When and how do ethnic majoritarian states produce de-majoritarian constitutional changes? Scholars studying ethnic conflict often argue that ethnic majorities are never so ‘self-abnegating’ as to give up power, but, in 2010, the 18th constitutional amendment was passed in Pakistan, with the majority ethnic group’s (Punjabi) support, that institutionalised provincial autonomy to reduce Punjab’s domination of the state. Very little, however, has been written about the political process that produced this amendment. More specifically, the question of why Punjab supported this amendment remains unaddressed. Drawing upon primary data (elite interviews and archives) and following a process tracing methodology, this paper offers a within-case analysis highlighting a causal conjunction of three factors – civil-military institutional tensions, social movements, and a politics of cross-ethnic consensus – underpinned by institutionalised divisions within Punjab that shaped Punjab’s support for the amendment. Beginning in the early 1990s and culminating in the 1999 military coup, Nawaz Sharif’s tussle with the Punjabi-dominated military establishment laid the basis for institutionalised divisions, as the Punjab-based leader – who was allied with the military in the 1980s – challenged the military’s dominance by changing the military-backed, ethnically centralised system. To explain how this happened, I first discuss the politics of the 2006 Charter of Democracy – especially, how (intra-ethnic) civil-military institutional tensions brought together Sharif’s Punjab-based Pakistan Muslim League – Nawaz (PML-N) and the Sindh-based Pakistan Peoples Party (PPP) against the military establishment and in favour of provincial autonomy to create, via a constitutional amendment, non-Punjabi power centres in (ethnic) provinces against the Punjabi dominated military. Secondly, I show how the Punjab-based, PML-N-funded Lawyers’ Movement (2007-09) mobilized Punjab, not merely to restore the deposed Chief Justice of the Supreme Court of Pakistan but mainly to defeat the military regime of Pervez Musharraf and push for provincial autonomy in light of the 2006 Charter. Thirdly, I show how institutionalised divisions within Punjab, in conjunction with civil-military institutional tensions and the Lawyers’ Movement, produced a cross-ethnic, multi-party consensus in the 18th amendment committee. Finally, I compare Pakistan with Sri Lanka to show how the absence of similar institutionalised divisions within the dominant Sinhalese casually contributed to the failure of its various political processes of de-majoritarian constitutional changes. The comparison makes a case for analysing internal political and institutional topographies of ethnic majorities to trace the politics and possibilities of de-majoritarian constitutional changes in ethnically majoritarian states.

Sociological theories of group dynamics offer nuanced and sometimes conflicting explanations for what influences tolerance (or intolerance) toward outgroups, including Allport’s contact hypothesis, Blalock’s theory of minority-group relations, Blau’s structural theory of heterogeneity, and Tajfel and Turner’s social identity theory. Existing research suggests religious identity can have a variable effect on tolerance; communal religious involvement can decrease tolerance, while personal piety may increase tolerance. This paper seeks to test these theories in Lebanon, where religious identity plays an outsized role in politics and many parties have an explicit sectarian identity. Using 2013 data from the Lebanese Association for Democratic Elections, the Arab Barometer, and the World Values Survey, I test the impact of religious identity and electoral district heterogeneity on political tolerance in Lebanon, the most religiously diverse state in the Arab region. Voter registration in Lebanon is mandatory, automatic, and includes confession, which enables me to determine majority or minority religious status within Lebanon’s 25 electoral districts and the country. I develop an index of political tolerance using survey questions from both the Arab Barometer and the World Values Survey and classify respondents as dominant majorities, local minorities, enclaved majorities, or enclaved minorities according to their confession and district. I hypothesize that, 1) in line with Allport’s contact hypothesis, diverse districts will have higher levels of tolerance, 2) enclaved majorities will be the least tolerant of outgroups, and 3) local minorities will be the most tolerant of outgroups. I also consider age during salient political conflicts as a potential mediator. Results will extend sociological theories of group dynamics in an understudied context and provide valuable insights into religion’s impact on tolerance in an area where data on religion is accessible. This research can inform efforts toward conflict resolution in Lebanon and, on a high level, evaluate the impact of religious salience on political tolerance.