Description

RACESING codes race responses into a simple, historically compatible scheme. Multiple-race responses in the 2000 census, the ACS and the PRCS are recoded (or ''bridged'') into single-race responses. Although this simplification was done carefully, it should not be assumed that it is useful or required in every situation. It may not be needed if the user can tolerate a "break" in the racial coding system of their data series, or if an analysis includes only data from 2000 and beyond. The full detail available for responses to the race question is preserved in the RACE variable.

RACESING assigns a single race to multiple-race people. Each multiple-race person is assigned to the single race response category deemed most likely, depending on the individual's age, sex, Hispanic origin, region and urbanization level of residence, and the racial diversity of their local area. Local areas are defined as super-PUMAs for the Census 2000 1% sample, PUMAs for the Census 2000 5% sample, the 2005-onward ACS, the 2005-onward PRCS, and states for the 2001-2004 ACS samples. RACESING was created using the methods described in Ingram, et al. (2003). Users wishing to replicate RACESING in any other dataset with state-level geographic identifiers, can use the "race bridge" STATA program. Note, however that the IPUMS USA RACESING variable uses more detailed geographic information, so this program will not perfectly replicate the variable.

For each multiple-race person, probabilities were calculated for each of five single-race responses: American Indian/Alaska Native (PROBAI), Asian and/or Pacific Islander (PROBAPI), black (PROBBLK), white (PROBWHT), and other race (PROBOTH). IPUMS used the highest of these probabilities to assign each multiple-race person to a RACESING group. The sum of values for PROBAI, PROBAPI, PROBBLK, PROBOTH, and PROBWHT equals 100 for all people.

Almost all Hispanics are classified as white in RACESING. This includes respondents who checked the "other" box for the race question and wrote in a Hispanic race. It also includes respondents who checked the "other" box for the race question but did NOT write in a Hispanic race, as long as they identified themselves as Hispanic in the direct question about Hispanic status (only asked from 1980 on, see HISPAN). We chose to explicitly classify Hispanics as white in the years from 1980 onward because our historical Hispanic variable suggests that more than 95% of Hispanics were classified as white prior to 1970 (see HISPAN). The only Hispanic persons who are not coded as white in RACESING are those who explicitly reported their race as Black, American Indian/Alaska Native, any Asian or Pacific Islander group, and those who reported multiple races but ended up getting bridged to one of these three categories.

Details about how the RACE and RACESING codes are related is offered in the table, Relationship between RACE and RACESING codes. For single-race responses the relationship between RACE and RACESING codes is a straightforward data translation. For multiple-race responses, the page linked above shows the range of possible destination categories for each combination of racial responses. For example, a person who indicates that they are both Black and White (code 801 in the RACE variable) could either end up being classified as Black (code 20) or White (code 10) in the RACESING variable.

Comparability

RACESING is comparable across years. Users should note that race questions were not asked in the Puerto Rican censuses of 1970, 1980 and 1990. They were asked in the 1910 and 1920 Puerto Rican censuses, and in the 2000 Puerto Rican census and the PRCS.

Universe

Availability

• 2023: --
• 2022: --
• 2021: --
• 2020: --
• 2019: --
• 2018: --
• 2017: --
• 2016: --
• 2015: --
• 2014: ACS
• 2013: ACS
• 2012: All samples
• 2011: All samples
• 2010: ACS; ACS 3yr; ACS 5yr
• 2009: All samples
• 2008: All samples
• 2007: All samples
• 2006: All samples
• 2005: All samples
• 2004: All samples
• 2003: All samples
• 2002: All samples
• 2001: All samples
• 2000: All samples
• 1990: All samples
• 1980: All samples
• 1970: All samples
• 1960: All samples
• 1950: 1%
• 1940: 1%
• 1930: All samples
• 1920: All samples
• 1910: All samples
• 1900: All samples
• 1880: All samples
• 1870: All samples
• 1860: All samples
• 1850: All samples

• 2023: --
• 2022: --
• 2021: --
• 2020: --
• 2019: --
• 2018: --
• 2017: --
• 2016: --
• 2015: --
• 2014: --
• 2013: --
• 2012: --
• 2011: --
• 2010: --
• 2009: --
• 2008: --
• 2007: --
• 2006: --
• 2005: --
• 2000: --
• 1990: --
• 1980: --
• 1970: --
• 1930: All samples
• 1920: All samples
• 1910: All samples

Flags

Editing Procedure

There is no editing procedure available for this variable.