Unpacking the Data to Tell Truer Stories About Ourselves

Getting your Trinity Audio player ready…

Back in 2020, a glance at public data might have left the impression that the Asian, Native Hawaiian, and Pacific Islander (ANHPI) community was somehow faring well during the Covid-19 pandemic.

“Really, when we looked at the data, it looked like nobody had Covid,” said Julia Liou, MPH, the chief executive officer of Asian Health Services (AHS), a Federally Qualified Health Center in Oakland’s Chinatown neighborhood. The clinic had become a ghost town. Few patients walked in for care, and patients who booked appointments routinely missed them. And, yet, in certain Asian subpopulations, the anecdotal evidence was strong: everyone knew people who were suffering with Covid.

Determined to figure out what was going on, AHS President Thu Quach and her staff reached out to community members and slowly pieced together a more nuanced picture that showed many Asian people were avoiding health care services entirely.

But there was more to it. The Trump administration threatened to label immigrants seeking public health assistance as “public charges,” potentially limiting their future eligibility to seek permanent residency or US citizenship. While the Biden administration later withdrew that proposal, it had a chilling effect on immigrants nationwide. On top of that, many Asian immigrants and ANHPI Americans were laying low because of increased hate crimes motivated by claims that Covid was the “China flu.”

Realizing that members of their community needed Covid testing and care but were not accessing it, Quach, Liou, and the AHS team endeavored to go to them. But who, precisely, was “them”?

AHS went back to the data for a closer look and found that while members of the Asian community collectively reported lower Covid rates than other groups, including the White population, a different story unfolded when that data was broken apart, or disaggregated. “We saw, for example, there was actually a huge Covid spike in our Vietnamese community,” said Liou. AHS quickly established pop-up Covid clinics in the Little Saigon neighborhood of Oakland, bringing care to where it was most needed and successfully quashing a local Covid surge.

“The lesson we learned,” said Liou, “is that to find the truth, we had to disaggregate.”

‘Undoing Decades of Harmful Stereotypes’

Over the last six decades, a mythology stereotyping the Asian community as America’s “model minority” has characterized Asians as unwaveringly successful, healthy, hardworking, and well educated.

While this label may appear flattering on the surface, it is harmful for several reasons — first, because it paints the diverse experiences of Asian groups inaccurately and with a broad brush, and second, because it has often been used as a cudgel against other racial and ethnic groups. As a result, ANHPI advocates have worked tirelessly to dispel this myth and convey the varied experiences of the more than 50 ethnic groups that comprise America’s Asian communities.

One lens into the diversity of Asian experiences in America is wealth. Asians as a whole are the most economically disparate racial or ethnic group in the US, according to the Pew Research Center. In the LA area, for example, the median value of liquid assets held by Indian households is $245,000 (PDF), compared to $130,000 for Chinese families and $3,000 for Korean households.

Understanding the impact that diversity might have on health needs has become a major force behind a new move toward data equity. Data disaggregation is a clunky term with profound implications that captures a strong desire in the public health world to stop treating Asian Americans as a monolith.

‘Disaggregating’ the Data

Advocates say granularity is the only way to understand ANHPI populations and to wisely allocate health resources.

“Disaggregating data offers a more accurate picture of the health needs of a community and allows for targeted interventions to improve outcomes,” said Ninez A. Ponce, PhD, MPP, director of the UCLA Center for Health Policy Research and a professor at the UCLA Fielding School of Public Health.

Using the California Health Interview Survey (CHIS), Ponce and her colleagues have found significant differences in health outcomes by separating Asian American subgroups. For example, rates of diabetes are significantly greater among Filipino Americans than Chinese Americans. “This difference would not be seen if not for the practice of data disaggregation,” said Ponce.

Ponce’s team also found disaggregation helpful in improving mental health (PDF) outcomes. In the survey results, 12% of Asians said they have thought about attempting suicide compared to 17% for Californians overall. When Asians were disaggregated, however, it emerged that 19% of Korean Americans reported thinking about suicide — a rate that was not only the highest of all Asian sub-communities, but that was also greater than the state average.

Ponce and colleagues also found a lack of detailed information on breastfeeding practices for disaggregated Asian communities. “The National Vital Statistics System does a great job of collecting and reporting disaggregated data, showing variation in breastfeeding initiation rates among Asian subgroups,” Ponce said. “But there aren’t datasets that offer disaggregated information on sustained or exclusive breastfeeding in these communities, for example exclusive breastfeeding at six months. We also lack knowledge about specific barriers to breastfeeding by community to effectively address needs.”

How Data Can Obscure Important Details

Katrina Heyrana, MD, PhD, assistant professor in family planning at Cedars-Sinai in Los Angeles, is also engaged in the data disaggregation movement. Her work digs into how the varied migration stories of Asian subpopulations provide insight into the health needs of communities — something that she says most people don’t understand.

“Cambodian people and Vietnamese people have often migrated to the United States because of a true refugee experience, like fleeing a war,” she said. But other groups immigrate from former US colonies, like the Philippines, or come to the US for economic or educational opportunities. A lot hinges on a person’s migration story, including economic and housing opportunities, baseline health status, and language abilities. “If we’re considering [Asian Americans] as a cultural monolith, not only is it racist, it fails to address the distinct needs of each cultural group,” Heyrana said.

One example of that phenomenon is the way Asian mothers and other birthing people have been presented in data concerning rising maternal mortality rates in the US. Reports show that Black people who give birth have the highest mortality rates, followed next by White, Latino/x, and Asian people. But this data point treats the Asian community as a singular entity, obscuring important details that policy makers need to understand.

Research on Asian subgroups in California found that ANHPI mothers and birthing people are at increased risk for poor maternal and infant outcomes — such as gestational diabetes, gestational hypertension, cesarean birth, and low birthweight — compared to the White population. But no two subgroups had the same red flags, and when their data was pooled, the red flags faded and the research did not tell their stories.

Heyrana notes that a predominant view among researchers is that members of the Asian community access selective abortions at a higher rate than other groups due to fetal genetic anomalies. But her research found problems with lumping together these data too. In reality, many subgroups obtain abortions at the same rates as other non-Asian groups — not higher — and for the same varied reasons.

There are obvious economic benefits to a health system that understands more precisely where resources are needed and that uses limited resources appropriately, Heyrana said. Disaggregated data can enable that, adding that the value of telling the stories of individuals in the community accurately and genuinely should not be overlooked, she said.

A Work in Progress for All

Data equity proponents are pleased by recent progress toward more granular data collection within the federal and California governments. Last March, the US Office of Management and Budget (OMB) published revisions to statistical policy directive No. 15. This was the first update since 1997, and it affects census data collection, federal research, and other data efforts. The process of revising the directive took OMB’s Interagency Technical Working Group on Race and Ethnicity Standards and the Equitable Data Working Group (PDF) two years and included 20,000 stakeholder comments across nearly 100 listening sessions.

The updated OMB standards make key revisions to questions used to collect information on race and ethnicity, including the separation of ANHPI categories for race and/or ethnicity. The guidelines also encourage questions that allow respondents to select as many options as needed to capture how they identify — a crucial change because of the increased number of people who identify as multi-racial.

“The OMB update helps solidify the need for [more] detailed groups and advocates for an option to write in data when those individuals do not see their identities in any of the listed categories,” Ponce said.

Still, data collectors need more guidance to effectively implement the standards, collect data in historically marginalized communities, handle small sample sizes, manage and categorize write-in data, and build long-term relationships with communities. For example, the Census Bureau misclassified the Hmong people as “east Asian” instead of “Southeast Asian” — an important distinction in the Hmong community for historical, social, and cultural reasons.

In California, efforts toward disaggregation are underway. Legislation implemented in 2022 requires the California Department of Public Health (CDPH) to break down broad demographic data into more specific Asian, Native Hawaiian, and Pacific Islander subgroups.

Stakeholders Collaborate with State Government

Proponents see these recent federal and state changes as victories, but they say more needs to be done. Implementation delays have limited community groups and researchers in California from seeing the impacts of the new policy on disaggregated data, said Quach, president of AHS.

In response, she has helped spearhead a coalition of more than 50 California organizations, researchers, and advocates to collaborate with CDPH to eliminate bottlenecks. “We have monthly meetings, and we’re going deep — working with data scientists, epidemiologists, and data flow issues.”  The coalition sees problems with how different health care systems share data as a huge challenge. But with human buy-in for greater collaboration, Quach said, the obstacles are surmountable.

Andrew Lee, the California policy and legislative advocate for the national nonprofit Southeast Asia Resource Action Center, said activists need to play the long game. “We’ve been one of the organizations advocating for data disaggregation since the 1970s as an approach to better understanding all sorts of issues,” said Lee. “It’s really the lens through which we look at all issues.”

Race and ethnicity are both social constructs. Race is often based on perceived or true biological variations among groups of people, whereas ethnicity is a broader category usually based on common language, religion, nationality, history, or other cultural factors. Race and ethnicity often overlap in complex ways. Definitions change continually over time. That’s why data equity isn’t just an Asian American issue, Heyrana said.

Researchers say data equity will become increasingly important as the US population becomes more diverse, and efforts to advocate for more granular data will go beyond Asian Americans to other communities, like those who identify as LGBTQ.

“Understanding the unique challenges our communities face is key to creating better solutions,” said Ponce. “A universal, one-size-fits-all approach has not effectively addressed health disparities. We need better solutions.”

Lee notes, however, that equity through data disaggregation sometimes fails to attract the spotlight given to other causes because it sounds so wonky. Folks don’t immediately want to talk about numbers, he said, but framed another way, it becomes clear that data disaggregation is a pathway to narratives that tell truths about people that are more compelling and relatable.

“Data equity is a civil rights issue,” Lee said. “In an ideal world, everyone counts, and everyone is counted. No community is too small, and no community is invisible. Because, really, every community has unique needs.”

Authors & Contributors