Aggregate Implications of Defined Benefit and Defined Contribution Systems

Financing retirement benefits is probably the most significant fiscal challenge that governments in industrial economies will be facing in the next few decades. Social security reform has therefore become an important public policy issue for many countries and various reform proposals have been recently put forth. Given the importance of understanding the aggregate and welfare implications of different social security systems existing in the OECD, a number of recent papers have investigated the general equilibrium implications of social security reform. We follow this general equilibrium literature to analyze the aggregate and welfare implications of social security arrangements in the presence of empirically relevant market frictions and individual heterogeneity, taking care to explicitly embed in the model the main institutional, social security arrangements observed in OECD economies. Specifically, we compare the aggregate implications of defined benefit (DB) versus defined contribution (DC) systems and also investigate the economic outcomes from varying the generosity of a particular system. That is, we perform a comparison both between DB and DC systems but also within a particular system. We first broadly describe different social security systems that exist in OECD economies and attempt to classify them into categories with broadly similar institutional features. We then embed aspects of these institutional arrangements in a realistically calibrated general equilibrium life-cycle model to quantify the implications for aggregate saving and capital formation. We find that the insurance provided by a DB system can outweigh the efficiency cost from higher taxes to finance the DB payments. As a result, social welfare is maximized at positive DB provision levels. On the other hand, the fully-funded DC system that taxes an individual and offers the benefits during retirement depending on the interest rate and the individual's contributions, does not improve social (aggregate) welfare for any positive tax rate. There are two main reasons for this surprising result. First, the constraint that forces young workers to save through the DC account distorts the consumption-saving allocation sufficiently to generate consumption profiles for the poor that are substantially different from what they would have preferred in the absence of forced saving. Second, the models generate higher capital accumulation and a lower interest rate implying that saving for retirement (either through the DC or non-DC account) earns a lower rate of return that outweighs the positive effect of higher mean wages in the economy