Is Social Security Worth Its Cost?

Report Budget and Spending

Is Social Security Worth Its Cost?

July 10, 2018 25 min read Download Report

Authors: Kevin Dayaratna, Rachel Greszler and Patrick Tyrrell


Americans would be better off keeping their payroll tax contributions and saving them in private retirement accounts than having to contribute to the government’s broken Social Security system. Social Security’s design has, over the decades, presumed that many Americans are too incompetent to make informed decisions for themselves, but few Americans believe that the government knows better than they do what is best for them and their families. Moreover, Social Security’s financial structure effectively guarantees that workers will receive extremely low—or even negative returns—on their payroll taxes.

Key Takeaways

This report compares what Social Security can provide and what workers could receive if they had ownership of their Social Security payroll taxes.

This information can help individuals of all ages understand what they can expect to receive from Social Security or from private savings.

Virtually all Americans would be better off keeping their payroll taxes and saving them in private retirement accounts.

Social Security began as an anti-poverty insurance program, aimed at preventing workers from outliving their savings when they were no longer physically able to work. As such, Social Security was limited in nature, beginning as only a 2 percent payroll tax—and promising to never take more than 6 percent of workers’ pay. Today, Social Security’s Old Age and Survivors Insurance (OASI) retirement program takes 10.6 percent of workers’ pay, and its Disability Insurance (DI) program takes another 1.8 percent, for a combined total of 12.4 percent. This is more than most Americans pay in income taxes.

As Social Security has grown in size and scope, it has become more than just an insurance and poverty prevention program—and with millions of seniors living below the federal poverty line, it is not doing a great job even at that. Having reduced the incentive to save for retirement, Social Security now represents a significant portion of most workers’ retirement savings. Despite the fact that Social Security was intended to be an insurance program, providing a secure retirement income, individuals have no legal claim to their scheduled Social Security benefits, as the program can only pay out as much money as it has on hand and Congress can change benefit levels if it wants. Not surprisingly, more than 60 percent of workers under the age of 50 do not think Social Security will be able to pay them a benefit when they retire.

With Social Security consuming such a large component of workers’ paychecks and offsetting their own private savings, it is important that workers receive a valuable benefit from Social Security—one at least as good as they, as a whole, could obtain from saving on their own. This analysis looks across the United States and across generations to see if Social Security does in fact provide that.

Utilizing federal government data on life expectancy and earnings in each state, Heritage analysts found that:

  • On average, personal retirement savings significantly outperform the current Social Security system. Taking an average of all 50 states and the District of Columbia, the average worker receives significantly less from Social Security than he would have if he had conservatively invested his Social Security payroll taxes in the market.
  • Foregone benefits vary across generations. For average-earner males in Florida (a median-income state), lost investment earnings equal over $600,000 for those born in 1955; over $700,000 for those born in 1975; and over $1.1 million for those born in 1995. For average-earner females living in Florida, Social Security will provide over $190,000 less in lifetime income than personal savings for those born in 1955; over $230,000 less for those born in 1975; and over $420,000 less for those born in 1995. (See Appendix Tables.)
  • Individuals with lower life expectancies often lose greatly. This occurs because they receive little or nothing in benefits and cannot pass along all their lost contributions to their surviving family members. Consequently, certain areas of the country, including Washington, DC, have significantly lower returns from Social Security.
  • Younger workers face lower, and even negative, returns from Social Security compared to older workers. This comes as a result of paying higher average Social Security tax rates over their lifetimes, coupled with a two-year increase in Social Security’s normal retirement age—as well as the benefit cuts that will occur if Social Security’s trust fund runs dry.

To see if Social Security is a worthwhile program for Americans—across generations and states—researchers at The Heritage Foundation’s Center for Data Analysis created a statistical model, The Heritage Foundation Social Security Rate of Return Model, to examine Social Security’s costs and benefits. We compare these results to what workers would have earned (including estimates on what younger workers likely would earn) in personal retirement savings accounts.

Social Security: Origin and Intent

Established during the Great Depression in the 1930s as part of President Franklin Delano Roosevelt’s New Deal, Social Security is a federal program designed to protect against poverty in old age. At that time, Social Security’s eligibility age of 65 was higher than the average life expectancy.REF Social Security’s payroll tax began at a rate of 2 percent and was never intended to rise above 6 percent.REF Those taxes seemed sufficient because life expectancy was 17 years lower then than it is today; not many people lived long enough to collect benefits, and those who did collected them for less time than retirees today.REF

However, what started out small has morphed into a nearly $1 trillion annual program that redistributes income to 61 million people—or about one out of every five people in the United States (including about 50 million OASI beneficiaries and 11 million DI beneficiaries).REFWhen Social Security first began, there were 42 workers paying into the system for every one retiree receiving retirement checks. Today, there are only 3.4 workers per OASI beneficiary.REF The program has long been on an unsustainable path and will run out of funds to pay promised benefits in 2029 according to the Congressional Budget OfficeREF and in 2035 according to the Social Security Trustees.REF Absent reforms, benefits will decline by over 20 percent across the board after the Trust Fund runs dry.REF Knowing the estimated benefits workers will get from Social Security versus what they could get by saving on their own can help workers and policymakers better assess what types of Social Security reforms would be most beneficial.

Social Security’s payroll tax consumes 12.4 percent of workers’ paychecks (10.6 percent for the OASI program and 1.8 percent for the DI program)—but that is not enough to sustain the programs.REF Maintaining current benefit levels for both OASI and DI would require the payroll tax to rise to 15.33 percent.REF Unfortunately, the payroll tax reduces workers disposable incomes and provides many with an incentive to save less. Therefore, a significant percentage of older individuals today rely primarily on Social Security for retirement income.REF Because Social Security no longer functions primarily as a poverty-prevention program for individuals who are too old to work, and since it consumes so much of workers’ savings capacity, it is important to quantify whether or not Social Security is a valuable savings program.

Why Rates of Return Matter

“Rate of return” is a widely used metric to measure the performance of an investment—that is, how much a given dollar returns over a specified period of time. If $100 invested today is worth $110 in one year, then it has a 10 percent annual rate of return that year. Since workers contribute payroll taxes and expect to receive something in return, Social Security is considered an investment by many people.

In reality, however, Social Security is not an investment. For starters, today’s cash-flow deficits within the program mean that all incoming payroll taxes go immediately out the door to pay promised benefits. Moreover, Congress ultimately determines what workers pay into the system and receive from it—leaving workers with no control.

Social Security’s rate of return, or what workers get back in comparison to what they pay in, is entirely determined by Social Security’s benefit calculation formula as well as its Trust Fund assets. Consequently, the “returns” individuals receive from their Social Security contributions vary wildly across individual workers and across generations. While Social Security provided a high return on payroll taxes to its early beneficiaries, it promises a much lower return to future beneficiaries—and, under certain scenarios, the actual return that it can currently afford to pay to millions of future retirees can even be negative.

We used the Heritage Foundation Social Security Rate of Return Model to compute the program’s internal rate of return under two scenarios: (1) current law (which assumes the trust fund runs dry in 2035 and benefits are cut by over 20 percent);REF and (2) maintaining promised benefits through an across-the-board payroll tax increase. Although some individuals do not live long enough to collect Social Security benefits, we only report rates of return for those who live to at least age 66 (the current full, or normal retirement age). Consequently, our results overstate Social Security’s actual rate of return because they exclude individuals who die before age 66 and receive little, if anything, in return for their Social Security contributions.REF

We compare these Social Security returns to a third scenario in which workers are hypothetically able to invest their payroll taxes in private accounts made up of stocks and bonds, as opposed to relying on benefits paid from other workers. These private account simulations assume that workers conservatively invest 50 percent of their existing payroll taxes in federal government bonds and the other 50 percent in large stocks.REF We applied historical rates of return for stocks and bonds through 2016 and projected forward the historical average (from 1954–2016) for 2017 and beyond (2.75 percent for government bonds and 7.04 percent for large-cap stocks). Full details are available in our methodological appendix.

Key Assumptions and Methodology of Heritage Foundation Social Security Rate of Return ModelREF

This analysis utilized the Heritage Foundation’s Social Security Rate of Return Model, which incorporates the following assumptions:

  • The hypothetical amount invested in personal retirement savings accounts equals Social Security’s payroll tax, which is 10.6 percent under current law (which requires benefit cuts beginning in 2035 to maintain solvency of the Social Security Trust Fund), and 13.28 percent under a financially solvent system that can maintain “promised” benefits. We do not make any changes to Social Security’s Disability Insurance (SSDI) program in this exercise, nor do we include the 1.8 percentage point SSDI payroll tax in workers’ hypothetical personal retirement account contributions.
  • We display results separately for both male and female individual workers with: (1) average earnings; (2) 50 percent of average earnings; and (3) the taxable maximum ($127,200 in 2017).
  • For state-by-state analysis, we make use of the fact that average incomes and life expectancies vary by state.
  • We account for future increases in life expectancy and wages. Unless otherwise stated, we use the intermediate assumptions reported in the Social Security Trustees’ 2017 annual report.REF
  • We adjust all Social Security benefits for inflation according to the Social Security Administration’s use of the Consumer Price Index for Urban Wage Earners and Clerical Workers (CPI-W). For discounting previous and future values of wages, investments, and returns, we use the Consumer Price Index for All Urban Consumers (CPI-U) to remain consistent with our source for inflation-adjusted returns (2017 SBBI Yearbook).REF We use the Congressional Budget Office’s January 2017 projections for future inflation in the CPI-U.REF
  • We include both the employee’s and employer’s shares of payroll taxes in the calculations.REF
  • The rate of return on private investments listed in the tables assume a conservative mix of 50 percent large-cap U.S. stocks and 50 percent U.S. Treasury bonds. We also provide information about how much workers pay in OASI payroll taxes over their careers and how much they receive in Social Security benefits (assuming they live the average life expectancy), compared to how much they could have accumulated in a private retirement account had they invested their payroll tax dollars.

National Analysis

We used the Heritage Foundation Social Security Rate of Return Model to determine how American workers born in 1995 and with differing life expectancies fare in terms of the expected rate of return they receive from Social Security (the amount they receive from Social Security compared to what they paid in payroll taxes, all in 2017 dollars):

BG3324 Table 1

BG3324 Table 2

These results illustrate that Social Security provides extremely poor—often negative—rates of return for younger workers all across the country, especially for individuals with lower life expectancies. In comparison to the current system, private retirement accounts would provide significantly higher returns to Americans, regardless of their life expectancies.

State-by-State Analysis

We also used The Heritage Foundation Social Security Rate of Return Model to conduct a state-by-state analysis of the rate of return of Social Security versus a hypothetical simulation assuming payroll taxes used for Social Security were instead invested in private accounts. The following tables provide, by state of residence, the rate of return of Social Security compared to private accounts.

These tables, as well as the more detailed tables in our appendix, show workers’ estimated total lifetime Social Security benefits compared to their accumulated personal retirement account balances under two scenarios—tax increases or benefit cuts—that would make Social Security solvent over the long run. We also break total benefits and accumulated account balances down into monthly benefits and monthly annuities that workers could purchase using their personal retirement account balances. For both measures, we show workers by state, gender, and income level:

BG3324 Table 3

BG3324 Table 4

BG3324 Table 5

BG3324 Table 6

BG3324 Table 7

BG3324 Table 8

BG3324 Table 9

BG3324 Table 10

Our results demonstrate that younger workers (born in 1995) of all income levels would receive between two and seven times as much in retirement income from personal savings as they would from Social Security. Because Social Security paid out more in benefits than it collected in tax revenues to earlier generations, the gains would not be as large for older workers (those born in 1955 and 1975); however, they would still receive more from personal retirement accounts than from Social Security.

Why Private Savings Produce Higher Retirement Incomes than Social Security

Our results establish that personal investment accounts provide a significantly greater rate of return compared both to what Social Security can afford to pay as well as what it has promised to pay. Prior research comes to the same conclusion.REF

1. Higher Returns and Larger Retirement Incomes. Investment returns in the private market are two to three times the rate of return of U.S. Treasuries (what Social Security used to invest in when it ran surpluses). Consequently, even most low-income earners who get the most back from Social Security (relative to what they contribute) would end up with higher retirement incomes if they had personal investment accounts. This phenomenon was not always true in the past under Social Security’s unsustainable promises, but it is true for anyone in their 40s or younger today.

For example, a 23-year-old female, born in 1995, living in Florida and earning the average wage can expect a monthly Social Security check of $1,393 in 2017 dollars when she retires at age 67 in 2062. However, if her payroll taxes had been invested half in large-cap stocks and half in Treasury bonds, based on historical averages she could buy an inflation-adjusting annuity (similar to Social Security) that would provide $2,524 per month in year 2017 dollars.REF This amount represents almost twice the amount Social Security will be able to pay.

The same is true for other earners. A 23-year-old male, born in 1995, living in Florida and earning only half the average wage throughout his working years would accumulate enough in a personal account to receive an annuity that would pay $3,093 per month in 2017 dollars. Social Security can only provide him significantly less at $1,551 per month.REF Even low-earning females (born in 1995) who tend to receive the most bang-for-the-buck from Social Security would receive 40 percent more from a personal account than from Social Security ($1,262 per month from a personal account vs. $902 from Social Security).

BG3324 Table 11

BG3324 Table 12

2. Continued Wealth Growth Post-Retirement. A second advantage of private retirement savings is that savings that are not withdrawn at retirement can continue to earn investment returns during retirement. Even with conservative investments, savings can continue to grow and provide for larger disbursements in subsequent years and bequests to help support family members, friends, or charities after death.

3. The Ability to Leave Bequests. Social Security prevents workers from passing on their “saved” payroll taxes to their heirs if they die before collecting benefits or shortly afterwards. Under the current system, a mean-income worker pays about $4,700 per year into Social Security’s retirement program.REF Over a 45-year career, that is up to $212,000 that he could potentially lose if he dies before collecting benefits. Personal savings, on the other hand, do not disappear if their owners die before using them. Shifting some or all of Social Security’s taxes to personal savings could have a particularly large and positive impact on lower-income earners as well as on many other groups that tend to have lower life expectancies and are more likely to get little to nothing back from their Social Security taxes. Bequests can serve as more than windfall benefits; they can change individuals’ and families’ lifetime trajectories by providing money that can help a child or grandchild attend college, start a business, or make other investments in their futures.

4. Larger Paychecks, Greater Incomes, and Increased Wealth. Social Security could accomplish the goal of preventing poverty in old age with significantly lower taxes than it currently extracts. A smaller Social Security program would leave workers with bigger paychecks that they could use for current consumption, gaining education, pursuing business opportunities, building wealth, and generating higher retirement incomes. More income and wealth would make low- and middle-income communities more dynamic and prosperous places to live and work.

5. Increased Productivity. Personal savings that support private investments allow companies to create productivity-enhancing capital in the form of new machines, technology, and facilities. This increases the output of workers, which leads to better jobs and higher wages.REF The current Social Security system—which immediately spends all incoming revenues on retirees’ benefits—fails to accomplish the same productivity- and opportunity-enhancing effects.


This report provides workers with a comparison between what Social Security can provide and what they could receive if they had ownership of their Social Security payroll taxes. This information can help individuals of all ages understand what they can expect to get from the program in the long run.

The results are overwhelmingly clear. Americans would be better off keeping their payroll tax contributions and saving them in private retirement accounts than having to sacrifice them to the government’s broken Social Security system. Social Security’s design has, over the decades, presumed that many Americans are too incompetent to make informed decisions for themselves, but few Americans believe that the government knows better than they do what is best for them and their families. Moreover, Social Security’s financial structure effectively guarantees that workers will receive extremely low, or even negative, returns on their payroll taxes.

—Kevin Dayaratna, PhD, is Senior Statistician and Research Programmer in the Center for Data Analysis, of the Institute for Economic Freedom, at The Heritage Foundation. Rachel Greszler is Research Fellow in Economics, Budget, and Entitlements in the Thomas A. Roe Institute for Economic Policy Studies of the Institute for Economic Freedom. Patrick Tyrrell is Research Coordinator in the Center for International Trade and Economics of the Kathryn and Shelby Cullom Davis Institute for National Security and Foreign Policy, at The Heritage Foundation.

Appendix: Basic Assumptions and Methodology

Contributions to Social Security and hypothetical personal retirement accounts are dependent upon workers’ earnings. Thus, we began by establishing average earnings profiles for American males and females over the course of their lifetimes based on 2015 data from the Current Population Survey (CPS).REF We enabled the age-based earnings profiles to vary by state by multiplying these profiles by the ratio of the average earnings in each state to the national average earnings and the national average wage index as reported in the 2017 Social Security Trustees report.REF We assumed workers have earnings between ages 21 and their full retirement age, which currently varies between ages 66 and 67.REF

For longevity, we utilized two data sets regarding life expectancy at age 65 to compute life expectancies by age and state. The Centers for Disease Control has compiled state-by-state life expectancies beyond the age of 65 in 2007, and the SSA has compiled overall life expectancies past the age of 65 on an annual basis. We estimated growth rates from the SSA data and applied those growth rates to the state-by-state life expectancies.REF Because we only report rates of return for individuals who live to at least age 65, our results overstate Social Security’s average rate of return because we leave out workers who die before reaching age 65 and who therefore receive little or nothing from Social Security.

Using our estimated earnings and life expectancies, we computed benefits according to Social Security’s benefit calculation formulas.REF We projected the earnings levels for Social Security’s bend points in future years based on the SSA’s Average Wage Index (AWI) projections. We calculated future benefits (and converted them to 2014 dollars) using adjusted index factors based on the growth rates from SSA’s AWI forecasts. The results for amounts paid in, and amounts accumulated in private accounts, Social Security benefit amounts, and the amounts paid out by annuities were converted from 2014 dollars to 2017 dollars using CPI-U changes that occurred through 2017.REF Thus, the retirement of someone who retires in the future (which is everyone we analyze in this model), is presented in 2017 dollars.

In one series of simulations, we used the Social Security Trustees’ projected insolvency date of 2035 and their projected shortfall in benefit payments, which begins at approximately 21 percent reductions in 2035, with the percentage reduction increasing over time according to the Social Security Trustees’ projections. We also assumed an OASI payroll tax rate of 10.6 percent from 2017 onwards. Based on these calculations, we computed Social Security’s internal rate of return.

Additionally, we also estimated the impact of maintaining Social Security’s current benefit formula through increased payroll taxes, beginning in 2017. According to the Social Security Trustees 2017 report, the current 10.6 percent OASI payroll tax would need to rise to 13.28 percent to maintain current benefits over the 75-year horizon.REF We model the effect of this increase alongside Social Security’s current benefit formula.

Finally, we simulated the impact of individuals using their OASI payroll tax to invest in private retirement accounts. We assume a conservative investment allocation consisting of 50 percent government bonds and 50 percent large cap stocks. We apply inflation-adjusted rates of return based on historical information contained in Ibbotson et al. (see footnote 20) and our estimated future real rates of return for future years (approximately 7 percent for stocks and 2.75 percent for bonds). Our estimated rates are based on the average real rates from 1954 to 2016.REF We also assumed a 0.7 percent administrative cost of stocks and bonds, which is a very conservative assumption also used in prior research.REF

The following tables contain additional results for other age cohorts beyond the 1995 cohort discussed earlier in the paper.

BG3324 Appendix Table 1

BG3324 Appendix Table 2

BG3324 Appendix Table 3

BG3324 Appendix Table 4

BG3324 Appendix Table 5

BG3324 Appendix Table 6

BG3324 Appendix Table 7

BG3324 Appendix Table 8

BG3324 Appendix Table 9

BG3324 Appendix Table 10

BG3324 Appendix Table 11

BG3324 Appendix Table 12


Kevin Dayaratna
Kevin Dayaratna

Chief Statistician, Data Scientist, Senior Research Fellow

Rachel Greszler
Rachel Greszler

Senior Research Fellow, Roe Institute

Patrick Tyrrell
Patrick Tyrrell

Former Research Coordinator