The budget compromise forged by Congress and the Obama administration at the end of last month makes two fundamental changes in Social Security. First, it denies a worker the opportunity to take a spousal benefit and simultaneously delay his or her own worker benefit. Second, it stops the “file and suspend” technique, where a worker files for retirement benefits then suspends them in order to generate a spousal benefit.
Unfortunately, neither of these changes gets to the root issue: that spousal and survivor benefits are unfair, although the reform redefines who wins and who loses. Social Security spousal and survivor benefits are so peculiarly designed that they would be judged illegal and discriminatory if private pension or retirement plans tried to implement them. They violate the simple notion of equal justice under the law. And as far as the benefits are meant to adequately support spouses and dependents in retirement, they are badly and regressively targeted.
As designed, spousal and survivor benefits are “free” add-ons: a worker pays no additional taxes for them. Imagine you and I earn the same salary and have the same life expectancy, but I have a non-working spouse and you are unmarried. We pay the same Social Security taxes, but while I am alive and retired, my family’s annual benefits will be 50 percent higher than yours because of my non-working spouse’s benefits. If I die first, she’ll get years of my full worker benefit as survivor benefits.
Today, spousal and survivor benefits are often worth hundreds of thousands of dollars for the non-working spouse. If both spouses work, on the other hand, the add-on is reduced by any benefit the second worker earns in his or her own right.
An historical artifact, spousal and survivor benefits were based on the notion that the stereotypical woman staying home and taking care of children needed additional support. That stereotype was never very accurate. And today a much larger share of the population, including those with children, is single or divorced. Plus, many people have been married more than once, and most married couples have two earners who pay Social Security taxes.
Where does the money for spousal and survivor benefits come from? In the private sector, a worker pays for survivor or spousal benefits by taking an actuarially fair reduction in his or her own benefit. In the Social Security system, single individuals and married couples with roughly equal earnings pay the most:
- Single people and individuals who have not been married for 10 years to any one person pay for spousal and survivor benefits, but don’t get them. This group includes many single heads of households raising children.
- Couples with roughly equal earnings usually gain little or nothing from spousal and survivor benefits. Their worker benefit is higher than any spousal benefit, and their survivor benefit is roughly the same as their worker benefit.
The vast majority of couples with unequal earnings fall between the big winners and big losers.
Such a system causes innumerable inequities:
- A poor or middle-income single head of household raising children will pay tens of thousands of dollars more in taxes and often receive tens of thousands of dollars fewer in benefits than a high-income spouse who doesn’t work, doesn’t pay taxes and doesn’t raise children.
- A one-worker couple earning $80,000 annually gets tens of thousands of dollars more in expected benefits than a two-worker couple with each spouse earning $40,000, even though the two-worker couple pays the same amount of taxes and typically has higher work expenses.
- A person divorcing after nine years and 11 months of marriage gets no spousal or survivor benefits, while one divorcing at 10 years and one month gets the same full benefit as one divorcing after 40 years.
- In many European countries that created benefit systems around the same stereotypical stay-at-home woman, the spousal benefits are more equal among classes. In the United States, spouses who marry the richest workers get the most.
- One worker can generate multiple spousal and survivor benefits through several marriages, yet not pay a dime extra.
- Because of the lack of fair actuarial adjustment by age, a man with a much younger wife will receive much higher family benefits than one with a wife roughly the same age as him.
When Social Security reform eliminated the earnings test in 2000 and provided a delayed retirement credit after the normal retirement age, some couples figured out ways to get some extra spousal benefits (and sometimes child benefits) for a few years. After the normal retirement age (today, age 66), they weren’t “deemed” to apply for worker and spousal benefits at the same time, allowing them to build up retirement credits even while receiving spousal benefits. Other couples, through “file and suspend,” got spousal benefits for a few years while neither spouse received worker benefits.
These games were played by a select few, although the numbers were increasing. Social Security personnel almost never alerted people to these opportunities and often led them to make disadvantageous choices. Over the years, I’ve met many highly educated people who are totally surprised by this structure. Larry Kotlikoff, in particular, has formally provided advice through multiple venues.
So is tightening the screws on one leak among many fair? It penalizes both those who already have unfairly high benefits and those who get less than a fair share. It reduces the reward for game playing, but like all transitions, it penalizes those who laid out retirement plans based on this game being available. It cuts back only modestly and haphazardly on the long-term deficit. As for the single parents raising children — perhaps the most sympathetic group in this whole affair — they got no free spousal and survivor benefits before, and they get none after.
The right way to reform this part of Social Security would be to first design spousal and survivor benefits in an actuarially fair way. Then, we need better target any additional redistributions on those with lower incomes or higher needs in retirement, through minimum benefits and other adjustments that would apply to all workers, whether single or married, not just to spouses and survivors.
As long as we keep reforming Social Security ad hoc, we can expect these benefit inequities to continue. I fear that the much larger reform required to restore some long-term sustainability to the system will simply consolidate a bunch of ad hoc reforms and maintain these inequities for generations.
This column originally appeared on PBS Newshour’s Making Sen$e.
When America’s Social Security system was first established in 1935, the public was deeply concerned about the fate of our elderly, who were then on average poorer than the rest of the population, less capable of working, more likely to work in physically demanding jobs, and less likely to live close to two decades past age 65. Today’s concept of Social Security was actually only one part of an act aimed at meeting the needs of the poor, old, needy, and unemployed of all ages.
In the early decades through the 1960s, Congress expanded old-age supports largely to cover important gaps such as spouses and survivors, disability, health insurance and inflationary erosion of benefits. Today, however, Social Security grows based on past laws that preordain increases in old-age support, largely independent of how the needs of the elderly and nonelderly have evolved or will evolve.
In a newly released study, Caleb Quakenbush and I find that a typical couple retiring today is scheduled to receive about $1 million in cash and health benefits; many millennials will receive $2 million or more. In effect, we’ve now scheduled many young adults to be future Social Security and Medicare bi-millionaires. And the growth continues; the succeeding generation, born early in the 21st century and sometimes referred to as the homeland generation or generation Z, is scheduled for significantly higher benefits. Add to these amounts additional Medicaid expenditures that also go to many elderly if in a nursing home for any extended period of time. (These figures are “discounted”—that is, they show what amount would be required in a saving account, at age 65, earning real interest, to provide an equivalent level of support.)
In fact, a very high proportion of all growth in federal government spending over the next several decades is currently scheduled for Social Security and Medicare. Almost all other spending, whether for children or defense, infrastructure, or the basic functions of government, already is held constant or in decline in absolute terms, and sometimes in a tailspin relative to the size of the economy and the federal government. Only other forms of health care and retirement support, interest costs, and tax subsidies are on the rise.
Such developments are hardly sustainable. Simple math tells us that they will continue to impose costs that the millennials and younger generations are already experiencing: cuts in other benefits for them and their children, higher taxes, and reduced government services when they are in school, working, or middle-aged.
Next time you read a headline on growth in student debt, the falling real value of the child credit, declines in federal spending on education and infrastructure support, or fewer soldiers and sailors, keep in mind that these stories all follow as a consequence of where past Congresses have directed almost all government growth. Of course, governments almost always spend more as an economy and the tax base expand, whether the size of government relative to the economy grows, stays constant, or declines. But past governments traditionally allowed future legislators and voters to choose what to do with those additional revenues; they weren’t stuck with leaving that decision to prior legislators.
How did we get here? As Congresses and presidents added to Social Security over the years, it became more generous. Health insurance was expanded to cover hospitals and doctors, then more recently under President George W. Bush, drug benefits. Cash benefits were raised through various enactments under Republican and Democratic presidents alike.
One big culprit is the retirement age, which, by remaining stable on the basis of chronological age, does not remain stable on the basis of years of support, which increase as people live longer. A typical couple retiring at the earliest retirement age now receives benefits for close to three decades, which is roughly the expected lifespan of the longer living of the two. Spend $25,000 (discounted) per year on each person, but then do it for 20 years or so per person, and you come up with a figure like $1 million for a couple.
Since the 1970s, real annual benefits have also been growing automatically as wages rise. In fact, the combination of “wage indexing” and failure to adjust for life expectancy schedules Social Security to rise forever faster than the economy.
Then, of course, there are the health care costs. People are getting more years of medical support as they live longer. Plus, the federal government has never effectively tackled the increasing costs that result almost inevitably in a system where you and I can bargain with our doctors over whatever everybody else should pay to support our next procedure or drug.
By the way, none of these calculations account for the decline in the birth rate and its effect on the number of workers available to support such benefit growth. Roughly speaking, the taxes available to support any system decline by about one-third when the ratio of workers to retirees falls from 3:1 to 2:1.
We’ve traveled a long distance from 1935’s legislation and its goal of addressing the needs of people of all ages.
Setting disability policy is tough. Very tough. It’s tough empirically to measure and distinguish among degrees of disability or need. It’s tough legally and administratively to draw boundaries without excluding some sympathetic person or giving an inappropriate level of benefits to someone whose needs can’t fully be assessed. It’s tough economically to transfer resources to people with disabilities without setting up perverse incentives that separate them from the workplace and their fellow workers. It’s tough socially because the needs are so great.
Disability policy has gotten increased attention recently because the Social Security Disability Insurance (SSDI) trust fund is unable to pay our current benefits through 2016. But reform should involve more than money. By defining eligibility for benefits partly by the inability to work, SSDI and other federal disability policies effectively discourage recipients from trying to support themselves. If they work, they lose their benefits. This needs to be fixed. But how?
In a recent conference sponsored by the McCrery-Pomeroy SSDI Solutions Initiative (disclosure: I helped organize the initiative and still serve as advisor), no one advocated reducing benefits to bring SSDI back into balance. Nor did anyone suggest merely raising taxes.
Most speakers talked about the need to modernize US disability policy—in particular, to offer opportunities for people who want to work, can work at some level, or can keep working if they receive help when they first develop a health problem or impairment. Speakers recognized that work is therapeutic and that disability policy should account for the factors that can affect someone’s disability differentially across his or her lifetime, such as the episodic nature of many mental illnesses and the kinds of rehabilitation that can prove helpful to different people at different times.
Disability policy is exactly the same as other policy in one respect: it contains a fairly precise, even if implicit, calculus of what and whom will and won’t be funded. So, any reform to the policy must address the balance sheet.
The President and Congress seem ready to punt on dealing with SSDI, effectively covering shortfalls by transferring money from Social Security Old Age Insurance (OAI). Despite the pretense, such a move is not costless. Today’s taxpayers and beneficiaries won’t need to pay for the growing costs of SSDI, but future taxpayers and beneficiaries—of SSDI or other federal programs—will inherit even higher SSDI or OAI deficits, along with their compounding interest costs. Meanwhile, today’s catalyst for reform is neutralized.
As I listened to the McCrery-Pomeroy conference speakers propose adding work incentives and supports to SSDI (a suggestion I lean favorably toward, despite many design issues), it struck me that the proponents were implicitly suggesting that there is a better way to spend the next SSDI dollars than simply expanding the current program. If those proponents genuinely believe that work incentives and supports are the right way to direct additional dollars, then they also imply that we ought to look at how Congress already has scheduled additional dollars to be spent.
Some advocates may try to claim that we shouldn’t make such marginal budget comparisons. When I co-wrote a book on programs for children with disabilities, my fellow authors and I were criticized in one review for noting simply that the principle of progressivity requires figuring out who needs support the most. Such a critique ignores that we have to make choices, so we may as well do them as best we can. No one, proponent or opponent of current or any reformed law, can get around the simple fact that dollars spent one way cannot be spent another.
Let me put this in terms of the politics. Politicians never want to identify losers because then we voters crucify them. They want to operate on the give-away side of the budget: spending increases and tax cuts. So how can we give politicians some protection to reform disability policy if you and I know that putting relatively more money into work supports changes the nature of SSDI and prevents some other use of the money?
Here’s one way: emphasize the long-term dynamic that reform in a growing economy makes possible. Counting everything from health care to education to disability policy, our social welfare budget now spends about $35,000 annually per household. As the economy grows over time, the number is going to increase—perhaps to $70,000 if the economy doubles in another few decades. We don’t need to cut back on disability programs absolutely in order to allocate a share of those new marginal dollars to different approaches. We simply need to focus the growth of those future budgets.
SSDI and OASI grow automatically over time because benefits are indexed for real growth in the economy over and above inflation. No legislator determines that today’s additional expenditures should be directed one way or another; it’s in a formula set decades ago. Why not consider reducing that automatic growth to finance more subsidies and supports for people who want to keep working? What about capping growth—at least for those getting maximum benefits? What about re-allocating some of the federal health budget, where so many of the dollars are captured by providers rather than consumers, to help pay for work supports?
Or what about cutting back on those features of SSDI that add to the anti-work incentives? For example, what about paring the ability to increase your benefits by about 30 percent if you retire at age 62 on disability insurance instead of old age insurance, at least for people at higher incomes? Or at least not increasing that disability insurance bump, as now happens automatically when the full retirement age increases?
We can shift toward a more modern disability insurance system, but only when we face up honestly to the trade-offs implicit or explicit in every system. We will never move disability policy away from its antiwork emphasis if we’re not willing simultaneously to address the way we put additional resources into the current prevailing system. And, as best I see it, that is just what a scared Congress and president are about to do.
This morning, I testified before the House Ways and Means subcommittee on Social Security. Below is a lightly edited transcript of my spoken remarks. A full copy of my written testimony can be found here.
Contrary to the popular argument that we live in an age of austerity, we live in an age of extraordinary opportunity. Yet, as I argue in a new book, Dead Men Ruling, we block progress by refighting yesterday’s battles and trying to control too much an uncertain future. As one reflection, in 2009 every dollar of revenue had been committed before that Congress walked in the doors of the Capitol.
Looking to Social Security, after three quarters of a century of continual growth, it has largely succeeded in providing basic protections to most, though not all, older people. Now, as psychologist Laura Carstensen at the Stanford Center on Longevity suggests, we should be redesigning our institutions around the new possibilities that improved healthcare, reduced physical demands, and long lives provide. But the eternal automatic growth of Social Security is not conditioned on any assessment of society’s opportunities or needs. Not making best use of the talents of people of all ages. Not child poverty or educational failures or the incidence of Alzheimer’s or autism.
Let me focus on three problems caused by this past, rather than future, focus:
Social Security redistributes in many ways, both progressive and regressive. And in many ways, it fails to provide equal justice.
Among the most outrageous, working single parents, often abandoned mothers, are forced to pay for spousal and survivor benefits they cannot receive, often receiving at least $100,000 fewer lifetime benefits than some who don’t work, pay less Social Security tax, and raise no children.
Similarly, the system discriminates against two-earner couples, spouses who divorce before ten years of marriage, long-term workers, and those who beget or bear children before age 40.
Middle Age Retirement
People today retire for about a decade longer than they did when Social Security first started paying benefits. The biggest winners of this multi-decade policy have been people like the witnesses at this table and members of Congress, who, if married, now get at least $300,000 in additional lifetime benefits.
But there are other consequences: a decline in employment, the rate of growth of GDP and personal income, as well as lower Social Security benefits for the truly old.
Meanwhile, within a couple of decades, close to one-third of the adult population will be on Social Security for one-third or more of their adult lives. There is no financial system, public or private, that can provide so many years of retirement for such a large share of the population without severe repercussions for individuals’ well-being in retirement and the workers upon whose backs the system relies.
The Impact on the Young
Today, lifetime Social Security and Medicare benefits approximate $1 million for a couple with average incomes throughout their working lives, Rising by about $18,000 a year, benefits for a couple in 2030 a couple are scheduled to grow to about $1 1/3 million.
Meanwhile, the rate of return on contributions falls continually for each generation. Each year of delayed reform shifts more burdens to younger generations from older ones, with the largest impact on groups like blacks and Hispanics, in part because they comprise a larger share of those future generations with lower returns.
In summary, each year of delay in reforming Social Security:
- Continues a pattern of unequal justice under the law;
- Threatens the well-being of the truly old;
- Increases the share of benefits paid to the middle aged;
- Leads government to spend ever less on education and other investments;
- Contributes to higher nonemployment, lower personal income and revenues; and
- Increases the burden that is shifted to the young and to people of color.
Our updated numbers for lifetime Social Security and Medicare benefits and taxes are now available, based on the latest projections from the Social Security and CMS actuaries for the 2013 trustees’ reports for OASDI and Medicare. Couples retiring today, with roughly the average earnings of workers in general, as well as average life expectancies, still receive about $1 million in lifetime benefits. This number is scheduled to increase significantly for future retirees and is higher for those with above-average incomes and longer life expectancies.
Little has changed on the Social Security side from our previous estimates, as the program has undergone no significant reform in recent years. Our estimates of present values of Medicare benefits for future retirees have decreased slightly from last year as slower health care cost growth has made its way into projections. By 2030, Medicare benefits (net of any premiums paid) are about 90 percent of last year’s estimates, still a significant multiple of Medicare taxes paid. (As in 2012, our numbers incorporate a Medicare cost scenario that assumes the “doc fix” and other adjustments will be extended, not the “current law” scenario in the trustees report.)
We have been publishing these numbers for a long time—and not without controversy over our intent. Our hope is simply that better and more complete information will help elected officials decide whether Social Security and Medicare are distributing taxes and benefits in the fairest and most efficient way possible, a decision we do not believe possible by looking only at annual numbers or how current, not future, retirees and taxpayers might fare. Therefore, we are delighted that in its most recent Long-Term Budget Outlook, the Congressional Budget Office for the first time also published estimates for lifetime Medicare benefits and taxes, as well as Medicare and Social Security combined. Using a slightly different methodology, CBO produces very complementary results. Differences derive from it using median-wage (rather than average-wage) workers, a 3 percent (rather than a 2 percent) real discount rate, and an assumption of Social Security claiming at 62 (rather than 65). As CBO also notes, expected benefits (and taxes, to a more limited extent) have grown over time for a number of reasons, including longer life expectancies, higher incomes, and rising health spending per person.
In a new brief, my colleagues and I examine how many features of Social Security combine to redistribute money among racial and ethnic groups over long periods of time, combining together generations. We find that the program as a whole, and especially its retirement portion, has likely redistributed from blacks, Hispanics, and other racial minorities to whites.
The major cause? When taxes are compared to benefits for each generation, the early generations of retirees got large windfalls, those retiring today come closer to breaking even, and tomorrow’s retirees on average will get back less than they pay in, assuming some modest interest rate could have been earned on their contributions or taxes. This phenomenon by itself wouldn’t cause interracial redistribution, but whites disproportionately occupy the high-return generations, while Hispanics, more recently immigrated groups, and blacks increasingly occupy the lower return generations.
The immigration part of the story is easy to understand. Hispanics and other recent immigrants weren’t around to receive the windfalls that came about as the system expanded over its early decades. Whenever Congress increased lifetime benefits for retirees and near-retirees, younger workers would be required to contribute for decades to support those higher benefits. Older workers would get those higher benefits with fewer years of additional contributions or, at times, none at all.
As for the redistribution from blacks, they disproportionately occupy the lower-return generations because of their larger family sizes. A simple analogy might be made with a two-family world where one couple has three kids and the second has one, but the kids all contribute the same $1,000 each to support the four parents, who all get the same benefit of $1,000 each. The larger family contributes $3,000 and gets back $2,000. The effect will be permanent unless some other redistributions over time offset this windfall gain for the smaller family.
Of course, these simple stories ignore Social Security’s other moving parts. Its progressive benefit formula, which I consider in many ways brilliant because it is based on lifetime rather than annual earnings, redistributes to those with lower lifetime earnings. At the same time, Social Security contains many regressive features. For instance, as a protection for old age it appropriately requires that payments be made in the form of annuities, but that ends up redistributing to those with higher incomes because they have higher-than-average life expectancies. Also, unlike private pensions where it would be illegal, single heads of household, often lower-income women, are required to contribute for spousal and survivor benefits they can’t receive.
Because of these various offsetting features, the retirement or old-age part of the system exhibits little net progressivity even before adding on this new multigenerational consequence of making each successive generation pay more for each dollar of benefit it receives.
When disability insurance is considered, it adds to significantly to progressivity in the sense of redistributing to those with lower incomes. From a broader perspective, of course, one can hardly consider it a plus that some racial and ethnic groups incur higher levels of disability, along with the huge losses of private income that Social Security does not replace.
The study does not contain policy prescriptions, though my own separate work has for very long time suggested the merits of substantively (not just symbolically) higher levels of minimum benefits and other progressive adjustments. Nor does the study contradict the success of Social Security in reducing poverty dramatically among the elderly, particularly in its early decades. It does suggest that as reform is being considered, we give serious attention to whether the system achieves its stated objectives, including the extent to which it really provides better protection for those individuals and classes who are less well off.
Both the left and right, I think it fair to say, have presumed that the system is far more progressive than it turns out to be—a more recent literature finding to which this study adds. Unfortunately, the Social Security debate, like so many we have today, tends to be argued on a thumbs-up or thumbs-down basis, rather than on how it might be better designed to meet society’s objectives.
Excerpt from “Reforming Social Security Benefits,” Testimony Before the House Ways and Means Subcommittee on Social Security.
In this testimony, I would like to focus on the need for Social Security benefit reform regardless of the current imbalances in the system or the taxes raised to support the system.
Why? Despite Social Security’s great success, its growth in lifetime benefits over time has been decreasingly targeted at its major goals. Even while programs for children and working families are being cut, combined lifetime benefits for couples turning 65 rise by an average of about $20,000 every year, so that couples in their mid-40s today are scheduled to get about $1.4 million in lifetime benefits, of which $700,000 is in Social Security.
Social Security has morphed into a middle-age retirement system. Typical couples are receiving close to three decades of benefits. Smaller and smaller shares of Social Security benefits are being devoted to people in their last years of life.
If people were to retire for the same number of years as they did when benefits were first paid in 1940, a person would on average retire at age 76 today rather than 64. Soon close to a third of adults will be on Social Security, retiring on average for a third of their adult lives.
While Social Security did a good job reducing poverty in its early years, it has made only modest progress recently, despite spending hundreds of billions of dollars more. The program discourages work among older Americans at the very time they have become a highly underused source of human capital in the economy.
The failure to provide equal justice permeates the system. It discriminates against single heads of household, spouses with relatively equal earnings, those who bear their children before age 40, long-term workers, and many others. At the same time, private retirement policy leaves most elderly households quite vulnerable.
Unfortunately, the Social Security debate has largely proceeded on the basis of being “for the box” or “against the box.” The contents themselves deserve scrutiny.
How might one break through the stalemate and find areas of mutual agreement? While I applaud the efforts of the Simpson/Bowles Commission and the Bipartisan Policy Commission I believe we can go much further to address the problems I just raised. How? We should start with a basic set of principles and see where they lead us.
Consider. Inevitably balance will be paid for mainly through benefit cuts or tax increases on higher income individuals who have most of the resources. That debate need not derail other needed reforms. I suggest proceeding in the following order:
First, consider reforms aimed at meeting Social Security’s primary purposes:
- providing greater protections for those truly old or with limited resources;
- supporting the work and saving base that undergird the system; and
- providing more equal justice for those suffering needless discrimination in the system, like single heads of household and longer-term workers.
Some of those fixes cost money, and some raise money; we don’t have to address trust fund and distributional consequences in each and every change.
Second, further adjust minimum benefits and the rate schedule and indexing of that schedule over time to achieve final cost and distributional goals. The extent of these adjustments will also depend upon the tax rate and base structure agreed upon.
My testimony provides a fairly detailed way to engage this type of reform process. It largely follows the logic I applied to taxation when serving as the economic coordinator of the Treasury effort that led to the bipartisan-supported Tax Reform Act of 1986, and in my testimony before the Simpson/Bowles Commission.
Arithmetic tells us we must either decrease the growth of Social Security spending or increase taxes as a share of gross domestic product.
But we should do it with an eye on fairness, growth and efficiency. We’re all in this together, so higher-income families must give up something to deal both with Social Security shortfalls and those in the budget more generally. A modest increase in the wage base for Social Security has some justification since that base has eroded in recent years. But if extended too far, it exacerbates the squeeze on other government programs. How? On the tax side, it tends to preempt other tax increases for non-Social Security purposes. On the benefit side, it attempts to maintain a growth rate of Social Security and other elderly programs that absorb more than all of the scheduled growth in government spending for decades to come, thus continuing a downward spiral in the share of the overall budget devoted to children, education and investment more generally.
Under current Social Security formulas, ending the cap on income would mean that some fairly wealthy individuals would get benefits in excess of $1 million. Though no one thinks that that makes sense as a benefit schedule, capping benefits goes against the Social Security tradition of being paid back for additional contributions. On the technical front, an unlimited Social Security tax would also encourage individuals to reclassify labor income as capital income not subject to Social Security tax. This would be a special problem for the self-employed and owners of partnerships, since Social Security now taxes both their capital and labor income as labor income.
Finally, the Social Security Administration’s Office of the Actuary found that even with a cap on benefits, the wage base expansion would still leave the program running future deficits. We shouldn’t pretend that it does otherwise.
This column was reposted from New York Time’s Room for Debate.