Factor Structures of Reasons for Immigration Among Older Asian and Latino Immigrants in the United States.

BACKGROUNDS AND
OBJECTIVES: Although reasons for immigration are significant predictors of immigrants' health, factor structures of reasons for immigration are still unclear among older immigrants. The present study examined the factor structure of reasons for immigration among older Asian and Latino immigrants in the United States. RESEARCH DESIGN AND METHODS: Drawn from the National Latino and Asian American Study, 396 Latino and 298 Asian immigrants over 55 years of age were selected for analysis. Exploratory factor analysis was conducted for nine items concerning reasons for immigration in each immigrant group.
RESULTS: Three factors were extracted from both Asian and Latino immigrant elders: (a) "voluntary reasons" to pursue development, (b) "involuntary reasons" due to uncontrollable situations, and (c) "semivoluntary reasons" regarding family/medical duties. While immigration to join family members was located in the "semivoluntary reasons" factor among older Asian immigrants, it was located in the "voluntary reasons" factor among older Latino immigrants. DISCUSSION AND IMPLICATIONS: These findings suggest that three underlying factors of reasons for immigration should be understood considering the different characteristics of two racial/ethnic groups of immigrants. In addition, a migratory reason to join the family should be considered differently for elderly Asian and Latino immigrants. This three-factor framework of reasons for immigration can help clinicians provide more culturally sensitive interventions for older minority immigrants.

Clinicians working with older Asian and Latino immigrants should be aware of three factors of reasons for immigration (“voluntary,” “semivoluntary,” and “involuntary” reasons) and use these factors to guide them to provide more culturally sensitive interventions for older immigrant.

Background and Objectives

According to the 2010 U.S. Census, 4.6 million people living in the United States were estimated to be foreign-born older adults, which constituted 12% of adults in the United States aged 65 years or older (Scommegna, 2016). Among these foreign-born older adults, older Latino and Asian immigrants are two of the largest groups, representing 38% and 29%, respectively (Wilmoth, 2012). The proportion of Latinos and Asians in the older population living in the United States is expected to increase more than 200% from 9 % in 2003 to 20% in 2050 (Federal Interagency Forum on Aging-related Statistics, 2006). Thus, understanding the characteristics of these two fast-growing elderly immigrant groups has emerged as one of the most urgent agendas to prepare for the rapidly aging U.S society.

Asian and Latino immigrants differ from each other in many ways, including language, culture, and immigration history. Since the 1965 Immigration and Nationality ACT, Asian immigrants have actively migrated to the United States from more than 20 countries, and Chinese (24%), Indian (20%), and Filipino (19%) immigrants comprise the three largest Asian immigrant ethnic groups in the United States (López, Ruiz, & Patten, 2017). Members of these ethnic groups often migrated to the United States in pursuit of opportunities for education and employment, and refugees from Southeast Asia (e.g., Vietnam) migrated for a safer environment (Chen, Gee, Spencer, Danziger, & Takeuchi, 2009). Despite the significant subethnic heterogeneity in religious and linguistic backgrounds, Asian immigrants, especially the Chinese, Filipino, and Vietnamese that we examine in this study—share characteristics such as ethical values based on Confucianism (i.e., filial piety) and a collectivistic culture emphasizing family and community (Toyokawa & Toyokawa, 2013). Asian immigrants as a group tend to demonstrate higher educational attainment, higher income, and better English proficiency, but slightly lower rate of employment, when compared to other immigrant populations such as Latino immigrants (Zong & Batalova, 2016).

Meanwhile, Latinos have a long immigration history in the United States. The Mexican-American War in the 19th century and the Cuban Refugee Program in 1966 led to a significant influx of Latino immigrants (Gutiérrez, n.d.; Browning, Portes, & Bach, 1985). Currently, the top three countries from which Latino immigrants migrate are Mexico (62.3%), Puerto Rico (9.5%), and Cuba (3.9%) (U.S. Census Bureau, 2017). It is notable that Latino immigrants share many common characteristics such as Roman Catholicism as a predominant religion, Spanish as a primary language, and a strong emphasis on a family and group-oriented values (Clutter & Zubieta, 2009). In sum, Asian and Latino immigrants are two distinctive groups in the United States characterized by each group’s unique cultural and migratory background.

Despite the well-known “healthy immigrant effect,” referring to the phenomenon where various health outcomes of immigrants tend to be better than those of their U.S.-born counterparts, health status of immigrants often tends to deteriorate over time and even converge to the level of that of their U.S.-born counterparts as their length of residence in the United States increases (Alegría, Álvarez, & DiMarzio, 2017; Breslau et al., 2007). Moreover, older minority immigrants are more likely to face a “double jeopardy” (Dowd & Bengtson, 1978), which posits that dual hardships—their status as “minority immigrants” and experience of “degenerative changes related to aging.”—could interplay to increase the likelihood of negative incidents, which, in turn, could negatively influence their health (Dowd & Bengtson, 1978). Regarding older Latino immigrants, they often demonstrate advantages in physical health (e.g., lower rates of hypertention and obesity) and lower mortality rates compared to non-Hispanic Whites in the United States (Palloni & Arias, 2004; Riosmena, Wong, & Palloni, 2013). However, compared to their U.S.-born counterparts, they report higher lifetime rates of dysthymia and generalized anxiety disorder (GAD), as well as higher 12-month rates of GAD (Jimenez, Alegría, Chen, Chan, & Laderman, 2010). In terms of cognitive health, Hill, Angel, Balistreri, & Herrera (2012) reported that older male Mexicans who migrated during the age of 20–49 years tend to have better cognitive baseline scores, as well as slower cognitive decline over a decade, than their U.S.-born counterparts, suggesting that “healthy immigrant effects” shown among older Latino immigrants may be due to certain demographic (i.e., gender, education) and acculturation variables (Hill et al., 2012).

In the case of older Asian immigrants, on the other hand, initial advantages in health and mortality compared to U.S.-born counterparts tend to diminish with duration of stay in the United States (Frisbie, Cho, & Hummer, 2001). With regards to mental health, older Asian immigrants demonstrate a higher lifetime prevalence of GAD compared to U.S.-born Asians (Jimenez et al., 2010). Previous studies suggested several factors relating to limited health care usage among older Asian immigrant populations. For example, a longer time since immigration was related to less frequent use of health care services by older Korean immigrants within the United States (Sohn & Harada, 2004). Also, in a study of older Chinese immigrants in Canada, predictors of limited health care access included female gender, single marital status, more recent immigration, less financial stability, and a strong adherence to cultural beliefs from one’s country of origin were (Lai & Chau, 2007).

Then, what makes people migrate from their countries of origin to other countries despite hardships such as “double jeopardy” as a minority (Dowd & Bengtson, 1978) and anticipated acculturative stresses due to differences in language and/or culture (Berry, Kim, Minde, & Mok, 1987)? To answer this question, researchers have proposed a few frameworks to explain the underlying reasons/motivations of immigration for decades. In general, immigration has been known to occur either to follow a single motivation or to pursue a combination of several different motivations encompassing economic, environmental, social, or political conditions (Tartakovsky & Schwartz, 2001). First, traditional migration theory has proposed two conceptual factors for migration: “push” and “pull” factors (Lee, 1966). Push factors refer to reasons which make an individual leave his/her own native country, such as political persecution, war, lack of safety, poverty, etc. In contrast, pull factors are elements which could drag people to other countries, such as economic prosperity or educational opportunity (Lee, 1966). Second, Tartakovsky and Schwartz (2001) suggested three theoretically distinct types of motivations for migration based on their confirmatory factor analysis on Jewish immigrants from Russia, including: (a) the preservation of physical, social, and psychological security; (b) the self-development of abilities, knowledge, and skills; and (c) the materialistic accumulation of financial well-being. Third, researchers often consider whether immigration was voluntary (i.e., whether they wanted to) or involuntary (i.e., whether they had to) (Guarnaccia & Lopez, 1998; Ogbu & Simons, 1998). Voluntary immigration happens when immigrants have a hope of a better future (i.e., better occupations, greater political or religious freedom) in the new host country than in their countries of origin, without the decision being forced upon them (Ogbu & Simons, 1998). In contrast, involuntary immigration happens to escape danger and seek safety for themselves and their families, without choosing or planning to settle in the new host country, such as in the case of political refugees (Guarnaccia & Lopez, 1998).

Reasons for immigration are known to be related to immigrants’ health (Giuntella, Kone, Ruiz, & Vargas-Silva, 2018) and mental health (Fabrega, Moore, & Strawn, 1969; Gong, Xu, Fujishiro, & Takeuchi, 2011; Kuo & Tsai, 1986; Ogbu & Simons, 1998; Tartakovsky & Schwartz, 2001). For instance, findings have consistently suggested that immigrants with more voluntary reasons for leaving their country of origin tend to have better mental health than their involuntary counterparts (i.e., refugee) (Berry et al., 1987; Pernice & Brook 1994). Regarding the three-factor frameworks of Tartakovsky and Schwartz (2001), preservation motivation for immigration (e.g., migration to pursue a safer environment and socially supported identity) was related to poorer subjective well-being, whereas self-development motivation (e.g., migration to advance individual strivings for growth and independence) was correlated with positive general mental health among immigrants. Previous research not only demonstrated a clear link between reasons for immigration and the mental health of immigrants, but also suggests that reasons for immigration can potentially act as a protective or risk factors for their mental health. Thus, understanding the reasons for immigration can potentially shed light on understanding the mental health problems of racial/ethnic minority immigrants.

Despite the reported role of reasons for immigration in immigrants’ health, there has been limited research to examine factor structures which can best represent the interrelationships among diverse reasons for immigration. For example, previous literature (Lee, 1966) proposed two factors of migratory reasons based on the theoretical assumption without a quantitative rationale based on statistical analysis. Also, even though three factors of reasons for immigration were proposed by confirmatory factor analysis (Tartakovsky & Schwartz, 2001), factors derived from their study possibly entail generalizability issues given their participants’ unique characteristics within the specific context (i.e., Jewish from Russia). In other words, factors of immigration reasons derived from a Jewish sample in Russia may be different from those of other racial/ethnic groups living in other countries. To our best knowledge, there has been no study exploring the underlying dimensions of the reasons for immigration specifically targeted at “older racial/ethnic minority immigrants” groups in the United States. Given that older Asian and Latino immigrants are the two major representative groups of racial/ethnic minorities in the United States with heterogeneous characteristics of their immigrant history/context/motivation related to their distinctive cultures, it is crucial to measure factor structures and psychometric properties of reasons for immigration in older Asian and Latino immigrants. Furthermore, we expect that identification of the latent constructs of various reasons for immigration should facilitate further studies by providing the sound conceptual framework to measure these constructs in predicting mental health risks among older racial/ethnic minority immigrants.

The aim of this study is to identify the underlying factor structure of reasons for immigration among older Asian and Latino immigrants in the United States by using exploratory factor analysis (EFA). Given its exploratory nature, we do not set a prior hypothesis about latent constructs of reasons for immigration among older Asian and Latino immigrants. However, we hypothesize that there will be different factor structures of reasons for immigration between older Asian and Latino immigrants, considering the heterogeneity of culture and history among the two groups in relation to their migratory backgrounds.

Research Design and Methods

Participants

The present study used the National Latino and Asian American Study (NLASS, 2002–2003). As a part of the Collaborative Psychiatric Epidemiology Studies (CPES), NLASS is designed to estimate the prevalence of mental disorders, as well as mental health service use by nationwide representative samples of noninstituionalized Latinos and Asian American adult populations(aged 18 years and older) in the United States. A total of 4,649 Latino and Asian Americans were recruited between 2002 and 2003 and participated in either a face-to-face interview, or a telephone interview, when a face-to-face interview was not feasible or requested by an interviewee. Interviews were conducted by trained bilingual interviewers and languages used for interviews in the NLASS were reported as English, Spanish, Vietnamese, Mandarin, Cantonese, and Tagalog. The median length of the NLASS interview was 2.4 hr, and the final weighted response rates were 73.2% for the total sample, 75.5% for the Latino Sample, and 65.6% for the Asian sample (Heeringa et al., 2004). More specific information on the NLASS dataset has been reported in several studies (Alegria et al., 2004; Heeringa et al., 2004). Using a publicly available version of the CPES including the NLASS has been preapproved by the University of Alabama Institutional Review Board.

Since this study focused on older Latino and Asian immigrants, we selected a total of 694 Latino (n = 396) and Asian (n = 298) immigrants who were aged 55 years or older, and who were born outside of the United States for the current analysis.

Measures

Reasons for immigration

Reasons for immigration were measured by nine items in the NLASS asking respondents to rate how important each of nine reasons were, when they decided to move to the United States (3 = very important, 2 = somewhat important, and 1 = not at all important, inversely recoded from original data). Those nine items were as following: (a) to find a job; (b) to join other family members; (c) to improve the future of children; (d) to look for better opportunities; (e) to escape political situations in their country of origin; (f) to escape persecution for political reasons; (g) to seek medical attention; (h) to pursue better educational opportunities; and (i) to avoid marital or family problems (Chen et al., 2009; Gong et al., 2011). Table 1 presents the descriptive statistics for nine items of reasons for immigration among elderly Asian and Latino immigrants. For both groups, “to look for better opportunities” was the most important reason for immigration (M = 2.75, and M = 2.76, respectively), followed by “to improve the future of children” as the second most important reason (M = 2.72, M = 2.69, respectively). Meanwhile, “to avoid marital or family problems” was the least important reason for immigration for both groups (M = 1.55, M = 1.22, respectively). Results from independent sample t tests demonstrated that Asian older immigrants were more likely than older Latino immigrants to consider following four reasons more importantly when deciding to move to United States (p < .001): “to join other family members,” “to pursue better educational opportunities,” “to seek medical attention,” and “to avoid marital or family problems.”

Table 1.

Description and Comparison of Reasons for Immigration Items by Racial/Ethnic Groups

	Asians (n = 298)	Latinos (n = 396)
Item	M ± SD	M ± SD
a. To find a job	2.34 ± 0.85	2.29 ± 0.89
b. To join other family members***	2.48 ± 0.81	2.23 ± 0.90
c. To improve the future of children	2.72 ± 0.64	2.69 ± 0.68
d. To look for better opportunities	2.75 ± 0.58	2.76 ± 0.57
e. To escape political situations in country of origin	2.15 ± 0.92	2.22 ± 0.95
f. To escape persecution for political reasons	1.76 ± 0.92	1.69 ± 0.91
g. To seek medical attention***	1.70 ± 0.90	1.46 ± 0.80
h. To pursue better educational opportunities***	2.50 ± 0.76	2.21 ± 0.88
i. To avoid marital or family problems***	1.55 ± 0.84	1.22 ± 0.59

Note: M = Mean; SD = Standard deviation.

***p < .001.

Analysis

An EFA was conducted with a nine-item instrument designed to measure the reasons for immigration for a total of 694 subjects (Asian = 298, Latino = 396). EFA is a method to compress information contained in original variables into a smaller set of composite factors with a minimum loss of information (Costello & Osborne, 2005). With the aim to identify the underlying latent structures of diverse reasons for immigration by each racial/ethnic group, we conducted assumption testing and EFA for the sample of older Asian immigrants (n = 298) and older Latino immigrants (n = 396), respectively. Before conducting EFA, several assumptions were checked to determine the feasibility of EFA by following the suggestion of Costello and Osborne (2005). For more detailed information about the procedure and criteria for assumption check in our study, please refer to the Supplementary Appendix section. The Kaiser–Meyer–Olkin test of sampling adequacy, the Bartlett test of sphericity, and linearity and outlier checks were also conducted in the process of the EFA. We used SPSS software (Version 24.0.0) for all aspects of the factor analysis in this study, except the multivariate normality assumption test conducted by R software.

Results

Sample Characteristics

Table 2 summarizes the descriptive profiles of these two racial/ethnic groups. In terms of ethnicity in our samples, Latino immigrant elders (n = 396) were comprised of 222 Cubans (56%), 74 Puerto Ricans (18.7%), 33 Mexicans (8.4%), and 67 all “other” Hispanics (16.7%). Among the Asian immigrant elders (n = 298), there were 104 Vietnamese (34.9%), 94 Filipinos (31.5%), 68 Chinese (22.8%), and 32 all “other” Asians (10.7%). It should be noted that Puerto Ricans were included as Latino Immigrants in this study for several reasons: First, even though they are technically U.S. citizens, Puerto Ricans moving to the United States are likely to face similar challenges that Latino immigrants experience, due to their racial/ethnic minority status in the United States and language barrier (U.S. Census Bureau, 2017). Furthermore, Puerto Rican immigrants share similarities with Latino immigrants, characterized by distinct Latin culture, tradition, and value (e.g., Familism), language (e.g., Spanish, as 94% of Puerto Ricans speak Spanish), religion (e.g., Catholicism), and geographic proximities (Clutter & Zubieta, 2009).

Table 2.

Descriptive Characteristics of the Sample

	Overall (n = 694)	Asians (n = 298)	Latinos (n = 396)
Variables	M ± SD or %	M ± SD or %	M ± SD or %
Sex
Male	45%	48.70%	42.30%
Female	55%	51.30%	57.70%
Age	65.32 ± 8.21	64.97 ± 8.15	65.58 ± 8.26
Race/ethnicity
<Asian>
Vietnamese	n = 104 (15%)	34.90%
Filipino	n = 94 (13.6%)	31.50%
Chinese	n = 68 (9.8%)	22.80%
All other Asians	n = 32 (4.6%)	10.70%
<Hispanic>
Cuban	n = 222 (32%)		56%
Puerto Rican	n = 74 (10.7%)		18.70%
Mexican	n = 33 (4.8%)		8.40%
All other Hispanics	n = 67 (9.5%)		16.70%
Total	n = 694 (100%)	n = 298 (43%)	n = 396 (57%)
Work status
Employed	37.70%	41.30%	34.90%
Not in labor force	2.60%	3.70%	1.80%
Unemployed	59.70%	55.00%	63.30%
Household income ($)***	41,466.61 ± 48,782.96	57,170.76 ± 56,832.40	32,390.68 ± 40,972.93
Marital status***
Married/cohabiting	65.40%	79.50%	54.70%
Divorced/separated/widowed	30.70%	17.40%	40.80%
Never married	3.90%	3.00%	5.50%
Educational attainment***
< high school	48.10%	34.60%	58.20%
High school	18.50%	18.10%	18.70%
Some college	14.30%	15.40%	13.40%
Graduate school	19.20%	31.90%	9.60%
Number of Years in US***
<5 years	5.20%	7.00%	3.80%
5–10 years	13.30%	18.10%	9.60%
11–20 years	15.90%	24.20%	9.60%
>20 years	65.70%	50.70%	77%

Note: ***p < .001.

As shown in Table 2, demographic characteristics between the two racial/ethnic elderly immigrant groups were compared. Older Asian immigrants had a mean age of 64.97 (SD = 8.15) and older Latino immigrants had a mean age of 65.68 (SD = 8.26). Among Asian immigrant elders, 48.7% were male and 51.3% were female. In this study, 42.3% of the Latino immigrants were male and 57.7% were female. In terms of work status, 41.3% of Asians and 34.9% of Latinos in the present sample were employed.

Compared to older Latino immigrants, older Asian immigrants showed a significantly higher rate of marriage or cohabitation (p < .001), as well as a lower rate of divorce, separation or widowed status (p < .001). Also, older Asian immigrants showed significantly higher educational attainment and higher household incomes compared to Latinos (p < .001). Seventy seven percent of Latino immigrants reported spending more than 20 years in the United States, but 50.70% of Asian immigrants spent more than 20 years in the United States, showing that Latino immigrants had spent significantly more years in the United States than Asian immigrants (p < .001). In conclusion, demographic analysis suggested the heterogeneous demographic characteristics of the two immigrant groups.

EFA for Older Asian Immigrants

Bartlett’s test of sphericity was statistically significant (χ 2 (36) = 800.282, p < .001), which indicated that the correlation matrix was not an identity matrix. The Kaiser–Meyer–Olkin Measure of Sampling Adequacy (KMO) was .737, suggesting that the data was appropriate to proceed with the factor analysis (Kaiser, 1974). The communalities were all above .3 (Table 3), further confirming that each item shared some common variance with other items. The determinant of the correlation matrix was positive (.065), so we can extract common shared variance.

Table 3.

Pattern Matrix for Reasons for Immigration among Older Asian Immigrants (n = 298)

Items	Factor 1	Factor 2	Factor 3	h2
A. Voluntary reasons
d. To look for better opportunities	0.88	0.04	−0.06	0.73
c.To improve future of children	0.76	−0.12	−0.04	0.57
h. To pursue better education	0.43	−0.24	0.12	0.35
a. To find a job	0.33	0.16	0.33	0.31
B. Involuntary reasons
f. To escape persecution for political reasons	0.01	−0.80	0.13	0.73
e. To escape political situations in country of origin	0.11	−0.77	0.02	0.63
C. Semivoluntary reasons
i. To avoid marital or family problems	−0.09	−0.19	0.74	0.62
g. To seek medical attention	−0.05	−0.09	0.63	0.42
b. To join other family members	0.24	0.14	0.41	0.31
Eigenvalue	3.24	1.67	1.08
% of variance	36.07	18.56	12.08
Cronbach α.	0.71	0.81	0.64

Note: Underlined values indicate a double loading on two factors. Loadings highlighted in bold indicate the factor on which the item was placed. Principal axis factoring with Oblimin rotation was used.

h2: Communality coefficient.

Based on an eigenvalue cut-off of 1.0 and the scree plot (Costello & Osborne, 2005), our EFA identified three factors that explained approximately 66.71% of the total variance in older Asian immigrants (Table 3). Specifically, factor one had a high eigenvalue of 3.24, accounting for 36.07% of the total variance. Factor two accounted for 18.56% of the variance, and factor three accounted for 12.08% of the variance. Table 3 exhibits the factor loadings after rotation, using a significant factor criterion of .32 (Tabachnick & Fidell, 2014) with eigenvalues and the percentage of variance.

As shown in Table 3, items of “to find a job,” “to improve the future of children,” “to look for better opportunities,” and “to pursue better educational opportunities” demonstrated salient loadings on factor one. Since those items were all related to the voluntary pursuit of intrinsic/extrinsic development through immigration, this factor was labelled as “voluntary reasons.” Factor two contained two items describing uncontrollable situations contributing to immigration (i.e., due to political situations and/or persecutory situations in country of origin). So, “involuntary reasons” were given as a label for this second factor among older Asian immigrants. Factor three contained three items related to responsibilities to resolve family or health-related issues (i.e., to join family members, to avoid marital or family problems, and to seek for medical attention). Thus, this third factor was labelled as “semivoluntary reasons,” because medical/marital/familial issues are rarely controllable in nature, yet still require a voluntary determination to migrate to another country in order to address them.

Because the item “to find a job” showed significant loadings (loadings > .32) on both voluntary and semivoluntary factors, we placed this item with the voluntary factor, which is more conceptually related (Pett, Lackey, & Sullivan, 2003; Tabachnick & Fidell, 2014). Cronbach α coefficients were reported as .71 (voluntary reasons, four items), .81 (involuntary reasons, two items), and .64 (semivoluntary reasons, three items) for each factor, which are considered as acceptable internal consistency.

EFA for Older Latino Immigrants

Bartlett’s test of sphericity (χ 2 (36) = 716.012, p < .001) and the KMO test of sample adequacy (value = .67) confirmed the adequacy of the current data for factor analysis (Kaiser, 1974). The determinant of the correlaxion matrix was positive (.16) which also confirmed the suitability of extracting common shared variance from the current data.

Using an eigenvalue cut-off of 1.0 and the scree plot, the factor analysis revealed a total of three underlying factors of reasons for immigration that explain a cumulative variance of 61.03% among older Latino immigrants (Table 4). Specifically, factor one accounted for 28.06% of the variance, factor two accounted for 19.39% of the variance, and factor three accounted for 13.58% of the variance.

Table 4.

Pattern Matrix for Reasons for Immigration among Older Latino Immigrants (n = 396)

Items	Factor 1	Factor 2	Factor 3	h2
A. Voluntary reasons
d. To look for better opportunities	0.83	0.03	−0.13	0.65
c. To improve future of children	0.76	0.05	−0.11	0.54
h. To pursue better education	0.51	0.02	0.19	0.35
a. To find a job	0.49	−0.28	0.06	0.35
b. To join other family members	0.34	0.06	0.12	0.16
B. Involuntary reasons
e. To escape political situations in country of origin	0.06	0.80	−0.01	0.64
f. To escape persecution for political reasons	−0.01	0.73	0.03	0.54
C. Semivoluntary reasons
g. To seek medical attention	−0.08	0.03	0.93	0.83
i. To avoid marital or family problems	0.04	−0.01	0.34	0.13
Eigenvalue	2.53	1.7	1.222
% of variance	28.06%	19.39%	13.58%
Cronbach α.	0.69	0.73	0.45

Note: Loadings highlighted in bold indicate the factor on which the item was placed. Principal axis factoring with Oblimin rotation was used.

h2: Communality coefficient.

A total of five items such as “to find a job,” “to improve the future of children,” “to look for better opportunities,” “to pursue better educational opportunities,” and “to join other family members,” demonstrated salient loadings on factor one when using the significant factor criterion of .32 (Tabachnick & Fidell, 2014). Since these items were related to voluntary pursuits of immigration, similar to the result from Asian sample, this first factor was also labelled as “voluntary reasons.” There was a notable difference that the “join other family members” item was loaded in the factor one (voluntary reasons) in the Latino sample, rather than factor three (semivoluntary reasons), as it was in the Asian sample.

Just as in the results from the Asian sample, factor two contained two items describing uncontrollable situations contributing to immigration (i.e., to escape political and persecutory situations in the country of origin), so “involuntary reasons” were given as a label for this second factor in older Latino immigrants. Factor three, “semivoluntary reasons,” included two items (i.e., to avoid marital or family problems, and to seek for medical attention), reflecting duties to address difficult family and medical issues. Cronbach α coefficients were reported as .69 (voluntary reasons, five items), .73 (involuntary reasons, two items), and .45 (semivoluntary reasons, two items) which are considered as acceptable internal consistency, except with respect to factor three (semivoluntary reasons). However, factor three was still retained in the model because three-factor structures showed a better model fit and and explained more variance compared to two-factor structures in the sample of older Latino immigrants.

Discussion and Implications

Even though reasons for immigration have been proposed as important factors to predict the adjustment and various health outcomes of immigrants (Fabrega et al., 1969; Gong et al., 2011; Giuntella et al., 2018; Ogbu & Simons, 1998; Tartakovsky & Schwartz, 2001), there has been no study examining the underlying structures of reasons for immigration among elderly racial/ethnic minority immigrant populations in the United States. Thus, the current study conducted EFA to identify the factor structures of diverse reasons for immigration among elderly Asian and Latino immigrants from a nationally representative dataset.

Our EFA extracted a total of three factors regarding reasons for immigration among both older Asian and Latino immigrant groups: (a) “voluntary reasons,” regarding willful motivations for intrinsic/extrinsic development through immigration, such as better opportunities in job, education, and future for oneself and family; (b) “involuntary reasons,” due to uncontrollable situations, such as political situation and persecution from the country of origin; and (c) “semivoluntary reasons,” reflecting duties to resolve marital, family, or medical issues.

Compared to existing literature on reasons for immigration, the “voluntary reasons” factor observed in this study has its conceptual similarities with a pull factor (Lee, 1966), voluntary immigration (Ogbu & Simons, 1998), and motivations for self-development and materialistic accumulation of financial well-being (Tartakovsky & Schwartz, 2001) in that they encompass intrinsic/extrinsic development through immigration. Also, given its involuntary and uncontrollable nature of immigration, the “involuntary reasons” factor derived from the present study is conceptually similar to such existing concepts as push factor (Lee, 1966), involuntary immigration (Guarnaccia & Lopez, 1998), and motivation for preservation of security (Tartakovsky & Schwartz, 2001).

Despite the aforementioned similarities with existing theories, it is notable that the current EFA found one additional distinctive factor structure of “semivoluntary reasons” for immigration, which has been overlooked in previous theories of migratory reasons (Guarnaccia & Lopez, 1998; Ogbu & Simons, 1998; Lee, 1966; Tartakovsky & Schwartz, 2001). In other words, the “semivoluntary reasons” factor derived in this study represents the gray area that classical bifactor conceptual frameworks (i.e., pull/push factors, or voluntary/involuntary factors) often overlook, because duties to address marital/familial/medical issues entail both voluntary and involuntary characteristics of immigration. That is, duty-related immigration requires a “voluntary” determination to resolve “forced” situations through immigration. Additionally, compared to the three factors of migratory reasons derived from Jewish immigrants from Russia (e.g., preservation of security; self-development; financial well-being) (Tartakovsky & Schwartz, 2001), our EFA results yielded qualitatively distinctive three-factor structures of migratory reasons among older Asian and Latino immigrants in the United States (e.g., voluntary, involuntary, semivoluntary reasons). Our finding suggests that reasons for immigration should be understood as a dynamic construct which can be heterogenous reflecting unique racial, ethnic, cultural, and historical contexts.

Another interesting finding is a racial/ethnic difference in one item (i.e., to join the family members in the United States), which is loaded into the different factor between older Asian and Latino immigrants. For older Asian immigrants, joining other family members was more likely to be associated with semivoluntary reasons related to duties to address marital, family, and medical problem in the United States. However, for older Latino immigrants, joining other family members was more related to voluntary reasons such as seeking for better opportunities, education, future, and jobs in the United States.

To elucidate this difference, heterogeneity in immigration background and constructs of traditional family values in these two racial/ethnic groups should be considered. First of all, potential explanations can be found from the differences in geographic proximity between the two immigrant groups’ countries of origin and the United States. In fact, 55% of the U.S. Hispanic population resides in the southwest (e.g., California, Florida, and Texas) which is geographically close to their countries of origin, such as Mexico and Latin America (Brown & Lopez, 2013). So, Latino communities located in adjacent southwest areas, sharing similar culture and the same language, could easily attract Latino immigrants to join family and relatives who are already settled in pursuit of better opportunities in the United States. However, in the case of Asian immigrants, moving across continents to join family members could be a huge and stressful decision accompanied by abrupt changes in perceived social status (Leu et al., 2008) and cultural differences in western society.

Second, we can also compare traditional family values between Asian and Latino immigrants to explain why an item “to join the family members in the United States” was loaded on different factors in the two racial/ethnic immigrant groups. Even though “family” plays a central role in both Asian and Latino cultures, traditional family values in two cultures differ in many ways. For example, Asian family value emphasizes filial piety—an obligation to achieve family harmony by prioritizing family interests over individual interests, and abiding by the opinions of one’s parents and elders in a hierarchical relationship (Chuang, 2005; Schwartz et al., 2010; Ying & Han, 2007)—as the ethics of behavior. In this context, even if one does not want to migrate, independent behaviors which can disturb family harmony tend to be highly discouraged in traditional Asian cultures. Therefore, for Asian immigrants, the decision to join other family members in the United States could be more likely to be perceived as “semivoluntary”; because it is related to duty and responsibility to keep a harmony, rather than a purely voluntary motivation for success and/or better opportunities.

In contrast, Latino family value, Familismo, stresses strong ties with immediate as well as extended family members, and dedication to family goals to pursue “prosperity” of the kinship networks (Burgess and Locke, 1945; Cauce & Domenech-Rodriques, 2002). For instance, Familismo puts a high value on sharing one’s resources with relatives who are in need, and physical proximity to family members as a pursuit of family interdependence (Sabogal, Marín, Otero-Sabogal, Marín, & Perez-Stable, 1987; Steidel & Contreras, 2003). Thus, it is possible that Latinos who migrated to “join the family in the United States” could be more likely to perceive their motivation as “voluntary” related to seeking for prosperity, with preexisting family support system that they can depend on. This potential explanation is also supported by previous findings that Familismo can buffer against several challenges that Hispanic immigrants encounter during immigration and acculturation process as a protective factor (Baca Zinn, 1994). Despite the aforementioned potential explanations, more research is still needed to fill the gaps and explain these different item loadings between the two older racial/ethnic minority groups.

A three-factor framework of reasons for immigration identified in the present analysis provides clinical implications to promote more culturally sensitive interventions for older racial/ethnic minority immigrants. First of all, clinicians and mental health professionals working with older Asian/Latino immigrants can utilize “three factors of migratory reasons” as a useful framework to conceptualize and intergrate patients’ unique immigrant history and cultural factors into the treatment planning. For example, according to the guideline for culturally competent evaluation and treatment (Center for Substance Abuse Treatment, 2014), clinicians working with the immigrant population are strongly encouraged to ask questions about their immigration history (e.g., migratory reasons; length of stay in the United States; perceived changes in social status), because immigration history can inform their level of support system and challenges which affect immigrants’ adjustment and health. Similarly, considering that migratory reasons are known to be a significant predictor of immigrants’ physical/mental health in previous studies (Gong et al., 2011; Giuntella et al., 2018), our three-factor structures of reasons for immigration (“voluntary”, “semivoluntary,” and “involuntary reasons”) can provide a rigorous theoretical framework to help clinicians conceptualize diverse immigration background for more culturally sensitive intervention for older Asian and Latino immigrants.

Second, it is essential for clinicians and mental health professionals to understand that heterogeneous immigration backgrounds and cultural values can potentially explain some racial/ethnic differences in the underlying structure of migratory reasons. For example, in the case of older Asian immigrants, motivation to join the family members in the United States could be more related to pressures and responsibilities stemming from Asian family values (i.e., filial piety), whereas the same motivation to join the family can be more related to the pursuit of better opportunities for older Latino immigrants in the context of Hispanic family values (i.e., Familismo).

Several limitations of this study could provide further discussion and direction for future research. First, due to the limited number of reasons for immigration provided by the secondary survey dataset, we could not entirely capture the broad range of migratory reasons in this study, such as seeking religious freedom or escaping natural disasters. It is important to note that numerous reasons for immigration can exist, depending on unique personal, social, economic, and environmental conditions in the country of origin (Rubenstein, 2016). In this sense, our findings might not be able to capture the complexity and/or dynamics of diverse reasons for immigration in each subethnic groups. Based on preliminary findings from current EFA study, further research is needed to collect the broader scope of reasons for immigration and conduct Confirmatory factor analysis (CFA) to verify the representativeness of the current three-factor structure.

Second, the current study aggregated the data of participants from diverse countries into two categories of “Asian” and “Latino” to obtain an appropriate sample size for statistical power. Considering the diversity within the same racial/ethnic immigrant groups in the United States (Ishii-Kuntz, 2000; Kim et al., 2010), further research should consider whether factor structures of migratory reasons differ by countries of origin by collecting more sample size from diverse ethnic groups.

Third, given that the current study was based on the cross-sectional dataset collected nearly a decade ago, caution should be exercised when generalizing our findings. It is important to note that temporal changes in immigration policies/circumstances can interact with reasons for immigration in a dynamic way. Thus, a three-factor framework derived from the present study may not represent contemporary migratory reasons reported by older immigrants. Despite a limitation as an old dataset, to our best knowledge, the NLASS is one of most recent nationally representative study which provides approximately 70% response rate for items of unique immigration experiences from a large sample of Asian and latino older immigrants in the United States (Heeringa et al., 2004). It is highly advised that future studies conduct a confirmatory factor analysis using more recently collected sample of older Asian and Latino immigrant to consolidate the generalizability of the three-factor structures from this study.

Fourth, reasons for immigration were measured based on the retrospective responses regarding older immigrants’ motivation at the time of immigration in the past. Thus, three factors observed in the present study may not fully capture “reasons why older adults decided to immigrate in old age,” without controlling the time factor of immigration. In addition, memory biases may have affected older immigrants’ responses (Chen et al., 2009). Future research needs to limit the sample only to older adults who migrated in their old age in order to examine potential differences in factor structures.

Lastly, although the sample was drawn from a nationally representative data set, characteristics of the present sample did not reflect the actual racial/ethnic demographics of the U.S. population, which may limit external validity (Onwuegbuzie, 2003). For example, Mexicans (62%) and Chinese (24%) are known as the largest ethnic subgroups in each immigrants population in the United States (López et al., 2017; U.S. Census Bureau, 2017), whereas in this study, Cubans (56% of older Latino immigrants) and Vietnamese (34% of older Asian immigrants) were the largest ethnic subgroups in each racial/ethnic immigrant group. Thus, the interpretation of current findings may require careful consideration of the potential biase in the sample. For example, because both Vietnamese and Cuban immigrants have shared unique immigration history as “refugees” due to political situation in the past, results are not generalizable to entire Asian/Latino immigrant population. Further research needs to be designed to maximize the external validity by collecting samples reflecting the actual demographic characteristics.

Despite these limitations and previous efforts to conceptualize different types of migratory reasons (Lee,1966; Tartakovsky and Schwartz, 2001), to our knowledge, this study is the first attempt to explore the underlying structures of diverse reasons for immigration among underserved elderly racial/ethnic minority populations in the United States: older Asian and Latino immigrants. Findings from our study provide a theoretical rationale for using three distinct categories of reasons for immigration (i.e., voluntary, involuntary, and semivoluntary reason) rather than depending on a bifactor structure (Lee, 1966) or factors derived from heterogeneous sample of immigrants such as Jewish immigrants (Tartakovsky & Schwartz, 2001), when studying or working with two major elderly immigrant groups in the United States. Moreover, our findings suggest that a migratory reason to join the family should be considered differently for elderly Asian (i.e., semivoluntary reasons related to duties) and Latino immigrants (i.e., voluntary reasons related to development), reflecting the two groups’ heterogeneous migratory background and unique value systems. Based on current findings, future research should be extended to examine the relationship between distinctive reasons for immigration and the physical/mental health of older Asian and Latino immigrants.

Funding

None reported.

Conflict of Interest

None reported.

Click here for additional data file.