Voice: Suhaila Khan, MD, PhD, MPH
Position: Program Director
Organization: Asian & Pacific Islander American Health Forum
Location: California
Keywords: Data collection

Efforts to collect information on the residents of the United States (U.S.) predate the Declaration of Independence and were signed into law with the Constitution in 1788.  Since 1790, when the first official U.S. Census was conducted, methods of data collection and reporting have grown increasingly robust.  In spite of this, a recent Institute of Medicine report, indicated that national surveys often fail to adequately depict the racial, ethnic, and linguistic diversity of the nation.  We spoke with Dr. Suhaila Khan of the Asian & Pacific Islander American Health Forum (APIAHF) to learn more about the importance of collecting data on multiple race/ethnicity, barriers to accomplishing this, and APIAHF’s efforts to promote programs and policies which support this endeavor to collect and report more reliable data.  

 

According to Dr. Khan, the rationale behind collecting population data is rather straightforward: for programs and policies to be effective and efficient they need to be grounded in evidence on the characteristics of the populations they are intended to serve. As racial/ethnic minorities are often underserved and have distinct needs—such as limited English proficiency and distinct cultural beliefs and customs—reliable information on these populations is particularly important to program planning. The Office of Management and Budget’s (OMB) Policy Directive No. 15 establishes minimum standards for collecting data by race/ethnicity, consisting of five racial categories (American Indian or Alaska Native, Asian, Black or African American, Native Hawaiian or Other Pacific Islander, and White) and two ethnic categories (Hispanic/Latino and not Hispanic/Latino).   Dr. Khan and others in the field believe that these minimum categories are incapable of fully capturing the ethnic diversity of the nation to enable effective programs and services.

 

Dr. Khan explained that there are over 50 different Asian American, Native Hawaiian, and Pacific Islander ethnic groups in the U.S., who together speak more than 100 languages. When data are reported on these populations in aggregate it masks important differences between these distinct communities. For example, 17% of Asians are uninsured. But when data are disaggregated, it becomes clear that certain Asian sub-groups differ substantially from Asians as a group—such as Koreans who have an uninsurance rate of 31%.

 

Such disparities are evident among a number of ethnic sub-groups across a wide range of socioeconomic indicators which contribute to a community’s ability to prepare for, respond to, and recover from emergencies.  For example, the aggregate poverty rate for Asians as a group is 8.2%, but nearly three times that for the Hmong (26.4%).  Information on the distinct needs of these ethnic sub-groups is critical to developing preparedness and response plans that are inclusive of the needs of the entire population.  

 

Failing to collect data on ethnic sub-groups is one part of the problem, but as Dr. Khan explained, data is often collected but not reported. The U.S. Census Bureau only reports on ethnic subgroups with populations over 65,000. This poses a major barrier to obtaining data on smaller ethnic sub-groups, especially at the state or county level.  For example, in 2006 there were only 69,687 Bangladeshi living in the U.S. and 86,246 Indonesians.  A reporting threshold of 65,000 makes it virtually impossible to acquire information on these populations beyond the national level.  While these reporting standards are intended to protect privacy, Dr. Khan and her colleagues at APIAHF have been vocal in advocating for innovative methods which expand reporting thresholds while maintaining privacy. 

 

Another issue surrounding ethnic sub-groups with small population sizes is that of statistical validity. Most population surveys, such as the U.S. Census’s American Community Survey (ACS), make population estimates based on a random sample of the population.  Samples of some racial/ethnic sub-groups are often so small in size that they are unable to produce population estimates that have statistical validity.  For example, the 2008 ACS estimates that there are 738 Vietnamese in Montana, with a margin of error of +/- 738.6.

 

Oversampling, the practice of intentionally surveying an elevated number of people from certain groups, is one technique to address this issue. APIAHF has advocated for oversampling practices in national surveys and a provision in the various 2009 health care reform bills (i.e. senate, house) would require oversampling for certain racial/ethnic sub-groups.  This technique produces more accurate estimates of a population’s demographic profile which can then be used to inform evidence-based programs and policies. APIAHF has also advocated for conducting standardized state level surveys across the nation.

 

In light of these challenges, APIAHF has launched a number of initiatives to improve data collection standards and centralize and distribute available data on Asian Americans, Native Hawaiians, Pacific   Islanders, and their ethnic sub-groups.  In addition to their advocacy work in the policy realm, APIAHF has developed a number of practical reports and online tools. One such resource is the Population, Growth and Socioeconomic Status tool.  This interactive resource uses 1990 and 2000 Decennial Census data to create portraits of Asian American, Native Hawaiian, and Pacific Islander sub-groups at national, state, and county levels. The tool provides information on a wide range of variables including language, education, income, immigration status, and household characteristics. 

 

Another valuable resource is a data chart entitled Asian American, Native Hawaiian and Pacific     Islander Population Demographics 2006.  This data chart uses 2006 ACS data (the most currently available and corrected) to depict the characteristics of Asian American, Native Hawaiian, and Pacific  Islander sub-groups in the U.S.  The report provides narratives, graphs, and figures on a wide range of categories such as English proficiency, income, and health insurance.  The report also provides a detailed table of data on the ethnic sub-groups in each state in the U.S. 

While barriers exist to obtaining reliable data on racial/ethnic sub-groups, there are a number of valuable resources to help maximize the utility of existing data.  As Dr. Khan emphasized, collecting and reporting data on racial/ethnic sub-groups is only the first step. 

After the characteristics of these communities are identified, their needs and assets must be incorporated into emergency preparedness, response, and recovery plans.  Such evidence-based actions are critical to reducing racial/ethnic disparities in disaster outcomes and protecting the health and safety of an increasingly diverse population.