Abstract
This article spotlights a widespread problem in research and practice: Asians are commonly categorized as a monolithic group in the United States. Regarding research, my 24-year archival analysis of Psychological Science shows that most U.S. studies did not specify which Asian subgroup(s) were examined. Regarding practice, my analysis of the diversity, equity, and inclusion (DEI) webpages and latest diversity reports of S&P 100 companies finds that none of them differentiated between Asian subgroups. Such use of the generic category “Asian” is problematic because it masks important differences among Asian subgroups: (a) Of all ethnic groups in the United States, socioeconomic inequality among Asian subgroups is the highest and fastest growing; (b) U.S. studies show that East Asians (e.g., ethnic Chinese)—but not South Asians (e.g., ethnic Indians)—experience a “bamboo ceiling” in consequential contexts, including leadership attainment, academic performance in law and business schools, and starting salaries. Thus, lumping Asians together can obscure the challenges faced by certain Asian subgroups and jeopardize the attention and resources they need. More broadly, this article demonstrates the importance of differentiating between ethnic subgroups in research (e.g., theorization, surveys, and data analysis) and practice (e.g., diversity reports) to foster DEI.
Asians are the fastest growing ethnic group in the United States, increasing from 2.8% of the U.S. population in 1990 to about 7% in 2021 (Budiman & Ruiz, 2021; U.S. Census Bureau, 2002). The U.S. Census Bureau (2002) defines “Asian” as “people having origins in any of the original peoples of the Far East, Southeast Asia, or the Indian subcontinent (for example, Cambodia, China, India, Japan, Korea, Malaysia, Pakistan, the Philippine Islands, Thailand, and Vietnam)” (p. 1). 1 As such, Asians in the United States comprise diverse subgroups, including East Asians (EAs; e.g., ethnic Chinese, Japanese), South Asians (SAs; e.g., ethnic Indians, Pakistanis), and Southeast Asians (SEAs; e.g., ethnic Filipinos, Vietnamese). These groups are categorized as “Asian” largely because they hail from the massive continent of Asia—regardless of their physical and cultural dissimilarities (Lu et al., 2020, 2022; Yao et al., 2017). The categorization is also political, as evidenced by the 1923 Supreme Court case United States v. Bhagat Singh Thind, which “classified Indians as part of the ‘Asiatic stock,’ thereby making them ineligible for naturalization” (J. Lee & Ramakrishnan, 2020, p. 1735). In other words, the category “Asian” has a shaky foundation to begin with.
In this article, I first present original analyses from both academic and business contexts to illustrate that Asians in the United States are commonly homogenized in research and practice. Next, I explain that the generic label “Asian” can result in scientific misunderstandings by masking important differences among Asian subgroups in (1) socioeconomic status and (2) social-psychological outcomes. In particular, I review studies showing that EAs—but not SAs—tend to experience a “bamboo ceiling,” which broadly refers to challenges that impede Asians’ advancement in the United States (Hyun, 2005). Finally, I conclude with action items and future directions.
Asians Are Homogenized in Research and Practice
Most U.S. studies do not specify which Asian subgroup(s) are studied
To examine how Asians are represented in psychology studies, I performed a 24-year archival analysis of Psychological Science, the flagship journal of the Association for Psychological Science. I conducted a literature search in PsycInfo for articles published between 1990 (inaugural year) and 2023 whose titles, abstracts, or keywords contained the word “Asian” or specific Asian subgroups (e.g., “Chinese,” “Vietnamese”). 2 After excluding commentaries and corrigenda, I arrived at 79 U.S. studies (in 29 articles) that involved Asian participants or stimuli.
Among the 54 U.S. studies that involved Asian participants, as many as 57% used “Asians” ambiguously without specifying which Asian subgroups were examined. For example, Johnson and Wilson (2019) wrote: “Our sample consisted of 106 Asian raters” (p. 557)—even though different Asian subgroups might provide significantly different ratings. Similarly, among the 25 U.S. studies that involved Asian stimuli (e.g., vignettes, photos, target groups rated by participants), as many as 76% used “Asians” ambiguously without subgroup specification.
Most U.S. organizations aggregate Asians as a monolith
Likewise, U.S. organizations routinely aggregate Asians as a monolithic group. In a March 2024 analysis, two research assistants coded the websites and latest diversity reports of the S&P 100 companies and found that these large-cap U.S. companies all lumped Asian subgroups together. Although Microsoft recently “expanded the options for Asian employees in the United States to identify their backgrounds in additional detail” (Microsoft, 2023, p. 22), it did not differentiate between Asian subgroups when presenting data on organizational issues such as leadership representation. Instead, Microsoft (2023) merely stated that “Asian representation rose year over year at all leadership levels” (p. 16).
Why “Asian” Is a Problematic Category
Having illustrated the widespread practice of aggregating Asians in academic and business contexts, I next demonstrate that this practice is problematic because it can mask important differences among Asian subgroups in socioeconomic status and social-psychological outcomes.
The category “Asian” masks socioeconomic inequality among Asian subgroups
In the United States, Asians are often stereotyped as the “model minority” (J. Lee et al., 2024); as a whole, they have the highest educational attainment and median income of all ethnicities (Budiman & Ruiz, 2021). For example, according to the American Community Survey, the 2022 per capita income was $53,030 for Asians and $47,943 for Whites (U.S. Census Bureau, 2022).
However, of all ethnic groups in the United States, socioeconomic inequality among Asian subgroups is the highest and fastest growing (Kochhar & Cilluffo, 2018). Whereas some groups emigrate from war-torn countries with little education, other wealthier groups self-fund their higher education in the United States and then obtain citizenship via employment, and even wealthier groups immigrate via monetary investment in the United States (e.g., the EB-5 Immigrant Investor Program). Compared with other Asian subgroups, SEAs from countries such as Cambodia, Laos, and Myanmar tend to fare worse economically and educationally. For example, the 2022 per capita income was far lower for ethnic Burmese ($23,715) than for ethnic Indians ($69,440; U.S. Census Bureau, 2022). This disparity is unsurprising given that among adults aged 25 and older, only 22% of ethnic Burmese had a bachelor’s degree or higher versus 78% of ethnic Indians (U.S. Census Bureau, 2022). These striking contrasts within the Asian population suggest that lumping all Asians together can produce scientific misunderstandings. When Americans view “Asians” as a monolith, they may erroneously assume that all Asians are “doing just fine.” Consequently, socioeconomically disadvantaged Asian subgroups may lack the attention and resources they need (e.g., federal funding, scholarships).
The category “Asian” masks social-psychological differences among Asian subgroups: insights from the bamboo ceiling
Even after controlling for socioeconomic status, there are still important social-psychological differences among Asian subgroups. These critical distinctions can be masked when Asians are aggregated into a monolith. Table 1 summarizes recent studies contrasting EAs and SAs, two large Asian subgroups in the United States. In particular, net of other factors (e.g., socioeconomic status, English fluency), ethnic EAs—but not ethnic SAs—experience a bamboo ceiling in key domains, including leadership attainment (Lu, 2022, 2024; Lu et al., 2020; Zhu, 2023), academic performance in law and business schools (Lu et al., 2022), and starting salaries (Lu, 2023).
Comparing EAs and SAs in the United States
Note: Where appropriate, the listed studies accounted for demographic variables such as English fluency, birthplace, American/international status, age, gender, education level, socioeconomic status, tenure at organizations, and so on. EA = East Asian; SA = South Asian; MBA = master of business administration.
Bamboo ceiling in leadership
Across various contexts, Lu and colleagues have consistently found a disproportionate underrepresentation of EAs—but not SAs—in leadership positions in the United States (Lu, 2022, 2024; Lu et al., 2020). For example, in S&P 500 companies, whereas EAs have a lower CEO-to-population ratio than Whites, SAs actually have a higher CEO-to-population ratio than Whites (Lu et al., 2020). Unlike the lack of prominent EA CEOs in the United States, SA CEOs have led influential U.S. companies such as Citigroup, Google, IBM, Mastercard, Microsoft, Motorola, PepsiCo, and Starbucks. Had the researchers lumped EAs and SAs into a single “Asian” group, EAs’ underrepresentation among CEOs would have been obscured. A similar pattern emerges when examining the deans of the top 100 U.S. business schools. 3 Although Asians appear well represented overall, making up 20 of the 100 deans as of March 2024, a closer examination reveals that there is only one EA dean versus 19 SA deans. Notably, many of these SA deans helm elite business schools (e.g., Cornell University, Harvard University, University of Chicago). These two examples demonstrate that lumping EAs and SAs together can mask EAs’ bamboo ceiling in leadership, especially given that SAs excel in this regard (Lu et al., 2020).
Bamboo ceiling in academic performance
In the United States, Asians are commonly stereotyped as excelling academically. After all, they are “A-sians” rather than “B-sians.” However, across six studies (N = 19,194), Lu et al. (2022) challenged this assumption by revealing the underperformance of EA students in two common and influential institutions of professional education: U.S. law schools and business schools. Net of other factors (e.g., performance on admission tests, birth country), EA students consistently earned lower grades than their SA and White counterparts, whereas SA students excelled. Again, EAs’ bamboo ceiling in academic performance would have been obscured if EAs and SAs had been lumped together into a single “Asian” category (Lu et al., 2022).
Bamboo ceiling in starting salaries
As noted earlier, the model minority myth is often attributed to Asians’ overall economic success, which is largely driven by their high educational attainment. However, Lu (2023) challenged this myth by revealing a bamboo ceiling in starting salaries. In one study, Lu (2023) analyzed 19 class years of master of business administration (MBA) graduates who accepted full-time job offers in the United States. Initially, “Asians” as a whole had similar starting salaries as their White counterparts. A remarkable gap emerged, however, when EAs and SAs were examined separately (controlling for socioeconomic status, international/American status, and so on): Whereas SAs secured the highest salaries among all ethnic groups, EAs found themselves near the bottom of the pay scale. The estimated starting salary gap between SAs and EAs in nonconsulting industries was a substantial $4,000 per year—a considerable difference that compounds over one’s career life (Lu, 2023). 4
Explaining the bamboo ceiling
Why do EAs, but not SAs, experience a bamboo ceiling in the contexts above? To answer this question, an emerging body of research is exploring underlying social-cognitive differences between EAs and SAs.
Assertiveness
One key difference between EAs and SAs concerns verbal assertiveness, a prized characteristic in the United States (Lu et al., 2020). A consistent finding is that EAs are less assertive than SAs and Whites in interpersonal communication (Chavez, 2021; Kitayama et al., 2022; Lu, 2022, 2023, 2024; Lu et al., 2020, 2022). This cultural difference in assertiveness has various behavioral manifestations and mediates EAs’ bamboo ceiling across different domains: EAs are disadvantaged in leadership partly because they tend not to speak up and voice opinions (Lu et al., 2020), disadvantaged in academic performance in U.S. law schools and business schools partly because they tend to have low class participation (Lu et al., 2022), and disadvantaged in starting salaries partly because they tend not to negotiate (Lu, 2023).
The face-honor culture framework (Leung & Cohen, 2011) provides insights into the cultural roots of the difference in assertiveness. In a measurement study, Yao et al. (2017) demonstrated that China, Japan, and South Korea represent “face” cultures, whereas India and Pakistan represent “honor” cultures. Face cultures (e.g., EA cultures), which are thought to originate from stable and homogeneous environments with centralized and strong governments, emphasize conformity, humility, and acceptance rather than assertiveness (Leung & Cohen, 2011; Lu, 2023). By contrast, honor cultures (e.g., SA cultures), which are thought to originate from unstable and heterogeneous environments with decentralized and weak governments, encourage individuals to act assertively to protect and advance their own and their in-group’s interests (Leung & Cohen, 2011). As explained by Sen (2005), SA cultures value debate and argumentation such that SAs “encounter masses of arguments and counterarguments spread over incessant debates and disputations” (p. 3).
Perceived creativity
A second difference concerns perceived creativity, another valued leadership characteristic in the United States (Lu, 2024). In essence, creativity requires individuals to diverge from established norms and methods (Goncalo & Staw, 2006), but this can conflict with EA cultures’ emphasis on conformity, acceptance, and humility (Lu et al., 2023). As a result, EAs who act in a more subdued and conforming manner may be perceived as lacking creativity in U.S. culture. Relatedly, EAs may be perceived as robots that simply follow orders rather than generate original ideas (Bain et al., 2009). Indeed, W. Y. W. Lee (2023) and Lu (2024) both found that, compared with other ethnic groups, EAs are stereotyped as lacking creativity, which contributes to their bamboo ceiling in leadership.
Perceived masculinity
A third difference concerns perceived masculinity, another prototypical leadership characteristic (Ensari et al., 2011). EA men tend to be perceived as less masculine than SA, White, and Black men (Goh & Trofimchuk, 2023; Lu, 2021). These ethnic differences in perceived masculinity are partly driven by two facial cues. The first is facial hair, which EA men tend to have less of than men from other ethnic groups (Goh & Trofimchuk, 2023). The second is facial width-to-height ratio, an indicator of facial masculinity (Geniole et al., 2015). Using cutting-edge facial analysis of headshots, Lu (2021) found that EAs tend to have a lower facial width-to-height ratio than other ethnic groups and that this lower perceived masculinity contributes to EAs’ bamboo ceiling in leadership.
Ethnic homophily
A fourth difference that contributes to EAs’ bamboo ceiling in leadership concerns ethnic homophily, defined as “the preference for interacting with individuals of the same ethnicity” (Lu, 2022, p. 960). Using social network analysis, Lu (2022) found that, compared with other ethnic groups, EAs—but not SAs—socialize more with ethnic in-group members (i.e., other EAs) rather than building cross-ethnic connections. As a result, EAs are less likely to attain leadership in multiethnic environments, which call for leaders who can effectively bridge and unite members of diverse ethnic backgrounds.
Together, the above studies highlight incongruence as an overarching reason for why EAs—but not SAs—experience the bamboo ceiling in leadership: EAs are less likely to attain leadership in the United States partly because their low assertiveness, low perceived creativity, low perceived masculinity, and high ethnic homophily are incongruent with the prototypical leadership characteristics valued in the United States (Lord et al., 2020). More importantly, these bamboo ceiling studies demonstrate why the generic category “Asian” is problematic and why it is essential to disaggregate between Asian subgroups in research and practice. Without such differentiations, many of the significant disparities among Asian subgroups would have been obscured.
Action Items
For researchers
Whenever possible, researchers around the world should avoid using the overly broad “Asian” category. In theorization, researchers should specify to which Asian subgroups their theory applies. During data collection, researchers should survey a person’s national origin (e.g., Japan) in addition to his or her ethnicity (e.g., Asian). This is particularly important given that Asian Americans are more likely to identify by national origin than by ethnicity (J. Lee et al., 2018). For example, in the 2012 National Asian American Survey—a nationally representative academic survey of Asian Americans—roughly 84% of participants identified with a specific national origin, whereas only 19% identified as Asian (J. Lee et al., 2018). In analyses, whenever statistical power permits, researchers should examine the differences among Asian subgroups and, if possible, heterogeneities within each subgroup. For instance, although EAs tend to be less assertive than both ethnic Indians and Pakistanis (Lu et al., 2020), these two SA groups also have meaningful cultural differences stemming from their distinct religions, languages, histories, and so on.
Similarly, researchers should examine intragroup heterogeneities within other ethnicities. For example, African Americans and Caribbean Black Americans are meaningfully different (Anderson, 2015), as are Mexican Americans and Spanish Americans (Moslimani et al., 2023). Although differentiating between ethnic subgroups can yield a deeper understanding of their distinctive challenges and opportunities, it is also important to emphasize intergroup solidarity (Burson & Godfrey, 2020). In fact, many groups rally under the panethnic “Asian” identity to facilitate collective action to improve equity and inclusion (Okamoto & Mora, 2014), especially given their “shared vulnerabilities in social standing” as ethnic minorities (Lin et al., 2024, p. 3).
For government and organizations
As exemplified by my analysis of S&P 100 companies, U.S. organizations commonly aggregate Asians into a single category. This problematic practice is unsurprising because the U.S. Equal Employment Opportunity Commission (EEOC) also treats Asians as one group on its standard EEO-1 form, which requires private-sector employers with at least 100 employees and federal contractors with at least 50 employees to report their workforce demographics by job category, sex, and race/ethnicity (U.S. EEOC, 2023). To address this issue, government agencies such as the EEOC should revise their demographic reporting requirements and mandate that U.S. organizations use more granular, disaggregated Asian categories when submitting workforce data. Even without such a regulatory requirement, organizations themselves should take the initiative to collect and present subgroup-specific data. In one of the few reports focusing on Asian American workers, McKinsey & Company presented separate statistics by Asian subgroups for perceived inclusion (operationalized as perceived acceptance, authenticity, belonging, camaraderie, fairness, and meaningful work): EAs and SEAs—but not SAs—perceived lower inclusion than their White peers (Chui et al., 2022). Only by understanding such distinct differences can organizations provide targeted support and interventions to address the unique challenges faced by particular Asian subgroups.
Future Research Directions
Understudied Asian subgroups
To date, most studies comparing Asian subgroups in the United States have focused on the more populous ethnic groups of EAs and SAs. More research is needed to study SEAs, especially given their lower socioeconomic status than EAs and SAs ( Kochhar & Cilluffo, 2018). For example, in the aforementioned 19-year study on salary negotiations and outcomes (Lu, 2023), the researcher was able to analyze not only EAs and SAs but also SEAs. The study found that SEAs exhibited similarities to EAs in terms of relational concerns, negotiation propensity, and starting salaries. These similarities are unsurprising because research suggests that like many EA cultures (e.g., Chinese), certain SEA cultures (e.g., Singaporean) are face cultures that prioritize values such as harmony, humility, and acceptance over negotiation (Yao et al., 2017).
Intersectionality
More research is also needed to examine how Asian ethnicity intersects with other demographic characteristics, such as gender, age, and sexual orientation (Rosette et al., 2018). In terms of ethnicity-gender intersectionality, some statistics suggest that SA men tend to fare better than SA women. In the aforementioned case of top 100 U.S. business school deans, only one of the 19 SA deans was female, hinting at a meaningful ethnicity-gender intersectionality effect. Similarly, although there are many SA S&P 500 CEOs, most of them are SA men. However, studies on leadership attainment have found inconsistent EA/SA × Gender interaction effects in different contexts, perhaps sometimes because of an EA floor effect (when there are too few EA leaders for the EA/SA × Gender interaction to be significant). More intersectional research is needed to unpack these complex dynamics.
Conclusion
With original analyses from both academic and business contexts, this article has spotlighted the problem that Asians are commonly categorized as a monolithic group in research and practice. As exemplified by the bamboo ceiling studies, the generic category “Asian” can mask important differences among Asian subgroups and lead to misleading findings. Disaggregating this category is essential for generating accurate scientific knowledge and fostering diversity, equity, and inclusion.
Recommended Reading
Kitayama, S., Salvador, C. E., Nanakdewa, K., Rossmaier, A., San Martin, A., & Savani, K. (2022). (See References). Moves beyond the common East-versus-West paradigm in its discussion of how non-Western cultural zones (East Asian, South Asian, Arab, and Latin American zones) are characterized by different forms of interdependence.
Lee, J., Ramakrishnan, K., & Wong, J. (2018). (See References). Offers a demographic perspective on why accurately counting Asian Americans is essential to diversity, equity, and inclusion.
Lu, J. G. (2023). (See References). Explores why East Asians but not South Asians experience a bamboo ceiling in starting salaries.
Lu, J. G., Nisbett, R. E., & Morris, M. W. (2020). (See References). Explores why East Asians but not South Asians experience a bamboo ceiling in leadership.
Lu, J. G., Nisbett, R. E., & Morris, M. W. (2022). (See References). Explores why East Asians but not South Asians experience a bamboo ceiling in U.S. law and business school classrooms.
Footnotes
Acknowledgements
I thank Jane Minyan Chen, Jennifer Lee, Angelina Li, Kyra Rodriguez, Lesley Luyang Song, Sophia Yat-Mei Suen, JoAnne Yates, Haoqing Sophie Ye, Doris Lu Zhang, and Anna Manyi Zheng for their helpful feedback and research assistance.
Transparency
Action Editor: Robert L. Goldstone
Editor: Robert L. Goldstone
