Tackling excessive pension withdrawals in Peru
While many governments globally have introduced policies aimed at promoting economic growth post-COVID, Peru has taken an unconventional path with its pension system. In some countries such as the UK worries around an ageing population and people dropping out of the labour market post-Covid prompted policies to increase the minimum pension withdrawal age, from 55 to 57 years. In contrast, Peru’s policy since 2020 has allowed all citizens, regardless of age, to dip into their pension funds through seven “extraordinary withdrawal programs”. This policy has led to a significant drop in national pension savings, threatening long-term financial security for millions of people. In order to reduce this risk this blog suggests an intervention inspired by John and Banerjee’s (2021) “Nudge plus” framework[TH1] . Helping to improve appreciation of the cost of these withdrawals, for better- informed decision-making.
Pension choices involve high levels of complexity, making it difficult for individuals to make decisions about their retirement savings. It is no surprise, therefore, that retail financial services have been a key focal point for behavioural scientists. Most significantly, Thaler and Banertzi’s “Save More Tomorrow” (SMarT) program, was designed to encourage individuals to both sign up and increase their savings as time progresses. The goal is to make valuable saving choices easier for people focused on present consumption decisions to reduce the risk of poverty in old age.
The COVID-19 pandemic fostered a shift in approach in order to re-boot economies, after a period in which many spent less and saved more, and many did not return to work at all. A global tendency emerged to use policy to boost consumption spending and prolong working lives (and thereby taxes on income and spend). The UK, for example, announced steps to partly reverse earlier pension rule relaxations, that had allowed people to dip into their pension pots from age 55, with plans to raise this age threshold. Peru has moved in the opposite direction.
Starting in 2020, the Peruvian government has approved six “extraordinary withdrawal programs” for all those with private or defined contribution pension plans, to all age groups. These were intended to mitigate the economic impact of the COVID-19 pandemic, the El Niño phenomenon, and inflationary pressures, offering immediate relief to households. In 2024, the government approved a seventh extraordinary withdrawal programme allowing all affiliates to withdraw up to S/. 20,600 (around $5,500) within 90 days after the program’s publication. Once again, the government has decided to implement policy that directly appeals to present bias, which most savings policies attempt to mitigate
The sum of the six withdrawals before 2024 totalled S/. 87.9 billion, reducing pension savings from 22% to 12% of GDP by December 2022[TH3] . This could have detrimental long-term consequences on Peru’s economy; increasing the risk of poverty in old age. The average pension savings per capita in these funds in 2022 has fallen (in real terms) to levels comparable to those of 2003. Average pension savings per capita halved between 2020 and 2022.
With the new 2024 pension withdrawal , it has been estimated that the number of people with depleted pension savings will increase from 2 million to 6 million – some 71% of affiliates.
These are worrying developments for public policy. Perhaps a behavioural public policy intervention Inspired by John and Banerjee’s (2021) “Nudge plus” policy might be appropriate in achieving a better balance between the competing policy goals – prompting people to think twice before acting, achieving a little thinking space around present bias impulses.
A nudge+ intervention?
Once the individual has selected online the amount of money they wish to withdraw, a simple table could be displayed, with three key pieces of information:
- the current amount of money saved in the individual’s account
- the projected balance of their savings after the withdrawal, and
- An estimate of the future loss incurred by withdrawing the money today (Similar to the regulated projections of retirement pension maturity value). Given that the average annual return on private pensions in Peru is 10% (Asociación de AFP, 2024), the value in this third box might be:
(withdrawal) * [1.10 ^ (years to retirement)]
Example pop-up for a person withdrawing the full sum of money allowed with 20 years to retirement age
| Your pension savings before withdrawing are: | S/. 50,000 |
| Your pension savings loss today: | S/. 20,600 |
| Your estimated savings loss at 65 will be : | S/. 138,586.50 |
The table has, of course, been created to use anchoring, loss framing and simplification:
The first behavioural component of the policy is the anchoring heuristic. The mental shortcut in which we focus on a prominent number (often the first or easiest to recall) when making decisions.Subsequent assessments are based on a comparison to that initial point. A reminder of the amount of money in the account before withdrawal, serves as the reference point for the thought process.
The next two rows in the pop-up table use the tendency towards “loss aversion” through a loss-based framing, rather than one based on the immediate gain. A decision frame can partly influence a person’s perception of a problem by how the problem is presented. Thus, the second row of the table reveals the withdrawal, emphasising the immediate loss in savings. Instead of portraying the withdrawal as a gain in cash, it is framed as a reduction in savings. This approach makes individuals experience the psychological impact of a loss, which is stronger than the psychological impact of a gain. This shift in the reference point makes retrieving the money less appealing.
The final row in the table simplifies the consequences of a withdrawal down to a single number to help individuals recognise the impact of present bias, and ease the cognitive challenge of compound interest and exponential growth.
Of course, such a simple intervention may struggle to accommodate very different financial circumstances and only testing could reveal the effects in the field. Individuals with higher wealth levels tend to have lower time discounting, meaning that they place more value on future money. Poorer individuals, who are often more present-biased, may not respond as strongly to the comparison between present and future gains because of a high discount on future wealth compared to current consumption needs. We may risk exacerbating inequality in old age through these asymmetric effects. When the objective of an action is survival, present bias may not be a cognitive mistake. Reminders of reduced savings and future losses can evoke and enhance feelings of shame in such situations, impacting the wellbeing of those affected.
As an intervention designed as no more than a “nudge+” using information provision, it is arguable that this ranks as nothing more than “soft paternalism” by the state. Nonetheless, it does propose another unavoidable step in the decision process, with some small impact on individual autonomy and presenting some risk of the reactance that accompanies paternalism. The current policy in Peru does, of course, carry significant risks for future wellbeing, and a small step towards a good individual appreciation of decision consequences may also help balance the power of marketing in favour of spending over saving.