Let's talk about ditching 401(k)s and starting over

(Bloomberg View) – There have been some modest attempts in the past few years to shore up and fill in gaps in this country's haphazard system of retirement savings. These attempts -- federal myRA accounts, state-run retirement accounts, and fiduciary standards for retirement account providers -- have also all been shut down, rolled back or delayed since Donald Trump moved into the White House a year ago.

So it may seem like a curious time to talk about tossing aside the current system of 401(k)s and individual retirement accounts in favor of an entirely new retirement savings system, as Teresa Ghilarducci and Tony James do in their book, "Rescuing Retirement: A Plan to Guarantee Retirement Security for All Americans," which after an earlier iteration in 2016 is now out in hardcover. But hey, since the little stuff isn't getting anywhere, why not go big?

Ghilarducci is an economics professor at the New School for Social Research in New York who became kinda-sorta famous after she told a House subcommittee in October 2008 that 401(k)s gave big tax breaks to the affluent while not helping most Americans save adequately for retirement -- which Rush Limbaugh, who went on to spend years harping on her ideas, translated as "she wants to basically eliminate the 401(k)." James is the president and chief operating officer of private equity giant Blackstone Group LP, as well as nonexecutive chairman of the board at Costco Wholesale Corp.

401k-401(k)-retirement

401The two unveiled their unlikely buddy act in 2015. The plan they have developed would supplant 401(k)s and individual retirement accounts with mandatory Guaranteed Retirement Accounts that have these defining characteristics:

Workers and employers would each contribute 1.5% of wages or salary, maxing out at $3,750 each (1.5% of a $250,000 salary); the self-employed would contribute 3%. Everyone contributing to a GRA would get a $600 refundable tax credit, meaning that for low-income workers, there would be no out-of-pocket cost to the accounts. On top of that, there would be a tax deduction for contributions maxing out at $3,750. Workers could still contribute to GRAs up to the current 401(k) limits of $18,500 and $24,500 and enjoy tax-deferred gains on those contributions; they just wouldn't get the upfront tax deduction. Workers would own their GRA accounts and choose a pension manager to invest the money. But they wouldn't be able to withdraw money from the accounts before retirement. Upon retirement, the accounts would be converted into annuities, with monthly payments calculated and administered by the Social Security Administration. These payments would also be guaranteed by the federal government. This guarantee would most likely never come into play, Ghilarducci and James argue, but the government could at very low cost create a fund to insure against this worst-case scenario. Also, account holders could choose not to annuitize up to 25%. of the account, but only if the annuity on the remaining amount would, along with Social Security, keep them above the poverty line.

This setup is similar in some ways to the "superannuation" system set up in Australia in the early 1990s, which has given the country what is now considered one of the world's best retirement savings systems (although, as my fellow Bloomberg View columnist Satyajit Das wrote last fall, even the best isn't perfect). The two key differences are that the GRA contributions would be much lower than the 9.5% of income currently required in superannuation (which is scheduled to rise to 12% by 2025) and that with GRAs, annuitization would be mandatory. Much of the difference between the contribution rates can be explained by the existence in the U.S. of Social Security, which would be left untouched by the Ghilarducci-James plan, already imposes a payroll tax of 12.4% (half paid by the employer, half by the employee), and provides more generous benefits than Australia's government-funded, means-tested Age Pension. But requiring annuitization also reduces the amount of money that needs to be saved.

That's because it costs a lot less per person to guarantee an adequate lifetime income for an entire generation than for an individual. When providing an annuity for millions of people, you can take the money saved when recipients die earlier than expected and use it to finance the retirements of those who live longer than expected. It's a morbid and in some ways inequitable wealth transfer, given that affluent people tend to live longer than poor people, but it costs a lot less than everybody saving for themselves. Americans have actually set aside far more money for retirement as a share of gross domestic product than the residents of most wealthy countries, but the general lack of annuitization means all that saving may not prove adequate for many.

Three percent of income is also a lot less than 10% of income or more that financial advice givers in the U.S. often say people should set aside for retirement. Annuitization plays a role here, of course. Also, those with higher incomes will need to set aside more than 3%, since Social Security replaces a smaller percentage of their incomes. Ghilarducci and James simply figure -- and there's recent economic research to back this up -- that affluent people don't need tax deductions to get them to save for retirement. Finally, they assume that the pension managers in the GRA system would get much higher returns than most people do now in their 401(k)s and IRAs. This is not a baseless assumption:

Defined-benefit pension funds get a lot of bad press, some of it deserved. But when a state pension fund runs into trouble these days, it's usually because elected officials have promised too much and/or failed to set aside enough money, not because the fund managers have blown it. In the Ghilarducci-James plan, "state pension funds, traditional money managers, or a federal entity such as the Thrift Savings Plan" could bid to manage GRAs. These pension managers would get higher returns than existing 401(k)s and IRAs do, they figure, because (1) they're more sophisticated investors than most 401(k) and IRA owners, (2) they'll be able to invest in illiquid assets such as infrastructure and private equity that are generally off-limits in 401(k)s and IRAs, and (3) their administrative costs would be lower, as would their fees.

No, none of this is going to come to pass anytime soon. And even over the longer run, incremental improvements to the current system seem a lot likelier than a massive overhaul such as the GRA. But the Ghilarducci-James plan raises several issues that really ought to be a part of any national discussion about retirement. One is annuitization, which I've been harping on for years now. Another is the wisdom (or lack thereof) of leaving individuals in charge of all their retirement saving, asset allocation and investment decisions, which has been addressed by some 401(k) sponsors in recent years (with default contributions into target-date funds, among other things) but is still a big issue for the IRAs that most 401(k) accounts eventually get rolled into. Finally, there's the question of whether the more than $200 billion in annual tax breaks for retirement savings should mainly benefit the top 20% of the income distribution, as is now the case, or be targeted to make it more likely that everybody has an adequate retirement income.