The Challenge

Around the world, governments encourage long-term savings as a cornerstone of financial security, yet many people lack the financial knowledge or tools to make informed decisions about their future.¹ A common problem is that individuals misjudge how much they will receive in retirement—either overestimating and saving too little, or underestimating and making unnecessary sacrifices in the present.² In Chile, where the pension system is based on mandatory contributions from formal employment and voluntary contributions to individual accounts, these challenges are pronounced. 

A national survey found that fewer than 10 percent of adults take active steps to plan for retirement, 82 percent do not know how their pensions will be calculated, and more than half of those who think they do know are mistaken.³ These gaps in understanding suggest that personalized, accessible information on expected pension benefits could help people make more accurate and proactive savings decisions. However, global evidence on the effectiveness of such interventions remains limited.

The Evaluation

In collaboration with IPA and the government pension supervisory board (Superintendencia de Pensiones), researchers conducted a randomized evaluation to measure whether personalized information increased payments to retirement savings among low- and middle-income workers. The intervention involved 2,500 low- and middle-income workers across the capital, Santiago, and its surrounding area.

All participants received information about the main ways to increase savings in their self-funded accounts: through increasing mandatory savings, increasing voluntary contributions, and delaying retirement. This information was available on online self-service modules in areas where social payments and services were available (Chile Atiende).

The individuals were randomly divided into the following groups:

• Personalized information: Individuals received personalized dollar-amount projections of their pension annuity payouts, plus forecasts of the difference in the payouts based on any of the three above savings and retirement actions, using their data in the system.
• General information (Comparison): Individuals received the percentage effect that any of the three savings and retirement actions is likely to have on average on their pensions.

Results

Providing personalized pension projections increased voluntary retirement savings. Eight months after the intervention, voluntary contributions rose by 10 percent on average and the share making voluntary contributions increased by 0.5-1 percentage point—a 30 percent rise.  However, the short-term increase in savings was too small and temporary to meaningfully increase future pension payments.

An analysis of the variation of impact suggests that the personalization allowed individuals to update their beliefs and not simply in making individuals think about pension savings more saliently. The increase in voluntary savings is concentrated amongst individuals who had previously overestimated their expected pension since this was the easiest way to increase pension savings. On the other hand, individuals who underestimated their pension decreased their mandatory savings (implying lower labor supply, lower formal employment, or lower taxable income) since the low initial take-up of voluntary savings implies that few can decrease their savings through this mechanism.

These results suggest that while personalized information can prompt some individuals to save more, it may also lead others to reduce their work or contributions if they learn their expected pensions are higher than anticipated. For policymakers, this highlights the importance of anticipating and addressing such offsetting behaviors when designing information-based interventions to boost long-term savings.  Results also emphasize that belief updating may be needed to be complemented with reminders and nudges if the impact is to be sustained over time.

Sources

^1. Gallego-Losada, Rocío, Antonio Montero-Navarro, José-Luis Rodríguez-Sánchez, and Thais González-Torres. "Retirement planning and financial literacy, at the crossroads. A bibliometric analysis." Finance Research Letters 44 (2022): 102109.

^2. Benartzi, Shlomo, and Richard Thaler. 2007. "Heuristics and Biases in Retirement Savings Behavior." Journal of Economic Perspectives 21 (3): 81–104.

^3. 2009 Encuesta de Proteccion Social (EPS). https://datos.gob.cl/uploads/recursos/EPS%202009.pdf

Implementing Partners