Document Type


Publication Date



teacher pensions


It is well-known that public pension plans exhibit substantial cross-subsidies, both within cohorts, e.g. from early leavers to those who retire at the “sweet spot”, and across cohorts, through unfunded liabilities. However, the cross-subsidies within and across cohorts have never been provided in an integrated format. This paper provides such a framework, based on the gaps between normal cost rates for individuals and the uniform contribution rates for the cohort. Since the unfunded liabilities and associated cross-subsidies across cohorts derive from overly optimistic actuarial assumptions, we focus on the historically most important such assumption, the rate of return. We present two main findings. First, an overly optimistic assumed return understates the degree of redistribution within the cohort. Second, persisting with an overly optimistic assumed return leads to steady-state contribution rates that exceed the true normal cost (let alone the low-balled rate), i.e. cross-subsidies from the current cohort to past cohorts. Using the case of California, we show how that negative cross-subsidy can easily swamp all positive cross-subsidies within the cohort, as contributions exceed the value of benefits received by even the most favored individuals – those who retire at the “sweet spot.”