem to be remorseful and maintained that his daughter had consented. 10For the first offense, we checked whether the person had been a victim of a crime involving stolen goods, but we did not include victims of theft because in the scenario it is a company that was robbed. For the second offense, we considered victims of murder or attempted murder. For the third offense, we considered victims of sexual assault (11%) and sexual touching of a child (15%). For the fourth offense, we considered everyone who had been the victim of a robbery (15.5%) or a car theft (22.5%) since the offender was convicted of both offenses. 11Taylor (2006) and Wallander (2009) discussed the debate regarding the necessity of using hierarchical regression while working with vignettes nested in respondents. With our data, models from ordinary logistic regression lead to the exact same results as were obtained with hierarchical logistic regression. 12When no account is taken of the dependency of the error terms then false degrees of freedom are used (based on the number of cases rather than the number of individuals) and this favors type II errors, i.e. the erroneous discovery of significant relationships. 13The analysis was performed using Mplus 7.4 software, adopting the penalized quasi-likelihood (PQL) as the estimation procedure. 14Chi-square tests reveal that the percentage of dissatisfaction in these four scenarios is statistically different from the percentage of dissatisfaction measure for the general question (p < .001). 15The absence of relation between knowing a lawyer and citizen’s opinion could possibly be explained by the absence of specification in the survey of the type of criminal lawyer. As people may answer that they know a lawyer referring to a civil or corporate lawyer, it is possible that this variable does not adequately capture the knowledge of a criminal lawyer. 16As some characteristics could have different effect depending on the type of crime, we test some interaction effects (e.g., being a woman or a parent and judging a case of incest; being a woman and judging domestic violence/homicide), but we found no significant effect. Acknowledgements Authors wish to thanks Carlo Morselli for his useful observations and suggestions and Joan McGilvray for her attentive revision of this article. References Applegate B. K. , Cullen F. T. , Turner M. G. , Sundt J. L. ( 1996 ). Assessing public support for three-strikes-and-you’re-out laws: Global versus specific attitudes . Crime and Delinquency , 42 , 517 – 34 . doi: 10.1177/0011128796042004002 Google Scholar CrossRef Search ADS Benesh S. C. , Howell S. E. ( 2001 ). Confidence in the courts: A comparison of users and non users . Behavioral Sciences and the Law , 19 , 199 – 214 . doi: 10.1002/bsl.437 Google Scholar CrossRef Search ADS PubMed Brillon Y. ( 1988 ). Punitiveness, status and ideology in three Canadian Provinces. In Walker N. , Hough M. (Eds.), Public attitudes to sentencing: Survey from five countries . Hampshire : Gower . doi: 10.1016/0047-2352(89)90038-x Casey S. , Mohr P. ( 2005 ). Law-and-order politics, public-opinion polls and the media (Australia) . Psychiatry, Psychology and Law , 12 ( 1 ), 141 – 151 . doi: 10.1375/pplt.2005.12.1.141 Google Scholar CrossRef Search ADS Chapman B. , Mirrlees-Blaws C. , Brawn C. ( 2002 ). Improving public attitudes to the Criminal Justice System: The impact of information . London : Home Office Research Studies . doi: 10.1037/e454602008-001 Cullen F. T. , Fisher B. S. , Applegate B. K. ( 2000 ). Public opinion about punishment and corrections. In Tonry M. (Ed.), Crime and Justice: A review of research (pp. 1 – 79 ). Chicago : University of Chicago Press . doi: 10.1086/652198 De Keijser J. W. , Van Koppen P. J. , Elffers H. ( 2007 ). Bridging the gap between judges and the public? A multi-method study . Journal of Experimental Criminology , 3 , 131 – 161 . doi: 10.1007/s11292-007-9031-3 Google Scholar CrossRef Search ADS Doob A. N. , Roberts J. V. ( 1983 ). Sentencing: An analysis of the public's view of sentencing . Ottawa : Department of Justice Canada . Doob A. N. , Roberts J. V. ( 1988 ). Public punitiveness and public knowledge of the facts: Some Canadian surveys. In Walker N. , Hough M. (Eds.), Public attitudes to sentencing: Surveys from five countries . Aldershot : Gower . doi: 10.1016/0047-2352(89)90038-x Durham A. M. , Elrod H. P. , Kinkade P. T. ( 1996 ). Public support for the death penalty: Beyond Gallup . Justice Quarterly , 13 , 705 – 736 . doi: 10.1080/07418829600093161 Google Scholar CrossRef Search ADS Dorfman L. , Schiraldi V. ( 2001 ). Off balance: Media coverage of youth crime . Guild Practitioner , 58 , 75 – 78 . Frost N. A. ( 2010 ). Beyond public opinion polls: Punitive public sentiment and criminal justice policy . Sociology Compass , 4 , 156 – 168 . doi: 10.1111/j.1751-9020.2009.00269.x Google Scholar CrossRef Search ADS Garland D. ( 2001 ). The culture of control, crime and social order in contemporary society . Oxford : Oxford University Press . Gelb K. ( 2006 ). Myths and misconceptions: Public opinion versus public judgment about sentencing . Melbourne : Sentencing Advisory Council . Gray J. M. ( 2009 ). What shapes public opinion of the criminal justice system. In Wood J. , Gannon T. (Eds.), Public opinion and criminal justice (pp. 49 – 72 ). Portland : Willan Publishing . doi: 10.4324/9781843926610 Gwin J. S. ( 2010 ). Juror sentiment on just punishment: Do the federal sentencing guidelines reflect community values? Harvard Law and Policy Review , 4 , 173 – 200 . Hough M. , Roberts J. V. ( 1998 ). Attitudes to punishment: Findings from the British crime survey . London : Home Office Research Study . Hough M. , Roberts J. V. ( 1999 ). Sentencing trends in Britain: Public knowledge and public opinion . Punishment and Society , 1 ( 1 ), 11 – 26 . doi: 10.1177/14624749922227685 Google Scholar CrossRef Search ADS Hox J. J. , Kreft I.G.G. ( 1994 ) Multilevel analysis methods . Sociological Methods and Research , 22 , 283 – 299 . doi: 10.1177/0049124194022003001 Google Scholar CrossRef Search ADS Hox J. J. ( 2010 ). Multilevel analysis, techniques and application ( 2nd ed.). New York, NY : Routledge . Hutton N. ( 2005 ). Beyond populist punitiveness? Punishment and Society , 7 , 243 – 258 . doi: 10.1177/1462474505053828 Google Scholar CrossRef Search ADS Indermaur D. ( 1987 ). Public perception of sentencing in Perth, Western Australia . Australian and New Zealand Journal of Criminology , 20 , 163 – 183 . doi: 10.1177/000486588702000304 Google Scholar CrossRef Search ADS Indermaur D. , Hough M. ( 2002 ). Strategies for changing public attitudes to punishment. In Roberts J. , Hough M. (Eds.), Changing attitudes to punishment: Public opinion, crime and justice (198-214) . Cullompton, Devon : Willan Publishing . doi: 10.4324/9781843924272 Jewkes Y. ( 2004 ). Media and Crime: Key approaches to criminology . Sage : London . Kahneman D. ( 2011 ). Thinking, fast and slow . Macmillan : London . Langworthy R. H. , Whitehead J. T. ( 1986 ). Liberalism and fear as explanations of punitiveness . Criminology , 24 , 575 – 91 . doi: 10.1111/j.1745-9125.1986.tb00391.x Google Scholar CrossRef Search ADS Leclerc C. ( 2010 ). La métrique de la juste peine: Une analyse des décisions de justice prises par les acteurs judiciaires et le public . Montreal : Doctoral dissertation . Mattinson J. , Mirrlees-Black C. ( 2000 ). Attitudes to crime and criminal justice: Findings from the 1998 British Crime Survey . Home Office Research Study No. 200. London : Research, Development and Statistics Directorate, Home Office . doi: 10.1037/e454372008-001 Mirrlees-Black C. ( 2002 ). Improving public knowledge about crime and punishment. In Roberts J. V. , Hough M. (Eds.), Changing attitudes to punishment: Public opinion, crime and justice (pp. 184 – 197 ). Portland : Willan Publishing . doi: 10.4324/9781843924272 Roberts J. V. ( 1988 ). Public opinion and sentencing: The surveys of the Canadian sentencing commission . Ottawa, ON : Department of Justice Canada . Roberts J. V. , Doob A. N. ( 1989 ). Sentencing and public opinion: Taking false shadows for true substances . Osgood Hall Law Journal , 27 , 491 – 515 . Roberts J. V. , Stalans L. J. ( 1997 ). Public opinion, crime, and criminal justice (pp. 35 – 52 ). Boulder, CO : Westview Press . doi: 10.1086/449205 Roberts L. D. , Indermaur D. ( 2007 ). Predicting punitive attitudes in Australia . Psychiatry, Psychology and Law , 14 ( 1 ), 56 – 65 . doi: 10.1375/pplt.14.1.56 Google Scholar CrossRef Search ADS Roberts J. V , Stalans L. , Indermaur D. , Hough M. ( 2003 ). Penal populism and public opinion: Lessons from five countries . Oxford : Oxford University Press . Rossi P. H. , Berk R. A. , Campbell A. ( 1997 ). Just punishments: Guideline sentences and normative consensus . Journal of Quantitative Criminology , 13 , 267 – 290 . doi: 10.1007/bf02221093 Google Scholar CrossRef Search ADS Rottman D. B. , Hansen R. , Mott N. , Grimes L. ( 2003 ). Perceptions of the courts in your community: The influence of experience, race and ethnicity . Williamsburg, VA : National Center for State Courts . Snijders T. A. B. , Bosker R. J. ( 1999 ). Multilevel analysis. An introduction to basic and advanced multilevel modeling . London : Sage Publishers . Snijders T. A. B. , Bosker R. J. ( 2012 ). Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling, second edition . Sage Publishers : London . Sprott J. B. ( 1996 ). Understanding public views of youth crime and the youth justice system . Canadian Journal of Criminology , 38 , 271 – 290 . Sprott J. B. , Doob A. N. ( 1997 ). Fear, victimization, and attitudes to sentencing, the courts and the police . Canadian Journal of Criminology , 39 , 275 – 291 . Sprott J. B. ( 1998 ). Understanding public opposition to a separate youth justice system, Crime and Delinquency , 44 , 399 – 411 . doi: 10.1177/0011128798044003004 Google Scholar CrossRef Search ADS Sprott J. B. , Webster C. M. , Doob A. N. ( 2013 ). Punishment severity and confidence in the criminal justice system . Canadian Journal of Criminology and Criminal Justice , 55 , 279 – 292 . doi: 10.3138/cjccj.2012.e.27 Google Scholar CrossRef Search ADS Stalans L. J. ( 1993 ). Citizens' crime stereotypes, biased recall, and punishment preferences in abstract cases: The educative role of interpersonal sources . Law and Human Behavior , 17 , 451 . doi: 10.1007/bf01044378 Google Scholar CrossRef Search ADS Stalans L. J. , Diamond S. S. ( 1990 ). Formation and change in lay evaluations of criminal sentencing . Law and Human Behavior , 14 , 199 – 214 . doi: 10.1007/bf01352749 Google Scholar CrossRef Search ADS St-Amand M. , Zamble E. ( 2001 ). Impact of information about sentencing decisions on public attitudes toward the criminal justice system . Law and Human Behavior , 25 , 515 – 528 . doi: 10.1023/a:1012844932754 Google Scholar CrossRef Search ADS PubMed Taylor B. J. ( 2006 ). Factorial surveys: using vignettes to study professional judgement . British Journal of Social Work , 36 , 1187 – 1207 . Google Scholar CrossRef Search ADS Tremblay P. , Cordeau G. , Ouimet M. ( 1994 ). Underpunishing offenders: Toward a theory of legal tolerance . Revue canadienne de criminology , 36 ( 4 ), 407 – 434 . Tufts J. ( 2000 ). Understanding Public Attitudes Toward Sentencing (Master thesis, Department of Criminology, Université of Ottawa, Ottawa). Unnever J. D. , Cullen F. T. , Fisher B. S. ( 2007 ). "A liberal is someone who has not been mugged": Criminal victimization and political beliefs . Justice Quarterly , 24 , 309 – 334 . doi: 10.1080/07418820701294862 Google Scholar CrossRef Search ADS Van de Walle S. ( 2009 ). Confidence in the criminal justice system: Does experience count? British Journal of Criminology , 49 , 384 – 398 . doi: 10.1093/bjc/azp001 Google Scholar CrossRef Search ADS Voumvakis S. , Ericson R. V. ( 1984 ). News account of attacks on women: A comparison of three Toronto newspapers . Toronto : Centre of criminology, University of Toronto . Wallander L. ( 2009 ). 25 years of factorial surveys in sociology: A review . Social Science Research , 38 , 505 – 520 . doi: 10.1016/j.ssresearch.2009.03.004 Google Scholar CrossRef Search ADS Warner K. , Davis J. ( 2011 ). Using jurors to explore public attitudes to sentencing . British Journal of Criminology , 51 , 1 – 20 . doi: 10.1093/bjc/azr066 Google Scholar CrossRef Search ADS Wemmers J.-A. ( 1996 ). Victims in the Criminal Justice System . Amsterdam : Kugler Publications . Zamble E. , Kalm K. L. ( 1990 ). General and specific measures of public attitudes toward sentencing . Canadian Journal of Behavioural Science , 22 , 327 – 337 . doi: 10.1037/h0078918 Google Scholar CrossRef Search ADS © The Author 2017. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal of Public Opinion Research Oxford University Press

Understanding the Relationship Between Public Opinion and Experience With the Criminal Justice System

Loading next page...
 
/lp/ou_press/understanding-the-relationship-between-public-opinion-and-experience-8b0OV4zXbs
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved.
ISSN
0954-2892
eISSN
1471-6909
D.O.I.
10.1093/ijpor/edx007
Publisher site
See Article on Publisher Site

Abstract

Abstract Surveys indicate that between 70% and 80% of people consider judges to be too lenient. This current research confirms the inadequacy of this general measure of public opinion by contrasting it with alternative opinion-based measures of court sentencing across specific criminal cases. The article also explores the impact of public’s experiences of the criminal justice system on their opinions about the courts. Having been a victim or attending a criminal case where the punishment was seen as being too lenient seems to cause greater dissatisfaction, whereas contact with people working with victims or who have been in prison improves the level of satisfaction. In September 2011, when he was being questioned about the rationale for a draft law on crime, a Conservative minister in Canada’s House of Parliament proclaimed, “We believe that the best experts in Canada are the voters who give us the mandate to do what we said we would do.”1 Indeed, and, as Garland (2001) once suggested, the dominant voice in matters of penal policy is no longer that of the experts, but that of the public. But what does the public really want in terms of sentencing? The results of surveys on criminal justice are disturbing: between 70% and 80% of the population feel that the sentences imposed by the courts are too lenient. According to Roberts, Stalans, Indermaur, and Hough (2003), this result is constant across both time and place, although rates are slightly lower in Canada (between 60% and 70% dissatisfied) and slightly higher in the United States (up to 85% dissatisfied). Politicians generally rely on such popular outlooks to justify measures that increase the length of sentences or more general tough-on-criminal stances (Sprott, Webster, & Doob, 2013). However, several researchers have demonstrated that the gap between public demand and provision by the courts is not considerable2 (Sprott, 1998; Hough & Roberts, 1998; Roberts & Stalans, 1997; Rossi, Berk, & Campbell, 1997; Zamble & Kalm, 1990; Roberts & Doob, 1989; Doob & Roberts, 1988). In several studies, it has been shown that when people were given a detailed vignette and subsequently asked to determine the suitable sentence, their recommendations were as severe or less severe than the sentences generally imposed by courts (Stalans & Diamond, 1990; Zamble & Kalm, 1990; Hough & Roberts, 1999). The same conclusion is reached when examining opinions about support for repressive policies. Applegate and his colleagues (1996) found that, when asked about their support for the “Three Strikes” policy, citizens may express clear support, but this drops dramatically when people are given a specific scenario and asked, for example, what sentence they would give someone convicted of a serious offense for the third time. Similarly, Durham, Elrod, and Kinkade (1996) found that when respondents were presented with 17 scenarios describing situations in which murder had been committed, the percentage of respondents supporting the death penalty was both higher (91%) and lower (31%) (depending on the scenario) than the percentage obtained with a general question about support for that penalty (80% in favor). Such nuances are rarely presented when presenting or discussing public opinion measures. Thus, there is a more general factor that is often overlooked. An individual’s knowledge base and experience with the criminal justice system will have a direct effect on how that person perceives leniency and severity in court cases and decisions. This article develops this general point. First, by assessing the limits of traditional measures of public opinion on sentencing. Second, by a review of past research that explained the public’s apparent dissatisfaction with courts and the criminal justice system. The remainder of this article then focuses on a series of analytical demonstrations that support alternative outlooks and measures of this phenomenon, with specific attention on the respondent’s knowledge-based and experience thesis. Limits of Public Opinion Measures There are three common criticisms of opinion polls. First, the general questions in surveys are subject to a variety of interpretations (Casey & Mohr, 2005). For instance, studies showed that people think of different crimes (often, but not always, violent crime) when expressing an opinion about court sentences in general3 (Indermaur, 1987; Sprott, 1996; Doob & Roberts, 1983). Second, polls are criticized for capturing the emotional responses of citizens, responses that are often volatile (Rossi, et al., 1997; Casey & Mohr, 2005) and more complex than suggested by the polls (Hutton, 2005; Roberts & Stalans, 1997; Cullen, Fisher, & Applegate, 2000). Third, citizens are required to give an opinion about a reality with which they are unfamiliar. Surveys of public knowledge show unequivocally that people almost systematically overestimate frequency and severity of crimes, as well as underestimating the severity of courts (Gray, 2009; Hough & Roberts, 1998; Mattinson & Mirrlees-Black, 2000; Roberts & Stalans, 1997; St-Amand & Zamble, 2001). Studies on heuristics suggest that in the absence of specific information or details, people fill the gaps by resorting to stereotypes (Stalans, 1993; Stalans & Diamond, 1990). Because most people get the bulk of their knowledge about the courts from the media, we can assume that people will generally take as their reference point cases that are serious, unusual, and not representative of the reality of the courts (Jewkes, 2004). These limitations have prompted several researchers to question the current use of opinion polls on criminal justice. It is frequently suggested that research should offer new ways of gauging public opinion by providing as much detail as possible to citizens so that they have the necessary information to formulate and support their opinions. However, there are only few studies that have ventured down this path. Warner and Davis’s study (2011) is both unique and innovative because it looked at a new way of measuring opinion about court sentences by questioning people who had recently served as jurors in one or more criminal cases. The first result of the study confirms the conclusions put forward by other researchers (see Gwin, 2010, for a comparison of juror recommendations and sentencing guidelines): the jury generally recommended sentences that were less severe than the one handed down by the judge in the same case. A more interesting result is that although the vast majority of jurors expressed a general dissatisfaction with the leniency of court sentences4 for the type of crime they had to judge, 90% found the final sentence imposed by the judge to be appropriate. While some jurors, whom the authors referred to as the “extrapolators,” used their experience and knowledge of the sentences of the courts to more favorably reevaluate their opinion about the broader work of the courts, others remained unaffected. What Warner and Davis’s study shows us is that there is a significant difference between a more general opinion expressed in a survey and an opinion that is more specific and placed in the context of the various contingencies that comprise a decision-making process. However, this study does not provide information about whether it was access to detailed facts, experience as a juror, or knowledge about the judge’s sentence that led to more positive opinions. Furthermore, as the study involved people who had recently served as jurors, we do not know its relevance for the majority of citizens, who typically have no experience with the criminal justice system. Explaining Public Opinion Knowing that incorrect knowledge about criminality and the courts encourages greater dissatisfaction with courts (Gelb, 2006; Roberts & Stalans, 1997; Hough & Roberts, 1998; Mattinson & Mirrlees-Black, 2000; Chapman, Mirrlees-Blaws, & Brawn, 2002), some researchers have suggested that, if the public knew more about the criminal justice system, they would be more satisfied with it (see Mirrlees-Black, 2002; Indermaur & Hough, 2002 for a discussion on that topic). However, empirical studies of this particular hypothesis provide contradicting results. Examining the effect of court experiences on public confidence in the criminal justice system, Van de Walle (2009) concluded that research in this area is “inconclusive” and noted a number of limitations. He and others (Benesh & Howell, 2001; Rottman, Hansen, Mott, & Grimes, 2003) pointed out the importance of taking into account different dimensions of experience with the criminal justice system: the level of implication (direct or indirect experiences), the timing (past or recent), the outcome (positive or negative), the role played; and the motivation to participate in the criminal justice system. Despite their recommendations, few researchers have considered the variability of experience (Van de Walle, 2009; Benesh & Howell, 2001). Most studies interested in opinions about sentencing described the typical “punitive” or “dissatisfied” citizen by presenting sociodemographic characteristics or attitudes. Roberts and Indermaur (2007), for example, tried to predict punitive attitudes (measured by support for death penalty, stiffer sentencing, and sentences reflecting public opinion) with different sets of factors. They concluded that, “It appears to be people’s knowledge and beliefs of crime and the criminal justice system that underpin punitive attitudes to a greater extent than demographics, political orientation, religious attendance or media use.” However, these kinds of studies that explain public opinion about court decisions all rely on general measures of public dissatisfaction. Even if these general measures or questions (e.g., is the sentence imposed by the court too lenient? Would you support stiffer sentencing?) are recognized to be severely biased, they continue to constitute the basis of research on public opinion about sentencing. Current Study In the first part of the article, three different measures of public opinion are proposed to highlight differences between general and specific opinions. According to Kahneman (2011), people think and process information by two different processes. The first process (referred to as System 1) is related to intuition and emotions. It treats information automatically, quickly, and with little or no cognitive effort. This process uses stereotypes and prior information to give a fast and simple answer. The second process (System 2) is more cognitively challenging and is mobilized when the situation requires analytical thinking. It processes information more slowly and involves a level of concentration and reasoning. Based on these ideas, we can hypothesize that when people are asked for a general opinion, they will answer more spontaneously (using System 1) and their answer is more likely to be based on stereotypes (most often inspired by the media representation of crime; Voumvakis & Ericson, 1984). On the other hand, when given detailed information about a specific case, individuals will be forced to use System 2 in dealing with the information, which will lead to a change in their point of reference and a more informed opinion. In the second part of the article, we use hierarchical regression analysis to examine the impact of different experiences with crime and the justice system on individual opinions about the leniency of the courts. In this article, we use two measures of public opinion that are put into context (reducing possible biases) and different types and levels of experience to gain a better understanding of what matters in court experiences. As the presentation of specific information would mobilize System 2, the respondent will make an effort to analyze the information provided to them. In such a context, past experiences or indirect knowledge acquired by contact with criminal justice actors might also be used, affecting an individual’s thought process. Method Data were collected during the summer of 2009 through a survey conducted with 200 French Canadian participants living in Montreal.5 Face-to-face interviews were conducted with each participant and respondent’s responses were noted on a paper questionnaire by the interviewer. The questionnaire consisted in 57 pages and lasted on average 90 min. The survey began with questions on opinions and knowledge about the court, the parole system, the crime, and offenders. Respondents were then provided three detailed criminal cases (see Appendix I for a sample description) and asked to sentence the defendant and give their view about the case (e.g., sentence provided by the court, sentencing goal). Questions about experiences with crime (as a victim or an offender) and the criminal justice or its actors (judge, police, etc.) were included before respondents had to evaluate three other crime scenarios. In these last three scenarios, respondents had to decide whether to grant parole or not and give their view about the case (e.g., opinion on the sentence originally imposed, recidivism risk). The survey ended with a set of sociodemographic questions (age, sex, education, etc.). Measuring Public Opinion At the beginning of the questionnaire, people were asked to answer the classic opinion poll question: “In general, would you say that sentences handed down by the courts are too severe, about right, or too lenient.” In the first three cases, participants had to make a series of decisions on the appropriate sentence to be imposed on three men convicted of various crimes: Peter, guilty of theft and handling stolen goods worth $8,000; Paul, guilty of a domestic homicide; and John, guilty of incest with his daughter. For each criminal case, respondents had to indicate the sentence they would impose, the length of sentence they considered acceptable (the minimum and the maximum sentence they would agree with), and the sentence they believed the courts would have imposed.6 Past research has demonstrated that citizens, as judicial actors, do not have a fixed concept of a just sentence, but are more likely to suggest wide ranges of sentences that are perceived as equally acceptable (Leclerc, 2010). It is therefore imperative to see whether the sentence they think the courts imposed for each case lies within this range. For each of the three criminal cases, we created a new opinion variable, the anticipated sentence of the court, to check whether the anticipated sentence was (1) too severe (more severe than the maximum acceptable sentence); (2) about right (between the minimum and maximum acceptable sentence); or (3) too lenient (less severe than the minimum acceptable sentence). This conception of public opinion is interesting for two reasons. First, it focuses on opinion and not on sentencing expectations. Researchers have often concluded that there is a difference between general and specific opinion by comparing responses with general questions on the work of the courts with the sentences that the public proposed for different types of offense (Stalans & Diamond, 1990; Zamble & Kalm, 1990; Hough & Roberts, 1999). Thus, they merged two different concepts: the opinion (“what I think of the courts”) and the sentence required (“what I hope the courts will impose”). The measure we propose avoids this problematic merge. Second, the measure reflects the willingness of the respondents to negotiate. It is not because an individual wants to impose 5 years in prison and believes that the court gave 3 years that he will necessarily be dissatisfied. This measure takes into account the openness of citizens to various penal options, which they consider as equally satisfactory. In the last three scenarios, participants had to decide whether to grant parole to three individuals who had requested it7: Thran, serving 2 years less a day in prison for a car theft and a robbery; Costa, jailed for 2 years in prison for dealing in hard drugs; and Hervé, sentenced to 18 months in prison for domestic violence. As one of the questions on the assessments made of each of these three cases, participants were asked what they thought of the sentence imposed by the courts. As with the earlier variable, they could say that the sentence seemed to be too severe, about right, or too lenient. This variable was then used as a measure of specific opinion on the actual sentence of the courts. This measure has the advantage of providing a sentencing “tariff” for respondents, making it possible to analyze opinion that is no longer biased by a lack of knowledge about the reality of the court. It also provides the opportunity to test the hypothesis that citizens are dissatisfied with judges’ decisions because they underestimate the severity of the courts. Explaining Public Opinion Tables 1 and 2 present all independent variables8 used to explain why some citizens consider sentences, in particular case, as being too lenient. As few people consider courts to be “too severe,” we decided to merge these people with the ones who consider the court decision to be “about right.” Moreover, people interested in public opinion are generally focusing on the “punitive person,” who considers the court to be too lenient. Models presented in this article thus compare sentences that are considered too lenient with all others (e.g., sentence considered severe enough and too severe). Additional analyses show identical results when excluding sentences considered too severe (7.8% of the sample). Table 1 Descriptive Measures of Independent Variables (Level 1) Type of crime Crime seriousness (1–100) (Median score) Victimization (% of respondents saying that a close family or themselves have been a victim of this type of crime) Theft and Handling stolen goods 40 10 Domestic homicide 90 3 Incest 90 21 Robbery and car theft 70 33 Hard drugs trafficking 60 14 Domestic violence 80 16 Type of crime Crime seriousness (1–100) (Median score) Victimization (% of respondents saying that a close family or themselves have been a victim of this type of crime) Theft and Handling stolen goods 40 10 Domestic homicide 90 3 Incest 90 21 Robbery and car theft 70 33 Hard drugs trafficking 60 14 Domestic violence 80 16 Table 1 Descriptive Measures of Independent Variables (Level 1) Type of crime Crime seriousness (1–100) (Median score) Victimization (% of respondents saying that a close family or themselves have been a victim of this type of crime) Theft and Handling stolen goods 40 10 Domestic homicide 90 3 Incest 90 21 Robbery and car theft 70 33 Hard drugs trafficking 60 14 Domestic violence 80 16 Type of crime Crime seriousness (1–100) (Median score) Victimization (% of respondents saying that a close family or themselves have been a victim of this type of crime) Theft and Handling stolen goods 40 10 Domestic homicide 90 3 Incest 90 21 Robbery and car theft 70 33 Hard drugs trafficking 60 14 Domestic violence 80 16 Table 2 Descriptive Measures of Independents Variables (Level 2) Characteristic % of respondents answering in the affirmative (N = 179a) Involvement in the criminal justice system     Were a witness in a case 12.1     Involved in a case as the accused or relative of the accused 11.7     Involved in a case as a victim or relative of the victim 14.7     Involved in a case where the sentence was considered to be too lenient 14.2 Indirect knowledge of the criminal justice system     Knows a lawyer 55.4     Knows a judge 13.1     Knows someone who works with criminals 39.2     Knows someone who works with victims 27.9     Knows someone who has been in prison 38.4 Personal characteristics     <30 17.3     >65 11.7     Male 48.0     Separated 19.6     Married or Living with 49.0     Children 55.7     Diploma of Collegiate Studiesb 21.6     University 43.3 Characteristic % of respondents answering in the affirmative (N = 179a) Involvement in the criminal justice system     Were a witness in a case 12.1     Involved in a case as the accused or relative of the accused 11.7     Involved in a case as a victim or relative of the victim 14.7     Involved in a case where the sentence was considered to be too lenient 14.2 Indirect knowledge of the criminal justice system     Knows a lawyer 55.4     Knows a judge 13.1     Knows someone who works with criminals 39.2     Knows someone who works with victims 27.9     Knows someone who has been in prison 38.4 Personal characteristics     <30 17.3     >65 11.7     Male 48.0     Separated 19.6     Married or Living with 49.0     Children 55.7     Diploma of Collegiate Studiesb 21.6     University 43.3 Note.aA seriousness scale ranging from 1 to 100 was added during data collection to compensate for the fact that another measure of seriousness (amplitude scale) seemed problematic for respondents. The first 19 respondents did not assess the severity of the different scenarios (n = 114) and therefore had to be excluded from the analyses. Twelve other cases were eliminated because they had one or more missing values on the variables under study. bThe school system in Quebec provides a college education (2 or 3 years) before university-level education. Table 2 Descriptive Measures of Independents Variables (Level 2) Characteristic % of respondents answering in the affirmative (N = 179a) Involvement in the criminal justice system     Were a witness in a case 12.1     Involved in a case as the accused or relative of the accused 11.7     Involved in a case as a victim or relative of the victim 14.7     Involved in a case where the sentence was considered to be too lenient 14.2 Indirect knowledge of the criminal justice system     Knows a lawyer 55.4     Knows a judge 13.1     Knows someone who works with criminals 39.2     Knows someone who works with victims 27.9     Knows someone who has been in prison 38.4 Personal characteristics     <30 17.3     >65 11.7     Male 48.0     Separated 19.6     Married or Living with 49.0     Children 55.7     Diploma of Collegiate Studiesb 21.6     University 43.3 Characteristic % of respondents answering in the affirmative (N = 179a) Involvement in the criminal justice system     Were a witness in a case 12.1     Involved in a case as the accused or relative of the accused 11.7     Involved in a case as a victim or relative of the victim 14.7     Involved in a case where the sentence was considered to be too lenient 14.2 Indirect knowledge of the criminal justice system     Knows a lawyer 55.4     Knows a judge 13.1     Knows someone who works with criminals 39.2     Knows someone who works with victims 27.9     Knows someone who has been in prison 38.4 Personal characteristics     <30 17.3     >65 11.7     Male 48.0     Separated 19.6     Married or Living with 49.0     Children 55.7     Diploma of Collegiate Studiesb 21.6     University 43.3 Note.aA seriousness scale ranging from 1 to 100 was added during data collection to compensate for the fact that another measure of seriousness (amplitude scale) seemed problematic for respondents. The first 19 respondents did not assess the severity of the different scenarios (n = 114) and therefore had to be excluded from the analyses. Twelve other cases were eliminated because they had one or more missing values on the variables under study. bThe school system in Quebec provides a college education (2 or 3 years) before university-level education. Real or perceived knowledge of the sentence A dichotomous variable enabled a distinction to be made between the three cases of sentencing where citizens had to determine the likely sentence of the courts and the three cases for parole where they were asked to comment on the quality of the sentence chosen by the judge. This variable is used to test the hypothesis often cited in the literature that people would be more satisfied if they had a better knowledge of the sentence handed down by judges. Crime seriousness The literature on public opinion suggests that citizens are more dissatisfied with sentences for serious crimes (Frost, 2010). We introduced a variable in order to test whether citizens’ levels of satisfaction is related to their assessment of the offense’s seriousness. The seriousness of the criminal scenarios was evaluated on a scale ranging from 1 to 100. The results in Table 1 show that citizens consider cases of incest and conjugal homicide as similar in severity (a median of 90)9 and more serious than other offenses. Crime experience Although several studies found no direct link between victimization and punitiveness (Tufts, 2000; Hough & Roberts, 1999; Brillon, 1988; Langworthy & Whitehead, 1986), we know little about the impact of victimization on opinions about court sentencing (with the exception of the work of Sprott & Doob, 1997 who discovered complex relationships between victimization and citizen opinions). Furthermore, it is possible that the lack of a correlation between victimization and punitiveness observed in the literature results from the use of a general measure of victimization. This variable tested whether experience of a similar crime,10 either personally or through a close family member, influenced the opinion someone had of sentencing by the courts in these particular cases. Table 1 shows that the rate of victimization varies depending on the crime (3–33%). Involvement in the court system Several researchers suggested that citizen’s opinions about the courts are problematic because they are based only on information obtained from the media—known to be a highly biased source (Jewkes, 2004). We wanted to examine whether having participated in or attended a criminal case would influence public opinion on sentencing by the courts. As shown in Table 2, participation in the court system is measured by having attended a criminal case as a witness (12.1%), as a defendant or a friend of the accused (11.7%), or as a victim or a friend of a victim (14.7%). These categories are not mutually exclusive (8% attended a trial in two different “statuses”), and we found that 27.6% of the respondents had attended a criminal trial in one or more of these three statuses. Among all respondents, 14.2% believed that the sentence in one or more of the cases they attended was too lenient. Indirect knowledge of the criminal justice system Indirect knowledge of the criminal justice system was measured on the basis of knowing someone who works directly in the criminal justice system (Table 2). A number of respondents knew a judge (13.1%), a lawyer (55.4%), an individual who works with victims (27.9%), an offender (39.2%), or a person who has directly experienced the criminal justice system by being in prison (38.4%). Most citizens have an indirect knowledge of the courts and several may have been exposed to different views because 21.1% of them knew two different types of participants, while 30% knew at least three. Personal characteristics Individual characteristics are also included to control for the fact that experiences with the court might not be randomly distributed (Van de Walle, 2009). We also took accounted for age, gender, and education level of respondents; variables often found in studies on opinion and attitudes of citizens. We also included marital status and whether respondents had children because these characteristics could potentially influence their opinion about certain crimes (e.g., domestic violence or incest). Analytical Framework The nested data structure (six cases evaluated per individual) requires the use of hierarchical regression analysis11 (or multilevel analysis, see Hox, 2010; Hox & Kreft, 1994; Snijders & Bosker, 1999). The opinions of an individual on court sentences for the six crime scenarios are not independent and they may be more similar than they would be if they were the responses of two different respondents. Hierarchical regression analysis has the advantage of avoiding the violation of the assumption of independence of the error terms (residuals) required by conventional regression analysis.12 The results of this analysis13 can be assessed in the same ways as those from an ordinary logistic regression. Results Measuring Public Opinion Table 3 presents different measures of public opinion on sentencing by the courts. Findings confirm the hypothesis based on the use of two different processes. Indeed, the percentage of respondents dissatisfied with the sentences of the courts is much higher when the question is asked in a general way without reference to a specific criminal case story. Information in a specific case will activate System 2 and ameliorate satisfaction regarding the work of the court. Table 3 Comparison of Different Measures of Public Opinion on Sentencing by the Courts Model 1 General opinion Model 2 Decision taking Model 3 Evaluation of the decision taken by a court In general Theft and handling stolen goods Domestic homicide Incest Robbery and car theft Hard drugs trafficking Domestic violence Court sentencea (2 years) (10 years) (3 years) 2 years–1 day 2 years 18 months Court too lenient 63% 25% 26% 53% 20% 31% 56% Court about right 36% 61% 61% 42% 74% 67% 42% Court too severe 2% 14% 13% 5% 7% 3% 3% Model 1 General opinion Model 2 Decision taking Model 3 Evaluation of the decision taken by a court In general Theft and handling stolen goods Domestic homicide Incest Robbery and car theft Hard drugs trafficking Domestic violence Court sentencea (2 years) (10 years) (3 years) 2 years–1 day 2 years 18 months Court too lenient 63% 25% 26% 53% 20% 31% 56% Court about right 36% 61% 61% 42% 74% 67% 42% Court too severe 2% 14% 13% 5% 7% 3% 3% Note. aSentence estimated by the public. Adapted from "Explorer et comprendre l'insatisfaction du public face à la clémence des tribunaux" by Leclerc (2010). Adapted with permission. Table 3 Comparison of Different Measures of Public Opinion on Sentencing by the Courts Model 1 General opinion Model 2 Decision taking Model 3 Evaluation of the decision taken by a court In general Theft and handling stolen goods Domestic homicide Incest Robbery and car theft Hard drugs trafficking Domestic violence Court sentencea (2 years) (10 years) (3 years) 2 years–1 day 2 years 18 months Court too lenient 63% 25% 26% 53% 20% 31% 56% Court about right 36% 61% 61% 42% 74% 67% 42% Court too severe 2% 14% 13% 5% 7% 3% 3% Model 1 General opinion Model 2 Decision taking Model 3 Evaluation of the decision taken by a court In general Theft and handling stolen goods Domestic homicide Incest Robbery and car theft Hard drugs trafficking Domestic violence Court sentencea (2 years) (10 years) (3 years) 2 years–1 day 2 years 18 months Court too lenient 63% 25% 26% 53% 20% 31% 56% Court about right 36% 61% 61% 42% 74% 67% 42% Court too severe 2% 14% 13% 5% 7% 3% 3% Note. aSentence estimated by the public. Adapted from "Explorer et comprendre l'insatisfaction du public face à la clémence des tribunaux" by Leclerc (2010). Adapted with permission. The percentage of respondents dissatisfied with the leniency of the courts drops from 63% (first column) to 20–30% in four of the six scenarios.14 In addition, when people are asked to make a decision (Model 2), they are more likely to think that the sentences of the courts were too severe and, therefore, longer than the sentence they deem acceptable in this specific case: the percentage increases from generally 2% to 7% to almost 15% (second and third column). When citizens are asked to put themselves in the place of the judge and make a decision, they seem much more aware of the severity of the decision than when asked to comment on the fairness of the sentences imposed by the courts (both in the abstract—Model 1—and in the concrete—Model 3). This result is similar to that of Warner and Davis (2011). Explaining Public Opinion: the Impact of Experiences and Contacts With the Court System Table 4 presents the results of different logistic regressions (multilevel analysis) in explaining public opinions about the courts’ leniency in each of the six cases. Parameters presented are odds ratios. The first model includes variables related to the criminal cases and the experiences of citizens. The second model adds the characteristics of the individuals. The last model only retains the variables that are significant or relevant from a theoretical point of view. Table 4 Multilevel Logistic Regression on the Dissatisfaction of Citizens With the Leniency of the Courts in Six Criminal Cases (N = 1,074 Cases, 179 Individuals) Model 0 Model 1 Model 2 Model 3 Model 4 Intraclass correlation coefficient (ICC)a 0.194*** 0.180*** 0.149*** 0.135*** 0.156*** Odds ratio Crime seriousness (1–100) 1.02*** 1.02*** 1.02*** 1.02*** Sentence of the judge known 1.02 1.03 1.03 1.03 Victim (direct or indirect) of the crime being assessed 2.05** 2.18*** 2.14*** 2.23*** Participated in a case as a witness 1.10 1.14 Involved in a case as the accused or relative of the accused 0.75 0.71 Involved in a case as the victim or relative of the victim 1.63 1.69 Involved in a case where the sentence was seen as being too lenient 1.69* 1.72* 1.78* Knows a lawyer 0.82 0.85 Knows a judge 1.54 1.70† 1.54 Knows someone who works with criminals 1.15 1.16 Knows someone who works with victims 0.60* 0.61* 0.63* Knows someone who has been to prison 0.66* 0.64* 0.67* <30 1.03 >65 0.73 Male 0.89 Separated 0.68 Married or living with 0.70 Children 1.44 Diploma of collegiate studies 1.50 University 1.03 Model 0 Model 1 Model 2 Model 3 Model 4 Intraclass correlation coefficient (ICC)a 0.194*** 0.180*** 0.149*** 0.135*** 0.156*** Odds ratio Crime seriousness (1–100) 1.02*** 1.02*** 1.02*** 1.02*** Sentence of the judge known 1.02 1.03 1.03 1.03 Victim (direct or indirect) of the crime being assessed 2.05** 2.18*** 2.14*** 2.23*** Participated in a case as a witness 1.10 1.14 Involved in a case as the accused or relative of the accused 0.75 0.71 Involved in a case as the victim or relative of the victim 1.63 1.69 Involved in a case where the sentence was seen as being too lenient 1.69* 1.72* 1.78* Knows a lawyer 0.82 0.85 Knows a judge 1.54 1.70† 1.54 Knows someone who works with criminals 1.15 1.16 Knows someone who works with victims 0.60* 0.61* 0.63* Knows someone who has been to prison 0.66* 0.64* 0.67* <30 1.03 >65 0.73 Male 0.89 Separated 0.68 Married or living with 0.70 Children 1.44 Diploma of collegiate studies 1.50 University 1.03 Note. †p ≤ 0.10; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001. aICC (logistic reg) = (Level2 Var)/(Level2 var + (π2/3)) as suggested by Snijders et al. (2012). Table 4 Multilevel Logistic Regression on the Dissatisfaction of Citizens With the Leniency of the Courts in Six Criminal Cases (N = 1,074 Cases, 179 Individuals) Model 0 Model 1 Model 2 Model 3 Model 4 Intraclass correlation coefficient (ICC)a 0.194*** 0.180*** 0.149*** 0.135*** 0.156*** Odds ratio Crime seriousness (1–100) 1.02*** 1.02*** 1.02*** 1.02*** Sentence of the judge known 1.02 1.03 1.03 1.03 Victim (direct or indirect) of the crime being assessed 2.05** 2.18*** 2.14*** 2.23*** Participated in a case as a witness 1.10 1.14 Involved in a case as the accused or relative of the accused 0.75 0.71 Involved in a case as the victim or relative of the victim 1.63 1.69 Involved in a case where the sentence was seen as being too lenient 1.69* 1.72* 1.78* Knows a lawyer 0.82 0.85 Knows a judge 1.54 1.70† 1.54 Knows someone who works with criminals 1.15 1.16 Knows someone who works with victims 0.60* 0.61* 0.63* Knows someone who has been to prison 0.66* 0.64* 0.67* <30 1.03 >65 0.73 Male 0.89 Separated 0.68 Married or living with 0.70 Children 1.44 Diploma of collegiate studies 1.50 University 1.03 Model 0 Model 1 Model 2 Model 3 Model 4 Intraclass correlation coefficient (ICC)a 0.194*** 0.180*** 0.149*** 0.135*** 0.156*** Odds ratio Crime seriousness (1–100) 1.02*** 1.02*** 1.02*** 1.02*** Sentence of the judge known 1.02 1.03 1.03 1.03 Victim (direct or indirect) of the crime being assessed 2.05** 2.18*** 2.14*** 2.23*** Participated in a case as a witness 1.10 1.14 Involved in a case as the accused or relative of the accused 0.75 0.71 Involved in a case as the victim or relative of the victim 1.63 1.69 Involved in a case where the sentence was seen as being too lenient 1.69* 1.72* 1.78* Knows a lawyer 0.82 0.85 Knows a judge 1.54 1.70† 1.54 Knows someone who works with criminals 1.15 1.16 Knows someone who works with victims 0.60* 0.61* 0.63* Knows someone who has been to prison 0.66* 0.64* 0.67* <30 1.03 >65 0.73 Male 0.89 Separated 0.68 Married or living with 0.70 Children 1.44 Diploma of collegiate studies 1.50 University 1.03 Note. †p ≤ 0.10; *p ≤ 0.05; **p ≤ 0.01; ***p ≤ 0.001. aICC (logistic reg) = (Level2 Var)/(Level2 var + (π2/3)) as suggested by Snijders et al. (2012). Crime seriousness The analysis shows that, as expected, the more serious someone judges a crime to be, the more likely they are to think that the individual’s sentence will be too lenient. This suggests that people are aware that their assessment of the seriousness of a crime can differ (e.g., they know they “overestimate” the crime’s seriousness) and that this can partially explain why they believe the sentences will not be sufficiently severe (they know that the judge chose the right sentence resorting on proportionality between the punishment and an assessment of the seriousness of the crime). However, it is important to note that not all violent and serious crimes lead to lenient opinions. As shown in Table 3, dissatisfaction regarding homicide or robbery is pretty low (respectively 26% and 20%). Knowledge of the sentence Research has shown a link between more general dissatisfaction and a lack of public knowledge about the work of the courts (Gelb, 2006; Roberts & Stalans, 1997; Hough & Roberts, 1998; Mattinson & Mirrlees-Black, 2000; Chapman et al., 2002). However, our results do not enable us to observe this relationship when the opinion is more specific and placed into context. Indeed, the variable that differentiates cases where respondents estimated the sentence with those in which they gave their opinion on the judge’s sentence is not significant. This result is contrary to what we expected but is in line with those obtained by St-Amand and Zamble (2001), who concluded that informing the public about the likely sentence of the courts in a detailed question does not significantly influence their acceptance of its severity. Our results suggest that informing the public about the real work of the courts (the solution usually proposed, see for example, the initiative of the British Home Office in the 2000s) would not necessarily have an impact on more specific views of the courts. This can be explained in part by the fact that, once freed from their traditional stereotypes (as expressed in the more general questions), respondents have a good knowledge of the court’s work (as shown in the findings in Table 3) and an opinion that is already more nuanced. Crime experience Our analyses of crime experiences also reveal that the odds of being dissatisfied with sentencing are twice as high if the respondent had been a direct or a vicarious victim of a crime. While this relationship may be expected, it contrasts with empirical literature that suggests that victims are not harsher or more negative in their view of justice or the judicial system (see, for example, Unnever, Cullen, & Fisher, 2007). The difference can probably be explained by our measure of victimization, which takes into account the type of crime assessed and does not restrict victimization experiences to the last year or to personal victimization (as it includes victimization of close family). Additional analyses concluded that a more general measure of victimization, which took into account victimization experiences for 15 crimes, has no effect on these opinions. Involvement in the court system Some authors have suggested that people’s dissatisfaction stems from a lack of knowledge combined, among other things, with an absence of involvement in the criminal justice system (Gelb, 2006; Mattinson & Mirrlees-Black, 2000; Chapman et al., 2002). Although this lack of direct contact can potentially explain the more general opinion of the courts found in some studies (see Van de Walle, 2009), it does not seem to be linked to opinions on more specific cases. Indeed, having been involved in a criminal case as a witness, a victim (or relative of a victim), or a defendant (or relative of a defendant) did not influence respondents’ opinions about the work of the courts in the six cases presented. These experiences have a significant impact only in cases where people considered the sentence imposed by the judge to be too lenient. Table 4 shows that people who had been involved in a case where they considered the sentence to be too lenient were one and a half times more likely to be dissatisfied with the severity of the courts. Indirect knowledge of the criminal justice system While knowing a lawyer or someone who works with offenders has no effect on the opinions of citizens,15 three types of contacts can have an influence. First, a person who knows someone who has been in prison is almost one and half times more likely to be satisfied with the work of the courts. This effect may be explained by the fact that people who have been in prison can break down some stereotypes about the judicial process, such as conditions of prison life and the effect of punishment on their criminal life. Knowing someone who works with victims has the same effect of increasing citizen satisfaction (odd ratio: 1.6). This result may, a priori, seem surprising, as victims often report dissatisfaction with their judicial experience. However, past researches showed that victims are rarely dissatisfied with the sentence and that their dissatisfaction most often concerns the place or the voice they did not have in the judicial process (Wemmers, 1996). Finally, people who know a judge are more likely to be dissatisfied with sentencing, but the effect is marginally significant and only emerges when the education level of citizens is considered (there is some covariance between these variables as people who know a judge are twice as likely to have a university degree—69% vs. 39%). Like the press, judges probably tend to talk about the more unusual or frustrating cases they presided over. In their personal relationships, judges have no obligation to inform and demystify the criminal justice system and, like other citizens, they may tend to focus on the more negative or darker sides of their work and of the criminal justice system. Given its marginal significance, this result should be interpreted with the usual caution. Personal characteristics Finally, Table 4 also shows that, contrary to our hypothesis, the characteristics of respondents are not related to their views on the work of the courts.16 However, this result is not surprising when one takes into account the variability of opinions among individuals. As shown in Table 5, because a respondent is dissatisfied with the work (actual or anticipated) of the courts in a particular case does not mean that that he or she will necessarily be dissatisfied with their work in another case. Most respondents (around 60%) reported being dissatisfied with one to three of the six scenarios, showing that their attitudes are not fixed and need to be explored in multiple contexts. This conclusion is important because it suggests that it may be unproductive to seek personal explanations (age, sex, or socioeconomic status) for public dissatisfaction because this dissatisfaction is nuanced and therefore subject to change depending on the context. Table 5 Distribution of Respondents by Number of Cases They Consider the Sentence Being too Lenient Number of cases % of respondents 0 17.4 1 20.5 2 22.6 3 19.5 4 13.8 5 4.6 6 1.5 Number of cases % of respondents 0 17.4 1 20.5 2 22.6 3 19.5 4 13.8 5 4.6 6 1.5 Table 5 Distribution of Respondents by Number of Cases They Consider the Sentence Being too Lenient Number of cases % of respondents 0 17.4 1 20.5 2 22.6 3 19.5 4 13.8 5 4.6 6 1.5 Number of cases % of respondents 0 17.4 1 20.5 2 22.6 3 19.5 4 13.8 5 4.6 6 1.5 Conclusion Some authors have emphasized the differences between general and specific attitudes of citizens, but they always rely on two different concepts (opinion about court decisions vs. sentencing expectations) to arrive at these conclusions (Zamble & Kalm, 1990, or Applegate et al., 1996). The analyses presented in this article differ in that they are based on alternative measures of the opinions of citizens. Following Warner and Davis (2011), the results confirm that there is a significant difference between the general opinion that citizens express on an abstract question about the work of the courts and the specific opinion that they give when asked to look at the work of the courts in a specific case. While the more general opinion seems to be an emotional response that can reveal perceived injustices and is often based on stereotypes about offenders or the criminal justice system (an automatic response from System 1 while processing information), specific views are more nuanced and result from a more analytic reasoning (System 2). This article suggests that dissatisfaction is not real because it disappears for most citizens when they are put in a decision-making situation (thus, activating System 2) where the use of stereotypes is limited by the provision of information about the details of the criminal case. It is therefore wrong to conclude that people want tougher penalties for all violent or serious crimes—citizens’ views are much more complex and cannot be easily generalized. Experiences with crime or the criminal justice system may also explain people’s opinions about sentencing. The variable with the most impact in this study was having been a victim of a crime (directly or indirectly through close family). As with the factor of having been involved in a case where the sentence handed down was considered too lenient, such negative experiences tarnish the respondent’s view of the court decisions. Only two variables lead respondents to more positive opinions: knowing someone who has been in prison and someone who works with victims. It is possible that direct experience with the court system (such as having attended a trial) does not change the image people have of courts because they do not believe that the case in question was representative of the daily activities of the courts (an argument noted by Warner and Davis, 2011) or because the experience is marginal when compared with the amount of information and experiences generated from the media (Dorfman & Schiraldi, 2001). Finally, studies that examine the degree of punitiveness of individuals should consider how this notion is conceptualized and integrate more flexible measures that vary depending on the context. If the results of previous studies are so contradictory and individual characteristics have such a low explanatory power (Frost, 2010), it is because the distinction between punitive individuals and nonpunitive (or satisfied vs. dissatisfied) individuals is not as simple as we hoped. Some individuals are potentially more punitive in some contexts, while remaining more lenient than their peers in other situations. Of the three cases presented, only 5% of respondents were consistently in the top quartile of the toughest respondents and 4% were consistently in the quartile of the less severe respondents. Measuring the punitiveness of an individual according to response to abstract questions or by adding up responses to several specific situations remain strategies that underestimate the adaptability of individuals. They fail to bring out the rational, analytical, and intelligent aspects of public perceptions and opinions. Funding The research had been funded by the Social Sciences and Humanities Research Council. Chloé Leclerc is an Assistant Professor of Criminology at Université de Montréal and a researcher at the International Centre for Comparative Criminology. Her research focuses on the penal and judicial system. She had worked on criminal justice professionals and citizens’ conceptualization of “just sentencing.” Anta Niang is a PhD candidate of the Department of Psychology in Université Rennes 2 at the Laboratoire de Psychologie: Cognition, Comportement et Communication (LP3C, EA 1285). Her research interests focus on social judgments about crime, criminals, and sentencing. She currently works on the impact of psychological and psychiatric expertise on jurors’ decision making. Marie-Chloé Duval has a Master’s Degree from the School of Criminology, at Université de Montréal. She had worked on media representation of criminal trials. Footnotes 1De grandpré, H. (2011). Les conservateurs prêts pour une loi omnibus. La Presse, 19 September. 2This result is contrary to the findings of De Keijser, Van Koppen, and Elffers (2007) and of Tremblay, Cordeau, and Ouimet (1994) who found that, as the public underestimates the severity of the courts, the gap between the two groups’ understanding of how severe sentences should be is even more important than one might have anticipated. 3A “general” opinion captures the overall views that people possess about a policy issue. “Specific” opinions are the views they express when the policy is applied to a case that has a certain set of attributes. These “specific” opinion/attitudes are especially relevant to criminal justice policy because decisions are made about cases that involve offenders, victims, and acts, which may vary on many dimensions and interact in unique ways (Cullen et al., 2000). 4The percentage varies depending on the type of crime. Just >50% of citizens considered sentences for drug offenses and crimes against property to be too lenient. This percentage increases to >75% for violent crimes or crimes of a sexual nature. 5The sample was randomly chosen from a list of voluntary participants drawn up using a telephone survey of the general public (randomly selected). The sample was stratified by gender and age to reflect the demographic importance of Francophone citizens living in the Montreal area (the survey was only conducted in French). The final sample reflects most of the characteristics of the population of Montreal, albeit with a slightly higher proportion of common-law or married people (49% as compared with 40% in the total population) and people with either a college education (21.6% compared with 15% in the total population) or a university degree (43.3% compared with 31.8%). Since socio-demographic variables are known to have a low impact on people’s opinions and attitudes (Roberts & Indermaur, 2007: 58) and given that the purpose of the article is less to provide a representative view of the population than to propose an alternative way of measuring public opinion and to determine the effect that some experience with the justice system can have on public opinion, these differences do not undermine the main conclusions of the article. 6Since sentences of different nature are not easy to compare, and since all respondent provided prison term without mentioning a problem, we limited our analyses to prison terms in this article. 7Respondent were told, that the individual was presenting his request on his first admissibility to parole, namely, after serving a third of his prison term. 8Multicollinearity was tested among all independent variables (the highest Spearman’s rho was <0.5). 9Although it may seem surprising that many citizens have evaluated the crime of incest as severely as the crime of homicide, it must be emphasized that the attitude of the accused can greatly influence the severity perceived by citizens. In the homicide case, the man regretted his actions and called the emergency services, whereas the father in the incest case did not seem to be remorseful and maintained that his daughter had consented. 10For the first offense, we checked whether the person had been a victim of a crime involving stolen goods, but we did not include victims of theft because in the scenario it is a company that was robbed. For the second offense, we considered victims of murder or attempted murder. For the third offense, we considered victims of sexual assault (11%) and sexual touching of a child (15%). For the fourth offense, we considered everyone who had been the victim of a robbery (15.5%) or a car theft (22.5%) since the offender was convicted of both offenses. 11Taylor (2006) and Wallander (2009) discussed the debate regarding the necessity of using hierarchical regression while working with vignettes nested in respondents. With our data, models from ordinary logistic regression lead to the exact same results as were obtained with hierarchical logistic regression. 12When no account is taken of the dependency of the error terms then false degrees of freedom are used (based on the number of cases rather than the number of individuals) and this favors type II errors, i.e. the erroneous discovery of significant relationships. 13The analysis was performed using Mplus 7.4 software, adopting the penalized quasi-likelihood (PQL) as the estimation procedure. 14Chi-square tests reveal that the percentage of dissatisfaction in these four scenarios is statistically different from the percentage of dissatisfaction measure for the general question (p < .001). 15The absence of relation between knowing a lawyer and citizen’s opinion could possibly be explained by the absence of specification in the survey of the type of criminal lawyer. As people may answer that they know a lawyer referring to a civil or corporate lawyer, it is possible that this variable does not adequately capture the knowledge of a criminal lawyer. 16As some characteristics could have different effect depending on the type of crime, we test some interaction effects (e.g., being a woman or a parent and judging a case of incest; being a woman and judging domestic violence/homicide), but we found no significant effect. Acknowledgements Authors wish to thanks Carlo Morselli for his useful observations and suggestions and Joan McGilvray for her attentive revision of this article. References Applegate B. K. , Cullen F. T. , Turner M. G. , Sundt J. L. ( 1996 ). Assessing public support for three-strikes-and-you’re-out laws: Global versus specific attitudes . Crime and Delinquency , 42 , 517 – 34 . doi: 10.1177/0011128796042004002 Google Scholar CrossRef Search ADS Benesh S. C. , Howell S. E. ( 2001 ). Confidence in the courts: A comparison of users and non users . Behavioral Sciences and the Law , 19 , 199 – 214 . doi: 10.1002/bsl.437 Google Scholar CrossRef Search ADS PubMed Brillon Y. ( 1988 ). Punitiveness, status and ideology in three Canadian Provinces. In Walker N. , Hough M. (Eds.), Public attitudes to sentencing: Survey from five countries . Hampshire : Gower . doi: 10.1016/0047-2352(89)90038-x Casey S. , Mohr P. ( 2005 ). Law-and-order politics, public-opinion polls and the media (Australia) . Psychiatry, Psychology and Law , 12 ( 1 ), 141 – 151 . doi: 10.1375/pplt.2005.12.1.141 Google Scholar CrossRef Search ADS Chapman B. , Mirrlees-Blaws C. , Brawn C. ( 2002 ). Improving public attitudes to the Criminal Justice System: The impact of information . London : Home Office Research Studies . doi: 10.1037/e454602008-001 Cullen F. T. , Fisher B. S. , Applegate B. K. ( 2000 ). Public opinion about punishment and corrections. In Tonry M. (Ed.), Crime and Justice: A review of research (pp. 1 – 79 ). Chicago : University of Chicago Press . doi: 10.1086/652198 De Keijser J. W. , Van Koppen P. J. , Elffers H. ( 2007 ). Bridging the gap between judges and the public? A multi-method study . Journal of Experimental Criminology , 3 , 131 – 161 . doi: 10.1007/s11292-007-9031-3 Google Scholar CrossRef Search ADS Doob A. N. , Roberts J. V. ( 1983 ). Sentencing: An analysis of the public's view of sentencing . Ottawa : Department of Justice Canada . Doob A. N. , Roberts J. V. ( 1988 ). Public punitiveness and public knowledge of the facts: Some Canadian surveys. In Walker N. , Hough M. (Eds.), Public attitudes to sentencing: Surveys from five countries . Aldershot : Gower . doi: 10.1016/0047-2352(89)90038-x Durham A. M. , Elrod H. P. , Kinkade P. T. ( 1996 ). Public support for the death penalty: Beyond Gallup . Justice Quarterly , 13 , 705 – 736 . doi: 10.1080/07418829600093161 Google Scholar CrossRef Search ADS Dorfman L. , Schiraldi V. ( 2001 ). Off balance: Media coverage of youth crime . Guild Practitioner , 58 , 75 – 78 . Frost N. A. ( 2010 ). Beyond public opinion polls: Punitive public sentiment and criminal justice policy . Sociology Compass , 4 , 156 – 168 . doi: 10.1111/j.1751-9020.2009.00269.x Google Scholar CrossRef Search ADS Garland D. ( 2001 ). The culture of control, crime and social order in contemporary society . Oxford : Oxford University Press . Gelb K. ( 2006 ). Myths and misconceptions: Public opinion versus public judgment about sentencing . Melbourne : Sentencing Advisory Council . Gray J. M. ( 2009 ). What shapes public opinion of the criminal justice system. In Wood J. , Gannon T. (Eds.), Public opinion and criminal justice (pp. 49 – 72 ). Portland : Willan Publishing . doi: 10.4324/9781843926610 Gwin J. S. ( 2010 ). Juror sentiment on just punishment: Do the federal sentencing guidelines reflect community values? Harvard Law and Policy Review , 4 , 173 – 200 . Hough M. , Roberts J. V. ( 1998 ). Attitudes to punishment: Findings from the British crime survey . London : Home Office Research Study . Hough M. , Roberts J. V. ( 1999 ). Sentencing trends in Britain: Public knowledge and public opinion . Punishment and Society , 1 ( 1 ), 11 – 26 . doi: 10.1177/14624749922227685 Google Scholar CrossRef Search ADS Hox J. J. , Kreft I.G.G. ( 1994 ) Multilevel analysis methods . Sociological Methods and Research , 22 , 283 – 299 . doi: 10.1177/0049124194022003001 Google Scholar CrossRef Search ADS Hox J. J. ( 2010 ). Multilevel analysis, techniques and application ( 2nd ed.). New York, NY : Routledge . Hutton N. ( 2005 ). Beyond populist punitiveness? Punishment and Society , 7 , 243 – 258 . doi: 10.1177/1462474505053828 Google Scholar CrossRef Search ADS Indermaur D. ( 1987 ). Public perception of sentencing in Perth, Western Australia . Australian and New Zealand Journal of Criminology , 20 , 163 – 183 . doi: 10.1177/000486588702000304 Google Scholar CrossRef Search ADS Indermaur D. , Hough M. ( 2002 ). Strategies for changing public attitudes to punishment. In Roberts J. , Hough M. (Eds.), Changing attitudes to punishment: Public opinion, crime and justice (198-214) . Cullompton, Devon : Willan Publishing . doi: 10.4324/9781843924272 Jewkes Y. ( 2004 ). Media and Crime: Key approaches to criminology . Sage : London . Kahneman D. ( 2011 ). Thinking, fast and slow . Macmillan : London . Langworthy R. H. , Whitehead J. T. ( 1986 ). Liberalism and fear as explanations of punitiveness . Criminology , 24 , 575 – 91 . doi: 10.1111/j.1745-9125.1986.tb00391.x Google Scholar CrossRef Search ADS Leclerc C. ( 2010 ). La métrique de la juste peine: Une analyse des décisions de justice prises par les acteurs judiciaires et le public . Montreal : Doctoral dissertation . Mattinson J. , Mirrlees-Black C. ( 2000 ). Attitudes to crime and criminal justice: Findings from the 1998 British Crime Survey . Home Office Research Study No. 200. London : Research, Development and Statistics Directorate, Home Office . doi: 10.1037/e454372008-001 Mirrlees-Black C. ( 2002 ). Improving public knowledge about crime and punishment. In Roberts J. V. , Hough M. (Eds.), Changing attitudes to punishment: Public opinion, crime and justice (pp. 184 – 197 ). Portland : Willan Publishing . doi: 10.4324/9781843924272 Roberts J. V. ( 1988 ). Public opinion and sentencing: The surveys of the Canadian sentencing commission . Ottawa, ON : Department of Justice Canada . Roberts J. V. , Doob A. N. ( 1989 ). Sentencing and public opinion: Taking false shadows for true substances . Osgood Hall Law Journal , 27 , 491 – 515 . Roberts J. V. , Stalans L. J. ( 1997 ). Public opinion, crime, and criminal justice (pp. 35 – 52 ). Boulder, CO : Westview Press . doi: 10.1086/449205 Roberts L. D. , Indermaur D. ( 2007 ). Predicting punitive attitudes in Australia . Psychiatry, Psychology and Law , 14 ( 1 ), 56 – 65 . doi: 10.1375/pplt.14.1.56 Google Scholar CrossRef Search ADS Roberts J. V , Stalans L. , Indermaur D. , Hough M. ( 2003 ). Penal populism and public opinion: Lessons from five countries . Oxford : Oxford University Press . Rossi P. H. , Berk R. A. , Campbell A. ( 1997 ). Just punishments: Guideline sentences and normative consensus . Journal of Quantitative Criminology , 13 , 267 – 290 . doi: 10.1007/bf02221093 Google Scholar CrossRef Search ADS Rottman D. B. , Hansen R. , Mott N. , Grimes L. ( 2003 ). Perceptions of the courts in your community: The influence of experience, race and ethnicity . Williamsburg, VA : National Center for State Courts . Snijders T. A. B. , Bosker R. J. ( 1999 ). Multilevel analysis. An introduction to basic and advanced multilevel modeling . London : Sage Publishers . Snijders T. A. B. , Bosker R. J. ( 2012 ). Multilevel Analysis: An Introduction to Basic and Advanced Multilevel Modeling, second edition . Sage Publishers : London . Sprott J. B. ( 1996 ). Understanding public views of youth crime and the youth justice system . Canadian Journal of Criminology , 38 , 271 – 290 . Sprott J. B. , Doob A. N. ( 1997 ). Fear, victimization, and attitudes to sentencing, the courts and the police . Canadian Journal of Criminology , 39 , 275 – 291 . Sprott J. B. ( 1998 ). Understanding public opposition to a separate youth justice system, Crime and Delinquency , 44 , 399 – 411 . doi: 10.1177/0011128798044003004 Google Scholar CrossRef Search ADS Sprott J. B. , Webster C. M. , Doob A. N. ( 2013 ). Punishment severity and confidence in the criminal justice system . Canadian Journal of Criminology and Criminal Justice , 55 , 279 – 292 . doi: 10.3138/cjccj.2012.e.27 Google Scholar CrossRef Search ADS Stalans L. J. ( 1993 ). Citizens' crime stereotypes, biased recall, and punishment preferences in abstract cases: The educative role of interpersonal sources . Law and Human Behavior , 17 , 451 . doi: 10.1007/bf01044378 Google Scholar CrossRef Search ADS Stalans L. J. , Diamond S. S. ( 1990 ). Formation and change in lay evaluations of criminal sentencing . Law and Human Behavior , 14 , 199 – 214 . doi: 10.1007/bf01352749 Google Scholar CrossRef Search ADS St-Amand M. , Zamble E. ( 2001 ). Impact of information about sentencing decisions on public attitudes toward the criminal justice system . Law and Human Behavior , 25 , 515 – 528 . doi: 10.1023/a:1012844932754 Google Scholar CrossRef Search ADS PubMed Taylor B. J. ( 2006 ). Factorial surveys: using vignettes to study professional judgement . British Journal of Social Work , 36 , 1187 – 1207 . Google Scholar CrossRef Search ADS Tremblay P. , Cordeau G. , Ouimet M. ( 1994 ). Underpunishing offenders: Toward a theory of legal tolerance . Revue canadienne de criminology , 36 ( 4 ), 407 – 434 . Tufts J. ( 2000 ). Understanding Public Attitudes Toward Sentencing (Master thesis, Department of Criminology, Université of Ottawa, Ottawa). Unnever J. D. , Cullen F. T. , Fisher B. S. ( 2007 ). "A liberal is someone who has not been mugged": Criminal victimization and political beliefs . Justice Quarterly , 24 , 309 – 334 . doi: 10.1080/07418820701294862 Google Scholar CrossRef Search ADS Van de Walle S. ( 2009 ). Confidence in the criminal justice system: Does experience count? British Journal of Criminology , 49 , 384 – 398 . doi: 10.1093/bjc/azp001 Google Scholar CrossRef Search ADS Voumvakis S. , Ericson R. V. ( 1984 ). News account of attacks on women: A comparison of three Toronto newspapers . Toronto : Centre of criminology, University of Toronto . Wallander L. ( 2009 ). 25 years of factorial surveys in sociology: A review . Social Science Research , 38 , 505 – 520 . doi: 10.1016/j.ssresearch.2009.03.004 Google Scholar CrossRef Search ADS Warner K. , Davis J. ( 2011 ). Using jurors to explore public attitudes to sentencing . British Journal of Criminology , 51 , 1 – 20 . doi: 10.1093/bjc/azr066 Google Scholar CrossRef Search ADS Wemmers J.-A. ( 1996 ). Victims in the Criminal Justice System . Amsterdam : Kugler Publications . Zamble E. , Kalm K. L. ( 1990 ). General and specific measures of public attitudes toward sentencing . Canadian Journal of Behavioural Science , 22 , 327 – 337 . doi: 10.1037/h0078918 Google Scholar CrossRef Search ADS © The Author 2017. Published by Oxford University Press on behalf of The World Association for Public Opinion Research. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

International Journal of Public Opinion ResearchOxford University Press

Published: Oct 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off