Birds of a feather trust each other The effect of contact on trust in foreigners in Qatar

Transcription

1 Birds of a feather trust each other The effect of contact on trust in foreigners in Qatar Majed Al-Ansari and Abdoulaye Diop Abstract : In the social prejudice literature, meaningful contact with people from other backgrounds is attributed with lessening prejudice. This is referred to as the Contact Hypothesis. The objective of this paper is to test the contact hypothesis using a Qatari sample. Contact is defined as having at least one non-qatari friend while trust in foreigner groups Arabs, Westerners and laborers- is used as an indicator of social prejudice. Using linear regression to control for other demographic variables, we did not find a significant effect of contact on trust in foreigners in Qatar. Gender and education level of the respondent were the only variables to exhibit an effect on trust in foreigners only in relation to trust in Non-Arab expatriates. Trust in Arabs was not affected by any variable included in the regression. The outcome of this paper is not in sync with findings of previous studies, stressing the effects of contact on prejudice and attitudes towards members of other groups. Keywords ---: Diversity, Qatar, social cohesion, trust I. INTRODUCTION: The Arabian Gulf area has been widely overlooked in research on diversity even though it is rapidly becoming a sought-after destination for job seekers from both developing and developed nations. Qatar, an Arabian Gulf country, with a population 1 of less than two million is a seemingly unique case, as the majority of the population more than 80 per cent is made up of expatriates of various nationalities. With most expatriates coming from Arab countries and South-East Asia, western expatriates are seen to be of a distinctive nature as their high income is similar 2 to their Qatari counterparts, while their social and linguistic credentials differ greatly. Arab expatriates, on the other hand are in tune with the Qatari population as they share religion, language and many social traditions. However, when it comes to income, the majority of them earn less than the average Qatari. This paper seeks to shed light on the relationships between Qataris and foreigners residing in Qatar through testing the contact hypothesis using data gathered by the Social and Economic Survey Research Institute of the University of Qatar, as part of a project aimed at establishing a record of social capital levels in Qatar. The research question addressed in this paper is does having meaningful contact with foreigners affect Qataris attitudes towards them? Using multiple linear regression, we test Qataris level of trust in foreigners against having at least one non-qatari friend, while controlling for a set of demographic variables. 1 According to the the 2010 government census by Qatar Statistical Authority, Qatar s population is 1, At least at the time of the survey data collection which is prior to the 60% increase in the salaries and pensions of Qataris. The remaining of the paper is divided into three sections. The first section of this paper identifies previous literature on the issue beginning by a conceptual overview of social cohesion, social capital and trust. This section also presents a summary of competing theories attempting to explain what affects attitudes towards members of other groups, minorities and migrant populations. The second section l presents the methodology including information on the source of data, the sample in question and details on variable selection. The final section illustrates results of the regression analysis of three different trust variables against contact and control variables and a brief discussion of the results detailing noteworthy findings, limitations and suggestions for further research about the issue. II.CONCEPTUAL BACKGROUND AND PREVIOUS RESEARCH The literature on social capital and social cohesion is as diverse as the definitions of both concepts. the following is an overview of scholarly work on the two concepts and explanations of the relationships between them. The bridges of a diverse county: defining cohesion Defining social cohesion is seemingly a problematic task as scholars differ greatly in conceptualising the term and its implications. Definitions range from the overly simplistic to the excessively complex. One definition that can be described as minimal and simplistic is Lockwood s [13] definition which reads, Social cohesion refers to a state of strong primary networks (like kinship and local voluntary organizations) at communal level. This definition neglects however, defining the communalities that are needed to achieve social cohesion. The main issue that segments views on the subject is defining the level of integration between members of a community sufficient to maintain and develop cohesion. A recent and better constructed effort to explain social cohesion was attempted by Joseph Chan et al. [3] who identified the term as a state of affairs concerning both the vertical and the horizontal interactions among members of society as characterized by a set of attitudes and norms that includes trust, a sense of belonging and the willingness to participate and help, as well as their behavioural manifestations. Employing elements from social capital, this definition provides a more grounded conceptualisation that explains the direction and shape of the communal bond. Also, Chan and colleagues- acknowledge the importance of social 22

2 capital in attaining cohesion, trust, a sense of belonging, participation and willingness to help as a component of civic virtue which are all measures leading to increased levels of social capital. Social capital is a complex and highly controversial term in its own right; nonetheless, it is widely used by scholars in various disciplines as a measure of connectedness between individuals in society. Social capital theorists came to differentiate between two types of social capital: bonding and bridging. Bonding reflects connections and networks within a homogenous group while bridging is concerned with social ties among people of dissimilar ethnicity [2]. As bridging is more related to the concept of social cohesion [4], it is a point of interest for this analysis. A comprehensive definition of bridging social capital is offered by Putnam [15] who indicates that Bridging social capital concerns voluntary associations and horizontal ties based on common interests that transcend heterogeneous differences of ethnicity, religion and socio-economic status. Scholars are at odds in assessing the relationship between social cohesion and social capital. Putnam [15] suggested that social cohesion is a means of getting ahead in society as opposed to bonding social capital which he describes as getting by. He theorises that, in order to advance one s social position, acquiring cross-border connections with members of other groups who possess different social capital is a necessity. A dependency on bridging social capital for positive social mobility suggests that social cohesion can be reached through realising the benefit of cross-community relations [8]. Putnam s claim is defended by Pieterse [14] who asserts that for bonding social capital to deliver requires bridging social capital. A different view is offered by critics of the positive relationship between social cohesion and social capital, Cheong and colleagues [4], for example, indicate that more social capital does not necessarily lead to better social cohesion. They continue by suggesting that what is usually viewed as good social capital cannot, in the context of social cohesion, remain so as more social capital within groups erodes connections with other groups. Moreover, as Leonard [10] affirms, in order to set in motion the framework for bridging social capital to emerge, the conditions that led to the development of bonding social capital need to be undermined. A growing body of literature supports the claim that social capital in both forms does not support social cohesion. Rajulton [17] illustrates this by affirming that a strong sense of belonging a measure of good social capital can be associated with low tolerance of diversity. Likewise, Cheong et al. advocate that the proposal to (re)create cohesive communities via social capital building may therefore be interpreted as a diversion from the fundamental injustices [4]. Leonard [10] agrees, adding that even the transition from bonding to bridging social capital will reinforce existing inequalities through affirming their existence. The socio-economic status of both individuals and communities are found to be the main influence behind weak social capital and social cohesion [5]-[11]-[12]. Therefore, it can be assumed that achieving social cohesion is a by-product of battling poverty and correcting social injustice. Assimilation of diverse groups into a national or communal identity as a means towards a cohesive society has been discredited by a great deal of scholarly work [5], which leads to the conclusion that a society that acknowledges diversity while insuring justice and prosperity for all will more likely be a cohesive one. To be in contact: explaining prejudice Many attempts have been made at explaining the relationship between diversity and social cohesion, using trust, mainly alongside other measures of social cohesion. The following is a brief discussion of the Contact theory, one of the explanations that relate contact with others to diminishing prejudice against them. The social contact hypothesis has been defined and redefined by various scholars. Despite the various definitions, the main idea, however, remained the same: more contact leads to more positive intergroup relations. Forbes [6] asserts that the argument of social contact theory is that more contact between individuals belonging to antagonistic social groups tends to undermine the negative stereotypes they have of each other and to reduce their mutual antipathies, thus improving intergroup relations. Similarly, Putnam [16] argues that the contact hypothesis argues that diversity fosters interethnic tolerance and social solidarity. As we have more contact with people who are unlike us, we overcome our initial hesitation and ignorance and come to trust them more. Many other arguments follow the same trend; contact with individuals from different groups will lead to a more positive perception of the groups they come from. Empirically, many results have been published supporting the contact hypothesis. Wagner et al. [20] found in their analysis of data collected in Germany that as the proportion of an ethnic group increases in society, prejudice towards this group diminishes and their conclusion was that this is an indication of the effects of more contact. As the group s members increase, the chance of social contact between them and members of other groups increases allowing for more positive views of the new group. The authors, however, fail to produce evidence of actual meaningful contact and build their conclusion on the assumption that increased contact is a function of growing numbers within minority groups. Sigleman and Welch [18] found that the effects of contact are significant in reducing hostility towards one s neighbours of other backgrounds. Hostility towards people of different backgrounds in general was not found to be a function of contact. Stein et al. [19], on the other hand, acknowledge the negative short-term effects of increased diversity. They assert that long-term effects have proved to be positive and their results show that living in a highly diverse area does produce higher tension between groups in the first stages. However, as more contact is facilitated by the increased number of residents belonging to other groups, 23

3 residents generally become more tolerating and show less prejudice against each other. Forbes [6] raises a valid point against contact theory as he asserts that the hypothesis is contradicted by the increasing and enduring prejudice reported in conflict situations, such as between Arabs and Jews in the Middle East, or between different ethnicities in the Balkans. In these situations, time and magnitude of contact were accompanied by increasing, and ultimately armed, conflict. Data III.METHODOLOGY AND RESEARCH DESIGN The Social and Economic Survey Research Institute (SESRI) at Qatar University conducted the first Social Capital survey in Qatar in 2011, as part of a larger project studying social capital and social cohesion in the country titled From Fareej to metropolis: a social capital survey of Qatar. The questionnaire included questions on trust, participation, neighbourhood life and social cohesion between Qataris, expatriates and labour migrant workers. The nationally representative sample used in the study included Qataris, White-collar expatriates and blue-collar workers. A total of 800 Qataris were interviewed through a random selection process within each Qatari household. Variables Contact-- The social capital survey of Qatar asked respondents about the nationalities of their closest friends (up to 5). The variable was then recoded and dichotomized into a new variable that considered respondents with at least one non-qatari friend as having contact with foreigners and those with Qatari friends only as having no contact. Trust--Risking generality and social desirability is almost inescapable when employing quantitative methods. For the SESRI survey at hand, the project team decided not to ask racially provocative questions and to measure prejudice. The team asked a specific set of questions on trust in different groups. The survey question asked How much do you personally trust each of the following groups of people? Respondents were provided verbally with a list of groups, which were rotated for each different interview. All questions used a scale from 0 to 10 where 0 meant don t trust at all and 10 meant trust completely. The groups included: Qataris, Arab expatriates, western expatriates, other expatriates, Arab labourers and other labourers. Not surprisingly, the highest level of trust reported by Qataris was towards Qataris with a mean rating of 8). Other groups received varying levels of trust, with Arab expatriates receiving the highest mean of trust by Qataris at 6. Migrant labourers received the lowest trust rating at 3. Trust in western expatriates was considerably low, scoring a lower score (rated at 3.5) than all other groups, with the exception of other labourers. Factor Analysis-- To provide a better view on the relationship between the trust variables used here, a factor analysis was conducted prior to the regression analysis. The results of the factor analysis yielded two possible components, the first component explained 59 per cent of variance observed, which, after rotation, was reduced to 44 per cent. The second component started weakly before rotation at only 14 per cent; however, after rotation the figure jumped to 30 per cent. As the results indicate, before rotation, factor loadings for the first component were high for all variables; the lowest was for Arab expatriates at.669 and the highest was for Arab labourers at.819. The second component produced low factor loadings for all variables with the exception of the variable about Arab expatriates (.684). These results indicate that respondents trust in Arab white-collar expatriates is, to some extent, different from trust in other foreigners. Rotated loadings provide a clearer picture as factor loading for trust in Arab expatriates is reduced to.144. While remaining variables maintain strong loadings, trust in western expatriates and non- Arab labourers receive almost identical factor loadings (.839 and.832). Trust in other expatriates non-arab or westernand Arab labourers also received near identical factor loadings (.601 and.639). The second component, when rotated, confirms associations portrayed by the first component, as trust in Arab expatriates receives a factor loading of.946 while trust in westerners and non-arab labourers receive low loadings (.097 and 0233). Trust in other expatriates and Arab labourers remained at.510 and.513, respectively. One explanation for these associations is that Arabs are generally viewed positively by Qataris due to language and religion similarities. As evidenced above, even when identified as labourers, Arabs still received higher trust levels than their non-arab counterparts. The association between other expatriates and Arab labourers might result from the fact that both possess positive and negative characteristics in their definitions. Other expatriates are of a similar economic class to Qataris, yet their identity is ambiguous as they are not defined as being of a certain ethnicity or religious background. Arab labourers share language and tradition with Qataris. However they are of a different economic class that is commonly portrayed negatively in the media and in local discourse. Significance of the factor analysis was high (p>.001) using Bartlett s test of sphericity and KMO test yielded a result of.807 confirming sampling adequacy. In response to the results of the factor analysis, three relationships were tested with two indices of trust in foreigners and using the variable on trust in Arab white-collar expatriates separately. The first variable is an index of all trust variables mentioned assuming it portrays trust in foreigners generally. The second variable is again an index of trust in all groups, however this time excluding trust in Arab white-collar expatriates. Finally trust in Arabs is used separately, in a third model. The index of trust in all groups yielded a result of 4.2 out of 10. Excluding trust in Arabs from the index reduced the mean of trust to 3.7 compared to trust in Arab expatriates is at a mean of 6.1. The analysis conducted below will include, alongside contact, a set of demographic variables which have been used in similar studies in the past. 24

4 IV. RESULTS AND DISCUSSION: Using a multiple regression, three index variables representing trust in foreigners were tested against contact while controlling for a set of demographic characteristics. We begin first with explaining the steps taken and results of the regression. Results The first step in the analysis was to test the three trust variables against contact. We begin with the trust variable including all groups (Arabs, white-collar expatriates and labourers, westerners and other white-collar expatriates and labourers). As shown in Table II, the result of this step in the model produces a very weak (b=.15) and non-significant (p=0.3) positive relationship between this measure of trust and contact. Secondly, we test the trust variable including all groups except Arab expatriates (See Table III). Here, the result is even less significant (p=0.8) yet still positive (b=.045). Lastly, when testing trust in Arab expatriates against contact, we find that there is a weak but highly significant positive relationship between the two variables. The data suggests that having a non-qatari friend increases reported trust by a little more than half a point on the ten point scale used for the trust question on average. These results offer little confirmation for the contact hypothesis, nonetheless, we continue by introducing some demographic characteristics to the model. After introducing a set of basic demographic characteristics (gender, age, education and marital status), we find some possible predictors of trust in foreigners, as both gender and education show a highly significant relationship with two of the three trust variables. For the variable including all groups, we find that women are more trusting of foreigners by half a point on the scale. In addition,, Qataris with post-secondary education are more trusting than their less educated counterparts. Both results were significant at the 5 per cent level and the model as a whole was also significant at that level. Contact however, remains as non-significant. (See Table I) Results are not significantly affected when removing trust in Arab, white-collar expatriates from the mean of trust in different groups. As seen in Table III, gender and education are the only variables exhibiting significance and contact remains non-significant. The levels of change in trust with a single standard deviation change in both variables are also almost identical to the previous regression results with both gender and education suggesting half a point more in trust for women and highly educated. When testing trust in Arab expatriates on its own against contact and the first set of demographic variables, the results are in favour of the contact hypothesis as the only significant relationship (p=.001) is exhibited by the contact variable. All other variables are not significant even at the 10 per cent level. Since significance of the model as a whole was not established, it is not possible to consider these results significant. The sum of squares is also considerably lower here than was shown by the two preceding regressions. To this point in our analysis, there is no confirmation that contact affects trust in foreigners. What remains is to add another set of demographic variables and to include the variable on the number of friends reported. (See Table III) In this last step in our model, we add to the set of control variables, four new variables. The first two are complementary to the marital status variable, one considering if the respondent has children or not, and the other is a dummy variable testing for the effects of a past divorce or separation. Also, a variable differentiating between residents of the capital city of Doha and residents of other towns was added to test the effects of area of residence. Alongside this variable is another variable recoding areas of residence to separate areas with a majority of Bedouin tribes from other areas, as Bedouin tribes are expected to exhibit higher levels of protectiveness and traditionalism. Finally, to measure the effect of the reported number of friends a respondent has, the variable on number of friends was added as a control variable. A test of multicollinearty affirmed that no strongly correlated variables affected the model in all accounts. The results however, show no significant change to previous observations. In the case of trust in all groups, as shown previously, gender and education are the two strongest relationships. However, significance is stronger in reference to gender (p<.005) while education was only significant at the 10 per cent level. Contact showed a weak positive relationship suggesting a slight increase in trust when having at least one non-qatari friend. However, significance is extremely low (p=.166). The model as a whole is significant at the 5 per cent level (p=.004). With Trust in all groups excluding Arab expatriates as the dependant variable, the significance of the model increases (p=.001) and the significance of the relationship between gender and trust also increased, now at the 1 per cent level. Other figures however, remain unchanged. The likelihood of women giving a higher score of trust slightly increases (unstandardized coefficient went up from b=.515 to.672) while education was relatively not affected. Trust of Arab expatriates against contact still produces a significant positive relationship even when controlling for all variables stated above. The model as a whole is, however, increasingly non-significant with p at.058. Also here, no other variable proved significantly related to trusting Arab expatriates. The best fit among the three versions of trust tested against was exhibited when removing Arab expatriates from the model as r=.217 and r^2=.047. In all versions of regression attempted within the model, gender and education are the only variables affecting trust in foreigners with the exception of trust in Arab whitecollar expatriates where gender and education failed to act as predictors and contact was the only variable exhibiting a significant effect. 25

5 TABLE I PREDICTORS OF TRUST IN ALL FOREIGNER GROUPS [1] [2]** [3]** Contact Gender (Male) -.521** -.515** Education (Post-secondary).462**.450*** Being married Age Being Divorced or separated Having children Living area Living area (Bedouin).230 Number of friends reported.005 TABLE II: PREDICTORS OF TRUST IN ALL GROUPS EXCEPT ARAB WHITE-COLLAR EXPATRIATES [1] [2]* [3]** Contact Gender (Male) -.628* -.627** Education (Post-secondary).501*.475*** Being married Age Being Divorced or separated Having children Living area Living area (Bedouin).297 Number of friends reported TABLE III PREDICTORS OF TRUST IN ARAB WHITE COLLAR EXPATRIATES [1]* [2] [3] Contact.654*.659*.615** Gender (Male) Education (Post-secondary) Being married Age Being Divorced or separated Having children Living area Living area (Bedouin).039 Number of friends reported.084 Discussion From the analysis above, we can conclude that, in reference to this sample, no significant connection has been found between having a non-qatari friend and levels of trust in foreigners. Using trust in non-arabs only did not change the results significantly, however, when testing trust in Arab expatriates only against contact, a weak positive relationship was found between the two variables. After introducing control variables, the model as a whole became insignificant while contact maintained its significance at the 5 per cent value. When regressing trust of all non-qatari groups with and without Arab expatriates on contact and control variables, two variables exhibited a significant relationship: gender and education. Women and the highly educated, according to this model, are more likely to trust foreigners. When considering trust in Arab expatriates only, these two variables, while maintaining the direction of the relationship, remained not significant. As argued by Alesina and La Ferrara [1], education is a factor of trust. However, contrary to their conclusions, divorce or past personal trauma did not exhibit a significant effect. The results are in agreement with Sigelman and Welch [18] in that contact does not necessarily facilitate better attitudes towards other groups. However, it does so in certain contexts. In our case, it is effective minimally when considering trust in Arab expatriates. Further research should include analysing non-qataris attitudes towards Qataris and if they are affected by contact. The data utilised for this analysis does offer in its complete form trust levels by all groups towards all groups and therefore could not be used for such analysis. REFERENCES [1] Alesina, A. and La Ferrara, E. (2002). Who Trusts Others? Journal of Public Economics 85, [2] Beaudoin, C. (2011) News Effects on Bonding and Bridging Social Capital: An Empirical Study Relevant to Ethnicity in the United States. Communication Research, [3] Chan, J et al. (2006). Reconsidering Social Cohesion: Developing a Definition and Analytical Framework for Empirical Research. Social Indicators Research 75.2: [4] Cheong, P. (2007), Immigration, Social Cohesion and Social Capital: A Critical Review." Critical Social Policy 27.1, [5] Fieldhouse, E. (2008) Social Capital and Ethnic Diversity. Ed. Nick Johnson. Citizenship, Cohesion and Solidarity. London: Smith Institute. [6] Forbes, H. D. (1997) Ethnic Conflict: Commerce, Culture, and the Contact Hypothesis. New Haven: Yale UP. [7] Giles, M. W., and A. Evans. (1986) The PowerApproach to Intergroup Hostility." Journal of Conflict Resolution 30.3: [8] Granovetter, M. (1973) The Strength of Weak Ties. American Journal of Sociology 78.6: [9] Hope Cheong, P. (2006) Communication Context, Social Cohesion And Social Capital Building Among Hispanic Immigrant Families. Community, Work & Family 9.3: [10] Leonard, M. (2004), Bonding and Bridging Social Capital: Reflections from Belfast. Sociology 38.5: [11] Letki, N. (2008), Does Diversity Erode Social Cohesion? Social Capital and Race in British Neighbourhoods. Political Studies 56.1: [12] Li, Y. (2003) Social Capital and Social Exclusion in England and Wales ( ). British Journal of Sociology 54.4: [13] Lockwood, D. (1999) Civic Integration and Social Cohesion. Capitalism and Social Cohesion: Essays on Exclusion and Integration. By Ian Gough and Gunnar Olofsson. New York: St. Martin's [14] Pieterse, J.(2003) Social Capital and Migration: Beyond Ethnic Economies. Ethnicities 3.1: [15] Putnam, R. (2000) Bowling Alone: the Collapse and Revival of American Community. New York: Simon & Schuster. [16] Putnam, R. (2007) E Pluribus Unum: Diversity and Community in the Twenty-first Century The 2006 Johan Skytte Prize Lecture. Scandinavian Political Studies 30.2: [17] Rajulton, F. et al. (2007) Measuring Social Cohesion: An Experiment Using the Canadian National Survey of Giving, Volunteering, and Participating." Social Indicators Research 80.3: [18] Sigelman, L, and Welch S. (1993) The Contact Hypothesis Revisited: Black - White Interaction and Positive Racial Attitudes. Social Forces 71.3: [19] Stein, R. et al. (2000) Reconciling Context and Contact Effects on Racial Attitudes." Political Research Quarterly 53.2: 285. [20] Wagner, U. (2006). Prejudice and Minority Proportion." Social Psychology Quarterly 69.4: