Submission 198
“Who saves for retirement.” A Survey experiment on determinants, preferences and saving behaviour.
PS2-G04-04
Presented by: Helia Marreiros
Saving adequately for retirement to face later-in-life expenses has become increasingly crucial, as life expectancy increases and birth rates and replacement rates decrease. This paper investigates savings habits and behaviours to identify the main determining factors for retirement savings decisions, in multiple dimensions, including time preferences, financial literacy, risk aversion, and other influencing factors like tax incentives, financial distress, and future expectations on retirement rate and life expectancy, controlling for the individual and socio-economic characteristics. Moreover, the paper examines the impact of salience of information on individuals' intention to save for retirement.
To this aim, we conducted a survey experiment administered to a representative sample of Portuguese households. Using a between-subjects design, participants were randomly assigned to one of three treatment groups, differentiated solely by the type of information presented to them prior to being queried about their retirement savings intentions. In the baseline or neutral treatment, participants received extracts from a published studied on a non-related saving matter. In the first or negative treatment, participants received study extracts highlighting the reduction in replacement rates, i.e., the pension value relative to the last salary in the near future and the second or positive treatment, participants received study extracts about the expected increase in productivity, which will be reflected in an increase of wages.
Additionally, our survey explores types of pension plans, public, employer and professional and future expectations, such as the probability of living until 100 years old, or the percentage of income they expect to receive after retirement.
Our results indicate that most households save regularly (67%), from which two-thirds also save for retirement. The probability of saving for retirement is higher for man, increases with age, income level, tenure, education, financial literacy and house ownership. The results also indicate that patient individuals, with preferences for higher future rewards, tend to save more for retirement.
The experimental results on the effects of salience of information indicate that: 1) the participants who save for retirement and were randomly allocated to the negative treatment are 14% more likely to raise contributions to their Retirement Savings Plan, in comparison to those allocated to the other treatments and 2) the participants who do not save and were randomly allocated to the positive treatment are 14% more likely to obtain information or to subscribe a Retirement Savings Plan, in comparison to those allocated to the Neutral treatment.
Given the critical importance of saving for retirement, the low levels of retirement savings observed in many OECD countries are a major concern for policymakers. While tax benefits have been effective in incentivizing savings, highlighting information through credible studies that emphasize the need to save for satisfactory standard of living may be a complementary strategy to encourage more inattentive individuals to start saving.
It is of crucial importance for policymakers and stakeholders to understand the relationship between time preferences, risk levels, behavioral biases, and savings for retirement to (1) predict savings behavior and (2) design products and policies for the different types of individuals.
To this aim, we conducted a survey experiment administered to a representative sample of Portuguese households. Using a between-subjects design, participants were randomly assigned to one of three treatment groups, differentiated solely by the type of information presented to them prior to being queried about their retirement savings intentions. In the baseline or neutral treatment, participants received extracts from a published studied on a non-related saving matter. In the first or negative treatment, participants received study extracts highlighting the reduction in replacement rates, i.e., the pension value relative to the last salary in the near future and the second or positive treatment, participants received study extracts about the expected increase in productivity, which will be reflected in an increase of wages.
Additionally, our survey explores types of pension plans, public, employer and professional and future expectations, such as the probability of living until 100 years old, or the percentage of income they expect to receive after retirement.
Our results indicate that most households save regularly (67%), from which two-thirds also save for retirement. The probability of saving for retirement is higher for man, increases with age, income level, tenure, education, financial literacy and house ownership. The results also indicate that patient individuals, with preferences for higher future rewards, tend to save more for retirement.
The experimental results on the effects of salience of information indicate that: 1) the participants who save for retirement and were randomly allocated to the negative treatment are 14% more likely to raise contributions to their Retirement Savings Plan, in comparison to those allocated to the other treatments and 2) the participants who do not save and were randomly allocated to the positive treatment are 14% more likely to obtain information or to subscribe a Retirement Savings Plan, in comparison to those allocated to the Neutral treatment.
Given the critical importance of saving for retirement, the low levels of retirement savings observed in many OECD countries are a major concern for policymakers. While tax benefits have been effective in incentivizing savings, highlighting information through credible studies that emphasize the need to save for satisfactory standard of living may be a complementary strategy to encourage more inattentive individuals to start saving.
It is of crucial importance for policymakers and stakeholders to understand the relationship between time preferences, risk levels, behavioral biases, and savings for retirement to (1) predict savings behavior and (2) design products and policies for the different types of individuals.