This paper analyzes macroeconomic policy in a two-party system as a repeated game, where parties have different objectives regarding inflation and unemployment. The model considers two parties, D and R, with differing preferences for economic outcomes. If policies are discretionary, economic cycles align with political cycles, leading to inefficiencies. However, reputational mechanisms and commitments to common policy rules can reduce economic fluctuations. The paper introduces a game-theoretic model similar to Kydland-Prescott (1977) and Barro-Gordon (1983), where wage-setters set nominal wages based on rational expectations. The policymaker faces an incentive to generate unexpected inflation to reduce unemployment, but wage-setters anticipate this and set wages to prevent it. Without binding commitments, the outcome is inefficient, but reputational forces can mitigate this. In a two-party system, parties have different policy goals and incentives. If parties are short-sighted, economic fluctuations tied to the political cycle occur. However, if parties recognize the long-term benefits of cooperation, they may adopt a common policy rule, reducing fluctuations. The paper shows that if binding commitments are unavailable, reputational forces can still improve outcomes. The model uses cost functions to represent the objectives of the two parties. Party D values lower unemployment more than inflation, while Party R values lower inflation more than unemployment. The efficient frontier of the game is derived, showing that cooperation can lead to better outcomes for both parties. The Nash bargaining solution determines the optimal policy, with the more popular party having more influence. The paper also discusses the sustainability of the efficient frontier, showing that cooperation is more likely when parties discount the future less and when the discount factor is higher. The results suggest that reputational mechanisms and repeated interactions can lead to more stable and efficient outcomes. The paper concludes that the two-party system can reduce policy volatility and improve long-term welfare through cooperation and commitment to common policy rules.This paper analyzes macroeconomic policy in a two-party system as a repeated game, where parties have different objectives regarding inflation and unemployment. The model considers two parties, D and R, with differing preferences for economic outcomes. If policies are discretionary, economic cycles align with political cycles, leading to inefficiencies. However, reputational mechanisms and commitments to common policy rules can reduce economic fluctuations. The paper introduces a game-theoretic model similar to Kydland-Prescott (1977) and Barro-Gordon (1983), where wage-setters set nominal wages based on rational expectations. The policymaker faces an incentive to generate unexpected inflation to reduce unemployment, but wage-setters anticipate this and set wages to prevent it. Without binding commitments, the outcome is inefficient, but reputational forces can mitigate this. In a two-party system, parties have different policy goals and incentives. If parties are short-sighted, economic fluctuations tied to the political cycle occur. However, if parties recognize the long-term benefits of cooperation, they may adopt a common policy rule, reducing fluctuations. The paper shows that if binding commitments are unavailable, reputational forces can still improve outcomes. The model uses cost functions to represent the objectives of the two parties. Party D values lower unemployment more than inflation, while Party R values lower inflation more than unemployment. The efficient frontier of the game is derived, showing that cooperation can lead to better outcomes for both parties. The Nash bargaining solution determines the optimal policy, with the more popular party having more influence. The paper also discusses the sustainability of the efficient frontier, showing that cooperation is more likely when parties discount the future less and when the discount factor is higher. The results suggest that reputational mechanisms and repeated interactions can lead to more stable and efficient outcomes. The paper concludes that the two-party system can reduce policy volatility and improve long-term welfare through cooperation and commitment to common policy rules.