In recent years, the sustainability of pension provisions has become an important element in the policy debate. The US social security fund is expected to run out of funds by 2040. Many European pension systems are considered to be unsustainable as well. These expected solvency problems must necessarily lead to future adjustment of pension systems of the countries involved. The postponement of these policy adjustments creates substantial uncertainty among participants in the schemes with respect to both future contribution rates and the real value of future pension benefits.
At the same time, the adjustments that have been made show a general movement towards a system where contributions are known - a defined contribution (DC) system - but benefits are uncertain. This conflicts with the original purpose of collective pension systems, where the benefit was defined (a defined benefit (DB) system) so as to guarantee retired workers a minimum standard of living. In addition, a shift can be observed towards private saving accounts. This diminishes the role of collective risk sharing in exchange for a larger element of private risk. As a result, future pensions are increasingly at risk -–and the general public is becoming aware of this.
Risk and uncertainty are important elements in people’s lives. Some risks are primarily of an individual nature and average out for the economy as a whole. Examples are individual mortality risk and, to a lesser extent, health risks. Other risks affect most people at the same time, such as stock market fluctuations. This difference is important because macroeconomic risks cannot be fully diversified away by mutual insurance.
This implies that they must affect people’s welfare. Economic agents respond to these risks by valuing income sources differently, depending on the actual or perceived risk of the source. If a complete set of markets exists where the claims to these income streams can be traded, this macroeconomic risk is borne by agents with the highest risk tolerance. This minimises the price of risk for society.
Some types of risk cannot be traded on markets, however, because the relevant market does not exist. This market failure may have different causes. For example, unemployment risk is difficult to insure with private insurance companies, because insured households have more information about their efforts to find a new job than the insurance company does. Therefore only the bad risks would insure themselves. With old-age insurance, the problem is different. Working-age households pay contributions to the pension scheme, which are used to help pay the pensions of the current pensioners if the savings of those pensioners collect a low return. In return, workers expect that when they retire future young households will help pay their pensions if they (as retirees) suffer an adverse shock. In this way the risk of shocks is shared between generations.
However, some of the households that are party to the contract have not yet been born. An explicit legal contract is thus not possible. A market in which pension contracts can be traded between generations therefore does not exist. As a result, macroeconomic risks are larger than what could ideally be achieved. Indeed, a large part of the value added of collective pension schemes is precisely to fill this missing market for trade between generations that are not alive at the same time.
The increased awareness that social security may not be secure has spawned a growing body of research into the detrimental effects of ageing on social security and the scope for pension reform to mitigate these effects. The bulk of this research has focused on the link between dependency ratios and contribution rates and on the distortionary effects of rising contribution rates. Less attention has been paid to the increased risk of pension arrangements as a result of ageing. The risk aspects of pension schemes have been analysed almost exclusively from the point of view of pension funds, using asset liability management (ALM) models. This approach largely neglects the demand for insurance on the part of the households and the response of households to changes in pension provisions and the risk of pension arrangements.
This is surprising, since insurance is a central aspect of pensions. Indeed, a number of important questions can be associated with pension risk. What is the size of the risk involved for pensioners? How do these risks depend on the characteristics of the pension contract (DB or DC)? Can these risks be mitigated by funding pension obligations? Addressing these questions requires an explicit representation of the sources of uncertainty about pension costs and returns as well as the effect that this uncertainty has on the behaviour of households.
Our research project focuses on the interaction between macroeconomic risks, pension arrangements and social security in an ageing society. For this, the project will employ both small analytic equilibrium models and large numerical ones to model macroeconomic risk and the response of economic agents to that risk. Important sources of risk that will be considered include demographic shocks and interest rate shocks. The project explicitly addresses the effect of risk on the market value of pension rights and pension obligations. Pension rights are an important part of lifetime wealth of households.
The risk profile of pension rights differs from that of financial assets, because these rights are linked strongly to future wage rates. This implies that future pension rights are discounted at a rate different from that of other assets. By offering an alternative asset, pension systems affect the behaviour of households both in terms of saving and in terms of human capital formation and labour supply. This project studies the impact of various pension arrangements on macroeconomic outcomes.
In this context, the project also considers the institutional design of pension systems. The DB character of many pension arrangements implies that economic shocks that depress the rate of return on assets of pension funds boost pension contribution rates. A rise in contribution rates, however, distorts labour supply and hampers economic growth. These institutional arrangements may therefore potentially be destabilising. Indeed, recent developments show that the DB character of pensions is already under attack and that pension schemes are moving towards a DC setting. This raises the question of how far we should proceed in that direction. Which features of an economy determine an optimal mix of DB and DC elements?
This project uses the current Dutch institutional setting, including the financial assessment framework (FTK), as a point of reference to analyse the merits of different pension systems. In particular, we will address the following issues:
q To what extent are current pension arrangements sustainable in the presence of macroeconomic shocks?
q What is the macroeconomic impact of a (partial) shift from DB systems towards DC systems?
q What are the welfare effects of the supervisory rules used by the FTK to value pension obligations and define the required level of financial buffers?
The project starts with an evaluation of steady-state risk and valuation of pension contracts in a small general equilibrium model with endogenous risk pricing. In 2006, the analysis will be extended to include transition paths. In 2007, we intend to explore the welfare and efficiency effects of pension arrangements.