Skip to main content
Log in

Discussion Networks, Issues, and Perceptions of Polarization in the American Electorate

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


Drawing on the sizable literature on polarization in the American public, we consider the link between discussion network composition and perceptions of polarization. Participants in the 2008–2009 ANES panel study were asked to complete an innovative battery; they interactively moved histograms to rate other groups’ positions on several prominent issues. These novel exercises provide data on individuals’ projections of the average attitudes of others, but critically, they also provide data on the variability of such attitudes. Thus, we use these “response-distributions” to thoroughly assess (1) the relationship between network characteristics and perceptions of the distance between party opinions, and (2) the relationship between network characteristics and perceptions of the homogeneity of opinions within parties. We find evidence that discussion networks track with individual perceptions of the parties in the electorate: exposure to interpersonal disagreement predicts the perception of less distance between (the mean opinions of) the parties, and the reporting of more heterogeneity of opinion within the parties. We conclude by discussing the implications of these findings for democratic functioning more generally.

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

Access this article

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

Instant access to the full article PDF.

Fig. 1

Similar content being viewed by others


  1. Questions of endogeneity are of perennial concern when it comes to examining the relationship between discussion networks and political behavior. As people exert some level of choice over their discussion partners, network characteristics can be shaped by the attitudes held/traits of individuals. That said, in the present effort we remain reassured by the degree to which individuals’ surroundings constrain their ability to self-select their networks—a dynamic articulated (perhaps most) prominently by Huckfeldt et al. (e.g., Huckfeld and Sprague 1995). We discuss these and other issues in detail in subsequent sections of the paper.

  2. Similarly, others have argued that networks—particularly, the disagreeable signals in networks—may produce “clarifying effects”; the presence of disagreement may serve as a contrast to oneself, thereby helping an individual to figure out where she stands (e.g. Sinclair 2012; cite redacted). If this is the case, then exposure to disagreement draws distinctions into sharper focus, and likely promotes more polarized views of both in and out-party attitudes.

  3. Data and replication code are available on the Political Behavior dataverse.

  4. Response rates at the recruitment stage were as follows: minimum response rate (AAPOR RR1): 26%; estimated response rate (AAPOR RR3): 42%. Response rates for the online ANES waves used in analyses: September 2008—16% (AAPOR RR1) and 26% (AAPOR RR3); January 2009—16% (AAPOR RR1) and 25% (AAPOR RR3); May 2009—15% (AAPOR RR1) and 24% (AAPOR RR3). (Additional information/study details are available at

  5. While the accuracy of main respondent perceptions of discussant attitudes certainly varies across individuals, there is ample evidence suggesting that people are fairly accurate when describing their discussants’ attitudes (see Huckfeldt et al. 2004 for a discussion and analysis).

  6. For a discussion of the conceptualization, operationalization and consequences of different measures of disagreement, see Klofstad et al. (2013). The alternative to the measure of general disagreement is to use partisan disagreement. We show results using this measure in the SI Document (the results are substantively similar).

  7. Average disagreement is calculated by summing the amount of general disagreement that is reported with each discussant, and dividing this number by the number of discussants reported. We first rescaled the original items to run from 0 to 1. The averaging then creates a measure that also ranges from 0 to 1, where 0 is those whose views are “not different at all” with all named discussants in the network, and 1 is those whose views are “extremely different” from all named discussants. The variable has a mean of 0.25 and a standard deviation of 0.22.

  8. This measure ranges from 0 to 14, where 0 describes those whose whole network had “no schooling completed,” and 14 is those whose entire network held “professional or doctorate degrees.” The mean is 8.6 (roughly corresponding to a high school graduate, or someone with some college but no degree), and the standard deviation is 4.6.

  9. See the SI Document for an alternative approach to looking at partisan perceptions.

  10. See the Supplemental Information Document for an exploration of models that combine these dependent variable items in different ways (the results are consistent with those presented in the next sections of the paper).

  11. To calculate this we first need to find the means of the Democratic and Republican distributions. We take the height that each bar on the distribution is moved to, divide that height by the overall height of all of the bars, and multiply the value by which number the bar was (1–5) following the measurement strategy in Judd et al. (2012). The formula is: ((1*(Bar 1 Height/Total Height of all Bars)) + (2*(Bar 2 Height/Total Height of all Bars)) + (3*(Bar 3 Height/Total Height of all Bars)) + (4*(Bar 4 Height/Total Height of all Bars)) + (5*(Bar 5 Height/Total Height of all Bars))). This produces the mean for the Republican or Democratic distribution. We then take the absolute value of the difference in the means for our measure of distance between the parties.

  12. To calculate something akin to standard deviation, we need to retain the distributional aspects of the task that respondents completed. Adjusting the weighting strategy in Judd et al. (2012) (itself an adjustment of the typical formula for calculating SD for a set of numbers), we take the square root of the following: ((((1 − mean of the distribution)^2)*(Bar 1 Height/Total Height of all bars)) + (((2 − mean of the distribution)^2) * (Bar 2 Height/Total Height of all bars)) + (((3 − mean of the distribution)^2) * (Bar 3 Height/Total Height of all bars)) + (((4 − mean of the distribution)^2) * (Bar 4 Height/Total Height of all bars)) + (((5 − mean of the distribution)^2) * (Bar 5 Height/Total Height of all bars))). The mean values that are used in this calculation are the same as those that were calculated using the formula in footnote 8. This is done for both parties on all four issues, and then coded as being in or out-party perceptions depending on the respondent’s reported party identification. The end result is eight variables (in and out-party, for four issues).

  13. Because we are coding for in and out-party dynamics, true Independents (those in the middle category of a 7 pt.) are dropped from these measures.

  14. The strength of partisanship measure comes from the variable ‘der08w9,’ which is in wave 9. The variable ranges from 0 (Independents) to 3 (Strong Partisans), with a mean of 1.95. The network size variable ranges from 0 (no named discussants) to 3 (three named discussants), with a mean of 2.28. A discussant was counted if there was a valid response to ‘w9zd4_1,’ ‘w9zd4_2,’ and ‘w9zd4_3.’ The political interest variable is created from ‘w9h1’ in wave 9, and is recoded such that 0 is those who are not interested at all, and 4 is those who are extremely interested. The variable has a mean of 2.69. Gender is a dichotomous measure where male is 1 and female is 0; it has a mean of 0.42. Race is coded such that non-white individuals are coded 1 and whites are 0, with a mean of 0.22. Education comes from variable ‘der05’ and ranges from 1 (no high school) to 5 (post graduate degree), with a mean of 3.36. Finally, age ranges from 18 to 90, with a mean of 50.78.

  15. We have replicated the models presented using survey weights to assess whether the constellation of results holds. While there are a few differences [namely, in the statistical significance—though not size or direction—of the standard deviation results (Table 3)], analyses using survey weights (following ANES documentation) reveal the same general pattern of findings (despite reducing the number of observations considerably). The weighted model estimates are provided in the Supplemental Information (SI) Document.

  16. See the Supplemental Information Document for a discussion of how the distribution of the dependent variables may produce violations of OLS assumptions. We find some evidence of heteroskedasticity, but note that the results largely hold in the presence of robust standard errors [as well other corrective strategies (e.g., estimation via FGLS)].

  17. The magnitudes of these effects sizes are also modest. Across the full range of each of the significant variables—across issues—we observe: between a 0.068 to 0.095 increase in perceived party difference for gender, between a 0.090 to 0.197 reduction in perceived difference for the race measure, between a 0.296 to 0.584 increase across the education measure, between a 0.216 to 0.648 reduction in perceived difference for the age variable, between a 0.88 to 1.4 point increase in perceived party distributions across the interest measure, between a 0.16 to 0.237 increase across the network size measure, and between a 0.084 to 0.2 increase in perceived differences across the strength of partisanship measure.

  18. In terms of effect sizes, moving from the least to the most educated individual produces a decrease in perceived standard deviation of between 0.037 to 0.083, depending on the model. Across the full range of age, there is between a 0.044 and 0.187 point increase in perceived standard deviation (again, depending on the model). Across the full range of the interest measure, there is between a 0.043 and 0.119 point decrease in perceived standard deviations.

  19. In addition to the results presented here, we have conducted subgroup analysis to see if the findings vary by strength of partisanship (i.e. it could be the case that we see different effects of disagreement for strong partisans compared to weaker partisans). With respect to the distance between the means, we find that strong partisans and weaker partisans both appear to respond to disagreement by perceiving smaller distances between the parties, though our statistical significance is diminished as the sample size decreases for such break-downs. For standard deviations we find that disagreement increases perceived heterogeneity within the parties for both strong and weak partisans, and a number of these effects remain statistically significant. In sum, disagreement appears to produce similar effects for both strong and weak partisans when it comes to structuring polarization perceptions.


  • Abramowitz, A. I. (2010). The disappearing center: Engaged citizens, polarization and American democracy. New Haven: Yale University Press.

    Google Scholar 

  • Althaus, S. (2003). Collective preferences in democratic politics: Opinion surveys and the will of the people. New York: Cambridge University Press.

    Book  Google Scholar 

  • Baldassarri, D., & Gelman, A. (2008). Partisans without constraint: Political polarization and trends in American public opinion. American Journal of Sociology, 114, 408–446.

    Article  Google Scholar 

  • Barabas, J. (2004). How deliberation affects policy opinions. American Political Science Review, 98, 687–701.

  • Berelson, B. R., Lazarsfeld, P. F., & McPhee, W. N. (1954). Voting. Chicago, IL: University of Chicago Press.

    Google Scholar 

  • Bishop, B. (2008). The big sort: Why the clustering of like-minded America is tearing us apart. New York: Houghton Mifflin Harcourt.

  • Brady, D. W., & Han, H. C. (2006). Polarization then and now: A historical perspective. In P. S. Nivola & D. W. Brady (Eds.), Red and blue nation? (Vol. 1). Brookings Institution Press and Hoover Press.

  • Brady, H. E., & Sniderman, P. M. (1985). Attitude attribution: A group basis for political reasoning. American Political Science Review, 79, 1061–1078.

    Article  Google Scholar 

  • Braumoller, B. F. (2006). Explaining variance: Or, stuck in a moment we can’t get out of. Political Analysis, 14(3), 268–290.

  • Bullock, J. G. (2009). Partisan bias and the bayesian ideal in the study of public opinion. The Journal of Politics, 71(3), 1109–1124.

    Article  Google Scholar 

  • Buttice, M., Huckfeldt, R., & Ryan, J. B. (2009). Polarization, attribution, and communication networks in the 2006 congressional elections. In J. Mondak & D.-G. Mitchell (Eds.), Fault lines: Why the republicans lost congress (pp. 42–60). New York, NY: Routledge.

    Google Scholar 

  • Chambers, S. (1996). Reasonable democracy. Ithaca, NY: Cornell University Press.

    Google Scholar 

  • DiMaggio, P., Evans, J., & Bryson, B. (1996). Have Americans’ social attitudes become more polarized? American Journal Sociology, 102, 690–755.

    Article  Google Scholar 

  • Enns, P. K., & Kellstedt, P. M. (2008). Policy mood and political sophistication: Why everybody moves mood. The British Journal of Political Science, 38, 433–454.

  • Evans, J. H. (2003). Have Americans’ attitudes become more polarized? An update. Social Science Quarterly, 84, 71–90.

    Article  Google Scholar 

  • Finifter, A. (1974). The friendship group as a protective environment for political deviants. American Political Science Review, 68(2), 607–625.

    Article  Google Scholar 

  • Fiorina, M. P. (2002). Parties and partisanship: A 40-year retrospective. Political Behavior, 24, 93–115.

    Article  Google Scholar 

  • Fiorina, M. P., Abrams, S., & Pope, J. (2008). Polarization in the American public: Misconceptions and misreadings. Journal of Politics, 70, 556–560.

  • Gimpel, J. G., & Schuknecht, J. E. (2001). Interstate migration and electoral politics. The Journal of Politics, 63(1), 207–231.

    Article  Google Scholar 

  • Granovetter, M. (1973). The strength of weak ties. American Journal of Sociology, 78, 1360–1380.

    Article  Google Scholar 

  • Greene, S. (1999). Understanding party identification: A social identity approach. Political Psychology, 20(2), 393–403.

    Article  Google Scholar 

  • Greene, S. (2004). Social identity theory and party identification. Social Science Quarterly, 85(1), 136–153.

    Article  Google Scholar 

  • Habermas, J. (1989). The structural transformation of the public sphere: An inquiry into a category of bourgeois society. Cambridge, MA: MIT Press.

    Google Scholar 

  • Hetherington, M. J. (2001). Resurgent mass partisanship: The role of elite polarization. American Political Science Review, 95, 619–631.

    Article  Google Scholar 

  • Huckfeld, R., & Sprague, J. (1995). Citizens, politics, and social communication: Information and Influence in an election campaign. New York: Cambridge University Press.

    Book  Google Scholar 

  • Huckfeldt, R. (2001). The social communication of political expertise. American Journal of Political Science, 45(2), 425–438.

    Article  Google Scholar 

  • Huckfeldt, R., Johnson, P. E., & Sprague, J. (2004). Political disagreement: The survival of diverse opinions within communication networks. New York: Cambridge University Press.

    Book  Google Scholar 

  • Huddy, L. (2003). Group identity and political cohesion. In D. O. Sears, L. Huddy, & R. Jervis (Eds.), Oxford handbook of political psychology (pp. 511–558). New York, NY: Oxford University Press.

    Google Scholar 

  • Huddy, L. (2013). Translating group identity into political cohesion and commitment. In L. Huddy, D. O. Sears, & J. Levy (Eds.), Oxford handbook of political psychology (2nd ed., pp. 737–773). New York, NY: Oxford University Press.

    Chapter  Google Scholar 

  • Iyengar, S., Sood, G., & Lelkes, Y. (2012). Fear and loathing in party politics: A social identity perspective on polarization. Public Opinion Quarterly, 76, 405–431.

    Article  Google Scholar 

  • Jacobson, G. C. (2000). Party polarization in national politics: The electoral connection. In J. R. Bond & R. Fleisher (Eds.), Polarized politics: Congress and the president in a partisan era (pp. 9–30). Washington, DC: CQ Press.

    Google Scholar 

  • Judd, C. M., Van Boven, D., Huber, M., & Nunes, A. (2012). Measuring everyday perceptions of the distribution of the American electorate. In J. Aldrich & K. McGraw (Eds.), Improving public opinion surveys: Interdisciplinary innovation and the american national election studies (pp. 195–219). Princeton, NJ: Princeton University Press.

    Google Scholar 

  • Kaufmann, K. M., Petrocik, J. R., & Shaw, D. R. (2008). Unconventional wisdom: Facts and myths about American voters. Oxford University Press.

  • Klofstad, C. A., Sokhey, A. E., & McClurg, S. D. (2013). Disagreeing about disagreement: How conflict in social networks affects political behavior. The American Journal of Political Science, 57, 120–134.

  • Krueger, J. (1998). On the perception of social consensus. In M. P. Zanna (Ed.), Advances in experimental social psychology (Vol. 30, pp. 163–240). San Diego: Academic Press.

    Google Scholar 

  • Levendusky, M. (2009). The partisan sort: How liberals became democrats and conservatives became republicans. Chicago, IL: University of Chicago Press.

    Book  Google Scholar 

  • MacKuen, M., & Brown, C. (1987). Political context and attitude change. American Political Science Review, 81(2), 471–490.

    Article  Google Scholar 

  • Marsden, P. V. (1987). Core discussion networks of Americans. American Sociological Review, 52, 122–131.

  • McCarty, N. M., Poole, K. T., & Rosenthal, H. (2006). Polarized America: The dance of ideology and unequal riches. Cambridge, MA: MIT Press.

    Google Scholar 

  • McClurg, S. D. (2006). The electoral relevance of political talk: Examining disagreement and expertise effects in social networks on political participation. American Journal of Political Science, 50(3), 737–754.

    Article  Google Scholar 

  • McClurg, S. D., Miller, J. M., Peterson, D. A. M., & Saunders, K. L. (2015). Polarized politics and political interest. Paper presented at the Annual Meeting of the Midwest Political Science Association, April 16–19, 2015. Chicago, IL.

  • Mutz, D. C. (2002). Cross-cutting social networks: Testing democratic theory in practice. American Political Science Review, 96, 111–126.

    Article  Google Scholar 

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

    Book  Google Scholar 

  • Poole, K. T., & Rosenthal, H. (1997). Congress: A political-economic history of roll call voting. NewYork, NY: Oxford University Press.

    Google Scholar 

  • Prior, M. (2007). Post-broadcast democracy: How media choice increases inequality in political involvement and polarizes elections. New York: Cambridge University Press.

    Book  Google Scholar 

  • Prior, M. (2010). You’ve either got it or you don’t? The stability of political interest over the life cycle. Journal of Politics, 72(3), 747–766.

    Article  Google Scholar 

  • Richard, F., & Bond, J. R. (2000). Partisanship and the president’s quest for votes on the floor of congress. In J. R. Bond & R. Fleisher (Eds.), Polarized politics: Congress and the president in a partisan era (pp. 154–185). Washington, DC: CQ Press.

    Google Scholar 

  • Rohde, D. W. (1991). Parties and leaders in the post reform house. Chicago, IL: University of Chicago Press.

    Book  Google Scholar 

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

  • Seyle, D. C., & Newman, M. L. (2006). A house divided? The psychology of red and blue America. American Psychologist, 61, 571–580.

    Article  Google Scholar 

  • Sinclair, B. (2012). The social citizen: Peer networks and political behavior. Chicago: University of Chicago Press.

  • Tajfel, H. (1959). Quantitative judgment in social perception. British Journal of Psychology, 54, 259–272.

    Google Scholar 

  • Tajfel, H. (1978). Social categorization, social identity, and social comparisons. In H. Tajfel (Ed.), Differentiation between social groups (pp. 27–60). London: Academic Press.

    Google Scholar 

  • Tajfel, H., & Turner, J. C. (1979). An integrative theory of intergroup conflict. In W. G. Austin & S. Worchel (Eds.), The social psychology of intergroup relations (pp. 33–47). Monterey, CA: Brooks/Cole.

    Google Scholar 

  • Tajfel, H., & Turner, J. C. (1986). The social identity theory of intergroup behavior. In S. Worchel & W. G. Austin (Eds.), Psychology of intergroup relations. Chicago, IL: Nelson-Hall.

    Google Scholar 

  • Tajfel, H., & Wilkes, A. L. (1963). Classification and quantitative judgment. British Journal of Psychology, 54, 259–272.

    Article  Google Scholar 

  • Testa, P., Hibbing, M., & Ritchie, M. (2014). Orientations toward conflict and the conditional effects of political disagreement. Journal of Politics, 76(3), 770–785.

    Article  Google Scholar 

  • Van Boven, L., Judd, C., & Sherman, D. (2012). Political polarization projection: Social projection of partisan attitude extremity and attitudinal processes. Journal of Personality and Social Psychology, 103(1), 84–100.

    Article  Google Scholar 

Download references

Author information

Authors and Affiliations


Corresponding author

Correspondence to Jeffrey Lyons.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary material 1 (DOCX 570 kb)

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lyons, J., Sokhey, A.E. Discussion Networks, Issues, and Perceptions of Polarization in the American Electorate. Polit Behav 39, 967–988 (2017).

Download citation

  • Published:

  • Issue Date:

  • DOI: