Tuesday, July 14, 2015

Private Savings Accounts versus Social Security

by Neil H. Buchanan

In my June 25 post here on Dorf on Law, I discussed the realistic options available for financing retirement, in the current and future U.S. economy.  The old three-legged stool -- defined-benefit pensions from employers, Social Security benefits, and private savings -- is now down to a very unstable two legs.  As I explained, the disappearance of pensions provided by private employers is reversible in theory, but now that people are no longer working for one employer per lifetime, it would be quite difficult and expensive to do so.  Furthermore, the most plausible strategy to work around the multiple-employer problem (requiring employers to contribute to a fund that disburses benefits to employees) looks like a needlessly complicated version of Social Security -- or, if it was not needlessly complicated, it would simply be another version of Social Security.

One conclusion, then, is that we could decide to enhance Social Security, expanding it to cover two-thirds of the old approach.  This is, in fact, a serious option that was recently proposed by now-retired Senator Tom Harkin, and which Senator Elizabeth Warren has championed in his absence.  Given the current political alignments in the U.S., of course, this is a non-starter.  Still, it is worth discussing the Harkin plan, and I plan to do so in a Verdict column and companion Dorf on Law post later this year (perhaps as early as next month).  Here, however, I want to return to the third leg of the stool, private savings, to discuss some interesting and important questions that are implicated by my June 25 post.

It would, of course, be possible to replace employer-provided pensions with private savings accounts, rather than by expanding Social Security.  And we have, in fact, gone a long way down that road.  What is euphemistically called a "defined-contribution pension" is, in fact, a savings account that is accompanied by various legal restrictions and saver protections.  We have created various tax incentives to encourage people to park their money in such accounts, with various limitations on total tax-free contributions, and so on.  Once one has used all of these available options, it is still possible to use the pure private saving option (if one's income can support that much total saving) by putting even more money in regular savings accounts, mutual funds, real estate, rare art, collectibles, a new business venture, and so on.

What makes defined-contribution plans different is that they cannot be cashed out early, except under very specific exceptions, and that the range of investment options is legally circumscribed.  Even so, I wrote in my June 25 post: "Research over the course of decades demonstrates over and over again that people are myopic, that they are easily confused by too many options, that private investment companies charge excessive management fees and generally figure out ways to divert money from savers, and so on."

On the comment board for that post -- in addition to yet another wonderful testimonial for a "spell caster" -- a reader questioned my dismissive attitude toward defined-contribution plans.  Because of the problems that I described in the quotation above, I had noted that such plans would need to be heavily regulated, to which the reader responded by asking whether I was referring to 401(k) and 403(b) accounts, adding: "Fortunately retirement plans are carefully regulated. ...  ERISA imposes heavy legal obligations on plan fiduciaries to select suitable investment options, monitor those investment options, limit plan fees, etc."

Yes, it is absolutely true that plenty of people invest in 401(k) and 403(b) accounts.  Indeed, because I have been an academic for my entire professional career (except for one year as a judicial clerk), my retirement savings consist of nothing but 403(b) accounts.  I hope and expect that they will be there when I am ready to retire in 50 or 60 years, but that depends on whether the financial system continues to function adequately, and on the continued commitment of the government to prevent financial institutions from plundering depositors' accounts.  Those seem like safe bets, but there are no guarantees.
 ERISA is, in fact, the cornerstone of the government's continued commitment to prevent financial institutions from plundering depositors' accounts.  As I noted above, ERISA imposes various legal rules on financial institutions to protect individual investors.  Does it do enough?  I have heard more than one person say that ERISA law is more complicated than tax law or even habeas law, but being difficult to practice does not necessarily tell us what we need to know about whether the law works as planned.  It does suggest that there is room for shenanigans by the sophisticated repeat players (that is, not you and me).  Even without a profit motive by the big players, complicated areas of law can be tilted against individuals, in fact if not by design.  For example, Social Security disability law is by all accounts a mess that appears often to be slanted against applicants.  Notably, that is quite different from Social Security retirement benefits (which are simple and automatic).

Imagine, in any case, that ERISA works perfectly, or as close to perfectly as is reasonably plausible.  That would simply mean that the law is enforced in a way that prevents people from being defrauded or from foolishly draining their accounts prior to retirement to buy a boat or to fritter away at the local casino.  When I wrote about the decades of research regarding savers' behavior, however, the possibility of fraud on the part of financial institutions was not the real point.  Saying "that private investment companies charge excessive management fees and generally figure out ways to divert money from savers, and so on," is not to say that they have violated ERISA, or any other law, but that they exist to make profits, and there are legal ways to do so.  Those profitable strategies must, as a matter of arithmetic, work at the expense of the depositors.

Now, one could argue that there is an invisible hand-like added surplus that is created here, with financial companies able to generate returns that more than justify their fees.  However, as I noted in my June 25 post, Social Security's returns can be set to mimic the average returns on the market, because the general prosperity of the economy generates the revenues necessary to finance Social Security benefits.  And because Social Security's administrative costs are minuscule, the average investor will be worse off with her net returns from even a well-managed 401(k)/403(b) fund than she would be with a well-managed expanded Social Security plan.

I am, however, very comfortable conceding that some individual investors could beat the market.  In fact, because there are standard deviations around the averages, there will almost always be some supra-normal returns to which one can point.  Over the course of a working lifetime, however, that will go away.  The guy who managed Yale University's endowment in the 1990's and 2000's (if I recall correctly) became a financial rock star because he seemed to have figured out how to defy the laws of gravity.  If ever there were an exception that proved a rule, however, that guy is it.

Of course, Social Security's benefit structure is also progressive, which means that people with higher incomes during their working lives earn a lower implicit rate of return than do people who have lower earnings.  That, however, is not an argument that private investors in the aggregate could beat the averages, but only that some higher-income people might prefer to opt out of a progressive system.  That is hardly news.

Finally, and most fundamentally, those decades of research uniformly indicate that most people are likely to make decisions that under-perform the market, in ways that the protections of ERISA and similar laws are not designed to discourage.  These unfortunate choices are not, moreover, a result of being stupid or lazy.  In this past Sunday's New York Times, an economist at Harvard named Sendhil Mullainathan wrote a fascinating article summarizing recent research about people's systematically poor decision-making in investing their retirement funds.

The most interesting aspect of the column was, however, not the systematic research that Mullainathan described, but rather his personal confessions of financial foolishness.  He admitted that he had been putting off even thinking about his retirement portfolio for years, and that he had somehow defaulted into an investment portfolio that literally paid nothing on some of his deposits.  He was so paralyzed by freedom of choice that he -- a tenured professor at the top economics department in the world, and a MacArthur Fellow to boot -- had passively decided for years on end to think about it later.

He also pointed out that the current incentives in the marketplace encourage fund managers to take excessive risks.  The best option for the people who rely on those fund managers?  Put one's money into funds that are diversified.  "But the market has also responded by charging high fees for this standardized product.  [I]ndex funds had as much variability in fees as their more labor-intensive actively managed counterparts. And these fees are nothing to be scoffed at — paying 1 percent more every single year in fees can compound over a lifetime to noticeably lower returns."  And Mullainathan does not even mention the troubling statistical regularity that holders of 401(k)'s systematically have over-invested in the stock of their employers.

There is nothing illegal about any of this, and that is the point.  Replacing traditional pension plans with highly regulated retirement savings accounts is better than replacing them with completely unregulated savings.  The more one looks at the experience of these types of investments, however, the more one is tempted to say, "Well, we could add a legal requirement o prevent this or that kind of myopic behavior, or to limit fees that fund managers can charge."  The more one moves in that direction, however, the more obvious it becomes that we are doing everything possible to try to mimic what Social Security already does.  Adding all of those protections, however, adds to the costs of the system, and it still does nothing to eliminate the "cut" that the fund managers must earn.

In an important way, therefore, the choice between private retirement savings and Social Security looks a lot like the choice between private health insurance and Medicare/single payer.  There are ways to make the private route less bad, but they are inherently expensive and wasteful (see, e.g., the Affordable Care Act).  Hence my dismissiveness about "pseudo-market alternatives."  As noted above, in an upcoming column and post, I will explain how we could expand Social Security.


Michael C. Dorf said...

W/r/t a comparison you draw at the end: I think the case for expanding Soc Sec is even stronger than the case for single payer medical care. Medicare, which is our version of single payer, has low administrative costs but high total costs, because it pays doctors and hospitals on a fee-per-service basis, which leaves them with incentives to provide expensive care and attracts fraudsters. By contrast, Social Security checks are cut directly to the individuals to whom the money is due, so it has low administrative costs AND low overuse/fraud rates. (As in the post, I'm talking about Soc Sec, not SSI disability). There is a way to avoid these problems on the health side too, which is not just to go to Canadian-style single payer, but to UK-style national health. However, that really would be "socialized medicine," and thus politically impossible in the US.

To be clear, I'm not arguing against single-payer health insurance. I think Medicare for All would be, all things considered, superior to our current system. My point is simply that single-payer health insurance (on the Medicare/Canadian model) has drawbacks that expanded Soc Sec for retirement does not.

David Ricardo said...

Two points.

First, although somewhat off the subject of the main topic here, with respect to Mr. Dorf’s comment

“There is a way to avoid these problems on the health side too, which is not just to go to Canadian-style single payer, but to UK-style national health. However, that really would be "socialized medicine," and thus politically impossible in the US.”

there is an alternative within the health care system that would move the U. S. away from the fee for service economics of health care and not be a UK style government operated health care system.

While the details of such a system are too large for discussion on a Forum like this, a good example where this is happening is where medical centers are purchasing physician practices and providing health insurance and where health insurers are purchasing medical practices/hospitals and providing direct services on a fixed cost basis. If a medical center, for example, is getting fixed payments per Medicare recipient then its incentives are to reduce costs in large part by prevention rather than treatment and by increasing efficiency. This is probably the future of American health care, and many aspects of such a system are already in place such as incentives in Medicare for more efficient treatment, but like much of what happens in America it is likely to be a difficult evolutionary process rather than an efficient planned process.

Those of us with expertise in medical economics know what to do to create a health care system in the U. S. that provides quality care with the right incentives that will reduce costs, there is simply not the will, the understanding or the political power to do such a thing.

Second, one problem with retirement programs such as Sec. 401 type plans is that they allow participants to borrow from their account. As a result many employees, particularly those with lower income will ‘raid’ their retirement accounts, often out of financial necessity but sometimes for non-necessary purchases. The result is that even though they may have contributed to such a plan for decades, when they retire there is very little to retire on. Social Security does not have this provision and this is one argument for its expansion as opposed to expanding Sec. 401 type plans. It will be interesting to see what Mr. Buchanan comes up with.