Annual changes in domestic discretionary spending are analyzed to test predictions from three distinct types of theories of U.S. policy-making: (1) preference-driven, or nonpartisan, theories such as the recently developed pivotal politics theory or the better-known median voter theory, (2) theories of majority-party strength within Congress, which highlight the ability of the majority party to set the agenda and craft procedures to achieve noncentrist outcomes, and (3) theories of presidential influence over Congress, which emphasize, among other things, strategic use of information, “going public,” an the veto. Findings are supportive of preference-driven and presidential influence predictions, but do not with majority-party strength predictions.