Skip to main content
Open Access Publications from the University of California

UC Berkeley

UC Berkeley Electronic Theses and Dissertations bannerUC Berkeley

Return Migration Among Latin American Elderly in the U.S.: A Study of its Magnitude, Characteristics and Consequences


The 1965 Immigration Act released a stream of immigration from Asia

and Latin America that continues to shape the U.S. population

composition. While some of these migrants promptly returned to

their countries of origin, many spent many years in the U.S. and

face retirement with truncated work histories, legal impediments to

old-age support programs, and social networks scattered in two

countries. This dissertation examines one issue in the aging

process for Latin American immigrants, namely the location of their

retirement. I examine the extent to which older immigrants return

to their home countries during later life and whether retirement

income plays a role in this decision.

A daunting challenge in studying this topic is data limitations.

The migration literature notes numerous inconsistencies across data

sources due to their different strengths and limitations. To

address this issue, I do an in-depth examination of the magnitude

and characteristics of return migration among older Mexican immigrants using multiple data sources to assess the consistency of the outcomes. In

chapter 2, I discuss the rate of return migration among Mexican

immigrants aged 50 years and their characteristics compared to their

U.S.-residing counterparts using the Integrated Public-Use Microdata

Series (IPUMS) for Mexico, the National Survey of Demographic

Dynamics (ENADID), and the Mexican Health and Aging Survey (MHAS).

I find that the five-year incidence of return migration from the U.S. to Mexico ranges from two percent when generated using IPUMS Mexico to six percent when using the MHAS. However, while the rate of return migration among

this population is inconsistent across data sources, certain

characteristics are not. All data sources suggest that return

migrants are predominantly male and have intermediary levels of

education. Characteristics that are inconsistent across data

sources are marital, employment, and citizenship status.

Aside from the magnitude and characteristics of return migration, I

also examine one possible reason for return migration during later

life, namely higher levels of retirement income. Mexicans with

greater retirement benefits may view this income stream as a means

toward greater luxury in the home country. Conversely, these migrants

may return migrate only upon concluding that they cannot make ends

meet in the U.S. Each scenario has vastly different implications for

the U.S. economy. I examine this question in two chapters in order

to take advantage of two forms of data: survey and administrative

data. Pooling IPUMS U.S.A. and IPUMS Mexico, I conduct logistic

regressions to determine if higher levels of retirement income are

associated with an increased probability of return migration. I also

do a sensitivity analysis to assess possible biases associated with

pooling two data sources. Results from this chapter suggest that

Mexican immigrants with lower levels of retirement income are more

likely to return to their home country during later life than those

with higher levels of retirement income. This pattern holds assuming

various rates of Hispanic undercount. However, in the absence of a

natural experiment, one cannot attach a causal interpretation to the results of this chapter.

The experimental nature of chapter 4 does enable a causal

interpretation. In chapter 4, I use a natural experiment whereby the

Social Security Administration substantially lowered the Social

Security benefits of the 1917-1921 birth cohorts due to a mistake in

their benefit calculation formula. These birth cohorts have since

been referred to as the ``notch" generation as graphs depicting

average benefit amounts by birth cohort show a visible notch for this

group. In chapter 4, I use this natural source of exogeneity to

observe whether the ``notch" generation was more likely to return

migrate than those who did not receive these lower benefits. Results

of this chapter indicate that Social Security benefits do not affect

the probability of return migration for Latin American primary Social

Security beneficiaries.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View