Natural gas – friend or foe of the environment? Evaluating the framing contest over natural gas through a public opinion survey in the Pacific Northwest

ABSTRACT

A fierce debate is raging about the role of natural gas in North America’s energy mix. Once viewed as a bridge fuel for renewable energy, it is increasingly being characterized as hindering the energy transition. We explore public opinions about natural gas, its use and export, among residents in Oregon, Washington and British Columbia. While many of our respondents supported the continued use of natural gas in electricity generation and viewed it as relatively environmentally friendly, they did not feel that the benefits of fracking (increasingly the main source of natural gas production) outweigh its risks. Males, political conservatives, those who prioritized the economy over the environment, and those who didn’t subscribe to anthropogenic global warming felt more favorably toward natural gas. Furthermore, those who saw gas as more environmentally friendly were more supportive of gas usage and export, while those with pro-environmental views were less likely to support it.

KEYWORDS:

• Public opinion
• natural gas
• framing
• survey research

Disclosure statement

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

Ethics review

This research was reviewed and declared exempt by the Institutional Review Boards at The Evergreen State College (Protocol #1819-046) and Oregon State University (Protocol# IRB-2019-0183).

Notes

1. Online administration is increasingly widespread in contemporary survey research, especially given declining response rates to mail, telephone, and in-person survey modes (Stedman et al. 2019). Online vendors like YouGov maintain large respondent pools and can obtain the desired number of responses through their proprietary on-line platform, then use quota sampling, sample weighting, and other strategies to ensure that a nonprobability sample best approximates a probability sample.

2. The matching and weighting procedure used a propensity score function that included age, gender, race/ethnicity, years of education, region, and metro/non-metro. The propensity scores were grouped into deciles of the estimated propensity score in the frame and post-stratified according to these deciles.

3. For Oregon and Washington, the weights were post-stratified first using 2016 Presidential vote choice and a four-way stratification of gender, age (4-categories), race (4-categories), and education (4-categories), then on metro/non-metro distribution from the November 2018 Current Population Survey (CPS) to produce the final weights. For British Columbia, the weights were first post-stratified on a four-way stratification of gender, age (4-categories), race (4-categories), and education (4-categories), and then on metro/non-metro distribution from 2018 Census.

4. Political ideology is a more complicated construct in Canada than in the United States. For example, the Liberal Party is a federal, left-of-center (liberal) political party in Canada, while the British Columbia Liberal Party is right-of-center (conservative) political party in the providence of B.C. Our measure asked respondents to self-rate on a five-point political ideology scale from Very Liberal to Very Conservative; as such, we were cautious in the interpretation. Our results suggested our respondents interpreted the question as we intended – our B.C. respondents who answered as ‘conservative’ were more likely to support natural gas usage and export, as were our U.S. respondents.

5. We used (weighted) logistic regression for our second and third regressions, since the dependent variables in these models are ordinal variables. We originally estimated ordered logistic regression; however, these models violated the proportional odds assumption, also called the parallel regression assumption (Brant 1990). Thus, we collapsed the dependent variables into binary measures of support and estimated binary logistic regression models instead.

6. McFadden’s R2 is often used in logistic regression models to indicate the amount of variation in the dependent variable accounted for by the predictor variables. McFadden states that pseudo R2 values in the range of 0.2–0.4 indicate quite a high amount of variation explained and ‘excellent’ model fit. However, McFadden himself warned that this (rho-squared) indicator produces values that are ‘considerably lower than those of the R2 index [in OLS regression]’ (McFadden 1973).

7. Though we did not report this in our results, we ran a weighted one-way Analysis of Variance (ANOVA) to determine that conservatives, moderates, and liberals had statistically different scores on the enviro_gas scale.

Additional information

Funding

This research was funded by the Evergreen State College, 2018 Sponsored Research Grant for Faculty, and by Oregon Sea Grant (#NA18OAR4170072).

Notes on contributors

Shawn Olson Hazboun

Shawn Olson Hazboun, PhD, is an environmental sociologist whose research focuses on the social dimensions of energy systems, including community impacts from resource extraction and energy production, as well as public perceptions on both renewable and fossil-fuels based energy production. She is particularly interested in the social impacts of the current global energy transition, especially within rural communities that have traditionally provided the nation’s energy resources. Shawn is an Assistant Professor of Environmental Social Science within the Graduate Program on the Environment at The Evergreen State College.

Hilary Schaffer Boudet

Hilary Schaffer Boudet, PhD, Associate Professor at Oregon State University, focuses her work on environmental and energy policy, social movements, and public participation in energy and environmental decision making. She is currently conducting research on public perceptions of and community response to energy development; community response to extreme weather events; and interventions with young people designed to encourage sustainable behavior. Before coming to OSU, Hilary was a Postdoctoral Scholar at Stanford University, where she also completed her PhD in Environment and Resources. She holds a BA in Environmental Engineering and Political Science from Rice University.