The effect of socioeconomic factors, fear of crime and social mobilization on the evaluation of judiciary decisions

O efecto dos factores socioeconómicos, o medo á delincuencia e a mobilización social na avaliación das decisións xudiciais

Manuel Ruiz-Adame mruizadame@go.ugr.es

University of Granada, Spain

Susana Martínez-Rodríguez susanamartinezr@um.es

University of Murcia, Spain

José Antonio Posada-Pérez jposadap@us.es

University of Sevilla, Spain

International University of La Rioja, Spain

University of Atlántico Medio, Spain

Fear of crime (FOC) is “an emotional response of dread or anxiety to crime or symbols that a person associates with crime” ( Ferraro, 1995). This emotion is connected to environmental conditions ( Fleming et al., 2016; Haans & de Kort, 2012) and socioeconomic profiles. This effect can also vary according to gender, age, and past experiences ( Maruthaveeran & Van den Bosh, 2015; Sreetheran & Van den Bosch, 2014). Additionally, neighborhood conditions are also important, with some authors highlighting that FOC is higher among the areas with higher crime rates ( Deka et al., 2018; Putrik et al., 2019; Snedker, 2015).

The links between vulnerability and FOC are strong ( Pantazis, 2000), some studies stating that it is positively associated with socio-demographic characteristics related to social marginality for women, the elderly, the poorly educated, and the unemployed ( Vieno et al., 2013). In any case, the role of certain demographical characteristics is unclear. Although gender and age are two typically mentioned variables linked with higher levels of FOC, it seems that the influence of socio-structural characteristics of the environment and social disorder ( Brunton‐Smith & Sturgis, 2011; Ferretti et al., 2019) are relevant factors to bear in mind. In this sense, some studies have stated that if “controlling for perceptions of community disorder, perceptions of collective efficacy, direct victimization experience and indirect victimization experience”, gender and age are largely unrelated to FOC ( Lee et al., 2020). In contrast, other papers have found that gender is a good predictor of FOC ( Nalla & Gurinskaya, 2020; Özascilar, 2013; Roberts, 2019).

Past experience is a predictor of FOC ( Bachman et al., 2011; Caliso et al., 2020; Hicks & Brown, 2013; Jalain et al., 2020; Riggs & Cook, 2015; Sironi & Bonazzi, 2016), but the vicarious experience provided by TV and social media must also be considered. Events such as kidnappings or child death covered extensively in the media, for example, might have an immediate negative impact, which may be long-lasting ( Prieto Curiel & Bishop, 2016). Social media can also alter the perception of FOC and the perception of safety (PS) ( Hollis et al., 2017). On the same topic, the influence of gender is very weak. In a study performed on television viewers, there was not any evidence to support that crime-related media consumption affected women’s perceptions of crime risk more strongly than men’s ( Callanan & Rosenberger, 2015).

Some researchers have highlighted that social media does not show a real image of the situation, and that there is a strong bias towards violent and sexual crimes ( Prieto Curiel et al., 2020). The association between television exposure and elevated perceptions of one’s own risk of victimization have been studied ( Custers & Van den Bulck, 2013), so identification plays an important role in FOC. Cultivation theory posits that television is an important storyteller and as such is a primary source of socialization in contemporary society ( Custers et al., 2017). In a recent study of the Finnish population, there was found to be a connection between criminal news consumption and the FOC ( Näsi et al., 2021), other authors underlining that the more time is spent consuming social media, the more intensive are fear-generating social and psychological factors ( Intravia et al., 2017). On the other hand, some academics have stated that the influence of the media is not particularly high. Fake news, current events programs, and reality TV shows are not a primary route through which FOC develops ( Chadee et al., 2019). Nevertheless, some authors ( Scherman Teitelboim & Etchegaray Thielemann, 2013), even though they consider that TV news programs play an important role in the perception of FOC or PS, highlight that “the variables with the greatest impact on fear are not related to the media, but include perceptions on the environmental problems, the respondent's gender, age and whether he or she has been the victim of a crime”.

Sexual violence is a prominent social problem ( Zatkin et al., 2021). In particular cases of sexual abuse and rape, perception of justice (PJ) is an issue that students have studied to a large degree ( Anderson et al., 2021; Fedina et al., 2018). Since perception is a subjective assessment, PJ is usually equated with the perception of the performance of the judicial system, particularly its decisions. The results of this research should be read with this in mind.

On this subject, there is a Finnish study that have assessed which punishments students considered should be applied, finding that the victim’s level of consciousness is an important factor. For example, in the case of perpetrators who drug their victims versus those whose victims willingly drink to become inebriated, students felt that the sanction should be lower in the latter case ( Follingstad et al., 2021). Some studies revealed that, in general, people believed that sex offenders should be punished more severely than other types of criminals ( Rogers & Ferguson, 2011). Regarding assessments of the crimes, some authors ( Black & Gold, 2008) found that there are differences between genders in terms of assigning blame based on the socioeconomic status of the perpetrator (e.g., women blamed the rapist more often than the victim if the former belonged to a low socioeconomic status, whereas men tended to assign more blame to the victim and less blame to the perpetrator).

In subjective PJ, especially in socially relevant events that cause alarm, the role of social media ( Blevins et al., 2019; Harlow, 2012; Kumar & Thapa, 2015) and social mobilization (SM) can have a great influence on the personal valuation of the judicial system and about feelings of safety. Silva & Caetano (2016), in a review of the literature about the concept of PJ, stated that socioeconomic factors are mentioned in 20 % of the studies. In any case, the risk of being sexually assaulted or sexually harassed seems to be more common in women and in social minorities ( Klein & Martin, 2019).

Certain cases can cause high social alarm and can create a great deal of SM with street protests, feminist groups playing a significantly important role, attracting attention across social media platforms such as Twitter and Facebook. They can even create legislative changes, as pointed out by Weldon & Htun (2013).

Gang rape is becoming a worrying trend in many societies and several cases are attracting attention around the world, one of which took place in Delhi (India) on December 16 ^th 2012 ( Lodhia, 2015; Roy, 2014; Shandilya, 2015) and could be likened to one known as “la manada” (“the wolf pack”), which our paper is based on. There have been similar cases in Australia occurring in the early 2000s ( Grewal, 2012; Warner, 2004), and in Japan ( Wijer-Hasegawa, 2003), and the felony has been specifically described in the context of military service in the USA ( Sadler et al., 2005). These types of rapes are not singular cases, as we can find them in many other countries, frequently linked with factors such as groups of male children and teenage boys ( Etgar & Prager, 2009), minorities and those living in vulnerable states (e.g. immigrants) ( Worthington, 2013).

In July 2016, in the city of Pamplona, Spain, an 18-year-old woman was claimed to have been sexually abused by a group of five men from Seville during a local festival known as San Fermín. The sentence for sexual abuse, which was announced on April 26th, 2018, was contested by many, who considered that it should have been considered as rape. This nuance has considerable consequences in the Spanish Criminal Code, the penalties being more severe in the second case. The court’s decision caused a furor of media coverage in Spain and in the international sphere ( Beatley, 2019; Comas, 2018; Díaz, 2020; Garcia Valdivia, 2019; Maestro & Zamira, 2020; Muñoz, 2019; News, 2019; Redacción, 2018; Rodríguez, 2020; Romo, 2019; Wikipedia, n.d.), consisting of a flurry of public protests throughout Spain, as well as an outcry on social media platforms ( Aurrekoetxea-Casaus, 2020; Idoiaga Mondragon et al., 2020; Larrondo et al., 2019; Navarro & Coromina, 2020). This sparked an open public, political and academic debate about how consent should be expressed in sexual relations ( Flecha et al., 2020).

The aim of this research paper is to provide empirical evidence on the role on SM on the perception of FOC, as well as to try to understand how the socioeconomic level can influence the subjective PJ and the PS.

We have performed a review of the literature and carried out empirical research. An ad hoc questionnaire was designed ( seeAppendix A), which was divided into three sections: A) Sociodemographic profile, B) Subjective evaluation of safety and justice, and C) Subjective evaluation of the criminal code and fines. Sections B and C were designed to collect information that related to FOC, the PS ( Bolger et al., 2021) and the PJ. We were specifically interested in the effect of gender ( Cops & Pleysier, 2011; Roberts, 2019; Snedker, 2012), the socio-economic environment ( Brunton‐Smith & Sturgis, 2011; Chon & Wilson, 2016; Hoolihan & Thomas, 2020; Podaná & Krulichová, 2018; Putrik et al., 2019; Snedker, 2015) and the compliance of the criminal code ( Yuning Wu et al., 2011) in the context of SM prompted by social media.

The data was gathered in Seville, where the perpetrators of the crime were from, meaning that this case had particular relevance. Only those over twenty years old were surveyed, the total population of this age group at that moment being 559,717 people ( Ayuntamiento de Sevilla [Seville City Hall], 2017). N=605 people answered a questionnaire. The total confidence error at 95% was 3.98%. The data was collected in two periods: the first one (n ₁=454) was performed in May 2018, after the social alarm was raised following the court’s decision; with this sample, the confidence error at 95% was 4.20%; the second one (n ₂=151) was performed four months later, in September 2018, to find out the effect of SM on the PS and the PJ, when the social alarm had decreased; with this sample, the confidence error at 95% was 7.97%. Due to budgetary restrictions, we had to reduce the sample size of the second period, although the confidence error allowed us to obtain statistically valid conclusions. It must also be highlighted that the confidence error for the second period was higher. Ethics approval was not required for this type of research.

The number of participants per city district was proportionally assigned to have a representative sample of all socioeconomic profiles in each period. Participants were randomly selected. The questions that required valuations were recorded on a Likert scale between 1 (very low) and 5 (very high).

This body of work was based on the study of associations between pairs of variables by means of contingency tables, contrasts on the means using the t-test, and contrast on whether the different categories of given variables were equally likely to occur (using the Binomial test if there were only two possible categories, and the Chi-square test for more than two). Additionally, we performed linear regression tests to determine the significance of the association of variables. SPSS version 21 software was used for data analysis.

The test used to determine the statistical significance of the possible relationship between pairs of variables was the Chi-square test. The hypothesis to contrast was the absence of association between the two variables analyzed in each case. The significant levels considered were 1% and 5%. The lower the significant level of the contrast, the greater the evidence of the sample was in favor of that relationship, or in other words, the more significant the result was. To measure the degree of linkage we used Cramer’s V for nominal variables (whose values were between 0 and 1, the closer the value to 1, the higher the linkage). Additionally, for ordinal variables we used Kendall’s Tau B (whose values were between -1 and 1, the nearer the value to 1, the more positive the linkage, the nearer the value to -1, the more negative the linkage).

The general sample (N=605) consisted of 305 women (50.4%) and 300 men. 54.0% were between 31 and 60 years old and 73.9% of the sample had an educational level between medium (high school, 35.7%) and high (university degree, 38.2%). 52.4% were in employment and the rest were retired (19.7%), unemployed (18.0%) or students (9.9%). The income level was divided into multiples of €500 starting from less than €500/month up to more than €2,500/month. The two main groups of workers consisted of those earning less than €500/month (27.8%), and those who earned between €1,000/month and €1,500/month (25.8%).

The PS of the city was generally good, 78.5% considering the safety level to be between normal (47.1%) and high (31.4%). In any case, the city was considered as safe as other cities in the country (36.0%) or less (27.4%) by most of the interviewees. Only 28.1% of the sample had fallen victim to a crime in the last ten years. Among those who had been affected by one, in most cases (74.1%), they were burglaries or robberies.

Meanwhile, the confidence in the judiciary system (mainly understood as judiciary decisions) was low. 54.7% reported that their confidence in it was very low (26.1%) or low (28.6%).

3.1. The effect of the proximity in time to an event that has generated social alarm

In the second period (September 2018), when the social alarm had decreased, the PS of the city had increased ( $\alpha$=0.001; T=-0.121). There were no statistically significant differences for crime victims between the two waves. The subjective PS of the city in comparison with other cities had also increased in the second period ( $\alpha$=0.000; T=0.120). Additionally, the perception of the effectiveness of the police had increased ( $\alpha$=0.000; T=-0.243). In contrast the trust in the legal system had decreased after the event. The general confidence was low for both time periods, but in the second period, it was lower ( $\alpha$=0.018; T=-0.049).

By gender and age, confidence in the courts was not associated with either of these variables (gender and age) in the first period, but there was an association with age in the second period. Confidence in the justice system had significantly fallen in the second period ( $\alpha$=.001). In the case of the effectiveness of the police, opinions were significantly higher by age in the first period ( $\alpha$=0.000), but not by gender ( $\alpha$=0.316), and the same association was seen in the second period. It was significantly higher by age ( $\alpha$=0.043) but not by gender ( $\alpha$=0.962).

The PS was associated with gender according to when (closer to when the social alarm was raised or farther from it) the measure was taken. There was a negative link for females in the first period ( $\alpha$=0.006) but not in the second ( $\alpha$=0.122). In contrast, the opposite phenomenon was seen with age. There was not a significant connection in the first period ( $\alpha$=0.156) but there was a significant ( $\alpha$=0.049) decrease in PS by age in the second period.

Despite the above correlation with the confidence in the justice system, the harshness of the Criminal Code changed between the two waves. In the sample that completed the questionnaire nearest to when the social alarm was raised, the harshness of the Criminal Code was mainly considered as being very low (33.5%) or low (31.9%), and in the second wave it was considered as low (31.2%) or normal (33.1%). The difference between these two points in time was statistically significant at 99% ( $\alpha$=0.000; T=0.249). There were neither any significant associations by gender or age in the valuation of the Criminal Code in the general sample, nor at the time closest to the social alarm but there was a significant drop ( $\alpha$=0.001) by age when the social alarm decreased (period 2). There were no significant differences by gender in either of the periods ( Table 1).

Table 1
Confidence in the justice system, the perception of the effectiveness of the police, the perception of safety, and the perception of the harshness of the criminal code by age and gender in the general sample and by periods (period 1 being the closest time to the social alarm).

Note: *Significant at 99%; **Significant at 95%

The valuation of then police, in general, also improved as the social alarm eased. The perception of the effectiveness of the police had gone up in the second period. In the first period, the effectiveness was considered mainly normal or high (66.8%), the exact values being normal (24.7%) or high (42.1%). Meanwhile, in the second period, more interviewees considered its effectiveness to be normal (41.7%). In general, in said period police effectiveness was considered to be between normal and high (76.1%, high). These differences were statistically significant at 99% ( $\alpha$=0.000; T=-0.243) (see table 2). However, the effectiveness of the police was positively associated with age at the time closest to the social alarm ( $\alpha$=0.000), but negatively associated with age in the second period ( $\alpha$=0.043). There were no significant associations with gender in either of the two periods.

Table 2
The effect of the social alarm on the PS and on trust in the judiciary system

Note: *Significant at 95%; **Significant at 99%Levels (1 to 5) 1.- Very low; 2.-Low; 3.- Normal – Equal (the same as in Seville); 4.- High; 5.- Very High

Regarding punishments for crimes (see table 3), those surveyed mainly agreed with permanent (reviewable) prison, but this opinion appeared to decrease as the social alarm decreased. In this case, the timing was relevant, with there being a positive association between gender ( $\alpha$=.016) and age ( $\alpha$=.006) at the moment closest to the SM (first period), but with a non-significant one at the second period for both of the variables.

As for the death penalty, there was general disagreement about it in all circumstances, but we must underline that the proportion of people who were against it grew as the social alarm fell. The consensus about the harshness of the criminal code improved as the social alarm decreased. In all cases, the mean differences were significant at 99%. The public’s main concerns (see table 3) appears to decrease in all cases with one exception, that of immigration. In all cases, the mean differences were significant at 99%.

Table 3
The effect of the social alarm on the opinion about permanent (reviewable) prison, the death penalty, the criminal code, and main concerns

Note: (*)Significant at 95%; (**)Significant at 99%Levels (1 to 5) 1.- Very low; 2.-Low; 3.- Normal; 4.- High; 5.- Very high

3.2. The effect of socioeconomic factors on the valuation of judiciary decisions and safety: educational and income level

The PS of the city was high in general, but it tended to be higher when the educational level was higher too. Among those with a low educational level, the safety of the city was considered to be between normal and high (73.1%). Those with a high level of education considered the safety to be between normal and high (81.9%) in a higher proportion. These differences were statistically significant at 95% ( $\alpha$=0.002; T=0.134).

There were no differences between interviewees who were previous victims of crime and those who were not.

Trust in judiciary decisions was low, but was the higher the educational level, the better regarded they were; lower educational level holders gave the legal system ratings that were mainly low or normal (53.7%); those with higher degrees of education scored it between low and normal (59.3%). These differences were statistically significant ( $\alpha$=0.046; T=0.121).

The PS of the city versus other cities was lower at both extremes of educational level in comparison with those who had a medium academic level. The differences were significant at 95%.

Table 4
The effect of the educational level on the perception of safety and confidence in the judiciary system

Note: (*)Significant at 95%; (**)Significant at 99%Levels (1 to 5) 1.- Very low; 2.-Low; 3.- Normal – Equal (the same as in Seville); 4.- High; 5.- Very High

In this case, there were no significant differences for confidence in the judiciary system (in general being mainly low or normal) or for the level of perception of the effectiveness of the police (in general being mainly normal or high). Regarding punishments, general consensus was in favor of permanent (reviewable) prison, and against the death penalty. There were no significant differences in these results among the different educational levels. Similarly, the subjective perception of the harshness of the criminal code did not have any significant differences among educational levels, which was between low and normal (see table 5).

There were significant differences (see table 5) for immigration concerns (significant at 99%) and for drugs and crime (significant at 95%). In the case of immigration, the significant differences were between the higher and lower levels of education in comparison to the average. The concern about drugs was higher among those who had lower academic qualifications. In the rest of the cases, for unemployment and terrorism, the concern was high but there were not any differences among the different educational levels.

Table 5
The effect of educational level on the opinion about permanent (reviewable) prison, the death penalty, the criminal code, and other serious concerns

Note: (*)Significant at 95%; (**)Significant at 99%Levels (1 to 5) 1.- Very low; 2.-Low; 3.- Normal; 4.- High; 5.- Very high

The perception of safety in the city was generally high and the greater the income level, the better the opinion of it was ( $\alpha$=0.002; T=0.134). The valuation of the effectiveness of the police showed statistical differences among each level of the variable, which was higher, the higher the income level was. By dividing the sample into two large groups, the first being below-average and average income earners (€1,000/month) and the other, above-average earners (see table 6), those receiving lower incomes considered the police to have a higher level of effectiveness.

The other significant effect regarded the perception of safety in comparison with other cities. It was significantly higher among those with higher incomes (see table 6). On this point, it must be highlighted that there were no significant differences for the confidence in the judiciary system, where for all economic levels the rating given was very low or low. The same result was obtained when we studied this variable for the educational levels.

Table 6
The effect of income level on the perception of safety and confidence in the judiciary system

Note: (*)Significant at 95%Levels (1 to 5) 1.- Very low; 2.-Low; 3.- Normal – equal (the same as in Seville); 4.- High; 5.- Very high

Regarding punishments, the only one for which we found significant effects among income levels was for permanent (reviewable) prison (see table 7). There was a general consensus for this measure, but the proportion of those who supported it was higher among greater-income earners. Again, there was general opposition to the death penalty without any differences among income levels and the perception of the harshness of the Criminal Code was low.

The concerns about terrorism, unemployment, immigration, and drugs as possible sources of crime were high in all cases without any significant differences among income levels (see table 7). The samples in the two periods were similar in terms of educational and socio-economic profiles.

Table 7
The effect of the income level on the opinion of life imprisonment, the death penalty and the criminal code, and other serious concerns

Note: (*)Significant at 95%Levels (1 to 5) 1.- Very low; 2.-Low; 3.- Normal; 4.- High; 5.- Very high

The inadequate response to gang rape cases according to social expectations of justice and governments cause enormous debates at both social and academic levels ( Belair-Gagnon et al., 2014) every time incidents like these occur. Victims often struggle with long-term consequences, such as posttraumatic stress disorder and constant suicide attempts ( Sadler et al., 2005; Ullman, 2007). When cases like these happen, they can encourage changes in the law to be made, but they do not always lead to a decrease in violence towards women ( Sharma & Bazilli, 2014). Instead, a change in mindset when it comes to gender roles and a cultural switch are needed ( Durham, 2013).

To consider the possible effects of social media, our study collected data at two different stages. In the first period, when posting activity was high, the influence of social media was also high, although it decreased in the second period. In our research, we saw that SMs as well as the messages and slogans published during the peak moments of the judicial sentence had a relevant effect on the population’s PS and FOC, but our data shows that these effects did not last for long, which is in line with other studies that have stated that social networks cause a temporal fluctuation of the correlation between the perception of crime and crime that actually occurred ( AitBihiOuali & Graham, 2021; Chaparro et al., 2021). The perception of danger decreased within just four months of the pivotal moment whereas its effect on the PJ lasted for longer. Some authors have highlighted that, regarding the jeopardy of sexual harassment in big cities such as Karachi and London, the PJ and PS were related with the perception of legal protection ( Hoor-Ul-Ain, 2020).

By age, the confidence in judiciary decisions and the PS was lower, even when the social alarm had decreased. Older people were more at risk of being crime victims. The judiciary system is in general low in its process, affecting older people more than young people, perhaps explaining why the former had a higher FOC.

In any case, the confidence in the justice system (basically understood as judiciary decisions) was, in general, low. It was lower during a break event such as a social case that caused social alarm but, in contrast with the PS, the PJ was not so resilient. After general mobilizations, the loss of confidence in justice persisted, which may have been related to what is called “the optimistic bias”, the belief that bad things happen to other people. It is assumed that those who have been victims of a sexual offense are more likely to fear being a victim again ( Chapin & Pierce, 2012).

Women have a lower PS, which is one of the factors that influences FOC. This perception is affected by SM. In contrast, we have not seen a significant effect by proximity in time with the SM among women with confidence in the justice system, the effectiveness of the police, and in the harshness of the Criminal Code.

As we have seen, most people have not been victims of any type of crime in the last ten years, which could explain why confidence in the police’s efficiency is high and increases when social alarm decreases. Even amongst those who have experienced a crime, in most of the cases it has not been sexual assault or rape. Additionally, the social alarm raised by the case of “la manada” did not have anything to do with the PS regarding policing. Despite that, public marches and further protests on social media caused an atmosphere of insecurity that did not reflect the actual findings. This feeling of insecurity was unrelated to past personal experiences but rather to the vicariousness of social media.

It is out of the scope of this body of research to study the characteristics of the messages disseminated by social media or the use of each social media platform in detail, but the effect of the high social alarm elevated by social networks is not out of the scope of this body of research. The solidarity phenomenon through Twitter (currently X) has been studied and some authors ( Aurrekoetxea-Casaus, 2020) have stated that “Twitter's role is to serve as a “loudspeaker” for the outrage of a population more impacted by a ruling considered unjust than by explicit support for the victim”. What is more, social networks can even lead to the “social myth around rape” being highlighted, as defined by Burt (1980), which are myths that serve to deny and justify male aggression against women. Obviously, supporting the cause via Twitter/X does not imply much of an effort when retweeting, although the degree of politicization can play an important role on social debate platforms like this one ( Larrondo et al., 2019; Navarro & Coromina, 2020) so the value of this social support cannot be denied. SMs have led to debates in the European Parliament ( Hedh, 2018; Parliament, 2018) and have forced the Spanish government to review and reform the sexual assault law, proposing one on sexual violence that introduces the concept of sexual consent in terms of “only yes means yes” ( Amigo, 2021; Idoiaga Mondragon et al., 2020; Redacción/Agencias, 2021)

The prejudice towards certain sociodemographic groups such as immigrants is curious in itself as, in the case of "la manada", immigration as a factor that increases insecurity were higher after the peak of the SMs, indicating that cultural prejudice does not need any evidence to rear its ugly head.

During the pivotal moments, protesters called for tougher sentences. One of the issues considered in our collected data that warrants particular reflection is the harshness of the criminal code in terms of to what level society would be willing to go. The death penalty was abolished in Spain in the 1978 Constitution (Art. 15), but was maintained as an exceptional punishment in the military code for war crimes before being completely abolished in 1995. Even though Spain has been targeted by terrorist groups for more than 40 years, in general the population has been opposed to this type of sentence. In our analysis, the data shows general disagreement with the death penalty during the height of the SMs and after that point as well. However, it should be pointed out that the number of those who were in favor during the peak of the protests, fell when the general mobilizations disappeared, albeit with a non-significant difference, which could be read as a warning sign; when negative events occur and pressure is felt, radical ideas can easily begin to appear.

Educational level should also be considered as a factor when the PS and the PJ are analyzed. Our findings were in the same line as previous studies, a low PS and PJ being associated with belonging to minority groups ( Gangoli et al., 2020; Thomas, 2018; Thomas & Mucherah, 2016). The more highly qualified had a greater PS and PJ but in spite of this, the PJ was low regardless of academic level. Similar results were obtained when income levels were analyzed; there were no differences for the PJ, which was low in general for all economic levels. In contrast, the PS was higher when the income level was higher too.

After reviewing the main results of the research, it is important to emphasize its limitations. Gang rape and gang sexual abuse are phenomena that warrant more attention from academics and policy makers. The issues in this paper are not about a singular case. As we have seen in the literature there have been similar cases in other countries and at different moments. However, our empirical results are related to a specific case. Further analysis is needed with larger samples and in more geographical areas, which would lead to a broader view of the topics of this research.

SM occurred all over Spain and with similarly high intensity. We could only gather data from one city, Seville, which was where the perpetrators were from. This could be a bias that needs to be given careful consideration when examining the results. It would have been useful to gather data from Pamplona, where the crime occurred, and from other cities, to compare them.

Regarding the concerns about immigrants, we consider that they should be analyzed due to society being more politically polarized than it used to be. Our study did not include any questions about ethnicity or political orientation. Additionally, it would be interesting to discover the effects of each social media outlet for each sociodemographic group. These goals are out of the scope of our research, nevertheless.

We have measured confidence in safety and the justice system. Safety is mainly understood as the personal feeling of non-vulnerability and justice is frequently understood as balanced decisions of the judiciary system, but these concepts are very complex, and they may have many other interpretations. There is a great deal of literature on these issues that are beyond the scope of this paper, but could be helpful to take into consideration. Additionally, confidence is a subjective perception, and it is clearly influenced by personal and social factors. To understand the effect of the special circumstances, we measured our variables around a peak moment and when the social alarm had decreased. However, these perceptions should be measured in a variety of points in time in order to gain a better understanding of which factors have the most influence and for how long these maintain said influence.

Our results and reflections may be helpful for policymakers so that they can consider how important the educational level and the socio-economic conditions are in contributing to the improvement of confidence in the justice system and in the security provided by the State.

Conceptualization: MRA, SMR, JAPJ, Data Curation: MRA, Formal Analysis: MRA, Funding Acquisition: SMR, Investigation: MRA, JAPJ, Methodology: MRA, JAPJ, Project Administration: MRA, Resources: MRA, SMR, JAPJ, Supervision: MRA, SMR, Validation: MRA, SMR, Visualization: MRA, Writing – Original Draft: MRA, SMR, Writing – Review & Editing: MRA, SMR All authors have read and agreed to the published version of the manuscript.

Project 21947/PI/22 funded by the Region of Murcia (Spain) through the Regional Program for the Promotion of Scientific and Technical Research of Excellence (Action Plan 2022) of the Seneca Foundation – Science and Technology Agency of the Region of Murcia.

The funders had no role in study design, data collection and analysis, the decision to publish, or the preparation of the manuscript.