Personal Retirement Accounts: Should They be Voluntary or Mandatory?

Share this page



Participation in Social Security is mandatory.  If a worker is employed in a job that is “covered” by the system, he or she must pay the Social Security tax.  This has been a requirement of the system since its inception.  The premise was that it had to be mandatory if it was to be successful in keeping the greatest number of people from becoming “dependent”–i.e., needing to be on welfare–in their old age.  A voluntary system would have been subject to gaming.  People who failed to join, thus evading payment of their “dues”, could eventually become dependent on those who did, creating a fundamental inequity.


President Bush and a number of Members of Congress have proposed that Social Security taxpayers be given a choice to invest part of their Social Security taxes in personal retirement accounts.  Although such a reform would not turn Social Security into a completely voluntary system–people would be required to either pay the FICA taxes or make equivalent deposits in their new accounts–the idea of a voluntary accounts option nonetheless carries many of the same vulnerabilities for the nation’s largest senior safety net as having a completely voluntary system.


Mandatory participation is basic to the concept of Social Security as a universal system of social insurance.  It makes sense, therefore, that if personal accounts are added to the system they should be universal and mandatory.  Given that Americans like the idea of choice, voluntary participation may be a good political selling point for personal accounts but it would not be good policy.  Mandatory participation-especially if it involves new money rather than a mere diversion of existing Social Security taxeswould be far more likely to boost national savings, maintain progressivity within the system, and ensure that workers build meaningful assets.


Compulsory Participation–A Basic Element of Social Insurance


Social Security was enacted in the 1930s as a system that sought to insure society as whole against poverty in old age.  By some accounts, half or more of the aged were living in some state of dependency.  As a form of social insurance, the program mandates participation while people work and dictates coverage of as many as the political system will permit so that as few as possible will enter their later years dependent on welfare programs and their families.  It requires contributions (taxes) and prescribes a benefit formula that attempts to balance the return each individual receives with the adequacy of the benefits it provides in general–the goal being to afford people with a minimal “floor of protection” when they come to that point in their lives when they can no longer work.  Today, 96 percent of the nation’s workers are covered by the system.


People don’t have to save on their own; they don’t have to open Individual Retirement Accounts (IRA)s or Keoghs; they don’t have to join pension plans or contribute to 401(k)s; they don’t have to buy houses or accumulate other assets; they don’t have to take out life or disability insurance; and they don’t have to set aside money for their families’ well being.  And their employers don’t have to provide private pensions or other benefits for their welfare.  It is prudent for workers and their employers to do those things, but they are all voluntary.  Social insurance is the one thing that society mandates so that workers will have at least a minimal income in retirement or when a disabling condition or death limits their ability to provide for themselves or their families.


One of the best commentaries on the subject came from a member of the business community of the 1930s.  Reinhart A. Hohaus, an actuary employed by the Metropolitan Life Insurance Company and an advisor and member of various Social Security advisory councils in the early years of the program, wrote what has long been viewed as a classic treatise on social insurance.  As excerpted from an article he wrote for the Society of Actuaries, Hohaus framed the concept this way:


Social insurance… aims primarily at providing society some protection against one or more hazards which are sufficiently widespread throughout the population and far-reaching in effect to be “social” in scope and complexion.  Usually these risks are not many in number. Yet if not guarded against through some organized means, they produce large dependency problems that take their toll in terms not only of financial but of human values as well.


Directed against a dependency problem, social insurance is generally compulsory–not voluntary–giving the individual for whom it is intended no choice as to membership. Nor can he as a rule select the kind and amount of protection or the price to be paid for all.  All this is specified in the plan, and little, if any, latitude is left for individual treatment.  Indeed, social insurance views society as a whole and deals with the individual only as so far as he constitutes one small element of the whole. Consistent with this philosophy, its first objective in the matter of benefits should, therefore, be that those covered by it will, as far as possible, be assured of that minimum income which in most cases will prevent their becoming a charge on society.  (emphasis added)


It has always been a condition for receipt of Social Security benefits that a person work for a specified period of time in covered employment (e.g., 10 years for retirement benefits).  Moreover, benefits are derived from a person’s wage history and, as a matter of equity, the more one earns in covered employment the greater the benefits will be.  However, with a benefit formula that provides a higher return on the taxes for low-wage earners than high-wage earners, automatic annuitization of benefits, no reduction in couples benefits even though survivor benefits are automatic, the provision of benefits to dependent spouses, aged parents, and children through high school, and inflation adjustments once one becomes eligible, the system is awash in social protections that are not ordinarily found in any other form of traditional pension or insurance plan.


Taken together, these features are rooted in the concept that Social Security should be a floor of protection against what President Franklin Roosevelt once described as the “vicissitudes of life.”  It is the one form of preparation for old age that people have as a safety net for life’s adversities.


But there is a condition for this benevolence: mandatory participation. The deliberate actions people take with their lives are important–i.e., whether they will be spenders or savers–but the other side of that is that society should not bear the burden when it is too late to alter one’s behavior or bad luck imposes financial hardship.  That’s the true nature of social insurance–it’s insurance for society as well as for the individual.


What Could Change With Voluntary Accounts


As noted, Social Security attempts to balance social adequacy with individual equity-meaning a fair return on contributions. But today, for the first time in the history of the program, large categories of newly retiring workers are due to get back less than the market value of prior contributions.


According to the Urban Institute’s calculations, the typical single male retiring at age 65 in 1970 earned an inflation adjusted return of 6.4 percent on his lifetime Social Security (Old-Age and Survivors) taxes. Today, the typical single male retiring in 2005 can expect to earn a return of 1.9 percent.  The typical single male retiring in 2040 is due to earn a return of 1.6 percent–and this assumes that current-law benefits can be paid in full without any increase in current-law taxes.  Social Security continues to offer a better deal to some categories of workers than to others.  But among younger Americans, virtually all categories–including low-earners–will earn a lower return on their Social Security taxes than they could if their taxes were invested in risk-free Treasury debt. 


The problem of declining “moneysworth” has led many to conclude that Social Security is not doing a good enough job of balancing individual equity with social adequacy.  Personal accounts could help in this regard by improving the return on workers’ contributions and creating a tighter link between contributions and benefits.


While details of the various proposals to create personal accounts under Social Security vary, most of those now being considered envision voluntary participation. Workers would be given the option of contributing a portion of their existing Social Security taxes into a personal account.  Money deposited in the new accounts would belong to the individual, not the government.  The individual would choose how much to deposit into the account (in some cases, up to a limit) and what asset class they wish to invest in.  Upon retirement, death, or disability, periodic withdrawals or annuities would be based on the accumulation of those deposits.


Under the current system, taxes are deposited in the U.S. treasury and the money supports the program’s financial needs in the aggregate.  Simply stated, the money does not belong to and is not put away for each individual.  Instead, the Social Security Administration keeps track of a worker’s earnings history–how long the person worked and how much he or she made each year. What the worker put into the system, though related to what they earned, is not directly relevant.  It is the earnings record that establishes eventual benefit levels.  Upon retirement (or death or disability), the individual or family members become “entitled” to a monthly benefit derived by applying a formula to an average of those earnings.  That’s the principle difference between what the existing system does and what President Bush and others have proposed.


The Consequences of All Things “Voluntary”


Voluntary participation in personal accounts conveys a very different meaning about the nature of Social Security contributions than if the accounts were enacted on a mandatory basis.  To be sure, any form of personal accounts-voluntary or mandatory-would introduce the element of asset ownership, which is not present in the current system. Voluntary accounts, however, would introduce the additional element of choice and that, in turn, could begin to erode the program’s underpinnings as a system of universal social insurance.


Choice implies control.  Indeed, advocates of voluntary accounts often tout “choice and control” as key selling points.  Yet to the extent that choice and control are emphasized in assessing the advantages of personal accounts, the social insurance nature of the system is diminished.


A first step toward a voluntary system opens it up to what is called “adverse selection.”  Those perceiving the most to gain will opt for the alternative.  Clearly those with high incomes would lean toward choosing personal accounts as they already perceive Social Security to be a bad deal and would hope to do better with their own accounts.  At some point, so will people of more moderate means.  Meanwhile, those with low incomes who benefit the most from Social Security’s current progressive tilt may be reluctant to opt for personal accounts even though this choice may leave them trapped in an unfunded pay-as-you-go system that must eventually cut back on promised benefits.


And while those who do opt to join the new system may start out with the perception that they are putting money away for their retirement, they may decide later that they want more discretion over the money and pressure Congress to weaken the retirement saving condition–something Congress has already done with tax preferred savings instruments. 


At that juncture, the safety net may begin to unravel.  The desire to improve Social Security’s individual equity by moving toward personal ownership need not and should not mean dismantling social insurance or “privatizing” Social Security.  The mandatory nature of the program doesn’t end with the requirement to pay taxes.  Its benefits are only available with the onset of one or another of three events:  retirement, death, or disability.  And those limitations have been time tested–no President or Congress has tried to challenge them.


This point could easily be lost if a part of the system were deemed to be voluntary.  People have a different perception of their own discretionary savings than they do with Social Security. They may choose to contribute to personal accounts with the idea that they are preparing for retirement, but they may also feel that because they have the option to contribute, they should be able to do what they want with the money.  In effect, what they may initially set aside for retirement they may some day want to use for something else… a family illness, a house, a car, a child’s education, a boat, or a vacation. People may well begin to ask: if we can choose to make deposits into our own accounts, why can’t we choose when to withdraw the money, and for what purpose?  Would workers then begin to demand the freedom of opting in and out as their circumstances or the markets shifted?


Questions such as these would be increasingly difficult for politicians to answer once they had sold the public on reforms explicitly designed to promote choice and control over Social Security contributions. With mandatory accounts, however, the purpose would be explicit and clear-to boost retirement saving, not just for the individual worker’s well being but for the greater good of society as well.  And because mandatory accounts would be consistent with the concept of universal social insurance, policy makers would find it much more acceptable politically to regulate investment options, administration, and conditions of payout.  All of this would be necessary if policymakers decide to include personal accounts as part of Social Security.


None of the perceived advantages of personal accounts-such as higher returns on contributions, greater national saving, accumulation of assets that can be passed to heirs, and a “lockbox” that politicians couldn’t pick to fund other governmental programs-requires that they be voluntary to be successful.  Just the opposite.  If personal accounts can help to achieve those things, there is no good reason to give workers the “option” of losing out on them or to deprive society in general of the gains.


Social Insurance Should Not Depend on a Coin Toss


Workers would face a profound choice under a system of voluntary accounts.  And it may be one that it is unfair or unreasonable to expect of them.  It is fine to have the hope and determination when one starts out in life that one will achieve financial security, but most people don’t know at the beginning of their careers how they will fare.  They don’t know if they will wind up rich or poor; whether they will become disabled or die at a young age; whether they will be unemployed for long periods; whether they will marry, divorce, or marry again; and if they do marry and have a family, whether their kids or spouses will have serious illnesses that drain their resources.


And yet a system of voluntary accounts would force workers at some early stage of their careers to choose between the market risk of personal accounts and the political risk of an unfunded pay-as-you-go system.  In effect, Americans would sort themselves into two vast segments-one making an enormous economic and policy bet going one way, and the other making an equally high stakes bet going the other way.  Mandatory personal accounts would carry the same potential advantages and risks for workers, but would not require workers to begin their careers by tossing a coin.


Would Voluntary Accounts Really Boost Saving?


American workers and employers have some experience with voluntary savings plans such as 401(k)s and Individual Retirement Accounts (IRAs).  To date, these plans have met with mixed results.  According to some estimates, fewer than 3 percent contribute annually to an IRA and only 8 percent take full advantage of an employer sponsored 401(k) plan.  More than 25 percent don’t take advantage of such plans at all.  Recognizing the problem of under utilization, an emerging trend in private sector plans is to make them less voluntary by providing automatic enrollment.  The worker could still opt-out but without such a deliberate action he or she would be enrolled in the plan and automatic deductions would be made from paychecks into savings accounts.


The experience with private sector plans suggests that participation in any new system of Social Security accounts may have to be mandatory to ensure that personal savings will actually increase.  If our national experiment with voluntary savings plans has demonstrated anything, it is that workers do not take full advantage of the “choice” offered to them.  As Dallas Salisbury, President and CEO of the Employee Benefit Research Institute (EBRI) recently stated in testimony before the House Ways and Means Committee, “Decades of data underline that compulsion in savings and distribution produce better retirement income results than open individual choice. If the policy objective is choice, that does not matter.  If the policy objective is life-long retirement income security, it does matter.”


Moreover, enacting voluntary carve out accounts would add even more uncertainty to long-term budget projections.  Estimates of a plan’s cost must begin with an estimate of participation. When the Social Security Administration evaluated the three models produced by the President’s Commission to Strengthen Social Security it produced several estimates for each model based on assumed participation rates.  The assumptions led to large differences in the fiscal impacts.  For example, Model Two was estimated to cost $2.5 trillion over its first three decades assuming 67 percent participation.  But assuming 100 percent participation, the plan was estimated to cost $4.5 trillion over the same period.  Voluntary carve out accounts would increase both the amount and the uncertainty of the government’s fiscal exposure.  This would not constitute prudent fiscal planning at a time when the government is already facing large chronic deficits.




Society has a legitimate interest in ensuring that people do not under-save during their working lives and become free riders on some future means-tested safety net the government feels compelled to provide.  Moving Social Security toward personal ownership–to help increase saving, sustain a healthy economy, and improve the outlook for future retirees-does not have to occur at the expense of the nation’s social insurance blanket.

Many personal account advocates, including the President, believe that personal retirement accounts should be voluntary.  That would be a mistake.  As a matter of principle, providing people with choice is good.  But choice is not without bounds.  People can’t choose whether their taxes finance the nation’s military, its roads and bridges, its promotion of public health and disease control, its police, firemen, and air traffic controllers, its coast guard, its borders, its food safety, its national parks, and so on.  As a nation, we act collectively through our various levels of government to support common interests that no individual has the resources to do on his or her own.

Any new personal accounts adopted as an element of Social Security should be mandatory.  It is true that by staying under the current system, workers face the risk that future Congresses will default on today’s unfunded pay-as-you-go benefit promises, and personal accounts may reduce that risk.  But how exactly are workers supposed to evaluate that risk against those inherent in personal accounts?  And why should they be expected to?  Choice is not important and can be potentially damaging in a compulsory social insurance program whose primary function is to protect people against poor choices.  Under a social insurance system, if personal accounts are deemed to be good for some, they should be good for all. 

 The largest group not participating is employees of certain state and local governments.

 See Policymaking for Social Security, Martha Derthick, The Brookings Institution, Washington, D.C. (1979).

 Reinhart A. Hohaus.  “Equity, Adequacy, and Related Factors in Old-Age Security,” The Record, American Institute of Actuaries, volume 37 (1938).

 “We can never insure one hundred percent of the population against one hundred percent of the hazards and vicissitudes of life, but we have tried to frame a law which will give some measure of protection to the average citizen and to his family …” Franklin Roosevelt’s statement on signing the Social Security Act, August 14, 1935.

 See, Statement of Dallas Salisbury, President and CEO, Employee Benefits Research Institute, May 19, 2005, testimony before the House Committee on Ways and Means. See also, “IRA Ownership in 2004,” Investment Company Institute, Research in Brief Volume 14, No. 1 February 2005, indicating that about 40 percent of households own an IRA with only about 10 percent of households contributing to an IRA in 2003.


Share this page

Related Issue Briefs