I recently came across a rather glib press release trumpeting the 25th anniversary of the 401(k) plan. It noted that 47 million Americans now actively participate in 401(k) plans, compared with 21 million who participate in traditional defined benefit plans. Essentially, we are well on our way to a wholesale shift of risk from large entities to the individual ... the magical ownership society … where *presto poof* we kill (or maim) three birds with one stone: 1) nettlesome pension deficits disappear; 2) 1,000 points of ownership light blink on; and, 3) financial middlemen enjoy an unprecedented boon.

Call me a skeptic, but I’ve met a lot of individual investors and financial advisers, and the level of knowledge, even among those who are paying any attention, is … well … appalling. In the 401(k) business, you end up with a potent cocktail of investor cluelessness and (generally) a horror-show lineup of investment options. It adds up to an old joke I remember from when I was growing up in rural Ohio: “He may be small, but he sure is slow.”

The 401(k) plan, or something like it, could actually work. But what we have right now is a bloated and inefficient system that forces investors to pay more than they should for investment options that aren’t what they ought to be. The way that mutual fund companies cling to 401(k) plans as massive distribution machines locks investors into a limited choice of generally high-expense-ratio, high-fee plans, putting many investors in a situation where they might do better in a taxable account.

The biggest cheerleader for the 401(k) is the Investment Company Institute (ICI). The ICI is delighted with how the 401(k) market has developed; as its benefactor, the mutual fund industry, has never lived better. Today, 54 million U.S. households are invested in mutual funds, at a median level of $48,000 and an average annual expense ratio of over 1 percent for equities funds.

I’m sure you’ve all seen the ICI pronouncements about how expense ratios have gone down over the years. And they have: The average equity fund charged an ungodly 2.19 percent in 1985, and “just” 1.13 percent in 2005.

But as John Bogle has noted, that drop in fees should be measured against the explosion in mutual fund assets—an explosion driven largely by the 401(k) boom. Mutual fund assets are up more than 1,500 percent since 1985, from $602 billion to $9.5 trillion. So while expense ratios have fallen, total fees have skyrocketed. In 1985, when the industry was gouging us with 2+ percent fees, mutual fund companies took home $12 billion in expense payments; today, with the new “low! low!” fees of 1+ percent, the register rings (and rings and rings) to the tune of $95 billion per year. So yes, Virginia, fees have fallen. They BETTER have … that’s what we call scale.

Unfortunately, the reduction in scale is about four times what it ought to be by one very rough measure. If you run the inflation calculator on the $12 billion in fees collected in 1985, it is the equivalent of $21 billion in 2005 dollars. Sure, the growth in equitization and assets is worth something in maintenance costs. But is it worth an industry that is nearly five times bigger in real terms?

The 401(k) program is significantly at fault. A system in which most—not just many—investors don’t even have options with expense ratios under 1 percent, and where investors generally pay chunky, built-in maintenance fees on top of that, is a broken system. The recent reforms are unfortunately too little, too late. We have worked ourselves into a morass where increasingly well-funded interests are essentially writing the laws governing retirement savings.

We would favor regulations that would force every plan to have low-fee options. Above all, we would favor legislation opening up all 401(k) platforms to competition. And if we ever take the Social Security system private, the plans need to default to 1) plan participation, and 2) investment in a low-cost, broadly diversified fund. One of the few things the government has actually done well—the federal Thrift savings plan—could be used as a model.

Right now, all we’re doing is building on 25 years of mediocrity.