Framing effects on personal opinion and perception of public opinion: the cases of physician-assisted suicide and social security *.

Objective. A growing body of literature on issue framing has demonstrated the conditional influence of issue frames on self-reported opinion. The effects of frames are conditioned by message content, the medium of communication, and the predispositions of respondents. However, the literature has yet to explore the influence of issue frames on respondents' perceptions of public opinion. We draw from the psychological literatures on cognitive accessibility biases and impersonal impact and construct competing hypotheses concerning the likelihood of issue frames affecting perceptions of opinion. Methods. We test hypotheses using data from an experimental field study that exposed respondents to opposing issue frames on two important issues--reforming Social Security and physician-assisted suicide. Results. Our results largely support the impersonal impact hypothesis. Conclusions. We find that available information from issue frames influences personal-level opinion but in general does not affect perceptions of pub lic opinion. We discuss the implications of these findings and suggest avenues for future research.

**********

Political issues and policy proposals arise from complex problems that are often separate and remote from the direct experiences of citizens (Lippmann, 1922; Edelman, 1964). Citizens' judgments about such issues rely crucially on the descriptions and rhetorical representations of political elites and other information sources, including the media and interest groups. The capacity to frame issues, that is, define the way an issue comes to be understood, is undoubtedly one of the most important and powerful communication strategies available to political elites and the media (Jones, 1994; Rochefort and Cobb, 1994; Stone, 1997). It is also a genuinely effective means to influence the distribution of public opinion on a variety of issues (Sniderman, 1993).

Considerable evidence supports the notion that opinions are shaped depending on how issues are framed. Issues may be framed by a set of competing values (Gamson, 1992; Chong, 1996), potential beneficiaries (Nelson and Kinder, 1996), or even different personalities (Kuklinski and Hurley, 1994). Alternative frames affect opinion by stressing specific considerations, making certain values or facts more accessible and more important (Iyengar, 1990); in turn these considerations carry significant weight in opinion processes (Zaller and Feldman, 1992).

However, while the framing literature has been preoccupied with the linkages between opinion change and issue frame, it has neglected the potential impact on other important political judgments, namely, individuals' perceptions of public opinion. Like opinions, perceptions about others' behaviors and attitudes are presumably guided by readily accessible information. If opinions are found to be sensitive to issue frames, are perceptions of public opinion equally affected?

To appreciate the significance of this question, it is important to recognize the role of perceived public opinion in contemporary political life. For example, representations of the electorate's preferences have been found to influence an individual's own preferences on specific policy issues (Mutz, 1992), candidate choice (Lazarsfeld, Berelson, and Guadet, 1944; Bartels, 1988), and institutional support (Joslyn, 1997). Perceptions of public opinion have also been linked to the willingness to engage in public debate (Noelle-Neumann, 1993). In addition, perceptions of how the country is performing economically are a major factor in vote choice (Hetheringron, 1996; Mutz, 1992; Kinder and Kiewiet, 1979). The decision calculus of whether to support a specific policy, political figure, or even political institution thus often includes a social imperative: the perceived opinions or behaviors of numerous others (Mutz, 1998). However, for all the attention given to this topic, to date no research has examined whethe r alternative frames of an issue affect individuals' perceptions of public support for that issue.

Our research contributes to the literature by examining the relationship between issue framing, personal opinion, and perceptions of public opinion. Alternative frames of Social Security reform and physician-assisted suicide were constructed and presented to several hundred respondents in a statewide opinion survey. After exposure to issue frames, respondents were asked their own opinion about the issue and then asked to estimate public opinion. Our research has several innovations. First, we utilize alternative depictions of two issues overlooked by previous research on framing effects. One issue can be characterized as an "easy" is sue (physician-assisted suicide), where opinion is based primarily on emotive beliefs. Social Security reform is our second issue, which can be classified as a "hard" issue because of the technical information that often informs opinion (Carmines and Stimson, 1980). We are thus able to examine the impact of alternative issue frames across two issue types. Second, we explore wheth er frames affect both individual and perceived opinion. Our analyses indicate that alternative issue frames are important and consistent determinants of individual opinion. However, the impact of frames on perceived opinion is neither impressive nor consistent across issue frames. We conclude with a consideration of how our results inform the literatures on political perceptions, media framing, and public policy.

Perception of Public Opinion

Conventional wisdom and empirical evidence alike underscore the important role of perceptions in political processes. Any reasonably complete picture of political behavior must take into account how individuals perceive their immediate and more distant political environments. Starting with the early voting work by the Columbia school, a substantial literature has established the powerful effects of perceptions of economic conditions (Kinder and Kiewiet, 1979; Mutz, 1992), candidate's issue positions (Conover and Feldman, 1989; Krosnick, 1990), and media impact (Perloff, 1996) on various political judgments. Drawing heavily from research on social conformity (Ash, 1951; Milgram, 1961; Noelle-Neumann, 1993) and personal influence (Katz and Lazarsfeld, 1955), researchers have focused on the intersection of individual citizen and the generalized "other" as represented by the media (Mutz, 1998). Such representations affect attitudinal and decision processes by spurring citizens to contemplate the attitudes and exp eriences of others.

But even as a considerable appreciation for the political effects of collective opinion has emerged, very few studies explore what determines perceptions of that opinion. What is known involves the role of …