Why Data Disaggregation Matters: Exploring the Diversity of Asian American Economic Outcomes Using Public Use Microdata Sample (PUMS) Data
In the U.S. Census and many official policy documents, data is often aggregated into a broad “Asian” category. This categorization serves many purposes, ranging from its simplicity to a more reliable reporting of sampled demographic estimates. However, this broad categorization also masks the significant diversity within the Asian American population, which includes individuals with distinct cultural, socioeconomic, and immigration histories from over 20 countries of origin. Many Asian subgroups such as Vietnamese, Hmong, Cambodian, Laotian, and Burmese largely settled in the U.S. as refugees with very few resources, yet their experiences are overshadowed by larger groups. This issue is particularly evident in academic and policy research that relies on the broad Asian category to study topics such as inequality and segregation. Researchers and advocates increasingly caution that a monolithic narrative of Asian Americans as a “model minority” can be misleading and harmful. And, as I show in this blog post, aggregating economic outcomes across the Asian category contributes to this stereotype by concealing the significant disparities among individuals with a wide range of backgrounds.
In this blog post, I highlight how aggregate statistics of Asian American economic outcomes may misrepresent the diverse experiences of this group by analyzing the Census Public Use Microdata Sample (PUMS). The PUMS data provides detailed, anonymized information about individuals and households from the U.S. Census, including variables on demographics, income, education, housing, and employment (Ruggles et al., 2024). It also allows for a deeper, more granular analysis than what is typically available in summary Census reports by allowing users to create more detailed crosstabs, including those that use more detailed race information. By showing how the stark variation in household income and homeownership rates along different ethnic subgroups, I argue that data disaggregation is essential for understanding the diverse economic experiences within the Asian American community and informing the development of more effective policies for addressing racial inequities.
Who Counts as Asian in the U.S. Census?
The Asian category in U.S. history is a social construction that has evolved over time. The first enumeration of Asian Americans occurred in the 1890 Census, which recorded 107,488 Chinese and 2,039 Japanese individuals (Lee, 1993). By the 1930 Census, this category expanded to include Filipinos, Hindus, and Koreans. Although these diverse categories existed in the Census, the term “Asian American” emerged as a socio-political identity during the Civil Rights era. This identity was shaped both by the racial categorization imposed by the U.S. government and by a shared experience of discrimination and exclusion (Espiritu, 1992). It wasn’t until 1980 that the Census began to enumerate Asian and Pacific Islanders as a combined group, also starting to record the Vietnamese population. In 2000, the Census allowed people to report more than one race, which had a significant impact on how the Asian category is defined and measured. Today, the Asian category in the Census includes individuals who identify with countries in East, Southeast, and South Asia.
According to the 2020 Census, the six largest Asian groups were Indian, Chinese (excluding Taiwanese), Filipino, Vietnamese, Korean, and Japanese (Rico et al., 2023). Figure 1 illustrates how the composition of the Asian population by ethnic group has shifted over recent decades. Notably, the “Other Asian” and “Two or More Groups” categories did not exist until 1990 and 2000, respectively. The prominence of the six major groups—especially Chinese and Indian populations—means that the characteristics of these groups can heavily influence aggregate statistics, as we will explore further below.
Figure 1. Composition of Asian Population By Reported Ethnicity (1980-2019)
What Does Aggregate vs. Disaggregated Data Tell Us About Asian American Income and Homeownership?
The easiest ways to access the PUMS data is by using the IPUMS USA website and the ipumsr package in R. If you are only working with the American Community Survey data, you can also use the get_pums function in the tidyCensus package for even more convenient access. For this blog post, I will not dig deeper into the details of assessing PUMS data but I certainly hope you explore it for your own projects!
One of the first things that we can easily notice about Asian American economic outcomes is their high income levels. Figure 2 illustrates the real median household income (adjusted to 1999 dollars) by four major ethnoracial groups used in most social sciences research and policy documents—Non-Hispanic White, Black, Asian, and Hispanic. The figure clearly illustrates that Asian households (defined by the household head) had the highest average household income among the four groups and showed marked growth during the 2010s in particular.
Figure 2. Median Household Income by Householder Race/Ethnicity (1980-2019)
However, when we leverage the granularity of PUMS data to examine how income levels of top and bottom earners have changed over time, a very different picture emerges. Figure 3 illustrates the real household income changes over time for the top 10% and bottom 10% earning households, broken down by racial groups. The data shows that low-income households have remained consistently at very low-income levels across all racial groups, despite the fluctuations for White and Asian households and some growth for Hispanic households. Strikingly, in contrast, the top-earning Asian households have experienced a significant increase in their income over time. In short, data disaggregation reveals that the overall growth in Asian income was primarily driven by top earners, who experienced sustained income growth.
Figure 3. Top 10% vs. Bottom 10% Household Income by Householder Race/Ethnicity (1980-2019)
Another important aspect of within-Asian disparities is ethnicity or country of origin. Nearly two-thirds of the Asian population in the U.S. are foreign-born immigrants, meaning that the socioeconomic conditions of their home countries at the time of immigration can significantly influence their economic outcomes, even in the U.S. Figure 4 highlights this dynamic by showing how median household income in 2019 varies when disaggregating the Asian category into more specific ethnic groups. While the median household income for Asians was $86,355, only three subgroups—Indian, Taiwanese, and Filipino—surpassed this figure, indicating a high level of income inequality within the Asian American population. In contrast, groups such as Burmese and Bhutanese earned less than half of the Asian median household income, reflecting a level of economic precarity that would be obscured if the Asian category were considered a monolithic group.
Figure 4. Median Household Income by Detailed Asian Householder Race/Ethnicity (2019)
Data disaggregation also provides important insights into Asian American homeownership, which is a key focus of my own research. There are very few studies on Asian American housing experiences, partly due to the perception that Asians are economically well-off and therefore face few challenges in the housing market. However, despite having substantially higher household incomes than their non-Hispanic White counterparts, as Figure 5 illustrates, Asians have significantly lower homeownership rates. This disparity highlights the need for a deeper exploration of the housing experiences of Asian Americans beyond the assumption of their general economic success.
Figure 5. Average Homeownership Rate by Householder Race/Ethnicity (1980-2019)
As with household income, homeownership rates also vary significantly across Asian subgroups. For example, Figure 6 shows that Taiwanese, Vietnamese, and Japanese households have some of the highest homeownership rates among Asian groups, exceeding 65%. In contrast, Nepalese, Bhutanese, and Burmese—groups with some of the lowest household incomes—also have the lowest homeownership rates. This further demonstrates how disaggregating data reveals the economic challenges faced by these groups, which might otherwise be overlooked if they were grouped under the broad “Asian” category, especially in policymaking. Additionally, the relationship between homeownership and household income is not straightforward for many Asian subgroups. For instance, Vietnamese and Laotian households earn less than the median Asian household but have higher homeownership rates than the average Asian group. These patterns raise important questions about the factors at the household and housing market levels that shape Asian American economic outcomes, providing valuable avenues for further research.
Figure 6. Average Homeownership Rate by Detailed Asian Householder Race/Ethnicity (2019)
Data Disaggregation Matters for Addressing Asian American Inequality
Research by the Pew Research Center highlights that Asian Americans now face the highest levels of income inequality among major U.S. racial groups (Budiman & Ruiz, 2021). However, these diverse economic experiences among Asian Americans are often overlooked in Census summary statistics that rely on aggregate race data. This blog also shows how disaggregating the “Asian” category into more specific ethnic groups reveals significant disparities in income and homeownership. In addition to ethnic differences, my ongoing research suggests that Asian American experiences vary widely depending on region, citizenship status, and immigration period—even among individuals from the same ethnic group. Organizations such as AAPI Data, which is based at UC Berkeley’s Asian American Research Center, also represent recent efforts to leverage disaggregated data for capturing these nuances and provide a clearer picture of Asian inequalities. These types of detailed analyses will help develop targeted policies that address the specific needs and challenges faced by different Asian American subgroups and advocate for policies that address disparities more effectively.
References
-
Budiman, A., & Ruiz, N. G. (2021). Asian Americans are the fastest-growing racial or ethnic group in the U.S. Pew Research Center. https://www.pewresearch.org/short-reads/2021/04/09/asian-americans-are-the-fastest-growing-racial-or-ethnic-group-in-the-u-s/
-
Espiritu, Y. L. (1992). Asian American Panethnicity: Bridging Institutions and Identities. Temple University Press.
-
Lee, S. M. (1993). Racial classifications in the US Census: 1890–1990. Ethnic and Racial Studies, 16(1), 75–94. https://doi.org/10.1080/01419870.1993.9993773
-
Rico, B., Hahn, J. K., & Spence, C. (2023). Chinese, Except Taiwanese, Was The Largest Asian Alone or in Any Combination Group; Nepalese Population Grew Fastest. United States Census Bureau. https://www.Census.gov/library/stories/2023/09/2020-Census-dhc-a-asian-population.html
-
Ruggles, S., Flood, S., Sobek, M., Backman, D., Chen, A., Cooper, G., Richards, S., Rodgers, R., & Schouweiler, M. (2024). IPUMS USA: Version 15.0. Minneapolis, MN: IPUMS.