Abstract: This dissertation begins from a desire to explain situations in which left-wing parties appear to adopt policies that are more typically associated with right-wing thinking. A standard explanation for such behaviour is that relatively weak left-wing parties are drawn to adopt those policies as a way of getting elected — commonly expressed
as convergence on the median voter. The puzzle, however, is that this explanation often seems to fall foul of the empirical reality that left-wing parties adopt these policies when they are relatively strong, not weak. The explanation for this advanced here is that parties, seeking to improve outcomes for their constituencies both now and in the future, often operate in political environments which lead them to assign a high probability that today’s policy choices will not survive the predations of government by opposing parties tomorrow. Where this is the case, there is incentive to pursue policies that are less efficient, but which have inbuilt political defence mechanisms: with the main such mechanism focused upon here being the power of organised public sector labour. The effect of partisanship is, therefore, conditioned by expectations about the future political power of parties. Where left-wing parties expect to be weak, they will tend to adopt the highly statist, bureaucratised, nationalised policies that are traditionally associated with the Left as these will tend to embody large amounts of organised labour that will be a counter to future right-wing governments. Where left-wing parties expect to be strong, the costs associated with such policies come to outweigh the beneﬁts, with the result that they do not need to pursue such ‘left-wing’ policies. These ideas are developed heoretically within an institutionalist framework, yielding a synthesis between the historical and rational choice institutionalisms. Empirically, the theoretical framework is applied to the development of welfare states and to the issue of privatisation of state-owned enterprises.