Common Sense on Social Security

Home of the Social Security Solvency Simulator

An Objective, Bipartisan Resource
on Social Security Reform

Our work is guided by three principles:   (1) Facts are more important than ideology;   (2) The public deserves access to the facts;   (3) Provide simulation tools so people can make sense of the issues.

        Sensibly-Asked Questions (SAQ's)

What do I need to know about how Social Security works?
How does Social Security calculate benefits?
Is there really a long-run financing problem?
"Insolvency." That sounds scary. What does it mean?
What's the best way to fix Social Security?
Wouldn't it be better to talk about Funding Social Security?
What does it take to make Social Security's benefit commitments affordable?
Slowing the growth in New Retiree Benefits - Won't this hurt?
A recap - do I have it right?
Why is this issue so polarized?
Is there anything I can do to reduce the polarization?

Return to Home Page

What do I need to know about how Social Security works?

Social Security collects money and it distributes money.

It collects money from employees (the payroll tax, at 6.2 percent of the employee's salary or wage). It collects money from employers (the employer match, also at 6.2 percent). It also collects a modest tax on retirement benefits, paid by higher income retirees. In addition, it earns interest on the bonds it has sold to the U.S. Treasury.

It distributes money. It distributes money to retirees who've paid in during their working careers. It pays money to the surviving spouses of eligible retirees who have died. And it pays money to working age individuals who have become disabled. (This was an abstract point to me until my brother Eric was hospitalized with congestive heart failure, then received a heart transplant. He received Social Security disability payments for a year, then was healthy enough to return to work.)

As of now, Social Security collects a bit more money than it pays out. By law, it uses the surplus to purchase special bonds from the U.S. Treasury. Within the next fifteen to twenty years, Social Security will find itself paying out more than it takes in. At that time, it will first ask the Treasury for cash interest payments on its bonds. When that's no longer enough, it will sell its bonds back to the Treasury. (Which, presumably, will be forced to sell new bonds in order to repurchase bonds from the Social Security Administration.)

Social Security's administrative costs for all this activity are quite low, roughly one half of one percent on retirement benefits, 1.5% on disability benefits. Top

How does Social Security calculate benefits?

Social Security has a set of formulas for calculating benefits. You don't need to know the details just now, but it's important to understand three key facts. When you retire, Social Security goes through an indexing process to figure out your average monthly earnings. They index all your earnings, from your first paycheck to your last, using an inflation-based index.

Next, Social Security goes through another indexing process to figure out what I will call "Your New Retiree Benefits." The New Retiree Benefit indexing process is NOT based on inflation. It is based on WAGE GROWTH.

It works roughly like this. If wages grow an average of 1% a year for an extended period of time, each year's New Retiree Benefit will be approximately 1% higher than last year's. If wages grow an average of 2% a year, each year's New Retiree Benefit will be approximately 2% higher than last year's.

Finally, those who are already retired will see a slow increase in benefits. Benefits for existing retirees grow at a rate which is indexed, again, to inflation.

To repeat. Social Security uses an inflation index to calculate your average monthly earnings for your career. It uses a wage index to ratchet up the New Retiree Benefit every year. And it uses an inflation index to boost retirement payments for those who have already retired. Top

Is there really a long-run financing problem?

Yes. There is a problem. But it's not a simple problem. There are two parts to it.

First of all, the American population is changing. Today, in 2004, only 12 percent of all Americans are 65 years old or older. That percentage is due to change, sharply, beginning in 2008. By 2033, a quarter of a century later, 20 percent of all Americans will be 65 years old or older.

Here's the sort of number that curls your hair if you're trying to figure out how Social Security is going to be able to pay benefits once the Baby Boomers all retire. From 2005 to 2035, the "65 and Over" crowd is due to grow in size from 37 million to 75 million, an increase of 38 million. In other words, the number of Americans 65 and over will double in size in the next thirty years. And, guess what? Most of them are going to expect Social Security benefits.

Meanwhile, the working age population, folks ranging in age from 20 to 64, is expected to expand by a measly 21 million, from 181 million to 202 million.

Remember who receives money from Social Security? Older folks. There'll be 38 million more older folks. Remember who pays money in? Working folks. There'll be 21 million more working age folks, and, most likely, only three-quarters of those will be in the work force.

So, yes, there is a problem, due to changes in the sizes of different age groups.

But, second, the perception of the problem is also an artifact of how Social Security puts together its long-range forecasts. Social Security's forecasters are trained actuaries, numbers folks, and, since I'm a numbers person too, I can relate.

They're also quite cautious. They expect rather slow GDP growth in the decades ahead. Maybe 0.4% a year for population growth. Maybe 1.2% a year for productivity growth. For a combined GDP growth rate of 1.6%. Not very fast.

Historical growth rates for the U.S. have been in the range of 3.3% a year (above and beyond inflation). Population growth of roughly 1.3% a year, productivity growth of roughly 2% a year.

Social Security actuaries analyze population growth rates, and they scale back the population portion of the GDP estimate. They look at slower productivity growth rates since 1970, and they scale back productivity estimates. All well and good, perhaps. And perhaps not.

When Social Security scales back its GDP growth estimates, Social Security's expected financial crisis ends up looking considerably worse. A faster GDP growth estimate produces a somewhat less drastic forecast of Social Security's impending insolvency. Top

"Insolvency." That sounds scary. What does Social Security mean when it says "insolvency"?

When Social Security uses the word "insolvency," my natural reaction is to say, "Oh, no! They're going broke. They're going to run out of money. And when they do, a squad of repo toughs in pickup trucks will land on their front doorstep and haul away their refrigerators and their office furniture." That's my natural reaction, and, perhaps, yours as well. But it's wrong.

When Social Security says "insolvency" that doesn't mean it'll be out of money. Remember, it still collects a payroll tax equal to 12.4% of all taxable wages and salaries. That's roughly five hundred billion dollars a year, and rising.

"Insolvency," as Social Security actuaries use the term, means that Social Security won't have enough money to cover ALL its benefit payments. It'll still have plenty of money to cover MOST of the benefit payments. Just not all. No burly repo guys hauling away Social Security's refrigerators. Top

What's the best way to go about fixing it?

There's a right way and a wrong way. Let's start with the right way. The right way to go about fixing it is to use logic and reason, open minds and a willingness to listen. Everybody has something to offer. Nobody has all the right answers.

Then we have to come up with a workable way to define the problem. Here's my suggested framing: As a result of significant population changes, Social Security has an affordability problem. It can't quite afford to pay benefits at the rate the law requires.

If the benefit formula could be adjusted, the problem could be resolved. A more affordable way of defining benefits can keep Social Security's two cash flows in balance. If money OUT to beneficiaries doesn't exceed money IN from payroll taxes, over the long run, Social Security's affordability problem has been taken care of.

The best way to have a reasonable discussion about the affordability problem is to set a benefit target for the future that's reasonably affordable. Maybe not fully affordable under Social Security's slow growth forecast, but fully affordable under a more optimistic (and realistic) forecast of future American GDP growth.

Then our President and our Congress have two sets of work to do.

One - figure out a reasonable set of options for hitting the target on the "money paid out" side. If the long run target were 80 percent, say, it would be Congress' job to develop four or five legitimate options for calculating benefits, so that all the options hit the affordability target.

Two - figure out a reasonable set of options for hitting the target on the "money coming in" side.

No restrictions on options allowed. Democrats must agree to listen respectfully to proposals for personal investment accounts. Republicans must agree to listen respectfully to proposals that make no use of personal investment accounts. Honest, collaborative dialogue is essential, with all parties obligated to work together to develop fair proposals that satisfy the affordability target. Top

Is "affordability" really the right target? What about "funding" our retirement system? Isn't it better to have a "funded" retirement system than an "unfunded" system?

That's an excellent point. Consider the following ways of slicing the issue. An "unfunded" retirement program depends on pay as you go cash flow. A "funded" retirement program depends on earnings from assets. A "partially funded" retirement program depends on both, pay as you go cash flow, plus earnings on assets cash flow.

Social Security is essentially set up as an "unfunded" retirement program. The 1983 Greenspan Commission muddied the waters with its neither fish nor fowl solution. Fund Social Security for awhile, then draw down the funds. Grow assets, then liquidate assets. Fund, then De-Fund.

Right now we're in the waning years of the "Grow Assets" cycle, with Social Security running a surplus, and investing its surplus in Treasury Bonds.

But everything turns around in the next decade or decade and a half. The surplus cash period ends. The asset liquidation period begins.

With all due respect to Mr. Greenspan, that's not a smart way to design a retirement system. Let's consider three options - unfunded, fully funded, and partially funded.

Unfunded. A system that's meant to be unfunded wouldn't bother to build up a surplus. It would strive for balanced cash flow, through ongoing benefits adjustments and payroll tax adjustments.

Fully Funded. The second hypothetical option, a fully funded Social Security system, one that pays all benefits from earnings on investments, may be an attractive notion but it's a real world impossibility. US financial assets are nowhere near sufficient to cover all of Social Security's long-run obligations and still have enough left over for other investors.

Partially Funded. The third option, a partially funded Social Security system is certainly possible, either through modest Trust Fund investments, or through a modest individual account system, or both. Asset earnings would then cover a small portion of benefits on a permanent basis. If Social Security is to be partially funded, though, it should accumulate a stable pool of assets, then hang onto its assets. It shouldn't spend three decades accumulating, then three more decades liquidating. That's wrong-headed.

Bottom line. In a rational world, with open-minded legislators willing to examine all the options, an unfunded strategy for Social Security can be a reasonable option. A partially funded strategy for Social Security can be reasonable as well. A fully funded strategy is impossible. The current strategy, though, "Build Assets, then Liquidate Assets," is very badly designed and takes us in a direction we really don't want to go. Top

Okay, I get the point about "funded retirement." A little funding could be a good thing. A lot of funding is out of reach. Back to affordability. What's it take to make Social Security affordable?

Here's the simplest way to think about it. Today's benefit formulas, which tie New Retiree Benefits to rising wages, as explained above, are wonderful in a (non-existent) world where the "65 and over" crowd is small and stays small.

Having New Retiree Benefits tied to rising wages in today's world is a big factor in putting affordability out of reach. Suppose the formula that defines New Retiree Benefits were to be adjusted downward just a bit. Suppose it were designed to grow New Retiree Benefits at a slightly slower rate than rising wages. If wages were to rise by 1% a year from now on, Social Security might grow its New Retiree Benefit level by one-third of a percent a year. If wages were to rise regularly by 2% a year, New Retiree Benefits could be grown by 1% a year.

Not forever. But for the next twenty or thirty years.

Then, after affordability has been achieved, after cash flowing out is in balance with cash received, after benefit payments flowing out and payroll taxes coming in are in balance, Social Security could return to its old practice. New Retiree Benefits could once again be grown at the same speed as wage growth. They wouldn't be as high as current formulas say they should, but they'd be higher than today, and they'd rise every year.

(This isn't the only path to affordability, of course. There are many others. But it's the right place to start, and it gives us a good frame of reference against which to measure other options.) Top

Slowing the growth in New Retiree Benefits. Won't this hurt?

Yes and maybe. If GDP growth is as slow as Social Security's actuaries cautiously forecast, this approach basically freezes the New Retiree Benefit for the next three decades or so. If, on the other hand, GDP growth moves forward a bit more smartly, there's room for a bit of upward growth in the New Retiree Benefit even as Social Security moves toward affordability.

As the Simulator demonstrates, a higher GDP growth economy makes it possible to achieve solvency, to achieve affordability, and still permit New Retiree Benefits to rise (though not as fast as now). Top

Okay. To recap. Social Security uses an inflation index to calculate my career wage average. It uses a wage growth index to calculate my New Retiree Benefit. And it uses an inflation index to adjust my retirement benefits upward once I'm already retired. If the New Retiree Benefit formula were changed to grow New Retiree Benefits a little faster than inflation but a little more slowly than real wage growth, affordability could be achieved. If GDP growth is as slow as Social Security's forecast, this will take a while. If GDP grows more quickly, Social Security affordability can be achieved more quickly and comfortably. Do I have it right?

That's right. You've got it.

If it's that simple, why has this issue become so polarized?

Regrettably, logic and analysis have been trumped by politics and ideology. Both camps are guilty on this one.

Conservatives have launched an attack on Social Security for ideological reasons. Conservative critics think Social Security's "social insurance" philosophy is immoral. Every future retiree for himself, they cry, but it won't hurt, because they'll all invest in the stock market and they'll all get rich. The upshot? Republicans favor a deeper cut in regular benefits than might be wise, because they're willing to bet that PRA's will more than make up the difference.

Liberals have countered by calling the solvency issue "a two percent problem." Raise the payroll tax by two percentage points, and, voila, Social Security is in actuarial balance. The upshot on the liberal side? Democrats resist making any adjustments to Social Security benefit formulas, on the premise that Social Security's pending shortfall isn't nearly as serious as critics claim.

Ideological smoke. Political games.

Conservatives are wrong about the stock market's capacity to absorb all the investments they hope to generate using Personal Retirement Accounts. They're also wrong about long-range stock market returns. (See "Are Seven Percent Returns Realistic?" elsewhere on this website.)

The problem, though, is not that they're wrong. The problem is deeper than that. The heart of the problem is that they seem not to know that they're wrong. They've never been curious enough to figure out if their individual investor logic will hold together if it's scaled up to macro-economic levels. Had they ever stopped to give their proposals the analytic scrub they deserve, they'd have found out long ago that their promises don't hold water. (See Simcivic Update #5.)

Liberals are wrong in a different area. They consistently misrepresent the size and structure of the problem.

And, again, the deeper problem is that apparently they don't even know they're wrong. They seem to believe an actuarial balance of zero, as calculated by Social Security, indicates lasting solvency. It doesn't. An Actuarial Balance of 0.0% indicates one thing, and one thing only - that Congress's current "Grow Assets, Then Liquidate" strategy reaches its insolvency point just AFTER Social Security's forecasting period ends. (See Simcivic Update #3.)

By endorsing Actuarial Balance as their test of solvency, liberal supporters of Social Security end up endorsing today's nonsense strategy of "Grow Savings, Then Liquidate Savings." Top

Is there anything I can do to reduce the polarization?

On this particular issue, polarization thrives on ideological smoke. The first task it to blow away the smoke. As always, in a genuine democracy, getting this done is up individual citizens.

Teach yourself how to see through the smoke. Then start assessing your public officials. Figure out whether they've taught themselves to see through the smoke too.

You might wish to develop out your own "Report Card" for elected officials and news editors on Social Security. An 'A' indicates in-depth understanding, an aversion to ideological smoke, a willingness to weigh different points of view, a sense of fairness. An 'E' indicates shallow understanding, soft thinking, a willingness to side with ideological partisans.

Then take some time to talk to opinion leaders and elected officials. Quiz them carefully.

Figure out how well, or poorly, each one understands the issues. As you figure this out, one interview at a time, assign grades to each. Once you've interviewed enough of your local leaders, on all sides of the political spectrum, announce the results of your research. Write letters. Hold a local press conference. Print leaflets. Get your friends to join you.

If enough citizens support a common sense understanding of Social Security, if enough citizens do what's needed to blow the smoke away, the climate will change.

The key is embarrassment. Once the nonsensical positions are adequately exposed, once a common sense view gains its footing, officials will find it embarrasing to defend the leaky arguments of the ideologues. This is the tipping point that's needed - the point at which common sense becomes stronger than nonsense.

A rejection of ideology that's meant to polarize rather than solve is the beginning of collaboration. It's the beginning of dialogue, the beginning of reform, the beginning of workable long-range solutions. Everything begins as it always does in a democracy, with individual citizens who take the trouble to understand the real issues. Top

Return to Home Page

Common Sense on Social Security
An initiative of The Wallcharts Workshop
A Non-Profit Successor to the Collaborative Democracy Project

Questions?  Reactions?  Feedback?
You can reach us at

For more information on Social Security, the following web sites are suggested

The Concord Coalition

The Social Security Administration

Page Version 1.02
Revision Date April 13, 2006