Social Security and Immigration: Part 3

By Edwin S. Rubenstein
Volume 21, Number 2 (Winter 2010-2011)
Issue theme: "The Economic Case for a Moratorium on Immigration"

Social Security is in trouble. The Great Recession has reduced payroll tax revenues while increasing the number of workers applying for benefits. As a result, last year (2010) the retirement trust fund had a cash outflow of $41 billion. If the economy recovers quickly, small surpluses could be achieved for 2012 through 2014, but by 2015 the deficit returns - permanently. That’s when baby boomers start retiring en masse.

According to the 2010 Trustees Report, the retirement “trust fund” will be exhausted in 2037. At that point, the system will have burned through all the assets accumulated since the inception of Social Security more than 70 years ago. It will be unable to pay all of the benefits that have been promised.

The long-term (75-year) actuarial deficit is estimated to be a whopping $7.9 trillion in present value terms. In other words, Congress would have to raise $7.9 trillion today and invest it in order to have enough money to pay all of Social Security’s promised benefits between 2015 and 2084. This money would be in addition to what Social Security receives during those years from payroll taxes.1

But not to worry! Many claim immigration can reduce Social Security’s long-term deficit. As we discuss below, this conclusion is based on flawed assumptions regarding the amount of payroll taxes immigrants pay into the system as well as the benefits they are likely to receive. Moreover the reduction in native wages — and payroll tax payments — stemming from competition with low-wage immigrants is invariably ignored by those who claim immigration can help shore up the retirement system.

Social Security today operates on what is known as a “pay-as-you-go” basis, in which current worker payroll taxes are used immediately to pay for the benefits of current retirees and other beneficiaries. In 1950, there were about 16 workers for every retiree. Today, there are slightly over three workers per beneficiary, and by the time today’s 20-year-olds retire, that number will fall to two workers for every beneficiary. Furthermore, Social Security is paying greater benefits for longer periods of time as life expectancy increases and baby boom retirees strain the system.2

The Social Security “problem” is largely the result of an aging population where fewer and fewer young workers are taking the place of older retirees. In the short run, immigration can help mitigate the funding crisis, but its impact is slight.

Social Security’s own analysis shows just how weak the immigration effect is.

A 350,000 increase in annual immigration (from 1,065,000 to 1,370,000) would reduce the 75-year actuarial deficit by about 10 percent (from -2.00 to -1.81 percent of taxable payroll), according to the 2009 Social Security Trustees Report.3

In other words, if we were to admit 23 million more immigrants than current immigration policy calls for over the next 75 years, the Social Security deficit would fall by less than one-tenth.

Think of it: 23 million additional immigrants to cut the retirement fund deficit by less than 10 percent.

Even this is optimistic. Sure, new immigrants are generally younger than natives. They pay payroll taxes long before they collect benefits. But that happy result is not as common as Social Security claims. Legal immigrants often enter the country years after they first apply. By then they are well past their prime earning years and closer to retirement than is commonly believed.

Older immigrants do well by Social Security. More than one-third of them who arrive after age fifty-five reportedly receive Social Security income.4 The vast majority of them receive benefits in excess of their contributions. More generally, typical immigrant workers arriving at any age receive a greater proportion of their Social Security contributions back in benefits than native-born workers. The subsidy is greatest for immigrants who earn at least $10,000 annually and have worked in the U.S. for 10 to 20 years. They receive 70 to 80 percent of the benefit received by native-born workers with comparable wages who contributed to the system over their entire working lifetime.

Special preference for immigrants? Not at all. It merely reflects the fact that immigrants make less than natives, and that Social Security benefits are “progressive,” i.e., they replace a larger share of the wages of low-income workers than high-income workers.

Do older immigrants come because of Social Security? It’s possible, even probable. George Borjas believes that many are well aware of the generosity of Social Security and other welfare-type programs, and that this is a factor in the immigration decision. The high welfare usage among those who enter the country after age fifty may reflect their inability to qualify for Social Security. These older immigrants might have entered expecting to receive Social Security, but with the knowledge that if they didn’t qualify, other safety net programs would take care of them.

And these are legal immigrants!

Many foreign workers are not immigrants, but are stationed in the U.S. on a temporary basis. Many of them, including the highly skilled (and often well paid) employees here on temporary H-1b work visas, pay no Social Security taxes whatsoever.5 Since the 1970s the Social Security Administration (SSA) has concluded so-called “Totalization Agreements” with about 24 different countries under which foreign workers may have their Social Security deductions sent to their home country program (rather than the SSA), and vice-versa. One of the primary objectives of the agreements is to eliminate dual Social Security taxation, as would occur when a worker sent to the U.S. by his employer must pay retirement fund tax to both the United States and his home country.

The agreements also protect people who have worked in both the U.S. and another country, but have not worked long enough in one country or the other to qualify for Social Security benefits. A U.S. worker employed temporarily in Canada, for example, shouldn’t lose those quarters of employment when filing for Social Security. Their contributions don’t end up in the host country’s coffers, but the employer is still obliged to pay them.

The U.S. collects the totalization taxes for each foreign worker who has asked for totalization benefits, and then a massive transfer of funds is made between the two governments. Therefore, while workers may have payroll taxes deducted from their paychecks, when they return home, those funds go with them. Whether they pay more or less in pension contributions depends on the policies of their home country and the specifics of the Totalization Agreements. It is a sure bet that most of the time the totalization tax is less than what would have been paid to Social Security, so that the U.S. employer saves money also.

The U.S. has entered into totalization agreements with the following countries:


India and China — the home countries of more than half of all H-1b’s — are conspicuously absent from the list, and for a very good reason: They have no Social Security systems. Exempting an Indian or Chinese from paying Social Security taxes would mean they would pay no Social Security taxes at all. That would be an insuperable competitive advantage for those H-1B’s and the companies that hire them. American workers with equal skills would be passed over in favor of tax-exempt foreigners.

Of course, if an Indian or Chinese national immigrates to the UK or Canada and gets their H-1b, they may be eligible for an exemption. Other techniques under which H-1Bs avoid paying payroll taxes are described in The Great American Tax Dodge, by Donald L. Barlett and James B. Steele:

Visit most any large American company and you’ll find two types of people working on the same computer project. One is a permanent company employee who pays taxes through withholding. The other is a temporary foreign employee who enjoys the kind of payday that an American worker can only dream about — a full paycheck with zero deductions.6

How is it done? Consulting and contracting firms recruit so-called temporary workers in other countries, especially India, and bring them to the United States under the temporary visa program. They farm out these foreign workers to large U.S. companies that do not want to add permanent employees to their payrolls or to replace higher-paid American workers who have been let go. Other big clients are state governments that contract out computer work.

Because they are employed by the consulting firm that recruited them, these temporary workers are paid either in cash or by check — and no money is withheld for U.S. income tax, Social Security, Medicare, or state or local taxes.

A potentially more important factor — and one which Social Security actuaries seem to ignore — is the impact of new immigrants on native wages. Native workers account for about 85 percent of today’s labor force, and because they earn more, an even larger share of payroll tax payments. Any dampening of native-income growth will worsen the financial condition of Social Security. But this is exactly what immigration does.

The relevant rule of thumb: native wages decline by 3.5 percent for every 10 percent increase in the foreign-born share of the workforce. (See wage section for details.)

Under current immigration policy, the foreign-born share of the labor force is projected to increase from 6 percent in 2010 to 23 percent in 2050. That 7-percentage-point increase will lower native wages by 2.5 percent according to the Borjas rule of thumb. If we should admit an additional 23 million, as Social Security’s high immigration scenario suggests, the immigrant share of the workforce could reasonably increase by 9 percentage points; native incomes would fall by 3.2 percent.

Conclusion: admitting an additional 350,000 immigrants per year could reduce native wages by an additional 0.7 percent. That would more than offset the (alleged) reduction in Social Security deficit from higher immigration.

An aging population:
Current policy vs. moratorium

It is sometimes said that immigration can “keep America young” and thereby “save Social Security.” Immigrants tend to be young adults; many have children after they arrive. And immigrants generally have larger families — higher fertility rates — than natives. But, surprise, surprise: immigrants age at the same rate as natives. So do their children. Over the long run, immigration has at best a modest effect on the median age of U.S. residents.

The U.S. median age, 35 in 2000, is projected by the Census Bureau to rise to about 39 by 2030, even with immigration of close to 1 million people annually. A moratorium would push median age up to 41 in 2030 and to 42 in 2050:


Median age for all races and Hispanic origin groups will increase from 2010 to 2050 under both moratorium and current policy. Asians are influenced most. In the absence of immigration to the U.S., an increase from 38 in 2010 to 51 in 2050 is projected, making them the oldest group in the United States. Under current policy, the median age of the Asian population increases from 36 in 2010 to 43 in 2050.

For non-Hispanic whites, immigration has a minimal impact on the pace of aging. For them, the median age is expected to rise to approximately 45 by 2050 in both scenarios.

The dependency ratio

The relationship between the size of the working-age population, on the one hand, and the young and elderly, on the other hand, is called the nation’s “dependency ratio.” Other things equal, a higher ratio portends a higher tax and economic burden on workers whose incomes are needed to care for Social Security recipients.

Pew Research Center’s “Main Projection” — roughly equivalent to our Current Policy projection — shows the dependency ratio rising from 58.6 dependents per 100 working age individuals in 2010 to 71.8 in 2050. There is no moratorium scenario in Pew’s population study, but under their low-immigration scenario, the ratio would rise even higher, to 74.9:


The rising ratio is used to stoke fears of Social Security insolvency and to promote higher immigration as a potential remedy. Despite all the hype, the reality is that the dependency ratio at mid-century will be lower than it was in 1970. As seen in the above graphic, this is true for both the high- (current policy) and low-immigration projections.

The rise in dependency ratio from 2010 to 2050 is not the issue. What is unprecedented is the source of the increase. The dependency ratio comprises two parts: those who are younger than working age (under 18) and those who are older (over 64). The youth dependency ratio is the ratio of people under age 18 to the number of people of working age (18 to 64), and the elderly dependency ratio is the ratio of those over working age to those in their working years.

Since 1970, the aged dependency ratio has risen while the youth dependency ratio has fallen off. Over the next 40 years, these trends will continue, and even accelerate for the elderly. But this is not an immigration policy problem:

The Center’s projections also indicate that between 2005 and 2050, the number of elderly will increase more rapidly than either the number of children or working-age adults. Immigration and births to immigrants will be responsible for growth in all other age groups but will have little impact on the growth of the elderly, which is affected mainly by the aging of the post-World War II baby-boom generation.7

While immigration is the most important determinant of U.S. population growth, it has relatively little impact on the age distribution. Even the most extreme immigration regime — a moratorium — would barely move the needle:

According to the UN Population Division’s Projection Branch, without any immigration at all between 2000 and 2050, the aged dependency ratio [elderly population/working age population] in the U.S. in 2050 would be 0.389. But having high immigration like we have now will barely make any difference in lowering the dependency ratio; continuing under present immigration policy, the U.S. will still have an aged dependency ratio of .355 in 2050 — a difference of .034, which is only 8.7 percent.

This 8.7 percent difference hardly justifies high immigration as a ‘magic cure’ for the country’s dependency ratio.8

A long-term rise in elderly dependency is locked into the system. It cannot be materially altered by immigration. This is explained by the fact that immigrants get old and “age out” of the working-age population. Like everyone else, they eventually become part of the problem.

The only way you can get a permanent immigration benefit is if you have not just high immigration, but exponentially increasing numbers of immigrants. In other words, the only way immigration can affect the age structure — and possibly “save” Social Security — is to enact an immigration Ponzi scheme, i.e., bring in more and more workers every year to compensate for the increasingly larger cohorts of foreign-born elderly.

That would send population through the roof, further erode the environment, and increase dependency among working-age natives displaced by massive numbers of immigrants. The cure would be worse than the disease.

Take a few steps back and it becomes clear there are much simpler and less traumatic ways of resolving the Social Security problem than importing 1.3 million immigrants a year into our society. What most alarmists fail to mention is that the long-term (75-year) funding shortfall is equal to a mere 2 percent of taxable payroll. A payroll tax increase of that magnitude would cover it. Benefit reductions would obviously reduce the required tax hike. A means test for Social Security eligibility has been suggested, though politically this option may be a non-starter at this time.

It also should be noted that the definition of “working age” is not fixed in stone. Many Americans work well into their sixties and seventies not because they have to, but because they want to. The reality is that people are not just living longer, they are healthier longer. Thanks to medical and lifestyle advances, many people are working — and paying payroll taxes — while they receive Social Security benefits.

Raising the age of eligibility from 65 to 67 — as was done recently — is simply a reflection of this trend. That change reportedly dwarfs any impact that higher immigration levels would have on long-term dependency. In fact, by raising the age of eligibility two years, we lowered elderly dependency by more than a complete cessation of immigration would have raised it.

Implication: We can impose a moratorium on all immigration and still cover the Social Security deficit by making modest adjustments in things like taxes and eligibility ages.

Illegal immigrants to the rescue?

Aliens who are not self-employed have Social Security and Medicare taxes automatically withheld from their paychecks. Since undocumented workers have only fake numbers, they’ll never be able to collect the benefits these taxes are meant to pay for. Last year, the revenues from these fake numbers — that the Social Security Administration stashes in the ‘earnings suspense file’ — added up to 10 percent of the Social Security surplus.
New York Times blog.

Background: Starting in the late 1980s, the Social Security Administration received a flood of W-2 earnings reports with incorrect Social Security numbers. It stashed them in what it calls the “earnings suspense file” in the hope that someday it would figure out to whom they belonged.

The file has been mushrooming ever since: $189 billion worth of wages ended up recorded in the suspense file over the 1990s, two and a half times the amount of the 1980s. By 2008, the file had risen to nearly $600 billion.

While Social Security does not know what fraction of the suspense file corresponds to the earnings of illegal immigrants, other researchers say illegal immigrants are the main contributors.9

Reality check: Totalization agreements

The assumption that most illegal immigrants will not collect Social Security — and that the suspense file money will be available to fund Social Security benefits for natives and legal immigrants — is unrealistic.

A law called the Social Security Protection Act of 2004 explicitly prohibits benefits to “ aliens residing in the United States unlawfully.” But a loophole in that law exempts illegals from any country “ …that has a social insurance or pension system under which benefits are paid to eligible U.S. citizens who reside outside that country....

That is exactly what Totalization Agreements do (see above). They are designed to protect workers who have divided their careers between the U.S. and a foreign country, but haven’t worked long enough under either social security system to qualify for benefits. The agreements allow workers to combine (“ totalize”) work credits earned in both countries to meet minimum eligibility requirements.

With the signing of the U.S.-Mexico Totalization Agreement on June 29, 2004, most of the illegal aliens living in the U.S. became potential Social Security recipients.

We say “ potential” because the U.S.-Mexico agreement has yet to be signed by the President — or even sent to Congress for review. Eligibility and costs will ultimately depend on specific terms and language of the final agreement.

Indeed, some observers fear Mexican totalization could metastasize into a de facto guest worker program, effectively legalizing millions of erstwhile illegal aliens.10 That devil could be in the details of the final agreement.

In any event, the Social Security Administration (SSA)’s preliminary cost estimates for Mexican totalization are absurdly low. In 2003, SSA’s actuaries projected those costs at $78 million in the first year, growing to $650 million (in constant 2002 dollars) by 2050. SSA claimed that the agreement would have a “ negligible impact” on the Social Security trust fund long-range actuarial deficit. (As noted above, the trust fund is expected to be exhausted in 2037 with or without Mexican totalization.)

However, SSA’s projections assume only 50,000 newly eligible Mexican beneficiaries would be added during the initial phases of totalization, with that number growing to 300,000 over time. Amazingly, these are the same numbers that SSA used to cost out the totalization agreement with Canada. Illegal aliens from Mexico make up about 70 percent of all illegals in the U.S. Illegals from Canada and other totalization countries combined are estimated to account for less than 3 percent of all illegal aliens in the U.S.11

Illegal alien headcounts don’t tell the whole story. Mexico’s retirement system is rudimentary compared to those of other totalization countries. Americans, for example, vest for Social Security benefits after working for 10 years; Mexicans must work for 24 years before vesting in their national pension plan. (Mexican aliens can vest for Social Security after working just 18 months in the U.S., and make up the difference by “claiming” to have worked in Mexico.)

Moreover, under the Mexican system, workers receive back exactly what they paid in, plus interest. (If it’s not stolen, that is. The men who paid into the Mexican Government’s Bracero Program in the 1940s haven’t been paid; the money just disappeared.)12 By contrast, Social Security is also an income-redistribution system, with low-wage workers receiving benefits far in excess of their contributions.

Another federal agency, the Government Accountability Office (GAO), warns that the prospect of easy Social Security eligibility could draw far more illegal aliens to the U.S. than SSA actuaries have projected:

…Although the actuarial estimate indicates that the agreement would not generate a measurable impact on the trust funds, an increase of more than 25 percent in the estimate of initial, new beneficiaries would generate a measurable impact. For prior agreements, error rates associated with estimating the expected number of new beneficiaries have frequently exceeded 25 percent. Because of the significant number of unauthorized Mexican workers in the United States, the estimated cost of the proposed totalization agreement is even more uncertain than for the prior agreements.”13

Bottom line: many illegal immigrants are likely to become Social Security pensioners after retirement. The much touted illegal immigration Social Security subsidy is just a temporary effect of new workers arriving who have no retired counterparts.

Today’s low-wage illegals will be tomorrow’s drain on the Social Security System. ?


1.  2010 social security trustees report continues to show the urgency of reform.




3. Sensitivity to Varying Net-Immigration (As a percentage of taxable payroll)


4. Robert Cherry, Immigration and Race: What We Think We Know, The Review of Black Political Economy, Fall 2003.


5. H-1b is a non-immigrant temporary work visa that allows employers to bring in highly skilled individuals to fill jobs where comparable domestic workers are (allegedly) in short supply. Critics claim that employers use the visa program to replace native workers with cheaper foreign labor.


6. Donald L. Barlett and James B. Steele, The Great American Tax Dodge, Google Books. 


7. Jeffrey S. Passel and D’Vera Cohn, U.S. Population Projections: 2005-2050, PewResearchCenter, February 2008. 


8. Federation for American Immigration Reform (FAIR), A Ponzi Problem: The U.S. Dependency Ratio, Social Security Solvency, and the False Panacea of Immigration, September 2000. 


9. Eduardo Porter, Illegal Immigrants Are Adding to Social Security With Billions, New York Times, April 5, 2005.


10. Phyllis Schlafly, Totalization: Sellout of American Workers, November 17, 2004.


11. Marti Dinerstein, Social Security ‘Totalization’: Examining a Lopsided Agreement with Mexico, CIS Backgrounder, September 2004.


12. The Bracero program allowed 1 to 2 million Mexican laborers to work temporarily in U.S. agriculture. Enacted in response to labor shortages during World War II, it was jointly operated by the U.S. and Mexican governments and lasted from 1942 to 1964. See Memo From Mexico.


13. Barbara D. Bovbjerg, Proposed Totalization Agreement with Mexico Presents Unique Challenges, General Accountability Office, September 2003. PDF



About the author

Edwin S. Rubenstein , president of ESR Research, economic consultants, has 25 years of experience as a business researcher, financial analyst, and economics journalist. Mr. Rubenstein joined Hudson Institute, a public policy think tank headquartered in Indianapolis, as Director of Research in November 1997. While at Hudson he wrote proposals and conducted research on a wide array of topics.

As a journalist, Mr. Rubenstein was a contributing editor at Forbes Magazine and economics editor at National Review , where his “Right Data”column was featured for more than a decade. His TV appearances include Firing Line, Bill Moyers, McNeil-Lehrer, CNBC , and Debates-Debates. In The Right Data (National Review Press, 1994), Rubenstein debunks many widely held beliefs surrounding the distribution of income, government spending, and the nature of economic growth.

From 1980 to 1986 he was senior economist at W.R. Grace & Co., where he directed studies of government waste and inefficiency for the Grace Commission. From 1978 to 1980 he was a municipal bond analyst for Moody’s Investors Service. Mr. Rubenstein has a B.A. in economics from Johns Hopkins and an M.A. in public finance from Columbia University .