Our investigation of the Senate politics of four major civil rights acts indicates that they did not result from winning coalitions bulldozing helpless minorities, nor did they result from some unpredictable chaotic process. These critical bills were the result of a flexible, multidimensional coalition-building process that proceeded by offering amendments carefully constructed to split off pivotal members of the winning coalition. Ideal point estimates of U.S. senators reveal that this coalitional negotiation process led to outcomes at some distance from the first choice of the winning coalition, testimony to significant compromise, both in early proposals and in refinements. This negotiation process resulted in outcomes apparently constrained by the boundaries of the uncovered set (McKelvey 1986; Miller 1980). “Closing the deal” in the U.S. Senate meant finding an outcome that could withstand robust attacks on pivotal coalition members—and that meant finding an outcome in the uncovered set.