June 28, 2005

HISPANICS AND SOCIAL SECURITY:
THE IMPLICATIONS OF REFORM PROPOSALS

By Fernando Torres-Gil, Robert Greenstein, and David Kamin[1]

Executive Summary

PDF of this report

Related Reports:

Hispanics’ Large Stake In The Social Security Debate
HTM | PDF

The Importance Of Social Security To The Hispanic Community
HTM | PDF

View Additional Social Security Reports

If you cannot access the files through the links, right-click on the underlined text, click "Save Link As," download to your directory, and open the document in Adobe Acrobat Reader.

KEY FINDINGS

Plans that rely largely on benefit cuts to restore solvency to Social Security are adverse for Hispanics, compared to plans that employ a balanced mix of benefit reductions and progressive revenue changes. 
Hispanics are likely to do worse under reform plans that make young workers and future generations bear the brunt of the sacrifices needed to preserve Social Security.
The President’s Social Security plan fails on both of these counts.  It relies entirely on backloaded benefit cuts that would hit young workers and future generations hard.
Replacing Social Security with private accounts would be especially harmful to Hispanics.  Hispanics benefit disproportionately from Social Security’s social insurance aspects and redistributive nature, which would be eliminated in a pure private account system.
Hispanics should be encouraged to save more for retirement by revamping and improving the current system of tax preferred saving accounts.

The Hispanic community has a great deal at stake in the debate over Social Security.  In a separate report, we have analyzed the particular importance of the Social Security system to the Hispanic community.[2]  Hispanics receive a higher rate of return on the taxes they pay into the system than the rest of the population, and elderly Hispanics rely on Social Security for a larger share of their income than other elderly Americans do.  For Hispanics, Social Security reform plans must be evaluated in light of the significant benefits they receive from the current system.

In this analysis, we outline the potential effects of Social Security reform on Hispanic Americans.  We consider how the Administration’s proposals to change Social Security benefits would affect the Hispanic community.  We also discuss alternative types of reforms.

In summary:

The Threat of Large Benefit Cuts and Imbalanced Reform

A number of plans under discussion would restore Social Security solvency primarily or exclusively through benefit reductions.  Over time, some of these plans would sharply reduce the portion of a worker’s pre-retirement income that Social Security replaces and would reduce Social Security survivor benefits and possibly disability benefits, as well.  This would have adverse effects on Hispanics.

Table 1
Proposed Sliding Scale Benefit Cuts

(For Worker Retiring At Age 65)

 

Dollar Reduction (In 2005 $)

Percentage Reduction

Earnings of $36,600*

 

 

      Worker retiring in 2045

   $-3,253

-16%

      Worker retiring in 2075

     -7,629

-28%

 

 

 

Earnings of $58,560*

 

 

      Worker retiring in 2045

     -6,444

-25%

      Worker retiring in 2075

   -15,154

-42%

 

 

 

Earnings of $90,000 or more*

 

 

      Worker retiring in 2045

     -9,324

-29%

      Worker retiring in 2075

   -21,808

-49%

* Note:  The earnings levels in the table are given in today’s terms.  Over time, workers’ earnings are assumed to grow at the same rate as average wages.

Source: Authors’ calculations based on Social Security Administration, Office of the Chief Actuary, “Estimated Financial Effects of a Comprehensive Social Security Reform Proposal Including Progressive Price Indexing -- INFORMATION,” February 10, 2005 and Social Security Trustees, 2004 Annual Report.  All percentage reductions in benefits are taken directly from the actuaries’ memo.

As noted, Hispanics receive a higher rate of return on the taxes they pay into the system than the rest of the population, and elderly Hispanics rely on Social Security for a larger share of their income than do other elderly Americans.  Thus, to the extent that a plan substantially reduces Social Security’s role in providing retirement security and redistributing resources, the Hispanic community would tend to be harmed.  Hispanics would be harmed disproportionately if large cuts are made in a system from which they disproportionately benefit.

The President’s Social Security proposals, as they have been outlined so far, rely solely on benefit cuts to restore solvency to the Social Security system.  The President has proposed “sliding scale benefit reductions” (also known as “progressive price indexing”) that would result in growing and, eventually, sharp benefit cuts for many Social Security beneficiaries.

Moreover, the Administration may eventually endorse deeper benefit cuts.  The benefit cuts that the Administration has proposed to date would close only about 59 percent of the 75-year Social Security shortfall, as measured by the Social Security Trustees. [5]  Further adjustments on top of the President’s proposed benefit reductions would be needed to close the shortfall.

When asking his Social Security Commission to outline options for reform in 2001, the President directed the Commission to avoid any measures that would raise additional revenues.  Since then, the President has softened his stance somewhat and has not ruled out legislation that would raise the maximum level of income that is subject to the Social Security payroll tax (currently set at $90,000).  But so far, the Administration has proposed only benefit cuts, and it seems clear that the Administration will support only measures to restore Social Security solvency that rely largely or entirely on benefit reductions.

Hispanics generally would fare better under plans that impose smaller benefit cuts on middle-income workers than the President has proposed, by making those with very high incomes shoulder a larger share of the load.  This could be done through progressive revenue changes, such as the following:

The President’s proposals do not include any progressive revenue adjustments that would enable the Social Security benefit reductions imposed on middle-income workers to be more modest.  The Administration’s proposals instead rely entirely on benefit cuts and place much of the burden of restoring Social Security solvency on middle-income people.  Hispanics would not be well served by such an approach. 

 

Future Generations Could Bear the Heaviest Load

The Hispanic population today is overwhelmingly young, but it is expected to age rapidly over the coming decades. 

Due to these demographics, Social Security changes that shield current retirees and spare baby-boomers much pain while making later retirees bear the heaviest loads would be especially detrimental to Hispanics.  To be sure, all generations — both old and young — benefit from Social Security, as it serves and is expected to continuing serving as a basic form of income security for the survivors of deceased workers, people with disabilities, and retirees.   But, the system faces a fiscal shortfall that needs to be closed.  For Hispanics, it is particularly important that the shortfall not be closed by measures that largely spare baby boomers and place the burden almost entirely on younger workers.  Hispanics would fare better under approaches that more equitably spread the burden of reform across generations.  The Administration’s Social Security plan fails to do this.

Under the Administration’s proposals, Social Security benefit reductions would be modest for those retiring in the next couple of decades, but would grow steadily deeper over time, with each new group of retirees facing sharper benefit cuts.  As a result, today’s younger workers, including many Hispanics, would face dramatically larger benefit reductions than earlier retirees. 

Specifically, the President’s plan exempts those aged 55 or older from any cuts in Social Security benefits.  For those retiring in or after 2012, the first year in which new retirees would begin to be subject to benefit reductions, the proposed cuts would start small but mount over time. As Figure 2 illustrates, an average wage-earner would face a benefit cut each year of:



 

The generational imbalance in the President’s plan is further exacerbated by the financing of the private accounts that the President has proposed.  To finance the “transition” to these accounts, the federal government would borrow trillions of dollars.  That would greatly magnify the debt burden on younger generations.

Because of the borrowing undertaken to finance the private accounts, the President’s Social Security proposals would add $4.9 trillion (in current dollars) to the national debt over the plan’s first 20 years, with several trillion dollars in additional debt added in the decades after that.  Under the President’s plan, this debt would eventually be paid off through additional reductions in the Social Security benefits of young workers.  In the words of economist Lawrence Kotlikoff, who himself favors private accounts, “the dirty little secret underlying most Social Security privatization schemes is that they head precisely down this road” of “dumping the entire…bill in our kids’ laps.”[9]  Doing that would not be in the interests of the Hispanic community.

 

The Danger of Private Accounts

As we have detailed in a separate analysis, the Social Security system disproportionately benefits people with:

Hispanics have all of these characteristics.[11]

Hispanics thus would be harmed by plans, such as many private account plans, that would shrink the social insurance aspects of Social Security by linking each person’s benefits to a much greater degree to that individual’s payroll tax payments, without regard to whether the person was paid low wages throughout his or her career, how long the person remains alive (and thus for how many years he or she needs income in retirement), or the misfortunes that may befall a person, such as disability or premature death.  In their most extreme form, private account plans would eliminate these redistributive aspects of Social Security, since the level of income that people could draw from the accounts would be determined solely by the size of a person’s contributions to his or her account and the rate of return on the account’s investments (rather than also being determined in part by circumstances such as whether the person had low earnings or whether his or her spouse or parent had died).  Scaling back Social Security benefits to pay for private accounts consequently could affect Hispanics adversely, since they benefit disproportionately from the social insurance and redistributive features of the current Social Security system and receive a higher rate of return than the rest of the population on the contributions they make to the system.

Despite this, some reports have claimed that Hispanics would be better off if Social Security were entirely replaced by private accounts.  For instance, a widely publicized 1998 Heritage Foundation report asserted that the “Social Security system’s rate of return for most Hispanic Americans will be vastly inferior to what they could expect from placing their payroll taxes in even the most conservative private investments.”[12]  In a separate analysis, we have explained the severe shortcomings with the Heritage report (which has been sharply criticized by the Social Security actuaries and widely discredited) and other reports making such claims. 

In summary, reports that claim the rate of return on private accounts would be significantly higher than the rate of return under Social Security suffer from two critical flaws.[13] 

The “bottom line” is that once transition costs and stock-market risk are taken into account, a system with private accounts would not produce a better rate of return for the population as a whole than the Social Security system.  And for Hispanics, the Social Security system is distinctly preferable to private accounts since Hispanics benefit disproportionately from Social Security’s social insurance aspects and redistributive nature.

 

Strengthening Retirement Security for Hispanics

Hispanics in general tend to be ill-prepared for retirement in terms how much they have accumulated in pensions and other retirement savings.  The solution to this problem is not to contract or dismantle the Social Security system, the one form of retirement security that works especially well for Hispanics.  Rather, the solution is to revamp and improve the system of tax-preferred saving accounts, such as IRAs and 401(k)s.  In their current form, those accounts provide the most powerful incentives and most generous tax benefits to people who have the highest incomes (and the least need for incentives and subsidies to help them save adequately for retirement).  To better secure Hispanics’ retirement, the current system of tax incentives for retirement saving should be reformed to provide greater incentives and opportunities for middle- and low-income Americans to save.

The current system of tax-preferred saving accounts is particularly ill-suited to helping Hispanic workers provide for their retirement.   To encourage participation in retirement saving plans, the current system relies principally on tax deductions and exclusions which shield from taxation the money that people save in various types of retirement saving vehicles.  The value of such deductions and exclusions depends on an individual’s tax bracket.  The higher the tax bracket, the greater the value of the deduction or exclusion; the lower the tax bracket, the smaller the value of the deduction or exclusion and thus the smaller the size of the tax subsidy.  In addition, the current system of saving incentives can be quite complicated.  For a population like Hispanics with lower-than-average income and education levels, this incentive system tends to be rather ineffective.

Brookings Institution economist Peter Orszag has proposed a number of policy changes that would improve retirement security for low- and middle-income families by simplifying the current system and strengthening the incentives.  Such reforms would be especially beneficial for Hispanics.  These reforms include:[17]

National Council of La Raza’s Proposals to Expand Coverage and Enhance Benefits for Hispanics

 As this analysis explains, the Social Security system provides better returns to the Hispanic community than to the population as a whole and to non-Hispanic whites or blacks.  Nonetheless, the system could be improved and strengthened.

In a new report, the National Council of La Raza suggests ways in which the Social Security system could be adjusted to widen its coverage and strengthen its support for elderly Hispanics.[21]  NLCR notes that “most Latinos who are eligible for Social Security benefits receive an ample amount of income support over their retirement and benefit greatly from the system’s progressivity and indexed benefits.”  But NCLR also finds that while those Hispanics who are eligible for Social Security benefits are helped more than other beneficiaries, on average, a smaller share of elderly Hispanics receive Social Security checks than of other retirees.

According to the Social Security Administration:

Social Security coverage is lower among elderly Hispanics for several reasons.  Although undocumented workers frequently pay taxes into the Social Security system (under false or non-work status Social Security numbers), many of these workers, who are disproportionately Hispanic, will never collect Social Security benefits based on these contributions.  The presence of undocumented workers lowers Hispanic participation rates. 

The Social Security system, through its ten-year work requirement, also lowers coverage rates for Hispanic workers who are here legally.  To qualify for Social Security retirement benefits, a worker must contribute to the Social Security system for at least ten years (40 quarters).  This creates a cliff in the system that can adversely affect workers with short work histories.  Despite the fact that they have contributed payroll taxes, those who work for less than ten years in covered employment receive no retirement benefits (unless there is a “totalization” agreement in effect with an immigrant’s native country; see footnote 22), while someone who works exactly ten years can receive substantial benefits.

New immigrants, of which there are many in the Hispanic community, tend to have relatively short work histories in the United States and thus would tend to be disproportionately harmed by the way the threshold is structured. [22]   The Social Security earnings requirements also make it more difficult for domestic workers (many of whom are Hispanic) than for others to get work counted toward the ten-year threshold.  Finally, a number of Hispanics work in sectors, such as agricultural labor, where some employers underreport earnings to the Social Security.  This makes it more difficult for some Hispanics, such as itinerant farm workers, to accrue the ten years necessary to qualify for benefits. 

To raise coverage rates among Hispanic workers, NCLR suggests a number of adjustments to the current system.  The proposed changes include:

NCLR also has added its voice to that of many independent groups and policymakers, including the President and a number of members of Congress, in expressing support for strengthening Social Security’s minimum benefit. The current system includes a minimum benefit, but the benefit is small and rapidly phasing out.  A meaningful minimum benefit in Social Security would help lift low-income workers out of poverty in retirement.

Strengthening the Social Security Administration’s Other Program:
Supplemental Security Income for the Elderly and Disabled Poor

 The Social Security Administration also administers the Supplemental Security Income program, which provides a basic safety net for poor seniors and people with disabilities who either receive small Social Security benefits or do not receive Social Security at all.  The SSI program lifts poor individuals who are elderly or disabled to about 75 percent of the poverty line and poor couples to about 90 percent of the poverty line.  In most states, receipt of SSI also qualifies an individual or couple for health insurance coverage through Medicaid.

As the recent report from the National Council of La Raza notes, SSI is of particular importance to Hispanics.  Some 13 percent of Hispanics aged 65 and over receive SSI.  This percentage is much higher than those for elderly whites, three percent of whom receive SSI benefits, and also exceeds the percentage for elderly blacks, 10 percent of whom receive SSI.*  The NCLR report explains that “SSI is particularly important as a safety net for Latinos who work intermittently for very low wages, who didn’t always work in Social Security-covered jobs, or who immigrated to the U.S. late in life.”**  Such people tend to qualify for small Social Security benefits that leave them well below the poverty line or not to qualify for Social Security at all.

SSI Improvements Needed to Prevent Well-intended
Social Security Reforms from Harming Vulnerable People

There appears to be broad consensus across the political spectrum in support of reforms that would assist poor Social Security beneficiaries by strengthening Social Security’s “minimum benefit” (discussed elsewhere in this report) and improving what is sometimes referred to as Social Security’s “widow’s benefit” — i.e., the benefit for surviving spouses who are age 50 or older if disabled and age 60 or older otherwise.    (Under the consensus proposal to strengthen the widow’s benefit, the Social Security benefits of a low- or moderate-income widow or widower would not fall below 75 percent of the combined Social Security benefit that the couple received when both spouses were alive.)  Both of these reforms would be desirable.  Yet they would have a severe side-effect on some poor elderly and disabled people who receive both Social Security and SSI — and would make these people significantly worse off — unless the reforms are accompanied by a key reform in SSI.

For people who receive both Social Security and SSI, an increase in Social Security benefits triggers a dollar-for-dollar reduction in SSI benefits.  If an increase in Social Security benefits lifts such people modestly above the SSI income limit, they lose eligibility for SSI — and consequently could lose Medicaid coverage in a majority of states.  The resulting loss of health care coverage and the increase in out-of-pocket health care costs these people would have to bear would far outweigh the small increase they would get in their monthly checks from the Social Security Administration.  With a larger share of elderly Hispanics receiving income from a combination of Social Security and SSI than of the rest of the elderly population, this is a matter of particular significance for the Hispanic community.

This matter could be addressed by including in Social Security legislation a provision requiring that people who would be eligible for SSI in the absence of improvements in Social Security’s minimum benefit and widow’s benefit be “deemed” to be receiving SSI for purposes of determining their eligibility for Medicaid.  This would enable the affected people still to qualify for Medicaid.  On several occasions in the past when Congress made changes in Social Security that would cause some people to become ineligible for SSI, Congress has included such a provision to prevent the loss of Medicaid coverage.

Other SSI Improvements

The SSI asset limits, set at $2,000 for individuals and $3,000 for couples, have not been adjusted since 1989.  They have been heavily eroded by inflation.  These limits have effectively been reduced by 36 percent since 1989, because of the lack of any inflation adjustment.

These limits are eroding further and becoming more restrictive with each passing year.  An upward correction in them, and a provision to adjust these limits for inflation in the future, would be of significant help to Hispanics. 

In a similar vein, when the SSI program was created in 1972, Congress directed that the first $20 in Social Security benefits, veterans’ benefits, or other unearned income be disregarded in determining SSI eligibility and benefit levels.  This level has not been adjusted for inflation in the 33 years since.  Consideration should be given to increasing it, as well, and adjusting it for inflation in the future. 

Finally, the harsh limitations enacted in the mid-1990s that severely restrict SSI eligibility for legal immigrants should be eased.  Legal immigrants who lawfully entered the United States on or after August 22, 1996 are categorically ineligible for SSI, unless they (or a spouse or parent) have amassed 40 quarters of work in this country.  Even a legal immigrant who has been unable to amass 40 quarters of work because he or she has become severely disabled after entering the United States, as a result of a workplace or other accident or the onset of a serious disease, is barred from SSI despite his or her obvious need.  The SSI rules regarding the eligibility of legal immigrants are more restrictive than the rules in food stamps, Medicaid, or the Temporary Assistance for Needy Families program and should be made less harsh.


* Social Security Administration, Income of the Population 55 or Older, 2002, March 2005, Table 1.3.

** National Council of La Raza, “The Social Security Program and Reform: A Latino Perspective,” 2005, p. 12.

It should be kept in mind, however, that not all minimum benefit proposals are alike.  Some proposed minimum benefits, including the minimum benefit endorsed by the President’s commission on Social Security, would phase out over time.  Such proposals would establish a minimum benefit floor, but the floor would rise more slowly than the Social Security benefits it is meant to augment.  Over time, the minimum benefit would become largely meaningless.  Other proposals would allow the minimum benefit to grow along with Social Security benefits as traditionally calculated.  This would produce a minimum benefit that would not phase out over time and be of far greater value in retirement to low-wage workers who are young today. 

These NCLR proposals deserve consideration as part of Social Security reform.  It should be noted that resolving these issues does not require making fundamental changes to Social Security.  They can be addressed through relatively modest adjustments within the current system.  Private accounts would do little to solve these problems and would end up harming Hispanics by undermining the features of Social Security that benefit the Hispanic population substantially.

 

Conclusion

The current Social Security system provides the Hispanic community with valuable benefits.  Hispanics receive better returns on the taxes they contribute to the Social Security system than other workers, and elderly Hispanics rely on Social Security benefits for a greater share of their income than the rest of the population.

Closing the Social Security shortfall is particularly important to the Hispanic community.  As a young population that is rapidly aging and that receives more in return for its contributions to Social Security than others, Hispanics have much invested in the future solvency of the system. 

However, as this analysis indicates, the Hispanic community should be wary of changes that seriously threaten the benefits that Hispanics receive from the current system, under the guise of restoring solvency.  Plans such as the President’s that rely solely on benefit reductions to close Social Security’s shortfall, and that would place the greatest burden on younger workers, would be more detrimental to the Hispanic community than alternative reforms.  Hispanics also would be harmed by private-account plans that shrink the social insurance aspects of Social Security by linking each person’s benefits more directly to that person’s tax contributions without regard to whether the worker has earned low wages or encountered some major misfortune.

Hispanics’ retirement security needs to be strengthened.  But substantially scaling back Social Security would have the opposite effect, as it is the one form of retirement security that now works well for the Hispanic community.   Hispanics should instead be given incentives to increase the amounts they save for retirement and thereby to bolster their retirement security.  This can be done by reforming the current system of retirement tax incentives to provide greater incentives and opportunities for low- and middle-income Americans workers and their families to build assets for retirement.


End Notes:

[1] Fernando Torres-Gil is Director of the UCLA Center for Policy Research on Aging and Acting Dean of the UCLA School of Public Affairs.  Robert Greenstein is Executive Director, and David Kamin is a Research Assistant, at the Center on Budget and Policy Priorities.

[2] See Fernando Torres-Gil, Robert Greenstein, and David Kamin, “The Importance of Social Security to the Hispanic Community,” Center on Budget and Policy Priorities, June 28, 2005. 

[3] National Council of La Raza, “The Social Security Program and Reform: A Latino Perspective,” June 2005, available at http://www.nclr.org/content/publications/download/32018.

[4] Although the President’s Social Security proposals have been widely reported as protecting benefits for all of those the bottom 30 percent of income earners, this is clearly not the case.  A substantial number of low-income beneficiaries would be subject to benefit reductions.  Those affected include elderly widows and divorced spouses, as well as low-income children of deceased workers.  Such beneficiaries would be subject to Social Security benefit cuts, regardless of their income level, if they receive a Social Security spousal or survivors benefit that is based on the earnings of another person (usually a deceased parent or a spouse or ex-spouse) who was not in the bottom 30 percent of wage earners.  Many spouses and children whose families were not poor while the family’s breadwinner was alive fall into poverty or low-income status after the death of the breadwinner; such people would face benefit reductions under the President’s plan.  For more details, see Jason Furman, “New White House Document Shows That Many Low-Income Beneficiaries Would Face Benefit Cuts,” Center on Budget and Policy Priorities, May 10, 2005, available at https://www.cbpp.org/5-10-05socsec2.htm.

[5] Including the cost of the private accounts that the President has proposed, the President’s proposals close an even smaller share of the 75-year gap.  Taken together, the President’s proposals would close only about 30 percent of the 75-year shortfall.

[6] This assumes an estate tax along the lines of the tax that is slated to be in effect in 2009, when the exemption from the tax will be $3.5 million per individual ($7 million per couple) and the top estate tax rate will be 45 percent.

[7] Census Bureau, Current Population Survey, March 2004, data available at http://ferret.bls.census.gov/ macro/032004/perinc/new10_001.htm.

[8] Census, U.S. Interim Projections by Age, Race, and Hispanic Origin, March 2004, available at http://www.census.gov/ipc/www/usinterimproj/.

[9] Lawrence Kotlikoff, “What’s Wrong?: Don’t Make the Kids Pay for Our Mess,” The Washington Post, January 23, 2005, Outlook p. B02.

[10] According to official government projections, Hispanics have substantially longer life expectancy than the rest of the population.  Based on Census Bureau data, the Social Security Administration reports that Hispanic men turning age 65 in 2004 can expect to live an additional 20 years, compared to 16 years for all men, and Hispanic women aged 65 in 2004 can expect to live an average of 23 years, compared to 20 years for all women.  The Government Accountability Office has come to similar conclusions, also using Census Bureau data, as have the Social Security actuaries.  Several researchers have raised questions as to whether Hispanics’ longer-than-average life expectancies are real or are a product of measurement error.  Even if Hispanics did not have longer-than-average life expectancies, they would receive well-above-average rates of return on their contributions to the Social Security system because they exhibit all of the other characteristics described here.

[11] The Hispanic population also has certain characteristics that would tend to lower Hispanics’ rate of return on their contributions to Social Security.  In particular, a slightly smaller share of the elderly Hispanic population aged 65 or older is or has been married than is true of the rest of the elderly population.  Thus, the Hispanic population may not benefit quite as much as the rest of the population from spousal benefits.  According to official government measures, Hispanics of working age also have lower mortality rates than others of the same age, which would tend to reduce the benefits Hispanics receive from Social Security survivors benefits.  The effects of these two characteristics, however, are substantially outweighed by the impact of the characteristics described in this analysis that significantly increase the rate of return that Hispanics receive on their Social Security contributions.  On net, Hispanics receive a substantially higher rate of return on Social Security than the rest of the population. 

[12] William W. Beach and Gareth G. Davis, “Social Security’s Rate of Return for Hispanic Americans,” Heritage Foundation, March 27, 1998, available at http://www.heritage.org/Research/SocialSecurity/CDA98-02.cfm.

[13] For detailed analysis of the problems with comparing the rate of return on private accounts to the rate of return on Social Security, see Jason Furman, “Would Private Accounts Produce a Higher Rate of Return Than Social Security?” Center on Budget and Policy Priorities, June 2, 2005, available at https://www.cbpp.org/6-2-05socsec.pdf.

[14] Craig Copeland, “Employment-Based Retirement Plan Participation: Geographic Differences and Trends,” Employee Benefit Research Institute Issue Brief, October 2004, available at http://www.ebri.org/ibpdfs/1004ib.pdf.

[15] Social Security Administration, “Income of the Population Aged 55 or Older, 2002,” March 2005, Tables 1.1 and 1.4, available at http://www.ssa.gov/policy/docs/statcomps/income_pop55/.

[16] Ibid., Tables 6.B1 and 6.B4.

[17] For a more detailed description, see Peter R. Orszag, “Improving Retirement Security,” Testimony Before the Committee on Ways and Means, May 19, 2005, available at http://www.retirementsecurityproject.org/pubs/File /Ways%20and%20Means%20Testimony%2020050517.pdf?PHPSESSID=d48df6046644e8926f38b1cbf74c337a.

[18] Brigitte C. Madrian and Dennis F. Shea, “The Power of Suggestion: Inertia in 401(k) Participation and Savings Behavior,” Quarterly Journal of Economics, vol. 116, no. 4, November 2001, Table 4.

[19] James J. Choi, David Laibson, Brigitte C. Madrian, and Andrew Metrick, “Defined Contribution Pensions: Plan Rules, Participant Decisions, and the Path of Least Resistance,” in Tax Policy and the Economy, vol. 16, James Poterba, ed. (MIT Press, 2002), pp. 67-113.

[20] Orszag, “Improving Retirement Security,” p. 8.

[21] NCLR, “The Social Security Program and Reform: A Latino Perspective.”

[22] This is not the case for citizens from the 20 countries with which the United States has “totalization” agreements in effect.  Under these agreements, the United States coordinates Social Security benefits with the public pension systems that exist in those countries.   Work in both countries can be counted toward the ten-year eligibility requirement for Social Security benefits, although to be eligible for “totalization” a worker must be employed in the United States for, at least, a year and a half.   Initial Social Security benefits are then pro-rated, reflecting the number years of employment in the United States.  Currently, Chile is the only Latin American country with which the United States has a totalization agreement that is in effect.  The United States has signed such an agreement with Mexico, but the President has yet to submit it for Congressional review.