Public perspectives and media reporting of wolf reintroduction in Colorado

Statewide survey

We adopted a stratified sampling approach to obtain a representative sample of Colorado residents, using three geographic regions, age, and gender categories as strata. All participants were at least 18 years of age and were recruited in August 2019 using Qualtrics (Provo, UT), a commercial sampling firm with a licensed online survey platform.

We set a minimum target sample size of 200 participants for each region in Colorado, including the Front Range (11 counties), Western Slope (35 counties), and Eastern Plains (18 counties), allowing for sufficient samples (estimates within ±7% at the 95% confidence level) for comparisons across the different regions. Our region classifications were adopted from Teel, Bright & Manfredo (2003). We were particularly interested in ensuring adequate sample size for the less populous Western Slope, given this is the region where wolves would most likely be restored and find suitable habitat, and thus impact residents more directly (Carroll et al., 2003). We also stratified our sample by age and gender to mirror the American Community Survey (ACS) (2017) 5-year (2013–2017) estimates retrieved from the Social Explorer database (Table S1). To reflect the 2017 ACS, our target stratification consisted of equal gender representation and equal distribution of the sample across the following age categories: 18–34, 35–54, and 55 years and older (Table S1).

Upon providing Qualtrics the target sample sizes, the firm then recruited and screened potential respondents from existing pools of online survey-takers contacted through panel partners. Qualtrics asked each potential respondent if they would be willing to participate in a survey; to avoid selection bias, they were not told about the topic of the survey. Potential respondents were incentivized to participate in different ways, depending on which panel they were recruited from; for example, panel partners compensate using monetary incentives, game points, gift cards, or other prizes. We first conducted a soft launch of the survey, in which we obtained 50 pilot responses. We checked responses to open-ended questions to ensure participants’ answers reflected that they had read the prompt, and checked responses to closed-ended questions for satisficing (i.e., participants selecting the same responses without considering the questions; Krosnick, Narayan & Smith, 1996). Final data collection then proceeded until target sample sizes across strata were met. Qualtrics removed respondents who did not pass a speed check (i.e., took the survey too quickly); who did not answer “yes” to screener questions (asking if they would agree to provide thoughtful and honest answers); or who did not provide informed consent. This led to a total of 734 valid responses.

We obtained 365 responses from the Front Range (49.73% of sample), 277 from the Western Slope (37.74%), and 92 from the Eastern Plains (12.53%) (Fig. 1). We received more than the 200 desired target responses from the Front Range and Western Slope because, once these quotas were reached, the Eastern Plains quota was still not reached and recruitment therefore remained open to try to increase sample size in that region. To accurately represent survey responses across the entire state, it was necessary to weight the data to reflect the actual population distribution of the three regions for state-level reporting (Vaske, 2008). Through our weighting procedure, we made our sample representative of the following geographic population distribution obtained from the ACS (2017): 82.56% of the population in the Front Range, 14.12% of the population in the Western Slope, and 3.31% of the population in the Eastern Plains. The weighting factor for each region was calculated as the population percentage in each region from the 2017 ACS divided by the sample percentage in each region (see Tables S1 and S2 and Supplemental Materials for more information and numbers used for weighting). A full table of weighted and unweighted proportions and weighted confidence intervals is given in the Supplemental Materials (Table S3); weighted and unweighted values were similar.

A limitation that should be considered when interpreting our results is that, because we used an online survey platform, it is possible that our sample may be biased towards individuals with high technology awareness (Ranchhod & Zhou, 2001) who are more comfortable with online surveys. Keeter & McGeeney (2015) suggest, however, this bias may be small. Past statewide surveys of Coloradans’ perspectives towards wolf reintroduction have been conducted by mail (Pate et al., 1996) and phone (Meadow et al., 2005); thus, it is possible that differences in findings across surveys may reflect differences in survey recruitment rather than respondent attitudes over time. Despite this limitation, however, we believe our sampling strategy retains a high degree of validity because our survey respondents were recruited independently of their interest in and knowledge of the survey topic, and our recruitment and data weighting procedures were designed to obtain a representative sample of the Colorado population with regard to age, gender, and geography (Wang et al., 2015). Furthermore, a growing number of studies are using online platforms for data collection due to increasing technology awareness among the U.S. population (Ryan & Lewis, 2017) and given low response rates that are increasingly a challenge for mail and phone surveys (Keeter et al., 2017; Stedman et al., 2019). Finally, research on alternative survey methodologies indicates that some of the same limitations (e.g., response bias, violations of probability theory) associated with online panel surveys apply to other commonly used survey sampling techniques (Rivers, 2013).

Measurement

The survey included a variety of questions on voting intentions, attitudes and beliefs, experiences, and behaviors related to wolves, wolf reintroduction, and wolf management in Colorado (adapted from Pate et al. (1996) and other studies, as described below; see Supplemental Materials for all survey questions reported on). Additionally, we measured identification with various interest groups (e.g., hunters, ranchers, gun rights advocates, wildlife advocates, conservationists), building on Bruskotter, Vaske & Schmidt (2009) and Slagle et al. (2019), who found that identification with interest groups influences perspectives towards wolf conservation and management. Identification with interest groups was measured as a four-point scale, ranging from identifying “not at all” to identifying “a great deal.” We also measured demographics (e.g., income, age, gender, pet ownership) and location of residence. Final survey and administration procedures were approved by Colorado State University’s Institutional Review Board (protocol #19-8942H). Informed consent was obtained from all survey participants; in particular, participants were given the consent language at the beginning of the survey and were then asked if they agreed to participate. If they agreed, they were moved on to the rest of the survey questions; if they did not, their survey was ended. Quantitative analyses were performed and figures were produced using the R statistical environment (version 3.6.1). The data, codebook, and commented R code are provided in full as Supplemental Files for this article.

To address our first two research objectives, we report descriptive statistics on voting intentions, acceptability of various management options, and how intentions varied by demographics, interest groups, and geography. Voting intention was measured by asking “If you were given the opportunity to vote for or against reintroducing the gray wolf into Colorado, how would you vote?”, with responses of either for or against (Pate et al., 1996). To identify public acceptance of management practices, subjects were asked “If wolf reintroduction were to occur and wolves became reestablished in Colorado, is it acceptable or unacceptable in the future for wildlife management agencies to…”; followed by six potential strategies for participants to rate on a 7-point scale from highly unacceptable to highly acceptable. These strategies were adapted from Dietsch et al. (2011) and included: (1) limit the number of wolves if they cause declines in deer and elk populations in certain areas; (2) capture and lethally remove a wolf if it is known to have caused loss of livestock; (3) compensate landowners for loss of livestock caused by a wolf; (4) use a portion of state hunting and fishing license dollars to compensate landowners for loss of livestock caused by a wolf; (5) use a portion of state tax dollars to compensate landowners for loss of livestock caused by a wolf; and (6) allow a recreational hunt of wolves once they have reached a certain population size that exceeds recovery goals.

To address the third research objective, we report results from the following two survey questions: (1) “To what extent do you feel that wolves would have a direct impact on your livelihood or quality of life?”, measured on a 7-point scale, ranging from strong negative to strong positive impact; and (2) “Briefly describe why you feel wolves would negatively/positively impact your life.” The second question was only asked of those who indicated that wolves would impact them in their response to the first question. The second question was open-ended, so we used a thematic content analysis (Braun & Clarke, 2006) to categorize the most common responses shared among individuals who believed wolves would negatively or positively impact their life. Thematic codes were developed through an iterative process. A three-person team reviewed the survey responses independently and identified a total of 18 themes. Two raters then individually coded all survey responses to these themes, resulting in 83.6% rating concurrence. All non-concurrent responses were examined until 100% concurrence was attained.

Media analysis

To examine our fourth objective, we collected articles from the digital archives of the top 10 Colorado daily newspapers by circulation (Agility PR Solutions, 2019): The Denver Post, Colorado Springs Gazette, The Pueblo Chieftain, Grand Junction Daily Sentinel, Boulder Daily Camera, Fort Collins Coloradoan, Daily Times-Call, Greeley Tribune, Lakewood Sentinel, and Loveland Reporter-Herald. We used the keyword “wolf” to search for articles that were published from January 2019 (when the ballot initiative for wolf reintroduction was announced) through January 2020 (when the announcement was made that enough signatures had been gathered to put the initiative on the ballot). We removed any articles that did not explicitly discuss wolf reintroduction or that were opinion pieces or letters to the editor. This search resulted in 35 articles. We used the thematic codes from the analysis of the open-ended survey responses on perceived positive and negative impacts of wolf reintroduction to code the media articles, because we were specifically interested in whether the media coverage of wolf reintroduction portrayed the diversity of perceived potential impacts we identified through the survey. Two of the three coders who had participated in the coding of open-ended survey responses coded each media article (codes are reported in Table S4 of the Supplemental Materials). This coding process allowed us to examine whether the media reported the same potential impacts from wolf reintroduction reported by the public.

Level of public support for wolf reintroduction

Overall, 84.0% of Coloradans reported intention to vote for wolf reintroduction, while 16.0% reported intention to vote against (Fig. 2). Voting intentions were similar across the different regions of Colorado: 84.9% of Front Range residents, 79.8% of Western Slope residents, and 79.3% of Eastern Plains residents would vote for wolf reintroduction (Fig. 3A). The proportion that would vote for wolf reintroduction was relatively similar among residents in cities, towns, or rural areas and individuals with and without children (Figs. 3B and 3C). Pet owners were more likely to vote for wolf reintroduction (88.3%) than non-pet owners (76.4%; Fig. 3D). Voting intentions were broadly consistent across demographic categories, including gender, age group, income, and education (Figs. 3E–3H). Voting intentions were consistently supportive of wolf reintroduction (>80%) among those who identified (i.e., slightly, moderately, or strongly identified) and those who did not identify as gun rights advocates, property rights advocates, hunters, and ranchers (Fig. 4). Support remained above 80% among those who slightly and moderately identified as ranchers and hunters, but was lower among those who strongly identified as ranchers and hunters (69.5% and 66.1%, respectively; Fig. S1). Individuals who identified as wildlife advocates, animal rights advocates, and conservationists indicated greater support for reintroduction in their voting intentions (89.4%, 90.4%, and 87.6%, respectively) compared to those who did not (70.5%, 70.8%, and 74.7%, respectively; Fig. 4).

Support for wolf management options

Generally, respondents were split on each of the proposed management options, but some relevant differences emerged (Fig. 5). Slightly more than 50% of Coloradans believed the following management options were slightly, moderately, or highly acceptable: limit the number of wolves if they cause declines in deer and elk populations in certain areas; compensate landowners for loss of livestock caused by a wolf; and, use a portion of state hunting and fishing license dollars to compensate landowners for loss of livestock caused by a wolf. Slightly less than 50% of Coloradans believed that the following options were acceptable: allow a recreational hunt of wolves once they have reached a certain population size that exceeds recovery goals; and, use a portion of state tax dollars to compensate landowners for loss of livestock caused by a wolf. Approximately 50% believed that capturing and lethally removing a wolf if it is known to have caused loss of livestock was acceptable.

Perceived impacts on livelihoods and quality of life

Responses to the question of how wolf reintroduction would impact personal livelihoods or quality of life showed that 28.6% of Coloradans would expect wolves to impact them positively, 11.7% would expect wolves to impact them negatively, and 59.7% would expect no impact. More residents on the Western Slope (19.1%) believed wolves would impact them negatively compared to the Front Range (10.4%; Fig. 6). Correspondingly, fewer residents on the Western Slope (48.0%) believed that wolves would have no impact compared to the Front Range (61.6%).

Our coding of open-ended responses revealed a variety of different reasons for why respondents believed wolf reintroduction would positively or negatively impact their livelihoods or quality of life (Fig. 7; Table S4 provides detailed descriptions and examples quotes for each coded theme). The percentages of respondents who mentioned each theme reported are out of the total number of respondents (n = 320) who indicated that wolves would impact them.

The most common way respondents believed wolves would positively impact their lives was through restoring balance to ecosystems and thereby enhancing ecosystem health. This positive impact was mentioned by 19.4% of the respondents who reported that wolves would impact them. For example, one respondent wrote, “Bringing wolves back into the ecosystem of Colorado would result in a healthier balance to nature.” The second most commonly reported positive impact was seeing or hearing wolves in the wild (15.6%). Respondents discussed how they would enjoy encountering a wolf, and many even reported that they would spend more time outdoors to try to see wolves. Respondents also discussed their emotional or cultural connection to wolves or their belief that wolves were “beautiful” or “majestic” (10.3%).” Some respondents (9.4%) provided moral arguments for reintroduction, including: wolf reintroduction can help correct past wrongs; wolves should have the right to live where they used to; humans should learn to coexist with nature; and, preserving a species is the right thing to do. Another reason reported among 6.9% of respondents was that wolves would reduce pest populations, particularly rodents and deer. Other, less commonly reported reasons included: respondents valued the existence of wolves now or for future generations (4.4%); wolf reintroduction would increase tourism (2.2%); wolves would increase the diversity or abundance of wildlife in Colorado (1.9%); wolves would help people learn about or appreciate nature (1.9%); wolves would help regulate disease, reduce the number of sick ungulates, or provide other ecosystem services to people (1.9%); and wolf reintroduction would enhance Colorado pride or make Colorado a better place to live (1.6%; Fig. 7; Table S4).

Those who indicated that wolves would negatively impact their lives most commonly reported that they were worried about wolves posing a threat to human safety or wandering into residential areas and causing conflict with humans (19.1%; Fig. 7; Table S4). Respondents also mentioned being concerned about wolves reducing hunting opportunities (5.0%), harming pets (4.1%), and threatening the livelihoods of ranchers through livestock depredation (3.8%). Some respondents (2.2%) believed wolf reintroduction would lead to a reduction in wildlife (e.g., elk and deer) abundance. Finally, some respondents (1.9%) mentioned concerns that wolves were “killers” or “cruel” animals, while others (1.6%) were concerned that wolves would be poorly managed or difficult to manage. For example, several respondents discussed how wolves were managed in other states and were concerned about ranchers not being fairly compensated for livestock losses from wolf predation.

Media analysis

News articles discussing wolf reintroduction in Colorado included many of the same themes identified within open-ended survey responses (Fig. 7; Table S4). In the 35 articles analyzed, negative impact themes from survey responses were emphasized more frequently compared to positive impact themes; for example, 20 articles mentioned two or more negative impact themes, while 10 articles mentioned two or more positive impact themes. The average number of positive themes per news article was 1.2 (range 0–6), while the average number of negative themes was 1.8 (range 0–5).

The most commonly reported negative impact theme discussed in the media was concern about livestock depredation, mentioned in 51.4% of media articles. While a reduction in hunting opportunities was only explicitly discussed in 14.3% of articles, many (45.7%) brought up potential reductions in wildlife abundance (particularly deer and elk). News articles less often discussed the perceived threats of wolves to human safety (25.7%) and pets (11.4%), the difficulty of compensation and management (22.9%), and perceptions of wolves as killers (8.6%).

The most commonly reported positive impact theme discussed in the media (reported in 54.3% of the articles) was the possibility of wolves restoring balance to ecosystems and enhancing ecosystem health. Some articles mentioned the desire to see or hear wolves in the wild (11.4%); wolves being an important part of state pride (11.4%); the potential positive effects of wolves controlling disease in wildlife populations or providing other ecosystem services to humans (8.6%); moral arguments for the return of wolves (8.6%; e.g., “it’s the right thing to do”); the existence value of wolves (8.6%); and wolves leading to increases in wildlife abundance or diversity (8.6%). Fewer articles discussed perceived emotional or cultural connections to wolves (5.7%) and opportunities for wolves to enhance environmental learning and awareness (2.9%). Control of pests and boosts in tourism with wolf watching were not discussed in the reviewed media articles.

While consistent themes were present in both survey responses and media, they were disproportionate in representation. Concerns about the threat wolves pose to livestock were expressed in only 3.8% of survey responses, but were mentioned in 51.4% of media articles (Fig. 7). The potential for wolves to provide balance to ecosystems was mentioned in 54.3% of media articles compared to only 19.4% of survey responses. Other arguments about the positive impacts of wolf reintroduction identified from surveys, such as emotional connections to wolves, reductions in pest populations, and increases in tourism, were rarely or never mentioned in the media.

Outside of the themes emerging from survey responses, media articles also often discussed the following negative impacts or concerns associated with reintroduction: the fact that the Colorado Parks and Wildlife Commission had voted against reintroduction in the past; the potential for wolves to disrupt ecosystems; the belief that making decisions about wildlife management through ballot initiatives is problematic; and the ability of wolves to recolonize Colorado naturally, potentially negating the need for reintroduction altogether. This last point was particularly salient in media articles from January 2020, given that a wolf pack was documented as naturally moving into the state at that time. Additional arguments in support of wolf reintroduction that the media brought up were that the majority of Coloradans were in favor of wolf reintroduction, reintroduction would connect the entire North American wolf population from Mexico to the Arctic, and that the ballot initiative would be a democratic approach for allowing the majority of the public to have a say in decision-making about wildlife.