Analysis

2.1 Knowledge about the causes and consequences of climate change

American high-school students do not understand climate change issues well. The average high-school student fails a nine-question quiz on the causes and consequences of climate change. As Table 1 shows, 50 percent of the respondents score 5 or fewer correct answers (out of nine).  It is more common that students ignore the causes of climate change than the consequences. For example, only 17 percent of the respondents know that nuclear plants do not contribute to climate change and only 25 percent know that toxic chemicals in landfills do not contribute to climate change.

Table 1: Distribution of correct answers

There are systematic differences in students' knowledge across socioeconomic groups and attitudes. Although these differences are relatively small (5 to 6 percentage points), statistical tests indicate that the differences are systematic and not driven by sampling errors. For example, children of parents with college education and students who do not affiliate with any religious denomination know more about climate change than their counterparts. African-American students answer correctly fewer questions on climate change than students of other races or ethnicities. The difference holds after controlling for additional characteristics such as parents' education, gender, and political preference, among others. Table 2 shows differences between means that are statistically significant at the 5 percent level or better.

Table 2: Climate Change Knowledge: Difference in means

High-school students who discuss issues about the environment at least occasionally with their friends and family know more about the causes and consequences of climate change than their counterparts do. However, discussing environmental issues in school does not appear to influence knowledge. Interest in politics and preference for the Democratic Party are also positively related to students' knowledge.

We also use regression analysis to estimate the effects of socioeconomic factors and learning environment on the score, assuming all other explanatory variables are held constant. We present the results in Appendix A, Table 1. Based on these estimates, we can compare the expected score of high-school students with different socio-economic profiles. For example, a male student who learns about climate change in a special class about the natural environment and who has at least one parent with college education scores approximately 13 percentage points higher than a female student who learns about climate change in a science class and who does not have parents with college education, everything else equal.

2.2 Sources of Information about the Environment and Climate Change

Almost 50 percent of the respondents say they learn the most about climate change in school. A similar proportion of students have learned the most in the media. Interestingly, there is no evidence that relying on the media rather than on school to learn about climate change affects a student's knowledge, on average. On the other hand, individuals who learn the most using the Internet do better than the average. Respondents who say they learn the most from the Internet also claim to be paying more attention to this issue than the rest.

Table 3: Learn the most about global climate change

Although in school the majority of students learn about climate change in science courses, those students who learn about climate change in a special class about the natural environment or in some other class (such as English or Social Studies) do better on the quiz than the rest of students.

Table 4: In school, which course has taught you the most?

2.3 Policy Preferences

The questionnaire also ask students whether they think the U.S. should start reducing greenhouse gas (GHG) emissions now or they think the U.S. should wait until there is more scientific evidence about the future benefits of reducing GHGs. Almost 71 percent of the respondents believe the U.S. should start reducing greenhouse gasses now. Children of households with at least one parent with college education are 9 percent more likely to believe efforts to reduce GHGs should start now. Students who identify as Democrats are almost 11 percent more likely, while students who are very interested in politics are 14 percent more likely to believe efforts to reduce GHGs should start now. Individuals who talk about environmental issues at least occasionally with friends and family are 19 percent and 10 percent more likely to believe efforts to reduce GHGs should start now, respectively, than their counterparts. Table 5 presents differences in proportions that are statistically significant at the 5 percent level or better.

Table 5: Difference in proportions: Start reducing greenhouse gasses now

In order to explore how politically important the issue of climate change might be in the future, we also ask respondents how likely it is that a candidate's position about climate change will strongly influence their vote the first time they vote. Only 20 percent of the students say it is very likely that a candidate's position about climate change will strongly influence their vote.

Table 6: Candidate's positions will strongly influence vote

There are also large and statistically significant differences across socioeconomic groups in the likelihood that climate change will be an important political issue for high-school students. Again, we find that there is a positive relationship between discussing environmental issues with friends and family, but not in school, and pro-environment preferences.

Table 7: Difference in proportions: Candidate's position will strongly influence vote

2.4 Risk Perceptions

Only 28 percent of the students say it is very likely that climate change will affect them personally in the future. African-American students are 12 percent more likely to believe that climate change will very likely affect them personally in the future than students from any other ethnic and racial background. Individuals who talk about environmental issues at least occasionally with friends are almost 18 percent more likely to believe that climate change will very likely affect them personally in the future than their counterparts. Students who show stronger pro-environment attitudes are more certain that climate change will affect them personally in the future. Consistent with this relatively low concern with the personal effects of climate change, only 11 percent of the respondents believe climate change is the most serious environmental problem, behind air pollution, water pollution and toxic waste.

Interestingly, high-school students who believe it is very likely that climate change will affect them personally in the future do not seem to engage in more pro-environment behaviors than high-school students who think it is somewhat likely or unlikely. However, how students perceive risk is correlated with their knowledge and political preferences. Those who believe they will experience the effects of climate change in their lives are 17 percent more likely to state the U.S. should start reduce GHGs now than their counterparts. In addition, these individuals are 19 percent more likely to say a candidate's position on climate change issues will affect their vote.

Table 8: Very Likely climate change will affect you personally in the future?

Table 9: Difference in proportions: Climate change will affect you personally

2.5 Environmental Attitudes

Although 66 percent of the respondents agree that humans have the right to modify the natural environment, a clear majority of respondents think that the earth's resources are limited and mankind is abusing the environment. Attitudes about the role of technology in solving these problems are evenly distributed. These responses suggest that the individuals in this sample of American youth are not hard-core environmentalists but share the belief that humans are depleting scarce natural resources.

Table 10: Distribution of Environmental Attitudes (responses in parentheses)

2.6 Pro-Environment Behaviors

Making efforts to conserve water is the most frequent behavior for the individuals in this sample. The least common frequent activity is to try to reduce the amount of waste the person generates. On average, men engage more frequently than women in these pro-environment behaviors. African-American students engage in fewer pro-environment behaviors in general than their counterparts. Students who are very interested in politics make more frequent efforts than students who are not that interested in politics. Talking at least occasionally about the environment with friends and family is positively related to frequent pro-environment behaviors, but talking in the classroom is not.

Table 11: Distribution of Pro-environment Behaviors (responses in parentheses)

There is no observable difference in pro-environment efforts between high-school students who strongly agree with the statement that we must consider the impact that our actions will have for the welfare of future generations (almost 83 percent of the responses) and those who mildly agree with the statement or disagree. On the other hand, students who strongly agree with the statement that we must consider the impact that our actions will have for the well-being of people living in other countries (almost 70 percent of the responses) engage in more efforts than those who mildly agree with the statement or disagree with the statement.

3. Conclusions

American high-school students do not understand climate change issues well. Despite the conventional assumption that the media might misrepresent the scientific consensus on climate change, the results of this survey suggest that whether students learn the most about climate change from the media or in school does not influence a student's understanding of the issues. Teaching students about climate change outside typical science courses, for example, in a special class dedicated to the natural environment, appears to increase the students' knowledge. Discussion of environmental issues in the schools is unrelated to pro-environment attitudes and behaviors. On the other hand, the responses to this survey consistently show that a student's social environment is correlated with his or her knowledge of climate change issues as well as with his or her environmental values and behaviors. High-school students who discuss issues about the environment at least occasionally with their friends and family engage in more pro-environment behaviors, know more about the causes and consequences of climate change, and are more likely to claim that the U.S. should start reducing greenhouse gasses now than their counterparts do.