The Origins and Consequences of Democratic Citizen’s Policy Agendas: A Study of Popular Concern About Global Warming


Click here to read the full report


This study examines the proximal causes of national seriousness assessments of global warming, the causes of attitudes, certainty, and existence beliefs, and the effects of national seriousness judgments on policy preferences using a pair of national surveys. First, we describe the results of a large-scale survey of a nationally representative sample of American adults. Then, using data from a survey of a representative sample of adults living in Ohio, the study explores many of the same issues and one new one.

Study 1: National Survey


Computer-assisted telephone interviews were conducted with a representative sample of 1,413 American adults by the Ohio State University Survey Research Unit between September 17, 1997, and February 13, 1998. The sample was generated via random digit dialing, and the cooperation was 70%. The resident of each household with the most recent birthday was asked to participate.


Respondents were asked about:

  1. The national seriousness of global warming
  2. Whether the winter and summer temperatures in their local areas had changed in recent years
  3. Their beliefs about global warming’s existence (i.e., whether global warming would occur in the future if nothing is done to stop it)
  4. Their attitudes toward global warming
  5. The certainty with which they held those beliefs and attitudes
  6. How much they felt they knew about global warming
  7. How much they had thought about global warming
  8. How much they trusted scientists
  9. How much they thought the US government should do to deal with global warming
  10. Whether they thought the US government should require air pollution reductions from US businesses and from countries that receive foreign aid from the US.
  11. Their educational level as a measure of cognitive skills

They were also asked whether the following possible consequences of Global Warming would be increased, decreased, or not changed by Global Warming:

  1. Sea Level
  2. Water Shortages
  3. Food Supplies
  4. Number of Types of Animals in the World
  5. Number of Types of Plants in the World
  6. The Frequency of Hurricanes and Tornadoes

If the respondent thought that global warming would produce a change in a phenomenon, they were then asked whether the change would be good or bad.


To test whether attitudes, existence beliefs, and certainty interacted as expected to influence national seriousness judgments, we conducted a series of OLS regressions predicting national seriousness judgments. Higher seriousness ratings were associated with belief in global warming’s existence, more negative attitudes, and higher certainty. High certainty exacerbated the effects of existence beliefs and attitudes and the posited three-way interaction between existence beliefs, attitudes, and certainty was statistically significant and positive, as expected.

People who believed that global warming will probably not occur in the future rated it as less serious than people who believed global warming probably will occur. Among people who thought global warming probably will occur, those who thought it will be neutral or good rated it as less serious than those who thought it will be bad. And among people who thought that global warming will occur and that it will be bad, those who were highly certain rated it as more serious than those who were less certain. Thus, national seriousness ratings were highest among people who believed in global warming’s existence, thought it would be bad, and were highly certain of those views.

Attitudes toward global warming most strongly predicted national seriousness ratings when respondents believed global warming probably will happen in the future and were certain of their global warming beliefs. Respondents placed significant weight on three possible consequences: changes in sea level, food supplies, and the numbers of animal species. The more negative was a person’s attitude toward global warming greater perceived national seriousness also enhanced support for specific policies to reduce air pollution by limiting air pollution from U.S. businesses (coded 1 if a respondent favored the proposed policy and 0 otherwise) and by requiring countries given U.S. aid to reduce pollution (

Ohio Survey

The Ohio survey allowed us to test the influence of media coverage on existence beliefs.



For the Ohio survey, the Ohio State University Polimetrics Laboratory conducted 40-min computer-assisted telephone interviews with 758 adults. A representative sample of private households with telephones in Ohio was generated by random digit dialing, and the adult member of each contacted household who had the next birthday was selected to be interviewed. The cooperation rate was 75%.


Many of the questions used in the Ohio survey were identical to those used in the National survey. In order to permit asking a wider array of questions, each respondent was randomly assigned to receive on of three partially different forms of the questionnaire (all responses were coded as in the national survey). All three forms asked respondents about global warming’s existence, their attitudes toward global warming, the certainty with which they held those beliefs and attitudes, how much they felt they knew about global warming, and how much they had thought about global warming.


Greater television exposure was indeed associated with an increase in belief in the existence of global warming, but only among people who trusted scientists and who were highly educated. And as predicted, newspaper exposure was not associated with existence beliefs among these respondents, who would have been able to retain both the initial newspaper assertions that global warming existed and the later skepticism.

Greater newspaper exposure was associated with less belief in global warming’s existence among people who were highly trusting and low in education. Television exposure was not associated with existence beliefs among these people, who would have been less able to retain the television story content from September and October.

The effects of sea level, food shortages, and number of animal species match the national survey’s findings, but the significant effect of water shortages here did not appear in the national data, which could be due to the different equation specification here or to different standards of judgment among Americans as a whole. The latter is certainly possible if Ohio’s importantly farm-based economy made its residents more sensitive to water supplies than residents of other parts of the country.

When we regressed certainty on knowledge, thought, and the demographics, both knowledge and thought had strong, positive effects, as expected. And as in the national survey, the effect of knowledge was twice the size of the effect of thought.