Rapidly aging populations are forcing policymakers to rethink pensions. Defined contribution schemes are quickly becoming the norm, but people don’t yet seem ready for them.
Academic evidence from economics and behavioural finance suggest that under many circumstances people may not be ready to take control of retirement savings investment. This is a challenging finding in a context where traditional defined benefit (DB) schemes, (which typically depend on individual salary history and length of service) are waning in favour of defined contribution (DC) schemes, where people can have a say on how to diversify retirement assets and future benefits. Pensions will come to depend on the accumulation of such assets and the returns they get.
The asset allocation decisions made by contributors to pension funds often seem naïve, which raises concerns about DC pension schemes because it is the individual who faces the investment risk (i.e. if the pot loses money, then the person will receive a lower pension). Since the seminal research work by Shlomo Benartzi from UCLA and Richard Thaler from the University of Chicago in the late 1990s, different scientific papers (many focused on the U.S. pension system) show that people’s investment decisions are often driven by rules of thumb or mental accounting. For example, many individuals will follow a “1/n strategy” by dividing their contributions evenly across the different funds offered in the pensions plan. Sometimes asset allocations can be extreme (either 100 percent or zero percent in equities) and there is inertia in asset allocations.
The number and characteristics of investment options on offer also influence the choices people make. The presence of a larger number of investment options has been associated with greater allocation to money market and bond funds at the expense of equity funds, which may hurt long-run returns. Large choice sets induce a preference for simpler menus of investment options. The number of funds used by a person is usually much lower than the number of funds offered by the pension plans, which suggests a degree of information overloading. The composition of the menu (and how the investment options are framed) also matters. If the menu offers more investment options that are intensive in equity, then people tend to demand more equity.
Despite the seemingly naïve investment decisions, simulations of wealth at retirement applied to both DC and DB schemes have shown that average retirement wealth accruals under current DC plans can exceed average accruals under private sector DB plans. Also, a paper published in the Journal of Public Economics finds that people who face investment choice are more likely to make annual contributions, they invest more in stocks and have accumulated more money in their accounts than comparable participants without investment choice. In short, people seem to benefit from DC schemes and from access to investment choice.
What is the problem then? Perhaps the biggest challenge in DC schemes is that the evidence usually describes people as “reluctant investors” who will opt for not making any decision at all. Different surveys conducted by academics on pension attitudes show that people often consider themselves financially unsophisticated and are reluctant to take control of retirement savings investment, even when offered the possibility to improve their financial education. People often do not have well-defined preferences and would opt for portfolios chosen by other people instead of making an independent decision.
The low extent of active choice in the industry of private pensions has been documented for some time. The large percentage of people whose money is invested in default funds (i.e. those funds where money is invested if people make no decisions) is generally taken as evidence that individuals are either incapable of making, or unwilling to make, decisions concerning the investment of their pensions. This is often called the default option bias. A paper released in 2001 by the National Bureau of Economic Research (NBER) showed that over 65 percent of 401,000 plan members were enrolled in default funds. More recent academic and policy papers corroborate that low active participation still prevails. In the U.K., a survey of Towers Watson reports that over half of FTSE 100 DC pension schemes had more than 80 percent of their membership in the default investment funds.
The size of the default bias seem to suggest that people are not yet ready to get involved in the administration of their pension assets. Although research suggests that the default bias may slightly decline over time as people get more tenure in the system, the design of an optimal default fund is a key aspect of pension funds. And by ‘optimal’ we mean funds that maximise returns on retirement money and minimise risk. Several mechanisms have been proposed, particularly the adoption and proliferation of life-cycle funds that become increasingly conservative as the pension plan member approaches retirement. However, do we really want an automatic mechanism that substitutes stocks with low return assets just when both people’s incomes (and thus contributions) and their accumulated assets in the pension fund are at their highest? There are still plenty of questions to address regarding investment choice and the design of default funds which I will address in future posts.
Eduardo Rodriguez-Montemayor is a Senior Research Fellow at INSEAD. He writes for INSEAD’s Work and Pensions blog.