A longstanding belief among many political scientists is that citizens have relatively few stable opinions. UCLA Professor John Zaller titled a chapter of his seminal book on public opinion, “Making it up as you go along,” implying that opinions on many (or perhaps most) issues are quite fleeting. This skepticism comes in large part from a long line of studies of “framing effects.”
A common framing study looks like this: Respondents to a survey are randomly divided in half. One half receives a framed message describing a proposed rally by a sometimes violent hate group in terms of the free speech rights of the rally participants. The other half of respondents receives information about the same rally, but this time emphasizing public safety concerns. Researchers consistently find that those receiving the free speech frame express more positive opinions about the rally and those receiving the public safety frame express more negative opinions.
The results aren’t limited to hypothetical rallies: framing effects have been demonstrated on across a wide array of issues. One interpretation of these findings is that opinions are highly malleable and context-dependent. Another possibility is that framing effects reflect reasoned updating of our beliefs in the face of new information.
We wanted to investigate this further and were particularly interested in whether the kinds of framing effects we commonly find really mean that opinions are “made up as we go along” or if they are more an artifact of how we study public opinion. We therefore devised a framing study with a twist. Prior to receiving the framed message our respondents were “pretreated” with some additional information about the issues at-hand.
Specifically, we gave some of our respondents information about the issues just to read without asking them to think about or form an opinion about the issues. We can call them “memorizers.” Another group was given the same “pretreatment” information but asked to form an opinion about the matters. Let’s call them “evaluators.” We repeated this pretreatment process three times for each group over a three week period. Then, the respondents participated in the framing study (just like in the hate rally example above).
The results were striking. Among “memorizers,” we saw the exact same results we typically find. Respondents’ opinions were easily moved by the short-framed message, even though they had been previously exposed three times to information about the issues. Among “evaluators,” we found something completely different. They were unmoved by the framed messages. Because they had already thought about and formed opinions on the issues during the “pretreatment” phase, the framed message had no impact on their opinions. Citizens might be less susceptible to framing effects than political scientists typically believe.
But whether this is good or bad depends on your perspective. If we think it is important for citizens to keep an open mind and be responsive to new information, then it is important for citizens to look more like “memorizers” than “evaluators.” At the same time, if we’re worried about opinions just being made up as we go along, it might be better for citizens to look like “evaluators” rather than “memorizers.” Our results suggest that citizens are a mix of both groups and that’s probably a good thing for the quality of our democracy.
About the Author: Thomas J. Leeper is a post-doctoral researcher in political science at Aarhus University, Denmark. His article with James N. Druckman, the Payson S. Wild Professor of Political Science at Northwestern University, entitled “Learning More from Political Communication Experiments: Pretreatment and Its Effects”, appeared in the October 2012 issue of the American Journal of Political Science.