Discrepancies between 1980 data: How can in-migrants be larger than total population?


I’m working with some 1980 data on black immigrants in counties in the Washington, DC metropolitan area. I was trying to create a count of the number of immigrants who identified their race as black, and initially the only option I saw was table NT004, Black Allocated In-Migrants by Country of Birth by Sex. I used this, but then realized I was looking at the wrong tables/wrong universe and needed the detailed breakout. I proceeded to download table NTPB25, Nativity and Place of Birth.

However, once I compared the values, I saw that for several counties, the values for NT004 were a lot larger than the values for NTPB25. How can the number of foreign born black in-migrants be larger than the value of TOTAL black immigrants living in the county?


It appears that in 1980 the “Black Allocated…” table refers to the fact that respondents who selected “Other” for race were then recoded to either Black or White.

“Many more persons of Spanish origin marked the category “Other” (race not specified) in 1980 than in 1970. These persons were transferred from the “Other” category to White or Black according to the distribution of persons of the same Spanish origin subgroup (Mexican, Cuban, Puerto Rican, other Spanish) who originally did specify White or Black…” (see 1980 census adjustements section of Intercensal Population Estimates, Bureau of the Census, 1970-1980).