18
ENDNOTES
1
The CPS uses a sample in which one respondent reports for all members of the household. This proxy reporting
seems not to affect estimates of turnout (J. Jennings 1990).
2
The choice of words in this sentence illustrates two data management decisions that differ from those made by the
Census Bureau in its biennial reports on registration and voting: 1) We deleted cases where information on registration
and voting was not ascertained, while the Census Bureau coded them as nonvoters. 2) Our analysis is confined to
citizens. Among other advantages, this precludes substantially underestimating the participation of Latinos and Asian-
Americans, not to mention turnout in states—like California—where these groups are a significant proportion of the
voting-age population but a smaller share of adult citizens (Citrin and Highton 2002).
3
The “non-response rate” for the basic November CPS was 7.5 percent; an additional 5.8 percent failed to respond to
the VRS (U.S. Department of Commerce 2001, 17-2).
4
Of the 2,982 people selected in 2000, the NES completed pre- and post-election interviews with 1,555. Completion
rates in 2000 for both surveys were lower than in the last decades of the 20
th
century: 95 percent for the CPS and 70
percent for the National Election Studies (Brehm 1993, p. 16).
5
Our results differ from those of Stoker and Jennings (1995) perhaps because of sample size limitations or cohort
variation. They studied participation in the 1980 presidential election with a sample of high school seniors drawn in
1965, while we analyze the 2000 contest with a large and nationally representative sample.
6
A generation earlier, one scholar attributed higher turnout by married people to “. . . similar motivations, outside
stimuli, and social norms which affect husband and wife simultaneously but independently . . . . interaction within the
family is probably an additional force which supplements the other predispositions of the individual members.” (Glaser
1959, p. 566).
Convention