Abstract: Nigeria with its multi-ethnic populace presents an appealing prospect for the testing of hypotheses on ethnic tolerance. This study isolates factors associated with portion of this phenomenon. Specifically, the study assesses the singular and joint effects of gender, age, marital status, religiosity and socio-economic status as predictors of ethnic tolerance. Ethnic tolerance in this study is indicated by a willingness to form cross-ethnic friendships accept cross-ethnic marriages and participate in cross-ethnic entertainment. Data for this study is derived from a cross-sectional survey of the Yoruba and Ibo ethnic adults. The results indicate that age, education, marital status and religiosity demonstrate varying and significant impact on ethnic tolerance. Implications of findings are discussed and recommendations for further studies in the area are suggested.

INTRODUCTION

The prevalence of ethnic antagonism with its concomitant catastrophic toll on time, resources and lives confronts us with a disquieting reality. The mania of tribal killings in Rwanda and Northern Nigeria; the fickleness of minds under the sway of ethnic passion evident in the tragic internecine battles in the Transylvanian region, the Nigeria’s civil war of 1967-1970 and the protracted conflict between factions of the Middle East (to mention only a few cases) cry out for urgent scientific scrutiny. This study seeks to understand the nature and dynamics of ethnic tolerance in Lagos, Nigeria.

The study of ethnic tolerance in Nigeria is especially relevant and important given its history. Not only is the country one of ethnic plurality but in 1967, it suffered one of the bitterest ethnic conflict recorded in Sub-Saharan Africa. Millions of lives were lost in the politically charged Biafran civil war.

Ethnic divisions constitute barriers to political coherence and social solidarity in most African countries and it is not easy to break new ground in the area of Nigerian ethnic relations.

Wilson (1998) at a more general level laments the deficiency of sophisticated quantitative studies in the related literature. The literature on ethnic relations in Nigeria clearly evidences this deficiency. Except for Landis (1993) mainly descriptive efforts effort on ethnic attitudes, little else is known about this topic.

As such, this study extends the level of analysis employed in the investigation of inter-ethnic attitudes in Lagos, Nigeria. It moves beyond mere descriptions to examine multivariate predictors of ethnic tolerance. Specifically, it assesses the singular and joint effects of gender, age, marital status, religiosity, income, occupation and education on inter-ethnic tolerance.

The nature of ethnicity: Fundamental to this study are questions about what is ethnicity and how it arises, persists and changes. We consider several approaches to answering these questions.

Some theorists treat ethnicity as a given (Brockman, 1998; Gordon, 2008). According to this non-rational/ primordalist perspective, ethnicity is essentially biological, innate and unchangeable. For Burgess (1998), ethnic attachment is based more on a sense of natural affinity than on social interaction. Gordon (2008)’s human nature theory underscores the importance of the biological component of behavior. He argues that since ethnicity cannot be shed by mobility, it becomes incorporated into the self.

Thus, when one rises up in defense of one’s ethnic group, one is in fact defending one’s self. Wilson (1998) concludes that in the light of this theoretical view, the potential for ethnic conflict will always be present given the narcissistic and often aggressive defender of selves human beings.

Rational/situational theorists regard ethnicity as a socially constructed phenomenon. Keyes (2002) notes that ethnicity is learned rather than genetically determined. For De Vas (2005), it is a feeling of continuity with the past which at its deepest psychological level, represents a sense of survival. It is an embodiment of shared beliefs, norms, values, preferences, in-group memories, loyalties and consciousness of kind (Schermethorn, 2008). Bonavich (2000) notes that ethnic identity is neither natural nor inherent to groups. Yancey (2006) argues, for example that European immigrants to North America developed a sense of ethnicity nor not so, much through the cultural communalities, they shared as a group but more so as a consequence of the external factors encountered. Thus, Banton (1993) observes that though ethnic groups constitute a community of culture, their resilience and power are contigent upon the vicissitudes of social change. This position is reinforced by Alba (1985) who contends that ethnic boundaries are not permanent, noting that without patterns of endogamy and some institutional base, ethnic attachments tend to attenuate and ethnic groups assimilate. The persistence of ethnic groups is possible only to the extent that they become composite communities, combining culture, affinity, institutional protection and a shared interest (Cornell, 1995; Yancey, 2006).

Factors associated with ethnic relations: The notion that ethnicity and ethnic relations are responsive to external forces has been well documented. Modernization theory links inter-ethnic rivalry with rapid economic development such that the concomitant industrialization and urbanization trigger ethnic/political rivalries noted in strong atavistic tendencies (Deutsch, 1999). In the view of ethnic competition theory, ethnic antagonism is linked to resource mobilization (Melson and Wolpe, 2000; Bates, 1974; Young, 2006; Olzark and Nagel, 2006).

The theory posits that as groups compete in the labor markets and thereby increase their access to the same range of resources, ethnic mobilization tends to occur. Drawing support from this and Korpi’s power balance theories, Horowitz (1995) observes that ethnic conflicts tend to increase as contending groups in a society come closer to parity in resources.

In contradistinction to this position, ethnic dominance theory proposes that it is the dominance of a single ethnic group that makes for ethnic instability. Thus, the monopoly of the power resources by a single ethnic group within a country excites ethnic group rapid economic development and the accompanying industrialization and urbanization, competition for scarce resources and increases in the parity of resources accessible to competing groups. In the case of Nigeria, post colonial political, economic and social change are all factors which could subtly or visibly influence the nature of ethnic relationships.

Ethnicity; the case of Nigeria: Opinions differ from one scholar to another on the status of ethnicity at the level of Nigerian society. Most scholars in the field subscribe to the theory that there are three dominant ethno-cultural or ethno-linguistic groups in Nigeria. Such researchers also recognize the considerable ethnic heterogeneity that researchers have observed as being characteristic of Nigerian society but their frequent emphasis is often upon three dominant ethnic groups the Hausas, the Ibos and the Yorubas as well as their pattern of interaction (Lloyd, 1967; Olorunsola, 1972). All these researchers recognize the ethnic heterogeneity in Nigerian society while at the same time, emphasizing the dominance of three ethnic groups. Also, both ethnic harmony and conflict in colonial societies like Nigeria are perceived as a function of the maneuvering of the dominant ruling class. Horowitz (1995) notes that colonial powers in multi-ethnic societies deliberately create and reinforce ethnic boundaries through missionary education, favoritism, importation of cheap labor and political maneuvering to ensure easy governance.

While the conditions referred to ensure easy governance, divisions and ethnic conflict in Nigeria, it is important to bear in mind that neither ethnicity nor the conditions that make for increasing consensus among social scientist that ethnicity is changeable (Keyes, 2002). In an attempt to facilitate clarity of understanding on ethnic boundaries researchers have drawn upon instrumentalists assumptions (Banton, 1993; Wilson and Williams, 2000). The basic premise is that members of ethnic groups behave like rational individuals and seek to optimize their social benefits in given situations. As such, the patterns of formation and dissolution of ethnic groups become predictable since in the light of this theoretical position, costs and benefits considerations assume primacy. Thus, Wilson (1998) notes that ethnic boundaries will be preserved only when it appears that greater advantages will be gained through group identifications.

However in spite of such evidence, the plural society model with its images of rigid, inflexible boundaries continues to provide the mainspring for the analysis of ethnic relations in Nigeria.

While one cannot deny the presence of conditions with counter productive potentials to the process of integration and ethnic tolerance in the Nigerian society, we argue that there are harmonizing measures operative in the system that seem to create a milieu that should be supportive of tolerance. Such measures include common schooling, multi-ethnic political parties and inter-ethnic marriages. Many of these measures happened in the post colonial era.

The conceptual model: Consistent with past researches, I hypothesize that intermingling between the Yoruba and Ibo ethnic groups varies by demographic and socio-economic factors. It has been observed that modernization has led to a mitigation of cultural differences in Nigeria (Lloyd, 1967). This position while running counter the modernization theory (Deutsch, 1999; Wilson, 1998) discussed elsewhere in this study accords with that of Keyes (2002) and Tuch (1997), both of whom identify migration and expansion of the boundaries of the state as sources of change that impact and modify ethnic identity.

The conceptual model builds upon the rational/ situational perspective that treats ethnicity as a social construction and the instrumentalist assumptions that conceive of ethnics as rational beings who calculate their behaviors and modify their responses to optimize social benefits. Analytically, the model suggests that one’s position in the social system would have direct impact on the nature of ethnic relations for the population under consideration. Religiosity as an indicator of one’s philosophic or ideological and practical orientation is also hypothesized as an influence on ethnicrelations.

MATERIALS AND METHODS

Data and measures: The study is based on data collected in Lagos, Nigeria during the summer of 2010. About 450 respondents were selected from a multi-stage probability samples of households in Lagos area. Matters relating to the specific sampling techniques utilized received ample treatment elsewhere. Response rate for the study was 87% and coder reliability ranging between 97-100%. For this analysis a subset of approximately 340 respondents is used.

Dependent variables: The construct, inter-ethnic tolerance, the dependent variable for this study is conceived as the preparedness on the part of one to accept others for whom they are and a predisposition to trust them. Such an attitude is thought to be indicated by the following:

Gordon (2008) identifies friendship as one of the indicators of structural assimilation. Respondents were asked to indicate how many of their close friends were from other specified ethnic groups. Five categories were provided for responses ranging from none to all (Table 1). We regard such a choice as indicative of a pre-paredness to trust and share loyalties and certainly an emotional and psychological transcending of ethnic boundaries.

In the case of cross-cultural entertainment, the Yoruba respondents were asked to indicate how often they listen to Ibo radio and television programs, attend Ibo ethnic group functions such as weddings, parties. Response codes ranged from never (1) to very often (4). We view these measures of cross-ethnic entertainment as indicative of ethnic tolerance.

Exogamy has long been regarded as a test of ethnic tolerance. Respondents were not asked to respond to the question from a personal level but instead were asked if they would mind if a very close relative marry someone of another ethnic group. This bogus pipeline approach (Jackman, 2008) was preferred in order to blunt the porgnancy of a rather sensitive question. Respondents were asked to indicate if it would bother them a lot (3), a little (2), not at all (1).

The analysis of ethnic tolerance in this study is confined to the two dominant ethnic groups of Yoruba and Ibo in Lagos. Insufficient cases preclude the analysis of any of the other groups.

Table 1:	Percent distribution of selected sample characteristics

Predictor variables: Religious practice (religiosity) has been consistently used to predict ethnic tolerance (Landis, 1993; Deutsch, 1999; Bates, 1974). Religiosity is a summated variable that includes the regularity with which one prays, attend s religious services, listens to religious programs and reads religious books. Responses were originally coded on a four point Likert scale ranging from 1-4 indicating increasing levels of frequency. Several researchers have found education to vary with ethnic tolerance (Melson and Wolpe, 2000; Landis, 1993).

Keyes (2002) in his attempt to illustrate the impact of social forces upon ethnic identity notes that the Malay government utilized educational means to address ethnic division of labor in Malaya. Accordingly, the position is taken in this study that inter-ethnic tolerance in Nigeria should vary with educational background. In order to determine their education backgrounds, respondents were asked to indicate the highest level of education reached. Education is analyzed as dummy variables indicating the major distinctions in educational accomplishment in Nigeria. These categories are primary (completion of a primary education or less); secondary and post- secondary are analyzed as dummy variables.

The capacity to deal with others is often constrained by the opportunities one has to interact. It is reasonable to expect that the greater the level of such involvement in terms of intensity and duration, the stronger will be the bonds formed. Some occupations increase the chances of meeting and interacting with others regularly and collaboratively while others do not. Occupational involvement in the formal and especially the public sector allows for such inter-ethnic interaction. We expect persons involved in occupations that require them to be meaningfully involved with others on a regular basis will be more tolerant than those who do not have such opportunities. In this study, occupation is collapsed into unemployed, service, clerical, skilled and semi-skilled, managerial/professional workers and are analyzed as dummy variables.

Income together with education and occupation is often used as an indicator of socio-economic status. Though under most conditions, the three variables tend to be highly correlated, a test for internal consistency among these variables did not support this position for the population in this study. For this reason, we analyze these variables separately. Because income is characteristically skewed, we utilize the inter-quartile ranges as cot off points for income levels and recorded these as dummy variables. Age is often an index of experience, maturity and adaptability. But, it may also be indicating rigidity and inflexibility of attitudes. In the area of ethnic relations, we expect to find these variables and therefore, included age (measured in years) in the analyses, though without any explicit directional predictions.

Gender can be a powerful predictor of behavior. Almost in every society, males and females experience different social pressure to conform based on the differential social expectations to which they relate. In Nigeria, there are gender differences in the level of social interaction with the wider society.

There is generally less interaction with the larger society by married women than among unmarried women. It is not unreasonable, therefore to expect that ethnic tolerance will vary by gender (coded 1 = male and 0 = female) in the sample being considered. Marital status was also dummy coded with 1 = married and 0 = unmarried and included in the analysis.

RESULTS AND DISCUSSION

Three levels of analysis are reported. The descriptive information highlights the sample characteristics, the bivariate analysis cross-tabulates ethnicity with each of the dependent variables. The multivariate analyses, ordinary least square and logistic regression models, gauge the singular and collective effects of the independent variables on the dependent variables.

Descriptive: Table 1 shows some of the demographic and socio-economic characteristics of the data. The Yorubas constituted about 50% of the sample and male representation for this group was higher (55%) than for the Ibos (44%). While the percentage of those completing a secondary education is almost similar, for both ethnic groups, the percentage of the Yorubas (26%) completing higher education is somewhat higher than the Ibos (19%). Most (54%) of the Yorubas’ respondents were muslims, though about a quarter of them were Christians. The Ibos’ respondents were predominantly Christians. Also, the Yorubas have about a 10% higher representation in managerial profession and in the higher income group.

Table 2 compares Yorubas’ responses to questions regarding the frequency with which they listened to Ibo programs or watch Ibo movies with Ibo responses to the question as to how often they attend Yoruba, Yoruba programs across the categories.

Table 2:	Percent distribution of participation in cross-ethnic entertainment

Whereas, about 21% of the Yoruba respondents said they attended Ibo parties very often, only 6 and 4% of the Ibo respondents said they listened to and attend Yoruba programs and parties very often. We observe also that a total of about 47% of the Yoruba respondents attend Ibo parties either very often or sometimes. When this is compared with Ibo participants in Yoruba programs and movies, only about 38 and 19% respondents indicated interest at these levels. When cross-ethnic friendship is examined for the two groups, the Yorubas reported a higher level of inter-ethnic friendship. About 6% of the Yorubas reported that about 40% or more of their friends are Ibos. In comparison, only 28% of Ibo respondents indicated that about 40% of their close friends are Yorubas. A slightly different picture emerges when the respondents to concern about inter-ethnic marriages are examined. While 51% of Yorubas expressed no concern at all about such marriages, 87.4% of Ibos expressed such concern.

Multivariate analysis: We employ both linear and logistic regression procedures to assess the impact of several predictor variables on inter-ethnic tolerance. The logistic regression models were used for the dependent variables assessing concerns about inter-ethnic marriages. The distributions for these variables permitted only simple dichotomies reflecting concerns or no concern for inter-ethnic marriages. Multivariate models for inter-ethnic friendships are not presented. None of the predictors significantly differentiated level of inter-ethnic friendships for this population.

The analysis reveals several interesting findings. The diferential impact of age on the Yorubas and the Ibos is somewhat puzzling. Since older Nigerians from both ethnic groups would have been exposed in ethnic conflicts, age was expected to be consistently negatively related to tolerance for both groups. The results indicate a differential effect for both age on the measures of inter-ethnic conflict employed in this study and across Yoruba and Ibo respondents. While older Yoruba were less likely to attend Ibo entertainment events, their greater tolerance for inter-ethnic marriages may suggest that the former may not be related to intolerance of such events. Rather, it may suggest that older Yorubas do not attend such events for other reasons.

On the other hand where statistical significance was reached, less tolerance for Yoruba events and inter-ethnic marriages was evident among older Ibos. While one may not be able to say with absolute certainty if one is willing to concede that we are not passive reactors to the social gestalt, the differences between Yoruba and Ibo is understandable. The exposed rational situationist theoretical perspective suggests that the experiences represent a calculating creative response to the environments. Thus, despite the common experience of an ethnic conflict, individuals could respond differently.

Alternatively, the propinquity factor would favor homogeneity for the Yorubas in the predominantly Yoruba Lagos’ area and exogamy for Ibo. In other words, while the Yorubas living in a predominantly Yoruba Lagos’ area find it easier to find a spouse of a similar ethnic group, the same may not be true of the Ibos. Hence, consistent with theories of minority/majority group influence, the pressure to appreciate exogamy may be greater among the Ibos. The effect of marital status on ethnic tolerance is consistent for both groups. Respondents who were married tended to be less tolerant of inter-ethnic marriages than those who were unmarried. It is plausible that married individuals were responding from the context of their own experiences since practically all were married to a partner of similar ethnic background. Alternatively, marriages more than age could be conceived as an indication of conservation for this sample, hence the preference for homogeneity among the married respondents.

The results also indicate that with respect to cross-ethnic entertainment, religiosity is a significant factor for the Ibos but not that much for the Yorubas. While the results are not consistent for all measures, the general trend of the findings seem to indicate greater levels of cross-ethnic tolerance on the part of the Ibos. This diverges from the traditionally held view of Ibo being more exclusive and ethnic bound. However, it should also be noted that the Ibo responses captured in this study represent the view of those from an urban setting that is set against norms and folkways that mark a deep correspondence with Ibo norm patterns. The ethnic composition of Lagos is such that Yorubas are in majority.

However, how much one is able to attribute this differential in cross-ethnic tolerance between these two groups to these relative members in the population will depend among other things upon parallel studies in a non-urban setting with a more representative Ibo presence.

CONCLUSION

In this study, the findings presented here, though limited represent an important addition to the literature on ethnic relationship in Nigeria. It should be borne in mind that the database from which this study derives was not primarily intended to study ethnic interaction. As such the variables for this study are limited. However, we contend that given the changing nature of the ethno-political situation in Nigeria, especially the country’s return to democratic form of government after decades of military rule, the study of ethnic relationships would continue to be a fruitful area for social investigation. The findings here should be an important point of reference for such endeavors.