Hur har reformerna av trygghetssystemen i Sverige påverkat välfärden och arbetskraftsdeltagandet bland äldre?
This project studies the economic effects of changes in the income security programs for older workers at the Swedish labor market. There is a somewhat contradictory development for the cohorts approaching retirement age in Sweden. On the one hand, these cohorts are more educated, healthier and remain at the labor market at increasingly older ages; on the other hand, there is a higher degree of economic inequality and strong political demands for changes in the public pension system through higher base-level benefits.
This research project is divided into four parts. Three of these are evaluations of major historical policy reforms – the introduction of the supplementary pension system (ATP); the series of changes in the eligibility rules of the Swedish Disability Insurance; and the 1998 reform of the Swedish old-age pension system. One of the sub-projects deals with a change in the Swedish population: the increased educational attainments of the birth cohorts approaching retirement. However, to get exogenous variation we will use yet another policy reform: the comprehensive schooling reform implemented in the 1950s and 1960s.
The overall aim of the project is to investigate to what extent the observed increase in labor force participation rates among older worker, changes in population health and economic inequality among the elderly can be linked to changes in the income security programs, and to what extent they can be attributed to changes in the population.
Final report
The effects of the recent changes in Sweden's income security programs on retirement and welfare
Mårten Palme, Department of Economics, Stockholm University
Purpose
Over the last decades, several aspects of retirement behavior in Sweden have changed. Male labor force participation decreased by more than 30 percentage points between the mid-1960s and the late 1990s, from around 85 to 55 percent. Since then, there has been a steady increase in the participation rate to levels only marginally smaller than in the early 1960s. Due to the general trend of female labor force participation, the downturn in the labor force participation rate from the 1960s to the late 1990s has been absent for females although the upsurge in labor force participation of older women since then has been largely parallel to the development for men in this age group.
Over the same period, there has been a steady improvement in health for both males and females, improvements in the work environment, and the population has become more educated. There have also been several policy changes in income taxes and public income security programs that have affected the economic incentives for workers to stay in the labor force. The overall aim of this research project is to explain to what extent the changes in labor force participation among older workers can be attributed to the policy changes that altered the incentives for remaining employed, or whether the development is driven by changes in the labor force, or the labor market in general, such as improvements in the physical work environment, the general health status of the work force and the educational attainments of the workers. Another aim is to shed light not only on the average development but also on income inequality among retirees and the relation to public policy.
Performance
In two projects, we focus on the public pension system and the disability insurance (DI) program and investigate to what extent these public programs have contributed to the development of labor force participation among older workers. In the first paper (Palme and Laun, 2021), we use earnings histories of different hypothetical workers to calculate changes in the economic incentives to stay in the labor force between 1980 and 2015. We calculate combined incentives measures from the old-age pension and DI systems that incorporate both changes in benefit calculations and in the strictness of eligibility rules for DI. In the second paper (Laun and Palme, 2025) we bring the analysis to the data and analyze the actual economic incentives during 1991–2012 facing the birth cohorts born between 1927 and 1950. We give a detailed description of the series of policy changes that have affected the incentives to remain in the labor force for older workers, describe how the economic incentives to stay in the labor force have changed because of these changes, and estimate an econometric model for the impact of financial incentives on retirement choice. These projects are part of the International Social Security (ISS) Project at the National Bureau of Economic Research (NBER). During the project period, we have also contributed to Coile et al. (2024), that summarizes the lessons from this long-term collaboration.
To shed light on income inequality among retirees, Hagen et al. (2022) analyzes the development of various types of income for the full population of retired individuals during 1991–2019. In ongoing work, we build on Laun and Palme (2025) and analyze the impact of the major pension reform on income inequality among retirees. Another ongoing project analyzes the impact of the introduction of the earnings-related ATP pension system in the 1960s. Preliminary results indicate that the reform substantially improved the economic situation of retired blue-collar workers compared to other groups who already had generous occupational pension schemes prior to the reform. Although these two projects have not yet resulted in publications, the results have been presented at research workshops. The analysis of the increased educational attainments of the birth cohorts approaching retirement, exploiting the comprehensive schooling reform in the 1950s and 1960s, has begun but become delayed due to the above commitments. This is one reason why we did not use the full research grant for the project.
Main results
One main takeaway is that public institutions matter, but to different extents. The results by Palme and Laun (2021) show very large effects on economic incentives of the transition rules for the implementation of the pre-reform old-age (ATP) pension scheme in the 1980s. Also changes in the disability insurance program appear important. However, the results show surprisingly small effects of the implementation of the new pension scheme decided in 1998. This means that our results do not lend support to the claim that the pension reform was the main reason behind the marked increase in labor force participation rates among older workers since the late 1990s that coincided with the implementation of the new pension system.
The results in Laun and Palme (2025), that deepens the analysis of the 1998 pension reform and changes to the DI system, confirm the results. The analysis shows that the 1998 pension reform changed the incentives to remain in the labor force ambiguously: although it induced an income effect towards later retirement through lower replacement levels, it also implied a lower price on leaving the labor market under some assumptions. By lowering pension benefits, the reform created an income effect towards later exit from the labor force, and the series of reforms of Sweden’s DI system implying gradually stricter eligibility rules reinforced the trend towards a lower social security wealth. However, the direction of the impact on the price of leaving the labor force highlights the importance of the assumption we make about the relationship between pension claiming age and labor force withdrawal age. Whereas the old pension system had strong actuarial adjustments in relation to the pension claiming age, the adjustment in the new pension relates more heavily to when the individual stops working. Assuming that pension claiming and labor force withdrawal is decided jointly implies a change towards weaker incentives to stay in the labor force in the post-reform pension system, whereas assuming pension claiming fixed at age 65 indicate stronger incentives to stay in the labor force because of the reform. Estimating the impact of the changes in economic incentives on labor force participation of older workers suggests that at most a small part of the increase in labor force participation of the elderly can be attributed to pension reform.
Another main takeaway is that retirees’ income has developed very positively across the income distribution over the last decades, but at the cost of increasing income inequality among retirees. Hagen et al. (2022) shows that real disposable incomes among retirees have increased significantly across the entire distribution during 1991–2019. Even relative to income developments for the working age population, pensioners’ incomes have risen throughout the distribution. Overall, however, the distribution of disposable incomes among pensioners has become more uneven. This is attributable to the upper part of the income distribution and is primarily due to increased labor income and occupational pensions.
A final takeaway is that in terms of retirees’ incomes in relation to their pre-retirement income, the development over the last decades is more complex. On the one hand, the average replacement rate in terms of disposable income has remained relatively high throughout the period. On the other hand, the average replacement rate in terms of pension income has declined significantly since the early 2000s. The decline in public pension replacement rates has been offset by increases in occupational pensions, labor income, and lower taxes. The declining replacement rates in public pensions coincide with the introduction of the new pension system but also with a drop in the retirement age, which has not kept pace with increases in life expectancy. Also, pre-retirement income levels have changed. In addition, the declining labor force exit through the sickness and disability insurance has increased labor force participation and pre-retirement incomes increased, which also resulted in a decline in pensions relative to pre-retirement incomes.
Communication of results
The work has been presented at several academic workshops, including three workshops within the National Bureau of Economic Research (NBER) International Social Security (ISS) Conference and one workshop within the international research collaboration ”Unequal ageing: life-expectancy, care needs and reforms to the welfare state” (PENSINEQ). The work has also been presented in popular versions at seminars at the Swedish Ministry of Finance, the Swedish National Financial Management Authority and the Swedish Ministry of Health and Social Affairs. The report by Hagen et al. (2022) was presented at a public seminar by the Swedish Fiscal Policy Council. It generated profound attention by the media, as evidenced by 19 newspaper editorials, 10 occurrences in radio or TV, and plenty of newspaper articles. Laun has also participated in several public discussions about the Swedish pension system, building on the combined research within the project, arranged by the Swedish Pensions Agency, the Swedish Government and representatives for the Swedish pensions industry. The results have been published in popular writing such as two reports for the Delegation for older labor under the Swedish government, an editorial in the popular journal Ekonomisk Debatt and a policy report for the Institute for evaluation of labor market and education policy (IFAU).
Mårten Palme, Department of Economics, Stockholm University
Purpose
Over the last decades, several aspects of retirement behavior in Sweden have changed. Male labor force participation decreased by more than 30 percentage points between the mid-1960s and the late 1990s, from around 85 to 55 percent. Since then, there has been a steady increase in the participation rate to levels only marginally smaller than in the early 1960s. Due to the general trend of female labor force participation, the downturn in the labor force participation rate from the 1960s to the late 1990s has been absent for females although the upsurge in labor force participation of older women since then has been largely parallel to the development for men in this age group.
Over the same period, there has been a steady improvement in health for both males and females, improvements in the work environment, and the population has become more educated. There have also been several policy changes in income taxes and public income security programs that have affected the economic incentives for workers to stay in the labor force. The overall aim of this research project is to explain to what extent the changes in labor force participation among older workers can be attributed to the policy changes that altered the incentives for remaining employed, or whether the development is driven by changes in the labor force, or the labor market in general, such as improvements in the physical work environment, the general health status of the work force and the educational attainments of the workers. Another aim is to shed light not only on the average development but also on income inequality among retirees and the relation to public policy.
Performance
In two projects, we focus on the public pension system and the disability insurance (DI) program and investigate to what extent these public programs have contributed to the development of labor force participation among older workers. In the first paper (Palme and Laun, 2021), we use earnings histories of different hypothetical workers to calculate changes in the economic incentives to stay in the labor force between 1980 and 2015. We calculate combined incentives measures from the old-age pension and DI systems that incorporate both changes in benefit calculations and in the strictness of eligibility rules for DI. In the second paper (Laun and Palme, 2025) we bring the analysis to the data and analyze the actual economic incentives during 1991–2012 facing the birth cohorts born between 1927 and 1950. We give a detailed description of the series of policy changes that have affected the incentives to remain in the labor force for older workers, describe how the economic incentives to stay in the labor force have changed because of these changes, and estimate an econometric model for the impact of financial incentives on retirement choice. These projects are part of the International Social Security (ISS) Project at the National Bureau of Economic Research (NBER). During the project period, we have also contributed to Coile et al. (2024), that summarizes the lessons from this long-term collaboration.
To shed light on income inequality among retirees, Hagen et al. (2022) analyzes the development of various types of income for the full population of retired individuals during 1991–2019. In ongoing work, we build on Laun and Palme (2025) and analyze the impact of the major pension reform on income inequality among retirees. Another ongoing project analyzes the impact of the introduction of the earnings-related ATP pension system in the 1960s. Preliminary results indicate that the reform substantially improved the economic situation of retired blue-collar workers compared to other groups who already had generous occupational pension schemes prior to the reform. Although these two projects have not yet resulted in publications, the results have been presented at research workshops. The analysis of the increased educational attainments of the birth cohorts approaching retirement, exploiting the comprehensive schooling reform in the 1950s and 1960s, has begun but become delayed due to the above commitments. This is one reason why we did not use the full research grant for the project.
Main results
One main takeaway is that public institutions matter, but to different extents. The results by Palme and Laun (2021) show very large effects on economic incentives of the transition rules for the implementation of the pre-reform old-age (ATP) pension scheme in the 1980s. Also changes in the disability insurance program appear important. However, the results show surprisingly small effects of the implementation of the new pension scheme decided in 1998. This means that our results do not lend support to the claim that the pension reform was the main reason behind the marked increase in labor force participation rates among older workers since the late 1990s that coincided with the implementation of the new pension system.
The results in Laun and Palme (2025), that deepens the analysis of the 1998 pension reform and changes to the DI system, confirm the results. The analysis shows that the 1998 pension reform changed the incentives to remain in the labor force ambiguously: although it induced an income effect towards later retirement through lower replacement levels, it also implied a lower price on leaving the labor market under some assumptions. By lowering pension benefits, the reform created an income effect towards later exit from the labor force, and the series of reforms of Sweden’s DI system implying gradually stricter eligibility rules reinforced the trend towards a lower social security wealth. However, the direction of the impact on the price of leaving the labor force highlights the importance of the assumption we make about the relationship between pension claiming age and labor force withdrawal age. Whereas the old pension system had strong actuarial adjustments in relation to the pension claiming age, the adjustment in the new pension relates more heavily to when the individual stops working. Assuming that pension claiming and labor force withdrawal is decided jointly implies a change towards weaker incentives to stay in the labor force in the post-reform pension system, whereas assuming pension claiming fixed at age 65 indicate stronger incentives to stay in the labor force because of the reform. Estimating the impact of the changes in economic incentives on labor force participation of older workers suggests that at most a small part of the increase in labor force participation of the elderly can be attributed to pension reform.
Another main takeaway is that retirees’ income has developed very positively across the income distribution over the last decades, but at the cost of increasing income inequality among retirees. Hagen et al. (2022) shows that real disposable incomes among retirees have increased significantly across the entire distribution during 1991–2019. Even relative to income developments for the working age population, pensioners’ incomes have risen throughout the distribution. Overall, however, the distribution of disposable incomes among pensioners has become more uneven. This is attributable to the upper part of the income distribution and is primarily due to increased labor income and occupational pensions.
A final takeaway is that in terms of retirees’ incomes in relation to their pre-retirement income, the development over the last decades is more complex. On the one hand, the average replacement rate in terms of disposable income has remained relatively high throughout the period. On the other hand, the average replacement rate in terms of pension income has declined significantly since the early 2000s. The decline in public pension replacement rates has been offset by increases in occupational pensions, labor income, and lower taxes. The declining replacement rates in public pensions coincide with the introduction of the new pension system but also with a drop in the retirement age, which has not kept pace with increases in life expectancy. Also, pre-retirement income levels have changed. In addition, the declining labor force exit through the sickness and disability insurance has increased labor force participation and pre-retirement incomes increased, which also resulted in a decline in pensions relative to pre-retirement incomes.
Communication of results
The work has been presented at several academic workshops, including three workshops within the National Bureau of Economic Research (NBER) International Social Security (ISS) Conference and one workshop within the international research collaboration ”Unequal ageing: life-expectancy, care needs and reforms to the welfare state” (PENSINEQ). The work has also been presented in popular versions at seminars at the Swedish Ministry of Finance, the Swedish National Financial Management Authority and the Swedish Ministry of Health and Social Affairs. The report by Hagen et al. (2022) was presented at a public seminar by the Swedish Fiscal Policy Council. It generated profound attention by the media, as evidenced by 19 newspaper editorials, 10 occurrences in radio or TV, and plenty of newspaper articles. Laun has also participated in several public discussions about the Swedish pension system, building on the combined research within the project, arranged by the Swedish Pensions Agency, the Swedish Government and representatives for the Swedish pensions industry. The results have been published in popular writing such as two reports for the Delegation for older labor under the Swedish government, an editorial in the popular journal Ekonomisk Debatt and a policy report for the Institute for evaluation of labor market and education policy (IFAU).