The Swedish pension reform and its aftermath

Increases in life expectancy, low birth rates, low economic growth and volatile financial markets have turned pension policy into a prominent public issue in many industrialized countries. In recent years, all OECD countries have introduced pension reforms to ensure the sustainability and, to a lesser extent, preserve the generosity of earlier commitments.1

Canada has not escaped these trends. Despite the 1997 reform that restored financial credibility to the Canada/Quebec Pension Plan (C/QPP) system by nearly doubling the required payroll tax rate, ongoing pension discussions have been occurring at both federal and provincial levels to address current shortcomings. The declining coverage of occupational pensions in the private sector is the primary concern. This has led to numerous debates on whether the Canada Pension Plan ought to be expanded or whether alternative solutions, such as the creation of a new private savings vehicle, should be implemented.

To cast light on these questions, I look at how Sweden has managed to reform its pension system in the past 20 years.2 In the field of social policy, many comparative analyses use Sweden as an explicit or de facto benchmark. The latest Swedish pension reform has featured prominently in numerous comparative social policy studies.3 It was also quickly embraced by the World Bank, which facilitated its diffusion across numerous countries such as Poland and Latvia.4

A social achievement

The introduction of a mandatory public, earnings-related pension scheme (ATP) in Sweden in the 1960s came after a lengthy and bitter political battle, which included an inconclusive national referendum that resolved nothing since citizens had three options. Two general elections followed. Eventually, one vote in Parliament, by a Liberal Party maverick from a blue-collar district in Gothenburg, tipped the balance in favour of the Social Democratic proposal.5 Adoption of the ATP came to represent the pinnacle of social achievement for many Swedish social democrats.

Combined with a universal basic pension, the mandatory ATP scheme produced a highly generous pension system. As a result, Sweden ranked at or near the top in the generosity of its pension system in a number of comparative analyses.6

The path to reform

However, actuarial reports conducted by the Pension Agency in the 1980s demonstrated that the system needed to be reformed.7 On one hand, with good economic growth, the Swedish system would have become a large flat pension benefit because the ceiling was indexed to wages. Thus any real gains in wages would result in most Swedes receiving the maximum pension. On the other hand, economic decline would require substantial increases in the contribution rates to sustain the generosity of the system. In one scenario, contributions rates above 30 per cent would be necessary.8

Following the 1991 election, which returned the centre-right coalition to power for the first time in nearly a decade, the minister responsible for pensions instituted a working group on pensions with representatives of all political parties in Parliament. At the time, Sweden was undergoing a deep economic crisis that led to widespread questioning of the sustainability of its socioeconomic model. Pension reform was low among contemporary concerns. Given that the political parties had not moved significantly from their original 1960s positions on pension reform, a compromise appeared unlikely, especially since most energy was geared toward fostering broad political agreement to enact policy solutions to deal with the economic crisis.

Nevertheless, in 1992 the working group came up with a compromise supported by five major parties (Social Democrats, Christian Democrats, Liberals, Conservatives and Centre). After the Social Democrats regained power in the 1994 election, the working group became an implementation group, and its work drew more public attention and debate among the parties, most notably within the Social Democrats and Conservatives.9 After lengthy negotiations to secure support for the proposed agreement, in 1998 the Riksdag adopted a law creating an entirely new public pension system.

The new system

The new system marked a significant departure from the previous combination of a generous and universal basic pension and the ATP scheme under which a full pension was given to individuals after 30 years of contributions (with the best 15 used to calculate one’s pension). These were some of the most important features of the new system:

  • The retirement age was eliminated de facto with introduction of the lifetime income principle, referred to as the nonfinancial defined contribution (NDC) scheme. The mandatory NDC scheme, with a 16.5 per cent contribution rate, introduced a close relationship between contributions and benefits. Much as in fully funded schemes, individuals receive regular statements providing a notional value of their contribution. This removes the incentive to retire early. It links every additional contribution to an increase in the pension. However, it also greatly diminishes the redistributive role of the pension scheme. To avoid penalizing individuals who withdraw from the labour market temporarily, built-in mechanisms are included to compensate for various leaves (such as child care or military service), with individuals receiving a contribution to their nonfinancial account.
  • A small mandatory contribution (2.5 per cent) to a private pension account was introduced.
  • An innovative indexing structure was created to depoliticize the indexation of pensions. Known as the automatic balancing mechanism, it takes into account fertility rates, life expectancy, economic growth and inflation.
  • A new income-tested guaranteed pension replaced the previous universal basic pension. This surprising element has been criticized by social policy experts, who assert that mandatory universal income programs, such as Sweden had before the reform, are far more successful at reducing income inequality and even combating poverty than means-tested programs that target the poor. However, in this case, it is important to note that it is an income test for pension benefits gathered within the public system. Rather than counting all pensionable income and individual assets, as is the case with most social assistance benefits, the income test is based on earnings accumulated in the NDC scheme.

Interestingly, the adoption of the new pension system differed in important ways from that in the 1960s. First, unlike the ATP scheme, the new pension system was devised and implemented with the support of five major political parties in the Riksdag. Also, rather than using traditional parliamentary commissions where experts play a prominent role, these five-party committees included only a few experts; politicians played a more substantive role in developing the new pension system.

Second, and surprisingly, the influence of the trade unions and employers was minimal. They were not involved directly in the negotiations. This seriously limited their ability to influence the content of the pension reform and their input occurred in large part after the original compromise was agreed on.10

Swedes become capitalists

With introduction of the new pension system, Swedes now have a choice of pension fund in which to place a portion of their contributions. This represents a longstanding demand of the right-wing parties to stimulate private savings and accentuate the private sector within the pension system. A controversial element concerns the consequences of making or not making a selection. If individuals choose not to make a selection, their contributions go automatically into a default fund managed by state authorities (akin to the Canada Pension Plan Investment Board). However, once an individual has made another choice, he or she cannot return to the default public scheme.

Each year since 2000, Swedes have been receiving distinctive orange envelopes informing them of the progress of their investment options. The private funds – there are more than 750 to choose from – charge very low administrative fees and are registered with and overseen by the state. Currently 57 per cent of Swedes are enrolled in private funds while 43 per cent remain with the default public option. So far, the default option has performed better than the average private fund.

Taking politics out of pensions – well, not quite

The core motivation behind the automatic balancing mechanism was to depoliticize the indexation of pensions while maintaining intergenerational balance. This innovative indexing tool has been the subject of numerous discussions in actuarial circles.11 It has not worked as intended.

As set up in 1998, the index meant that retirees could experience a rise or fall in their pension benefits and current contributors a rise or fall in the accrual value of their contributions. These changes depended on year-to-year changes in the Swedish economy and other elements of the indexing mechanism (such as fertility rate and life expectancy).

The 2008 economic crisis generated intense political pressure to revise the index. If unchanged, it would have resulted in a 3.3 per cent decrease in pension benefits starting in 2010, a few months before an election. Eventually, the five-party working group agreed to revisit the formula, using changes in a three-year moving average instead of annual changes when making adjustments. This modification did not violate the initial agreement’s spirit since no new public money was injected into the system to negate the potential decline in benefits.12

Is Sweden a model for Canada?

The new Swedish pension system is a notable departure from the previous system. Many elements suggest a retrenchment. The current system assumes a career length of 40 years as opposed to 30; it features a mandatory private component; and pension benefits are based on contributions made throughout one’s career. Thus, individuals experiencing lengthy spells of unemployment or underemployment are unlikely to obtain a generous pension in the future. Nonetheless, it is important to stress that the previous system favoured white-collar workers over blue-collar workers. In addition, Swedish fiscal policies continue to redistribute income over citizens’ lifetimes.

Many features of the Swedish systems have been discussed in recent Canadian debates. In Canada the growing number of private-sector workers without pension coverage is alarming. Introduction of a mandatory Registered Retirement Saving Plan has been discussed. Funding such a mandatory plan would require a payroll tax above the Swedish contribution rate of 2.5 per cent – unless such a plan is accompanied by an increase in the generosity of the C/QPP. The QPP faces more financial problems than the CPP. The Régie des Rentes du Québec (RRQ) tested political waters with a consultation document that featured a modified NDC scheme. It faced strong opposition from multiple actors, including unions. The government eventually shied away from these solutions, preferring instead to raise contribution rates and maintain the current system.

Comparative assessments of pension systems have frequently demonstrated that Canada is as successful in alleviating poverty in old age as Scandinavian countries.13 With core features – Old Age Security, the Guaranteed Income Supplement and the C/QPP – firmly in place, our pension system may have more capacity to alleviate poverty among current and future retirees than the current Swedish pension system.


1 Organisation for Economic Co-operation and Development, Pensions at a Glance 2009: Retirement–Income Systems in OECD Countries (Paris: Author, 2009).

2 Many of the elements used for this article originate from a more in-depth analysis featuring lessons from other countries for a new Canadian reform. See Patrik Marier, Improving Canada’s Retirement Saving: Lessons from Abroad, Ideas from Home, IRPP Study No. 9 (Montreal: Institute for Research on Public Policy, 2010). Retrieved from here

3 See for example, Patrik Marier, “Where did the bureaucrats go? Role and influence of the public bureaucracy in the Swedish and French pension reform debate,” Governance, Vol. 18, No. 4 (2005), pp. 521–44; Olli Kangas, Urban Lundberg and Niels Ploug, “Three routes to pension reform: Politics and institutions in reforming pensions in Denmark, Finland and Sweden,” Social Policy & Administration, Vol. 44, No. 3 (2010), pp. 265–84; Kent Weaver, “Paths and forks or chutes and ladders? Negative feedbacks and policy regime change,” Journal of Public Policy, Vol. 30, No. 2 (2010), pp. 137–62.

4 Robert Holzmann and Edward Palmer, Pension Reform: Issues and Prospects for Non-Financial Defined Contribution (NDC) Schemes (Washington: World Bank, 2006).

5 Björn Molin, Tjanstepensionsfrågan (Göteborg, Sweden: Akademiförlaget, 1965).

6 Joakim Palme, Pension Right in Welfare Capitalism: The Development of Old-Age Pensions in 18 OECD Countries 1930 to 1985 (Stockholm: Swedish Institute for Social Research, 1990); Lyle Scruggs and James Allan, “Welfare-state decommodification in 18 OECD countries: A replication and revision,” Journal of European Social Policy, Vol. 16, No. 1 (2006), pp. 55–72.

7 For a detailed account of the reform process, see Patrik Marier, Pension Politics: Consensus and Social Conflict in Ageing Societies (London: Routledge, 2008), chapter 4.

8 Riksförsäkringsverket, “Atp och dess finansiering i det medel – och långsiktiga perspektivet,” RFV Anser 1987:9.

9 Urban Lundberg, Juvelen I Kronan: Socialdemokraterna Och Den Allmänna Pensionen (Stockholm: Hjalmarson & Högberg, 2003); Anders Lindbom, “De borgerliga partierna och pensionsreformen,” in Joakim Palme, ed., Hur Blev Den Stora Kompromissen Möljig? Politiken Bakom Den Svenska Pensionsreformen (Stockholm: Pensionsforum, 2001), pp. 50–87.

10 This diminished influence of labour market partners is in line with recent findings on the decline of corporatism in Sweden. See Torsten Svensson and Per Ola Öberg, “Labour market organisations’ participation in Swedish public policy-making,” Scandinavian Political Studies, Vol. 25, No. 4 (2002), pp. 295–315.

11 For a nontechnical overview of the automatic balance mechanism, see Ole Settergren, The Automatic Balance Mechanism of the Swedish Pension System, Working Papers in Social Insurance – Riksförsäkingsverket 2001:2.

12 See Barry Bosworth and R. Kent Weaver, Social Security on Auto-Pilot: Internaitonal Experience with Automatic Stabilizer Mechanisms, Center for Retirement Research Working Paper 2011–18, pp. 19–20.

13 Régie des Rentes du Québec, Évaluation du système québécois de sécurité financière à la retraite (Quebec City: Author, 2004); Scruggs and Allan, “Welfare-state decommodification.”

Toronto photo courtesy Photo courtesy John Tavares/Flickr