For an entirely predictable event, retirement is going to come as a shock to tens of millions of Americans — a financial shock, that is. Many people in their 50s and 60s are about to find that the money they've set aside for retirement is too meager to support the standard of living they'd hoped for. Soon, the long-expected U.S. retirement crisis will no longer be a forecast; it's becoming a brutal fact.

For those at the leading edge of the baby boom, the prospect of an extended retirement with limited means leaves few options. Many will have to work longer than they'd wished, in jobs they don't like. Many will sell homes they'd hoped to keep — to tap the equity, such as it is. Household budgets will be squeezed, then squeezed again. Workers further from retirement, though, still have time to address the problem — and public policy must start encouraging them to do so.

The retirement-savings deficit has grown so large partly because U.S. employers have stopped providing adequate pensions. Between 1980 and 2006, the share of U.S. private-sector workers covered by an employer's defined-benefit plan fell from an already low 40 percent to just 15 percent. A savings vehicle intended to supplement traditional pensions — the defined — contribution 401(k) retirement plan — has instead replaced them. This diminished the stock of retirement savings in itself; now, many companies are scaling back their contributions to 401(k) plans, too. Roughly half of all Americans have no private retirement savings at all.

There's Social Security, but on its own it isn't nearly enough. It will replace too small a proportion of most Americans' incomes after they retire. It's sufficient to avoid poverty in old age, but not much more. One way or another, private saving for retirement needs to rise.

In January, President Barack Obama offered a modest step forward when he ordered the Treasury to create low-cost, government-run MyRA accounts for low-wage workers who lack access to employer plans. It's good that the White House recognizes the problem, but the MyRA really isn't much use. Workers will have to opt in, and most of them probably won't bother: The investment will give only a modest return and, for the low-paid, the MyRA's tax advantages are beside the point.

A better way would be to promote auto-enroll savings plans: Instead of opting in, workers would have to opt out. As behavioral economists have shown, that would make a big difference. All the evidence says that most workers would stay with the program.

Last month, Democratic Sen. Tom Harkin of Iowa introduced legislation to this effect. It would automatically enroll workers not already covered by a retirement plan into portable, tax-sheltered savings plans run by private managers and overseen by the government. Contributions would be set at 6 percent of pay up to a ceiling. This would build a fund which, together with Social Security, would allow a middle-income couple to retire with 80 percent of its pre-retirement income — about enough to maintain its standard of living.

Harkin's proposal can be improved. It lets workers opt out or vary their contribution rate; in the same spirit of flexibility, savers should also be allowed to choose their own investments. Once the program was up and running, it would be a good platform for doing more. A matching contribution from taxpayers at large to workers on low incomes would be worth considering.

The main drawback of this or any similar initiative is that it would further complicate a labyrinthine system of tax preferences. Savers in the United States are already blessed with 401(k)s (traditional or Roth), 403(b)s, 457s, TSPs, Individual Retirement Accounts (traditional or Roth), SIMPLE ("Savings Incentive Match Plan for Employees") IRAs, SEP ("Simplified Employee Pension") plans and more — not to mention the new MyRAs — all with their own rules and conditions, which frequently defy understanding.

So many plans, so little effect on financial security. That's partly because these aren't auto-enroll programs: Savers have to opt in. For many, the complexity is surely paralyzing: The policies offer incentives to save, but you need an accountant to guide you through the options. Most important, these programs weren't designed with low-wage workers in mind.

It's a ridiculous mess. Harkin's proposal is a good one, but it should ideally be adopted as part of a thorough simplification of incentives for saving — one that gives subsidies and tax preferences to the savers who need them most and, as income rises, withdraws this support at a measured and deliberate pace.

Comprehensive reform of anything is a lot to ask these days — especially when it requires foresight, because the worst of the problem lies ahead. Failure to confront this issue in the past has already condemned many to a severely pinched retirement. For the rest, it isn't too late.