The Payout Decision in the Swedish Occupational Pension System

Johannes Hagen, Researcher at Department of Economics, Uppsala University


podcast (in Swedish)

summary-payout-decision-in-the-swedish-occupational-pension-system_2-1.pdf 223.2 KB PDF

In Sweden, all four major occupational pension plans have introduced so called “fixed-term payouts” as an alternative to life-long annuities. A life annuity guarantees the retiree a constant monthly income for the remainder of his or her life, whereas payments cease after a certain date under the fixed-term payout options. A life annuity could thus be viewed as an insurance for (unexpected) longevity. Because the occupational pension is an important component of the individual’s total pension income, the payout decision is an important determinant
of old-age economic security.

This report uses microdata from the two largest pension companies in Sweden, AMF and Alecta, to study the payout decision at retirement. Data include real payout decisions of about 370,000 individuals who started to collect their pension at some point between 2008 and 2013. This data is merged with national administrative data – from Statistics Sweden – of individual background information, such as labor market history, education level, retirement benefits, and health status.

In the empirical analysis, the population is divided into two groups: those who receive their pension from ITP2 and those who do not. The former group consists of white-collar workers and the latter group is mainly made up of blue-collar workers (SAF-LO) and local government workers (KAP-KL). In the first part of the report, the two groups are studied how they differ in terms of individual characteristics and the economic implications of the payout decision. As expected, the ITP2 group is on average better educated, in better health, and has higher earnings than the SAF-LO/KAP-KL group. Because white-collar workers face no restriction on the fraction of wealth that can be withdrawn during a fixed number of years, the payout decision involves more money relative to total pension income for this group. Under the 5-year payout, total pension income would on average fall by 38 per cent after 5 years. For the SAF-LO/KAP-KL group, the corresponding decline would amount to 12 per cent.

The second part of the report describes how payout choices have evolved over time. The popularity of the fixed-term payouts has increased in all pension plans. Between 2008 and 2013, the fraction of people that chose the life annuity decreased from 81 to 70 per cent. The corresponding decline in SAF-LO and KAP-KL was even larger – about 22 percentage points. About two-thirds of the fixed-term payouts are accounted for by 5-year payouts. However, the 10- and 15-year payoutoptions have grown more in relative terms. Very few people choose the 20-year payout.

In the third part of the paper, the determinants of the payout decision are examined. Attention is paid to both individual-level determinants and institutional regulations that may affect the payout decision. Below follows a brief summary of the main results. An individual’s health condition and life expectancy should influence the annuitization decision because annuities hedge longevity risk. For white-collar workers, clear evidence is found that individuals who have shorter life expectancies and individuals in bad health are more likely to choose the 5-year payout. The sign of the relationship between health and payout length is the same for individuals in the SAF-LO/KAP-KL group, but the magnitude is smaller. Other demographic factors, such as education, marital status and children, only marginally affect the payout decision.

Payout preferences vary significantly by account balance and pre-retirement income. Among white-collar workers, there is a positive relationship between income and pension capital on the one hand, and the probability of choosing a life annuity on the other. Taxes are likely to play an important role here. The negative tax effect on the value of the fixed-term payouts is particularly large for high income individuals with large capital stocks, due to the progressivity of the tax schedule. Among blue-collar workers and local government workers, the
relationship has the opposite sign: those with high incomes and large capital stocks are less likely to annuitize.

The relatively high demand for a life annuity is most likely driven by the fact that individuals are defaulted into it if they take no action before age 65. In the population being studied, 7 out of 10 individuals stick with the default. Survey evidence suggests that many individuals interpret the default provision as implicit advice from the plan sponsor. They might also be uninformed about or even unaware of the imminent payout decision.

Finally, individuals who retire before the normal retirement age of 65 are much more likely to choose one of the fixed-term payouts. Among white-collar workers, the combination of short payout horizons and early retirement seems to be driven by individuals who want to maximize their consumption possibilities during the years of early retirement. Compared to those who claim their pension at age 65, early retirees receive larger pensions, have significantly higher pre-retirement income, and are in better health. In the SAF-LO/KAP-KL group, the differences
between early claimers and those who claim at age 65 are less pronounced, yet point in the same direction. How the payout phase should be designed is a matter of controversy. Fixed-term
payouts allow individuals to invest, leave bequests, or increase consumption during the early years of retirement, but also raise concerns that they may trigger individuals to spend money too rapidly for their own good. This report shows that many individuals do seem to make an informed decision. However, ignorance is widespread and the default effect is large. Future research will have to shed light on the long-term consequences of fixed-term payout options for individuals as well as society. To avoid the problem that some individuals are defaulted into the life annuity against their will, I argue that individuals should be required to make an active choice in order to start receiving the pension.

This report is part of the research program “New Challenges for the Pension System.” The program takes an overall perspective of the pension system and deals with questions that concern both the occupational pension and the state pension.

The report received a great deal of media attention, with headlines including, “Risks with premature retirement”, and “More people take out occupational pension during a limited period.”
The report was presented at an SNS seminar in Stockholm on January 17, 2017 where Staffan Ström, economist at Alecta, and Dan Adolphson, economist at AMF participated.
“Those who choose fixed-term payouts for their occupational pensions are often not aware of the choice they made. After a few years when payouts end, many are surprised”, Staffan Ström economist at Alecta, said.
“Four out of ten are worried that their retirement and private savings will not suffice”, Dan Adolphson, economist at AMF, said.