Sage Journals: Discover world-class research

Abstract

Previous studies have shown nonpartisan direct mail interventions have a small turnout effect. It is less clear, however, what effect these interventions have with regard to cognitive mobilization of young voters. We report the results of a large (N=52,324) experiment conducted during the February 2015 mayoral election in Chicago. A random sample of 22,179 registrants between 18 and 30 were sent a voter guide containing two sets of orienting political information for the five candidates: a list of their endorsements and their policy positions on five issues in the campaign. We find turnout among the treated group was 0.9 points higher than the control group which did not receive a guide (33.6% compared with 32.7%; SE is 0.003). We find, however, that treated registrants in above-median household income census tracts are significantly more likely to vote than treated registrants in lower-income census tracts.

Keywords

Turnout Chicago mobilization experiment

Low rates of participation in elections are often described as a chronic ailment of the American body politic.¹ There are, nonetheless, a variety of proposed interventions to increase participation in American elections. We experimentally assess the effect of an increase in political knowledge on turnout among young voters in a local election. In advance of the first round of the 2015 mayoral election in Chicago we mailed a guide listing endorsements and policy positions of each of the five candidates to 25,000 registrants between the ages of 18 and 30. After the election we located as many of these treated registrants and the pre-specified control group as possible in voter registration lists to measure the effect of the treatment of receiving a nonpartisan voter guide. We find, first, that this direct mail treatment significantly increased turnout by about 0.9 percentage points (the standard error around this treatment effect is 0.003). When we interact the mailer treatment with median income at the census tract level, however, we find the treatment had a significant effect on turnout only in census tracts with above-median household income.

We begin with some general observations about turnout in the United States, the potential for voting reforms to increase turnout, the confounding effects of cross-cutting information on political activity, and observe how this study contributes to the largely experimental literature of voter mobilization studies. We then describe the experimental design and our findings before making a few concluding remarks.

Turnout has markedly declined since the heyday of participation in the 1960s (Wattenberg, 2002; Teixeira, 1987). Presidential races generally attract the most attention—and the highest turnout—while local elections tend to feature comparatively low rates of participation (Hajnal and Lewis, 2003, Holbrook and Weinschenk, 2014; Oliver, 2012). Regardless of electoral context, turnout is low among people below the age of 30 (Gimpel et al., 2004; Highton and Wolfinger, 2001; Kaufmann et al., 2008; Leighley and Nagler, 2014; Shea and Green, 2006; Wattenberg, 2012). Focusing our experiment on young people in a local election serves to isolate registrants least likely to vote in an election already less likely to exhibit high rates of participation. This approach is not entirely misguided; we know voting is a habit-forming activity (Coppock and Green, 2015; García Bedolla and Michelson, 2012; Gerber et al., 2003; Green and Shachar, 2000; Plutzer, 2002) and thus this experiment may prove fruitful to demonstrate an intervention that pays long-term dividends in terms of turning young voters into frequent voters.

There are a wide array of reforms designed to make the casting of a ballot easier. In a review of the effectiveness of a variety of electoral reforms, Berinsky (2005) finds these institutional avenues of reform are limited. “The true costs of participation lie not just in the expression of opinion but also in the formation of political opinions. Existing electoral reforms do nothing to minimize the cognitive costs associated with voting” (Berinsky, 2005: 485, original emphasis). Cognitive mobilization, Berinsky argues, will increase interest in politics, a well-known predictor of participation. One study in this area (Wolfinger et al., 2005) finds states that provide pre-election sample ballots and information about polling places have higher turnout, particularly among young and less-educated registrants.² Our experiment is designed to test the proposition that a nonpartisan mailer that includes information related to endorsements and policy positions of mayoral candidates in Chicago will increase turnout among 18–30-year-old registered voters who are not voting for the first time.

This proposition is not universally supported in the literature. Mutz (2002, 2006) demonstrates that encountering political information and discourse contrary to one’s preferences can reduce participation in politics. Shi (2016) finds demobilization—on the order of 1 to 2 percentage points—follows from a mailed postcard expressing a view on same-sex marriage in North Carolina that is contrary to the recipient’s partisan affiliation. Furthermore, efforts to correct misperceptions based on predispositions are not always successful (Nyhan and Reifler, 2010). It may be the case that this mailed voter guide may reduce turnout among the treated observations. This experiment, in short, provides a context to shed light on the utility of objective information in spurring greater political participation along the lines of Berinsky’s call for greater cognitive mobilization.

Direct mail experiments and turnout

Random assignment to a treatment is a pillar of the causal empiricist approach to quantitative research (Samii, 2016). Voter mobilization experiments are one subfield in which this approach has taken root and clarified our understanding of the drivers of turnout. A central finding in this research agenda is that nonpartisan direct mail increases turnout by a significant—but small—amount (Green and Gerber, 2015; Green et al., 2013). The early experiment by Gosnell (1927) found, for example, a mailed reminder about an upcoming Election Day increased turnout in the 1925 municipal election in Chicago by 9 percentage points. Miller and colleagues (1981) find voters aged 21–30 were more likely to vote in a local primary election after receiving a written letter with campaign brochures. In a more recent study, Gerber and Green (2000a) found postcards encouraging voting in the 1996 general election increased turnout by 7 percentage points among unaffiliated voters and have no effect on partisans. An experiment conducted in the 1998 general election found direct mail increased turnout by about 0.6 percentage points (Gerber and Green, 2000b). As García Bedolla and Michelson observe, however, “we know little about the underlying mechanisms that make these short sociocultural exchanges so effective in changing voter behavior” (García Bedolla and Michelson, 2012: 11).

Experimental research designs have been used to estimate mobilization effects among targeted populations, like age and racial (or ethnic) groups. Sociocultural cognition—the claim that an individual’s sociocultural context has a role in cognitive outcomes—provides a theoretical link between mobilization experiments and observed outcomes (García Bedolla and Michelson, 2012). Nickerson (2006) finds young voters—once they are registered to vote—are equally responsive to appeals as other voters, but that it is about three times as difficult to contact a young voter. The extent to which minority populations can be mobilized by these appeals is a common topic in the literature, though significant effects are not frequently obtained (Green, 2004; Matland and Murray, 2012; Trivedi, 2005; Wong, 2005). García Bedolla and Michelson (2009), however, find direct mail appeals increase turnout by about 0.7 percentage points among Asian populations in California. The present study is designed to assess if, and to what degree, a mailed voter guide with information that helps orient a voter in relation to candidates in an election increases turnout. In that sense, the present study is most similar in design and intent to García Bedolla and Michelson (2009), though it differs in that the electoral context is a municipal election and the population of interest is young voters.

Experimental design

We assess the mobilizing effect of direct mail solicitations using a random sample of registered voters residing in Chicago. These data are drawn from a list of all registered voters in Chicago provided by a third-party vendor prior to the election (1,603,822 observations in total).³ For the purposes of our experiment we exclude the following observations from our analyses: those registrants who reside outside of the city of Chicago, or who are younger than 18 or older than 30. We excluded registrants who did not participate in the 2014 general election—but retained registrants who did not participate in the 2011 municipal elections—to identify moderate propensity voters, a population thought to be most responsive to the treatment we deployed in this experiment (Eldersveld, 1956; Highton, 2004; Zaller, 1992).⁴ These criteria exclude 1,543,308 observations originally in the dataset. We used a computer-based random number generator to assign 25,000 registrants from the remaining 60,514 observations to the treatment group, with the balance kept as a control group. In the post-election registration lists provided by the same vendor 8,190 observations—2,821 within the treated group and 5,369 within the control group—were lost to follow-up.⁵ In sum, the data for this experiment include 22,179 treated registrants and a control group of 30,145 observations.⁶ The dependent variable in our statistical models is a dichotomous indicator of voting in the 2015 mayoral election, as reported for each observation in the voter registration files.⁷

The control variables in our models are sourced from two datasets. First, the voter registration list itself includes individual-level data on a registrant’s gender, age, and race. To control for neighborhood-level effects on turnout (Cho et al., 2006; Gimpel et al., 2004) we collected the total population, median age, median household income, proportion of the population that is white, black, Latino, or other, and the proportion of the population in one of five educational attainment categories from the 2013 American Community Survey for each Chicago census tract in our data. We report variable balance across the control and treatment groups in Table 1. We find no significant differences across the experiment groups, demonstrating randomization in the assignment to the treatment.

Table 1.

Variable balance in control and treatment groups.

Variable	Control group (N=30,145)		Treatment group (N=22,179)
Variable	Mean	Std. Dev.	Mean	Std. Dev.
Individual-level variables
2015 mayoral election vote	0.327	0.469	0.336	0.473
Female	0.597	0.491	0.593	0.491
Age	25.44	3.41	25.48	3.43
Whites	0.489	0.5	0.492	0.5
Blacks	0.247	0.431	0.246	0.431
Latinos	0.138	0.345	0.136	0.343
Census tract variables
Log of population	8.2	0.5	8.2	0.49
Median age	34.09	5.8	34.07	5.81
White population %	47.27	34.52	47.25	34.48
Black population %	37.14	40.65	37.16	40.7
Latino population %	21.02	26.3	21.09	26.4
Population % with a high school diploma	21.86	12.12	21.79	12.05
Population % with some college	23.81	10.11	23.84	10.08
Population % with a college degree	22.23	15.33	22.26	15.28
Population % with a professional degree	15.51	13.63	15.49	13.56
Median household income	30.54	15.79	30.38	15.67

Notes: Hotelling’s two-group F test statistic (1.14) is nonsignificant (p<0.31) indicating the control and treatment groups are not significantly different.

The treatment group was mailed an informational booklet in advance of the election.⁸ This treatment condition is intended to test the idea, presented by Berinsky (2005), that a barrier to participation in elections is low levels of political information among potential voters. The English-language mailer contained two relevant sets of political information. First, it listed endorsements from prominent groups and individuals for four of the five candidates for mayor.⁹ Previous research has shown endorsements are helpful heuristics for voters in cases where party affiliation is unavailable (Bowler and Donovan, 1998; Schaffner et al., 2001). Mayoral elections in Chicago have been contested on a nonpartisan basis since 1999. Second, the mailer reported the position of each of the five candidates for mayor with regard to five salient policy issues in the campaign.¹⁰ This information was intended to aid voters in finding a candidate who shares their policy preferences (Lau and Redlawsk, 1997, 2006). The guide itself did not convey polling place locations—instead it directed the reader to the Chicago Elections Board website. The guide mentions the span of the early voting window, the date of the election, and advised registrants to bring some form of acceptable identification to the polls to avoid any potential problem with ballot access.¹¹

Our hypothesis that the mailer will increase turnout in the election will be confirmed, then, if our analyses report a positive and significant result for the treated observations. We use logit and multilevel models (Cho et al., 2006; Gelman and Hill, 2006; Gimpel et al., 2004; Steenbergen and Jones, 2002) to estimate the effects of this treatment condition.¹²

Findings

Our findings demonstrate the treatment increased turnout in the Chicago municipal election, all else being equal. Table 2 contains the results of four statistical models; the coefficients are reported as log odds in each specification. We first discuss the results in the first three models. The first and least complicated of these models includes only the treatment variable. As we observed in Table 1, there is no significant difference between the control and treatment groups with regard to the battery of explanatory variables. Model 1 shows the voter guide significantly increased turnout in the mayoral election. The treatment effect remains positive and significant as additional individual-level controls are incorporated (see Model 2) and when we estimate a multilevel model of individuals within Chicago census tracts and include demographic data from each census tract (see Model 3). Black registrants are least likely to vote, relative to other racial or ethnic groups, but this difference, with regard to Latino registrants, vanishes in Model 3.¹³

Table 2.

Turnout effects of voter guide.

Predicting voting in 2015 mayoral election	Model 1	Model 2	Model 3	Model 4
Individual variables
Voter guide treatment	0.042**	0.043**	0.043**	−0.017
Voter guide treatment	(0.019)	(0.020)	(0.020)	(0.038)
Female registrant		−0.096***	−0.074***	−0.075***
Female registrant		(0.019)	(0.019)	(0.019)
Age (18–30)		0.001	−0.001	−0.001
Age (18–30)		(0.004)	(0.003)	(0.003)
White registrant		0.422***	0.067*	0.066*
White registrant		(0.035)	(0.036)	(0.036)
Latino registrant		0.737***	0.076	0.079*
Latino registrant		(0.040)	(0.048)	(0.048)
Other registrant		0.523***	0.097**	0.096**
Other registrant		(0.042)	(0.045)	(0.045)
Census tract variables
Log of population			−0.100***	−0.105***
Log of population			(0.031)	(0.030)
Median age			0.017***	0.014***
Median age			(0.003)	(0.003)
White population %			0.000	0.001
White population %			(0.001)	(0.001)
Black population %			−0.007***	−0.005***
Black population %			(0.001)	(0.001)
Latino population %			0.007***	0.006***
Latino population %			(0.001)	(0.001)
% with high school diploma			0.003	0.001
% with high school diploma			(0.003)	(0.003)
% with some college			0.012***	0.006**
% with some college			(0.003)	(0.003)
% college graduates			−0.000	−0.000
% college graduates			(0.003)	(0.003)
% graduate degrees			0.006*	0.005
% graduate degrees			(0.004)	(0.003)
Median household income			−0.008
Median household income			(0.027)
2^nd household income quartile				0.050
2^nd household income quartile				(0.054)
3^rd household income quartile				0.142**
3^rd household income quartile				(0.067)
Top household income quartile				−0.217**
Top household income quartile				(0.089)
Guide * 2nd household income quartile				0.006
Guide * 2nd household income quartile				(0.054)
Guide * 3^rd household income quartile				0.094*
Guide * 3^rd household income quartile				(0.053)
Guide * top household income quartile				0.125**
Guide * top household income quartile				(0.058)
Constant	−0.721***	−0.552***	−0.745**	−0.506
Constant	(0.022)	(0.095)	(0.370)	(0.369)
Observations	52,324	52,324	52,324	52,324
Log pseudolikelihood	−33,221	−32,888	−32,322	−32,294
Wald chi-square	4.686**	393.2***	864.2***	990.9***
Number of census tracts			794	794

Notes: Robust standard errors clustered by census tract are in parentheses. Cells report logit coefficients. The reference racial/ethnic group is black registrants. The income quartiles in Model 4 are in reference to the 1st (bottom) household income quartile.*** p<0.01, ** p<0.05, * p<0.1.

We use the results in Model 2 to estimate the change in the predicted probability of voting in the 2015 mayoral election to get a better sense of the effect of the voter guide, as logit coefficients are difficult to interpret directly. Figure 1 shows the predicted probability of voting, and 95% confidence intervals, for the treatment and control groups. Turnout among the treated observations was about 0.9 points higher than in the control group, 33.6% compared with 32.7% (the standard error is 0.003). A difference of means test shows this absolute difference in turnout is statistically significant (two-tailed p<0.025). The relative change in the probability of voting among the treated observations, which can be obtained by exponentiating the treatment coefficient in Table 2, is about 4.4%.

Figure 1.

Predicted probability of voting in experiment groups.

This effect is large relative to the mobilization literature. Green and Gerber (2015: 186–189) report, on the basis of a meta-analysis of 85 studies between 1998 and 2014, the average treatment effect of a nonpartisan mailer without an element of social pressure is 0.523 points. If we limit our consideration to the seven studies in the meta-analysis conducted in the context of a municipal election, the average treatment effect of a mailer is 0.04 points. When compared with other mobilization strategies specifically targeting young voters, these results are more cost-effective than other direct mail experiments, but less effective than phone banking or canvassing and leafleting efforts (see Young Voter Strategies 2006 for a summary of this research).

Is this increase in turnout constant across all observations? We know, for example, that wealthier and better educated citizens are more politically active (Delli Carpini and Keeter, 1997) and more likely to vote (Leighley and Nagler, 2014). Green and Gerber (2015: 63) also find mailers have weak turnout effects among low-propensity voters. It may be the case, then, that the mailer did not increase the likelihood of voting among observations in less wealthy neighborhoods of Chicago. Model 4 presents subgroup analyses of the treatment effect by each quartile of the median household income quartiles in the 794 census tracts in these data. Interestingly, the probability of voting is highest among observations in the 3rd income quartile and lowest among observations in the top quartile, when compared with the reference bottom income quartile. The effect of the voter guide treatment is also only discernible among the top two income quartiles. These results suggest the effect on turnout is limited to observations in more affluent census tracts in Chicago.¹⁴ Additional analyses are reported in the online appendix.

Last, we assess the cost-effectiveness of this mailer. Green and Gerber (2015: 16) state “[i]n order to know whether a campaign tactic is cost-effective, it is necessary to determine how many votes are produced for each dollar spent.” We estimate the cost-effectiveness of the mailer, on an intent-to-treat basis, is about $56 (or about 111 mailers were required to produce a single vote). The effectiveness of this mailer is quite high relative to other studies considered by Green and Gerber, who find “one additional vote is generated for every 273 people who receive a conventional nonadvocacy GOTV mailer” and that mailers with a production cost of $0.50 cost $91 per vote (Green and Gerber, 2015: 66).

Conclusions

We sought to empirically test Berinsky’s (2005) suggestion that cognitive mobilization is the key to increasing participation in American elections. We narrowed our study to young voters in the 2015 mayoral election in Chicago. Our results indicate the mailed voter guide significantly increased turnout. When we account for income levels in different parts of Chicago, however, we find the treatment only increased turnout among wealthier neighborhoods in the city. We find, like much of the literature on minority mobilization, no significant effect across racial or ethnic groups. The scale of the effect of an increase in political information reported here is similar to recent studies of direct mail (García Bedolla and Michelson, 2009; Gerber and Green, 2000b) though the context of a local election and the focus on young voters is distinctive.

This study suggests additional experimental designs that may shed further light on the role of cognitive mobilization with regard to voter turnout. The present study is limited to young voters with some prior experience with voting. It is an open question to what degree a similar experiment may mobilize newly registered voters with no prior voting experience. Moreover, a similar experimental framework in conjunction with pre- and post-election surveys could tell us something about the potential for additional political information to change a voter’s choice in the course of an election campaign. Replicating this experiment in a variety of electoral contexts, for example open seat races, when an incumbent fails to win reelection, or in localities with a tradition of high turnout in local elections could demonstrate the robustness of these findings and further assess the effects of cognitive mobilization. This experiment only tests the effect of direct mail. Future work should take account of multiple “treatment channels” (Green and Gerber, 2015; Miller et al., 1981) to generate cognitive mobilization and therefore turnout. This study—and other experimental designs—have shown chronic low turnout in American elections by young voters is a treatable condition.

Footnotes

Our thanks to Kevin Arceneaux,Maneesh Arora,Josh Berezin,Neil Chaturvedi,Donald Green,Dan Hopkins,Kerem Kalkan,and Henry Kraemer for helpful comments on an earlier draft of this project. Peter Miller is the Fulbright-University of Tampere Scholar. Rebecca Reynolds is a past executive director of Chicago Votes. Matthew Singer is a past executive director of the Alliance for Youth Action. The dataset used in this analysis,the replication files,and the voter guide are stored at a Harvard Dataverse repository available here:

Correction (June 2025):

The article has been updated with correct dataverse link in the supplementary material section. For more details,please see the correction notice

Authors’ Note

This project was previously presented at a Behavioral Ethics Lab seminar at the University of Pennsylvania,at a Global Studies Research Seminar at the University of Tampere,and at a meeting of the Midwest Political Science Association.

Declaration of Conflicting Interest

The authors declare that there is no conflict of interest.

Funding

This research received no specific grant from any funding agency in the public,commercial,or not-for-profit sectors.

Supplementary Material

The supplementary files are available at http://journals.sagepub.com/doi/suppl/10.1177/2053168017738410 . The replication files are available at:

Carnegie Corporation of New York Grant

This publication was made possible (in part) by a grant from Carnegie Corporation of New York. The statements made and views expressed are solely the responsibility of the author.

References

Berinsky

(2005) The perverse consequences of electoral reform in the United States. American Political Research 33: 471–491.

Bowler

Donovan

(1998) Demanding Choices: Opinion, Voting, and Direct Democracy. Ann Arbor: University of Michigan Press.

Brady

McNulty

(2011) Turning out to vote: The costs of finding and getting to the polling place. American Political Science Review 105: 115–134.

Cho

Gimpel

Dyck

(2006) Residential concentration, political socialization, and voter turnout. Journal of Politics 68: 156–167.

Coppock

Green

(2015) Is voting habit forming? New evidence from experiments and regression discontinuities. American Journal of Political Science 60: 1044–1062.

Delli Carpini

Keeter

(1997) What Americans Know about Politics and Why it Matters. New Haven: Yale University Press.

Eldersveld

(1956) Experimental propaganda techniques and voting behavior. American Political Science Review 50: 154–165.

García Bedolla

Michelson

(2012) Mobilizing Inclusion: Transforming the Electorate through Get-Out-the-Vote Campaigns. New Haven: Yale University Press.

García Bedolla

Michelson

(2009) What do voters need to know? Testing the role of cognitive information in Asian American voter mobilization. American Politics Research 37: 254–274.

10.

Gelman

Hill

(2006) Data Analysis Using Regression and Multilevel/Hierarchical Models. New York: Cambridge University Press.

11.

Gerber

Green

(2000a) The effect of a nonpartisan get-out-the-vote drive: An experimental study in leafletting. Journal of Politics 62: 846–857.

12.

Gerber

Green

(2000b) The effects of canvassing, telephone calls, and direct mail on voter turnout: A field experiment. American Political Science Review 94: 653–663.

13.

Gerber

Green

Shachar

(2003) Voting may be habit-forming: Evidence from a randomized field experiment. American Journal of Political Science 47: 540–550.

14.

Gimpel

Dyck

Shaw

(2004) Registrants, voters, and turnout variability across neighborhoods. Political Behavior 26: 343–375.

15.

Gosnell

(1927) Getting out the Vote: An Experiment in the Stimulation of Voting. Chicago: University of Chicago Press.

16.

Green

(2004) Mobilizing African-American voters using direct mail and commercial phone banks: A field experiment. Political Research Quarterly 57: 245–255.

17.

Green

Gerber

(2015) Get Out the Vote: How to Increase Voter Turnout. 3rd edition. Washington: Brookings Institution Press.

18.

Green

Shachar

(2000) Habit formation and political behavior: Evidence of consuetude in voter turnout. British Journal of Political Science 30: 561–573.

19.

Green

McGrath

Aronow

(2013) Field experiments and the study of voter turnout. Journal of Elections, Public Opinion, and Parties 23: 27–48.

20.

Hajnal

Lewis

(2003) Municipal institutions and voter turnout in local elections. Urban Affairs Review 38: 645–668.

21.

Highton

(2004) Voter registration and turnout in the United States. Perspectives on Politics 2: 507–515.

22.

Highton

Wolfinger

(2001) The first seven years of the political life cycle. American Journal of Political Science 45: 202–209.

23.

Holbrook

Weinschenk

(2014) Campaigns, mobilization, and turnout in mayoral elections. Political Research Quarterly 67: 42–55.

24.

Kaufmann

Petrocik

Shaw

(2008) Unconventional Wisdom: Facts and Myths About American Voters. New York: Oxford University Press.

25.

Lau

Redlawsk

(1997) Voting correctly. American Political Science Review 91: 585–598.

26.

Lau

Redlawsk

(2006) How Voters Decide: Information Processing in Election Campaigns. Cambridge: Cambridge University Press.

27.

Leighley

Nagler

(2014) Who Votes Now? Demographics, Issues, Inequality, and Turnout in the United States. Princeton: Princeton University Press.

28.

Matland

Murray

(2012) An experimental test of mobilization effects in a Latino community. Political Research Quarterly 65: 192–205.

29.

McDonald

Popkin

(2001) The myth of the vanishing voter. American Political Science Review 95: 963–974.

30.

Miller

Bositis

Baer

(1981) Stimulating voter turnout in a primary: Field experiment with a precinct committeeman. International Political Science Review 2: 445–460.

31.

Mutz

(2002) The consequences of cross-cutting networks for political participation. American Journal of Political Science 46: 838–855.

32.

Mutz

(2006) Hearing the Other Side: Deliberative Versus Participatory Democracy. Cambridge: Cambridge University Press.

33.

Nickerson

(2006) Hunting the elusive young voter. Journal of Political Marketing 5: 47–69.

34.

Nyhan

Reifler

(2010) When corrections fail: The persistence of political misperceptions. Political Behavior 32: 303–330.

35.

Oliver

(2012) Local Elections and the Politics of Small-Scale Democracy. Princeton: Princeton University Press.

36.

Primo

Jacobsmeier

Milyo

(2007) Estimating the impact of state policies and institutions with mixed-level data. State Politics and Policy Quarterly 7: 446–459.

37.

Plutzer

(2002) Becoming a habitual voter: Inertia, resources, and growth in young adulthood. American Political Science Review 96: 41–56.

38.

Samii

(2016) Causal empiricism in quantitative research. Journal of Politics 78: 941–955.

39.

Schaffner

Streb

Wright

(2001) Teams without uniforms: The nonpartisan ballot in state and local elections. Political Research Quarterly 54: 7–30.

40.

Shi

(2016) Cross-cutting messages and voter turnout: Evidence from a same-sex marriage amendment. Political Communication 33: 433–459.

41.

Shea

Green

(eds) (2006) Fountain of Youth: Strategies and Tactics for Mobilizing America’s Young Voters. Washington: Rowman & Littlefield.

42.

Squire

Wolfinger

Glass

(1987) Residential mobility and voter turnout. American Political Science Review 81: 45–66.

43.

Steenbergen

Jones

(2002) Modeling multilevel data structures. American Journal of Political Science. 46: 218–237.

44.

Teixeira

(1987) Why Americans Don’t Vote: Turnout Decline in the United States 1960–1984. Westport: Greenwood Press.

45.

Trivedi

(2005) The effect of identity-based GOTV direct mail appeals on the turnout of Indian Americans. Annals of the American Academy of Political and Social Science 601: 115–122.

46.

Wattenberg

(2002) Where Have all the Voters Gone? Cambridge: Harvard University Press.

47.

Wattenberg

(2012) Is Voting for Young People? 3rd edition. Boston: Pearson.

48.

Wolfinger

Highton

Mullin

(2005) How postregistration laws affect the turnout of citizens registered to vote. State Politics and Policy Quarterly 5: 1–23.

49.

Wong

(2005) Mobilizing Asian American voters: A field experiment. Annals of the American Academy of Political and Social Science 601: 102–114.

50.

Young Voter Strategies (2006) Young Voter Mobilization Tactics. Available at: http://civicyouth.org/PopUps/Young_Voters_Guide.pdf (last accessed 2 May 2017).

51.

Zaller

(1992) The Nature and Origins of Mass Opinion. Cambridge: Cambridge University Press.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.32 MB

0.00 MB