How Much Is Investor Autonomy Worth?

There is a worldwide trend towards defined contribution savings plans, where investors are often able to select their own portfolios. How much is this freedom of choice worth? We present retirement investors with information about the distribution of outcomes they could expect to obtain from the portfolios they picked for themselves, and the same information for the median portfolio selected by their peers. A majority of our survey participants actually prefer the median portfolio to the one they picked for themselves. We investigate various explanations for these findings and offer some evidence that the results are partly attributable to the fact that investors do not have well-defined preferences. A major trend in defined contribution savings plans is the expansion in the choices available to participants. A decade ago most plans offered very few choices, often just a money market fund, a bond fund, a stock fund, and stock in the company sponsoring the plan. Now the plans offer an average of 11 funds (Hewitt Associates (1999)), and some plan participants are even permitted to pick individual securities through a direct brokerage account. And, in Sweden, a recent social security reform giving workers the right to direct 2.5 percent of their salary to individual accounts, offered a stunning 450 different funds to choose from. Do investors benefit from this ability to choose portfolios for themselves? It is a basic principle of economic theory that expanding the choice set cannot make a consumer worse off (at least ignoring decision-making costs). Here, where the financial stakes are quite high, and choices are made infrequently, many would argue that more choices are unambiguously a good thing. Still, choice comes at a cost. For example, in discussing various options for a fully or partially privatized social security system, an important design issue is how much choice to offer participants. Diamond (forthcoming) estimates that administering individual accounts with even just a limited set of investment choices will cost between $40 and $50 per participant per year. With the commonly proposed deferral rate of two percent of income earned, typical individual account balances will be negligible for at least a few years. Consequently, Diamond estimates that the administrative costs will be higher than the investment gains for some time. To minimize administrative costs, some have proposed that individual accounts will initially be invested in a single fund (perhaps a balanced fund that divides money between diversified portfolios of stocks and bonds), with choices only introduced once balances grow. More generally, adding choices to 401(k) and 403(b) plans also increases the costs of administering these plans. Do participants gain from this expansion of their choice set, and if so,