The Effects of the Differential Undercount on Survey Estimates of Race Differences in Marriage

Abstract

The fact that Blacks, and especially Black men, are not as well represented in survey data as Whites is well known among demographers. What is less well known is the potential impact of the differential undercount on estimates of race differences in various demographic behaviors. This research examines the potential effects of the differential undercount by race and marital status on estimates of race differences in marriage. The results show that the differential undercount has a surprisingly small effect on population distributions and on life-table estimates of race differences in marriage formation. The author concludes that underrepresentation of unmarried Black men in survey data cannot explain why the race gap in marriage is smaller among men than among women.