But wait, there's more! In addition to these very real problems, 401(k) plans are generally run by the asset management industry, which (surprise!) does not always have your interests at heart.

Most defined contribution plans allow participants to select from a menu of investment options, largely mutual funds. The question is, who decides what's on the menu? About three-quarters of the time, these decisions are made by a mutual fund company acting as the plan's trustee.** And (surprise again!) mutual fund companies are more likely to push their own funds onto employees than other companies' funds--despite the fact that they are legally obligated to act in the best interests of plan participants.

Although many people have suspected this all along, we now have convincing evidence from a paper by Veronika Pool, Clemens Sialm, and Irina Stefanescu aptly titled "It Pays To Set the Menu" (which was sent to me by two different readers). The paper uses a clever empirical approach. In any given year, a mutual fund may be included in 401(k) plans overseen by that fund's sponsoring company (e.g., the Fidelity Magellan Fund may be included in a plan whose trustee is Fidelity) and also in other plans not overseen by that company. It turns out that there are many such funds.

Let's say such a fund has a bad year. If plan trustees are acting solely in the interests of their participants, we would expect the identity of the trustee not to affect the chances that the fund is dropped from the investment menu. But that's not the case: a poorly-performing fund is much less likely to be dropped from a menu controlled by its sponsoring fund company than from a menu controlled by a third party. Fund companies are also much more likely to add their poorly performing funds to plan menus that they control.

But maybe the fund companies are keeping their poorly-performing funds because they know that they are likely to do better in the future. After all, past performance is no guarantee of future results, and maybe they are anticipating mean reversion. Not so much. Those poorly-performing funds that fund companies keep on plan menus continue to do worse than other funds--by 3.6% per year, after adjusting for risk. Finally, we can't count on individual rational actors to bail us out here: plan participants do not correct for fund companies' bias by proactively shifting their money out of these poorly-performing funds.

This is a special and particularly maddening example of the problem that the vast bulk of our retirement savings, other than Social Security, are funneled through the asset management industry, which charges often-hefty fees to manage our money, despite its proven inability to beat the market. (Theoretically speaking, since the asset management industry is most of the market, in aggregate it should match the market--before fees.) This would be a necessary inconvenience if competition drove fees down to reasonable levels, like the 6 to 20 basis points that Vanguard charges to invest in just about any major segment of the global stock market. But instead, structural factors, like mutual fund companies' control over investment options in DC plans, take the edge off of competition, allowing these rents to persist. (As of 2009, domestic stock mutual funds in 401(k) plans, in aggregate, charged 74 basis points in expenses--and cut into returns further with transaction costs.)