The quantitative strand of social policy research suffers from a triple deficit: analyses of aggregate expenditure dominate, most of the few studies of replacement rates focus on unemployment or sickness benefits while pensions are excluded, and the interdependence between public and private pension plans is often ignored. This article addresses the said deficits, first, by discussing the pension sectors' theoretical peculiarities and by proposing two hypotheses: one on the role played by political parties in implementing public pension retrenchment, and the second on their role in extending private pension plans. Second, the article presents regression results of public pension replacement rate changes in 18 developed democracies. The findings show considerably smaller cuts to pensions than to unemployment or sickness benefits, and striking differences regarding partisan effects between the sectors. Lastly, the article assesses partisan effects on private pension plans, detecting some rather surprising effects. Most noteworthy is the fact that those parties which reduced public pension generosity during the 1990s (i.e. Social Democrats) cannot claim responsibility for compensating these cuts by eliciting higher private engagement.