Abstract

The impact of immigrant parents’ premigration family background on their second-generation children residing in destination countries remains underexplored in the literature on historical social mobility. Using multigenerational historical survey records from the Japanese American Research Project, this study investigates the influence of premigration socioeconomic and cultural background of Japan-born grandparents and parents on the social mobility of second-generation Japanese Americans born in the continental United States in the early twentieth century. The analysis reveals the enduring effects of family premigration socioeconomic status, as indicated by occupation and education, and culture conducive to upward mobility, proxied by samurai ancestry, on second-generation Japanese Americans’ educational and income levels. These effects may extend back to their nonmigrant grandparents and possibly contrast with their European second-generation immigrant counterparts, who typically experienced upward mobility regardless of their family background. The results point to the critical role of origin-country socioeconomic status and culture in immigrant social mobility research, particularly for populations whose negative reception has hindered their resource access in their new countries.

Introduction

Considering family socioeconomic and cultural resources from the parental countries of origin is imperative for explaining the socioeconomic attainment of second-generation children of immigrants born in destination countries (Feliciano and Lanuza 2017; Luthra et al. 2018; Pong and Landale 2012; Potochnick and Hall 2021). An important exception has been documented in the literature on historical social mobility during the Age of Mass Migration (mid-nineteenth to early twentieth century), a period in which the United States incorporated several million immigrants and their children from diverse socioeconomic and cultural backgrounds (Abramitzky and Boustan 2017; Hatton and Williamson 1998). Recent studies have shown that U.S.-born children of these European immigrants were upwardly mobile regardless of family resources from their parental country of origin, measured by the premigration socioeconomic status of their immigrant parents or nonmigrant grandparents (Catron 2020; Connor 2020). Favorable reception contexts in the United States—such as the expansion of schooling, local labor market, and easy access to parental naturalization—have been linked to successful upward mobility (Catron 2019, 2020; Connor 2020).

This study extends the literature by investigating the socioeconomic achievements of second-generation Japanese Americans in the continental United States,1 born to immigrant parents who arrived during the Age of Mass Migration. The historical Japanese American population offers a unique lens to investigate the roles of premigration parental socioeconomic and cultural background in upward social mobility. Similar to their European counterparts, second-generation Japanese Americans are considered a successful group: they had an advantage in educational attainment (Hirschman and Wong 1986) and achieved occupational attainment comparable to their U.S.-born White counterparts by the 1960–1970s (Hirschman and Wong 1984; Varon 1967). However, the Japanese were subjected to particularly harsh socioeconomic and legal discrimination and exclusion. For example, first-generation Japanese immigrants faced restrictions on landownership and citizenship acquisition, and the entire West Coast Japanese community (including U.S.-born second-generation children) experienced forced relocation and internment after the 1941 Japanese attack on Pearl Harbor (Broom and Riemer 1973; Daniels 1988: chaps. 6–7). Such difficulty in accumulating or capitalizing on U.S.-based resources may have led some Japanese families to draw more heavily on the resources they inherited from Japan for upward social mobility.

In particular, although European immigrants often experienced negative socioeconomic selection from their origin countries (e.g., Abramitzky et al. 2012; Connor 2019; Spitzer and Zimran 2018), owing to restrictions imposed by both the Japanese and U.S. governments, Japanese first-generation immigrants in the continental United States experienced positive socioeconomic selection2 relative to their peers in Japan (Hirschman and Wong 1986:9; Kihara 2022; Sawada 1991; Suzuki 1995; Wakatsuki 1979). This group comprised not only laborers but also a sizable number of entrepreneurs, better-off farmers, and students who eventually settled permanently (Azuma 2005: chap. 1; Ichihashi 1932: chap. 5; Wakatsuki 1979). Furthermore, Japanese immigrants may have brought with them cultural values conducive to upward mobility. A segment of Japanese youths who made it to the United States was influenced by a strong upward mobility–oriented culture rooted in the self-advancement ideals of the former samurai class (i.e., shizoku, individuals in Meiji-era Japan whose ancestry was traced to the warrior class but who no longer held privileges), conveyed through popular publications promoting emigration to the United States (e.g., Azuma 2005:24–25; Tachikawa 1986 for emigration publications; Kinmonth 1982; Takeuchi 1975; Yasuda 1969 for samurai culture). Considering the persistent samurai-ancestry advantage through the family transmission of mobility-oriented values in Japan (e.g., Yasuda 1969, 1971), such values might have similarly facilitated upward mobility in the United States.

This study analyzes linked historical micro-records from the Japanese American Research Project (JARP), a nationally representative social survey of Japanese immigrants and their descendants in the continental United States (Levine 2006). Unlike the census (e.g., Suzuki 1995), the JARP survey offers detailed measures that consider the socioeconomic and sociocultural contexts of the Japanese population in the late nineteenth and early twentieth centuries. Specifically, this study leverages a feature in the JARP survey that asks first-generation immigrant parents about their premigration socioeconomic and cultural background and those of their parents in Japan. I construct a three-generation family data set linking zero-generation grandparents (born in Japan), first-generation immigrant parents (born in Japan and migrated to the United States), and second-generation (grand)children (born in the United States). Here, zero-generation grandparents refers to the parents of first-generation Japanese immigrants and the grandparents of U.S.-born second-generation Japanese Americans (see also Figure 2). Notably, most zero-generation grandparents are unlikely to have migrated internationally and likely spent their entire lives in Japan. Given the availability of extensive variables—such as occupation, farm size, Japanese school types, and samurai ancestry—this study can integrate highly relevant contextual information about premigration family socioeconomic and cultural backgrounds in the Japanese context.

In contrast to findings of recent research on European immigrants and their children reporting weak or null effects of zero-generation or first-generation socioeconomic status on second-generation attainment (Catron 2020; Connor 2020), I find the enduring effects of family premigration socioeconomic status and mobility-oriented culture on U.S.-born second-generation education and income. These associations are substantially meaningful and exist net of U.S.-based resources, such as parental postmigration socioeconomic status, parental Japanese prefectures of origin, or U.S. birth counties.

Background

Social Mobility Among the Immigrant Population: Origin-Country Perspective

The status attainment model (Blau and Duncan 1967) typically views mobility as a process whereby parents’ occupational and educational statuses significantly affect their children's socioeconomic status. Various studies of immigrant assimilation have incorporated this model (Alba and Nee 2003:28). However, a long-standing puzzle has been the limited applicability of the status attainment model to immigrant intergenerational mobility in the United States, where the transmission of immigrant parents’ socioeconomic status to their children is weak (Drouhot and Nee 2019; Luthra and Soehl 2015). Moreover, related research has stressed that the prediction of the status attainment model does not align well with the U.S. Asian population, which typically enjoys high upward intergenerational mobility regardless of family background (Fishman 2020; Tian 2023: chap. 1).

Several explanations, not necessarily conflicting, have addressed immigrant populations’ deviation from the typical social mobility processes. First, framing and measuring parental socioeconomic status exclusively in the destination country and ignoring the family socioeconomic status in origin countries’ context is problematic because immigrant parents experience significant downward mobility upon arrival (Feliciano and Lanuza 2017; Luthra et al. 2018; Pong and Landale 2012; Potochnick and Hall 2021). Most first-generation immigrant parents are educated and begin their careers in their country of origin. At the time of migration, they often experience positive selection from their home communities (Feliciano 2005, 2020). However, upon arrival in destination societies, they experience downward mobility because of language barriers, devaluation of origin-country education, or outright discrimination (Akresh 2008; Zeng and Xie 2004). Despite these challenges, they retain their human capital and self-awareness of social status based on their premigration background and strive to pass this status to their children by providing psychological or educational resources (Brunori et al. 2020; Ichou 2014; Pong and Landale 2012). Studies on second-generation immigrants in contemporary Europe and the United States have found that parental premigration socioeconomic status or their socioeconomic attainment in the context of their origin country, such as premigration occupation or education, explains a significant part of second-generation attainment, independent of postmigration status (Feliciano and Lanuza 2017; Ichou 2014; Pong and Landale 2012; Potochnick and Hall 2021).

Second, some scholars have contended that immigrant social mobility necessitates considering culture as a driver of mobility. Because immigrant parents are socialized in their families in their origin countries, they bring specific cultural values and norms to the destination country (Luthra et al. 2018; Polavieja 2015). Specifically, research on Asian Americans found that a strong cultural belief or mindset regarding education at least partially explains the higher socioeconomic attainment of Asians over other racial and ethnic groups (Chen and Stevenson 1995; Hsin and Xie 2014; Liu and Xie 2016; Sakamoto et al. 2009) and thus the deviation from typical status attainment models (Fishman 2020; Kim and Kim 2023).

A similar cultural perspective is reflected in the works of Jennifer Lee and Min Zhou (Lee and Zhou 2015; Zhou and Lee 2014, 2017), who posited that the “success frames” shared among Asian immigrants and their second-generation children facilitate their socioeconomic success in the United States regardless of parental socioeconomic background (Lee and Zhou 2015:51–68). However, they challenged the notion that culture operates independently, proposing that these success frames are influenced by migration patterns and ethnic socioeconomic structure. They introduced the term “hyper-selectivity” to describe the trend among certain immigrant groups, such as contemporary Asian immigrants, who not only exhibit higher education levels than their nonmigrant peers in their home countries (i.e., positive educational selection) but also exceed the education levels of the U.S.-born White population (Lee and Zhou 2015:30–32). Lee and Zhou argued that such migration patterns accumulate ethnic capital that can benefit the success of coethnic peers (Lee and Zhou 2015; Zhou and Lee 2014, 2017). However, some researchers have questioned this mechanism, arguing that culture is an independent force that should not be reduced to socioeconomic resources or ethnic capital (e.g., Kim and Kim 2023; Sakamoto 2017).

Historical Patterns of Immigrant Mobility in the Age of Mass Migration

Historical research has also highlighted the deviation of immigrant populations in the United States (predominantly from Europe) from the traditional status attainment model, noting immigrants’ strong upward mobility since the late nineteenth century (e.g., Abramitzky et al. 2023; Alba and Nee 2003; White and Mullen 2016). Recent studies using linked microdata from census and other records have corroborated this view. By analyzing linked 1880–1910 and 1910–1930 U.S. decennial census records of European immigrant parents and their children, Abramitzky et al. (2021) revealed that second-generation children with immigrant parents at the lower end of the income distribution had a much higher income rank in adulthood than the children of U.S.-born parents from the same income bracket, indicating much weaker patterns of intergenerational income associations (i.e., high-income mobility) among immigrant families. Similarly, by linking data from families across the U.S. Census, Current Population Surveys, and American Community Surveys from the 1940s to 2017, Lowrey et al. (2021) explored long-term educational mobility and reached a similar conclusion regarding European immigrants, their children, and grandchildren. Tian (2023: chap. 1) examined Japanese and Chinese populations before World War II using a linked census and found that second-generation Japanese and Chinese Americans were more likely to attain higher absolute and relative education than their parents relative to native-born Whites; this trend did not extend to occupational achievements, however. Offering a notable counterpoint, Ward (2020) provided evidence of the strong persistence of parental and grandparental status across immigrant generations: the high mobility perceived in prior studies (i.e., small intergenerational and multigenerational coefficients) partly results from attenuation bias stemming from measurement errors related to parental status in the census data.

Unfortunately, data limitations have prevented most historical studies on immigrant intergenerational mobility from incorporating the premigration socioeconomic status of parents or culture in their models. The rare exceptions are two studies (Catron 2020; Connor 2020) that used novel data and found little persistence of premigration parental (first-generation) or grandparental (zero-generation) socioeconomic status on the status attainment of U.S.-born children of European immigrants. By linking the 1901 Irish census with the 1920 and 1940 U.S. decennial censuses, Connor (2020) studied whether grandparental occupation in Ireland was associated with second-generation Irish Americans’ socioeconomic status. Connor found that although grandparental socioeconomic advantage in Ireland did not confer an educational advantage to second-generation Irish Americans, it did confer an income advantage. However, this income advantage was small and observed only at the top of the distribution, among less than 10% of second-generation children whose grandparents had a white-collar background (as opposed to a poorer farmer background) in Ireland. Further, after first-generation fathers’ U.S. occupation was added as a predictor, the coefficient for the grandparental background weakened to a level not statistically different from zero. Likewise, Catron (2020) studied Italian, German, Russian, and Jewish immigrant populations by linking the U.S. decennial census to transatlantic ship passenger records with information on the immigrant parents’ premigration occupation and ship travel compartment (e.g., first class, steerage). Catron found that although the first-generation's premigration status strongly predicted their postmigration U.S. occupation (i.e., intragenerational mobility), the status differences among the first generation became quite small among the second generation owing to the high upward mobility of those from lower socioeconomic backgrounds.

Studies on historical immigrant social mobility generally concur that strong upward mobility is attributable to the favorable U.S. reception contexts. Using neighborhood fixed effects, Abramitzky et al. (2021) found that much of the upward mobility of children of low-income immigrant parents can be explained by immigrants’ sorting to states and neighborhoods more favorable toward them and their children's mobility. Catron (2019) found that easy access to citizenship for European immigrant parents in the early twentieth century may have facilitated their children's mobility: children of naturalized immigrant parents were more educated and had better labor market outcomes than children of noncitizen immigrant parents. Catron (2020) and Connor (2020) highlighted expanded schooling and U.S. labor market opportunities as important factors facilitating the upward mobility of U.S.-born children of European immigrants regardless of premigration family background. Moreover, Ward (2020), who contrarily argued for a stronger persistence of grandparental and parental status, concurred with other researchers on the significant role of geographic mechanisms.

The Current Study: The Case of the Japanese Population

Pre–World War II migration of first-generation Japanese immigrants to the continental United States, commonly known as the Issei, mostly occurred over a period of less than 40 years. An inflow of young men beginning in the late 1880s was followed by young women, including “picture-brides,” in the 1900s and 1910s (Daniels 1988: chaps. 4–5; Ichioka 1988: chap. 5). The Immigration Act of 1924, also known as the “Japanese Exclusion Act” in Japan, significantly curtailed new migration from Japan. Consequently, the majority of pre–World War II second-generation Japanese American children, commonly known as the Nisei, were those born in the 1910s and 1920s to immigrant parents who had arrived in the continental United States before 1924 (Daniels 1988: chap. 5). Mirroring this broader historical context, approximately 75% of the second-generation sample in this study were born in the United States between 1910 and 1930.

The U.S.-born Japanese second generation likely experienced a stronger negative reception than their European American second-generation counterparts. While attending public schools, they faced strong interpersonal and institutional discrimination, including pressure and attempts to segregate them from White children (Daniels 1988: chaps. 4–5). Despite these circumstances, second-generation Japanese immigrants were so socioeconomically successful that their attainment was, on average, comparable to that of the native-born White population. Using census data, Hirschman and Wong (1986) examined educational attainment by sex and birth cohort and found that all birth cohorts of U.S.-born Japanese second-generation men had comparable or better educational attainment relative to native-born Whites; findings for U.S.-born Japanese women, except the pre-1915 birth cohort, were similar. In the labor market, which most of these second-generation immigrants entered in the 1930s to 1950s, they faced strong discrimination (Bonacich and Modell 1980: chaps. 5–6; Ichihashi 1932: chap. 22). However, after the war, a significant portion of Japanese second-generation immigrants entered professional/white-collar jobs and reached similar occupational levels to those of Whites by 1960, although they were still underrepresented in these occupations given their high educational attainment (Hirschman and Wong 1984; Varon 1967).

Premigration Family Socioeconomic Background

The literature on immigrant social mobility (Feliciano and Lanuza 2017; Pong and Landale 2012) suggests that the success of second-generation immigrants might stem partly from their parents’ premigration socioeconomic status in Japan rather than their parents’ postmigration status in the United States. First-generation Japanese immigrants—the parents of the second generation—included not only laborers but also entrepreneurs, wealthier farmers, and a significant number with secondary education (Azuma 2005: chap. 1; Kihara 2022; Wakatsuki 1979). Such educational opportunities were typically available to those from the middle or higher socioeconomic strata, such as sons of white-collar or better-off farming families, in Japan's Meiji and Taisho eras (e.g., Tsuburai and Sato 1995).

Even among their peers in the United States, Japanese first-generation immigrants were some of the most educated. Using the IPUMS Census 1% file (Ruggles et al. 2022), Figure 1 plots the average years of education (x-axis) against the standard deviation (y-axis) for men born in 1871–1910 (aged 30–69 in 1940). It includes the 20 largest groups of foreign-born men by birth country, native-born Whites and Blacks, and Japan-born immigrant men. More than 95% of the Japanese first-generation parents in my later analysis belong to these birth cohorts. It is clear that among their foreign-born peers, Japanese fell into the most educated category; however, at the same time, Japanese immigrants did not hold an education advantage over native-born Whites and exhibited significantly more variability than other well-educated groups, suggesting potential socioeconomic diversity and the usefulness of education as a socioeconomic marker.

The negative reception of Japanese parents in the United States prevented many from fully realizing their socioeconomic potential regardless of their premigration backgrounds. Using linked historical census data from 1920 and 1940, Kihara (2023: chap. 2) revealed that Japanese first-generation immigrants faced lower initial incomes upon arrival, experienced limited upward mobility, and were at a significant disadvantage relative to European immigrants with similar education. Furthermore, most Japanese immigrants and their U.S.-born children were interned in relocation centers in the early 1940s (i.e., when most of their second-generation children were in their 20s), significantly disrupting their lives and leading to economic loss (Broom and Riemer 1973; Daniels 1988: chaps. 6–7). Such host-society contexts likely weakened status transmission through economic mechanisms—for example, by disrupting wealth accumulation for Japanese small business owners and subsequent transfers to their children. Nevertheless, immigrant parents still might have been able to pass down their premigration family socioeconomic advantages by mobilizing resources through their family ties in Japan or providing psychological and educational resources. Therefore, I hypothesize the following:

  • Hypothesis 1A: Second-generation Japanese Americans whose first-generation parents had a higher premigration socioeconomic status in Japan were more likely to achieve a higher socioeconomic status in the United States, regardless of their parents’ postmigration education and occupation.

Additionally, I propose that the socioeconomic status of the grandparents (i.e., the zero generation) in Japan of second-generation children persisted in the United States:

  • Hypothesis 1B: Second-generation Japanese Americans whose first-generation parents came from better-off families in Japan (i.e., had zero-generation grandparents with higher socioeconomic status) were more likely to achieve higher socioeconomic status in the United States.

Premigration Family Sociocultural Background

Another possibility is a cultural explanation. The long-standing social psychology and sociology literature has often cited culture to explain the higher achievement of Japanese Americans relative to other racial and ethnic groups (Chen and Stevenson 1995; Kitano 1962, 1976; Sakamoto et al. 2012). Furthermore, Tian (2023: chap. 1) suggested that independent cultural forces might be a driver of the significant upward mobility of the Japanese (and Chinese) relative to U.S.-born Whites. Although this extant literature more strongly emphasizes the cultural differences between ethnic groups, culture also varies probabilistically within ethnic groups (Polavieja 2015).

Samurai heritage, referred to as shizoku status in the Meiji period, is an effective indicator to capture variability in mobility-oriented cultural values among Japanese families. The social, legal, and economic privileges associated with being a samurai were eliminated in the 1870s after the Meiji Restoration, impoverishing the samurai class and leading to revolts that were quelled by the new Imperial Army (Buck 1973; Harootunian 1959; Ochiai 2015). Although some ex-samurai transitioned to white-collar or professional occupations under the new Meiji government, most faced economic hardship, prompting social programs, such as shizoku-jusan, which included promoting migration to the northern island of Hokkaido to alleviate poverty (Harootunian 1960; Ochiai 2015: chaps. 5–6). Despite these extreme instabilities and hardships, samurai descendants predominantly experienced upward mobility (Yasuda 1969, 1971). Japanese sociologists have partly attributed this upward mobility to the intergenerational and multigenerational transmission of cultural values (e.g., Yasuda 1969:33–36).

Sometimes referred to as the ideology of self-advancement (risshin-shusse), a set of strong mobility-oriented cultural values emerged after the economic and legal privileges of being samurai were lost; these values were transmitted through the family line (Takeuchi 1975; Yasuda 1969, 1971). This mobility-oriented creed emphasized self-discipline related to attaining high social status through education and was especially prevalent among those from lower-ranking former samurai families (Kinmonth 1982; Takeuchi 1975, 1981). Indeed, Yasuda's analyses of 1960s data from Tokyo speak to the importance of mobility-oriented values by showing that structural factors alone could not explain samurai descendants’ upward mobility; familial transmission of values also played a role (Yasuda 1969:35–36).

Although Japanese sociologists have examined samurai heritage only in the context of Japanese stratification, samurai descendants in the United States are also likely to hold such an advantage because values and beliefs are transferrable across national boundaries. Samurai ancestry was not uncommon among Japanese immigrants in the continental United States (Wakatsuki 1979:481–483). Therefore, I hypothesize the following:

  • Hypothesis 2: Second-generation Japanese Americans with samurai ancestry were more likely to achieve higher socioeconomic status in the United States.

Properly interpreting the cultural influence of samurai heritage requires accounting for ancestors’ occupations by jointly testing Hypotheses 1B and 2. As noted earlier, despite experiencing widespread setbacks during the Meiji period, some ex-samurai secured emerging roles as primary school teachers or civil servants, new types of professional/white-collar occupations that became available under the Meiji regime (Sonoda et al. 1995).

Data and Methods

Data

JARP Survey and Sample Construction

This study used the first- and second-generation surveys of the JARP conducted in the 1960s. Housed at the University of California, Los Angeles, the JARP survey was part of a larger academic project by Japanese American communities, sociologists, and historians to preserve the Japanese experience in the continental United States.3 The survey's initial phase, aided by Japanese American community organizations, involved compiling a comprehensive list of every first-generation Japanese immigrant residing in the contiguous United States in 1962 who had lived through World War II (Levine 2006). The project excluded two groups from the sampling frame because of their distinct migration patterns and World War II experiences: (1) Japanese immigrants in Hawaii; and (2) post–World War II Japanese immigrants, such as scientists, businessmen, and Japanese spouses of U.S. military personnel (Levine and Rhodes 1981; Minamikawa 2007:213–214). Approximately 18,000 first-generation immigrants were identified (Levine 2006). The online appendix (section A) presents a detailed analysis of the coverage and representativeness of the initial list of 18,000 using information from the census, showing a reasonable estimated coverage of 75% or higher.

From this extensive list, the survey drew a first-generation sample of 1,043 individuals, most of whom were interviewed between 1962 and 1965. These respondents were predominantly older than 70 in 1962. Their second-generation Japanese American children were surveyed in 1967 using names and addresses the first-generation parents provided. Of the 1,043 first-generation respondents, 993 provided valid responses on all relevant variables. Among these 993 individuals, 850 (86%) had at least one child who participated in the second-generation survey, contributing an average of 2.4 and a maximum of 10 second-generation Japanese Americans to the analytic sample. The online appendix (section B) compares the profiles of the 850 first-generation respondents with at least one linked second-generation child and the 143 first-generation respondents without linked second-generation children.

The analytic sample was 2,072 second-generation Japanese American adults (aged 25 or older at the time of the survey) nested in 850 first-generation parents.4 The income analysis was restricted to 1,036 second-generation men nested in 678 first-generation parents. Information about the zero-generation grandparents was obtained from the first-generation immigrants’ responses about their parents. Figure 2 provides information on data structure, variables, and demographic contexts.

Note that Levine and colleagues, the original members of the JARP project, used the JARP survey to examine questions of social mobility and acculturation (Levine and Montero 1973; Levine and Rhodes 1981; Montero 1980). However, their analyses comprised cross-tabulations of two variables and did not capitalize on the richness of the survey questions or the unique data structure by linking the parent–child pairs and applying modern statistical analyses. Wong (1982) also raised this point in his critical review essay of then-recent works on Japanese Americans. This study responds to Wong's (1982) criticisms and fully uses the JARP data structure and the information available.

Construction of Weights: Comparing the JARP Second-Generation Sample With the 1970 Census

The JARP data representativeness was verified by cross-checking with the census data, which indicated that the bias is probably modest and can be addressed with weights. First-generation interviews, conducted in person by a Japanese-speaking interviewer, had a refusal rate of less than 1% (Levine and Rhodes 1981:18–19).5 Although the survey in 1967 attempted to collect responses from all second-generation children of the first-generation survey respondents, the second-generation data had a lower response rate (60%) as a result of using mail questionnaires for some respondents and the emphasis on reaching at least the eldest son (Levine and Rhodes 1981:18–20). The original JARP researchers were concerned that the second-generation sample was overeducated because of selective nonresponse of the less educated (Levine and Rhodes 1981:23).

My analytic sample and the corresponding second-generation population in the 1970 census did not statistically differ in sex ratio, birth year, or years of education. However, a statistically significant difference emerged when education was operationalized as a categorical variable. To ensure representativeness, I used a probability weight at the second-generation level in the analysis based on the distribution of education by birth cohort and sex (see the online appendix, section C, for details).

Variables and Measures

Dependent Variable: Second-Generation Socioeconomic Status

The first key outcome was educational attainment. In the JARP second-generation survey (Levine 2006), the highest grade completed was categorized as no schooling, 1–4 years, 5–7 years, 8 years, 9–11 years, 12 years (high school graduates), 13–15 years (some college), 16 years (college graduates), and more than 16 years (some graduate school or more). To transform these categories into a continuous variable, I used the midpoint for the categories with intervals; I categorized those with some graduate education as having 18 years of schooling.

The second outcome was the natural log of the total family income in 1967. Second-generation respondents were asked to report total family income using the following categories: less than $2,500; $2,500–$4,999; $5,000–$7,499; $7,500–$9,999; $10,000–$14,999; $15,000–$19,999; $20,000–$29,999; and $30,000 or more. Here, too, I transformed the variable by assigning a midpoint to the categories with intervals. I categorized a family income of less than $2,500 as $1,250; I top-coded $30,000 or more as $30,000.

Key Zero- and First-Generation Premigration Variables

First-generation parents’ premigration background included their education and occupation in Japan. Education was operationalized according to the Japanese educational system in the Meiji and Taisho eras and was measured as the higher of the father's or mother's Japanese school type: (1) lower primary education (Jinjō-shōgakkō) or less; (2) upper primary education (Kōtō-shōgakkō); and (3) secondary education (e.g., Kyūsei Chūgakkō, Jogakkō, or Jitsugyō-gakkō) or higher (e.g., Senmon-gakkō, Kōtō-gakkō, or Daigaku).6 For the Japanese birth cohorts that include the first-generation parents in this study, only lower primary education (i.e., modern-day elementary schools), lasting four or six years depending on the birth year, was compulsory, although earlier cohorts likely had limited attendance. Opportunities for secondary education (i.e., modern-day senior high schools) or above were limited and are sometimes regarded as a marker of socioeconomic status (Kihara 2022; Tsuburai and Sato 1995; Yoshida 1985). For a likely explanation of the relatively high proportion of first-generation immigrant parents with secondary education in the JARP sample, see footnote 2.

Parental premigration occupation in Japan was operationalized as a father's occupation before migration: (1) professional/white collar, including managers and clerical and sales occupations; (2) farmers; (3) blue collar and others, including craftsmen, operatives, and military and domestic service occupations; and (4) not in the labor force, including minors and students who immigrated before starting a career. Grandparental occupation was categorized as (1) professional/white collar, (2) wealthy farmers, (3) average or poor farmers, and (4) blue collar or others. Farmers were deemed wealthy if their farms in Japan were as large as one chō (traditional Japanese unit of land measurement) or more.7

Finally, based on the sociology literature in Japan (e.g., Sonoda et al. 1995; Takeuchi 1975; Yasuda 1969), the family sociocultural background was measured with a dummy variable indicating samurai ancestry (shizoku status). A second-generation immigrant was deemed to have samurai ancestry if they had at least one zero-generation paternal or maternal grandfather who was shizoku.

Key First-Generation Postmigration Variables

First-generation parents’ postmigration socioeconomic status was measured by education acquired in the United States and their primary U.S. occupation. Education in the United States reflects the highest level achieved by either parent: (1) no U.S. education, (2) informal U.S. education only, and (3) formal U.S. education. Informal education included English language schools, trade schools, and citizenship classes. Most individuals with formal education in the United States attended elementary or high schools, and a few had U.S. college or graduate education.

The primary U.S. occupation was first-generation fathers’ principal profession. The variable assumed three values: (1) professional/white collar, (2) farmer, and (3) blue collar or other. Most fathers in white-collar occupations were urban proprietors (self-employed managers of small shops) in the West Coast states in the 1920s and 1930s (Bonacich and Modell 1980).

Other Variables: Demographic and Other Controls

First-generation-level controls8 included age (converted to age in 1967), age squared, gender,9 year-of-migration dummy variables, and a fixed effect based on the 47 Japanese prefectures of origin. Second-generation-level controls included age at the time of the survey (in 1967); age squared; birth order (dummy variable for eldest son or daughter); and a fixed effect based on 65 U.S. birth counties, some of which with small numbers were aggregated with neighboring counties or at the state level. The analysis of educational attainment (for second-generation men and women) included second-generation gender as an additional control. The analysis of family income (for men only) added controls for marital status, relative coresidence, and child status.

Statistical Analysis

To test the hypotheses, I estimated multilevel regression models predicting second-generation socioeconomic attainment (education and income) using their parental (first-generation) and grandparental (zero-generation) socioeconomic or cultural backgrounds with controls and weights. I incrementally added parental Japanese-origin prefecture and U.S. county fixed effects to the models. Including Japanese prefecture fixed effects enabled me to estimate intergenerational transfers net of ethnic community resources based on the parental Japanese-origin prefectures. Japanese immigrants in the United States hailed from different parts of Japan and created mutual aid associations based on their origin prefectures (ken) called kenjin-kai, which provided social and economic support (Daniels 1988:166; Light 1972: chaps. 3–4). Including U.S. county fixed effects enabled me to estimate the intergenerational mobility net of U.S. neighborhoods where the second-generation grew up. The U.S. counties differed in their reception contexts, as reflected in coethnic network density, the average schooling level, labor market conditions, and hostility toward immigrants (Connor 2020). The standard errors were cluster-robust at the family (first-generation) level (Cameron and Miller 2015:11).

Results

Summary Statistics

Table 1 presents the unweighted and weighted summary statistics for the variables used in the analyses. The same (grand)parents may have been counted multiple times in the zero- and first-generation statistics because the second-generation sample included many siblings. For data specific to individual parents (counting each parent only once), see Table B1 (online appendix). The distribution of Japanese-origin prefectures and U.S. birth counties is reported in the online appendix, section E.

Multilevel Regression Results

Tables 2 and 3 present coefficients from multilevel linear regression models with probability weights. Table 2 presents results predicting years of education using the full sample, and Table 3 presents results predicting the log of family income using the male sample. All models controlled for the variables described earlier. For both tables, Model 1 presents results predicting the outcome with only the zero-generations’ (grandfathers’) occupation, Model 2 adds family sociocultural background (zero-generation samurai ancestry), Model 3 adds first-generation premigration socioeconomic status (occupation and education in Japan), Model 4 adds first-generation postmigration socioeconomic status (U.S. occupation and education), Model 5 adds the fixed effects of the 47 origin prefectures at the first-generation level, and Model 6 adds the fixed effects of the 65 U.S. birth counties at the second-generation level. As a reference point, Model 7 predicts the socioeconomic outcome using first-generation parents’ postmigration socioeconomic status (U.S. occupation and education) without including premigration (grand)parental socioeconomic and cultural predictors. This last model provides a conventional result often found in analyses of immigrant social mobility that considers only destination-country measures of parental social status.

In addition, I tested the robustness of the findings to alternative specifications (online appendix, section D). Specifically, I explored the effects of unweighting my dataset and incorporating maternal grandparents’ occupations. The results were generally robust to these alternative approaches.

Educational Attainment

Second-generation educational attainment is positively linked to having grandparents who were in professional/white-collar occupations or wealthy farmers rather than average or poor farmers (Model 1 in Table 2). The advantages of grandparents being in professional/white-collar occupations are significantly reduced after I include the parents’ premigration Japanese education and occupation (Model 3). However, the advantages of grandparents being wealthy farmers persist even after I account for a full set of parents’ Japanese and U.S. education and occupation (Model 4) and after I include fixed effects for origin prefecture (Model 5) and U.S. birth county (Models 6). In a fully specified model, the predicted magnitude of education advantages from grandparents being wealthy farmers (vs. average or poor farmers) is 0.3 years (Model 6), partially supporting Hypothesis 1B.

Family sociocultural background (as measured by samurai ancestry) confers strong educational advantages to the second generation born in the United States, net of zero-generation occupation (Model 2), first-generation socioeconomic status (Models 3 and 4), Japanese prefecture fixed effects (Model 5), and U.S. county fixed effects (Models 6). In a full model (Model 6), the predicted magnitude of education advantage from having grandfathers of samurai ancestry is 0.6 years, supporting Hypothesis 2.

First-generation parents’ premigration socioeconomic status in Japan significantly predicts second-generation educational attainment. Specifically, having parents with upper primary or more education or who had been in a professional/white-collar occupation in Japan before migration confers persistent advantages (Model 3), even after I account for parents’ U.S. education and occupation (Model 4), parental origin prefectures (Model 5), and U.S. birth counties (Model 6). In Model 6, the predicted magnitude of education advantages from parents’ Japanese education is 0.3 for upper primary education and 0.7 years for secondary or more education. The magnitude of the education advantage of having fathers in professional/white-collar occupations is 0.6 years. These results strongly support Hypothesis 1A.

Consistent with theories of intergenerational mobility, parents also transmitted their postmigration U.S. education and occupation to their children. Those whose parents gained additional education in the United States are more educationally advantaged than others, and those with fathers in professional/white-collar or blue-collar occupations are more educationally advantaged than those whose parents were U.S. farmers.

Income

Second-generation income (sample restricted to men), as measured in 1967 in the United States, was positively linked to having grandparents who were in professional/white-collar occupations or wealthy farmers in Japan as opposed to average or poor farmers (Model 1 in Table 3). Although the advantages of having grandfathers in professional/white-collar occupations decline after I account for parents’ premigration education and occupation, income advantages conferred by grandparents being wealthy farmers in Japan are marginally significant in some models. In Model 6, the magnitude of advantage is approximately 9% (i.e., exp(0.089) – 1 ≈ 0.09), lending some credence to Hypothesis 1B. Mobility-oriented family cultural background, measured by samurai ancestry, also confers an income advantage. In Model 6, this advantage is approximately 14% (i.e., exp(0.130) – 1 ≈ 0.14), consistent with Hypothesis 2.

Parents’ premigration education and occupation also predict second-generation income. Having parents who had a Japanese upper primary or greater education (vs. lower primary education) and worked as professionals/white-collar or blue-collar workers (vs. farmers) confers persistent advantages (Model 2), even after I account for parents’ U.S. education or occupation (Model 4), origin prefecture (Model 5), and the second generation's birth county (Model 6). In Model 6, the predicted magnitudes of income advantages persisting from parents’ Japanese upper primary and secondary or more education are 14% and 12%, respectively. The premigration occupation advantage is 18% for professional/white-collar and blue-collar workers. These results uphold Hypothesis 1A.

First-generation postmigration U.S. education is positively associated with second-generation income at marginally significant levels when premigration socioeconomic status variables are not included (Model 7 in Table 2). However, in contrast to the analysis of second-generation educational attainment in Table 2, this analysis of second-generation income shows that first-generation fathers’ postmigration U.S. occupation does not confer advantages to their children (Model 7 in Table 3).

Discussion

This study's results reveal the long-term effects of premigration family background on the outcomes of second-generation immigrant children in the United States. Premigration family background includes grandparents’ and parents’ socioeconomic status in Japan measured as premigration occupation and education and mobility-oriented family culture proxied by samurai ancestry. These associations are substantially meaningful and independent of U.S.-based resources, such as postmigration parental socioeconomic status, origin community ties in the United States, or U.S. county–level resources. My results diverge from extant studies finding null or weak associations between European immigrant (grand)parents’ premigration status and their second-generation (grand)children's socioeconomic outcomes (Catron 2020; Connor 2020). Although providing a mechanism for such divergence is beyond the scope of this work, I note several possibilities.

One potential mechanism is the extremely negative reception context for first-generation parents in the United States. Japanese immigrant parents in the United States suffered from discrimination in all aspects of life, including labor market discrimination, exclusion from citizenship, landownership restrictions, and forced relocation (Broom and Riemer 1973; Daniels 1988). These experiences made it difficult for them to accumulate and transmit U.S.-based resources or take advantage of U.S.-based opportunities for their children that aligned with their human capital or socioeconomic status, perhaps increasing the importance of premigration resources. In fact, my results reveal that postmigration parental socioeconomic status (as measured by U.S. education and occupation) might not be as crucial as their premigration status (measured by Japanese education and occupation). Although both pre- and postmigration parental occupational statuses confer educational advantages to second-generation children, the analyses of second-generation income showed that such advantages are conferred only by the premigration Japanese education and occupation. In terms of postmigration occupation, I found no advantage of parents being in professional/white-collar occupations over farmers: the coefficient is slightly negative (but statistically insignificant). Such counterintuitive results might be driven by the first-generation parents in white-collar occupations (e.g., small business owners in 1930s Los Angeles) experiencing relatively greater hardships because of financial losses during internment relative to farmers (see also Arellano-Bover 2022:163–166).

Another possible explanation for the divergence of my results from those involving European immigrants is the higher data quality and measurement precision of the JARP survey, which likely enabled a more accurate capturing of multigenerational/intergenerational associations. As a result of its research design,10 the JARP survey likely has significantly fewer missed or incorrect linkages between parents and children than census linking, which typically relies on name, age, and birthplace and achieves match rates of less than 30% (e.g., Abramitzky et al. 2021). Further, the JARP survey's socioeconomic measurements are more refined and tailored toward the immigrant population, whereas census and ship records (Catron 2020; Connor 2020) offer only basic measures of premigration status and are more susceptible to enumeration and transcription errors. These conditions could render previous studies on European immigrants more prone to attenuation bias under classical measurement error assumptions—a possibility Ward (2020, 2021) explored extensively.

Regarding multigenerational associations between grandparents (the zero generation) and grandchildren (the second generation), the results suggest that grandparents’ occupational status persists among the second generation but is mostly transmitted indirectly via first-generation parents. I draw this conclusion from the observed attenuation of the coefficients for professional/white-collar grandfathers in Japan (Model 2 vs. Model 3, Tables 2 and 3).

Yet, having wealthy farmer grandparents might also have offered a direct advantage (Model 6 in Tables 2 and 3). The multigenerational mobility literature (e.g., Mare 2011) found the direct effects of grandparents’ socioeconomic status on grandchildren among some populations, especially when there is an opportunity for direct contact between them (e.g., Song and Mare 2019; Zeng and Xie 2014). In the 1930s, the formative period for most U.S.-born Japanese Americans, multigenerational contact was not uncommon. For instance, it was a popular practice for Japanese immigrant parents to send their U.S.-born second-generation children to Japan for a few years to live with their grandparents or other relatives, enabling the children to learn the Japanese language and culture (Azuma 2005: chap. 6; Kitano 1976:159–160).

However, the omission or measurement errors in parental education or occupation variables could plausibly inflate the multigenerational association, a methodological concern increasingly recognized in multigenerational mobility research (Anderson et al. 2018; Modalsli and Vosters forthcoming;,Solon 2018). Thus, caution is needed in the interpretation of the multigenerational coefficient. In any scenario, my research results stress the importance of incorporating premigration family socioeconomic history measures into the analysis of the social mobility of immigrant populations.

The persistent advantages for individuals of samurai ancestry, who constitute approximately 14% of the second-generation Japanese American population, highlight the importance of intangible and cross-nationally transmissible cultural values in explaining second-generation children's socioeconomic attainment. The samurai ancestry advantage remains even after I account for immigrant parents’ and grandparents’ occupations and education. Notably, the magnitude of this educational and income advantage is large and meaningful.

Japanese sociologists (e.g., Yasuda 1969, 1971) have maintained that samurai descendants embodied upward mobility–oriented values and transmitted them to their children and grandchildren. Here, it is crucial to clarify that I am not suggesting that such cultural values are exclusive to the descendants of the samurai. Instead, I propose that descendants of samurai were one of the groups more likely to hold stronger mobility-oriented values: the Japanese sociology literature suggests that such mobility-oriented values gradually spread from the ex-samurai to the masses (Takeuchi 1975, 1981; Yasuda 1969, 1971). These mobility-oriented values also helped facilitate the migration of some Japanese youths through popular publications (Azuma 2005:24–25; Sawada 1996: chap. 5; Tachikawa 1986). It is possible that even non-samurai Japanese descendants held stronger upward mobility–oriented values than other ethnic groups in the United States, contributing to the second-generation's remarkable upward mobility (Hirschman and Wong 1986; Sakamoto et al. 2012; Varon 1967).

The finding on samurai ancestry advantage aligns with studies maintaining that certain types of cultural values might facilitate mobility (Chen and Stevenson 1995; Hsin and Xie 2014; Liu and Xie 2016; Sakamoto et al. 2009, 2012; Tian 2023). Although most studies have utilized variations between ethnic groups, such as between White and Japanese individuals, my results support the importance of cultural values using within-group cultural variations among the Japanese, with samurai ancestry as an observable marker.

The ongoing debate about whether the immigrant culture is a function of a broader ethnic socioeconomic structure (e.g., Hsin and Xie 2014; Kim and Kim 2023; Lee and Zhou 2015; Sakamoto 2017; Zhou and Lee 2014) is beyond the purview of my study because I use data from a single ethnic group. However, the advantage associated with samurai ancestry remains positive after I account for (grand)parental socioeconomic status, parental origin prefectures, and U.S. counties. Furthermore, as Tian (2023: chap. 1) highlighted and as shown in Figure 1, first-generation Japanese immigrant parents in the early twentieth century were not “hyper-selected” because they did not have educational advantage over their U.S.-born White counterparts, although they were more educated than their nonmigrant peers in Japan (positive educational selection). Yet, their second-generation descendants consistently outperformed in education relative to native-born Whites (Hirschman and Wong 1986). These observations hint that some of Lee and Zhou's (2015) propositions might need reevaluation, although it would be unwise to discount the role of ethnic capital.

Including county-of-birth fixed effects (Model 6) hardly changed the key coefficients of premigration family socioeconomic and cultural background. This outcome diverges from extant literature underscoring the role of U.S. neighborhoods in mediating or moderating the relationship between parents’ and children's backgrounds (e.g., Abramitzky et al. 2021; Catron 2020; Connor 2020). More than 85% of the sample was raised in the West Coast states of California, Washington, and Oregon. Consequently, the sample did not exhibit large variations in terms of U.S. states of birth. The West Coast states are known to have been conducive to social mobility because of their early adoption of compulsory schooling (Connor 2020). Second-generation Japanese Americans, in general, might thus have benefited from their parents’ choice of settling in the West relative to other immigrant and native-born groups; however, the analysis does not capture this effect because of the research design of studying a single immigrant group and their children. Additionally, had the initial Japanese settlement been in less mobility-favoring regions, the impact of premigration family socioeconomic and sociocultural backgrounds might have been more pronounced because of the fewer available resources in the United States.

Conclusion

Do second-generation immigrants draw from the family's socioeconomic and sociocultural resources from the parental country of origin? If so, to what extent? By incorporating an origin-country perspective into the study of immigrant social mobility, this study examined the case of the Japanese population in the United States in the early twentieth century. While facing extremely negative socioeconomic contexts, the second generation experienced strong socioeconomic upward mobility. The analyses found strong and enduring effects of their Japanese immigrant parents’ premigration socioeconomic status and cultural backgrounds that facilitated second-generation upward mobility. These effects may extend back to their zero-generation grandparents and possibly contrast with the European second-generation immigrants, who typically experienced upward mobility regardless of their family background because of relatively favorable reception contexts.

Note that this study provided only a partial account of Japanese upward mobility. The study relied solely on data on Japanese individuals, without comparable data on native-born Whites or other immigrant groups. In addition, it focused on family- or individual-level socioeconomic resources and a specific marker of cultural background. Therefore, I could not explore additional mechanisms, especially those operating at scales beyond the family, such as at the ethnic group level (e.g., cultural practices or ethnic capital shared by all Japanese individuals) or geographic region level (e.g., economic development in the West Coast states). Furthermore, the study's methods do not permit conclusions about the causality of the associations.

Nevertheless, this study highlights the importance of employing historical and origin-country perspectives. Understanding immigrant social stratification hinges on grounding migrants in their origin societies’ contexts and culture, considering the premigration socioeconomic statuses of migrants, and examining how these factors have historically interacted with the host society contexts. Future research should explore different immigrant populations. In the United States, it is particularly important to study the historical incorporation of non-White groups, such as Chinese, Filipino, Mexican, and Caribbean immigrants and their descendants. Future studies should examine immigrants’ attainment in the context of origin countries; variations in cultural social mobility norms; emigration processes; and the legal, political, social, and economic conditions immigrants have historically faced.

Acknowledgments

This study received funding from the JSPS KAKENHI (grant JP23KJ0338), with additional logistical and research support from Keio University, the Institute of Social Science at the University of Tokyo, and the Population Studies and Training Center at Brown University, which receives funding from the Eunice Kennedy Shriver National Institute of Child Health and Human Development for training support (T32 HD007338) and general support (P2C HD041020). This article is a significantly revised version of a chapter from my dissertation (“International Migration and Social Mobility Across the Pacific: A Historical Analysis of the Japanese Population in the Continental United States”), which I completed at Brown University in 2023. Early versions of this article and related papers were presented at the Yale Project on Japan's Politics and Diplomacy Series at the Council on East Asian Studies, Yale McMillan Center (2022); the spring 2022 meeting of the Research Committee 28 of the International Sociological Association in London, United Kingdom; and the 2022 annual meeting of the Population Association of America in Atlanta, Georgia, among others. I am grateful for the insightful comments provided by the participants of these research meetings and by the four external reviewers, as well as by Margot Jackson, Michael White, Emily Rauscher, John Logan, and Kaoru Sato. I thank Sage Author Services and the Demography editorial team for their excellent English language editing and proofreading.

Notes

1

In the late nineteenth and early twentieth centuries, Japanese immigrants were mainly living in Hawaii and West Coast states. However, because my main analysis data are from the JARP Survey, which excluded Hawaii, I focus on the continental United States.

2

The Japanese government at the time, mindful of its national image in Western countries, coupled with pressure from the United States, endeavored to restrict emigration to those with the financial means to maintain the “quality” of its emigrants, particularly in the continental United States (Daniels 1988: chap. 4; Sawada 1991; Wakatsuki 1979:475–476). These policies, together with selective return migration, led to a form of positive socioeconomic selection where a higher proportion of immigrants had at least some Japanese secondary education than nonmigrants in Japan (Kihara 2022; Suzuki 1995), an opportunity more common to those from the middle or higher social strata (Tsuburai and Sato 1995; Yoshida 1985). This selection was nonlinear and likely favored individuals from the middle to upper middle of the Japanese socioeconomic strata rather than from the very top (Kihara 2022:2595).

3

The JARP also surveyed the third generation (i.e., children of the second generation), but I excluded them from my analysis because most were too young at the time of the survey to have completed education or have income. See Minamikawa (2007: chap. 7) for a critical historical overview of the JARP. For details on survey implementation, see Levine (2006: documentation files) and Levine and Rhodes (1981: chap. 3).

4

From the original JARP second-generation sample (n = 2,304), I excluded those who were not U.S.-born (n = 58), were younger than 25 at the time of the second-generation survey (n = 105), and had missing or invalid responses on at least one item (n = 69).

5

See the online appendix (section A) for further discussion on the representativeness of the first-generation survey.

6

The literal translations of Jinjō-shōgakkō and Kōtō-shōgakkō are “ordinary” and “higher” elementary schools, respectively. However, for readers’ convenience, I translate these terms to “lower” and “upper primary education.” I did not distinguish between secondary and tertiary education because it is not relevant for the first-generation parents in this study who were born in the 1860s and 1870s or who migrated before age 18. Among parents in the secondary or more category, 86% have Japanese secondary and 14% have Japanese tertiary education. Education for zero-generation grandparents is unavailable in the JARP, likely because such inquiries were irrelevant: formal education began to take form in the 1870s, when most grandparents were already adults.

7

In the survey, the first-generation immigrants whose fathers were farmers were asked about the size of their fathers’ farms: (1) “Poor” (less than 5 tan), (2) “Average” (5 but less than 10 tan), (3) “Above average” (1 chō but less than 5), (4) “Well-to-do” (5 chō but less than 10), or (5) “Jinushi” (landowner, more than 10 chō) (Levine 2006:37). In Japanese, one tan corresponds to approximately 1,000 m2, and one chō is approximately 10,000 m2 (equivalent to about 2.5 acres). Respondents in the latter three categories were considered as having fathers who were wealthy farmers. Those in the first two categories or with missing responses on the father’s farm size question (but whose fathers were farmers) were assumed to have had fathers who were average or poor farmers. The survey did not ask about farm sizes of first-generation immigrants themselves, most likely because many were too young to have owned farms at the time of migration. The operationalization for grandparental occupation preferred the paternal grandfather’s over the maternal grandfather’s occupation. When the information for the paternal grandfather was missing, the maternal grandfather’s information was used.

8

For demographic controls, father’s information was preferred over mother’s information (e.g., first-generation father’s age rather than mother’s age). When father’s information was missing, mother’s information was used.

9

In the first-generation survey, the JARP researchers approached the first-generation family and interviewed the husband about both him and his wife. When the husband was not available (e.g., deceased), the wife was asked about her and her husband (Levine and Rhodes 1981:18–19).

10

The JARP study was designed specifically for Japanese American families, and the two generations can be linked by family ID. Furthermore, unlike the real-time data in censuses or ship manifests, which are often used in studies on European immigrants (e.g., Catron 2020; Connor 2020), the JARP survey asked the immigrants to recall their “last important jobs before leaving Japan” and their father’s “principal occupation.” Although retrospective measures are potentially subject to errors, extant studies have shown that factual information such as occupation is less prone to recall inaccuracies (Bielby et al. 1977; Torche 2015). The JARP survey measures are more theoretically relevant and valid for my study. For example, an individual with a decade of work experience in Japan, officially recorded as a coal miner during a census year but primarily a farmer for nine years before migrating, should be considered a farmer. Thus, a retrospective question on “primary occupation” before migration aligns more closely with the study’s objectives. This distinction is analogous to that made in the historical income mobility literature between average and current income (e.g., Ward 2020, 2021).

References

Abramitzky, R., & Boustan, L. (
2017
).
Immigration in American economic history
.
Journal of Economic Literature
,
55
,
1311
1345
.
Abramitzky, R., Boustan, L., Catron, P., Connor, D., & Voigt, R. (
2023
).
The refugee advantage: English-language attainment in the early twentieth century
.
Sociological Science
,
10
,
769
805
.
Abramitzky, R., Boustan, L. P., & Eriksson, K. (
2012
).
Europe's tired, poor, huddled masses: Self-selection and economic outcomes in the age of mass migration
.
American Economic Review
,
102
,
1832
1856
.
Abramitzky, R., Boustan, L. P., Jacome, E., & Perez, S. (
2021
).
Intergenerational mobility of immigrants in the United States over two centuries
.
American Economic Review
,
111
,
580
608
.
Akresh, I. R. (
2008
).
Occupational trajectories of legal US immigrants: Downgrading and recovery
.
Population and Development Review
,
34
,
435
456
.
Alba, R., & Nee, V. (
2003
).
Remaking the American mainstream: Assimilation and contemporary immigration
.
Cambridge, MA
:
Harvard University Press
.
Anderson, L. R., Sheppard, P., & Monden, C. W. S. (
2018
).
Grandparent effects on educational outcomes: A systematic review
.
Sociological Science
,
5
,
114
142
.
Arellano-Bover, J. (
2022
).
Displacement, diversity, and mobility: Career impacts of Japanese American internment
.
Journal of Economic History
,
82
,
126
174
.
Azuma, E. (
2005
).
Between two empires: Race, history, and transnationalism in Japanese America
.
New York, NY
:
Oxford University Press
.
Bielby, W. T., Hauser, R. M., & Featherman, D. L. (
1977
).
Response errors of Black and non-Black males in models of the intergenerational transmission of socioeconomic status
.
American Journal of Sociology
,
82
,
1242
1288
.
Blau, P. M., & Duncan, O. D. (
1967
).
The American occupational structure
.
New York, NY
:
John Wiley & Sons
.
Bonacich, E., & Modell, J. (
1980
).
The economic basis of ethnic solidarity: Small business in the Japanese American community
.
Berkeley
:
University of California Press
.
Broom, L., & Riemer, R. (
1973
).
University of California publications in culture and society: Vol. 4. Removal and return: The socio-economic effects of the war on Japanese Americans
.
Berkeley
:
University of California Press
.
Brunori, C., Luijkx, R., & Triventi, M. (
2020
).
Immigrants’ selectivity and their socio-economic outcomes in the destination country: The Italian case
.
Population, Space and Place
,
26
,
e2352
. https://doi.org/10.1002/psp.2352
Buck, J. H. (
1973
).
The Satsuma Rebellion of 1877: From Kagoshima through the siege of Kumamoto Castle
.
Monumenta Nipponica
,
28
,
427
446
.
Cameron, C. A., & Miller, D. L. (
2015
).
A practitioner's guide to cluster-robust inference
.
Journal of Human Resources
,
50
,
317
372
.
Catron, P. (
2019
).
The citizenship advantage: Immigrant socioeconomic attainment in the age of mass migration
.
American Journal of Sociology
,
124
,
999
1042
.
Catron, P. (
2020
).
The melting-pot problem? The persistence and convergence of premigration socioeconomic status during the age of mass migration
.
Social Forces
,
99
,
366
397
.
Chen, C., & Stevenson, H. W. (
1995
).
Motivation and mathematics achievement: A comparative study of Asian American, Caucasian American, and East Asian high school students
.
Child Development
,
66
,
1215
1234
.
Connor, D. S. (
2019
).
The cream of the crop? Geography, networks, and Irish migrant selection in the age of mass migration
.
Journal of Economic History
,
79
,
139
175
.
Connor, D. S. (
2020
).
Class background, reception context, and intergenerational mobility: A record linkage and surname analysis of the children of Irish immigrants
.
International Migration Review
,
54
,
4
34
.
Daniels, R. (
1988
).
Asian America: Chinese and Japanese in the United States since 1850
.
Seattle
:
University of Washington Press
.
Drouhot, L. G., & Nee, V. (
2019
).
Assimilation and the second generation in Europe and America: Blending and segregating social dynamics between immigrants and natives
.
Annual Review of Sociology
,
45
,
177
199
.
Feliciano, C. (
2005
).
Educational selectivity in U.S. immigration: How do immigrants compare to those left behind?
Demography
,
42
,
131
152
.
Feliciano, C. (
2020
).
Immigrant selectivity effects on health, labor market, and educational outcomes
.
Annual Review of Sociology
,
46
,
315
334
.
Feliciano, C., & Lanuza, Y. R. (
2017
).
An immigrant paradox? Contextual attainment and intergenerational educational mobility
.
American Sociological Review
,
82
,
211
241
.
Fishman, S. H. (
2020
).
Educational mobility among the children of Asian American immigrants
.
American Journal of Sociology
,
126
,
260
317
.
Harootunian, H. D. (
1959
).
The progress of Japan and the samurai class, 1868–1882
.
Pacific Historical Review
,
28
,
255
266
.
Harootunian, H. D. (
1960
).
The economic rehabilitation of the samurai in the early Meiji period
.
Journal of Asian Studies
,
19
,
433
444
.
Hatton, T. J., & Williamson, J. G. (
1998
).
The age of mass migration: Causes and economic impact
.
New York, NY
:
Oxford University Press
.
Hirschman, C., & Wong, M. G. (
1984
).
Socioeconomic gains of Asian Americans, Blacks, and Hispanics: 1960–1976
.
American Journal of Sociology
,
90
,
584
607
.
Hirschman, C., & Wong, M. G. (
1986
).
The extraordinary educational attainment of Asian Americans: A search for historical evidence and explanations
.
Social Forces
,
65
,
1
27
.
Hsin, A., & Xie, Y. (
2014
).
Explaining Asian Americans’ academic advantage over Whites
.
Proceedings of the National Academy of Sciences
,
111
,
8416
8421
.
Ichihashi, Y. (
1932
).
Japanese in the United States
.
Stanford, CA
:
Stanford University Press
.
Ichioka, Y. (
1988
).
The Issei: The world of the first generation Japanese immigrants, 1885–1924
.
New York, NY
:
Free Press
.
Ichou, M. (
2014
).
Who they were there: Immigrants’ educational selectivity and their children's educational attainment
.
European Sociological Review
,
30
,
750
765
.
Kihara, T. (
2022
).
Socioeconomic selectivity of Japanese migration to the continental United States during the age of mass migration
.
Journal of Ethnic and Migration Studies
,
48
,
2577
2600
.
Kihara, T. (
2023
).
International migration and social mobility across the Pacific: A historical analysis of the Japanese population in the continental United States
(Doctoral dissertation). Department of Sociology, Brown University, Providence, RI. Retrieved from https://repository.library.brown.edu/studio/item/bdr:nkm33mv2/
Kim, C., & Kim, A. T. (
2023
).
Is hyper-selectivity a root of Asian American children's success?
Social Science Research
,
113
,
102886
. https://doi.org/10.1016/j.ssresearch.2023.102886
Kinmonth, E. H. (
1982
).
The self-made man in Meiji Japanese thought: From samurai to salary man
(1st ed.).
Berkeley
:
University of California Press
.
Kitano, H. H. L. (
1962
).
Changing achievement patterns of the Japanese in the United States
.
Journal of Social Psychology
,
58
,
257
264
.
Kitano, H. H. L. (
1976
).
Japanese Americans: The evolution of a subculture
.
Englewood Cliffs, NJ
:
Prentice-Hall
.
Lee, J., & Zhou, M. (
2015
).
The Asian American achievement paradox
.
New York, NY
:
Russell Sage Foundation
.
Levine, G. N. (
2006
).
Japanese-American Research Project (JARP): A three-generation study, 1890–1966
(Version 2) [Dataset, documentation file, and codebook]. Inter-university Consortium for Political and Social Research [distributor]. https://doi.org/10.3886/ICPSR08450.v2
Levine, G. N., & Montero, D. M. (
1973
).
Socioeconomic mobility among three generations of Japanese Americans
.
Journal of Social Issues
,
29
(
2
),
33
48
.
Levine, G. N., & Rhodes, C. (
1981
).
The Japanese American community: A three-generation study
.
New York, NY
:
Praeger Publishers
.
Light, I. (
1972
).
Ethnic enterprise in America: Business and welfare among Chinese, Japanese, and Blacks
(2nd ed.).
Berkeley
:
University of California Press
.
Liu, A., & Xie, Y. (
2016
).
Why do Asian Americans academically outperform Whites? The cultural explanation revisited
.
Social Science Research
,
58
,
210
226
.
Lowrey, K., Van Hook, J., Bachmeier, J. D., & Foster, T. B. (
2021
).
Leapfrogging the melting pot? European immigrants’ intergenerational mobility across the twentieth century
.
Sociological Science
,
8
,
480
512
.
Luthra, R. R., & Soehl, T. (
2015
).
From parent to child? Transmission of educational attainment within immigrant families: Methodological considerations
.
Demography
,
52
,
543
567
.
Luthra, R. R., Waldinger, R., & Soehl, T. (
2018
).
Origins and destinations: The making of the second generation
.
New York, NY
:
Russell Sage Foundation
.
Mare, R. D. (
2011
).
A multigenerational view of inequality
.
Demography
,
48
,
1
23
.
Minamikawa, F. (
2007
).
“Nikkei Amerika-Jin” no rekishi shakaigaku
[Historical sociology of “Japanese Americans”].
Tokyo, Japan
:
Sairyusya
.
Modalsli, J., & Vosters, K. (forthcoming). Spillover bias in multigenerational income regressions.
Journal of Human Resources
. Advance online publication. https://doi.org/10.3368/jhr.0520-10919R2
Montero, D. (
1980
).
Japanese Americans: Changing patterns of ethnic affiliation over three generations
.
Boulder, CO
:
Westview Press
.
Ochiai, H. (
2015
).
Chitsuroku shobun: Meiji ishin to buke no kaitai
[Abolition of hereditary stipend: Meiji restoration and dissolution of samurai families].
Tokyo, Japan
:
Kodansha Gakujyutsu Bunko
.
Polavieja, J. G. (
2015
).
Capturing culture: A new method to estimate exogenous cultural effects using migrant populations
.
American Sociological Review
,
80
,
166
191
.
Pong, S., & Landale, N. S. (
2012
).
Academic achievement of legal immigrants’ children: The roles of parents’ pre- and postmigration characteristics in origin-group differences
.
Child Development
,
83
,
1543
1559
.
Potochnick, S., & Hall, M. (
2021
).
U.S. occupational mobility of children of immigrants based on parents’ origin-country occupation
.
Demography
,
58
,
219
245
.
Ruggles, S., Flood, S., Goeken, R., Schouweiler, M., & Sobek, M. (
2022
).
IPUMS USA (Version 12.0)
[Dataset].
Minneapolis, MN
:
IPUMS
. https://doi.org/10.18128/D010.V12.0
Sakamoto, A. (
2017
).
Book review: “The Asian American achievement paradox” by Jennifer Lee and Min Zhou
.
American Journal of Sociology
,
122
,
2011
2013
.
Sakamoto, A., Goyette, K. A., & Kim, C. (
2009
).
Socioeconomic attainments of Asian Americans
.
Annual Review of Sociology
,
35
,
255
276
.
Sakamoto, A, Kim, C., & Takei, I. (
2012
).
The Japanese American family
. In Wright, R., Mindel, C. H., Habenstein, R. W., & Van Tran, T. (Eds.),
Ethnic families in America: Patterns and variations
(pp.
252
276
).
New York, NY
:
Pearson-Prentice Hall
.
Sawada, M. (
1991
).
Culprits and gentlemen: Meiji Japan's restrictions of emigrants to the United States, 1891–1909
.
Pacific Historical Review
,
60
,
339
359
.
Sawada, M. (
1996
).
Tokyo life, New York dreams: Urban Japanese visions of America, 1890–1924
.
Berkeley
:
University of California Press
.
Solon, G. (
2018
).
What do we know so far about multigenerational mobility?
Economic Journal
,
128
(
612
),
F340
F352
.
Song, X., & Mare, R. D. (
2019
).
Shared lifetimes, multigenerational exposure, and educational mobility
.
Demography
,
56
,
891
916
.
Sonoda, H., Hamana, A., & Hirota, T. (
1995
).
Shizoku no rekishi shakaigakuteki kenkyū: Bushi no kindai
[Historical sociology of the Shizoku: Modernity of the samurai class].
Nagoya
,
Japan: Nagoya University Press
.
Spitzer, Y., & Zimran, A. (
2018
).
Migrant self-selection: Anthropometric evidence from the mass migration of Italians to the United States, 1907–1925
.
Journal of Development Economics
,
134
,
226
247
.
Suzuki, M. (
1995
).
Success story? Japanese immigrant economic achievement and return migration, 1920–1930
.
Journal of Economic History
,
55
,
889
901
.
Tachikawa, K. (
1986
).
Meiji-kōhanki no tobei netsu–Amerika no ryūkō [“To-Bei” (going to America) boom in late Meiji era: American fever]
.
Shirin
,
69
,
383
417
.
Takeuchi, Y. (
1975
).
Rissin-Shusse shugi no keifu to ronri: Meiji-ki wo chūsin-ni [The change and the structure of the notion ‘Rissin-Shusse’ in Meiji, Japan]
.
Kansai University Departmental Bulletin Paper
,
7
(
1
),
33
49
.
Takeuchi, Y. (
1981
).
Kyosō no shakaigaku: Gakureki to shō-shin
[Sociology of competition: Credentials and promotion.].
Kyoto, Japan
:
Sekaishiso-Sya
.
Tian, Z. (
2023
).
Three essays on the educational investments of Asian American families
(Doctoral dissertation). Sociology Department, Princeton University, Princeton, NJ. Retrieved from http://arks.princeton.edu/ark:/88435/dsp01db78tg26x
Torche, F. (
2015
).
Analyses of intergenerational mobility: An interdisciplinary review
.
Annals of the American Academy of Political and Social Science
,
657
,
37
62
.
Tsuburai, K., & Sato, T. (
1995
).
Senkanki Nihon ni okeru shokugyo to gakureki: 1896–1925 nen shussei kohōto ni miru kyūsei chūtō kyōiku no shakai-teki ichi [Occupation and educational level in interwar Japan: Social state of secondary education under old system according to 1896–1925 birth cohort]
.
Journal of Educational Sociology
,
56
,
81
98
.
Varon, B. F. (
1967
).
The Japanese Americans: Comparative occupational status, 1960 and 1950
.
Demography
,
4
,
809
819
.
Wakatsuki, Y. (
1979
).
Japanese emigration to the United States, 1866–1924: A monograph
.
Perspectives in American History
,
12
,
387
516
.
Ward, Z. (
2020
).
The not-so-hot melting pot: The persistence of outcomes for descendants of the age of mass migration
.
American Economic Journal: Applied Economics
,
12
(
4
),
73
102
.
Ward, Z. (
2021
).
Intergenerational mobility in American history: Accounting for race and measurement error
(NBER Working Paper 29256).
Cambridge, MA
:
National Bureau of Economic Research
. https://doi.org/10.3386/w29256
White, M. J., & Mullen, E. J. (
2016
).
Socioeconomic attainment in the Ellis Island era
.
Social Science History
,
40
,
147
181
.
Wong, P.. (
1982
).
Issei, Nisei, and Sansei: Three generations of Japanese Americans
.
Contemporary Sociology
,
11
,
267
273
.
Yasuda, S. (
1969
).
Shizoku to shakai-idō [Samurai descendants and social mobility]
.
Japanese Sociological Review
,
19
(
4
),
21
40
.
Yasuda, S. (
1971
).
Shakai-idō no kenkyū
[A study of social mobility] (2nd ed.).
Tokyo, Japan
:
University of Tokyo Press
.
Yoshida, A. (
1985
).
Meiji Taishōki no chīkisyakai ni okeru chūtō kyōiku no shakai-teki-kinō [The social function of secondary education in the local society during the Meiji and Taisho period]
.
Journal of Educational Sociology
,
40
,
150
164
.
Zeng, Z., & Xie, Y. (
2004
).
Asian-Americans’ earnings disadvantage reexamined: The role of place of education
.
American Journal of Sociology
,
109
,
1075
1108
.
Zeng, Z., & Xie, Y. (
2014
).
The effects of grandparents on children's schooling: Evidence from rural China
.
Demography
,
51
,
599
617
.
Zhou, M., & Lee, J. (
2014
).
Assessing what is cultural about Asian Americans’ academic advantage
.
Proceedings of the National Academy of Sciences
,
111
,
8321
8322
.
Zhou, M., & Lee, J. (
2017
).
Hyper-selectivity and the remaking of culture: Understanding the Asian American achievement paradox
.
Asian American Journal of Psychology
,
8
,
7
15
.

Supplementary data