Cui Prodest - the Distributional Impact of the Riester Scheme

During the last two decades, governments have introduced or developed programs of financial aid to foster private saving for retirement purposes. Such programs complement public pay-as-you-go (PAYG) pension systems and are often justified by the increasing difficulty of financing PAYG pension systems because of demographic change.

In this vein Germany introduced the Riester scheme in 2002, which promotes certified financial products for retirement saving by means of  subsidies and tax deductions. After several studies have investigated whether the program fosters uptake in private savings contracts - which it does not -, we investigate its distributional impact.

The question is open-ended, since the Riester scheme incorporates a mix of seemingly progressive and regressive poicies. While the German PAYG system is of the Bismarckian variety and relies on the equivalence principle, the provisions of the Riester scheme entails the following direct subsidies: a basic allowance that is equal for everybody and generous child allowances that favor multi-member households. Yet its opacity is mainly due to the way in which beneficiaries self-select into participation in the scheme and the tax deductions also available to participants. Eligible persons are entitled to the financial aid of the government only if they invest a certain mimimum amount in so-called Riester contracts, the amount being determined as a fixed fraction of a person’s wage income. This boils down to requiring a minimum saving propensity in order to have enough income that can be shifted into a Riester contract. Furthermore, the impact of available tax deductions is by construction most relevant to high-income households. 

Whether the progressive effect from the subsidy provisions outweighs the regressive effect from self-selection and tax deductions is an empirical issue. Determining the distributional impact of the Riester scheme requires a dataset that is representative of the German population and contains enough information to compute the subsidy received by each household. Such a dataset, the Panel of Household Finance (PHF) has recently been released by the German central bank. It is constructed along similar lines as the Survey of Consumer Finances (SCF) in the U.S. and, differently from the German SOEP, reveals the amounts individuals contribute to their Riester contract if they have one. In order to recover the subsidy effectively received by a household, we develop a microsimulation model that allows us to determine whether the government grants a direct subsidy or a tax deduction. We carefully distinguish between tax unit - which is key for determining the type and level of the subsidy - and household - which is the reference unit for our distributional analysis.