No impact---best studies.

With respect to internal explanations, the effects of income inequality and ethnic diversity are presented in Table 3. Models 3.1 and 3.2 indicate that neither directly affects chauvinism. H4 is therefore not supported. The results suggest, however, that both have a negative effect on the national-identification slopes. Contrary to our expectations, countries with higher levels of economic and ethnic division appear to exhibit a weaker relation between national identification and chauvinism. While these findings might seem to contradict H5, the pattern was caused by outliers. After excluding South Africa—the most unequal and ethnic diverse country in our sample—the effect of ethnic diversity is not even of borderline significance. After excluding Chile—the most unequal country in our sample—the interaction effects for economic inequality were also far from significant. The results, therefore, do not support H5.21

<<TABLE 3 OMITTED>>

Conclusions

During the historic phone call between President Obama and Iranian President Sheikh Hasan Rouhani in September 2013, the latter stated that his country's nuclear program ‘represents Iran's national dignity’.22 This declaration reflects the common perception that Iran's nuclear program mobilizes Iranians in support of resisting further national humiliation at the hands of foreigners (Moshirzadeh, 2007). This reflects the important role national feelings play in the contemporary international arena. Evidence from other examples—such as the Israeli-Palestine conflict—indicates that national identity serves as a key factor in conflict resolution. The prominence of national feelings is not limited to the Middle East, their effect on public attitudes towards international issues, and conflicts also being manifest in the West (Billig, 1995; Kinder & Kam, 2010).

It is thus hardly surprising that scholars seeking to develop a better understanding of conflicts adopt a social-psychology perspective, replacing the deterministic view that identification with one's in-group necessarily leads to antagonism towards out-groups with an examination of the broader social context. In line with this approach, the present paper focuses on the way in which political and social contexts encourage chauvinistic views towards the international arena and how they affect the relation between national identification and chauvinism.

Integrating various social and psychological theories, we investigated two external contextual explanations (globalization and conflict) and an internal explanation (social division). Employing cross-national survey data, we examined the relation between national identification and chauvinism across 33 countries. The findings indicate that a positive relationship exists between national identification and chauvinism across most of the countries, although the level differs from country to country. Using a multilevel regression analysis, we tested to see whether globalization, conflict, and social division correlate with this variation. The results indicate that social and political contexts are related to chauvinism and the ways national identification and chauvinism are linked. Although a closer relation exists between national identification and chauvinism in more globalized countries, globalization failed to explain the variation in chauvinism itself. These findings support the notion that globalization highlights the importance of national identity (Calhoun, 2007; Castells, 2011). While those sections of globalized societies that are attached to their country also tend to resist international cooperation and endorse hostile views, the complexity of the phenomenon—as evinced by the divergent findings of previous studies (e.g. Jung, 2008; Norris & Inglehart, 2009)—calls for further research of this interpretation. The fact that the current study is cross-sectional must also be taken into account, the findings adducing the relation but not the causal relations between the variables. In contrast to experimental studies, the present design is similarly limited in its ability to offer a robust control for alternative explanations.

Another external factor found to be relevant—to a certain degree—was conflict. Countries that suffered large numbers of deaths in conflicts and mobilized resources and personnel exhibited higher levels of chauvinism. When other indices for conflict were used, however, these results were not replicated. A possible explanation for this finding lies in the inherent limitation in the way in which conflicts are measured across various countries. Measuring international conflicts is a challenging task (Anderton & Carter, 2011). While the ways of measuring conflict were chosen because they reflect different dimensions of conflict in order to be representative of a wide range of countries, the problem of comparability cannot be ignored. An alternative explanation may derive from the fact that only deaths from conflict and resources/personnel mobilization are sufficiently significant to contribute to chauvinism. The limitations of our measurements of conflict and research design mean that this idea must remain speculative, however. In addition, it is important to emphasize that the sample of countries is also limited as many countries are not involved in conflict and there is also limited variation in the types of conflicts.

Contrary to what the divisionary theory of national mobilization would lead us to expect, neither economic inequality nor ethnic diversity were related to chauvinism or affected the relation between national identification and chauvinism. This finding might also be explained by the limitation of the current research design. The number of countries included in the ISSP 2003 National Identity Module being relatively small and the sample only covering countries with available survey data, the results relate solely to this specific sample of countries. Across another set of countries, social division might play a far more significant role. Another explanation might be the meaning given to national identification and chauvinism across the countries. While evidence exists for the comparability of the scales across most of the countries, the divergent meaning probably attributed to them in Germany, the United States, and Israel might form an additional limitation.