Skip to main content
SpringerLink
Log in

Testing the Reciprocal Effects of Campaign Participation

  • Original Paper
  • Published:
Political Behavior Aims and scope Submit manuscript

Abstract

Questions persist regarding the robustness of cross-sectional estimates of effects of variables that are themselves endogenous to the participation process. On one hand, the consequences of working on a campaign have interesting implications for democratic society. Less benign, however, is the possibility that failure to control for reciprocal processes leads to biased estimates of the causes of campaign participation. I use a panel of Democratic and Republican contributors interviewed following each of the past three presidential elections (1996, 2000, and 2004) to explore the relationships between campaign participation and three variables typically parameterized as predictors of participation: receiving a contact, ideological extremism, and strength of party identification. The effect of strength of party identification on campaign participation proves robust; however, I find that nearly all of the associations between contacts and participation and ideological extremism and participation appear to extend from, not into, participation and past participation.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Subscribe and save

Springer+ Basic
$34.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or eBook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1

Similar content being viewed by others

Notes

  1. The data come from Walter J. Stone’s and Ronald B. Rapoport’s, Party Leadership and Presidential Selection Survey. Prior to the 1996 presidential election, they obtained samples of contributors to the Democratic and Republican National Committees who had received a direct mail appeal between October 1994 and October 1995 (the average contributor in these groups contributed less than $50). These were surveyed following the elections in 1996, 2000, and again in 2004. To be sure, not every activist contributes to the national parties, but it is difficult to conceive of a better group to sample if one wants a sample that is representative of activists in the mass public. Convention delegates, for example, are probably more involved than activists in the mass public, on average, because being a delegate requires an arduous selection procedure and convention attendance requires expensive travel. Lists from individual candidates enumerate activists in the mass public, but the groups may be unique in idiosyncratic ways related to the specific candidates. In short, the Party Leadership and Presidential Selection Survey follows a unique approach to gain new insights into a group that is difficult to enumerate.

  2. See Koger et al. (2005) for an interesting study of the dissemination of donors’ contact information.

  3. Political experience was proxied by age. See Converse (1969, 1976).

  4. In addition, the three-wave models enable me to study both the reciprocal and cross-lagged effects while also correcting for measurement error.

  5. Year

    DNC N

    DNC response rate

    RNC N

    RNC response rate

    1996

    666

    46.4%

    789

    39.9%

    2000

    419

    62.9%

    494

    62.6%

    2004

    167

    39.9%

    216

    43.7%

    1. For an analysis of attrition and potential response bias, see Appendix A. Lisrel’s full information maximum likelihood estimation option enables me to use all respondents who participated in all three waves of the study, for an N of 383. While which party individuals identify with is not a factor in my models, I note that the ratio of Democrats to Republicans in these data approximates the analogous ratio of Democratic contributors to Republican contributors in the NES (in most presidential election years, 1952–2004). I pool them in my analyses, but I note that the partisan balance in this sample reflects the partisan balance of individual contributors in the electorate. For more information see The Presidential Selection Study website: http://pss.iga.ucdavis.edu/

  6. While the combined scale no doubt misses some differences in the determinants of each act, I ran separate models for “convincing friends” and “contributing money” and the estimates were very similar to the models based on the combined scale. I report only the results of the combined scale analyses because the underlying processes that I study do not appear to differ dramatically across the various acts and because total involvement captures variation in participation in ways that separate models of each act would miss. For example, an individual could forego contributing money one year, but still increase participation by adding other acts.

  7. See Appendix B for descriptive statistics.

  8. The error variances of the observed indicators are represented by the ɛ terms and the structural disturbances of the latent variables are represented by the ζ terms. I follow the single indicator procedure for three-wave data described in Wiley and Wiley (1970). Wiley and Wiley identify the single indicator model by constraining the error variances to equality in each wave ɛ1 = ɛ2 = ɛ3 and ɛ4 = ɛ5 = ɛ6 and setting each of the structural coefficients linking the latent variables to their observed indicators to one (which also ensures the units of the latent variables retain the units of the observed indicators). In addition, I impose several “consistency” constraints to gain additional degrees of freedom. Finkel (1995) recommends setting the structural coefficients to equality when panel waves are equally spaced (as these are). For example, I set the structural coefficient linking c00  →  p00 equal to that of c04  →  p04. I do the same for the lagged effects (e.g. c96  →  c00 is set equal to c00  →  c04), the cross-lagged effects, and the reciprocal effects.

  9. Finkel (1995) describes this approach to handling measurement error problems,

    When an indicator with random measurement error is used as an independent variable in a regression equation, the result is biased estimation of the true causal effect of the latent variable…[a]lthough measurement error leads to serious problems in panel analysis, it can also be handled much more easily than in the cross-sectional context… In panel designs, the repeated measurement of the indicators over time increases the power of this [error correction] approach considerably, as additional waves of data provide more information with which to estimate relevant structural and measurement coefficients. In fact, measurement properties and structural effects in models with only one indicator of a latent variable can be estimated with at least three waves of data, and thus all multiwave panel models can be treated as variants of the multiple indicators [error correction] approach (47–49).

  10. Fit for each model is measured by the chi-square statistic, with higher values (relative to the number of degrees of freedom) indicating worse fit (see Finkel 1995). The models are compared by subtracting the chi-square of the unconstrained model from that of the constrained model. The degrees of freedom for this comparison are equal to the difference of the degrees of freedom of the two models (Bollen 1989). The null hypothesis in these tests is that model fit for both the constrained and unconstrained models are equal.

  11. It might also be worth noting that these problems are unique to activism—the contact questions probably work quite well in terms of turnout because voting does not involve extensive interaction with a party or candidate.

References

  • Abramson, P. R., & Claggett, W. (2001). Recruitment and political participation. Political Research Quarterly, 54(4), 905–916.

    Google Scholar 

  • Brady, H. E., Schlozman, K. L., & Verba, S. (1999). Prospecting for participants: Rational expectations and the recruitment of political activists. American Political Science Review, 93(1), 153–168.

    Article  Google Scholar 

  • Bollen, K. A. (1989). Structural equations with latent variables. New York: John Wiley & Sons, Inc.

    Google Scholar 

  • Campbell, A., Converse, P. E., Miller, W. E., & Stokes, D. E. (1960). The American voter. New York: John Wiley & Sons, Inc.

    Google Scholar 

  • Carmines, E. G., & Stimson, J. A. (1989). Issue evolution: Race and the transformation of American politics. Princeton: Princeton University Press.

    Google Scholar 

  • Carmines, E. G., & J. Woods (2002). The role of party activists in the evolution of the abortion issue. Political Behavior, 24, 2002.

  • Carsey, T. M., & Layman, G. C. (1999). A dynamic model of political change among party activists. Political Behavior, 21, 17–41.

    Article  Google Scholar 

  • Claggett, W. (1981). Partisan acquisition versus partisan intensity: Life-cycle, generation, & period effects, 1952–1976. American Journal of Political Science, 25, 193–214.

    Article  Google Scholar 

  • Converse, P. E. (1969). Of time and partisan stability. Comparative Political Studies, 2, 139–71.

    Article  Google Scholar 

  • Converse, P. E. (1976). The dynamics of party support: Cohort-analyzing party identification. Beverly Hills, CA: Sage.

    Google Scholar 

  • Finkel, S. (1985). Reciprocal effects of participation and political efficacy: A panel analysis. American Journal of Political Science, 29, 891–913.

    Article  Google Scholar 

  • Finkel, S. (1995). Causal analysis with panel data. London: Sage Publications.

    Google Scholar 

  • Freie, J. F. (1997). The effects of campaign participation on political attitudes. Political Behavio, 19(2), 133–156.

    Article  Google Scholar 

  • Gershtenson, J. (2002). Political participation in campaign activities, 1952–1996. Political Research Quarterly, 55, 687–714.

    Google Scholar 

  • Grant, J. T., & Rudolph, T. J. (2002). To give or not to give: Modeling individuals contribution decisions. Political Behavior, 24(1), 31–54.

    Article  Google Scholar 

  • Green, J. C., & Guth, J. L. (1989). The missing link: Political activists and support for school prayer. The Public Opinion Quarterly, 53, 41–57.

    Article  Google Scholar 

  • Isenberg, D. J. (1986). Group polarization: A critical review and meta-analysis. Journal of Personality and Social Psychology, 50, 1141–1151.

    Article  Google Scholar 

  • Koger, G., S. Masket, & H. Noel. (2005). We appreciate your support: Information exchange and party networks. Paper presented at the annual meeting of the American Political Science Association.

  • Leighley, J. E. (1995). Attitudes, opportunities and incentives: A field essay on political participation. Political Research Quarterly, 48(1), 181–209.

    Google Scholar 

  • McCann, J. A. (1995). Nomination politics and ideological polarization: Assessing the attitudinal effects of campaign involvement. Journal of Politics, 57(1), 101–120.

    Article  Google Scholar 

  • Miller, W. E., & Jennings, M. K. (1986). Parties in transition: A longitudinal study of party elites and party supporters. New York: Sage.

    Google Scholar 

  • Mutz, D. (2006). Hearing the other side: Deliberative versus participatory democracy. Cambridge: Cambridge University Press.

    Google Scholar 

  • Rapoport, R. B., & Stone, W. J. (1994). A model for disaggregating political change. Political Behavior, 16(4), 505–532.

    Article  Google Scholar 

  • Rapoport, R. B., & Stone, W. J. (2005). Three’s a crowd: The dynamic of third parties, ross perot, and republican resurgence. Ann Arbor: University of Michigan Press.

    Google Scholar 

  • Rosenstone, S. J., & Hansen, J. M. (1993). Mobilization, participation, and democracy in America. New York, NY: Macmillan.

    Google Scholar 

  • Saunders, K. L., & Abramowitz, A. I. (2004). Ideological realignment and active partisans in the american electorate. American Politics Research, 32, 285–309.

    Article  Google Scholar 

  • Sunstein, C. R. (2000). Deliberative trouble? Why groups go to extremes. The Yale Law Journal, 110, 71–119.

    Article  Google Scholar 

  • Sunstein, C. R. (2002). The law of group polarization. The Journal of Political Philosophy, 10, 175–195.

    Article  Google Scholar 

  • Verba, S., & Nie, N. H. (1972). Participation in America: Political democracy and social equality. New York: Harper & Row.

    Google Scholar 

  • Verba, S., Schlozman, K. L., & Brady, H. E. (1995). Voices and equality: Civic voluntarism in American politics. Cambridge, MA: Harvard University Press.

    Google Scholar 

  • West, D. M. (1988). Activists and economic policymaking in congress. American Journal of Political Science, 32, 662–680.

    Article  Google Scholar 

  • Wiley, D. E., & Wiley, J. A. (1970). The estimation of measurement error in panel data. American Sociological Review, 35, 112–117.

    Article  Google Scholar 

  • Wong, J. S. (2000). The effects of age and political exposure on the development of party identification among asian americans and latino immigrants in the United States. Political Behavior, 22, 341–371.

    Article  Google Scholar 

Download references

Acknowledgements

This research would not be possible without data from The Party Leadership and Presidential Selection Survey. The author thanks Walt Stone and Ron Rapoport for making their study available and for advice and assistance.

Author information

Authors and Affiliations

  1. Department of Political Science, Kent State University, P.O. Box 5190, Kent, OH, 44242-0001, USA

    Ryan L. Claassen

Authors
  1. Ryan L. Claassen
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Ryan L. Claassen.

Appendices

Appendix A: Analyzing Attrition

In Table A1 I describe respondents and nonrespondents, on each of my four variables, in the 2000 wave and the 2004 wave, by comparing their baseline 1996 responses. Fortunately, I only detect a potential problem with one of the eight comparisons. The 95% confidence intervals associated with the mean 1996 participation reports of respondents in 2000 compared to the mean 1996 participation reports of nonrespondents in 2000, do not overlap. However, the difference of the two means is only .40. While the difference is statistically significant, on a scale with a theoretical maximum of 18, the difference is less significant substantively. Further mollifying concern about response bias, the confidence intervals associated with the mean 1996 participation reports of respondents in 2004 compared to the mean 1996 participation reports of nonrespondents in 2004, do overlap. Finally, I detect no statistically significant response bias on the other variables.

Variable

Response group

Mean

95% CI

N

1996 participation*

2000 nonrespondent

2.04

1.88–2.21

542

2000 respondent

2.44

2.29–2.59

913

1996 participation

2004 nonrespondent

2.21

2.07–2.33

1072

2004 respondent

2.53

2.31–2.76

383

1996 contact

2000 nonrespondent

.75

.71–.79

542

2000 respondent

.79

.77–.82

913

1996 contact

2004 nonrespondent

.78

.75–.80

1072

2004 respondent

.77

.73–.81

383

1996 strength

of PID

2000 nonrespondent

2.51

2.45–2.58

527

2000 respondent

2.49

2.44–2.54

891

1996 Strength

of PID

2004 nonrespondent

2.50

2.45–2.54

1046

2004 respondent

2.51

2.43–2.59

372

1996 ideological

extremism

2000 nonrespondent

1.71

1.64–1.78

510

2000 respondent

1.73

1.68–1.77

880

1996 ideological

extremism

2004 nonrespondent

1.69

1.65–1.74

1022

2004 respondent

1.79

1.71–1.86

368

  1. * The 95% confidence intervals for respondents and nonrespondents do not overlap

Appendix B: Descriptive Statistics

Variable*

Observations

Mean

SD

Min.

Max.

1996 participation

383

2.53

2.23

0

11

2000 participation

383

2.40

2.28

0

15

2004 participation

383

2.94

2.31

0

15

1996 contact

383

.77

.42

0

1

2000 contact

383

.70

.46

0

1

2004 contact

383

.76

.43

0

1

1996 Id. Ext.

368

1.79

.70

0

3

2000 Id. Ext.

371

1.85

.70

0

3

2004 Id. Ext.

381

1.82

.74

0

3

1996 Str. PID

372

2.51

.76

0

3

2000 Str. PID

373

2.41

.85

0

3

2004 Str. PID

378

2.50

.82

0

3

  1. In order to assess the ideological extremism and strength of party identification variables for possible measurement problems (especially concern about their high means), I compared them to identical scales constructed using National Election Studies respondents who reported making a campaign contribution (I pooled the 1996, 2000, and 2004 surveys). The means reported above in samples drawn from contributors to the national parties were very similar to the analogous NES means

Rights and permissions

Reprints and permissions

About this article

Cite this article

Claassen, R.L. Testing the Reciprocal Effects of Campaign Participation. Polit Behav 30, 277–296 (2008). https://doi.org/10.1007/s11109-008-9052-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s11109-008-9052-2

Keywords

Search

Navigation