In the first chapter, Political Business Cycles with Policy Compromise, I consider a dynamic model of political decision-making where policy decisions are the result of bargaining between two political parties. The focus of this paper is on how unobservable actions by parties (e.g., the whip encourages certain voting behavior in private strategy sessions) generate inefficient political business cycles. The paper supposes that parties place a high value on future policy outcomes and considers the set of (constrained) surplus-maximizing equilibria. As the future surplus, or government budget, to be divided amongst the parties is stochastically determined by parties' hidden actions, a moral hazard problem arises. Due to this moral hazard problem, any constrained surplus-maximizing equilibria is necessarily inefficient. Furthermore, if efforts towards cooperation are complementary, constrained surplus-maximizing equilibria generate policy outcomes that exhibit political cycles. This result therefore provides a rationale for political business cycles in an environment where cooperative, patient parties negotiate policy. In the second chapter, Gridlocks, Extreme Policies, and the Proximity of an Upcoming Election, I analyze how the proximity of an upcoming election affects the path of policy proposals before that election. Policy outcomes before an election date depend on the proximity of this election date and on current and discounted expected future political power. When there is a common expectation that policy outcomes immediately after the election will generate high social surplus, phases of legislative gridlock, where agreement is infeasible, will either be nonexistent or occur immediately before the election. When the distribution of political power is highly asymmetric, implemented policies favor the party with higher power. When the distribution of power is fairly symmetric, implemented policies can favor either party and intervals of disagreement, and thus legislative gridlock, often occur more frequently.