Abstract
Over the course of the last three presidential elections, young voters and minorities have become an increasingly sought after segment of the electorate. In particular, youth are also the most likely to be the beneficiaries of advances to technology, while minorities often lag behind in access. The 2008 election provided a number of examples of campaigns utilizing online technology as a means of targeting young voters. This article examines the influence of the Internet on young and minority voters, focusing specifically on Internet use for political purposes, such as visiting a candidate's Web site or engaging in political discussions on blogs, as predictors of youth political participation offline. The authors find that engaging politics online leads to increases in political participation offline and that among younger voters, racial minorities are as connected as whites. However, among older voters, whites are far more likely to have access to and use the Internet politically.
Notes
Note. Standard errors are in parentheses.
+Significant at 10%; *significant at 5%; **significant at 1%.
Dependent variable is 7-item participation index: attending a political meeting, volunteering for a political party or campaign, donating money to a political party or campaign, attending/participating in a protest, persuading friends/family/others about politics, writing a letter to a political official, and voting in the 2008 election. Ordinary least squares regression coefficients and standard errors reported.
Note. Standard errors are in parentheses.
+Significant at 10%; *significant at 5%; **significant at 1%.
Note. Standard errors are in parentheses.
+Significant at 10%; *significant at 5%; **significant at 1%.
Though in 1996 both Bill Clinton and Bob Dole had Web sites, they were hardly used or mentioned in the media, other than for their novelty.
Asian youths are excluded here from the Pew analysis because of their small numbers; they account for only 4 percent of the millennial population (Pew 2010Please add the Pew 2010 citation to the reference list or delete the citation).
The Asian American sample includes the six largest national origin groups: Chinese, Asian Indian, Filipino, Korean, Vietnamese, and Japanese.
States were identified at the time of the sample frame design as battleground, in September 2008.
A cell phone–only sample was not conducted; however, some registered voters have provided their cell phone number on the voter registration record and those numbers were eligible to be dialed.
We rely on tables 2 and 6 from the 2008 CPS, which can be found here: http://www.census.gov/population/www/socdemo/voting/cps2008.html.
We additionally ran Poisson regression models in consideration of the “count” nature of our dependent variable and found the results to be almost identical.
Also, to assess potential endogeneity between political interest and online political activity, we ran a two-stage least squares model in which we first modeled online activity as a function of political interest. Those results are consistent with those presented here and can be found in the Appendix.
Though Tolbert and McNeal's paper models both vote propensity and participation, the CMPS is a sample of registered voters across the nation across a single election. The distribution of voters versus nonvoters in this particular sample was highly skewed: 207 respondents did not vote, while the remaining 4,356 claim to have voted. Data limitation also limited our ability to incorporate the environmental variables (minority diversity index and number of initiatives) used by Tolbert. Nonetheless, given that this is a single election analysis, these variables would not have contributed to our interpretation as the authors utilized them to illustrate change over election cycles.
Excluded from this index but present in Tolbert and McNeal's political participation index are whether the respondent displayed political buttons or signs and whether he or she gave money to interest groups. These are items that are unavailable on the CMPS, but given the amount of detail available regarding the online political behavior of respondents, the loss of these two items in the dependent variable is negligible.