Community multiculturalism and self-reported immigrant crime: Testing three theoretical mechanisms
Abstract
EconStor is a publication server for scholarly economic literature, provided as a non-commercial public service by the ZBW.
Full text
Leerkes, Arjen; Fokkema, Tineke; Bening, Jonathan Working Paper Community multiculturalism and self-reported immigrant crime: Testing three theoretical mechanisms UNU-MERIT Working Papers, No. 2020-055 Provided in Cooperation with: Maastricht Economic and Social Research Institute on Innovation and Technology (UNU-MERIT), United Nations University (UNU) Suggested Citation: Leerkes, Arjen; Fokkema, Tineke; Bening, Jonathan (2020) : Community multiculturalism and self-reported immigrant crime: Testing three theoretical mechanisms, UNUMERIT Working Papers, No. 2020-055, United Nations University (UNU), Maastricht Economic and Social Research Institute on Innovation and Technology (UNU-MERIT), Maastricht This Version is available at: https://hdl.handle.net/10419/326757 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-sa/4.0/
#2020-055 Communitymulticulturalismandself‐reportedimmigrantcrime: Testingthreetheoreticalmechanisms ArjenLeerkes,TinekeFokkemaandJonathanBening Published8December2020 MaastrichtEconomicandsocialResearchinstituteonInnovationandTechnology(UNU‐MERIT) email:[email protected]u|website:http://www.merit.unu.edu Boschstraat24,6211AXMaastricht,TheNetherlands Tel:(31)(43)3884400
UNU-MERIT Working Papers ISSN 1871-9872 Maastricht Economic and social Research Institute on Innovation and Technology UNU-MERIT UNU-MERIT Working Papers intend to disseminate preliminary results of research carried out at UNU-MERIT to stimulate discussion on the issues raised.
Communitymulticulturalismandself‐reportedimmigrantcrime:testingthree theoreticalmechanisms Prof.dr.ArjenLeerkes,MaastrichtUniversity&ErasmusUniversityRotterdam ([email protected]) Prof.dr.TinekeFokkema,NetherlandsInterdisciplinaryDemographicInstitute (NIDI),ErasmusUniversityRotterdam&UniversityofGroningen ([email protected]) JonathanBening,MSc([email protected]) Abstract There is considerable contextual variation in crime among immigrants and their native-born descendants, and this study aims to understand that variation better. It examines whether municipal variation in self-reported crimes among Turkish- and Moroccan-Dutch men living in 35 representative Dutch cities (N=911), including the four largest cities, is associated with municipal variation in multicultural attitudes, or ‘community multiculturalism’, among the native-Dutch (N=2,556). We propose, and test, a mechanism-based theoretical model that links Berry’s acculturation theory to general strain theory, social bonding theory, and collective efficacy theory. Evidence is found for a protective effect of community multiculturalism for immigrant crime, which is mostly explained by collective efficacy theory with somewhat weaker evidence for general strain theory and social bonding theory. We discuss implications for the discussion on the (dis)advantages of multiculturalism, and suggest various avenues for further inquiry into immigrants’ ‘context of reception’, and how the acculturation attitudes among established groups affect social cohesion outcomes in multi-ethnic societies. Keywords:acculturationtheory,immigrantcrime,contextofreception,local‐level variation JELClassification:K13,I30,Y80
1 Community multiculturalism and self-reported immigrant crime: testing three theoretical mechanisms Introduction Immigrants’ context of reception – the economic, legal and social conditions in the country of immigration (Portes and Rumbaut (2019) – codetermines immigrant incorporation patterns, and related outcomes such as crime. That context also involves attitudes and policies on migration-related diversity, and there is now an especially rich academic and public debate on the (dis)advantages of multiculturalism – typically understood as a national policy approach. Bloemraad and Wright (2104), for instance, highlight possible advantages of multiculturalism and report that immigrants report less discrimination and higher level of social and institutional trust under multiculturalism. Others dispute that multiculturalism is necessarily advantageous, and argue that it may unintendedly hamper immigrants’ positioning on the labour market to the extent that it puts less pressure on newcomers to learn the destination country’s formal language(s) (cf. Koopmans, 2010, 2013). In Berry’s (2011, 2017) acculturation theory, which informed our analyses, multiculturalism does not necessarily pertain to policies (also see Bourhis et al., 1997); it is an ‘acculturation orientation’ or ideology in the receiving society that – through policies or otherwise – seeks to facilitate immigrants’ and ethnic minorities’ equitable participation, while allowing these groups to maintain heritage cultures to a notable extent.1 Internationally comparative studies can collect no more than suggestive evidence on the (dis)advantages of multiculturalism, as it cannot identify relevant confounders easily: if international variations in a given outcome are correlated with differences in integration policies, it is typically hard, if not impossible, to assess whether the relationship is caused by these policies or, for instance, reflects international differences in the ethnic composition of
2 immigrant populations, labour markets, different admission policies, and so forth. Such methodological limitations are among the reasons why scholars have increasingly become interested in intra-national variation in the context of reception (Bean et al., 2012; Glick Schiller and Çaǧlar, 2009). One intriguing contextual difference, both internationally and intra-nationally, pertains to what could be called ‘immigrant (origin) crime’ (for reasons of brevity shortened to ‘immigrant crime’ in this article): crime among foreign-born residents (‘first-generation immigrants’) and their native-born children (conventionally labelled ‘second-generation immigrants’).2 While both generations tend to be overrepresented among suspected and convicted offenders in Europe – in some countries more than in others – they are not, or considerably less, overrepresented in traditional immigration countries, including the US, Canada and Australia (Bersani, 2014; Bucerius and Tonry, 2014; Kubrin, Kim and Hipp, 2019). Furthermore, there is relevant local-level variation (Piopiunik and Ruhose, 2017; Van San and Leerkes, 2001). In the Netherlands, for example, the degree of overrepresentation of residents with a first- or second-generation Moroccan background among local crime suspects differs between municipalities from 1.3 to 4.2, and is generally lower in cities than in smaller towns (De Boom et al., 2013). We carried out two interrelated research projects with a view to better understanding contextual variation in immigrant crime and its relation to multiculturalism. Both projects focus on local-level multicultural attitudes, or what we will call ‘community multiculturalism’ (hereafter: ‘CM’), rather than multicultural policies, which are typically determined at the national level. In discussions of national policy models, local-level variation in such attitudes is usually overlooked, and the importance of attitudes is often neglected altogether. There is nonetheless considerable local-level variation in acculturation attitudes, even in small countries like the Netherlands, our focal country. There is substantial variation between cities and more
3 rural environments (Maxwell, 2019, Schwarz et al., 2014) but also among municipalities with comparable demographics (Martínez-Ariño et al., 2019). The first research project examined how local-level variation in CM among native-Dutch residents living in the four largest cities and 31 other representative municipalities was associated with differences in the number of registered suspected crimes among all male inhabitants (ages 12–65) with a first- or second-generation Turkish or Moroccan background living in these municipalities in 2010. This was done by adding aggregated survey information as a municipal-level variable to a population database with micro-level police data and other administrative information on all registered residents. CM was measured using Wave 1 of the Netherlands Longitudinal Survey (NELLS), a large sociological survey on the living situation of the Dutch population, carried out in the 2008–2010 period. It was found to be associated with a lower number of crimes suspects in each of the two immigrant groups, especially among first-generation immigrants (no significant effects were found for second-generation immigrants with a Moroccan background). The results of the second research project are reported here. For that project, we similarly used the NELLS to measure CM, but also used it to measure criminal behaviour, as the questionnaire included various items on self-reported criminal offending. Furthermore, by limiting the analysis to the NELLS, we could conduct additional, more direct tests of the three main causal mechanisms that, in our view, promote a negative relationship between CM and immigrant crime, namely: (1) strain reduction (cf. Agnew, 2005, 2016), (2) facilitation of individual-level social bonds (cf. Hirschi, 1969), and (3) facilitation of the extent to which the immigrant group experiences, and is involved in, collective efficacy (cf. Sampson and Wilson, 2012). Triangulating police data and self-reported offending data is crucial, as both are imperfect indicators of criminal behaviour. Police suspect data evidently measures criminal behaviour,
4 but also capture residents’ reporting behaviour and law enforcement practices (cf. Newburn, 2003). Self-reported offending data, in turn, is subject to underreporting, and is limited to lighter forms of crime among a sample of people who are willing and able to participate in surveys. A main strength of the NELLS data is that it can rule out two alternative interpretations of the observed negative association between CM and crime found in the first project, namely (1) that it merely points at local variation in residents’ reporting behaviour and law enforcement (e.g. that a lower punitiveness towards immigrants under CM leads to less formal punishment), and (2) that the association is merely due to reverse causality as immigrant crime may decrease CM. If an association between CM and self-reported crime can be demonstrated, the first of the two alternative interpretations apparently cannot (fully) account for the association. If we can empirically demonstrate paths leading from CM to immigrant crime, both alternative interpretations lose plausibility. Data on self-reported offending is generally considered suitable for such theory testing (cf. Junger-Tas and Marshall, 1999). We thus ask: Is there a negative relationship between CM and the degree of self-reported offending among first- and second-generation immigrant men with a Turkish or Moroccan background, and, if so, can that relationship be explained by municipal differences in the degree to which the men experience strain, social bonds and community collective efficacy? The analysis pertains to the same municipalities as the in first project, similarly focuses on men – as men commit most crimes3 – and similarly focuses on crime among residents with Turkish or Moroccan migration background. The latter focus is partially a consequence of the NELLS, which was designed to also be representative for the two immigrant groups – the two largest ethnic minorities in the Netherlands – not for other immigrant groups. As a result, a considerable number of men with a Turkish or Moroccan background (n=911) answered
11 Data came from the Netherlands Longitudinal Lifecourse Study (NELLS) Wave 1, a largescale sociological study on the opinions and living situation of people living in the Netherlands (Tolsma et al., 2013). Basic demographic and socio-economic variables, such as age, household structure and education, were measured using face-to-face interviews, which were followed by a self-completion questionnaire. Most information used here, including the information on criminal offending, is from the self-completion questionnaires, which are less sensitive to social desirability bias than interviews. The NELLS is representative for the Dutch population (N=5,312), including the Turkish- and Moroccan-Dutch minorities, which were oversampled. The survey was carried out between December 2008 and May 2010 and was collected in the four largest cities and 31 other, randomly selected, Dutch municipalities. About half of the Turkish- and Moroccan-Dutch minorities are residing in these 35 municipalities. The NELLS researchers enriched the data by adding information from Statistics Netherlands on certain neighbourhood and municipality characteristics, including level of urbanization (municipality) and ethnic composition (neighbourhood). We made several choices when deciding on our analytical sample. CM was measured using the native-Dutch respondents only (n=2,556), while the regression analysis only pertain to male participants with a first- or second-generation Turkish or Moroccan migration background (n=1,107). In order to deal with missing data, listwise deletion was used. Multiple imputation was not feasible as most missing data pertained to the dependent variable – criminal offending. We deleted all cases that had missing data on any of the variables included in the regression models, but respondents who did not provide information about their income were kept in the analysis by creating a dummy ‘income missing’. Our final analytical sample consisted of 911 observations across all models. These men were living in 138 different neighbourhoods, as defined by Statistics Netherlands.
12 Dependent variable Five items measure criminal behaviour. The participants were asked if, in the last 12 months, they had been involved in the following activities: ‘stolen something from a person or a shop’, ‘damaged or demolished property of others’, ‘carried a weapon (knife, gun)’, ‘threatened someone’, and ‘kicked or punched someone or participated in a fight’. Respondents could answer on a four-point Likert-scales (‘never’, ‘once’, ‘two–three times’ and ‘four times or more’). Participants could also answer, ‘I don’t want to say’, which was coded as missing. We calculated the sum of reported crimes for each participant with the category ‘never’ coded as 0, the category ‘once’ as 1, ‘two–three times’ as 2.5, and ‘four times or more’ counting as 4. Independent variables There is considerable debate about the definition and operationalization of ‘community’ (Shearer et al., 2007), and we eventually used measures at both the municipal and neighbourhood level, basing our decisions on the level used on both substantive arguments (e.g. ‘will this aspect of collective efficacy mostly operate at the municipal level or at the neighbourhood level?’) and statistical arguments (‘which operationalization leads to the best model fit?’). The dominant group’s acculturation orientation was operationalized by averaging relevant scores of all inhabitants with a native-Dutch background (operationalized as having two parents born in the Netherlands) by municipality. We had several discussions on Berry’s notion of the ‘dominant group’: there is considerable ethnic diversity in the Netherlands, especially in cities, and ‘established’ first- and second-generation immigrants (e.g. Dutch citizens) should perhaps also be included. We decided to only use the scores of respondents with two Netherlands-born parents as Berry describes the dominant group in ethnic rather than
13 civic terms, and we wanted to remain close to his acculturation theory. The municipal level is the most appropriate level for this community measure, as most people spend a considerable amount of time outside of their direct neighbourhood, while relevant policy decisions are also taken at the municipal level – e.g. whether institutions that matter to the immigrant community are facilitated or not. The following items measured tolerance: (1) It would be better for our country if all inhabitants had the same customs and traditions, (2) It would be better for our country if different beliefs exist, (3) It would be better for our country if all inhabitants spoke the same language, and (4) Immigration to our country has to stop to reduce tensions. These items were measured on five-point Likert scales ranging from 1 ‘disagree strongly’ to 5 ‘agree strongly’. After reverse coding some of the items, we obtained a scale with a fair internal consistency ( = .65). Reliability could not be improved by dropping items. To measure perceived social distance to Turkish- and Moroccan-Dutch communities two similar sets of three items were used. The items asked: ‘I have a problem with someone of Turkish/Moroccan origin (1) becoming my boss, (2) moving next door, and (3) marrying my son/daughter’. The questions were measured on three-point scales: 1 ‘no problem at all’, 2 ‘not a problem’, and 3 ‘would be a problem’. The Cronbach’s alpha for perceived social distance indicated excellent internal consistency ( = .93). The aggregated scores on tolerance correlated negatively and strongly with aggregated social distance (r = -0.78), and the items loaded strongly on one underlying construct. We therefore decided to calculate a single score on a 1–3 scale, which we called community multiculturalism (CM), by averaging the scores for tolerance, which we recoded to a 1–3 scale, and the inverted scores for perceived social distance. In Berry’s model, variation in CM thus runs ‘diagonally’ from exclusionism to multiculturalism. Segregationalism and assimilationism cannot be separately identified at the municipal level in the Netherlands.
14 To operationalize strain theory, we used perceived ethnic discrimination and SES. Perceived ethnic discrimination was operationalized by averaging five items: ‘Did you in any of the following situations experience that you were discriminated on the basis of your ethnic background (1) during a job application, (2) at work, (3) at school, (4) on the street, in stores, or in public transport, (5) in associations or sports clubs, (6) during nightlife or discotheques?’ The three answer options were 0 ‘no, never’, 1 ‘yes, once in a while’ and 2 ‘yes, quite often’. The Cronbach’s alpha of these items indicated a very good internal consistency ( = .82). Participants’ SES was assessed using monthly income and educational attainment. Monthly income of the respondent, and his possible partner, was classified into ‘low’ (net monthly income up to €1,499), ‘middle’ (€1500–€2,499) and ‘high’ (€2500 or more). For the 91 participants who did not report income we created a dummy called ‘missing income’. Highest educational attainment (in the Netherlands or elsewhere) was classified into: ‘no formal education or primary education’, ‘secondary education’ (high school, lower professional education), and ‘tertiary education’ (higher professional education, university degrees). When participants were still attending school their unfinished level of education was used. Social bonding theory was operationalized using four individual-level variables: particularized trust, generalized trust, institutional trust and religiosity. Particularized trust was measured with ‘I know many people I can trust completely’ with answers ranging from 1 ‘totally applicable’ to 4 ‘totally not applicable’. Generalized trust was based on ‘Nowadays you really do not know who you can trust’, measured on a Likert-scale ranging from 1 ‘completely agree’ to 5 ‘completely disagree’. Institutional trust was based on ‘Please indicate how much trust you have in the police and justice’, with four answering options ranging from ‘a lot’ to ‘very little’. Religiosity was based on ‘How important is religion to you personally?’ with answering options ranging from 1 ‘very important’ to 5 ‘not important at all’. Some items were reverse coded, so that higher scores indicated higher trust and religiousness.
15 Collective efficacy theory was operationalized using four indicators, two at the neighbourhood level and two at the municipal level. At the neighbourhood level, we used the local immigrant groups’ perceived neighbourhood collective efficacy, which is average score of all Turkish- and Moroccan-Dutch respondents, including female respondents, living in a given neighbourhood on six 4-point items, each ranging from 0 to 3, namely people in this neighbourhood: (1) ‘greet each other’, (2) ‘trust each other’, (3) ‘go along well’, (4) ‘know each other’, (5) ‘like to help each other’ and (6) ‘would say something against youth causing nuisance’ (= 0.84). As an additional neighbourhood indicator, we used the local immigrant community’s degree of neighbourhood contact, which is average score of all Turkish- and Moroccan-Dutch respondents living in a given neighbourhood on a seven-point scale from never (0) to daily (6). At the municipal level, we used the local immigrant community’s trust in the police and justice system and its religiousness. These measures were obtained by averaging the individual scores of all residents of Turkish or Moroccan origin by municipality respectively.We also experimented with measures that were based on the scores of all local respondents (including those of native-Dutch origin), which led to comparable findings, but a somewhat worse model fit. Immigrant crime thus seems to be controlled less by general local levels of local collective efficacy than by forms of collective efficacy with a considerable involvement of the local immigrant group, which partially seem to transcend the immediate neighbourhood.5
16 Confounders As individual-level controls we used age (and its square divided by one-hundred to take the age-crime curve into account (Hirschi and Gottfredson, 2017)), immigrant generation, national origin and household composition. Immigrant generation is categorized into two groups. Firstgeneration immigrants were born in either Turkey or Morocco; second-generation immigrants were born in the Netherlands and had at least one parent born in Turkey or Morocco. Immigrants are often found to have a lower prevalence of criminal behaviour than their nativeborn children (Bersani, 2014), although only limited generational differences in self-reported offending have been found (Leerkes et al., 2019). We controlled national origin to account for possible ethnic differences in offending; the Turkish-Dutch are typically found to have somewhat lower crime rates than the Moroccan-Dutch, although ethnic differences in selfreported offending are small (Leerkes et al., 2019). These ethnic differences are usually attributed to differences in intra-ethnic social cohesion (cf. Bovenkerk and Fokkema, 2015). Three dummies measure household composition: ‘living with partner/child’, ‘living with parents’ and ‘other’. Household composition is a potential confounder since living with a spouse or having children is generally considered to prevent crime (Glowacz and Born, 2015). At the municipality level we controlled degree of urbanization, which is taken from Statistics Netherlands and has four categories: ‘small village’, ‘big village’, ‘small city’ and ‘big city’. Criminal behaviour is generally more prevalent in big cities (Glaeser and Sacerdote, 1999), possibly suppressing the effects of CM, which also has a positive relationship with urbanization. We also controlled the percentage of co-ethnics at both the municipal and neighbourhood level, using Statistics Netherlands data. The percentage of co-ethnic is defined as the number of first- and second-generation immigrants of Turkish (for ‘Turkish’ respondents) or Moroccan (for ‘Moroccan’ respondents) origin as a percentage of the number of residents in the municipality and neighbourhood respectively. The latter controls we used in
17 order to rule out the possibility that CM-related differences in social bonds merely reflect municipal or neighbourhood differences in the opportunity and/or need to interact with coethnics.6 Analytical strategy The hypotheses were tested in Stata 16 using weighted negative binomial regression with robust standard errors allowing for clustering of observations at the municipal level. Robust regression was preferred to multilevel regression, because the NELLS includes individual-level weights only, making proper weighted multilevel analyses difficult (highly similar results were obtained using unweighted multilevel analysis). Negative binomial regression was used because the dependent variable – the amount of self-reported crimes – was highly skewed (cf. Land, McCall and Nagin, 1996). We estimated five models. The first specified the relation between CM and criminal behaviour, while controlling the demographic variables mentioned. It tested H1 and served as a benchmark model. The strain indicators were added to Model 1 in Model 2, the social bonds indicators were added in Model 3, and collective efficacy measures were added in Model 4. Finally, we estimated a fifth model (M5) combining all the variables in M1–M4. Hypothesis 2–4 were tested by comparing the more complex models to the benchmark model (e.g. we tested H2 by comparing the coefficient of CM in Model 2 and Model 5 to the coefficient of CM in Model 1). Postestimation tests did not indicate multicollinearity issues for the variables of interest.7 As an additional control, Model 1 and Model 2 (without perceived discrimination) were also estimated for native-Dutch men and found no effect of CM among these men (not reported). Table 1 shows the means, standard deviations and the level of measurement per variable. [Insert Table 1 here]
18 Results [Insert Table 2 here] Table 2 presents the regression results. In Model 1, CM has a negative association with the number of self-reported crimes. A one-unit increase in CM decreases the difference in logs of the expected number of criminal offences by 12.33 (p < .001). We thus accept H1 that CM is associated with lower levels of immigrant crime. The effect is quite strong: an increase in CM with one standard deviation reduces the incidence rate ratio by 0.29 (not shown). Model 2 adds perceived ethnic discrimination and SES, leading the coefficient of CM to go down from -12.33 to -10.94 (p < .01), a decrease of 11.3%. A notable part of the association between CM and criminal behaviour is thus explained by differences in strain, leading us to accept H2. Additional analyses (not shown) indicate that perceived discrimination, educational attainment and income all contribute to the decrease. We also find that a low income has a significantly weaker effect on crime when CM increases, which could indicate lower vulnerability to certain strains under CM. Model 3 adds the individual-level indicators of social bonds, leading the coefficient of CM to go down from -12.33 to -11.00 (p < .001). That decrease of 10.7% suggests that the negative association between CM and crime is partly explained by individual-level social bonds, leading us to accept H3. Only particularized trust and religiousness (significant in Model 5) turn out to contribute significantly and independently to the decrease. Individuallevel generalized trust and institutional trust are positively associated with CM, but do not predict immigrant crime.
19 The collective efficacy indicators are added in Model 4, leading the coefficient of CM to decrease from -12.33 to -7.02 (p < .05), a decrease of 41.1%. That observation leads us to accept H4. All four indicators of collective efficacy independently contribute to the decrease. When testing all theories simultaneously in one comprehensive model, the coefficient of CM is no longer significant at conventional levels of significance (b = -2.97, p = .06), suggesting that the mechanisms of strain, social bonds and collective efficacy indeed explain (most of the) the negative association between CM and immigrant crime. Discussion Immigrants’ context of reception codetermines immigrant incorporation patterns, and related outcomes such as crime. Departing from acculturation theory, we conducted two interrelated research projects. The projects aim to (1) contribute to the discussion on the (dis)advantages of multiculturalism for immigrant incorporation and related outcomes, (2) help solve the puzzle of the contextual variation in immigrant crime through linking Berry’s acculturation theory with relevant sociological-criminological theory, and (3) show the usefulness of going beyond the country-level and formal policy models when studying immigrants’ context of reception and its influences. The projects focused on municipal variation in CM in the Netherlands and municipal variation in criminal behaviour among men with a first- or second-generation Turkish or Moroccan background. In the first project, we focused on police data, and conducted indirect tests of mechanisms that, in our view, promote a negative relationship between CM and immigrant crime. The second project, which is reported here, focused on self-reported offending and involved more direct tests. We find significant negative effects of CM on immigrant crime in both projects, and also find direct and more indirect evidence that three main mechanisms cause these effects: CM reduces crime (1) by reducing strains in the immigrant group, (2) by strengthening social bonds
20 in the immigrant group, including intra-ethnic bonds, and (3) by promoting the immigrant groups’ involvement in collective efficacy. All these observations indicate that CM reduces immigrant crime, and that the effects observed in the first project do not, or not only, reflect local differences in crime reporting and enforcement. Furthermore, in spite of not having conducted longitudinal analyses, we also feel safe to conclude that the negative association between CM and immigrant crime is not, or not only, caused by reverse causality: the evidence for the hypothesized paths between CM and crime make strong reverse causality unlikely (in the first project, two-stage regression analyses also pointed in that direction). The findings do not lend much support to Koopman’s (2010) claim that multiculturalism impedes immigrants’ labour market positioning, at least not when community attitudes rather than national policies are considered. Evidently, we do not claim that CM annuls all immigrant crime, or that it necessarily reduces all types of crime. When comparing the two projects, we observe that the association between CM and self-reported offending is stronger than the association between CM and registered crime. This may indicate that CM primarily reduces relatively common and ‘lighter’ forms of crime: surveys measure less serious types of crime than are typically registered by the police. Possibly, CM may also trigger mechanisms that increase specific forms of offending. Higher levels of trust and neighbourhood contacts may increase opportunities to commit certain crimes, such as certain ‘organized’, non-street crimes that police data and surveys measure poorly. Higher level of generalized trust and less institutional distrust may diminish the fear of formal punishment, which for some reduce the threshold to commit certain crimes (e.g. by weakening the fear of deportation). Possibly, such opposite mechanisms explain why individual-level institutional trust and generalized trust did not have significant negative effects on the men’s crimes. Should such opposite mechanisms matter, they were clearly too weak to
27 Table 1. Descriptive statistics of the variables for Turkish and Moroccan participants. Level of measurement Minimum Maximum Mean Std. Dev. Criminal behaviour Individual 0 14 0.36 1.44 Community multiculturalism Municipal 1.84 2.28 2.11 0.10 1s t generation Individual 0 1 0.66 0.47 Turkish Individual 0 1 0.52 0.50 Age Individual 14 49 31.79 9.14 Age2/100 Individual 1.96 24.01 10.94 5.57 Big city Municipality 0 1 0.50 0.50 Small city Municipality 0 1 0.29 0.45 Big village Municipality 0 1 0.21 0.41 Living with partner/child Individual 0 1 0.57 0.50 Living with the parents Individual 0 1 0.27 0.44 Percentage of co-ethnics Municipality 0.18 9.18 5.38 2.68 Percentage of co-ethnics N eighbourhoo d 0 31.00 8.58 7.38 Perceived discrimination Individual 0 2 0.44 0.45 Low income Individual 0 1 0.45 0.50 Middle income Individual 0 1 0.28 0.45 Missing income Individual 0 1 0.10 0.30
28 Primary education Individual 0 1 0.54 0.50 Secondary education Individual 0 1 0.28 0.45 Particularized trust Individual 1 4 2.94 0.75 Generalized trust Individual 1 5 2.54 1.04 Institutional trust Individual 1 4 2.38 0.86 Religiosity Individual 1 5 4.33 0.97 Perceived neighbourhood efficacy N eighbourhoo d 2.33 4 3.21 0.21 N eighbourhood contacts N eighbourhoo d 1.79 6 4.08 0.58 Community trust in police and justice Municipality 1.8 3.25 2.38 0.11 Community religiosity Municipality 3.5 5.0 4.45 0.27
29
30 Table 2. Regression coefficients for the independent variables on criminal behaviour Model M1 M2 M3 M4 M5 Community multiculturalism -12.33*** -10.94*** -11.00*** -7.02*** -2.97+ Generation (ref. 2nd generation) 1st generation -0.18 -0.34 -0.25 0.08 -0.50 Ethnicity (ref. Moroccan) Turkish -0.18 -0.14 -0.15 0.01 0.11 Age -0.25* -0.32** -0.25* -0.22* -0.30** Age2/100 0.42* 0.52** 0.40** 0.32* 0.42** Urbanization (ref. large city) City -1.69** -1.47** -1.70** -1.09*** -0.87*** Small town -2.58** -2.34** -2.50*** -1.23* -0.82+ Village -3.15** -2.65** -2.54* -0.42 1.54+ Household composition (ref. other composition) Living with a partner/child -1.43*** -1.38*** -1.18*** -1.13*** -0.68**
31 Living with the parents -0.58* -0.96** -0.46 -0.38 -0.73* Percentage of co-ethnics municipality 0.11 0.08 0.12 0.16 0.12 Percentage of co-ethnics neighbourhood -0.10* -0.10* -0.10** -0.09* -0.10** Perceived discrimination 0.78** 1.34*** Monthly income (ref. high income) Low income 2.07** 2.27* Middle income 2.39** 2.53* Missing income 1.97+ 2.72* Educational attainment (ref. tertiary education) Primary education 1.87** 2.15** Secondary education 1.58** 2.14*** Particularized trust (individual) -0.63** -0.81*** Generalized trust (individual) 0.08 0.15 Trust in the police (individual) -0.02 0.25 Religiosity (individual) -0.29+ -0.28*
32 Perceived neighbourhood efficacy -1.21+ -1.24* Neighbourhood contacts -0.09 -0.41* Community trust in police and justice -1.21+ -4.55*** Community religiosity -1.97*** -2.16*** Constant 30.08** 24.54** 30.00*** 35.89*** 34.43*** #Note. ***p < .001, **p < .01, *p <.05, + p<.10
33 1 While we focus on attitudes, CM may lead to local integration policies (Jørgenson, 2012; Scholten, 2013) and/or affect the implementation of national policies. 2 We acknowledge that native-born sons and daughters of immigrants never migrated themselves and should perhaps be called first-generation native-born rather than ‘second-generation immigrants’, but we stick to established conventions here. 3 In the NELLS, the incidence density of self-reported crimes is about four times higher among Turkish- and Moroccan-Dutch men than among Turkish- and Moroccan-Dutch men (0.36 against 0.09) 4AstheNELLSdidnotmeasureself‐esteem,wecouldnottestwhetherthereisalowervulnerabilitytostrainunderCM,whichpossiblyattenuatestherelationship betweenstrainandcrime. 5 First-generation immigrants in particular indeed have more contact with co-ethnics than with other ethnic groups, and some forms of external social control in particular, such as community religiosity, will transcend the neighbourhood. 6 We also estimated models with percentage of co-ethnics and native-Dutch in the neighbourhood, which did not lead to substantively different conclusions. 7 In Model 5, all VIF values were between 1.14 and 3.60 with the exception of age (because age square is also included). The VIF value for CM ranges from 2.7 in Model 1 to 3.6 in Model 5. 8 When estimating Model 1 by generation and origin separately, the coefficient of CM was -22.5 for first-generation ‘Moroccans’ (p<0.01, n=286), 8.2 for second-generation ‘Moroccans’ (p=0.06, n=151), -31.6 for first-generation ‘Turks’ (p<0.01, n=316), and -9.54 for second-generation ‘Turks’ (p<0.05, n=158). When we estimated Model 1 separately for 18 municipalities with and 17 municipalities without at least 1,200 co-ethnics (as we did in project 1), the coefficient of CM is -16.77 (p<0.01, n=817) and -4.55 (p=0.5, n=94) respectively.
TheUNU‐MERITWORKINGPaperSeries 2020-01 DebatingtheassumptionsoftheThirlwallModel:AVECManalysisoftheBalanceof PaymentsforArgentina,Brazil,Colombia,andMexicobyDaniloSpinola 2020-02 TheLaMarcaModelrevisited:StructuralistGoodwincycleswithevolutionarysupply sideandbalanceofpaymentsconstraintsbyDaniloSpinola 2020-03 Unevendevelopmentandthebalanceofpaymentsconstrainedmodel:Termsof trade,economiccycles,andproductivitycatching‐upbyDaniloSpinola 2020-04 Time‐spacedynamicsofreturnandcircularmigration:Theoriesandevidenceby AmelieF.Constant 2020-05 MappingindustrialpatternsandstructuralchangeinexportsbyCharlotteGuillard 2020-06 Forreal?Incomeandnon‐incomeeffectsofcashtransfersonthedemandforfood byStephanDietrichandGeorgSchmerzeck 2020-07 Robotsandtheoriginoftheirlabour‐savingimpactbyFabioMontobbio,Jacopo Staccioli,MariaEnricaVirgillitoandMarcoVivarelli 2020-08 STI‐DUIinnovationmodesandfirmperformanceintheIndiancapitalgoodsindustry: Dosmallfirmsdifferfromlargeones?ByNandithaMathewandGeorgePaily 2020-09 TheimpactofautomationoninequalityacrossEuropebyMaryKaltenbergandNeil Foster‐McGregor 2020-10 Whatmattersinfunding:Thevalueofresearchcoherenceandalignmentin evaluators'decisionsbyCharlesAyoubi,SandraBarbosu,MichelePezzoniand FabianaVisentin 2020-11 TheproductivityimpactofbusinessvisitsacrossindustriesbyMariacristinaPiva, MassimilianoTaniandMarcoVivarelli 2020-12 Technologicalrevolutions,structuralchange&catching‐upbyJanFagerbergand BartVerspagen 2020-13 Semi‐endogenousgrowthmodelswithdomesticandforeignprivateandpublicR&D linkedtoVECMswithevidenceforfivecountriesbyThomasZiesemer 2020-14 TheeconomicimpactofpublicR&D:aninternationalperspectivebyLucSoete,Bart VerspagenandThomasH.W.Ziesemer 2020-15 Takingthechallenge:AjointEuropeanpolicyresponsetothecoronacrisisto strengthenthepublicsectorandrestartamoresustainableandsocialEuropebyJo Ritzen,JaviLopez,AndréKnottnerus,SalvadorPerezMoreno,GeorgePapandreou andKlausF.Zimmermann 2020-16 MigrationofhighereducationstudentsfromtheNorthAfricaRegiontotheUnited KingdombySamiaSattiOsmanMohamedNour 2020-17 OverviewoftheSudanUprisingbySamiaSattiOsmanMohamedNour 2020-18 Inter‐countryDistancing,GlobalisationandtheCoronavirusPandemicbyKlausF. Zimmermann,GokhanKarabulut,MehmetHuseyinBilginandAsliCansinDoker 2020-19 Howdoesinnovationtakeplaceintheminingindustry?Understandingthelogic behindinnovationinachangingcontextbyBeatrizCalzadaOlvera&MichikoIizuka 2020-20 Publicuniversities,insearchofenhancedfundingbyJoRitzen 2020-21 Ph.D.researchoutputinSTEM:theroleofgenderandraceinsupervisionbyGiulia Rossello,RobinCowanandJacquesMairesse 2020-22 LabourmarketeffectsofCOVID‐19insub‐SaharanAfrica:Aninformalitylensfrom BurkinaFaso,MaliandSenegalbyRackyBalde,MohamedBoly,andElvisAvenyo
2020-23 OccupationalsortingandwagegapsofrefugeesbyChristopherFBaum,HansLööf, AndreasStephanandKlausF.Zimmermann 2020-24 PolicyopportunitiesandchallengesfromtheCovid‐19pandemicforeconomieswith largeinformalsectorsbyRajneeshNarula 2020-25 EconomicgendergapintheGlobalSouth:howinstitutionalqualitymattersbyElena Bárcena‐Martin,SamuelMedina‐ClarosandSalvadorPérez‐Moreno 2020-26 HowimportantisGVCparticipationtoexportupgradingbyGideonNdubuisiand SolomonOwusu 2020-27 Patternsofgrowthinstructuralistmodels:Theroleoftherealexchangerateand industrialpolicybyGabrielPorcile,DaniloSpinolaandGiulianoYajima 2020-28 COVID‐19,Lockdowns,andAfrica’sInformalSector:LessonsfromGhanabyElvisK. Avenyo,JohnNanaFrancoisandTatendaP.Zinyemba 2020-29 Thepoliticaleconomyofpublicresearch,orwhysomegovernmentscommitto researchmorethanothersbyAndreaFilippettiandAntonioVezzani 2020-30 Economicpreferencesacrossgenerationsandfamilyclusters:Alarge‐scale experimentbyShyamalChowdhury,MatthiasSutterandKlausF.Zimmermann 2020-31 Internationalstudentmobilitydecision‐makinginaEuropeancontextbyTalitha Dubow,KatrinMarchand,MelissaSiegel 2020-32 SupplyanddemandinKaldoriangrowthmodels:aproposalfordynamicadjustment byGuilhermeR.MagachoandDaniloSpinola 2020-33 Productiveefficiency,technologicalchangeandcatchupwithinAfricabyEmmanuel B.Mensah,SolomonOwusuandNeilFoster‐McGregor 2020-34 OptimalsocialdistancinginSIRbasedmacroeconomicmodelsbyYosephGetachew 2020-35 Towardsanewindexofmobilemoneyinclusionandtheroleoftheregulatory environmentbyGodswayKorkuTetteh,MichelineGoedhuys,MatyKonteandPierre Mohnen 2020-36 SegmentedpathsofwelfareassimilationbyYip‐ChingYuandZinaNimeh 2020-37 Self‐selectioninphysicalandmentalhealthamongolderintra‐Europeanmigrants byAmelieF.ConstantandNadjaMilewski 2020-38 Theroleofinnovationinindustrialdynamicsandproductivitygrowth:asurveyof theliteraturebyMpehmetUgurandMarcoVivarelli 2020-39 Doesgendermatterforpromotioninscience?EvidencefromphysicistsinFranceby JacquesMairesse,MichelePezzoniandFabianaVisentin 2020-40 Automation,globalisationandrelativewages:Anempiricalanalysisofwinnersand losersbyAntonioFrancescoGravinaandNeilFoster‐McGregor 2020-41 StagnantmanufacturinggrowthinIndia:Theroleoftheinformaleconomyby GbenoukpoRobertDjidonouandNeilFoster‐McGregor 2020-42 Intra‐EUmigration:Sheddinglightondrivers,corridorsandtherelativeimportance ofmigrantcharacteristicsbyMiriamMack,SarahRoeder,KatrinMarchandand MelissaSiegel 2020-43 Rootsofdissent:TradeliberalizationandtheriseofpopulisminBrazilbyFrancesco Iacoella,PatriciaJustinoandBrunoMartora 2020-44 Theroleofdomestic‐firmknowledgeinforeignR&Dcollaborations:Evidencefrom co‐patentinginIndianfirmsbyNandithaMathew,LorenzoNapolitanoandUgo Rizzo 2020-45 Issub‐SaharanAfricadeindustrializing?byEmmanuelB.Mensah
2020-46 Doesvaluechainparticipationfacilitatetheadoptionofindustry4.0technologiesin developingcountries?byMicheleDelera,CarloPietrobelli,ElisaCalzaandAlejandro Lavopa 2020-47 Inter‐sectoralandinternationalR&DspilloversbyReneBelderbosandPierre Mohnen 2020-48 EconomicadjustmentduringtheGreatRecession:Theroleofmanagerialqualityby GilbertCette,JimmyLopez,JacquesMairesseandGiuseppeNicoletti 2020-49 SchumpeterandKeynes:Economicgrowthinasuper‐multipliermodelbyÖnder Nomaler,DaniloSpinolaandBartVerspagen 2020-50 Addressingtheproductivityparadoxwithbigdata:Aliteraturereviewand adaptationoftheCDMeconometricmodelbyTorbenSchubert,AngelaJäger,Serdar TürkeliandFabianaVisentin 2020-51 SocialassimilationandlabourmarketoutcomesofmigrantsinChinabyShuCaiand KlausF.Zimmermann 2020-52 ParentalgenderstereotypesandstudentwellbeinginChinabyShuaiChu,Xiangquan ZengandKlausF.Zimmermann 2020-53 Rethinkinghumanitarianaid&makingthecaseforhumanitariansocialprotection: Aresponsetothe2019GlobalRefugeeForumbyZinaS.Nimeh,TamaraA.Kool, FrancescoIacoellaandAlexanderHunns 2020-54 Women,leadershipandviolentextremism:Apotentialsecurityrisk?byAurélie WertzandDorcasMbuvi 2020-55 Communitymulticulturalismandself‐reportedimmigrantcrime:Testingthree theoreticalmechanismsbyArjenLeerkes,TinekeFokkemaandJonathanBening