Torben M Andersen, Joydeep Bhattacharya, Anna Grodecka-Messi, Katja Mann 05 May 2022
Rising income and wealth inequality have recently attracted attention among scholars and public discourse. International institutions like the European Commission, the OECD, and the IMF have raised concerns that these developments, especially the spectacular increases in incomes and wealth at the top, may leave enduring blotches on the social fabric. Piketty (2014) fuels these worries by pointing to the self-reinforcing effects of the rich saving more and owning more wealth, thereby appropriating the return to capital. In turn, these effects are possibly magnified and propagated by intergenerational transfers of wealth (Elinder et al. 2016, Boserup et al. 2016, Nolan et al. 2020).
The observed developments in income and wealth inequality are complex and have multiple explanations (De Nardi and Fella 2017, Hubmer et al. 2020, Benhabib et al. 2019, Jakobsen et al. 2020). A core element in this debate is the role of retirement savings – i.e. savings to ensure adequate living standards post-retirement. Pension systems intersect with individual-level decisions to save. Yet, somewhat surprisingly, their role in this debate has been neglected even though this link was prominently featured in a seminal and widely influential 1994 World Bank report.
The prototype pension systems are (a) pay-as-you-go (PAYG), where the earnings of the current working generations are taxed and transferred to existing retirees; and (b) fully funded (FF), where workers contribute mandated amounts to individualised accounts which are managed by professionals and returned to the worker post-retirement. In a PAYG system, some responsibility for retirement savings is shifted to the public sector, which means individuals save less on their own, and their financial wealth post-retirement is lower. Although voluntary savings are also crowded out in a mandated FF scheme, total savings increase due to contributions to the pension funds. The definite implication is wealth accumulation, and its distribution may critically depend on whether the pension system is funded or not. It follows that a country may witness significant changes in both the wealth level and its distribution, transitioning from a dominant PAYG pension system to a mandated FF system.
In a recent paper (Andersen et al. 2022), we study the role of the pension system in wealth accumulation and its distribution. We situate our laboratory in Denmark mainly because the evolution of the pension system there comes close to the textbook reform of introducing mandated funded pensions as an add-on to an existing PAYG pension scheme. Such reforms are also underway or being actively contemplated in many OECD nations. The choice of Denmark is also meaningful because it currently has the highest accumulated pension wealth as a share of GDP in the OECD (240% in 2020) and is one of the few OECD countries that has experienced a decline in wealth inequality in recent years. It also bears emphasis that Denmark generates high-quality registry wealth data.
The pension reforms
A significant change in the Danish pension system was initiated in the late 1980s with the increased penetration and phasing-in of occupational pension schemes as an outcome of collective agreements. Hence, contributions to the plan are involuntary for the individual but result from voluntary bargaining between the parties involved. The scheme is a funded, defined-contribution scheme. Contribution rates were gradually increased and, since 2010, have been held steady at between 12% and 18%. The pension funds are run as non-profit organisations, contributions are collectively invested, and benefits are paid in annuities and lump-sum payments. The Danish public pension scheme, a PAYG system, consists of a universal base pension and means-tested supplements. It remained essentially unchanged during the phasing-in period (Andersen et al. 2022).
The reform, which has enjoyed and continues to enjoy broad support, is believed to have had several drivers. It was generally perceived that Denmark had an under-savings problem with low savings rates, which showed as systematic current account deficits at the macro level. The reform was expected to cure this savings problem. Also, the existing system seemed unfit to deliver acceptable living standards to retirees from a forward-looking perspective. The ‘solution’ – raising taxes even further – was not palatable. Finally, the new scheme resulted from a compromise between unions wanting firm-specific wage-earner pension plans and employers who opposed the same.
Effects of the pension reform
The above-discussed reform has had significant effects on the Danish Economy. The savings rate has increased significantly (see Figure 1). And so has the wealth accumulated in pension funds (see Figure 2).
Figure 1 Net savings rate in Denmark, 1980-2019
Source: Statistics Denmark, www.statistikbanken.dk
Figure 2 Aggregate pension assets in Denmark, 1986-2019
Source: Danish Ministry of Taxation, www.skm.dk and www.statistikbanken.dk
The current account has shifted from systematic deficit to surplus.1 While Denmark in the late 1980s had a foreign net debt position close to 40% of GDP, there is now a net wealth position of about 75% of GDP.
Figure 3 Cross-sectional wealth holdings of Danish households, 2017
Source: Andersen et al. (2022)
Today, private pension assets play a dominant role in the asset portfolio of near-retirees. Wealth inequality has declined over this period. The Gini coefficient, defined over wealth computed for the entire population, has fallen from 0.75 in 1992 to 0.69 in 2017. Restricted to the age group 60-69, the Gini declined from 0.63 to 0.51. At the same time, the wealth share of the top 10% has not increased.
We link these developments to the Danish pension reforms using a lifecycle model with overlapping generations. We embed a two-pillar pension system (the second, fully funded pillar on top of the public pension one). Calibrating the model to Danish data, we find that the pension reform – i.e. the phasing-in of funded pensions – does an excellent job accounting for these developments. The Lorenz curves over wealth for the entire population and the age group 60-69, as shown in Figure 4, document the importance of pension wealth in determining the wealth distribution and making a difference for the lower wealth deciles.
Figure 4 Lorenz curves for net wealth in Denmark, 2017
Source: Andersen et al. (2022)
Why did the pension reform have an equalising effect on the wealth distribution? Within the context of our model, pension reform raised retirement savings (the sum of personal savings and the portion done by the government) especially for low- and middle-income people who, left to their own devices, would want to borrow. While we do not get into the details of the underlying reasons in our paper (other than to assume people place different weights on their future utility), one can, going beyond the confines of our model, attribute the higher saving to at least three factors. First, people may be present-biased and discount their retirement years heavily, saving insufficiently for their retirement. Here, the mandated aspect of the new scheme forces such households to save more, an effect which is most potent for low-income and highly- impatient groups. Second, collective financing gives low-wealth groups access to financial expertise and scope for risk diversification, which is difficult to attain for individuals with modest wealth levels (see also Scharfstein 2018). Financial literacy barriers, often unsurmountable for the poorer segments of society, may also be overcome. Finally, the average return in the pension funds (about 5%) has been higher than the implicit return via PAYG pensions (about 2 %).
The analysis presented here shows the importance of pension system design for the level and distribution of wealth. It is a separate issue to discuss the normative aspects of wealth distribution and possible policy steps to affect it. Still, the present analysis shows that the design of the pension system should not be neglected in this discussion as it plays an important role. Our work suggests that pension systems must be considered in cross-country comparisons of wealth inequality and its distribution.
Andersen, T M, J Bhattacharya, A Grodecka-Messi and K Mann (2022), “Pension reform and wealth inequality: evidence for Denmark”, CEPR Discussion Paper No. 17078.
Andersen, T M, S-E H Jensen and J Rangvid (Eds.) (2022), The Danish Pension System: Design, Performance and Challenges, Oxford University Press.
Benhabib, J, A Bisin and M Luo (2019), “Wealth distribution and social mobility in the US: A quantitative approach”, American Economic Review 109(5): 1623–47.
Boserup, S, W Kopczuk, C T Kreiner (2016), “Bequests and wealth inequality: Evidence from Denmark”, VoxEU.org, 11 March.
De Nardi, M and G Fella (2017), “Saving and wealth inequality”, Review of Economic Dynamics 26: 280–300.
Elinder, M, O Erixson and D Waldenström (2016), “How inheritances influence wealth inequality”, VoxEU.org, 20 April.
Hubmer, J, P Krusell and A A Smith (2020), “Sources of US wealth inequality: Past, present, and Future”, NBER Macroeconomics Annual 2020 (35): 391–455.
Jakobsen, K, K Jakobsen, H Kleven and G Zucman (2020), “Wealth Taxation and Wealth Accumulation: Theory and Evidence from Denmark”, Quarterly Journal of Economics 135(1): 329–388.
Nolan, B, J C Palomino, P Van Kerm, S Morelli (2020), “The intergenerational transmission of wealth in rich Countries”, VoxEU.org, 19 September.
OECD (2021), Pensions in Figures, Paris.
Piketty, T (2014), Capital in the Twenty-First Century, Cambridge, MA: Harvard University Press.
Poschke, M, and B Kaymak (2016), “US wealth inequality: Quantifying the driving factors”, VoxEU.org, 17 April.
Scharfstein, D S (2018), “Presidential Address: Pension Policy and the Financial System”, Journal of Finance 73 (4): 1463-1512.
Waldenström, D (2021), “Wealth and history: A reappraisal”, VoxEU.org, 17 November.
World Bank (1994), “Averting the old age crisis: Policies to protect the old and promote growth”, Washington, D.C.: World Bank Group.
1 This has also been important in supporting the credibility of the Danish exchange rate peg to the euro.