Recent studies show that pre-electoral commitments and the ideological distance between parties influence government formation. But do pre-electoral pacts or rejections of party combinations really have an independent impact on the outcome of the government formation game? Which policy areas matter when parties agree to build a coalition? This paper addresses these questions by applying dataset that includes information on preferred/rejected coalition partners and the policy-area specific programmatic heterogeneity of all potential coalitions. The results show that pre-electoral commitments have a significant impact on government formation after controlling for endogeneity problems. There is also evidence that not only diversity in economic issues determines the partisan composition of governments.