Retirement and Supplemental Income Programs for Low-Income Older Mexican-Origin Adults in the United States and Mexico

By 2050, one in five persons in the United States and Mexico will be at least 65 years of age (World Bank, 2019). Older adults in both nations face high poverty rates: among adults at least 66 years of age, nearly one in four in both nations have household incomes less than half the median levels in their nation (Organisation for Economic Co-operation and Development, 2018). Among U.S. older adults, poverty rates are higher for Hispanics—nearly two-thirds of whom are of Mexican origin (Noe-Bustamante et al., 2019)—and are nearly triple those for non-Hispanic Whites (Flores and Radford, 2017).

In both countries, middle-aged and older Mexican-origin populations with lower levels of education in low-paid occupations are often self-employed; work in the “gig” economy; work in the informal sector as babysitters, house cleaners, gardeners, taxi drivers, or contingency workers; and work in agriculture, hospitality, and construction (Aguila et al., 2011; Pisani, 2021). As such, they are more likely to face involuntary job separation before retirement, with scarce options to re-enter the labor market (Flippen and Tienda, 2000).

Within the United States, the informal sector represents 11%–40% of the labor force; in Mexico, it represents 58% (Aguila et al., 2011; Smith Nightingale and Wandner, 2011). In 2015, U.S. Hispanics were also more prevalent (37.8%) in the informal sector than were all workers (36.0%) or non-Hispanic Whites ((35.1%) (Pisani, 2021). Because informal sector wages tend to be lower than formal sector ones, poor workers in the informal sector find it harder to escape poverty (Smith Nightingale and Wandner, 2011) and may be more likely to need public assistance. Furthermore, because the informal sector in Mexico employs most of the labor force, most workers in Mexico lack access to social security benefits (Aguila et al., 2011). In 2016, only 31% of Mexican workers received social security benefits (Comisión Nacional del Sistema de Ahorro para el Retiro, 2016). Work in the informal sector typically leaves workers ill-prepared, financially, for retirement (Aguila et al., 2011; Smith Nightingale and Wandner, 2011).

Research in the United States identifies several reasons why, when close to retirement age, Hispanics have relatively high rates of withdrawal from the labor force. These include deficits in human capital and health, as well as discontinuous work histories and informal employment (Flippen and Tienda, 2000). Lower socioeconomic status, reflecting adverse effects of discrimination and residential segregation, limits the education and job opportunities of older Hispanics and contributes to poor living and working conditions throughout their life course (Angel and Angel, 2015).

The United States and Mexico have pursued divergent safety net income policies for low-income older adults. Supplemental Security Income (SSI) programs, also called noncontributory pensions or social pensions, have filled some of the gaps in retirement income for low-income older adults in both nations. In this paper, we analyze the characteristics of Mexican-origin population participants of SSI programs in both countries.

SSI Programs for Older Adults in the United States and Mexico

Supplemental Security Income in the United States originated with Title I of the Social Security Act, the Grants to States for Old-Age Assistance Act in 1935 (Grundmann, 1985). This old-age assistance program was the main source of public income support for older, blind, and, since 1950, disabled persons. States were allowed to set their own standards for SSI payments (Congressional Research Service, Library of Congress, 1984). In 1972, amendments to the Social Security Act established a federal minimum income guarantee, set at three-quarters of the poverty line. In 1974, the SSI program established uniform, nationwide eligibility requirements for the program (Social Security Administration, 2017a).

Today, the U.S. program is means-tested for citizens living in the United States who are blind, disabled, or at least 65 years of age. Recipients cannot have assets exceeding $2,000 for individuals or $3,000 for couples. SSI coverage among adults 65 and older decreased from more than 9% in 1973 to less than 5% in 2019, largely due to the means-testing rules on income and to assets not being adjusted for inflation over time. Among SSI recipients age 65 or older, 42% have income (including SSI benefits) below the poverty line (Center on Budget and Policy Priorities, 2020). United States benefits have increased an average of 2.4% for each year, reaching a level that could not exceed U.S. $771 for an individual and U.S. $1,157 for a married couple in 2019 (Social Security Administration, 2019a), though individual states may augment these benefits (Social Security Administration, 2017b). Nonetheless, because SSI payments are set below the poverty threshold, the program lifts few recipients out of poverty (Herd et al., 2008).

Mexico established a federal SSI program for older adults in 2007 (Aguila et al., 2016). The Mexican program has varied in its eligibility age; in 2015, for example, it was 65, while in 2019 it was 65 for indigenous communities and 68 for other Mexicans. Benefits have grown over time, from $55.90 in U.S. dollar purchasing power parity (equivalent to 33% of minimum monthly wages and an amount 30% below the poverty threshold) in 2007 to $142.60 (equivalent to 41% of minimum monthly wages and an amount 15.9% above the poverty threshold) in 2019 (Aguila et al., 2016; CONEVAL, 2021; Sangod, 2021). The Mexican program has expanded its geographic coverage in stages over time. From its first coverage of rural localities with fewer than 2,500 inhabitants, it grew to cover all localities while excluding persons with social security benefits by 2012, with the means test being removed in 2019 (Aguila et al., 2016; Sangod, 2021). In other words, and in contrast to the U.S. program, the Mexican supplemental income program has expanded to include virtually all older adults.

Characteristics of Mexican-Origin Recipients of Supplemental Income Programs in the United States and Mexico

Using nationally representative data from the U.S. Health and Retirement Study (HRS) and the Mexican Health and Aging Study (MHAS), we analyzed the characteristics of several groups of adults 50 or older who received supplemental income payments in the United States and Mexico. In this study, we use data from HRS Wave 12 collected in 2014 and MHAS Wave 4 collected in 2015. For the United States, we analyzed (a) non-Hispanic Whites, (b) Mexican Americans born in the United States, and (c) Mexican immigrants. We used HRS Wave 12 because it has a larger sample size of Mexican immigrants than any other HRS wave. For Mexico, we analyzed (a) persons who migrated to the United States but later returned (return migrants), and (b) persons who never migrated (nonmigrants). We used the MHAS wave for 2015 because it was collected within a year of our HRS data. By using the MHAS wave for 2015, when Mexican supplemental income programs were available to all persons age 65 or older who did not participate in other social security programs, we can also compare the older populations in both nations at a time when their supplemental income program policies were more alike.

The study was conducted according to the guidelines of the Declaration of Helsinki and was approved by the Institutional Review Board of The University of Southern California (UP-21-00199).

Table 1 shows that in the United States, a higher proportion of Mexican Americans received SSI benefits than did Mexican immigrants and non-Hispanic Whites. In Mexico, a higher proportion of return migrants received supplemental income than did nonmigrants. All these differences were statistically significant (p < .05). Among our sample of persons 50 and older, Mexican immigrants to the United States were, on average, younger than Mexican Americans and non-Hispanic Whites. Among older persons in Mexico, return migrants and nonmigrants had a similar average age. Females comprised similar percentages of the three U.S. groups, but were a significantly (p < .05) lower proportion of return migrants than of non-migrants in Mexico.

Table 1.

Descriptive Characteristics

Variables	United States			Mexico
	Non-Hispanic Whites, % or mean (SD)	Mexican American, % or mean (SD)	Mexican immigrant, % or mean (SD)	Return migrants, % or mean (SD)	Nonmigrants, % or mean (SD)
Receive SSI	3.29	14.52	11.72	15.66	10.75
Age	70.36 (10.79)	66.35 (9.72)	63.44 (9.10)	68.89 (10.69)	66.82 (9.77)
Female	57.82	56.29	55.74	26.86	60.43
Education years	13.34 (2.60)	11.03 (3.67)	6.74 (4.51)	5.33 (4.53)	5.64 (4.73)
0 education years	0.40	1.80	10.41	18.19	18.49
1 to 12 education years	46.67	64.87	79.41	73.83	71.89
>12 education years	52.93	33.33	10.18	7.98	9.62
Monthly individual income
Tercile 1 (low)	27.54	47.86	67.93	23.94	25.61
Tercile 2 (middle)	34.65	30.01	24.02	40.60	41.29
Tercile 3 (high)	37.81	22.13	8.05	35.46	33.11
Working status
Working	25.64	30.29	43.31	45.97	36.34
Unemployed	1.34	1.66	2.84	0.84	0.44
Retired	68.80	60.86	34.56	17.50	15.32
Not in the labor force	4.22	7.19	19.29	35.69	47.90
Occupation most of life
White collar	50.04	32.78	8.05	8.14	10.82
Services	22.06	25.03	33.49	43.51	45.36
Agriculture	1.03	2.90	6.15	0.54	0.65
Blue collar	17.26	25.86	34.20	38.91	23.51
Military	0.67	0.55	0.24	0.00	0.05
Never worked	8.94	12.86	17.87	8.90	19.61
Contributed SS in the U.S.	-	-	-	28.93	-
Contributed SS in Mexico	-	-	-	44.13	42.02
Receive SS benefits in the U.S.	67.06	55.33	41.54	5.45	-
Receive SS and SSI benefits in the U.S.	2.31	10.65	9.94	-	-
Receive SS benefits in Mexico	-	-	-	23.64	27.77
U.S. citizen	-	-	20.71	3.61	-
Number of observations	11,090	723	845	1303	12,852

Note. SD = standard deviation; SS = Social Security; SSI = Supplemental Security Income.

Average education was lower for Mexican immigrants than for Mexican Americans or non-Hispanic Whites; these differences were statistically significant. Within Mexico, return migrants and nonmigrants had similar levels of education. Non-Hispanic Whites in the United States were most likely to have at least 12 years of education. Non-Hispanic Whites were most prevalent in the top income tercile; Mexican Americans and Mexican immigrants were most prevalent in the bottom one. Among the two groups in Mexico, income distribution was similar.

Mexican immigrants, return migrants, and non-migrants were most likely to be working, typically in service or blue-collar jobs. Non-Hispanic Whites and Mexican Americans were more likely to be retired and to have been white-collar workers. Mexican Americans, Mexican immigrants to the United States, and nonmigrants in Mexico were more likely than others to have never worked for pay.

Mexican immigrants, return migrants, and nonmigrants were most likely to be working, typically in services or blue-collar jobs. Non-Hispanic Whites and Mexican Americans were more likely to be retired and to have been white-collar workers.

Most non-Hispanic Whites and Mexican Americans reported receiving U.S. social security benefits, as did nearly half of Mexican immigrants and a small proportion of return migrants. Most SSI beneficiaries also received U.S. social security benefits. Differences in receiving SSI and U.S. social security benefits for Mexican immigrants and return migrants may be attributable to U.S. citizenship: 20.7% of Mexican immigrants reported U.S. citizenship, but only 3.6% of return migrants did so. In contrast, more than 40% of nonmigrants and return migrants to Mexico reported having contributed to the Mexican social security system. About one in four nonmigrants and return migrants reported receiving Mexican social security benefits.

Table 2 shows the results of logit models in odds ratios for the estimated probability of receiving supplemental income programs in the United States or Mexico. Mexican Americans were 3.4 times and Mexican immigrants 1.8 times more likely to receive SSI benefits than were non-Hispanic Whites. Return migrants in Mexico were 1.2 times more likely to receive supplemental income benefits than nonmigrants, which may reflect the impact of the means-testing rule of the program in 2015. Return migrants had labor histories split between the two nations, making them less likely to be entitled to social security benefits in either (Aguila and Zissimopoulos, 2013).

Table 2.

Logit Models of the Estimated Probability to Receive Supplemental Security Income

Variables	United States			Mexico
	OR	SE	95% CI	OR	SE	95% CI
Mexican American	3.377**	0.466	2.577–4.426	—		—
Mexican immigrant	1.898**	0.328	1.353–2.662	—		—
Return migrants		—	—	1.298*	0.149	1.037–1.624
Age	0.935**	0.006	0.924–0.947	1.079**	0.003	1.073–1.086
Female	1.448**	0.147	1.186–1.768	1.049	0.065	0.930–1.183
0 education years	3.919**	1.150	2.206–6.964	3.311**	0.816	2.043–5.367
1 to 12 education years	2.073**	0.241	1.651–2.603	2.561**	0.611	1.604–4.088
Unemployed	1.747	1.087	0.516–5.912	0.471	0.294	0.138–1.601
Retired	13.845**	2.894	9.191–20.855	0.530**	0.063	0.419–0.670
Not in the labor force	10.247**	2.408	6.466–16.240	1.022	0.079	0.878–1.190
Services	2.429**	0.331	1.859–3.173	1.813**	0.314	1.291–2.547
Agriculture	5.276**	1.418	3.116–8.933	2.325*	0.922	1.068–5.059
Blue collar	2.817**	0.439	2.076–3.823	2.129**	0.380	1.501–3.020
Never worked	4.651**	0.702	3.461–6.252	1.909**	0.350	1.332–2.735
Contributed SS in the U.S.		—	—	1.274	0.232	0.891–1.820
Contributed SS in Mexico		—	—	0.423**	0.037	0.356–0.503
Number of observations	12,658			14,155

Notes. CI = confidence interval; OR = odds ratio; SE = standard error; SS = Social Security. Benchmark categories are working, 12 or more education years, white collar, non-Hispanic White for the U.S. sample and nonmigrants for the Mexican sample.

*p < .05; **p < .01.

Older adults in Mexico were also more likely to receive supplemental income as they aged, which is not surprising given the age qualification for the program. Females were more likely than males in the United States to receive SSI benefits, but there is no difference in supplemental income in Mexico for older males and females. Individuals with lower levels of education in both the United States and Mexico were more likely to receive supplemental income benefits, as were those who worked in service, agriculture, or blue-collar jobs and those who never worked. Supplemental income recipients in the United States were more likely to be retired and those in Mexico were more likely to be working.

Discussion

While Mexican immigrants to the United States were more likely to have lower incomes than Mexican Americans, they were less likely to receive SSI benefits. By contrast, in Mexico return migrants were more likely than non-migrants to receive supplemental income benefits.

These difference stem from strict means-testing rules for the U.S. program, as well as higher participation rates in the U.S. Social Security program. As a result, supplemental income is the largest source of support for older persons in Mexico, while Social Security benefits are the largest source of support for older persons in the United States. Even when the majority of SSI recipients (56.5%) receive Social Security benefits, these adults face poverty and deprivation (Social Security Administration, 2016).

Policymakers may wish to understand how supplemental income programs improve health and wellbeing of older adults relative to other programs. United States SSI recipients are typically eligible for Medicaid, a health insurance for lower-income and disabled persons that provides health-care services and long-term nursing care, while Medicare typically provides health-care services for citizens 65 or older. Hispanics are twice as likely as non-Hispanic Whites to receive both Medicaid and Medicare (Medicare Payment Advisory Commission, 2017). The United States also provides a Supplemental Nutrition Assistance Program (SNAP; commonly known as food stamps) and a Meals-on-Wheels program (U.S. Department of Agriculture, 2014), an Older American Act program for older adults with disabilities (Meals on Wheels America, n.d.). In Mexico, older adults are eligible for Seguro Popular, a public health insurance program (Parker et al., 2018). There are no federal long-term care or nutrition assistance programs for older adults in Mexico.

Research on whether supplemental income and health insurance programs complement or substitute for each other is scarce. Taubman and Sickles (1983) found that the U.S. SSI program improved the health of older adults within three years of introduction. Herd et al. (2008) found higher SSI benefits led to lower disability rates. Riumallo-Herl and Aguila (2019) found that supplemental income benefits for older adults in Mexico may increase the use of higher-quality health care. There is, however, no clear evidence of the effects of the Seguro Popular program on the health of older adults. Barros (2008), for example, found no impact on self-reported health and hypertension.

In Kenya, Haushofer et al. (2020) found minor effects of health insurance for informal sector workers on health-care utilization or health outcomes, but also a decline in stress, consistent with previous U.S. studies (e.g., Finkelstein et al., 2012). Health insurance seems to provide peace of mind to recipients, which reduces stress. Haushofer et al. (2020) also found minor short-term effects of a supplemental income program on health and health-care utilization. By contrast, research in Brazil, Mexico, and South Africa found large short-term effects of supplemental income programs on health and health-care utilization for older adults (Case, 2004; Galiani et al., 2016; Schwarzer and Querino, 2002). We are not aware of previous studies for the United States comparing the contributions of SSI and other programs to recipient health. Exploring this topic, including whether supplemental and health programs substitute or complement each other, would be a worthy focus of future research.

United States nutrition assistance programs, including SNAP and Meals-on-Wheels, have produced modest improvements on health and nutrition (e.g., Bartfeld et al., 2015; Roy and Payette, 2006). We are not aware, however, of research on whether supplemental income and nutrition assistance programs complement or substitute for each other, or which type of program is more effective. The balance of research on income and nutrition assistance programs suggests supplemental income programs may be an effective policy to improve the health of older adults, but more research is needed.

The differences we noticed in supplemental income program participation also reflect differences in educational levels. In both nations, the least educated and those who never worked or who held blue-collar, agriculture, or service jobs were more likely to receive supplemental income benefits. In Mexico, supplemental income recipients were more likely to continue working, while those in the United States were more likely to retire. Daly and Burkhauser (2003) note that U.S. SSI recipients receiving other safety net programs have fewer incentives to continue working because of the complex eligibility requirements, creating a poverty trap for recipients.

Within the United States, nativity matters for SSI program participation. While Mexican immigrants were poorer and less educated than Mexican Americans, they were more likely to work in occupations that may not provide access to Social Security benefits or to have legal status preventing them from claiming such benefits or SSI benefits, resulting in financial insecurity in old age (Angel, 2003).

Policy Recommendations

Our findings point to several possible initiatives for U.S. and Mexican policymakers. First, U.S. policymakers could expand safety net programs such as SSI for low-income older adults. Research from other countries increasingly shows that SSI programs reduce poverty and food insecurity, increase health-care utilization, and improve the health and well-being of older adults (e.g., Aguila et al., 2018). The easiest way to increase SSI coverage would be to reduce eligibility barriers and increase benefits. Of course, in the long run, and especially with a growing population of older adults, an extensive SSI program could strain government budgets. Policymakers therefore could combine a short-run expansion of SSI with long-term programs to help workers save for retirement, with SSI then later being reduced.

Second, Mexican policymakers could promote policies to encourage participation in the social security system. Mexico’s virtually universal supplemental income programs may become increasingly difficult to sustain as the older population grows. Increasing social security program participation can help alleviate pressure on supplemental income programs. Providing ease of access, particularly for migrants who go to and return from the United States, will help encourage overall participation in the program. Allowing these migrants to contribute to the Mexican social security system and save for retirement while they work in the United States would also help their own wellbeing at the end of their working years.

Third, U.S. and Mexican policymakers might make social security contributions portable. The United States and Mexico have bilateral social security agreements with several other countries, but not with each other. The United States, for example, has such agreements with 30 European countries, as well as with Canada, Chile, Brazil, Uruguay, Japan, and South Korea (Social Security Administration, 2019b). Such a totalization program would allow workers who may not have worked long enough to qualify for social security benefits in either country to do so by combining work credits from both countries (Aguila and Zissimopoulos, 2013). Instituting such an agreement would help Mexican immigrants and return migrants access social security benefits and reduce their dependence on supplemental income programs.

Funding

This work was supported by the National Institute on Aging (R03AG063183; P01AG022481).

Conflict of Interest

None declared.