An Agenda for Retirement Security

Andrew G. Biggs

Current Issue

The movement for Social Security expansion began just a few years ago among a small cadre of progressive activists, and it has grown by leaps and bounds. Today, a majority of Democratic Party voters — and, based on a 2015 resolution in the Senate, an even larger share of Democratic members of Congress — favor increasing Social Security benefits, despite the program's large and growing funding shortfalls. Prominent commentators on the left, such as the New York Times editorial page, have endorsed Social Security expansion. And calls merely to "strengthen" Social Security, as former President Obama made in his 2016 State of the Union address, are not enough. "Strengthening," stated James Russell, a progressive activist, was a euphemism that "dodged taking a position within the Democratic Party's increasingly central debate over expansion of Social Security."

In their push to expand benefits, Democrats and progressives have propagated the narrative of a "retirement crisis": Millions of Americans, they say, have failed to save adequately for retirement. A closer look at the evidence, however, reveals that most Americans have actually increased their retirement savings and that Social Security replaces a larger share of pre-retirement earnings than previously thought. Moreover, progressive leaders like Senator Bernie Sanders, who enjoyed significant support during his 2016 presidential campaign, have neglected to consider the pitfalls and unintended consequences of their plans to expand Social Security.

The truth is, increased Social Security benefits and other progressive reforms would actually aid the highest earners the most, fail to make the program solvent or pay for higher benefits, and prompt Americans to reduce their retirement savings. Another progressive approach, retirement plans run by state governments, risks lowering private saving and forcing Americans to rely on public officials with poor track records of delivering on the benefits they promise. The overall result would be future retirees receiving a substantially greater share of their total income from the government, which has shown itself to be a poor steward of citizens' money.  

This is an unacceptable outcome for conservatives who care about America's tradition of limited government and personal responsibility. Too often in the past, however, conservatives have failed to articulate a compelling vision for Social Security reform that would gain political support. They have treated reform as merely an accounting exercise requiring tax-and-benefit adjustments, rather than as an opportunity to truly strengthen the program and America's private retirement-savings system.

Most Americans aren't looking for a handout. But they understandably worry about their retirement incomes, and calls for more government involvement will only proliferate in the absence of reform. Given that President Trump has so far said little about Social Security other than his pledge to protect the program, forward-looking members of Congress now have an opportunity to seize the policy discussion and show what the right approach to protecting the elderly and the next generation of Americans might look like.

They must begin by articulating why the progressive narrative of a retirement crisis — and the need for Social Security expansion — is not only inaccurate but also an inadequate vision for guaranteeing retirement security in the 21st century. They must heed the concerns of older Americans, who are crucial to the success of any reform proposal. They must advance principled reforms that improve the lives of older Americans and make Social Security financially solvent by strengthening the program for low-income seniors and bolstering its work incentives. And to counter the progressive narrative, conservatives can demonstrate that more Americans are saving for retirement today than ever before — and, with better plan designs, private retirement saving can be made accessible to all. Only then will all Americans — young and old, rich and poor — be confident in a secure retirement.

A RETIREMENT CRISIS?

Supporters of expanding Social Security rely on a small number of easy-to-understand statistics to demonstrate the inadequacy of private retirement saving and the need for higher Social Security benefits. For many, these figures are all they need to conclude that a larger role for government is needed. But a tug on the threads of these claims reveals a far less persuasive case.

Consider three claims. First, in a January 2016 editorial supporting expanded Social Security, the New York Times noted that "36 percent of retirees rely on Social Security for 90 percent or more of their income; over all, 65 percent of retirees rely on it for more than half of their income." Second, the Times cited a Government Accountability Office study finding that "52 percent of American households with someone 55 or older have nothing saved for retirement and that only half of that 52 percent will get anything from a company pension." Third, policy analysts such as Alicia Munnell, who heads the Center for Retirement Research at Boston College, note that households' total wealth has declined in recent years relative to those households' incomes. These "basic data," Munnell concludes, "[reveal] a significant decline in retirement preparedness."

On the surface, these claims are compelling. Yet they do not hold up to scrutiny. For instance, the figures the Times cites regarding Americans' reliance on Social Security come from a government dataset — the Census Bureau's Current Population Survey — that ignores most of the income retirees derive from Individual Retirement Accounts, 401(k)s, and lump-sum withdrawals from traditional pensions. In that survey, "income" is defined as a regular payment, such as a monthly check from Social Security. Irregular payments, such as as-needed withdrawals from a 401(k) account, don't count as "income." In other words, supporters of expanding Social Security point to these statistics that define away much of retirees' non-Social Security incomes, then lament that Americans rely so heavily on the main income source — Social Security — that is included in the data. But recent Census Bureau research that relied on IRS administrative data, which counts IRA and 401(k) withdrawals in whatever form they are made, found that from 1984 to 2007 the percentage of new retirees receiving private retirement-plan benefits doubled and median benefit payouts more than doubled. (This study focused on retired women, but it looked at total household incomes and included male spouses, if present.) Thanks to rising private retirement benefits, real total incomes for the median retiree household rose by 58%. In the CPS, which undercounts private retirement benefits, total household incomes rose by only 21%.

Similarly, when examined closely the GAO's figures on the number of near-retirees without retirement savings accounts are less than frightening. Half of the group cited by the GAO has a traditional pension plan, which progressives tout as the ideal form of retirement saving. The other half of this non-saving group has very low incomes, of about $20,000 per household or $11,000 per person. These are precisely the people Social Security was created for and who will, according to a December 2016 Congressional Budget Office analysis, receive Social Security benefits that replace most or all of their inflation-adjusted career-average earnings.

Likewise, Federal Reserve data do show that household wealth has declined relative to household earnings. From 2007 to 2013, for instance, average household wealth fell from 576% of earnings to 469%. Commenting on the decline in net worth for near-retirees, a Wall Street Journal financial editor said, "In one chart, you can see the standard-of-living for America's senior citizens slowly withering away." What proponents of the retirement-crisis theme don't mention, however, is that none of this decline was due to lower retirement saving. In fact, retirement savings increased slightly relative to earnings. The decline in wealth was almost totally accounted for by the bursting of the housing bubble. A housing-market decline says nothing about the design of America's system of retirement saving. In fact, the decline in housing wealth may not even say much about Americans' overall level of preparation for retirement. For individuals who stayed in their homes during the rise and fall of housing prices from 2001 through 2009, the expansion and collapse of the housing bubble are of limited importance. Likewise, while retirees who sold their homes as they downsized may have taken a loss following the housing downturn, they also could purchase a new home at a lower cost.

In reality, the "basic data" tell a very different story. Americans' saving for retirement occurs mostly in three forms: retirement accounts such as 401(k)s and IRAs; benefits accrued under traditional defined-benefit pensions; and benefits accrued under Social Security. The Federal Reserve tracks the first two sources of retirement savings while the Social Security Administration tracks the third. Unfortunately, comprehensive data for all three exist only as far back as 1996, but the trend is clear: In 1996, total retirement assets were equal to 269% of total personal incomes, and by 2015 assets had risen to 414% of incomes. While pockets of under-saving no doubt exist today, as they did in the past, it is difficult to look at these figures and conclude that retirement saving has stagnated in the United States, much less that we face a retirement crisis.

Recent research has also begun to tell a different story regarding the generosity and adequacy of Social Security benefits, which form the foundation of retirement income for nearly all Americans and are a lifeline for many low earners. For years, Americans have received this message from the SSA: "Most financial advisors say you'll need about 70 percent of your pre-retirement earnings to comfortably maintain your pre-retirement standard of living. If you have average earnings, your Social Security retirement benefits will replace only about 40 percent."

In truth, the SSA's 40% "replacement rate" figure doesn't compare the average retiree's Social Security benefits to that retiree's past earnings. SSA's calculations — while convoluted — effectively compare the benefits paid to the average new retiree today to the earnings of the average worker today. This number tells us nothing about a given retiree's ability to maintain his pre-retirement standard of living, which is taken to be the goal of retirement saving both in conventional financial planning and in academic economics.

The real picture is more encouraging. Using Social Security Administration data on earnings and benefits, I calculated a fairly straightforward and understandable measure: for all non-disabled beneficiaries, average Social Security benefits as of the full retirement age as a percentage of those beneficiaries' average earnings between the ages of 46 and 60, adjusted for inflation. The resulting "replacement rate" was 90%. The reality is that Social Security benefits are higher relative to true household earnings than many people realize, a fact which helps explain Americans' saving and retirement habits.

EXPANDING SOCIAL SECURITY

Nevertheless, proponents of an expanded government role in retirement have not been deterred. Senator Bernie Sanders has offered the most prominent plan to expand Social Security. The plan he proposed as part of his 2016 presidential campaign would expand benefits for all retirees and disabled workers, and would be financed through increases in payroll taxes and investment-income taxes on high-income households.

Sanders's proposal would increase benefits across the board, while enhancing benefits for low-wage workers with long careers. His plan would also increase annual cost-of-living adjustments for both current and future beneficiaries. To finance these new benefits, Sanders would phase out the maximum taxable wage, currently $127,200, applying the 12.4% payroll tax to all earnings. In addition, Sanders would institute a 6.2% tax on investment income for households with incomes above $250,000, with the proceeds funneled to the Social Security trust funds.

From a static perspective, Sanders's plan would raise more than enough revenue to make Social Security solvent for 75 years, though other issues — discussed later — make it likely that revenues would fall short of projections. According to the SSA's actuaries, Sanders's plan raises enough taxes to address 107% of the long-term Social Security shortfall.

But rather than use those funds to pay the benefits Social Security has already promised, Sanders expands benefits by more than $34 trillion in the next 75 years. As a result, instead of keeping Social Security solvent for 75 years and beyond — which has long been the goal of reformers, and which the plan comfortably achieves with Sanders's new revenues — Sanders's proposal would keep Social Security solvent until about 2065, at least two decades less than if he could resist the temptation to increase benefits. In other words, despite Social Security tax increases totaling $2.4 trillion in the first 10 years and even more thereafter, a young person entering the workforce today could not expect to receive his full Social Security benefits in retirement.

The distribution of the new benefits is also an important factor to consider. The lowest-earning workers, particularly those with short careers, may receive little in the way of benefit increases due to the way Sanders structures his reforms: Strangely, Sanders's plan increases the number of retirees who receive the most progressive elements of Social Security's benefit formula, which replaces 90% of an individual's pre-retirement earnings, but it does not increase benefits for most of those already in that most progressive zone.

Overall, the Sanders plan does raise most low earners' benefits by a higher percentage than it raises benefits for higher earners, and, combined with Sanders's tax increases on high earners, the plan as a whole is progressive. But the largest dollar increases in Social Security benefits under the Sanders plan do go to the highest earners: Fully 40% of benefit increases under his plan flow to the highest-earning fifth of households; nearly 60% of all benefit increases under his plan go to the top two quintiles of earners.

Those who think of Sanders as a European-style socialist shouldn't be surprised. The difference in generosity between U.S. and European social-transfer programs isn't generally in their treatment of the poor. It's in how they treat middle and high earners. In the United States, fiscally conservative Social Security reforms trade lower benefits among high earners for maintaining current tax rates on high earners, while progressive reforms increase Social Security taxes on high earners in order to pay greater benefits to generally higher-earning households.

From a practical perspective, Sanders's approach to Social Security ignores how difficult it is to pass substantial tax increases in the United States. Of all the potential uses for a dollar of tax increases — including progressives' favorite options, such as subsidizing college or expanding health coverage — it seems bizarre to dedicate those hard-won revenues to expanding Social Security benefits for middle- and high-income households that are easily able to save on their own.

RAISING THE CAP

A core element of Bernie Sanders's Social Security plan — and indeed all progressive reform proposals — is eliminating the limit on earnings used to levy taxes and calculate benefits. Since Social Security's creation in 1935, both payroll taxes and benefits have been calculated on earnings up to a "cap." Currently, individuals pay taxes on their first $127,200 in earnings, and those same earnings are plugged into Social Security's benefit formula to calculate the benefits they will receive in retirement. Eliminating the payroll-tax ceiling would, all other things being equal, increase the effective top marginal tax rate on earned income by 12.4 percentage points as well as raise the benefits paid to high earners once they retire.

The payroll-tax ceiling was put in place at Social Security's origin for a reason: It established Social Security as a social-insurance program, in which everyone paid taxes and received benefits in some reasonable proportion, rather than a "welfare" program in which most taxes are paid by one group while most benefits are received by another. In fact, the Committee on Economic Security appointed by Franklin Roosevelt to draft his Social Security plan recommended that individuals earning above $3,000 — about $145,000 in terms of today's wage distribution — be exempt from Social Security taxes and not participate in the program. That is, the committee recommended that Social Security contain no redistribution from very high earners to lower earners.

While Congress chose to include high earners in Social Security, the cap on earnings subject to payroll taxes reflected Roosevelt's intent that Social Security resemble a mandatory individual-savings program, with a supplement to low earners, rather than a welfare program transferring resources wholesale from rich to poor. The fact that Social Security is seen as an "earned" benefit rather than a handout is central to the program's continuing political strength. It is ironic that the progressive left, which holds that the very rich have disproportionate power in Washington, D.C., wishes to enact tax increases that, for the first time, would give the very rich a strong reason to oppose the Social Security program.

Beyond changing the character of the program, there are other practical problems with eliminating the Social Security payroll-tax ceiling. To begin, while eliminating the ceiling would raise Social Security's annual revenues by more than 1% of GDP, doing so would address only 40% to 70% of the long-term program deficit, depending upon whether one puts greater stock in the Social Security Trustees' or the CBO's projections. Based on the CBO's projections, eliminating the payroll-tax celling would delay Social Security's trust-fund insolvency date by only a decade to 2039. Even if, as Sanders recommends, we eliminated the payroll-tax ceiling and paid no additional benefits to high earners, that would not be sufficient to make Social Security solvent.

Moreover, even these figures may be optimistic. Both the SSA and CBO assume, in their scoring of proposals to eliminate the taxable limit, that employers would reduce the wages they pay employees to cover the employers' increased payroll-tax liability. That's a fairly standard reading of how the costs of employee benefits are borne. What the budget scorekeepers don't note, however, is how this wage reduction would affect other tax revenues, namely federal and state income taxes and federal Medicare payroll taxes. When employee wages are reduced, governments collect lower taxes on those wages. In a 2006 study, economists Emmanuel Saez and Jeffrey Liebman estimated that the reduction in federal and state income-tax and Medicare-tax revenues would reduce the net revenues from eliminating the tax cap by about 20%. Since 2006, the top marginal federal tax rate on earned income has risen by five percentage points, according to the OECD, which implies that the non-Social Security revenue loss associated with eliminating the payroll-tax ceiling would cut about 25% from the gross Social Security revenue increase.

Furthermore, the figures above take no account of the behavioral effects of higher taxes. Using what Saez and Liebman take to be a middle-of-the-road estimate for the way taxable income responds to changes in tax rates, total revenues raised by eliminating the payroll-tax ceiling would be only about half of what is projected using the SSA or CBO static projections, or even less using today's higher marginal income- and payroll-tax rates. And recall that these are not the writings of supply-side economists; Saez, in particular, is a darling of the left for his work analyzing income inequality and advocating for higher taxes, while Liebman served in a top budget job in the Obama administration. In simple terms, eliminating the payroll-tax ceiling would not be nearly sufficient to pay currently promised Social Security benefits, much less increase them, while non-Social Security revenues would fall significantly due to income shifting and behavioral effects. This is no recipe for a larger Social Security program or a healthier federal budget.

TOTAL RETIREMENT INCOMES

Despite panicked rhetoric about expanding Social Security to respond to a "retirement crisis," it is not clear how much total retirement incomes would actually increase in response to raising Social Security benefits. The reason is that most households save for retirement beyond the amounts they expect to receive from Social Security, and many Americans will adjust their retirement saving downward if they expect to receive higher benefits from the government.

Low earners, to be sure, are highly dependent on Social Security and don't save much on top of the program. But it is well established that middle- and high-income households react to anticipated changes in Social Security taxes or benefits. Middle- and upper-income earners, who anticipate a combination of Social Security benefits and income from savings and employer-sponsored retirement plans, likely would offset Social Security benefit increases by saving less elsewhere. Both micro-level analysis of individual-household savings and macro-level comparisons of one country to another strongly indicate that middle- and upper-income households are saving about what they wish to save for retirement. If forced to save more for retirement through one means, they are likely to save less elsewhere.

For instance, a 2015 study published by Canada's Fraser Institute found that, when Canada raised the payroll-tax rate to fund its pension system, households reduced their own saving by about a dollar for each dollar of extra taxes they paid. Similar data was found in analyses of reforms in the United Kingdom, Italy, and Poland. Likewise, research on retirement saving in the United States suggests that better-educated households and households with retirement savings accounts — both groups who are effectively equal to higher-earning households — are likely to offset much or all of a Social Security benefit increase via reduced household saving. These are precisely the households who receive the lion's share of the benefit increases under a benefit-expansion plan like that advocated by Sanders.

Data across countries suggest similar conclusions. The OECD reports on the total incomes of retirees in developed countries, while breaking those incomes down between government-pension benefits and incomes that retirees provide for themselves through saving or work in retirement. The average American aged 65 and older has a total income equal to about 92% of that of the overall U.S. population. Given that the cost of living declines in retirement, average U.S. retirement incomes should provide a standard of living at least as high as that of working-age Americans. The United States has a less generous Social Security program than government pensions in other countries, but Americans make up for it by saving more on their own and retiring later. And despite claims of inadequate saving in the United States, total U.S. retirement incomes are above the OECD median of 88% of countrywide average incomes.

Looking from one country to the next also provides clear evidence that government pensions are a substitute for other sources of retirement income. Across OECD countries, retirement incomes derived from work and saving are lower by 83 cents for each dollar of additional income provided by government-run pension plans. In other words, we could expect that much of the increased benefits from an across-the-board Social Security expansion would be offset by reduced saving elsewhere.

Targeted Social Security expansions, as I will discuss later, could benefit low-earning households who have little or no personal savings for retirement. But for middle- and upper-income households, who both save for retirement on their own and expect to receive Social Security, raising Social Security benefits is likely to reduce each household's own retirement savings. From those households' perspectives, the tradeoff is likely a wash: They will receive fewer benefits from their personal savings, but more from Social Security. But from a larger economic and political perspective, it is far from a wash: Reducing real saving by households, which can be used to finance research and investment, lowers future economic growth, and raising taxes to fund Social Security benefits lowers incentives to work. In the broader political sense, it cannot be helpful to America's culture of limited government and individual responsibility to make middle- and upper-income households — who could, should, and would save more on their own — more dependent on government payments in old age.

Further complicating this issue is the progressive left's solution for inadequate personal saving: state-run retirement plans. So-called "Secure Choice" programs would require employers that don't offer a retirement plan to enroll employees in IRAs managed by the state. More than half of states already have considered or passed legislation to develop these retirement plans, including California, Illinois, and Connecticut. These plans aren't without merits, as they would both make available a new savings option and enroll employees who don't currently have a retirement savings account. And if progressives proposed them as alternatives to expanding Social Security, rather than additions to it, the drawbacks and potential dangers might be worth the tradeoff. For now, however, there are several reasons for caution.

First, state-run retirement programs sometimes exclude the small employers who are least likely to offer a retirement plan. For instance, the Illinois Secure Choice plan applies only to businesses with 25 or more employees, but roughly three-quarters of these employers already offer plans. Meanwhile, only about half of businesses nationwide with 10-24 employees offer plans, and among those with fewer than 10 employees, only 28% offer a 401(k) or other retirement vehicle. Proponents cite the larger number of employees currently not offered a retirement plan, without noting that their solution doesn't always address the problem. But applying the savings mandate to very small employers might generate resistance.

The second danger is that some employers who already offer 401(k) plans will eliminate them or restrict eligibility and place employees into the state-run system instead. Doing so could reduce costs and regulatory headaches for employers, but could also result in dramatic reductions in the amounts workers save for retirement. The average total 401(k) contribution including employer contributions is more than 8% of a worker's wages, while the default contribution for most state-run plans looks to be about 3% of earnings. If employers shift employees to state-run plans, total savings by those employees could drop significantly. It is hard to say how large these effects might be, but the fact that the mutual-fund industry is opposing state-run retirement plans indicates that they perceive such a threat.

Third, while proponents of Secure Choice plans often cite the fees associated with private 401(k)s, it's not at all clear that state-run plans will be much cheaper. Fees on 529 college-savings plans run by the states are highly variable. And costs for the Illinois Secure Choice plan could run as high as 0.75% of assets per year, whereas the median participant in 401(k) plans today pays 0.67% — and large plans charge even less. Likewise, several prominent state officials, including the former head of the California Public Employees' Retirement System and the former New York state comptroller, have been jailed for "pay to play" agreements with investment managers for state pension funds. All of these facts provide reasons to be wary.

Finally, state-run retirement programs could fall prey to the temptation to offer participants a guaranteed return on their investments without accounting for the costs of providing such guarantees. This is precisely what drove state and local government-employee pension plans into financial trouble: promising employees benefits based on the assumption of high investment returns, with the taxpayer bearing the costs if and when investments fail to perform as predicted.

A recent report commissioned by the state of Connecticut found that guarantees against market risk are prohibitively expensive. But some of the most prominent advocates for state-run retirement plans continue to propose guaranteeing returns of 2% to 3% above inflation. If state governments provided guarantees, there would be no legal requirement to account for their costs, just as state and local pensions do not account for the market risk they currently bear. Luckily, most states are moving away from providing guarantees within state-run retirement plans as they come to terms with the costs.

State-run retirement plans have also been designed to be exempt from the Employee Retirement Income Security Act, or ERISA, a federal law that grants protections to workers' retirement savings. The Obama administration's Department of Labor did the states a favor in setting regulations exempting state plans from ERISA. But the states' poor fiscal management of their own employee-retirement plans, which themselves are not subject to ERISA, does not generate a great deal of faith.

Recently, congressional Republicans have moved to repeal the Obama-era Labor Department regulations exempting state-run retirement plans from ERISA. This presents policymakers with a very difficult choice: State-run plans indeed would help expand the universe of employees with access to a retirement-savings vehicle, a highly desirable outcome. And yet these same plans may not be the best vehicles for many workers and could potentially poach participants from superior 401(k) plans. Congress should shift from a yes-or-no stance on state-run retirement plans to a more engaged policy process that could enable these plans while providing both adequate regulation and new, better retirement-saving options to employees who currently lack them.

REFORMING SOCIAL SECURITY

It is important to push back against unfounded claims of a retirement crisis and to critically examine proposals to expand the government's role in the provision of retirement income. But it is even more important to put forward proposals to address saving shortfalls where they do occur and make Social Security a solvent and effective social-insurance program for the future.

There's not a lot of political appeal in trying to fix Social Security, in the sense that any proposal to fill a $10- to $15-trillion funding hole is either going to raise taxes or cut benefits. That is the harsh reality that reformers of any political party must eventually face and explains why, in spite of politicians' claims regarding "what Americans want," reform has not taken place. What Americans want is a Social Security fix that doesn't involve sacrifice, but that fix does not exist.

For Social Security reform to have any hope of passing, policymakers must find a way to step beyond the usual policy and political constraints that — under the Clinton, Bush, and Obama administrations — have made reform impossible. A serious Social Security reform plan cannot be merely an accounting exercise of putting together a collection of off-the-shelf changes to taxes and benefits in order to keep Social Security's trust funds solvent. Rather, policymakers must be clear-eyed about the goals they want to achieve with Social Security reform and the necessity of securing support from current and future retirees.

The basic story is simple. If we were inventing a Social Security program from scratch, we would design it for today's world. There is no reason a person retiring in 2035 should participate in essentially the same Social Security program that was devised in 1935. Designing a system today, we would recognize what only government can do — provide a robust and reliable backstop against poverty in retirement — but also recognize the limits of what government can do, in particular government benefits' tendency to displace private saving and governments' almost-universal habit of promising pension benefits without adequately funding them. We would know that individual retirement saving can work, but that plan design matters a lot: For instance, a voluntary plan that automatically enrolls employees will have much greater participation than one that requires them to manually sign up. We would also know, from years of academic research, that the work incentives presented by retirement plans are important. When retirement plans discourage delayed retirement, people retire earlier. Plans that encourage longer work lives result in longer work lives, which are beneficial both to individuals' retirement security as well as to the broader economy.

Based on this blank-slate exercise, I previously outlined a broad proposal to reform Social Security in these pages (see "A New Vision for Social Security" in the Summer 2013 issue), which I refer to as the "flat-benefit plan." Under the plan, Social Security would guarantee that all retirees, regardless of work history or earnings, are lifted out of poverty in old age. Thus, while Social Security currently offers no minimum benefit, a strong minimum benefit would be established at the poverty threshold. Over time, however, the maximum Social Security benefit would be reduced so that eventually all retirees would receive essentially the same monthly benefit. The philosophy behind this plan draws on the pension systems in New Zealand and the United Kingdom, which establish strong minimum benefits for retirees. Under my proposal, Social Security's new minimum benefit would be set at the single, over-age-65 poverty threshold of about $950 per month. This benefit level is similar to the base benefits provided in New Zealand and the United Kingdom. Like Canada's basic pension system, the full minimum benefit would be paid to retirees who had lived in the country for 40 years, with scaled reductions based on the number of years a resident had resided legally in the United States.

The result is that every long-term legal U.S. resident would receive a Social Security benefit that would lift them above the poverty line. At first glance this may appear a modest goal, but around one-third of retirees currently receive a Social Security benefit that is below the poverty threshold, and 10% of Americans over age 65 have total incomes below the poverty line. The flat-benefit plan would effectively eliminate poverty in old age for long-term U.S. residents. Moreover, the minimum benefit would increase by about 1% above inflation each year, building a stronger base of retirement income over time.

The flat-benefit plan would include additional provisions to boost benefits for some current retirees and to encourage delayed retirement. While those already in retirement would not be eligible for the new minimum benefit, COLAs for current retirees with monthly benefits below $950 would be increased, financed by applying the slightly lower inflation-adjustment measure known as "chained CPI" to retirement benefits above $1,750. In addition, the 12.4% Social Security payroll tax would be eliminated beginning at age 62, increasing the rewards to delayed retirement.

On top of eliminating poverty in old age, the flat-benefit plan would address Social Security's $10 trillion-plus funding shortfall and restore the program to long-term solvency. Social Security is currently running payroll-tax deficits and will need to draw on its trust funds to make full benefit payments through the funds' exhaustion in the early 2030s. The flat-benefit plan would extend the lives of the Social Security trust funds and, by mid-century, put Social Security back into cash surpluses. These surpluses would give future Americans the option to reduce taxes or increase benefits. Under current law, by contrast, Social Security will be running annual $400 billion deficits (in today's dollars) by 2050, unless benefits are cut across the board by roughly 25%.

ENGAGING OLDER AMERICANS

The fiscal benefits of Social Security reform shouldn't be expected to resonate strongly in a political discussion. American voters embrace fiscal responsibility about as enthusiastically as American politicians do. But the flat-benefit plan includes a number of provisions that could alter the usual political dynamic surrounding Social Security reform.

Current and near-retirees are a key group affecting the political viability of a Social Security reform proposal. On paper, this group has very little at stake in the Social Security reform debate. Most reform proposals would have only small effects on current and near-retirees, and many exempt them from any benefit changes whatsoever. Reformers sometimes hope that, once exempted from significant changes, the old and near-old will simply stand aside and let the debate continue among the younger Americans whom it will most heavily impact.

In reality, though, current and near-retirees are interested and engaged in Social Security and retirement issues. A reform plan's chances of success are dramatically increased if advocates can engage older Americans in its favor. And the flat-benefit approach has a number of aspects that, if introduced on their own merits, may be attractive to retirees and those nearing retirement.

As discussed above, the flat-benefit plan would eliminate the payroll tax for anyone age 62 and older. This isn't merely a political sweetener. Rather, it eliminates a distortion in the current Social Security benefit formula in which workers who delay retirement continue to pay the 12.4% Social Security payroll tax but receive almost no additional benefits in return. To the individual retiree, eliminating the payroll tax means an immediate 6.2% increase in the pay they receive from working. Economic research has found that near-retirees are particularly sensitive to changes in after-tax wages because, unlike younger workers, anyone age 62 and older can simply retire and claim Social Security. Based on that research, I have calculated that reducing or eliminating the payroll tax on older workers would significantly increase their labor supply, raising non-Social Security tax revenues — from federal and state income taxes and Medicare payroll taxes — to offset most of the Social Security payroll-tax losses.

Likewise, the new poverty-level minimum benefit would be implemented immediately, providing a benefit boost for many low-income individuals on the verge of retirement. Reductions in the maximum benefit, by contrast, would be phased in over several decades.

And for those who are already retired, COLAs would be increased for any retiree with benefits below the poverty threshold. In addition to reducing poverty in old age, these progressive COLAs would help mitigate the effects of a growing life-expectancy gap between rich and poor, which otherwise would reduce the progressivity of lifetime Social Security benefits.

The flat-benefit plan is good public policy that makes Social Security financially sustainable, strengthens the safety net by reducing poverty in old age, and improves incentives to delay retirement. On top of that, these changes would make Social Security reform far more politically attractive to low-income retirees, low earners on the verge of retirement, and retirees who wish to continue to work. Those groups would have more reason to listen to fiscally conservative Social Security reforms and would benefit personally if these reforms were passed.

INCREASING RETIREMENT SAVING

Benefit increases and tax reductions, however, don't come without a cost. And that cost is reduced benefits for middle- and high-earning households. In the long term, Social Security would pay a flat benefit pegged to the poverty threshold, which is currently about $950 for a single individual older than age 65. According to SSA data, about 38% of new retirees receive a benefit under that level, before any reduction for early retirement. They're clear winners under the flat-benefit plan. But the median new retiree receives a benefit of about $1,150. So, once fully implemented, a typical retiree's Social Security benefit would decline by about 17%. A retiree at the 95th percentile — meaning, one who receives a benefit higher than 95% of other retirees — receives about $2,075 per month. That's about a 54% difference. Make no mistake, this is a fiscally conservative plan.

Benefit reductions for middle- and high-earning individuals would be implemented gradually, for two reasons. First, these households need time to adjust their retirement saving and their expectations about how long to work. Precipitous benefit changes make such planning difficult. Second, the new flat benefit is phased in so that, to a good approximation at least, all benefits that have been earned under the current benefit formula would be honored. The flat-benefit plan would not default on the benefits middle- and high-earning workers had already accrued, but would alter the rate at which they earn future benefits to make the program more financially sustainable. Many younger Americans do not believe they will receive any Social Security benefits at all. If reductions to their future benefits were implemented to give them plenty of warning — in a way that did not reduce the benefits they already have earned, and as part of a larger reform plan that ensures Social Security will be able to keep the promises that it makes — these younger Americans may look more kindly on reform.

But the approval of younger and middle- and upper-earning households depends on their ability to easily make up for reduced Social Security benefits by increasing their own retirement saving. I have calculated that, on average, if every household saved about 3% more of its earnings in a 401(k)-style account, holding only government bonds, those additional savings would be sufficient to make up for the losses to their traditional Social Security benefits.

For high earners, this is simply a matter of dialing up their saving rate — something they are both capable of and willing to do, certainly if it spares them the expense of higher taxes. Higher earners tell researchers that they are likely to save more or delay retirement in response to Social Security benefit reductions, and research on individuals' reactions to pension reforms in other countries confirms that finding.

It's in the middle where this question is more challenging. From both a political and a practical standpoint, middle-income households need time to increase their retirement savings. There is no fundamental reason that middle-income households today cannot save as much as the average U.S. household of previous generations. But on a practical level, some households have a hard time ensuring that money gets set aside out of each paycheck. For instance, at the time President George W. Bush was promoting Social Security reform that would reduce benefits for middle and high earners, Bush's then-spokesman acknowledged that he had himself never participated in an employer-sponsored 401(k) plan.

But this is not an unsolvable problem. Indeed, the biggest parts of the retirement-saving problem have already been overcome. For instance, as the U.S. retirement system shifted from traditional defined-benefit plans, which effectively had mandatory employee participation, to voluntary defined-contribution 401(k)s, many employees simply didn't sign up to participate. But research beginning in the late 1990s found that when employees are automatically enrolled in 401(k)s, the vast majority continue to participate and are glad they did so. Today, 59% of employers offering 401(k)s use so-called "auto-enrollment." Overall, more Americans participate in employer-sponsored retirement-savings plans today than during the so-called "Golden Age" of traditional pensions. Requiring all employer-sponsored pensions to use auto-enrollment as a condition of the tax preference they receive could expand retirement saving even further.

Similarly, employers today are applying "auto-escalation" to employees' 401(k)-contribution rates. Employees who are automatically enrolled in a 401(k) are assigned a relatively low initial contribution rate, such as 3% of earnings. Over time, however, contributions automatically rise to 6% or so of wages, an amount that is more appropriate for most savers. This gradual increase effectively means slower pay increases for several years, but following that most participants remain at the higher saving rate.

The final concern is employees who aren't offered a retirement plan at work. In 2012, 75% of all private-sector employees were offered a retirement plan at work, according to research published by the SSA, up from 72% in 2006. In theory, employees who are not offered a plan could set up an IRA. In practice, however, few will do so and the contribution limits for IRAs are much lower than for 401(k) plans. The problem with plan availability lies principally with small employers, where the complexity of regulations regarding which employees are covered and how much they can contribute — along with the need to retain lawyers, accountants, consultants, bookkeepers, communication specialists, and investment advisors to remain on the right side of the law — can make 401(k)s more trouble than they are worth. It is not that simpler options do not exist, but many employers are not aware of them, and they come with restrictions on employee contributions that make them less attractive to small-business owners.

Republican senator Orrin Hatch of Utah has proposed legislation to spread the availability of retirement plans, especially among small employers. Among other things, Hatch's proposal would make it easier for small employers to join together in Multiple Employer Plans with defined contributions, which would reduce the cost and administrative burden of establishing 401(k)s. Hatch's legislation would also allow for "Starter 401(k)s" with higher contribution limits than IRA accounts but with fewer administrative burdens.

Together, these steps have increased participation, boosted employee contributions, lowered administrative costs, and could potentially spread access to 401(k) plans. Great progress has been made, but until more workers have the option of an employer-sponsored retirement plan, threats from expanded Social Security or retirement plans run by state governments will remain.

A NEW APPROACH TO RETIREMENT

Retirement income security is a smaller and less challenging problem than many imagine, and there is certainly no crisis. More Americans are saving for retirement and more money has been saved than at any time in U.S. history. Private retirement plans are more widespread than during the so-called "Golden Age" of traditional pensions, and 401(k)s are being improved to increase participation, raise contribution rates, and reduce administrative costs. While this progress needs to continue, the remaining challenge is to make access to retirement plans even more widespread, particularly by encouraging smaller employers to offer 401(k)s. This is not an easy task, but proposals exist to simplify the regulatory burdens and lower the costs of offering a retirement plan in a small-business setting. These should be promoted and pursued with vigor, by both Congress and the Trump administration.

Indeed, while progressives have promoted their "retirement crisis" narrative to buttress their belief that government must dramatically expand its role in providing income to retirees, it is clear that government is the weakest source of retirement funding. While 401(k)s have improved and private saving has increased, and even the dying private-sector defined-benefit pensions are reasonably well funded, government-run retirement plans seemingly grow less solvent with each passing year. Social Security's long-term deficits rose by more than $5 trillion during President Obama's two terms, and neither he nor progressives in Congress did much to fix the system. At the state and local level, government plans for public employees languish in multi-trillion-dollar unfunded liabilities, which governments have tried to address by taking increased investment risk. Even years past the end of the Great Recession, nearly six in 10 state and local pension plans have still failed to receive their full "required" contributions. In this context, it is puzzling to hear that individual retirement saving cannot work and that responsibility must be handed over to the government.

A fiscally conservative approach to Social Security reform has more merit today than ever before, but to succeed it needs to show more imagination. Social Security is a program that Americans believe in, at least at an abstract level, but one that in practice gives them very little confidence regarding their own retirements. Fiscal conservatives must talk about making Social Security solvent, but also about how to make Social Security and the broader system of retirement saving stronger and more effective. It is within our reach to effectively eliminate poverty in old age, but not with either the liberal or conservative reform approaches of the past.

Andrew G. Biggs is a resident scholar at the American Enterprise Institute.


Insight

from the

Archives

A weekly newsletter with free essays from past issues of National Affairs and The Public Interest that shed light on the week's pressing issues.

advertisement

Sign-in to your National Affairs subscriber account.


Already a subscriber? Activate your account.


subscribe

Unlimited access to intelligent essays on the nation’s affairs.

SUBSCRIBE
Subscribe to National Affairs.