Replacement rates in the new Swedish pension system – a Danish perspective

313 Replacement rates in the new Swedish pension system - a Danish perspectiveNFT 4/2006 The editor has asked us to provide a comment on the article” Replacement rates in the new Swedish pension system” by KG Scherman in NFT 2/20061. We are certainly inclined to meet the demand of the editor. However, we must stress that we are not experts in the Swedish pension reform. Moreover, any pen- sion system and its division of responsibility between the tax payers, the government, pri- vate individuals saving for the future and the social partners is the result of a political pro- cess, and, not least, of considerations on in- come distribution. We have no intention – directly or indirectly – to become part of a Swedish political debate and our comments below must be seen in this perspective. Replacement rates in the new Swedish pension system – a Danish perspective by Per Bremer Rasmussen and Peter Skjødt The authors comment on the article in NFT 2/2006 on replacement rates in the Swedish pension system. The authors are not surprised by the financing problems arising in Sweden, and they claim that the organization of the Swedish pension system as a tax financed pay-as-you-go system makes it vulnerable to budgetary considerations, even though the system is a notional defined contribution scheme. In Denmark the responsibility for topping up social pensions lies with the private sector. This gives rise to different risk sharing features than in Sweden. However, we have experience with the es- tablishment of the Danish pension system which gives rise to some qualitative com- ments on the Swedish system. The Swedish challenge It is clear from KG Scherman’s article that the Swedish pension system faces some impor- tant challenges. Like in many other countries, life expectancy is increasing and this puts strain on the pension system – either replace- Per Bremer Rasmussen is Msc Econ and Chief Execu- tive Officer of Forsikring&Pension. Peter Skjødt is Msc Econ and Deputy Chief Executive in Forsikring& Pension. Per Bremer Rasmussen pbr@forsikringenshus.dk Peter Skjødt psk@forsikringenshus.dk 314 Replacement rates in the new Swedish pension system - a Danish perspective ment rates will be significantly lower than projected, or people must spend more years on the labour market, as clearly analyzed in the article. For an outsider, the financing or replace- ment rate issues facing the Swedish pension system are not that surprising. The reform, decided upon in 1994 and phased in over a number of years, basically replaced one pub- licly (taxpayer) financed pay-as-you-go (PAYG) pension system with another public- ly financed PAYG pension system. One of the important changes was that pension rights would no longer be calculated on the basis of the 15 years with the highest wages (out of 30 years) but would rather be based on contribu- tions and with the establishment of a balanc- ing mechanism to secure future pension rights. As KG Scherman states in his article, the political expectation was that the reform would not in itself necessitate a reduction in the future pension level compared to the former system. However, it seems that concern about the ability to service the former ATP pension system with taxpayers money was a clear rationale behind the reform. In other words, there was a need to cut the costs to taxpayers of the former pension system. All else equal, this would imply that acquired pension rights under the new system had to be lower than under the old system. The problems now documented by KG Scherman must be seen against this background. Hence, it is no big surprise that the replace- ment rates and relative pension levels do not reach the target levels which were formulated in 1994. It seems to us that the statement in the bill introducing the system in 1994 – “there are no reasons why the pension levels in general should need to be reduced” (Scher- man in NFT 2/2006, page 101) is more based on wishful thinking than economic analysis. As mentioned, the public Swedish pension system is a tax financed PAYG system, ex- cept for the funded Premium Pension, which plays a minor role in the overall picture. The ability of the Swedish pension system to honour the expectations of the future pension- er’s therefore rests only to a marginal extent on the stipulated contribution rate in the Pre- mium Pension (which is to be held constant), and primarily on the development in the Swed- ish tax base over time, life expectancy and the average time spent on the labor market. According to KG Scherman, the reformed Swedish system was “completely rearranged” compared to conventional PAYG schemes because it is a notional defined contribution (NDC scheme). It is to be financially balanced over time through the buffer mechanism. However, in our view, even an NDC scheme can not be viewed in isolation from the gene- ral public budget. If the general budget is under strain with (structural) deficits increasing, there will be a tendency to look to areas outside the official budget for financing. In respect to the NDC scheme this would probably imply a pressure for lower benefits without lowering contribu- tions, hence alleviating the general budgetary problems. . In Denmark, there have been examples of financing certain public expenditures by way of an “earmarked” tax where the tax rate was to be lowered in case of expenditures being lower than projected. Politically, however, the “surplus” tax proceeds found other expen- ditures to meet and rates were not lowered. In our view, this political problem could be a challenge facing the Swedish NDC scheme together with the longevity issue as described by KG Scherman. The Danish pension system KG Scherman does not comment in detail upon the role played by private (pillar II and III) pension schemes in Sweden. It is our impression, however, that they play a less significant role in Sweden than in Denmark. 315 Replacement rates in the new Swedish pension system - a Danish perspective The Danish pension system relies to a lesser extent than in Sweden on tax financing and more on funded savings. Like in many other OECD countries, the Danish pension system has a multi-pillar structure. Pillar I consists of the social pensions, which are unfunded and financed from general tax revenues (i.e. a PAYG system). At the core of the public pillar, and of the whole system, the social pension scheme pays benefits to people over 65. It consists of two parts: a flat univer- sal pension that is subject to a residency test and proportionality rule as well as an employ- ment earnings test and a supplement that is paid to qualifying people subject to an income test. The public pillar also has a smaller compo- nent that is fully funded, is financed from employee and employer contributions (or the tax payers for unemployed workers and those on parental leave etc.), and operates as a defined contribution plan. This is known as ATP (Labor Market Supplementary Pension). Despite being fully funded and based on indi- vidual accounts, ATP is classified as a first pillar scheme because it was established by law and entails social security features. The second pillar comprises occupational pension plans that are quasi-mandatory and nearly universal. Most have been established by collective agreements between employer organizations and labor unions. They are managed by life insurance companies and multi-employer pension funds as well as – on a small scale – company pension funds and banks. The vast majority of these operate as defined contribution plans. The third pillar comprises voluntary per- sonal pension plans. These are created by life insurance and pension companies as well as banking institutions. The latter are not permit- ted to offer annuity products. In general – but subject to certain regula- tions – premiums paid to pension savings are income tax deductible while benefits are sub- ject to income taxation. Yields on the assets invested are taxed at a flat rate of 15 per cent. This system provides some tax incentive to save for retirement, yet does not leave a door wide open for tax evasion. Coverage of the three pillars is very high. It is universal or nearly universal in the public pillar components, almost 80 percent of wage earners under occupational schemes (outside the mandated supplementary schemes), and 40 percent of wage earners in the third pillar. Overall, more than 90 percent of wage earners participate in at least one occupational pen- sion scheme or individual scheme. Another characteristic of the Danish pen- sion system is the extensive use of guaranteed minimum benefits in the second and third pillars. Plans operated by insurance compa- nies and multi-employer pension funds both offer guaranteed minimum investment returns (in the sense that future benefits are guaran- teed), while banks do not have permission to offer guaranteed minimum investment returns. The use of guaranteed benefits in occupa- tional pension plans has been promoted by the active involvement of labor unions in collec- tive bargaining and a strong emphasis on risk sharing arrangements that aim to protect reti- ring workers from large fluctuations in invest- ment returns. In recent years, demand for more individualized products, with no or a low guarantee attached to the benefits, has been increasing fast. Hence, unit link pro- ducts and specialized products tailored to the life cycle of the pension savers, are being developed and brought to the market with great success, in both pillar II and pillar III schemes. Contributions to occupational pension plans increased steadily over the past ten years or so. Their annual growth rate was remarkably sta- ble, ranging between 10 and 12 percent in nominal terms while during the same period (1995-2004) inflation averaged 2 percent per year. 316 Replacement rates in the new Swedish pension system - a Danish perspective The increase in contributed amounts are partly due to expanding coverage and partly to a gradually rising contribution rate. While contribution rates vary among different schemes, the upward trend in contribution rates is illustrated by the following figure, which represents the evolution of the average contribution rate for schemes covered by the labor market agreement between the Danish Confederation of Trade Unions (LO) and the Danish Employers’ Confederation (DA). The contribution rate has crept upwards from 1 percent in 1993 to over 10 percent in 2006 (Figure 1). Replacement ratios Projections of current and future replacement ratios are based on assumptions about future performance and bonus payments and take into account all types of pension benefits and allow for tax payments. A report from the Ministry of Economic and Business Affairs2 provides some details of current and expected replacement ratios. The average replacement ratio is expected to increase in the future irrespective of educa- tion (and income). For persons with a shorter education, the replacement ratio will increase from 80 per cent in 2000, reaching almost 100 per cent in 2045 (figure 2). For highly educat- ed persons it is foreseen to reach a little less than 90 per cent in 2045. The reason for the shift in replacement ratios is the widening coverage of occupational pensions, which will affect in particular the lower income groups. For all groups, private pensions will play a more important role in the future, but the social pension will still represent the major source of income for pensioners with a mod- est income even in 2045. Risk sharing in the Danish pension system The Danish pension system faces to some extent the problems which the Swedish sys- tem is exposed to. The social, tax financed pensions are set to increase significantly over Figure 1: Evolution of Contribution Rate 0 2 4 6 8 10 12 1993 1995 1997 1999 2001 2003 2005 2007 Percent Source: The Danish Insurance Association 317 Replacement rates in the new Swedish pension system - a Danish perspective 2000 2045 2000 2045 2000 2045 the decades to come because of demographic changes. Fewer people will be active on the labor market, while the number of retirees will increase – and the retirees will live longer, adding to the public financing pressure. However, in the Swedish case this will to a very large extent be a political problem. Hard political decisions will have to be made on how to pay the bill – more reliance on tax payers money may be inevitable, or future pensioners expectations will not be fulfilled. In the Danish case, a public financing di- lemma also arises in the context of the social pensions. However, when it comes to the funded schemes of pillars II and III, the pro- blem lies not in the first place with politicians. It lies with the pension savers and sharehold- ers of the pension institutions. Annuity insu- rance is quite widespread, and where guaran- tees of future benefits have been provided as part of annuity insurance, these guarantees must be met – in the end by shareholders of the pension companies paying the bill. For lump sum benefits (benefits are paid out at one time as a lump sum) and phased withdrawals (benefits are paid out over a specified period) the pensioner bears the risk of increased lon- gevity. In pillar II and III schemes the market risk (financial risk) is split between pension savers and the shareholders – in the sense that they share losses or yields that are considered in- adequate. However, in relation to products based on guaranteed benefits the shareholders solely bear the risk that the long term yield can be lower than the guarantees, while for unit link products the market risk is (primarily) borne by the pension saver. Basically, then, the Danish system has broad- er risk sharing features than the Swedish one. In particular, the risk that political promises to both tax payers and pensioners can not be kept is probably somewhat lower in the Danish than in the Swedish system. Of course there are political risks to the Danish pillar II and III schemes, however, of a different nature than in the Swedish system. We shall not elaborate in detail on this. Less educated Skilled Highly educated Source: Ministry of Economic and Business Affairs (2003). Figure 2: Average expected replacement ratios for different groups of education in 2000 and 2045 Percent 0 20 40 60 80 100 Private pension Public pension 318 Replacement rates in the new Swedish pension system - a Danish perspective Tax wedges The Danish pillar II and III schemes are de- fined contribution plans. Since they do not rely on tax financing, there is no problem of introducing tax wedges – which would reduce economic efficiency – associated with the financing of these two pillars. The need to avoid tax induced efficiency problems may be one reason why the Swedish system has been reformed into one based on a NDC scheme. If the system is credible and pension savers believe that their future pen- sion income reflect their tax contributions and some “interests earned”, this system may suc- ceed in reducing efficiency problems usually associated with tax financing. However, if it is not credible – which KG Scherman leads one to presume – this system may distort econo- mic decisionmaking because of distortions introduced by the tax financing. Conclusion The Danish and the Swedish pension systems are quite different. In both pension systems the public pension system plays a key role, aiming at avoiding poverty in old age. How- ever, it seems to us that the further role of not only avoiding poverty, but also securing de- cent incomes upon retirement, is to a greater extent a public task in Sweden than in Den- mark. The reliance on tax payer financing is greater in Sweden, hence calling upon politi- cians to make difficult choices when the sys- tem comes under strain as documented by KG Scherman. The Danish pension system as a whole must meet some of the same objectives as the Swe- dish system. But the task of providing ade- quate incomes for pensioners and for bearing the associated risk lies more with the future pensioners and shareholders of pension insti- tutions than in Sweden. In any way, if the problems highlighted by KG Scherman are real, serious and difficult political choices must be made. And politi- cians are not keen to make those choices. In Denmark, a reform of the public early retire- ment system has just been passed in parlia- ment alongside with an increase of the retire- ment age, which is set to increase in line with longevity. Such changes are warranted and needed. But in order to gain political support, they are introduced with a long time horizon – the increase in the retirement age will only have effect starting in 2024 and all changes will only affect people under 48 years at the end of 2006. Hopefully Swedish politicians are better at making tough decisions and introducing real welfare reforms than the Danish ones! Notes 1 This article is one in a series about the Swedish pension reform. Earlier articles published in the NFT are written by Hagberg and Wohlner (4/ 2002), Könberg (1/2004), Casey (2/2004), Barr (3/2004), Lezner and Tipperman (4/2004), McGillivray (3/2005), Scherman (2/2006), Set- tergren (3/2006) and Andresen (4/2006). These articles can all be found at www.sff.a.se/ ?avd=forlag&sida= pension. lasso 2 “Increased freedom of choice in the pension saving”, May 2003.