The 'new politics of the welfare state,' the term coined by Pierson (1996) to differentiate between the popular politics of welfare expansion and the unpopular politics of retrenchment, emphasizes a number of factors that distinguish countries' capacities to pursue contentious measures and avoid electoral blame. Policy structures, vested interests, and institutions play a prominent role in accounting for cross-national differences in leaders' abilities to diffuse responsibility for divisive initiatives. One important omission from the 'new politics' literature, however, is a discussion of partisan politics. 'Old' conceptualizations of the political right and left are implicitly taken as constants despite radical changes in the governing agenda of many leftist parties over the last decade. Responding to this oversight, Castles (1998) has recently probed the role of parties with respect to aggregate government expenditures, only to concludethat parties do not matter under 'conditions of constraint.' This article contends that parties are relevant to the 'new politics' and that, under specified institutional conditions, their impact is counterintuitive. In some notable cases the left has had more effect inbruising the welfare state than the right. One explanation for these cross-cutting tendencies is that parties not only provide a principal source of political agency, they also serve as strategies, thereby conditioning opportunities for political leadership. By extension, they need to be situatedwithin the 'new politics' constellation of blame-avoidance instruments.