Marco Rubio wasn't the only one who cashed out an IRA last year

It is easy to mock Senator Marco Rubio, who cashed out $68,241 in IRA retirement funds last September. The GOP presidential hopeful, who made about $230,000 last year, told Fox News he needed the dough to prepare for his campaign, buy a new $3,000 refrigerator, and fix his busted a/c. As it happens, he is far from alone.

Using retirement money to buy a high-end appliance may not be the wisest financial decision. Investing the cash in a presidential run may be more prudent. After all, as the Clintons have shown, post-White House money-making opportunities can be quite lucrative. And even also-rans can turn failed races into money-making career enhancers. Just ask Mike Huckabee.

But it turns out that Rubio is not the only 40-something draining a retirement account. Substantial assets leak because people under age 59 ½ take early withdrawals or borrow against their IRAs or 401(k). And the problem raises an important and challenging policy question:  Should the money in these accounts be available for non-retirement purposes?

Exactly how much is lost is unclear, but it is not a trivial sum. My Urban Institute colleagues Barbara Butrica, Sheila Zedlewski, and Philip Issa estimate that more than 8 percent of working-age retirement account owners made at least one withdrawal between 2004 and 2005. And they pulled out about 20 percent of their total balances.

Overall, at least 1.5 percent of all retirement savings leaks out of the system each year, although some studies estimate it is much more than that. And even withdrawals at the low end of the estimates can result in a substantial reduction in retirement savings.According to Alicia Munnell and Tony Webb at the Center for Retirement Research at Boston College, early withdrawals of about 1.5 percent ultimately reduce total IRA and 401(k) wealth at retirement by about one-quarter.

In most respects, Rubio is atypical. Unlike many who take money out early, he’ll still probably have a secure retirement. The Urban study found that those most likely to withdraw early were age 25-34, had low-incomes, and had little net worth. African-Americans are more likely to withdraw than whites. And those who pull out money when they are young can significantly lower their standard of living in retirement.

Interestingly, several of the studies found that divorce, death of a spouse, or job loss correlate strongly to withdrawals. Not surprisingly, these shocks have their biggest effects on low-income workers.

In 2013, Robert Argento, Victoria Bryant, and John Sabelhaus of the Federal Reservelooked that how the Great Recession affected withdrawals. After all, if income shocks increase withdrawals, you’d think the effects would be especially strong when millions of Americans lost their jobs.

It turned out that while early withdrawals were higher after 2007 than before, they were not that much higher. And they seemed to merely continue a long-standing national pattern of steady increases. Still, in 2010, according to the Fed study, nearly half the value of new contributions by working-age households were offset by early withdrawals.

What to do? How do policymakers distinguish between low-income people who dip into retirement savings to cushion a financial shock and someone earning a six-figure income who wants a new Sub-Zero? How does government encourage retirement savings—the purpose of the tax subsidy, after all—and still give workers some financial flexibility when they really need it?

Early withdrawals are not a cheap way to get money. They are subject to both tax and, in most cases, those younger than 59 ½ also owe a 10 percent penalty.  In addition, people of working age can only pull money from 401(k)s when they change jobs or for reasons of hardship (for instance, for medical or funeral expenses or to prevent foreclosure or eviction). IRA withdrawals can be made for any reason, though they are still subject to taxes and usually penalties.

Should the rules be tougher?

We know that even modest policy adjustments can change behavior. In a 2012 National Tax Journal article, my Tax Policy Center colleagues Len Burman and Bill Gale, along with Norma Coe and Michael Dworsky, looked at how policy changes in 1986 and 1992 encouraged people to hang on to their retirement accounts when they changed jobs. They found that raising taxes on cash-outs encouraged people to roll over their accounts to IRAs instead of pocketing the money. Merely relabeling the added tax as a penalty or withholding taxes on cash-outs also increased roll-overs, though neither changed the amount of tax owed.

Munnell and Webb think it’s time for Congress to reconsider tax-advantaged savings: Should they be focused entirely on boosting retirement assets or should they play the dual role of encouraging savings for retirement and for unexpected hardship at working age?

It is a good question. And the personal experience of Marco Rubio—who, after all, has proposed a major tax reform of his own--may prove a useful spark for the conversation.