Number 148
March 2009
Office of the Chief Actuary
Baltimore, Maryland
Projections of Immigration for the 2008 Social Security
Trustees Report

by Alice H. Wade, J. Patrick Skirvin, Alvin K. Winters, and Stephen C. Goss


As required by law, the Old-Age and Survivors Insurance and Disability Insurance (OASDI) Board of Trustees submits an annual report to Congress on the long-term actuarial status of the Social Security program. Both the income and cost of these programs are directly related to the size and composition of the United States (U.S.) population and the broader Social Security area population.1 For this reason, we devote considerable effort to develop the best possible projection of the population distributed by age, gender, and other characteristics that influence employment and benefit eligibility. This note describes the projections of immigration developed for the intermediate assumptions of the 2008 Trustees Report.

Developing reasonable and appropriate immigration projections is important because immigration will continue to play an important role in determining the overall growth rate and distributional characteristics of the U.S. population. For example, since 1990 the number of births in the U.S. was about 4 million persons per year, while net immigration2 into the U.S. exceeded 1 million persons per year, a level that is expected to continue for the foreseeable future. Furthermore, the immigrant population has tended to be younger and less likely to work in covered employment and therefore less likely to be eligible to receive benefits. Thus, developing reasonable net immigration assumptions requires more than simply setting annual numbers of people who will enter and exit the Social Security area. It also requires setting their age/gender distribution, length of stay, employment, and likelihood of benefit receipt.

For the projections, the immigrant population is divided into two broad groups. The first is the "legal-immigrant population", which includes people born outside the U.S. who are legal permanent residents (LPRs) or naturalized citizens. Individuals with this status enjoy most of the rights of native-born citizens and are assumed to participate in employment and receive benefits in essentially the same manner as the native-born population. The second group is the "other-immigrant population", which includes people born outside the U.S., excluding those who are LPRs or naturalized U.S. citizens. Some of these individuals reside in the U.S. on a legal temporary basis. However, most of the other immigrants residing in the U.S. are undocumented.

For the intermediate assumptions of the 2008 Trustees Report, the following diagram illustrates how the annual flow of immigrants and emigrants affect the size of the Social Security area population.


Legal-Immigrant Population

Legal immigrants are admitted into the U.S. for permanent residence based on certain categories and preferences as defined by the Immigration Act of 1990. These categories include employment-based preferences, family-based preferences, immediate relatives of U.S. citizens, refugees, asylees, diversity categories, and specifically legislated categories. The law places numerical limits for each of these categories, except for the category of immediate relatives of U.S. citizens. For the projections, we assume that current law will be retained and the current numerical limits will not be changed. Even with this assumption, there is still variation in the actual number gaining LPR status each year. For example, the number of LPRs who are immediate relatives of U.S. citizens has increased significantly in recent years. Also, the number of persons attaining LPR status in a year can be affected by the ability of the Department of Homeland Security to process applications. Numbers processed in a year can vary if backlogs of applications are building up, or if special efforts are made to reduce backlogs.

The number of U.S. citizens and LPRs leaving the Social Security area population (legal emigrants) reduces the overall size of the Social Security area population. Furthermore, children born to LPRs in the country are themselves citizens and, thus, increase the size of the overall population. However, since the average employment and benefit entitlement characteristics of the legal-immigrant population are projected to essentially match those of the native-born population, no special identification of the category is maintained in our projections. Thus, they are treated the same as the native-born population for employment and benefit entitlement purposes.

Other-Immigrant Population

Unlike the legal-immigrant population, the other-immigrant population differs from the native-born population substantially in employment activity and benefit eligibility. Other immigrants who are in the country on an authorized temporary visa are subject to strict legal restrictions on their activity and time limits on their residence in the U.S. Undocumented immigrants have no legal authorization to work or to collect benefits in the U.S. Thus, we identify the other-immigrant population as a distinct group with explicitly different employment and benefit entitlement assumptions. While the other-immigrant population is assumed to be equally as likely to be employed as the legal permanent residents of the same age and gender, they are less likely to be paying taxes to the Social Security program, and even less likely to ever receive benefits based on their earnings.

The number of other immigrants in the population is difficult to estimate because they cannot be counted directly. Most estimates are based on residual calculations that subtract (1) an estimate of the population (based on using only the counted births, deaths, and legal immigrants and an estimated number of legal emigrants) from (2) the actual enumerated population. This difference is attributed to the other-immigrant population.

In the late 1980s, the size of the other-immigrant population was estimated to be in excess of 5 million, or about 2 percent of the total population, and rising. Between 1988 and 1992, about half of this population was granted LPR status under the Immigration Reform and Control Act (IRCA), reducing the other-immigrant population to roughly 1 percent of the total population. Since then, however, the other-immigrant population is estimated to have grown substantially, reaching over 13 million today (over 4 percent of the total population), and is projected to grow further to 24 million, or about 6 percent of the total population by 2050. The graph below displays the size of the other-immigrant population. But the total population that is classified as the other-immigrant population is only a part of what is needed to understand the effects on Social Security financing. The age and employment patterns reflecting timing of entry into the country are critical.


For projections, we make specific assumptions about other-immigrant flows and earnings. Historically, the number of persons entering the country each year in this status has grown and is currently estimated to be about 1.5 million. We assume that in the future the number of persons entering in that status will continue to be about 1.5 million per year. However, since other immigrants are a relatively transient group, we also assume that about one-third of them will gain LPR status within a few years, and that the majority of the remaining 1 million other immigrants will eventually leave the country.3 The annual net other-immigrant flow (the number entering the country minus the number becoming LPR and minus the number leaving the country) is estimated to be currently about 400,000 and is projected to decline to 300,000 by 2040 and somewhat lower thereafter (275,000 for 2061 and later).

As the size of the other-immigrant population in the U.S. increases, the number who can, and will, depart from this country to return to their country of birth is assumed to increase. Thus, with a sustained 1.5 million other immigrants assumed to be entering each year, and an increasing total number currently residing in the U.S., the annual net flow of other immigrants will decline. Part of the reason the net other-immigrant flow has been so large since 1990 is that, after IRCA, the total number of other immigrants in the population was greatly reduced and so relatively few were leaving the country. Unless the number of other immigrants entering the country rises substantially above the assumed 1.5 million per year, and continues rising indefinitely, this decline in the net other-immigrant flow should be expected.

Other immigrants enter the country each year generally in their twenties. Only a relatively small number are assumed to remain for life in this status because of the substantial emigration from the country and transfers to LPR status. Thus, we project that a relatively large proportion of the other-immigrant population will be in their thirties and forties, compared to the total population. The two graphs below compare the proportions of the Social Security area population that are in other-immigrant status, by age and sex, as of January 1, 2000, and as of January 1, 2050. The relatively young age distribution of the other-immigrant population, along with the children born to them in the United States, tends to add more workers than beneficiaries in all future years, and thus significantly improves the financial outlook for the program over the 75-year projection period.



Employment of the Other-Immigrant Population

As stated above, we assume that those in the other-immigrant population are as likely to be employed as legal permanent residents of the same age and gender. However, their temporary-legal or undocumented status makes them less likely to be properly registered with the Social Security Administration (SSA) and far more likely to be working in the underground economy or working with an unregistered name and Social Security Number (SSN). Thus, we divide those in the other-immigrant population who are working into the following four classes.

Class 1. Those covered by OASDI and included in the registered SSA records with an SSN associated with their name. These workers have earnings that are subject to payroll taxes and credited to their own earnings history in SSA records. Historically, most of the workers in this class are assumed to have obtained SSNs in their name using fraudulent identification, particularly birth certificates. After 2001, however, SSA became far more vigilant on identification and these cases are assumed to be relatively rare. This class also includes those with temporary work visas, which allow them to establish a legitimate SSN associated with their name. Additional workers in this class include those who have overstayed a temporary visa, (work or student) and subsequently work using their legitimately acquired SSN.

Class 2. Those not covered by OASDI and included in the registered SSA records with an SSN associated with their name. These workers have earnings that are not subject to payroll taxes or credited toward benefits. This class includes mainly students working for state universities in non-covered employment.

Class 3. Those providing an unregistered SSN to their employer or a registered SSN not associated with their name. These workers (and their employers) pay payroll taxes, but their earnings are not credited to them for the purpose of eventually establishing benefit eligibility. Instead, their earnings are placed in a "suspense" file. In some rare cases, earnings placed in the suspense file for undocumented workers are retrieved if legal work authorization is later obtained and adequate records of the prior work activity are produced.

Class 4. Those working in the underground economy. These workers pay no taxes and their earnings are not reported to the Social Security program.

For projections, we make specific assumptions about the proportions of the other-immigrant workers in each of the classes identified above.

The table below shows the percentages of the employed other-immigrant population that are assumed to be in each of these four classes. The percentages assumed to be in the suspense file or in the underground economy are clearly expected to be much greater for cohorts entering the country after 2001.

Employment of Other Immigrant Population
OASDI Covered Class 1
Noncovered Student Class 2
Suspense File
Class 3
Underground Economy Class 4
Year 2000
Arrive after 2001

Benefit Entitlement for the Other-Immigrant Population

Individuals who enter the country as other immigrants and remain in that status for life are relatively unlikely to receive benefits from the OASDI program. Those who work in the underground economy have no basis for expecting to be entitled for benefits. Those who have worked and paid payroll taxes without a matched SSN will have had their earnings placed in the suspense file and will have only a relatively remote possibility of retrieving these earnings for the purpose of becoming entitled to a benefit. The relatively small and declining number of other immigrants with earnings credited in their name and SSN may be able to receive benefits in the future if all of the following conditions are met:

(1) They worked long enough to become insured under the program (including insured based on a totalized agreement).

(2) They were legally work-authorized at some time.

(3) They received LPR status before benefit entitlement or, if not, are willing to leave the U.S. to receive a benefit.

We estimate that around 30 percent of the relatively small number of other immigrants age 62 in 2000 is fully insured and could potentially become eligible to receive retired-worker benefits in the future. However, due to the restrictions outlined above, relatively few of these individuals will actually file for and receive benefits. We project that the percent eligible to receive a retired-worker benefit will decline to around 10 percent at the end of the 75-year projection period, and the number actually receiving a benefit will be very small. Children born to other immigrants in this country, however, are native-born citizens and are assumed to be as likely as others to work in covered employment and ultimately receive OASDI benefits.


Immigration assumptions and methods play a highly important role in projecting the future population and financial status of the OASDI program. For the intermediate assumptions of the 2008 Trustees Report, the projected ultimate level of net legal immigration is 750,000 persons per year. Including these individuals in the projections is relatively simple because this group has employment characteristics very similar to native-born citizens.

Since many of the workers in the other-immigrant population are not legally authorized to work, it is far more difficult to include them in the projections. We project that the ultimate number of other immigrants entering the country each year is 1.5 million, but ultimately less than 20 percent, or about 275,000, remain in this status for life. Thus, the age distribution of this group is and will continue to be relatively young, which has a positive effect on the financial status of the OASDI program. Furthermore, the proportion of the other-immigrant population that remains in this status until retirement age and becomes entitled to receive benefits is small and projected to decline substantially in the future. However, these positive effects are reduced by the assumption that relatively few pay taxes based on their employment, and that this proportion will further decline.

Both legal and other immigration tend to have positive effects on the financial outlook for Social Security. And perhaps most important for both groups is that their children born in the U.S. are recognized as U.S. citizens and thus contribute a further positive effect on the financial status of the Social Security program.

1 Besides U.S. residents, the Social Security area population includes armed forces and U.S. citizens overseas, residents of U.S. territories, and non-citizens living abroad who are insured for Social Security benefits.

2 Net immigration is defined as total immigration less total emigration.

3 The annual number of other immigrants who leave the Social Security area is estimated based on a model of departures over the period 2000-2006, disaggregated into two groups. The first departing group is calculated by applying a set of annual departure rates, by age and sex, to the other-immigrant population in the Social Security area. The second departing group is set at fixed annual numbers of departures, by age and sex, which remain constant throughout the projection period.