This thesis is a comparative investigation of how federal institutions accommodated linguistic and religious identities in India and Pakistan. There are three explanatory variables. The first is the history of self-rule for the principalities within South Asia; tracing continuities in territorial autonomy from the Mughals up to independence. The second is the distribution of linguistic and religious identities within the states of India and Pakistan, both at the provincial and national levels. The third is the articulation of a national identity in India and Pakistan. These explanatory variables are not independent of one another; their interaction accounts for the different strategies adopted by India and Pakistan in the formation and stabilisation of their federations. The differences in federal design are calculated according to a scoring system that measures the degree of consociationalism within the federal plans proposed before independence, and the constitutions created after independence. The state-sponsored national identities are distinguished according to their recognition of identities in the public and private spheres. They are further categorised according to the costs for a non-dominant group of being managed by this strategy. The three explanatory variables explain why linguistically homogeneous states were created in India but not in Pakistan. It is argued that this variable explains the stabilisation or otherwise of their federations. It therefore confirms Wilkinson's rebuttal of Lijphart's claim that India under Nehru was consociational. Unlike Wilkinson, it argues that the degrees of consociationalism that emerged since the formation of the constitution have enhanced federal stabilisation within India. It defines federal stabilisation according to continuity in state borders, the number and type of secessionist movements, but more importantly by correlating the effective number of linguistic groups at state level with the effective number of parties in national elections. It concludes that federal accommodation of linguistic groups in homogeneous provinces has enabled the party system to fractionalise in India and Pakistan; an indication of the security of these groups. Where secessionist movements have existed in India and Pakistan, their emergence is explained by the lack of security for a group - defined on either linguistic or alternative criteria.