This dissertation consists of 3 chapters that study topics in the politics of bureaucracy, institutions and political survival.Chapter 1 studies why governments sometimes do not develop meritocratic bureaucracy facing electoral competition. Meritocracy improves the efficiency of public goods provision, which aids governments in elections. However meritocracy hinders the government's ability to conduct targeted redistribution and patronage. Hence the development of meritocracy is affected by the salience of public goods provision versus redistribution. Moreover, the bureaucracy is a long-lasting institution, where earlier decisions constraints future governments' ability to make adjustments. Hence incumbents can strategically ``sabotage" the efficiency of the bureaucracy by excluding talented bureaucrats from their own groups. When redistribution is moderately salient, challengers cannot commit to hire these talented bureaucrats in the future due to redistribution concerns. But the incumbent can commit to such hiring. This gives the incumbent an electoral advantage at the expense of not developing meritocracy and lower bureaucratic efficiency. Chapter 2 studies how governments and politicians can ensure their survival by choosing policies and institutions that endogenously affect their challengers' payoffs from rebellions. Rebellions often require the coordination of multiple challengers. But a conflict of interest often exists among challengers regarding their post-rebellion payoff. The governments can use policies and institutions to affect such payoffs to hinder coordination among challengers and ensure survival. Specifically, the government should target the weaker challengers. Over time the government should begin an institution path with an institution that puts the weaker challengers at a disadvantage, and then transit to an institution that makes them more powerful. The transition time point cannot be too early so that the weaker challengers are easily satiated, or too late such that the transition is not enough to compensate them. The government's ability to commit to such an institution path is key for its survival. Chapter 3 studies why governments sometimes do not regulate long-term corruption despite its negative effect on social welfare. Long-term corruption allows bureaucrats to capitalize on current favorable environment for corruption to guarantee future benefits, even though the future environment becomes less desirable for corruption. The increased corruption reduces social welfare. Hence ex ante it is optimal to institute policies that prevent them from existing. However corruption also creates distributional effects by altering who has access to public goods and services. If corruption directs public goods and services to those with higher valuations, long term corruption may increase social welfare by increasing the efficiency of public goods distribution. When positive distributional effects are realized, politicians may find it interim optimal to allow long term corruption relationships to exist. This creates a time inconsistency problem.