The potentially adverse labor market effects of severance pay mandates are a continuing source of policy concern. In a seminal study, Lazear (1990) found that contract avoidance of severance pay firing costs was theoretically simple – a bonding scheme would do – but that empirically the labor market distortions were large. Subsequent empirical work resolved the apparent paradox – firing cost effects are modest even without firm avoidance activities. To explore why that should be so, formal measures of severance-induced firing costs and hiring costs are derived. Firing costs are, it turns out, systematically less than benefit generosity alone would imply. Moreover their interrelationship with hiring costs, often employed in empirical studies as a substitute measure, is complex, with co-movements varying in sign and magnitude across policy parameters and the economic environment. Although the analysis assumes a fixed benefit mandate, the cost measures are easily extended to assess the impact of service-linked severance benefits on age-specific employment levels. The model permits design of a cohort-neutral severance mandate – which is not a flat rate structure.