Simulating the impact of the pension bonus on the financial implications of working longer
In 2005, the federal government presented the ‘Generation Pact’, containing a number of measures designed to strengthen the financial sustainability of the Belgian system of social security in the light of demographic ageing. One of these measures, the introduction of a pension bonus, is designed to encourage older workers to postpone retirement. This working paper discusses the effect of this bonus on the financial consequences of retirement simulated for four fictitious older workers, representing various types of workers.
A pension bonus is introduced in the first-pillar old-age pension system for private sector employees and that of the self-employed, granting them an additional pension benefit of EUR 2 per working day that they continue to work after reaching the age of 62. The minimum benefit itself is not increased by the bonus.
The option-value approach to the retirement decision assumes that rational individuals look beyond what benefit they will get immediately after retirement, but instead consider their full stream of potential discounted future earnings and pension benefits. The individual considers the gains and losses pertaining to every year that he or she could retire. The Micro Economic Pension model MEP simulates these gains and losses by applying pension rules and fiscal regulations to a fictitious individual representing a male or female white- or blue-collar employee. It allows for the simulation of specific changes of these rules on pension wealth, which is the sum of total discounted future pension benefits. The results indicate whether or not these changes will encourage older workers to postpone retirement.
The introduction of the bonus changes the results from the retirement age of 63 onwards. A first and obvious result is that the replacement rate of older workers gradually increases as a result of the bonus. Without the introduction of the bonus, there is a positive balance of gains (extra earnings) and losses (a loss in pension wealth) associated with postponing retirement by one year. The introduction of the bonus reinforces this positive balance. Finally, the implicit tax on working longer, which is the ratio of the loss in pension wealth over the extra earnings, decreases considerably. A first conclusion, therefore, is that the introduction of the pension bonus might effectively decrease the implicit tax on working longer.
However, the paper also suggests that there are important differences between the fictitious individuals representing various types of workers. The proportional effect of the pension bonus is strongest for the female white-collar worker, followed by the male blue-collar worker. Finally, the effect is the smallest for the male white-collar worker. Overall, it seems that the decreasing effect of the introduction of the pension bonus on the implicit cost of retirement of white-collar women is higher than that for men. This can be explained by the higher life expectancy of women relative to men. The greater effect of the pension bonus for a typical male blue-collar worker than for his male white-collar colleague can be explained by the fact that the benefit is a lump sum.
The paper describes the simulation results for the female blue-collar worker in more detail, as these are to the largest part determined by a nonlinearity caused by the interaction between the minimum pension benefit and the bonus. It should be remembered that the minimum pension benefit itself is not increased by the bonus. So, any individual whose pension-benefit-plus-bonus is lower than the minimum pension benefit will see his or her pension benefit being scaled upwards to this minimum level, and the effect of the bonus will be absent. The minimum pension benefit introduces a pivotal effect of the bonus. If the pension plus the bonus is below the minimum pension benefit, then the implementation of the bonus will have no effect. If, on the other hand, the pension before the bonus is equal to the minimum pension benefit, the implementation of the bonus will have a full effect on the financial consequences of postponing retirement. In between, there is a small bandwidth where the bonus will have a partial effect. A very small change in income or in bonus can, in this nonlinear case, have a large effect on the financial consequences of retirement, and therefore on the simulation results. This is especially so for female blue-collar workers, as it is obvious to assume that they have the highest probability of finding themselves in this situation, and their high life expectancy increases the effect of these small changes in income or bonus on the results of the simulation.