Pension design is a challenge for many countries. Additionally, many people make bad choices, not saving enough, making bad investment decisions, paying high administrative charges, and/or retiring too soon. Nicholas Barr writes that exposure to risk must decline with age, changes must be phased in gradually, and there must be no shocks for workers close to retirement. He recommends offering workers a choice architecture with fewer options. His research has shaped Sweden and Finland’s initiatives to strengthen long-term financial stability in their pension systems.
|Impact Case -- Research Excellence Framework (REF)|
What was the problem?
Pension reform needs to take account of imperfect information, non-rational behaviour, and incomplete markets.
Longer life expectancies in developed countries create pressures for policymakers seeking to manage pension systems in ways that are sustainable and fair within and across generations. Spending on Sweden’s 2.25 million pensioners absorbed 7.2 per cent of GDP (2017 figures). In 2016, Finland spent 13.4 per cent of GDP on its 1.6 million pensioners.
Given the number of people and the costs involved, it is important that pensions adjust to changing economic and demographic circumstances. The design of a country’s pension system, including benefit levels, pension age, and incentives that influence the choices individuals face when planning for their retirement, matter both for safeguarding old-age security and to protect the long-term financial stability of the system.
What did we do?
In a series of books and articles, Professor Peter Diamond (MIT, Nobel Laureate) and I set out the analytics of pension finance and their implications for pension design, drawing on developments in economic theory and demographics. A core element was to frame analysis in second-best terms. Simple theory assumes that individuals make optimal choices, and that labour markets, savings institutions, and insurance markets exist and function ideally. Formulating policy within that first-best framework is a useful analytical benchmark but a bad guide to pension design. Instead, analysis needs to take account of market imperfections such as imperfect information, non-rational behaviour, and incomplete markets.
These explorations yielded a series of conclusions about pension design. Firstly, longer healthy lives are good news but impose strains on pension finance. If benefit levels are maintained, the stress is on the financial sustainability of the system. If, instead, benefits adjust to available finance (as in Sweden), the stress manifests itself through lower pensions. In both cases, an implication is that pension age needs to adjust to changes in life expectancy. Since pensions exist to help people to plan over the life course, and since older people have less time to adjust to changes and, in the case of pensioners, fewer ways of adjusting, an important principle is that exposure to risk should decline with age. Thus a further implication is that changes should be phased in gradually, with no shocks for workers close to retirement.
A second set of implications draws on the findings of behavioural economics. In particular, faced with the need to make decisions about complex financial products, many people make bad choices: for example, they do not save enough, they make bad investment decisions, pay high administrative charges and/or retire too soon. A central – and seemingly paradoxical – conclusion is that pension design is improved by offering workers a choice architecture with fewer options.
Implemented with cross-party support, Sweden’s strategic reform of its pension system in 1998 became an exemplar for other countries and has been emulated and widely studied. The system proved robust in the face of the 2008 economic crisis. In 2012, the pension authorities in Sweden and, separately, Finland invited me to produce an assessment of their respective pension systems, not as a crisis response but as prudent in-flight adjustment. The resulting reports applied the analytical principles to the particulars of Sweden and Finland.
Following my evaluations, Sweden and Finland both introduced changes to their pension systems.
In Sweden, the country introduced two reforms, the first significant adjustments since 1998. Firstly, in 2017, the Pension Group, the long-established cross-party body that acts as “guardian” of the pension system, agreed to raise retirement ages to reflect rising life expectancy. Earliest and normal pension ages will rise to 63 and 66 respectively in 2023. From 2026, these ages will be based on a formula relating pension age to the remaining life expectancy of a worker’s birth cohort at the time they reach earliest pension age.
Secondly, I had concluded that Sweden’s Premium Pension (in which workers choose from over 800 private pension providers) offers too much choice, a point I repeated at a seminar hosted by the Minister of Health and Social Affairs at the Swedish Parliament in December 2017, speaking alongside Richard Thaler, the 2017 Nobel Laureate. In 2018, the Pension Group established a working group to make proposals for reforming this element, specifically reducing the number of funds and replacing the previous system (which any provider could enter) with a procured platform with a simpler choice architecture.
In November 2018, at their request, I met the chair and members of the working group to discuss details of this reform, and over the course of 2019, implementation was discussed by the Pension Group, parliamentarians, and officials, during which time I continued to comment on proposals. In December 2019, the working group’s report was submitted to the Minister of Social Security, with recommendations to establish a new agency to procure pension funds, with the three-fold task of reducing the number of funds, simplifying and improving the choice architecture for workers, and strengthening quality assurance. Peter Diamond and I were invited to comment on the proposals, in response to which a senior official in Sweden’s Social Insurance Division wrote that, “I have been using a lot of yours and Peter´s brilliant quotes from your response, which has considerably increased the quality of the arguments in the Draft Bill”. The resulting legislation passed in early 2022.
In the case of Finland, following my recommendations, the government entered extensive negotiations about pension reform with representatives of employers and workers. In 2014, an agreement was reached and legislation passed to raise the general retirement age by three months per year until it reaches 65 years. The brochure explaining these reforms stated that:
“The main objective of the reform is to encourage people to work longer. Extended working lives ensure adequate pensions and pension financing and intergenerational fairness. Although working lives are extended, the time spent in retirement will also grow as people live longer.”
The reforms contribute to that objective in several ways. Gradually raising pension age protects the ability of the system to continue to pay adequate benefits. The reforms also adopted my proposal to allow for a partial pension deferral, so pensioners can draw part of their pension while still working, with the deferred element continuing to grow. Introducing this option facilitates longer working life, with benefits both to the economy and individual well-being.
The body of research by Peter Diamond and me has had international influence beyond Sweden and Finland. I have advised on pension reform in several countries, including as a member of Chile’s Presidential Advisory Commission on the Pension System and the Network for Pensions in Latin America and the Caribbean; and Peter Diamond and I have provided evidence to an Australian inquiry into the design of default pension arrangements for workers.
- This blog post appeared originally as an LSE Research Excellence Framework (REF) impact case study.
- The post represents the views of its author(s), not the position of LSE Business Review or the London School of Economics.
- Featured image by Rod Long on Unsplash
- When you leave a comment, you’re agreeing to our Comment Policy.