Photo by Greta Punch on Unsplash

Many social security systems across the world link pension benefits to earnings. The rules vary significantly across countries and pension schemes, ranging from purely unfunded (Pay-As-You-Go) to fully funded. In many countries, means-testing has been used as a simple device to provide adequate insurance to low-income households in a budget-neutral way. The idea of means-testing is that a fiscally neutral reallocation of benefits targets poorer households and ensures an adequate and more equitable post-retirement income without compromising any of the core insurance properties these programs have for high-income households.

Unfortunately, means-testing does not come without a cost because it is very likely to erode the self-financing potential of the system by discouraging individual savings and undermining work effort. This, in turn, may increase the long-term fiscal cost and effectively offset one of the main motivations behind means-testing, namely, maintaining fiscal neutrality.

Recent research by behavioural economists has shown that people are often unable to make rational decisions about how much to save. It is important to understand how means-testing retirement benefits is affected by these behavioural biases.

Self-control problem: temptation to consume now versus long-run interest

In our recent paper, we explore quantitatively the welfare consequences of introducing means-testing of pensions as an alternative to an earnings-dependent and progressive Pay-As-You-Go program when the population has self-control preferences.

We conduct our analysis in two stages involving two economies that only differ in agents’ preference specifications, but are otherwise identical. In the first economy, all individuals have standard preferences; while in the second economy, individuals have self‐control preferences. Our aim here is to compare the aggregate and welfare consequences of the replacement of the Pay-As-You-Go social security system with a means-tested old age pension programme, without altering the expected present value cost of the former scheme.

To capture our agents’ temptation towards current consumption, our model economies use self-control preferences. The key theme here is that self-control preferences assume that agents maximise a utility function that is a ‘compromise’ between the standard utility (or ‘commitment’ utility) and a ‘temptation’ utility. The conflicting ways in which agents derive utility in this setting are the device through which the trade-off between the temptation to consume on the one hand and the long-run self-interest of the agent on the other is captured.

Our research is novel for two reasons. First, it enhances the scope of studies on means-testing by taking self-control issues into consideration. This allows us to analyse the possible interaction between means-testing and self-control issues, and to explore the possibility to identify separately the disincentives to savings that are due to either of those.

Second, in contrast to earlier studies on means-testing, our paper compares means-tested programs in two different settings. In the first setting, we keep the expected present value cost of all the programmes including the Pay-As-You-Go program constant. This allows us to isolate the implications of the phase-out rate. Later, we also look at the implications of the means-testing, varying the benefit rate while keeping the minimum guaranteed benefit fixed, that is, varying the expected present value cost.

When individuals have self-control problems, they would like to avoid high temptation

Our results show that individuals’ self-control problems would have a role in determining the welfare ranking of the programs.

When the cost of the system fixed, lower taper rates do not generate additional tax distortions. If individuals have standard preferences, then a benefit reduction rate of 80 per cent generates the highest welfare among the means-tested programs by improving the distribution of income. Interestingly, when individuals have self-control preferences, a zero per cent benefit reduction rate generates the highest welfare among means-tested programs.

When individuals have self-control preferences, they would like to avoid high temptation. Higher taper rates enlarge the choice sets of certain individuals and increase their self-control costs. A zero per cent benefit reduction rate provides a lower amount of benefits to all individuals without increasing certain individuals’ self-control costs. In other words, when individuals have self-control preferences, the self-control cost-reducing benefit resulting from a zero per cent benefit reduction rate exceeds the re-distributional benefits of the higher taper rates.

When the cost of the program is allowed to vary, lower taper rates come up with huge tax distortions, and hence, in both cases, means-tested programmes with 100 per cent benefit reduction rates maximise the welfare. In this case, all means-tested programs generate higher welfare benefits than the Pay-As-You-Go pension program. When individuals have self-control preferences, higher benefit reduction rates generate larger welfare gain. Since the cost of the system varies, even individuals with self-control preferences prefer the system with low tax distortions.


Further reading

Kumru, CS, Piggott, J & Thanopoulos, AC 2019, ‘A note on means testing and tempatation’, Economic Papers, vol. 38, no. 3, pp. 247-260.

Leave a comment

Your email address will not be published. Required fields are marked *