Grading on a Curve: Why Americans Support the Military Above Other Institutions

• In Event: Challenges in Comparative Civil-Military Relations

Abstract

Why has public confidence in the U.S. military increased over the last three decades while confidence in other government institutions has stagnated or declined? Previous attempts to answer this question have focused primarily on public attitudes about the military itself, often using survey experiments to isolate alternative explanations such as partisanship. This body of research has proposed potential explanations for military confidence, but also exposed seeming contradictions. The competence or ethical behavior of those in uniform might explain public admiration, for example, but if so, why has it survived infamous scandals and uncertain results in the wars the U.S. has waged over two decades? Politicization of the military or perceived military partisanship also may shape public attitudes, but why has confidence remained high across both parties? High confidence also could be the result of a substitutionary effect as regard for other institutions declines, but why is the military the obvious repository for esteem? In this paper, we analyze original proprietary data collected expressly to explore the causes and consequences of public confidence in the military in comparison to other American institutions. The survey was administered on-line by the National Opinion Research Center (NORC) as part of their AmeriSpeak panel to some 4500 adults in the United States. Drawing on this data, we explore the determinants of public support for government institutions and assess the strength of competing and complementary explanations. We find that public perceptions of institutional performance and partisanship drive public confidence in the military and help explain why public assessments of other government institutions have diverged. We consider the potential implications of these findings on democratic accountability and close with suggestions for extending this research, in light of these new findings.

Authors

• Jim Golby, University of Texas at Austin

• Peter D. Feaver, Duke University