An Innovative Model for Generating Limited English Proficient Population Estimates
By Christy Hoff and Erica Gardner
In an effort to improve meaningful access to Washington State government information and services as required by Title VI of the Civil Rights Act, state agency staff collaborated to create an innovative model for identifying the most common languages spoken across the state and generating Limited English Proficient (LEP) population estimates. The model uses data on the primary language spoken at home for all public school students in grades 1 through 12 combined with race/ethnicity-specific estimates of household size. The data have been used by individual state agencies to target resources for language assistance services and to inform the state’s COVID-19 Language Access Plan.
Why did Washington State design a methodology for estimating its LEP population?
In June 2014, the Governor’s Interagency Council on Health Disparities issued a report with recommendations to assist state agencies in providing meaningful access to information and services for LEP residents. At that time, council staff also convened an informal network of state staff working to promote language access in their respective agencies. Many agencies were developing, updating, and/or implementing their written language access policies and plans and identifying the most common languages for translation and interpretation services.
While some agencies collected their own data on limited English proficiency directly from clients accessing their services, others were relying primarily on Census data. The U.S. Census Bureau’s decennial census and annual American Community Survey (ACS) collect data on whether and what languages other than English are spoken at home and language proficiency. The data are coded into 42 language classifications; however, many of the classifications are broad and, therefore, not very helpful with informing decision-making (e.g., “Swahili or other languages of Central, Eastern, and Southern Africa”).
Council and other agency staff identified the need for more complete data on limited English proficiency and languages spoken in the state to better inform language access resources. Washington State’s Office of Superintendent of Public Instruction (OSPI) collects data on the primary language spoken at home for all students in grades 1 through 12. It is an ideal data source because it offers virtually complete coverage throughout the state. Languages listed are not limited, allowing for the detection of languages that are emergent or of lesser diffusion. In 2016, the second and most recent time these LEP data were analyzed, 246 languages were identified as spoken by the state’s primary and secondary student population.
The Washington State Office of Financial Management (OFM) is the agency responsible for supplying data and information to inform government decision making, including statewide population data. Council staff and OFM analysts discussed the need for better language data in the state and whether the OSPI student data could be used to extrapolate population estimates.
What is the methodology?
OFM has published a detailed description of the methodology. In brief, OFM analysts use OSPI’s data as a base for their analysis and combine it with census data on household size for the racial/ethnic group that most closely matches each language spoken. Population estimates are obtained by multiplying the number of students who speak each language by household size. Estimates are provided at the statewide level as well as any county with at least 250 students (or at least 5 percent of the student population) identifying the language as the primary language spoken at home. The OSPI results are checked against the latest ACS data on limited English proficiency and individuals requesting translation services via the Temporary Assistance for Needy Families/Supplemental Nutrition Assistance Program/Medicaid programs to ensure languages are not missed. Any additional languages that meet a threshold of 500 people within a county or the state are added to the list of estimated languages (a higher population threshold for inclusion is used as these data sources include adults and children while OSPI data include only the student population).
How are LEP population estimates used in Washington State?
The population estimates are used by state agencies to ensure compliance with Title VI of the Civil Rights Act and to target limited resources for language assistance services at the state and local levels. As one specific example, the estimates were used to inform the state’s COVID-19 Language Access Plan. The plan directed all state agencies to translate their vital COVID-19 information into the top 37 languages spoken by people with limited English proficiency. The 37 languages were selected based on the Department of Justice’s safe harbor provision, which requires that, at a minimum, vital information is translated into languages spoken by at least 5 percent of the population or 1,000 people (whichever is less). The languages were identified from the 2016 OFM estimates and the governor’s office secured the funding for all translations of COVID-19 vital information.
Christy Hoff was Manager of the Governor’s Interagency Council on Health Disparities at this writing; Erica Gardner is a Senior Forecast Analyst with the Washington State Office of Financial Management.