Abstract
This study evaluates the competing influences of motivated reasoning and personal experience on policy preferences toward the Affordable Care Act. Using cross-sectional and panel survey data, the findings reveal that healthcare attitudes are responsive to information that individuals receive through personal experience. Individuals who experienced a positive change in their insurance situation are found to express more positive views toward the health reform law, while individuals who lost their insurance or experienced an otherwise negative personal impact on their insurance situations express more negative views. The results point to personal experience as a source of information that can influence individuals’ preferences. However, although attitudes are responsive to the quality of one’s personal interactions with the healthcare system, the results also suggest that partisan bias is still at work. Republicans are more likely to blame the health reform law for negative changes in their health insurance situations, while Democrats are more likely to credit the law for positive changes in their situations. These motivated attributions for their personal situations temper how responsive partisans’ attitudes are to information acquired through personal experience.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
Kaiser Family Foundation tracking polls include nationally representative surveys of more than 1000 U.S. adults conducted nearly each month from 2011 to 2015.
From 2011 to April 2014, respondents were asked, “So far, would you say you and your family have been negatively affected by the health reform law, or not?” and “So far, would you say you and your family have personally benefited from the health reform law, or not?” Those indicating “not negatively affected” and “not benefited” were coded as no personal impact. From May 2014 to 2015, respondents were asked, “So far, would you say the health care law has directly (helped) you and your family, directly (hurt) you and your family, or has it not had a direct impact?” Respondents indicating “no direct impact” were coded as no personal impact.
The survey also asks the nature of the change in insurance: “Which best describes the change in your health insurance situation? Did you lose or drop your health insurance coverage, change health insurance plans, get health insurance after being uninsured, or was it some other type of change?”.
There are few differences in race/ethnicity, gender, marital status, income, unemployment, political ideology, objective knowledge about the ACA, and general health between individuals with and without recent changes in their insurance, and minimal differences in age, education, and partisanship. (Individuals with recent changes in insurance are slightly younger, more educated, and less likely to be Democrats.) Individuals with recent changes in insurance are more likely to have indicated that have a pre-existing chronic health condition and are slightly more likely to indicate they have “enough” information about how the ACA will affect them personally.
In this analysis, partisans include Independents who lean partisan; results hold in robustness checks that exclude partisan leaners. The KFF data do not make it possible to distinguish between strong and not very strong partisans. The Amazon Mechanical Turk replication analysis controls on this, and the results hold (Online Appendix Table 13).
To the extent that subjective and objective perceptions do not align—that some individuals view their insurance purchased through the marketplace in a negative light—it should only make it harder to detect more positive views toward the ACA among this subgroup, relative to those who have no direct experience with the ACA.
Individuals with employer-sponsored insurance are chosen as the comparison group because they constitute the largest proportion of insurance-holders in the sample.
The dependent variable is dichotomized to isolate whether variation in favorability is driven by support versus opposition. Results are generally consistent when applying sampling weights, in a linear probability model, when treating ACA favorability as dichotomous but omitting ‘don’t know’ responses, as a 4-point continuous scale (omitting ‘don’t knows’), or a 5-point scale (treating ‘don’t knows’ as the midpoint). See Online Appendix Table 1.
This analysis is unable to employ all control variables included in the analysis of the KFF January 2014 poll to follow because not all questions were asked across each monthly poll.
Respondents are classified as Democrats or Republicans based on their responses to a 7-point party identification question in the 2012 wave. Partisans include leaners.
One could alternatively compare those who gained insurance to those who reported having no insurance in both waves (those who did not gain insurance). Democrats and Republicans who reported having no insurance in 2012 and 2014 became significantly less likely to support the ACA in 2014. Thus, although there is not a significant growth in support of the ACA among those who gained insurance, one could interpret the effect of gaining insurance as preventing the decrease in support for the ACA that would have occurred if respondents had continued be uninsured. Respondents without insurance in both waves might have perceived “failing to gain insurance” similarly to a loss. In the analyses to follow, those who gained or lost insurance are instead both compared to those who always had insurance.
The two approaches require two plausible but distinct sets of assumptions. The LDV approach assumes that conditional on pre-existing views and other covariates, respondents who gained and lost insurance between waves would emerge with indistinguishable views from those who reported having insurance in both waves if not for the changes in their insurance status. The difference-in-difference assumption is that individuals who gained versus lost insurance would experience the same degree of change in their views between waves (relative to those who always reported having insurance) if not for the changes in their insurance. The analyses control for age, employment status (‘unemployed’ or ‘retired’), education, gender, race, marital status, having a child, political ideology, and partisan strength. In the difference-in-difference specification, covariates that do not plausibly change over time are dropped (e.g., gender, race).
These regressions are weighted. Unweighted models yield similar results (Online Appendix Tables 4 and 6).
In Bayesian updating, an individual’s opinion toward the ACA would be a weighted average of their pre-existing beliefs and the new information gleaned from personal experience. The paper does not take a strong position on whether a Bayesian model of learning is the most appropriate model. A model of Bayesian learning is discussed simply as one approach for determining whether partisans acquire and process information in a biased or unbiased manner. As Bullock discusses, “Partisans are not Bayesians, but Bayesian updating matters because it is increasingly the normative standard against which partisan updating is judged” (2009, p. 1110).
The interaction coefficient in the model is significant. Predicted probabilities hold all covariates at observed levels, while fixing partisanship and the nature of the insurance change at desired levels. The results presented use unweighted regression models, but results hold when including survey sampling weights. These analyses use listwise deletion for missing data. Respondent income (10 % missing due to an option where respondents can indicate they “prefer not to say”) is the only covariate with more than 1.2 % missing data. Results are not sensitive to including or excluding income as a covariate.
See Online Appendix p. 14 and Online Appendix Table 10 for robustness checks that vary survey question order.
The analyses focus on the role of partisanship, not ideology, because partisanship appears to be the theoretically more accessible elite cue and, empirically, the clearer signal of respondent preferences on the ACA. It is possible that political ideology could also affect preferences toward the law and temper the effect of personal experience in a similar way as partisanship. Indeed, a model that interacts ideology with personal experience instead of partisanship yields similar results; however the inclusion of the partisanship interaction renders the interaction between ideology and personal experience indistinguishable from zero while the partisanship interaction remains significant.
Online Appendix Table 9 presents a cross-tabulation of reported objective changes in insurance situation by subjective perception of whether the changes were for better or worse: 81 % of those who lost or were dropped from coverage saw this as a change for the worse, and 87 % who saw a cost increase saw this as a change for the worse. In contrast, 80 % of those who gained insurance saw this as a change for the better.
References
Achen, C. H, & Bartels, L. M. (2012).Why shark attacks are bad for democracy presented in the Center for the Study of Democratic Institutions, Vanderbilt University.
Ansolabehere, S., & Schaffner, B. (2015). Cooperative Congressional Election Study 2010-2014 Panel Study [Computer File]. Release 1: June 10, 2015. Amherst, MA: University of Massachusetts, Amherst [producer]
Arceneaux, K. (2003). The conditional impact of blame attribution on the relationship between economic adversity and turnout. Political Research Quarterly, 56(1), 67–75.
Bartels, L. M. (2002). Beyond the running tally. Political Behavior, 24(2), 117–150.
Bullock, J. G. (2009). Partisan bias and the bayesian ideal in the study of public opinion. Journal of Politics, 71(3), 1109–1124.
Bullock, J. G. (2011). Elite influence on public opinion in an informed electorate. American Political Science Review, 105, 496–515.
Bullock, J. G., Gerber, A. S., Hill, S. J., & Huber, G. A. (2015). Partisan bias in factual beliefs about politics. Quarterly Journal of Political Science, 10(4), 519–578.
Campbell, A., Converse, P. E., Miller, W. E., & Stokes, D. E. (1960). The American voter. New York: Wiley.
Citrin, J., & Green, D. P. (1990). The self-interest motive in American public opinion. Research in Micropolitics, 3, 1–28.
Delli Carpini, M. X., & Keeter, S. (1996). What Americans know about politics and why it matters. New Haven, CT: Yale University Press.
Fiorina, M. P. (1981). Retrospective voting in American national elections. New Haven: Yale, University Press.
Gaines, B. J., Kuklinski, J. H., Quirk, P. J., Peyton, B., & Verkuilen, J. (2007). Same facts, different interpretations: partisan motivation and opinion on Iraq. Journal of Politics, 69(4), 957–974.
Gilens, M. (2001). Political ignorance and collective policy preferences. American Political Science Review, 95, 379–396.
Kaiser Family Foundation: January Kaiser Health Tracking Poll. (2014). The Henry J. Kaiser Family Foundation.
Kaiser Family Foundation: Trends Dataset. (2015). The Henry J. Kaiser Family Foundation.
Key, V. O. (1966). The responsible electorate: Rationality in presidential voting, 1936–1960. Cambridge: Harvard University Press.
Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108(3), 480–498.
Lenz, G. S. (2009). Learning and opinion change, not priming: Reconsidering the priming hypothesis. American Journal of Political Science, 53(4), 821–837.
Lenz, G. S. (2012). Follow the leader? How voters respond to politicians’ policies and performance. Chicago: University of Chicago Press.
Lerman, A. E, & McCabe, K. T. (2016). Personal experience and public opinion: Conditional policy feedback and support for public health insurance. Unpublished.
Malhotra, N., & Kuo, A. G. (2008). Attributing blame: The public’s response to Hurricane Katrina. The Journal of Politics, 70(1), 120–135.
Mettler, S., & Stonecash, J. M. (2008). Government program usage and political voice. Social Science Quarterly, 89(2), 273–293.
Nawara, S. (2015). Who is responsible, the incumbent or the former president? Motivated reasoning in responsibility attributions. Presidential Studies Quarterly, 45(1), 110–131.
Nyhan, B. (2010). Why the death panel myth wouldn’t die: Misinformation in the health care reform debate. The Forum 8(1).
Pierson, P. (1994). Dismantling the welfare state? Reagan, thatcher, and the politics of retrenchment. New York: Cambridge University Press.
Prior, M., Sood, G., & Khanna, K. (2015). You cannot be serious: The impact of accuracy incentives on partisan bias in reports of economic perceptions. Quarterly Journal of Political Science, 10(4), 489–518.
Rudolph, T. J. (2003). Who’s responsible for the economy? The formation and consequences of responsibility attributions. American Journal of Political Science, 47, 698–713.
Rudolph, T. J. (2006). Triangulating political responsibility: The motivated formation of responsibility judgments. Political Psychology, 27(1), 99–122.
Sears, D. O., & Funk, C. L. (1990). Self-interest in Americans’ political opinions. In J. J. Mansbridge (Ed.), Beyond self-interest. Chicago: University of Chicago Press.
Sears, D. O., Lau, R. R., Tyler, T. R., & Allen, H. M, Jr. (1980). Self-interest vs. symbolic politics in policy attitudes and presidential voting. American Political Science Review, 74, 670–684.
Sekhon, J. S. (2011). Multivariate and propensity score matching software with automated balance optimization: The matching package for R. Journal of Statistical Software, 42(7), 1–52.
Sirin, C. V., & Villalobos, J. D. (2011). Where does the buck stop? Applying attribution theory to examine public appraisals of the president. Presidential Studies Quarterly, 41(2), 334–357.
Stoker, L. (1994). A reconsideration of self-interest in American Public Opinion. Paper presented at the annual meeting of the Western Political Science Association.
Taber, C., & Lodge, M. (2006). Motivated skepticism in the evaluation of political beliefs. American Journal of Political Science, 50(3), 755–769.
Taber, C., Lodge, M., & Glather, J. (2001). The motivated construction of political judgments. In J. Kuklinski (Ed.), Citizens and politics: Perspectives from political psychology (pp. 198–226). New York: Cambridge University Press.
Zaller, J. (1992). The nature and origins of mass opinion. New York: Cambridge University Press.
Acknowledgments
Thank you to Markus Prior, Martin Gilens, Amy Lerman, members of the Princeton Behavior Group, and the anonymous reviewers at Political Behavior for their guidance and helpful feedback. Thank you to the Center for the Study of Democratic Politics at Princeton University for financial support. Replication materials can be found at http://dx.doi.org/10.7910/DVN/COZDRV.
Author information
Authors and Affiliations
Corresponding author
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
About this article
Cite this article
McCabe, K.T. Attitude Responsiveness and Partisan Bias: Direct Experience with the Affordable Care Act. Polit Behav 38, 861–882 (2016). https://doi.org/10.1007/s11109-016-9337-9
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11109-016-9337-9