Why do we collect data on women officeholders by race and ethnicity?
Since our founding in 1971, a primary initiative of the Center for American Women and Politics (CAWP) has been our collection and public reporting of data on women officeholders and, more recently, candidates. This data has been imperative to revealing the underrepresentation of women across levels of elected office in the United States and motivating efforts to increase women’s political power. Consistent with these goals, CAWP began collecting data on women officeholders by race and ethnicity in 1997 to document the disparities in women’s representation by race and ethnicity, as well as make clearer the persistent underrepresentation of women within historically marginalized racial and ethnic groups. These motivations continue to guide CAWP’s data collection and public reporting.
What do we mean by race and ethnicity?
CAWP’s data collection and public reporting has historically combined race and ethnicity based in our recognition that these identifications, as well as their influence on individual and group experiences, are not wholly separable and are rooted in complex social constructions. As our earliest coding aligned with the U.S. Census (see below), we approach our racial and ethnic identification measurement with the U.S. Census recognition that the categories we use “reflect a social definition of race recognized in this country and not an attempt to define race biologically, anthropologically, or genetically.” Ethnicity, when distinguished from race, often denotes shared culture, such as language, ancestry, practices, and beliefs. In a slight departure from the U.S. Census, to date we have not distinguished race from ethnicity in our reporting, as is most notable in our inclusion of Latina among all racial/ethnic categories. The U.S. Census has considered including Latino/a as a racial category. Likewise, they have considered but not adopted inclusion of Middle Eastern/North African among racial categories, which we now include in our data.
How do we report data by race and ethnicity? Has this changed over time?
CAWP has reported summary data on women officeholders and candidates by race/ethnicity since 1997. Prior to 2021, we reported this data under the category of “women of color,” presenting aggregate counts of women with self-identified races and ethnicities other than white with sub-counts by each racial/ethnic group (including one group for any women who identified as multiracial). While this approach was consistent with CAWP’s interest in challenging the myth that women are not monolithic, it centered whiteness as a default racial/ethnic category.
In our current reporting, beginning in 2021, we provide more detailed counts of women by racial and ethnic groups, including white women. In our current pages for women officeholders by race and ethnicity, we also offer more detailed current and historic data on Asian American/Pacific Islander, Black, Latina, Middle Eastern/North African, and Native American/Alaska Native/Native Hawaiian, and white women officeholders. CAWP added Middle Eastern/North African as a category of self-identification starting in 2019. We also included Alaska Native as a category for self-identification for the first time in 2019 and Native Hawaiian as a category for self-identification for the first time in 2021 providing greater specificity for native peoples within the United States. While these racial and ethnic categories are themselves imperfect and subject to change, we adopt these categories provisionally and in alignment with both the U.S. Census and CAWP’s historical race and ethnicity data collection. We will periodically revisit these categories to reflect current understandings of racial and ethnic identification.
In our current reporting, women who self-identify as more than one race and ethnicity are included in each group with which they identify. If officeholders choose to identify as “multiracial” alone, without specifying any unique racial and ethnic identifications, they are categorized as such in CAWP’s database and reporting. We strongly caution against adding totals from each racial/ethnic group should, as it will double count officeholders. To conduct more detailed calculations, users should refer to CAWP’s Women Elected Officials Database. For those choosing to report an aggregate count of “women of color,” including any women with racial/ethnic identities other than white, please refer to the database search tool by race/ethnicity (and select all but white and unavailable) or contact CAWP staff directly.
How do we collect racial/ethnic identification data?
Officeholder race/ethnicity is coded by a team of CAWP researchers in two ways. First, we rely on officeholder self-identification through direct contact (via email, phone, or social media). In direct exchanges with officeholders (whether at the candidate or officeholder stage), we provide them with information about why we collect this data, where it will be publicly reported, and how it will be used. We then provide a list of racial/ethnic categories, asking them to choose which (one or multiple) best reflect their racial/ethnic identification. Currently, we include the following racial/ethnic categories in our requests for self-identification: Asian or Pacific Islander, Black/African American, Hispanic/Latina, Middle Eastern/North African, Native American/Alaska Native, Native Hawaiian, White. We note that if none of these categories capture how an individual identifies that they should provide their preferred self-identification.
Where self-identification is not provided to us directly, we rely on historical and/or contemporary public records (e.g. websites, social media, interviews) that include statements of racial/ethnic self-identification by the officeholder. In a small number of cases where other indicators of racial/ethnic identification are publicly available, we use a multiple source verification process for coding. After exhausting these efforts, we label officeholders for whom we were unable to determine racial/ethnic identification as “unavailable.”