Increasing Polarization since 1970 in Trust for American Institutions

• In Event: Political Polarization and the American Public: Fact or Fiction?

Abstract

Using over time data from the Harris Poll, Gallup Poll, and the General Social Survey back to the early 1970s, we show that public trust in even previously apolitical institutions such as higher education, journalism, science, the military, and religion is polarizing along party lines in the same way that public trust in business, unions, and bankers has historically been polarized. We show that this historic polarization is related to the long-standing class cleavage in American politics while the polarization of the previously apolitical institutions is related to the social-cultural cleavage (related to religious attendance) that emerged in American politics from the 1980s onward. We then use new questions and new data from the 2019 CCES (a national Internet sample) to test various hypotheses about the basis for trust or distrust of institutions. These questions ask about various features of these institutions including their costliness, fundamental importance, perceptions about their indoctrination of those involved with them, and perceptions about their partisan and ideological composition. We provide evidence that these factors matter (e.g., college costs are viewed as very high which decreases trust in colleges; journalism is viewed as indoctrination by some groups which reduces trust; the military is viewed as fundamentally important which increases trust, etc.). We also show that perceptions of the ideological and partisan make-up of these institutions have very significant impacts on trust for the institution with Democrats distrusting institutions (e.g., the military, religion) they regard as involving largely Republican partisans and Republicans distrusting institutions (e.g., colleges, journalism, science) they regard as involving largely Democratic partisans. We show that the polarization in trust for these institutions is related to (and perhaps the cause of) people’s evaluations of those they deem appropriate as marriage partners for their kin and what would be appropriate roles for their kin. Democrats, for example, are less likely to endorse having their kin be highly religious, in the military, or in the business sector; Republicans are less likely to endorse having their kin be college professors, journalists, or scientists. We formulate a novel cognitive balance model to account for this polarization, and we test it using our data. We show that partisanship greatly shapes people’s trust in institutions. We end by considering the implications for American politics and democracy, and we conclude that it is one thing to have politics contesting the division of national income and producing conflict over taxation and expenditures as with a class cleavage but it is another thing when politics is producing polarization in the evaluations of basic knowledge producing institutions (journalism, science, and colleges) and basic values institutions (religion and the military) because this provides little grounds for agreement on facts or values.

Authors

• Henry E. Brady, University of California, Berkeley

• Thomas Bradley Kent, UC Berkeley