Disaggregating Data in AAPI Communities — A Model for Representing Minorities
Big data has become inextricably woven into so many essential services in today’s world. For industries ranging from healthcare to finance, it is indispensable.
The political world is no exception; politicians have been using massive amounts of voter information to shape the direction of their campaigns during each election cycle. Populations with high ratios of eligible voters are targeted more frequently by candidates, as are populations who vote together in a phenomenon known as “bloc voting.”
Other populations which have not clearly demonstrated bloc voting in past elections, such as Asian-American communities across the country, are more likely to be dismissed as disunited and statistically insignificant, receiving less attention from candidates and incumbents alike. However, AAPI (Asian American and Pacific Islander) early voting rose by almost 300% in the 13 most contested presidential battleground states of 2020, and AAPI voters proceeded to have a large impact on the margin of victory in the 10 most contested states in November.
These overlooked communities have proven themselves to be significant enough to warrant attention during elections. Mary Lou Akai-Ferguson, who served as the Asian American and Pacific Islander outreach director for Sen. Elizabeth Warren’s presidential campaign, understands the importance of appealing to the AAPI electorate: “I think that people are finally starting to take it more seriously. I think that people are understanding that we just can’t rely on the anonymous Midwestern white man to win the election for us.”
So, how can the use of data be adapted to accurately represent communities which are significant, yet do not satisfy the traditional pattern of bloc voting? The solution: disaggregated data. Analysis of disaggregated data as opposed to traditional aggregated data can more accurately model the nuances of voting behavior among changing demographics, which will translate into better and more informed policymaking in the future.
Data Doesn’t Lie, Right?
Data doesn’t lie, but it doesn’t always tell the whole story. Most data have represented Asian-American communities as homogeneous, ignoring their diversity, and the assumption has critically inhibited the effectiveness of most analyses in assessing the political behavior of those communities. In the words of Dr. Wendy Tam, senior research scientist at the National Center for Supercomputing Applications at UIUC: “Past studies of Asian voting behavior have more often than not treated Asians as a single homogeneous group…the underlying assumption of homogeneity can produce fallacious results when the group is not homogeneous.”
Here is the crux of the issue: the flaw is not in the data, but in its interpretation. Differences in voting patterns are misinterpreted as disunity, when they actually reflect different Asian-ethnic groups pursuing their individual interests. Similarly, when these Asian-ethnic groups do demonstrate bloc voting, analysts and pundits are caught off guard. The traditional patterns of data used to model voting behavior for AAPI communities are prone to inappropriate aggregation, and consequently result in misinterpretation.
Disaggregated data, on the other hand, is data which has been divided into detailed sub-categories to more accurately represent the sample population. In the case of mapping voting behavior in AAPI communities, disaggregated data identifies and categorizes the data into subgroups (Chinese, Japanese, Indian, Vietnamese, etc.) for those communities. This process is much more comprehensive than aggregated data analysis, and renders the data much more useful.
Why Disaggregate Data?
The National Asian American Survey (NAAS) and AAPI Data, an organization dedicated towards studying voting behavior in Asian-American communities, conducted a survey of voters after the 2016 presidential election. The results: 70 percent of the surveyed AAPI-identifying voters reported that they were not contacted by either political party throughout the election. This low level of engagement on the part of political campaigns reflects the lack of understanding of AAPI voters.
In order to appeal to voters, candidates need to identify the issues those communities are most passionate about. Campaigns have long been reluctant to invest in that kind of outreach in AAPI communities, where there is neither precedent nor accurate data representation of the voters. However, as AAPI populations grow at a significantly higher rate, the lack of understanding and procedure is making the AAPI voter population a wild card in elections, which has been evident in the experiences of politicians competing in battleground states such as North Carolina, Pennsylvania, and Georgia, where AAPI voters comprise 3.5 percent, 4 percent, and 3 percent of the electorate, respectively. While still small in comparison to other groups, those numbers are increasing rapidly, and are already substantial enough to sway close contests.
Disaggregated data is the key to understanding the voting patterns of diverse AAPI communities. Traditional aggregated data incorrectly treats the AAPI electorate as a monolithic voting bloc; Pomona College professor of politics Sara Sadhwani recognizes that “such research has been conducted by aggregating all Asian-Americans, and making the assumption of a pan-ethnic Asian-American identity.” In clear contrast, disaggregated data recognizes and documents the critical nuances of the population. By categorizing the voting population into distinct entities with individual interests, more effective outreach strategies can be identified and implemented, resulting in increased voter engagement which benefits politicians and voters alike.
Outreach into Asian-American communities increased throughout the most recent elections. During the 2020 presidential election, the campaigns of both presidential candidates dedicated extensive resources to attract AAPI voters, especially the Biden campaign. The outreach took the form of support for affinity groups, the hiring of translators at rallies and campaign fundraising events, targeted advertising, translated campaign literature, and outreach through ethnic media platforms. Amit Jani, the Biden campaign’s national AAPI director, understands the importance of connecting with AAPI communities through appropriate means. “We’re proud of our individualized and culturally relevant outreach to engage with some of the AAPI voters who may have felt neglected before,” he says, “We’ve worked hard to ensure the AAPI community feels galvanized, empowered, and heard — that their voices and votes matter.”
It is evident that resources are increasingly being directed towards the ever-growing AAPI electorate. Therefore, it is more important than ever to know how to use those resources efficiently, which requires accurate data collection and representation. Operating on aggregated data and fallacious assumptions results in non-specialized, ineffective engagement strategies which end up appealing to no one. By disaggregating the data, the door is opened to clearer understanding and more effective communication between candidates and the electorate.
Understanding AAPI Communities
The lack of understanding in AAPI communities has laid bare the need for better communication between politicians and their minority constituents. Furthermore, as minority populations grow in the United States, a robust methodology for data analysis is necessary for both political and civic planning.
Asian-American communities include individuals representing approximately 50 different ethnic groups and speaking over 100 different languages, each coming from distinct countries and cultural backgrounds. As mentioned before, these communities do not necessarily have the same interests, and to represent them as a single entity is conducive to the misallocation of resources and the ignoring of groups with overlooked needs. This issue is aptly summarized by UCLA PhD candidate Vivien Leung: “Asian-American voting preferences present a tougher test of theories of racial bloc voting and descriptive representation because of the added layer of national-origin differences.” Without appropriately disaggregated data, policymakers do not have the necessary information to adequately recognize and address their diverse constituencies’ needs.
Take a look at the following data illustration, which is a side-by-side comparison of disaggregated data for 10 AAPI subgroups and the aggregated AAPI data (the representative figure representing Asian-Americans in general) measuring the percentage of the population with a Bachelor’s degree or higher, the percentage in poverty, and the percentage without health insurance.
It is evident that while many groups are close to the representative average of AAPI aggregated data, others fall well below it. This example illustrates the danger of assuming that aggregated data adequately represents the needs of all members of the larger AAPI group.
Aggregated data oftentimes fail to convey the fact that Asian-American communities face different and unique challenges. For example, most studies fail to recognize that Hmong, Vietnamese, and Cambodian communities are disadvantaged in economic, higher education, and health outcomes as compared to other Asian-ethnic groups. In 2016, California passed AB-1726, which tasked the state Department of Public Health with disaggregating health data coming from AAPI communities. This pattern of legislation recognizes the inherently different realities faced by different Asian-American populations.
AAPI communities make up the fastest growing group in the United States, which has been recognized by most governing bodies, and data needs to more accurately reflect the subject populations in order to remain effective. The White House Initiative on Asian Americans and Pacific Islanders (WHIAAPI) was created in order to promote, among its many other goals, better data collection, dissemination, and disaggregation. During President Barack Obama’s administration, WHIAAPI worked with 24 federal agencies and offices in developing data collection methods to more accurately model the rapidly growing AAPI populations. Work of this nature has revealed overlooked statistics which could have implications in governing and in addressing the needs and expectations of certain AAPI communities. For example, a 2012 report published by the United States Department of Veterans Affairs to assess the demographic and socioeconomic characteristics of minority Veterans found that Asian American veterans are among the oldest and least educated across all racial groups. Furthermore, a United States Department of Labor study shows that Pacific Islanders suffered the highest rise in unemployment rates among AAPI demographics during and following the recession which began in January of 2009.
These reports can also serve as indicators of growth and measures of impact. A 2011 report released by the United States Census Bureau indicates that the number of U.S. businesses owned by members of the AAPI community increased 40.4 percent to 1.5 million between 2002 and 2007, increasing at more than twice the national rate, prompting Census Bureau Deputy Director Thomas Mesenbourg to state that “Asian-owned businesses continued to be one of the strongest segments of our nation’s economy, bringing in more than half a trillion dollars in sales in 2007 and employing more than 2.8 million people.” In the same year, the Department of Homeland Security reported that over 40% of all immigrants who obtained legal permanent residence last resided in an Asian country. Evidently, these statistics are significant and lay the vital groundwork for more effective understanding of not only the growing AAPI communities, but of the United States as a dynamic and evolving nation.
Despite these efforts, AAPI communities are still recognized as some of the most understudied groups in the country. Obtaining more disaggregated data will not only help policymakers serve their AAPI constituents more effectively, AAPI communities will be more accurately represented in general. Data should accurately reflect the realities of their subject populations, and disaggregation will refine the data to better represent the realities of AAPI communities, from voting behavior to healthcare trends to education. As California Attorney General Rob Bonta stated: “Our goal is to make progress, and we think tremendous progress can be made and knowledge learned that’s incredibly valuable through disaggregated data.”
The Future
Asian-American communities are growing faster than ever before, and although the voter turnout rate for AAPI communities remained glaringly low throughout the history of the United States, that has changed dramatically as well. Turnout rates have steadily crept to all-time highs, culminating in the 2020 presidential election, when AAPI voter turnout increased by 310%. Furthermore, the popularity of Asian-ethnic candidates in the 2020 presidential race like Andrew Yang and Vice President Kamala Harris reflects the increased civic participation of Asian-American communities across the country. All of these factors contribute to the AAPI electorate’s status as a rising and important force in the American democratic system.
Disaggregated data lays the foundation for policymakers to recognize the broader AAPI community for what it truly is: diverse and distinctive. This new pattern of data analysis paves the way for better understanding of not only AAPI communities, but also other significant minority communities that have also grown considerably. As more minority communities participate in the political process, they will play pivotal roles not only in flipping battleground states during presidential elections, but also in redefining long-standing patterns in congressional districts and the composition of the federal legislature for years to come.
The American electorate at large will only become more fractured and bloc voting will become less common. Throughout this time of change, disaggregated data will provide the information policymakers need to better understand, and more importantly serve, their constituencies.
