Reconsidering the Impact of Informational Provision on Opinions of Suspended Sentences in the Netherlands

Introduction

In a recent article, published online before appearing in print in Crime & Delinquency, Van Gelder, Aarten, Lamet, and Van der Laan (2015) report on two studies that focus on Dutch public opinion of suspended sentences. They assessed why some people support more strongly than others sentences that are “imposed but [do] not have to be served on the condition that during a specified period of time an offender abstains from committing criminal acts” (p. 2). Van Gelder et al. convincingly demonstrate that public views of suspended sentences are rooted in general penal attitudes, and that the perceived effectiveness and perceived punitiveness of suspended sentences mediate this relationship. The more one supports punitive ideals, the less one is inclined to perceive suspended sentencing as a punitive and effective form of punishment and be generally supportive of this type of sentencing as a result.

In addition, the most important argument in Van Gelder et al.’s article focuses on issues of how people’s opinions of suspended sentences relate to their knowledge of this type of sentence (Study 1: correlational), and how these opinions are influenced if knowledge increases because information about suspended sentences is provided (Study 2: experimental). This is a promising approach, as one of the key assumptions in the scholarly debate on the so-called “knowledge-deficit model” is that educating people by providing them with information about an issue will help them overcome a lack of knowledge, described as “knowledge deficit,” leading to more positive attitudes toward this issue (Knight, 2007). This assumption is shared by numerous policy makers, politicians, and institutional experts who believe that educating the public about certain measures will lead to them being supported more.

Van Gelder et al.’s experimental study unexpectedly

found that providing information about the punitive elements of suspended sentences did not lead to changes in either opinion or attitude. However, providing information about the goals of suspended sentences did lead to stronger beliefs in the effectiveness of suspended sentences and also to more positive attitudes. (p. 16)

The authors rightly stress that these findings are “modest” in terms of their strength. Indeed, in short, their research does not offer strong support for the knowledge-deficit model, in that increasing people’s knowledge of suspended sentences does not greatly contribute to them having more supportive opinions of this type of sentence.

Although we applaud the efforts of Van Gelder et al., we argue that elaborating on their findings helps explain why they did not identify clear support for their hypothesis, namely that providing people with information leads to more support. In the next section, we discuss a cultural-sociological approach to communication that informs a re-analysis of Van Gelder et al.’s data. We demonstrate that effects of informational provision exist that are far stronger and more salient than those reported by Van Gelder et al.

A Cultural-Sociological Approach to Communication

Although Van Gelder et al. laudably scrutinize the impact of informational provision, they do not take into account the fact that increases in knowledge may have different effects on different types of people: facts can be interpreted in another way by those who have different mindsets and worldviews. Van Gelder et al. implicitly assume that factual knowledge unequivocally affects people’s support for suspended sentences in a positive way to roughly the same degree. This is, however, a rather strict assumption, which is unlikely to be tenable (see Achterberg, Houtman, Van Bohemen, & Manevska, 2010). The supposition that the same stimulus, information, or social phenomenon produces universal reactions among the public-at-large is fiercely criticized within both cultural-sociological literature (e.g. Fiske & Taylor, 1991; Hall, 1973/1980; Mead, 1934) and the communication science literature on framing (Chong & Druckman, 2007; Nisbet & Mooney, 2007; cf. De Koster, Achterberg, Van der Waal, Van Bohemen, & Kemmers, 2014), which commonly argues that people’s cultural predispositions mediate and interpret social life.

Chong and Druckman (2007) explain, “Framing refers to the process by which people develop a particular conceptualization of an issue or reorient their thinking about an issue” (p. 104; cf. Gamson & Modigliani, 1989; Scheufele, 1999). Theorizing on framing emphasizes the importance of cultural frames for the interpretation of everyday life (Goffman, 1974). The basic argument is that these frames—“principles of selection, emphasis and presentation composed of little tacit theories about what exists, what happens, and what matters” (Gitlin, 1980, p. 6)—vary across people of different backgrounds. This underlies the expectation that groups of people re-interpret the knowledge available to them using their wider cultural predispositions as frames: “Every opinion is a marriage of information and predisposition: information to form a mental picture of a given issue, and predisposition to motivate some conclusion about it” (Zaller, 1992, p. 6). The fundamental argument is that the same information does not mean the same to everyone—some will be culturally predisposed to accept certain information readily, or to reach a positive conclusion about it, whereas others will find the same information problematic and will discard it or reach a negative conclusion.

Research in communication science demonstrates that this mechanism is empirically tenable (e.g., Achterberg, 2014; Nisbet, 2005) and the same cultural framing logic may well be applicable to the translation of information about suspended sentences into opinions about this type of sentencing. In other words, it is likely that information that is provided about suspended sentences is also culturally framed instead of uniformly processed. Van Gelder et al.’s experimental study includes two types of information (pp. 13-14; see also Marchand, 2010, p. 5). The first of these is aimed at providing information about the goal of suspended sentencing, explaining that the aim is to push criminals into exhibiting less criminal behavior by means of an “individualized sanction” that is “specifically tailored to suit the offender and the offence” (p. 13). The second type of information explains the punitive element of suspended sentencing:

A conditional sentence still has to be served if a perpetrator breaks the conditions . . . The special conditions that are often set can have a serious impact and adherence to them is monitored. Moreover, committing a new criminal offence will result in also being prosecuted for this new criminal act . . . (Marchand, 2010, p. 5)

The two types of penal attitude included in Van Gelder et al.’s study may well be the cultural predispositions that are important for interpreting these two kinds of information. The first is punitiveness, and as Van Gelder et al. observe, “More punitively oriented individuals are likely to have a less favorable opinion of suspended sentences” (p. 5). More importantly, those who embrace punitive ideals will probably also embrace new information revealing that there are actually punitive elements in suspended sentencing. According to our cultural-sociological framework, people with a more punitive orientation are more likely to accept information indicating that suspended sentences have a punitive nature, leading to them adopting more supportive opinions of this type of punishment (Hypothesis 1).

The second cultural frame consists of an orientation toward restoration and rehabilitation, which, according to Van Gelder et al.’s research, is related to a greater belief in the effectiveness of suspended sentencing (p. 10). More important than this direct effect of restorative and rehabilitative orientation is how it may moderate the impact of informational provision on opinions of suspended sentences. Those who embrace ideals of restorative justice, namely, compensating a victim for the harm caused by the offence (cf. De Keijser, 2000, p. 9) and rehabilitation, that is, decriminalizing offenders (cf. De Keijser, 2000, p. 18), will probably not translate the information provided about the goal of suspended sentencing into more favorable opinions. This is because this information does not focus on compensation for victims, who play “a central role in restorative justice” (De Keijser, 2000, p. 9), but on individualized sanctions for offenders. In addition, these individuals are not expected to react positively to information about the punitive nature of suspended sentences, as this does not aid the decriminalization of offenders. Following our cultural-sociological framework, we therefore expect that information provided about the goal and punitive character of suspended sentences leads less to favorable opinions about this form of punishment among those who are more strongly oriented toward restoration and rehabilitation (Hypothesis 2).

Testing these hypotheses may enable us to explain why Van Gelder et al. did not identify a general, strongly positive effect of informational provision on favorable opinions about suspended sentencing. No such uniform effect is expected; informational provision is likely to have different outcomes for groups with different cultural predispositions—in this case, different penal attitudes.

Data, Measures, and Results

We re-analyzed the experimental data used by Van Gelder et al. (for details, see Marchand, 2010), ¹ which are especially suitable as they include information on opinions on suspended sentences, two types of penal attitude, and an experimental manipulation, whereby, based on random selection, the respondents either did or did not receive goal-related information and/or information about the punitive elements of suspended sentences. This led to four conditions: no information provided, only goal-related information provided, only information about punitive elements provided, and both types of information provided.

We used the same measures as Van Gelder et al. ² and also followed these authors in excluding respondents who completed the survey in less than 2 min (too fast to give serious answers) or took more than 25 min (which is too much time to complete the questionnaire without distractions). Like Van Gelder et al., we also excluded respondents who indicated that the questions were not sufficiently clear. ³ Combining all variables used in our analyses while using listwise deletion produced a data set of 354 individuals.

To test our hypotheses, we performed a series of regression analyses for each dependent variable. All analyses control for gender, age (in years), gross household income (in euros), and level of education (in six categories), and include measures for both types of penal attitudes and for both types of informational provision. For reasons of brevity, we only discuss the coefficients directly relevant for testing our hypotheses. ⁴

The direct effects of both types of informational provision are in line with the findings reported by Van Gelder et al. Provision of information about the goal of suspended sentences and about their punitive elements is not related to the perceived punitiveness of these sentences (the unstandardized regression coefficients and associated p values are, respectively, b = 0.08, p = .28; b = 0.04, p = .59). In addition, information about the punitive elements of suspended sentences does not affect either people’s belief in their effectiveness (b = 0.12, p = .13) or their general attitude toward this type of sentence (b = 0.13, p = .24). In contrast, people who receive information about the goal of suspended sentences are more likely to express belief in their effectiveness (b = 0.21, p = .01) and have a more positive general attitude toward this type of sentence (b = 0.26, p = .02). However, as Van Gelder et al. (p. 16) acknowledge, these effects are modest in strength. Our analyses demonstrate that the difference between the respondents who received goal-related information and those who did not is only 0.21 and 0.26 points, respectively, for variables measured on a five-point scale. In addition, these analyses suggest that information about the punitive elements of suspended sentences does not have any effect on the public’s evaluation of this sanction.

It remains to be seen, however, whether these absent or modest “direct” effects do indeed mean that informational provision does not matter much when it comes to people’s attitudes toward suspended sentences. After all, our hypotheses predict that the effect differs between those who have different penal attitudes, suggesting that the overall effect may seem to be small or non-existent, but might actually be considerable for individuals who have specific penal attitudes. Therefore, we have tested several models including interaction terms.

We find limited support for our first hypothesis, which predicts that people with a more punitive orientation are more likely to accept information indicating that suspended sentences have a punitive nature, leading to more supportive opinions thereof. There are no significant interactions of punitiveness with goal-related information, ⁵ and information about the punitive elements of suspended sentences does not significantly interact with the respondents’ punitiveness in explaining either the perceived punitiveness of suspended sentences or the general attitude toward this sanction. ⁶ However, a meaningful positive interaction effect between punitiveness and information about punitive elements exists when it comes to explaining the belief in the effectiveness of suspended sentences. ⁷ As interaction terms are best interpreted visually (Brambor, Clark, & Golder, 2006), we have depicted in Figure 1 how the effect of informational provision depends on the punitiveness of the penal attitudes of those involved (keeping all other variables at their mean scores).

OPEN IN VIEWER

Figure 1.

Conditional marginal effects of information about punitive elements on belief in effectiveness, by punitiveness (95% confidence interval).

The effect of informational provision proves to be strikingly moderated by punitive penal attitudes, even to the extent that the effect is negative for those with the least punitive views, and positive for those who are most punitive. Among the former, the provision of information about the punitive elements of suspended sentences leads to a decline of 0.93 in the 5-point scale measuring the belief in their effectiveness, whereas it leads to an increase of 0.32 among the latter group. This underlines the relevance of cultural framing: If we had only taken into account the overall effect, we would have echoed Van Gelder et al.’s conclusion that information about the punitive elements of suspended sentences does not have any effect on people’s evaluations thereof.

The second hypothesis states that information provided about the goal and punitive nature of suspended sentences leads less to favorable opinions about them among those who are more strongly oriented toward restoration and rehabilitation. Although information about the punitiveness of suspended sentences does not significantly interact with a restorative and rehabilitative orientation, ⁸ the interactions of this orientation with goal-related information are in line with this second hypothesis when explaining belief in the effectiveness and the general attitude toward suspended sentences. ⁹ We have again graphically depicted these findings, keeping the other variables at their means (see Figure 2).

OPEN IN VIEWER

Figure 2.

Conditional marginal effects of information about goal by restorative and rehabilitative orientation (95% confidence intervals).

The two panels in Figure 2 indicate that the provision of information about the goal of suspended sentences leads to more favorable attitudes toward them among those who are least oriented toward restoration and rehabilitation, whereas it leads to less favorable attitudes among those who are most strongly oriented toward restoration and rehabilitation. Providing information about the goal of suspended sentences to the former group leads to a 1.08 point increase in the belief in the effectiveness of this type of sanction, but leads to a (statistically non-significant) decrease of 0.48 points among the latter individuals. When it comes to the general attitude toward suspended sentences, these effects are an increase of 1.21 and a (non-significant) decrease of 0.49, respectively.

These results strongly underline the importance of cultural framing. First, for both the belief in the effectiveness of and the general attitude toward suspended sentences, the impact of informational provision about the goal of these sentences is much stronger than suggested by analyses of the direct effects; instead of a modest increase in the dependent variable of 0.21 or 0.26 points, respectively, informational provision can make a difference as high as 1.08 or 1.21 points in the five-point dependent variables. Second, our analyses indicate that informational provision does not have the same effect for all people. Indeed, both the size and the direction of the effect depend on the penal attitudes of those involved. Although overall analyses of informational provision suggest that effects are either absent or modestly positive, our findings demonstrate that they are substantially negative instead of positive among respondents who have non-punitive penal attitudes (Figure 1) and substantially (yet non-significantly) negative among those with a restorative and rehabilitative orientation (Figure 2).

Conclusion

Our elaboration on Van Gelder et al.’s study demonstrates that merely investigating the direct effect of informational provision on support for suspended sentencing leads to a severe underestimation of the importance of providing people with information. By discarding the implicit assumption that information about suspended sentences has universal effects, we have been able to scrutinize how informational provision is differently related to opinions on suspended sentences among different groups of people.

The rather modest and sometimes even absent direct effects reported by Van Gelder et al. prove to be the result of two countervailing effects that cancel each other out. Our most important finding is that among people with specific penal attitudes, the effects are not only much stronger than the mere inspection of direct effects suggests, but also in opposite directions (that is, positive for some groups and negative for others). More specifically, although more information about the punitive elements of suspended sentences leads to a stronger belief in their effectiveness among those with very punitive penal attitudes, the opposite is true for people who score low on punitiveness. Moreover, although information about the goal of suspended sentences both increases the belief in their effectiveness and leads to a more positive general attitude toward them among those who do not have a restorative and rehabilitative orientation, such information does not have a positive effect at all among people who embrace such ideals.

Van Gelder et al., of course, discuss the absence of a clear confirmation of their principal hypothesis (p. 16). They relate this to their research design: The same “baseline” information about suspended sentences was given to all respondents, which could have caused the entire sample of people to have changed their mind uniformly. Van Gelder et al. also suggest that the amount of information provided was too limited. However, a cultural-sociological perspective informs two comments on this discussion. First, it is implausible that the baseline information provided to all respondents would have led to a uniform reaction among all sections of the public—some groups are simply culturally predisposed to react differently to the same text than other groups. Second, although discussion on the strength of stimuli is common in experimental research, cultural-sociological studies suggest that it may be at least as important to consider the cultural sensitivity of the public involved (cf. Mutz, 2011). Indeed, whereas some will be culturally predisposed to react to even the slightest stimulus, others with different cultural frames might not respond at all, even if they are stimulated very strongly.

Our findings have implications for the way policies are communicated to the public. Van Gelder et al.’s final remark that “a lack of familiarity with sanctions can indeed make them ‘unloved’ by the general public” (p. 17) should not lead to the conclusion that providing the public with a lot of information will simply help increase support for these policies; among significant parts of the population, it may well have the opposite effect. Although it is a popular approach among scholars studying “knowledge-deficit” effects, providing the public with the same information is likely to yield very different responses, ranging from an enthusiastic embrace of the policies involved to a strong rejection of them, with all possible reactions in between.