Pension systems refer to institutional and occupational settings that provide retirement income during retirement. Pension systems are commonly described in a multi-pillar pension setting (Holzmann et al. 2008). Pension systems combine government and privately sponsored support to finance a suitable standard of living during retirement. Governments provide income security to maintain a minimum standard of living and privately sponsored plans, through employers or individual savings, provide supplemental retirement income. Benefits differ across pension systems in two ways, depending on who the main bearer of the risks is. Some schemes provide guarantees in the benefit payout and transfer the financing risk to the sponsor. Others guarantee a minimum level of financing, transferring the risk to the retiree. The benefit design has an impact in the long-term sustainability of the pension system, the standard of living during retirement, and the actuarial fairness that represents the link between lifelong contributions and benefits. Pension systems can be financed through pay-as-you-go (PAYG) or through pre-funding. PAYG pension schemes finance retirement benefits in one particular period with the contributions of the working-age population. Pre-funding, on the other hand, relies on the capital markets.
Pension Structure and Financing
In PAYG, the sponsor, typically the government, faces labor, demographic, and longevity risk (Bloom et al. 2015; Alonso-García and Rosado-Cebrian 2019). Indeed, labor and demographic risk affect the contribution base of the pension system, as the number of employed, their wages, and the fertility influence whether there is sufficient liquidity to fund pension liabilities. However, longevity risk affects the liability base of the pension system, as unexpected improvements in life expectancy will extend the period that pensions are paid. In pre-funding, accrued benefits, typically individualized, are funded through employer and/or employee contributions and earn a market return. The risks faced by the sponsor are labor, market, and longevity risk. Accrued benefits typically depend on wages, earned while employed, and the market returns. Similar to the PAYG schemes, the sponsor needs to finance retirement contingent on survival and hence faces longevity risk too.
Within a multi-pillar pension setting, the zero pillar works as a poverty alleviation mechanism that guarantees a minimum benefit during retirement. The first pillar is earnings-based and aims to provide replacement income. The second and third pillars are privately managed individual savings accounts. The second pillar is typically managed by the employer through occupational schemes, pension funds, or insurance companies. In some countries the second pillar is mandatory. On the other hand, the third pillar is a voluntary retirement savings scheme.
The zero and first pillar are typically provided by the government through PAYG financing. However, in some countries in Asia and South America, these pillars are provident funds or mandatory individual accounts that rely on the financial markets through government intervention (Queisser 1995). An additional distinction can be made in how the government-sponsored pension setting is designed (Esping-Andersen 2013). Beverigdian pension systems provide flat-rate zero pillar pension payments that serve as a poverty alleviation mechanism. The first pillar is combined with the second pillar, which is commonly mandated or has high coverage and is pre-funded. The Bismarckian model combines the zero and first pillar and offers earnings-based retirement benefits to replace earnings during the retiree’s working life. In this model, second and third pillar arrangements are less prevalent.
Some countries combine PAYG and pre-funded schemes within the zero and first pillar as to diversify the abovementioned risks. The PAYG rate of return tends to be lower than the financial market. Feldstein (1996) argues that the after-tax financial returns have been as much as four times as high than social security returns. However, diversification benefits appear when accounting for the variability of the return favoring a mix of PAYG and pre-funded schemes (De Menil et al. 2006; Alonso-García and Devolder 2016). Furthermore, full transition from PAYG to pre-funded suffers from an obvious flaw. PAYG systems were introduced by starting paying retirement benefits to individuals who did not contribute explicitly to the system, and hence no buffer fund was accumulated. This is commonly called the “first-generation gift.” This carries forward and presents a problem for current generations if they were to transit fully from PAYG to pre-funded systems. Indeed, in such a scenario, current generations would have to pay twice: first to finance their own future pension through pre-funded schemes and twice to finance PAYG liabilities to current retirees (Barr and Diamond 2006).
There are two main ways of defining pension benefits, both in funding and pay-as-you-go, namely, defined benefit (DB) and defined contribution (DC) (See Triangle “Defined Contribution/Defined Benefit”). DB schemes provide guarantees in the level of benefits paid at and during retirement and provide full certainty to the beneficiaries. However, financing risk is borne by the sponsor during the accumulation phase, as sufficient funds need to be accumulated to finance the liabilities, and during the decumulation phase when paying lifetime benefits. On the other hand, DC schemes guarantee a minimum level of financing, e.g., a proportion of the wages, but do not provide certainty on the level of the accumulated capital as it depends on economic or market performance indices. Hence, the sponsor does not bear financing risk during the accumulation phase but still does during the decumulation when paying lifetime benefits. Indeed, both benefit structures are subject to longevity risk, that is, the risk that individuals may live longer than anticipated and financially accounted for.
PAYG and pre-funded systems can be DB, DC, or hybrid. The benefit design is comparable between PAYG and pre-funded systems. However, pre-funded schemes aim to supplement the zero and first pillars. Hence, it is common to only accrue DB rights or pay DC contributions above a certain wage threshold or cap. The idea is that the first part of your wage contributes to the government-organized schemes through social security contributions and that any excess wages earned contribute to your private pre-funded schemes.
DB systems have a benefit formula that depends on wages, accrual rates, and replacement rates (Poterba et al. 2007). The idea is that individuals with full careers will have a certain replacement rate during retirement. This replacement rate is based on either the last wage, the average of the last contribution years, the average of the best contribution years, or the average of the lifetime wages during their working career. DC systems accrue return on lifelong contributions. In pay-as-you-go, DC systems are known as notional, or nonfinancial, defined contribution (NDC). In pre-funded systems, they are known as financial defined contribution (FDC). NDC schemes mimic individual savings accounts as the accumulated capital depends on contributions and its returns (Palmer 2006). The notional return earned is based on an index that reflects the financial health of the system and is based on the GDP growth, average wages, or covered wage bill. The system is called notional as the system is not pre-funded and accumulated savings are only used for record keeping. Upon retirement, the virtual capital is converted into a lifetime payment stream. The conversion rate, or annuity rate, depends on the cohort life expectancy, pre-charged discounting rate, and indexation rate. FDC schemes, compared to NDC schemes, earn a return based on financial markets.
Hybrid systems appear in pay-as-you-go and pre-funded schemes. They incorporate risk-sharing mechanisms either between employer and employee or between different cohorts of workers. Hence, they are seen as a middle way between classical DB and DC. In pay-as-you-go, hybrid schemes are known as “point” systems. Points are earned based on the relationship between the wages earned and the average wages of the economy during the same period (Legros 2006). For instance, those earning exactly the average wage during their careers will accumulate 1 point per worked year. Upon, and during, retirement the pension is calculated based on the number of accumulated points, the value of the point, and any adjustments based on early or late retirement. This value can incorporate balancing mechanisms that increase or decrease the payments based on the sustainability of the system. In pre-funded schemes, hybrid plans are found in the form of cash balances that guarantee a minimum return, conditional indexation plans that aim to provide a DB benefit subject to solvency requirements, nursery plans that combine DC and DB payments during retirement, and floor or underpin plans that pay the higher between a DC and DB benefit (Pugh and Yermo 2008).
The goodness of the benefit design can be analyzed in at least three different manners. First, sustainability factors inform about the short- and long-term feasibility of retirement benefit structures. The scheme, both for PAYG and pre-funded, should have sufficient income from contributions or liquid funds to finance retirement benefits. Additionally, the current level of liabilities should be backed by sufficient inter-temporal assets. This analysis raises the need to compile actuarial balances, compulsory for pre-funded schemes, increasingly relevant for PAYG schemes, to highlight the true pension risk (Vidal-Meliá and Boado-Penas 2013). If the system is not sustainable, automatic balancing mechanisms that adjust key parameters in the system, such as accrual or indexation rates, could be introduced to ensure the sustainability (Godínez-Olivares et al. 2016; Alonso-García et al. 2018b). Secondly, retirement income adequacy should be ensured. Adequacy is measured in relative terms and compares the first pension benefit to the last wage earned. An alternative way of assessing the adequacy is to compare the pension benefit to the average wages in the economy, known as “benefit ratio” in aggregate accounting studies (Boldrin et al. 1999). The first measure reflects how the first pension smooths lifetime income, whereas the second measure compares pensions paid to average productivity. Lastly, actuarial fairness can be analyzed as a longitudinal measure of the pension system. It measures the relationship between lifetime contributions during the individual’s working life and lifetime benefits during retirement. A system is deemed fair if these two values coincide.
Classical DB schemes have high replacement rate levels as they do not adjust payments to demographic factors. This yields to unsustainable levels of payments, exacerbated by demographic shocks such as baby booms or sudden rise in life expectancy (Alonso-García et al. 2018a). Pure NDC pension and point systems are deemed more actuarially fair than DB schemes. NDCs achieve this by factoring the life expectancy at retirement in the first retirement income benefit. However, the choice of the annuity pays a crucial role and can yield to unsustainable payment levels, especially in the presence of systematic longevity increases (Alonso-García and Devolder 2019). Points achieve this through their adjustment factors and value of the points. It should be noted that carefully designed DB schemes that adjust the payments to life expectancy improvements would have comparable actuarial fairness to (N)DC (Cichon 1999).
Future Directions of Research
Recent literature has focused on the sustainability of pension systems, especially those PAYG funded. This is due to the great increase in life expectancy in most countries. Globally, life expectancy has increased from 66.5 to 72 between 2000 and 2016 (World Health Organization 2019). The financial impact of ageing is worsened by the demographic transition caused by a fertility boom followed by a fertility bust. Some countries have made structural pension reforms and changed their benefit structure from DB to (N)DC. Others have explored the option to pre-fund benefits instead of financing them through pay-as-you- go. The advantages and disadvantages of these reforms have been discussed in the previous paragraphs. Most, however, have introduced parametric reforms to adjust the retirement age or accrual rates. Alonso-García et al. (2018a) show that raising retirement age successfully reduces the dependency ratio, compared to a scenario where the retirement age is fixed. The effect of such reforms is much more pronounced in DB than DC schemes due to benefit design. Overall, reforms have aimed to reduce benefit levels to achieve jointly actuarial fairness and long-term sustainability. However, these reforms are based on the assumption that there are no differentials in mortality.
Future research should focus on the impact of policy design in inequality. Diverging life expectancies, as noted in the literature (e.g., Chetty et al. 2016), need to be incorporated in the policy design (Holzmann et al. 2019) as it may create labor market distortions for those living less and longer than average. Those with poorer health might retire earlier to ensure a minimum retirement period, whereas those in good health might retire earlier as well to maximize the implicit subsidy received by exceeding the average life expectancy. Furthermore, the gender dimension should be incorporated into pensions. Women often have a lower labor participation rate, work less hours, and live longer than the average population (Deere and Doss 2006). Finally, pension adequacy should be strengthened. Recent parametric or structural reforms have aimed to provide sustainable, but often significantly lower, retirement benefits. Future reforms should aim at providing sustainable but adequate retirement income. A possible venue of research is the role of capital markets to further complement retirement income benefits. Tax reforms, together with behavioral insights, could be used to engage individuals into supplementing their retirement through pension savings schemes.
Pension systems refer to institutional and occupational settings that provide retirement income during retirement. Retirement benefits can be funded through pay-as-you-go where income from contributions finances pension expenditures during a given period. Alternatively, they can be pre-funded where pension contributions mimic financial savings accounts earn a financial market return. There are various benefit designs that guarantee pension levels, contribution levels, or a combination of both. The benefit design, and financing strategy, impacts the financial health of the pension system and its adequacy to maintain a standard of living during retirement and fairness. Future research should focus on the effect of pension systems and reforms on heterogeneous agents, be it in terms of gender or socioeconomic status, as well as how financial markets can further complement retirement income benefits.
- Alonso-García J, Rosado-Cebrian B (2019) Financial crisis and pension reform in Spain: the effect of labour market dynamics. J Econ Policy Reform. https://doi.org/10.1080/17487870.2019.1599718
- Esping-Andersen G (2013) The three worlds of welfare capitalism. Wiley, CambridgeGoogle Scholar
- Feldstein M (1996) The missing piece in policy analysis: social Security reform. Am Econ Rev 86(2):1–14Google Scholar
- Holzmann R, Hinz RP, Dorfman M (2008) Pension systems and reform conceptual framework. World Bank discussion paper, 824. http://documents.worldbank.org/curated/en/716871468156888545/Pension-systems-and-reform-conceptual-framework
- Holzmann R, Alonso-García J, Labit-Hardy H, Villegas A (2019) NDC schemes and heterogeneity in longevity: proposals for redesign. In: Holzmann R, Palmer E, Palacios R, Sacchi S (eds) Progress and challenges of nonfinancial defined pension schemes –volume 1: Addressing marginalization, polarization, and the labor market. The World Bank, Washington DCGoogle Scholar
- Legros F (2006) NDCs: a comparison of the French and German point systems. In: Holzmann R, Palmer E (eds) Pension reform: issues and prospects for Non-Financial Defined Contribution (NDC) schemes. The World Bank, Washington DC, pp 203–225. http://documents.worldbank.org/curated/en/713861468120840510/Pension-reform-issues-and-prospects-for-Non-Financial-Defined-Contribution-NDC-schemesGoogle Scholar
- Palmer E (2006) What’s NDC? In: Holzmann R, Palmer E (eds) Pension reform: issues and prospects for Non-Financial Defined Contribution (NDC) schemes. The World Bank, Washington, DC, pp 17–35. http://documents.worldbank.org/curated/en/713861468120840510/Pension-reform-issues-and-prospects-for-Non-Financial-Defined-Contribution-NDC-schemesGoogle Scholar
- Queisser M (1995) Chile and beyond: the second-generation pension reforms in Latin America. Int Soc Secur Rev 48(3–4):23–39. https://doi.org/10.1111/j.1468-246X.1995.tb00444.xCrossRefGoogle Scholar
- World Health Organization (2019) Life expectancy and healthy life expectancy. http://apps.who.int/gho/data/node.home. Accessed 15 Feb 2019