We analyze the optimal balance between social security taxation and private saving in the provision of retirement income in dynamically efficient economies, a question at the center of policy debates in Europe and the United States. We consider the relative importance for this question of the return to capital, the internal return of the pay-as-you-go system, and the variabilities and correlation (or independence) of labor earnings and the capital return. We analyse these influences theoretically in the context of a two-period, overlapping generations model with uncertainty. We use a new method to calibrate the model using annual data on GDP per worker and the total real return on equities, from 1950 to 2002, from which we infer the stochastic characteristics of lifetime labor income and the return to lifetime savings in the US, UK, France and Japan. We obtain a range of optimal, steady-state values of the social security tax and the rate of lifetime savings. When the relative rate of risk aversion is assumed to be 2.5, the computed optimal tax varies from 5% in the United States to 22% in Japan. France is similar to Japan, and the UK is in between.

The second issue of the fifth volume of the JPEF features 4 research articles, and a book review section. The first article is by Estelle James, Guillermo Martinez and Augusto Iglesias and looks at: “The payout stage in Chile: Who annuitizes and why?” Low rates of voluntary annuitization are a major issue in defined contribution pension designs. Yet, in Chile, about 2/3 of individuals choose to annuitize instead of taking programmed withdrawals. The paper examines the reasons for this high rate of annuitization and concludes that it is due to the combination of a competitive market in annuities and the relatively constrained alternatives available to individuals. The paper nevertheless finds some evidence of adverse selection in the Chile annuity market.

The last issue of the fifth volume of the JPEF features 4 research articles, an issues & policy paper and a book review section. The first paper in the issue looks at how unemployment affects retirement savings for those with DC plans. Jason Seligman and Jeffrey Wenger of the University of Georgia find that unemployment magnifies pension losses because unemployment occurs at the same time as equity markets decline so that workers systematically lose not just from unemployment but also from declines in the stock market. Because unemployed workers do not have funds to invest when the stock market is relatively low, they stand to lose from subsequent increases in equity markets. The authors look at who is affected most and find that nearly ten percent workers will lose between 8 and 13% of their retirement savings due to unemployment. The results overall indicate the importance of incorporating human capital factors such as unemployment spells and changes in the wage distribution into retirement planning models.

Retirement savings in defined contribution plans vary as a result of the timing and frequency of unemployment spells. We hypothesize that unemployment is coincident with negative shocks to equities prices, implying workers may systematically miss investment opportunities. First we match historic stock returns to unemployment hazards by gender, and earnings quartile. Next we test the relationship between unemployment, equity returns, and pension savings, by repeated simulation. Finally, we find that the timing of unemployment spells amplifies retirement savings losses on average for all worker-types in our analysis. Timing impacts are observed to be largest for high earnings workers and to increase with unemployment losses disproportionately.

In 1981 Chile adopted its new multi-pillar system, which featured privately managed individual accounts. Starting in 1983 payouts from the accounts were permitted and detailed rules about payouts were put in place. The Chilean scheme therefore gives us an opportunity to examine how pensioners and pension providers react when individual accounts replace DB systems, and how detailed regulations shape these reactions.

Retirees in Chile have a choice between early versus normal retirement (before or after age 65M/60W) and between annuitization versus programmed withdrawals; lump sum withdrawals are largely ruled out. Almost two-thirds of all retirees have annuitized – a very high proportion compared with other countries. This paper argues that this high rate of annuitization is the result of guarantees and regulations that constrain payout choices, insure retirees through the minimum pension guarantee, eliminate other DB components, and give a competitive advantage to insurance companies selling annuities. The minimum pension financed by the government provides insurance to workers with small accumulations, who retire at the normal age with programmed withdrawals, while those with large accumulations retire early and must purchase annuities to acquire longevity and investment insurance. Insurance companies further induce annuitization by marketing aggressively, facilitating early retirement for those who annuitize and offering a high money's worth ratio for price-indexed annuities. We find evidence of adverse selection based on asymmetric information about short-run health status, but this does not seem to deter the high rate of annuitization.

Our first issue of the fifth volume of the JPEF features 4 research articles, an issues & policy piece and a book review section. The lead article is by Georges de Menil (PSE, Paris and Stern School, NYU), Fabrice Murtin (CREST(INSEE) and PSE, Paris) and Eytan Sheshkinski (Hebrew University of Jerusalem and Princeton University). The paper on “Planning for the Optimal Mix of Paygo Tax and Funded Savings” looks at the classical question of what the right balance is between pay-as-you-go and funded pension provision. Using a stochastic two period overlapping generations model, the authors show how optimal pension provision depends on the variances and correlations of labour earnings, return on capital and the internal rate of return on the pay-as-you-go pension system.

This paper empirically examines the strategic asset allocation and the asset/liability issues in the Finnish defined benefit pension funds. The results indicate that there is a relationship between the liability structure and the asset allocation. While pension funds with younger participants have more equity exposure, more mature pension funds have more fixed income investments.

Wide dispersion in asset allocations is also found between the funds. One fund holds its entire portfolio in fixed income securities, whereas other funds have none or only few fixed income holdings. Equity investments also vary dramatically, ranging from 0% to over 70% of the asset allocation. The same applies to investments in a sponsor, real estate investment, and money market investments. A portion of these different asset allocations is explained by the liability structure, but another part remains unexplained. The other variables affecting strategic asset allocation of a pension fund are not obvious, but they could include factors such as regulatory environment, historical reasons, mean-variance optimization instead of ALM, sponsor's own preferences or pension fund's irrationality. Analyzing these factors would be a fruitful topic for further research. Additionally, international comparisons would be a fruitful topic for further investigation.

This paper presents evidence from the British Household Panel Survey (BHPS) on the persistency of contributions to individual, private pensions. The BHPS contains a wealth of information on individuals' socio-economic and demographic characteristics, allowing detailed analysis of how individual circumstances – and changes in those circumstances – can affect whether or not someone continues to contribute. The paper finds variation in persistency rates by gender, earnings, and household income. As might be expected, changes in income and consumption needs (for example, becoming unemployed or the arrival of a new baby) increase the probability of lapse, but, conditional on these changes, the level of household income also matters, suggesting that pensions may be less affordable for those on low incomes, even in the absence of shocks.

The independence of outside directors is critical to corporate board effectiveness. We examine a unique period in corporate governance when outside directors' defined benefit pensions are replaced with increases in equity. Firms with pension plans significantly underperform their industry in terms of stock returns. Firms terminating the pension plans in exchange for equity have significant increases in stock returns relative to their industry subsequent to the change. All samples outperform the ROA and ROE industry medians both before and after the change in compensation, indicating pressure from organized investors likely comes from stock performance, not accounting performance. Investor rights pressure and outside director compensation and not takeover risk or institutional ownership best explain firms altering outside director compensation, with board of director effectiveness improving.

We investigate the pension choices made by over 700 firms between 1981 and 1998 when DC plans expanded and overtook DB plans. Their average pension contribution per employee dropped in real terms from $2,140 in 1981 to $1,404 in 1998. At the same time, the share of their pension contributions attributed to defined contribution plans was 23% in 1981 and increased to 68% in 1998. By analyzing pension plan data from the IRS Form 5500 and finances of the plan's sponsoring employer from COMPUSTAT with a fixed-effects ordinary least squares model and a simultaneous model, we find that a 10% increase in the use of defined contribution plans (including 401(k) plans) reduces employer pension costs per worker by 1.7–3.5%. This suggests firms use DCs and 401(k)s to lower pension costs. Lower administrative expenses may also explain the popularity of DC plans. Although measuring a firm's pension cost per worker may be a crude way to judge a firm's commitment to pensions, this study suggests that firms that provide both a traditional defined benefit and a defined contribution plan are the most committed because they spend the most on pensions. Further research, especially case studies, is vital to understand employers' commitment to employment-based pension plans.

Increasingly, individuals are being required to take more responsibility for their own retirement saving. Lifecycle theories of resource allocation provide a framework to examine work, retirement, consumption, and saving decisions. However, optimal decision making requires adequate knowledge of financial mathematics, risk and return properties of investments, and expectations concerning wage growth and tax policy. This paper explores the response of individuals to financial education seminars. Using data from three surveys of participants in seminars offered by TIAA-CREF, we estimate changes in retirement goals and saving behavior after the respondents have attended a seminar which discusses keep components of saving for retirement. The results indicate that financial education can produce significant changes in how individuals think and plan for retirement. Throughout the analysis, women were found to be more responsive to the seminar and were more likely to raise their desired retirement age, increase their target income replacement goal, and alter their savings behavior.

Pension systems are complex and diverse, so comparing them is consequently difficult. Yet there are valuable lessons to be learned from the pension experiences of other nations. International comparisons of pension systems have focused almost wholly on the fiscal aspects of ageing populations. This paper provides consistent data on pension entitlements for the 30 member countries of the OECD, allowing cross-country analysis of the adequacy and distribution of pension promises.

The following section provides a brief description of the 30 retirement-income regimes and compares key parameters, such as pension eligibility ages, ceilings on pensionable earnings, contribution rates to defined-contribution schemes and accrual rates for defined-benefit plans. Section 2 sets out the methodology for and assumptions used in modelling pension entitlements. Section 3 presents replacement rates, the most commonly used indicator in pension analysis. Section 4 calculates pension wealth, the present value of the flow of pension entitlements, which captures the effects of cross-country differences in life expectancy, pension eligibility ages, and indexation of pensions in payment. The concept of weighted averages is introduced in Section 5: a way of summarizing the main indicators across the range of individual earnings. The structure of the pension package, that is the role that different components of the retirement-income system play, is explored in Section 6. Section 7 presents a measure of the progressivity of the pension benefit formula, while Section 8 concludes.

The aim of this paper is to explain the ‘annuities puzzle’ in greater depth by introducing the bequest motive. It will try to determine whether this motive really is a relevant feature influencing the demand for life annuities from married couples. With this aim in mind, we develop an optimization model of the utility provided by purchasing a life annuity with contingent survivor benefit or a joint survivor life annuity. Our model is based on that first put forward by Brown and Poterba (2000), to which we have added elements from other models, such as Friedman and Warshawsky's (1990) and Vidal and Lejárraga's (2004), which include the bequest motive. This will enable us to calculate the annuity equivalent wealth and the optimal percentage of wealth to annuitize in various contexts: the possibility of access to actuarially fair annuity markets, the inclusion of so-called market imperfections, and the assumption that couples already have part of their wealth in pre-existing life annuities. Numerical results are presented for the case of Spain. The bequest motive is found not to be a significant factor influencing the demand for annuities from couples. Indeed very few couples would be willing to purchase them once we take into account the combined effects of market imperfections, the possibility of pre-existing annuities and the bequest motive. These findings have repercussions for policy makers regulating defined contribution capitalization systems, which are complementary to defined benefit systems.

The present paper compares the distributional and risk-sharing consequences of two pension reform options in Germany, which both aim to improve the sustainability of the current system by introducing demographic variables to the benefit calculation. While the first reform proposes a so-called ‘sustainability factor’, which measures the changes in the dependency ratio, the second reform proposes a so-called ‘demographic factor’, which takes into account the changes in life expectancy. Our simulations indicate that both reforms imply a double burden for currently middle-aged generations and a double relief for future living generations. On the one side, resources are redistributed from currently towards future living generations. In addition, part of the risk from demographic uncertainty is shifted from future living towards currently living middle-aged generations. The reforms differ, however, with respect to the magnitude of the resource distribution and risk implications. Therefore, future generations are much better off with the ‘sustainability factor’, while it is not clear whether middle-aged generations are better off with the ‘demographic factor’ or the ‘sustainability factor’.

This paper evaluates the overall investment performance of Australian pension funds by examining the determinants of risk-adjusted performance, and the relationship between risk, returns, and expenses. Using quarterly return data for 225 pension funds comprising 68% of total prudentially regulated pension fund assets, we find significant differences exist across fund types. On both a net return and risk-adjusted performance basis, not-for-profit funds significantly outperformed for-profit funds over the seven years to June 2002. We suggest that the performance difference is consistent with the hypothesis that agency costs in for-profit funds (due to non-representative trustee board structures and potential board member conflicts of interest) are greater than agency costs in not-for-profit funds (with representative trustee boards).

Investors generally face inflation-linked obligations – a fact contributing to the popularity of TIPS and other inflation-linked bonds. With TIPS, one characterization of inflation, the Consumer Price Index, applies to all investors. Investors, however, face different inflation. To date, these heterogeneous needs have not been addressed by the inflation-linked marketplace. The paper describes the case for and mechanics of splitting TIPS into disaggregated TIPS matched to components of the Consumer Price Index. Disaggregated TIPS better address the risks of investors' specific real liabilities. A case study highlights disaggregated TIPS applicability to retirees with heavier post-retirement medical needs.

Government-sponsored inquiries into trustee competence, and legislation regarding the protocols and practice of trustee decision making, have raised questions about the competence of trustees to make investment decisions consistent with the long-term interest of defined benefit pension plan beneficiaries. In this paper, we report the results of an analysis of trustee competence in solving problems relevant to their investment responsibilities. Based upon a set of widely recognized problems drawn from the psychology literature, we assess their discount functions, their willingness to risk their own money and others' money, their appreciation of probability, and their use of evidence to solve problems. For comparison, where appropriate we report the results of the same testing regime applied to a group of Oxford undergraduates. Our goals are fourfold: first, to demonstrate the nature of trustee competence in decision making; second, to demonstrate the range of trustee responses to problems relevant to investment; third, to assess trustees' risk appetites in relation to their own and others' money; and fourth, to draw implications from these results for the governance of trustee boards and their relationships with advisers and service providers. It is shown that trustee competence is surprisingly heterogeneous, and the lack of common approaches to problems relevant to investment practice has significant implications for fund governance.