Wildlife Society Bulletin 45(3):465 – 472; 2021; DOI: 10.1002/wsb.1220 Research Article Determinants of Support for Funding Options for Fish and Wildlife Conservation in Wisconsin KOFI NKANSAH, 1 Bureau of Environmental Analysis and Sustainability, Wisconsin Department of Natural Resources, 101 S Webster St, Madison, WI 53707 USA BEN BEARDMORE, Bureau of Environmental Analysis and Sustainability, Wisconsin Department of Natural Resources, 101 S Webster St, Madison, WI 53707 USA ROBERT H. HOLSMAN, Bureau of Environmental Analysis and Sustainability, Wisconsin Department of Natural Resources, 101 S Webster St, Madison, WI 53707 USA ALAN R. COLLINS, West Virginia University, 4403 Agricultural Sciences, Morgantown, WV 26506 USA ABSTRACT Declining sales of hunting and angling licenses create budget challenges for state fi sh and wildlife management agencies. Using a mail survey, we obtained Wisconsin residents ’ opinions regarding funding for fi sh and wildlife management including options to raise licenses fees or create new funding mechanisms. We used an adjacent category, ordered logit regression model to explain respondents ’ support for the funding options contrasted against budget cuts. Our model revealed that respondents ’ self ‐ interest and demographic variables in fl uenced support for the revenue generating options. Most importantly, re - spondents ’ perceived bene fi ts of public lands and high interest in fi sh and wildlife consistently in fl uenced their support for increased funding options relative to requiring the Wisconsin Department of Natural Resources to reduce its operating costs. Rural respondents, however, opposed all options associated with raising revenues to fund conservation e ff orts. We suggest that gaining the support of rural hunters and anglers for any of the funding options will be imperative for the long ‐ term sustainability of fi sh and wildlife conservation funding in Wisconsin. © 2021 The Wildlife Society. KEY WORDS Conservation funding, fi shing, hunting, urban ‐ rural divide. State agency funding for fi sh and wildlife management pro - grams, with few exceptions, remain dependent on a user ‐ pay model (Jacobson et al. 2010). The 2 primary sources for agency budgets are license sales revenue and allocations of federal excise taxes on the manufacturers of fi rearms, hunting, and angling equipment (Williams 2010, Southwick and Associates 2012). Likewise, about 90% of the Wisconsin Department of Natural Resources ’ (WDNR) budget for fi sh and wildlife management comes from these 2 sources (Holsman 2016). Declining sales of hunting licenses across the U.S., com - bined with fl at sales of fi shing licenses, have raised concerns about the viability of the current conservation funding model (Shipley et al. 2019). Declining revenue from license sales impacts the ability of state agencies to generate the matching funds required to receive federal funding from the Pittman ‐ Roberson and Dingell ‐ Johnson excise taxes (Crafton 2019). Another aspect of the funding challenge is the lack of suitable funding mechanisms for manage - ment directed at nongame programs (Anderson and Loomis 2006, Holsman and Dunfee 2014). Revenue from licenses and federal excise taxes have been primarily applied to activities that directly bene fi t consumptive recreation opportunities, leaving other ecological priorities under - funded (Jacobson et al. 2007, Holsman and Dunfee 2014). State wildlife programs often lack fl exibility in directing funds generated from traditional constituents toward man - agement objectives not directly tied to game species man - agement. Even when federal funding has been allocated for state nongame wildlife programs, these allocations generally are not adequate to address the needs of nonhunted species (Paige 2000, Association of Fish and Wildlife Agencies 2015, Stein et al. 2018). In Wisconsin, declining license sales have resulted in budget shortfalls of several million dollars annually. As a result, WDNR has reduced many conservation activities (e.g., habitat management, conservation warden patrols, and invasive species control) by $ 20 million over the past 5 years, mostly by keeping vacant positions un fi lled (WDNR 2016). The Wisconsin Legislature directed the WDNR to develop a report in 2016 with options for addressing a structural Received: 3 April 2020; Accepted: 2 June 2021 Published: 21 September 2021 1 E ‐ mail: ko fi .nkansah@wisconsin.gov Nkansah et al. • Conservation funding 465 de fi cit in the state ’ s fi sh and wildlife account. The directive included the need to address both revenue increases and potential spending cuts. We conducted a statewide survey to assess Wisconsin residents ’ level of support for various fi sh and wildlife conservation funding strategies to assist in ad - dressing structural de fi cits in WDNR ’ s operating budget. Strategies considered by the WDNR ranged from poten - tial budget cuts to options for increased revenue. Devel - oping funding sources that reach audiences beyond hunters and anglers is an idea that has been advanced for both practical and normative reasons (Jacobson et al. 2010, Holsman and Dunfee 2014). Enactment of policy to raise revenue requires public support, especially among groups most vested in fi sh and wildlife conservation. We used data from a Wisconsin household survey to model factors pre - dicting support for potential fi sh and wildlife conservation funding policy options: (a) raising hunting and fi shing li - cense fees; (b) requiring access fees for all users of state fi sh and wildlife properties; (c) requiring all Wisconsin residents to pay for the conservation and management of fi sh and wildlife; (d) increasing the number of hunters, anglers, and trappers; or (e) requiring the WDNR to reduce its operating costs (e.g., cut sta ff and reduce services). Conceptual Framework of Respondents ’ Support Model Policy research has consistently shown that an individual ’ s policy preference is motivated by self ‐ interest (Deacon and Shapiro 1975, Hewitt 1985, Schokkaert 1987, Variyam et al. 1990). The concept of self ‐ interest as a primary driver of an individual ’ s policy preference is rooted in the random utility maximization theory. An individual is assumed to derive some level of utility from a choice among exhaustive and mutually ‐ exclusive alternatives in a choice set (Mukherjee and Rahman 2016). In the process of sup - porting or opposing a policy among alternatives, an in - dividual is assumed to have full information regarding al - ternatives available and the cost and bene fi t of each alternative is considered before choosing an alternative that maximizes utility (Lankford 1985, Mukherjee and Rahman 2016). Stakeholders ’ acceptance of, and support for, any policy change partly depends upon an under - standing of the bene fi ts that they derive from fi sh and wildlife conservation. Recreational outdoor activities, like any other products, have attributes that confer some level of bene fi t to the user and in fl uence demand. Hence, under - standing how these attributes in fl uence the demand for recreational outdoor activities and their relative import - ance to consumers ’ choices are important (Bennett and Blamey 2001). Within the self ‐ interest framework, an individual will be more likely to oppose a conservation funding option if it has a negative consequence to their self ‐ interest. In other words, an individual will oppose a funding policy if the costs in - curred exceed the bene fi ts derived from the option. Con - versely, individuals will support a fi sh and wildlife con - servation funding option if the bene fi ts derived from the option exceed the costs incurred. In assessing the demand for public spending policies, Ferris (1983) demonstrated that a random utility maximization theory can be adopted when modeling a public policy choice if the variables identi fi ed within the public good are perceived by in - dividuals as o ff ering bene fi ts to them. An appropriate proxy for an individual ’ s personal bene fi ts should be related either directly or indirectly to their interaction with the public good (Ferris 1983). METHODS Four policy options presented in this survey involved raising revenue from speci fi c groups of natural resource users or all citizens to support conservation e ff orts: (a) raising hunting and fi shing license fees, (b) requiring access fees for all users of state fi sh and wildlife properties, (c) requiring all Wisconsin residents to pay for the conservation and man - agement of fi sh and wildlife, and (d) increasing the number of hunters, anglers, and trappers. In contrast, a fi fth option directed WDNR to reduce its operating costs should real dollar revenue continue to decline (e.g., cut sta ff and reduce services; Fig. 1). The angler ‐ hunter variable in our model was assumed to capture the perspective of these consumptive users of Wisconsin fi sh and wildlife within the model. In this study, participation in fi shing or hunting was identi fi ed as a proxy to identify individuals whose personal bene fi ts derived from fi sh and wildlife resources are exclusive by their consumptive nature. In other words, bene fi ts from fi sh and wildlife re - sources to individuals who hunt or fi sh are both directly tied to the resource and rendered private through the harvest of that resource. We hypothesized that individuals engaged in hunting or fi shing would be more likely to value the bene fi ts of conservation to ensure continued replenishment of the resources they consume. However, we expected this same group of hunters and anglers to be less likely to support raising hunting and fi shing license fees, preferring instead options that distribute the cost of conservation to others. This expectation of hunters and anglers being less likely to support raising hunting and fi shing license fees is also consistent with the literature on the price elasticity of Figure 1. Fish and wildlife conservation funding options presented to Wisconsin, USA, residents in a mail survey conducted in 2016. 466 Wildlife Society Bulletin • 45(3) 23285540, 2021, 3, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1220 by University Of Florida, Wiley Online Library on [29/11/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License demand for hunting and fi shing licenses (Teisl et al. 1999, Poudyal et al. 2008). We expected that individuals with higher levels of interest in fi sh and wildlife would be more likely to value the ben - e fi ts of fi sh and wildlife conservation than those with little to no interest. We also expected such individuals to be less likely to support a reduction in services provided by the agency. Likewise, we expected individuals who perceived that public lands in Wisconsin bene fi t all citizens, not only users of public land, to value the bene fi ts of fi sh and wildlife conservation more than the cost of funding the e ff orts. Such individuals were hypothesized to be less likely to choose reductions in operating costs. Sociodemographic variables were also included in our model to capture di ff erences in preferences among in - dividuals (Ferris 1983). Variables included income level, age, sex, and race. Following Poudyal et al. (2008), we used 2 age classes ( < 35 or > 35 years old). Respondents were not asked to report income level in the survey, so we used the per capita income level for each respondent ’ s zip code as a proxy for income level, adjusted to 2016 dollars. Teel and Manfredo (2010) found that urbanization tends to be as - sociated with a higher prevalence of mutualistic value ori - entations towards wildlife, suggesting that urban residents may be more likely to support wildlife protection over human interest and oppose consumptive uses of wildlife than rural residents. Following Ratcli ff e et al. (2016), we used respondents ’ zip codes to classify them as urban or rural which was included in the model to test its in fl uence on policy choice. Questionnaire A 4 ‐ page questionnaire was mailed to a strati fi ed random sample of 2,000 Wisconsin residents with an oversampling of younger adults (18 to 40 years of age) in anticipation of lower response rates among that stratum (Gigliotti and Dietsch 2014). The sample of Wisconsin residents was purchased from a commercial marketing and survey fi rm (Survey Sampling Inc., Shelton, CT, USA). A follow ‐ up reminder letter was sent a week after the initial survey was mailed. We sent a second survey to nonrespondents 2 weeks after the initial survey. The survey was conducted during April and May 2016. We had a response rate of 35% (614 respondents). Hunters responded to the survey at a higher rate than expected based on statewide per capita partic - ipation rates. At the time of survey, the average age of adult hunters in the state was 47 years and 92% were male. As a result, older (ages ≥ 50 years old) and male respondents were over ‐ represented in our sample compared to the state census demographics (Table 1). The questionnaire consisted of 3 main sections: (a) re - spondents ’ perceptions and interests in Wisconsin public land, fi sh, and wildlife; (b) fi sh and wildlife conservation funding options; and (c) sociodemographic characteristics and respondents ’ participation in outdoor recreational ac - tivities. A statement that described how fi sh and wildlife management is funded in Wisconsin was presented to re - spondents before presenting the alternative funding options in the survey. The dependent variable was respondents ’ level of support or opposition for each of the 5 fi sh and wildlife conservation funding options measured on a 5 ‐ point scale (i.e., de fi nitely oppose to de fi nitely support). Explanatory variables used to capture respondent self ‐ interest motivations and direct participation in the con - sumption of fi sh and wildlife as a public good included response to questions that captured respondents ’ (a) level of interest in Wisconsin fi sh and wildlife, (b) perceived bene - fi ts of Wisconsin public land to all citizens, and (c) partic - ipation in hunting and fi shing within the last 2 years. Pre - sented with a statement “ What is your level of interest in Wisconsin ’ s fi sh and wildlife ” , respondents ’ level of interest were measured on a 5 ‐ point scale (i.e., great interest to no interest). Presented with a statement “ I think public lands in Wisconsin bene fi t all citizens of the state, regardless of whether they visit any public land ” , respondents ’ opinions were measured on a 5 ‐ point scale (i.e., strongly agree to strongly disagree). Presented with several outdoor recrea - tional activities (hunting, fi shing, trapping, and bird/wildlife watching or photography away from home), respondents were asked to report which activity they had participated in during the 2 years prior to the survey. Respondents who reported that they had participated in angling or hunting within the past 2 years were dummy coded to represent active participation in fi sh and wildlife. Empirical Model To predict support or opposition to the policy options re - garding conservation funding, we used an adjacent category ordinal logit model to predict support ratings as a function of independent variables due to the ordinal nature of the dependent variable. Although respondents were not pre - sented with a discrete choice question regarding funding options, the transitive principles of utility maximization choice are assumed to apply when comparing support rat - ings across funding options. The logit model for a policy option (Q) can be formulated as follows (Vermunt and Magidson 2005): Table 1. Characteristics of survey respondents from mail questionnaire on fi sh and wildlife conservation funding options administered in 2016 in Wisconsin, USA, to statewide demographics estimated from census data the same year. Demographic Respondents Census data Males 62% 49% White 95% 89% 18 ‐ 34 year ‐ olds 17% 26% 35 ‐ 49 year ‐ olds 22% 26% 50 ‐ 64 year ‐ olds 30% 28% 64 years and older 32% 19% Urban residents 56% 68% Mean per capita income by zip code $ 29,691 $ 28,297 Interested in natural resources 85% (weighted) Perceives bene fi t of conservation 94% (weighted) Hunt* 37% 15% Fish* 57% 20% * Hunting and fi shing rates of respondents were provided over a 2 ‐ time scale compared with annual participation rates. Nkansah et al. • Conservation funding 467 23285540, 2021, 3, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1220 by University Of Florida, Wiley Online Library on [29/11/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License ∑ β β η = + × × ⁎ = y z m m con m q Q q att q att 1 In this equation, η m is the systematic component of the support rating of category m , β m con is the category ’ s alternative speci fi c constant, ⁎ y m is the fi xed category score (here, sup - port rating ranges from 1 = de fi nitely oppose to 5 = de fi - nitely support), and β q att is the estimate of the contribution to the support rating of the option ’ s attributes, including the speci fi c policy option, characteristics of the respondent as well as any interaction e ff ects ( z q att ). The analysis accounted for the panel structure of the dataset (5 observed ratings made by each of 614 respondents) and was conducted using Latent Gold Choice 4.0 software (Vermunt and Magidson 2005). The fi nal model included 50 parameters, with continuous attributes coded using linear terms, and categorical attrib - utes dummy ‐ coded if they were binary or e ff ects ‐ coded to center each attribute ’ s values at zero if they contained 3 or more values (Bech and Gyrd ‐ Hansen 2005). While this is a large model, it is based on 2,926 observed rating events from 602 individuals, which is similar to other choice models reported in in the literature (e.g., Greene and Hensher 2003). Included parameters fell into one of 4 groups. The fi rst and second types of parameters included the alternative speci fi c constants representing the relative likelihood of a given rating in the absence of additional variables, and the main e ff ect of the policy option, modeled categorically. Respondent characteristics were brought into the model in the third and fourth groups of parameters. The third group of parameters captured the main e ff ects of re - spondent characteristics on funding option support, whereas the fourth group of parameters captured interaction e ff ects with 4 of the policy options (excluding the cost reduction option, which was chosen as the basis for comparison). Within the main e ff ect parameters of the model, a positive (negative) coe ffi cient is interpreted as a respondent ’ s propensity to support (oppose) the funding option. Where appropriate, the coe ffi cients of the parameters are presented in parenthesis. The main e ff ects of the funding options, however, assume baseline characteristics for the respondents; the inclusion of respondent characteristics is central to the predictive value of the model. Respondents ’ characteristics captured in the models in - clude 2 self ‐ interest variables (perceived bene fi ts derived from public land and interest in Wisconsin fi sh and wild - life), socio ‐ economic demographic variables (age, gender, income level, and urban/rural residence of respondent), and respondents ’ active participation in hunting and angling. Calculating 2 ‐ way interactions between these 8 respondent characteristics and 4 of the policy options was central to achieving our research objective, by capturing the di ff er - ential e ff ect of respondent characteristics on support for each option. In calculating interactions, the baseline option was to reduce the WDNR ’ s operating costs. Given that under the status quo, monetary in fl ation, and declining li - cense sales are e ff ectively reducing the purchasing power of the WDNR ’ s operating costs, this option was chosen to represent a do ‐ nothing approach to funding conservation. As a fi nal step, distributions of predicted support ratings for each of the 5 policy options were calculated. To make inferences about support for these options in the Wisconsin population, we had to account for di ff erences between our sample responses and characteristics of the general pop - ulation (Table 1). Therefore, we used mean population values derived from the U.S. Census Bureau rather than those of the sample. Where population data were not available (i.e., measures of perceived bene fi ts derived from public land and interest in Wisconsin fi sh and wildlife), the values used in the model were weighted to re fl ect census ‐ based socio ‐ demographic characteristics (U.S. Dept of In - terior et al. 2016). RESULTS Sixty ‐ two percent of the sampled population participated in fi shing or hunting in Wisconsin in the previous 2 years Approximately 24% of the survey respondents did not par - ticipate in any wildlife recreational activity (hunting, fi shing, trapping, and bird/wildlife watching or photography) in Wisconsin. Most respondents (85%) had moderate to great interest in Wisconsin ’ s fi sh and wildlife. Nearly all (93%) agreed with the statement “ I think public lands in Wisconsin bene fi t all citizens of the state ” , regardless of whether they visit any public lands. The majority of re - spondents were urban residents and had a mean per capita income higher than the state average (Table 2). Fewer urban respondents were engaged in hunting and fi shing than rural respondents. We observed a signi fi cant negative trend within the al - ternative speci fi c constants of the ordered logit model for respondents ’ support ratings (Table 3). The e ff ects of parameters associated with the policy op - tions indicate that all else being equal, respondents were less supportive of funding options that involved access fees for hikers, bird watchers and other users ( − 0.32) and all citizens paying for fi sh and wildlife conservation ( − 1.17). Con - versely, all else being equal, respondents were most sup - portive of reducing WDNR operating costs (0.92). The main e ff ects of respondents ’ characteristics in the model estimate their in fl uence on support for the base funding option (reducing WDNR operating costs). Re - spondents ’ interest in Wisconsin fi sh and wildlife and per - ception that Wisconsin ’ s public lands bene fi t all citizens decreased support for this option ( − 0.29 and − 0.46 Table 2. Demographic statistics (Rural vs. Urban Respondents) of survey respondents to a mail questionnaire on fi sh and wildlife conservation funding options administered in 2016 in Wisconsin, USA. Rural respondents Urban respondents Number of Respondents 269 336 Mean per capita income by zip code $ 26,424 $ 31,013 Hunters 48% 29% Anglers 63% 53% 468 Wildlife Society Bulletin • 45(3) 23285540, 2021, 3, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1220 by University Of Florida, Wiley Online Library on [29/11/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License respectively). Support for the base funding option was also lower among urban respondents ( − 0.18) than rural re - spondents. Respondents ’ income, gender, age, race, and participation in hunting or fi shing had no statistically sig - ni fi cant e ff ects on support for reductions in WDNR ’ s budget. To address the e ff ect of respondent characteristics on preferences for the remaining 4 funding options, interaction terms were included in the model. These parameters are additive to the main e ff ects, and the statistical signi fi cance of these parameters re fl ects di ff erences relative to the main e ff ect. The summative e ff ect is also included to assess the overall e ff ect of a respondents ’ characteristic on their sup - port for these funding alternatives, (Table 3; see the inter - actions column). A positive (negative) summative e ff ect for an attribute represents a positive contribution to re - spondents ’ support (opposition) for the funding option relative to the base funding option (i.e., reducing operating costs). Where appropriate the summative e ff ects are pre - sented in parentheses. Table 3. Adjacent categories ordinal logit regression to predict public support for conservation funding options based on a mail questionnaire of fi sh and wildlife conservation funding options administered in 2016 in Wisconsin, USA. Attributes β SE Interactions Model Constant (ASC) (1) de fi nitely oppose 0.985*** 0.361 (2) probably oppose 0.597*** 0.185 (3) unsure − 0.024*** 0.041 (4) probably support − 0.198*** 0.183 (5) de fi nitely support − 1.359*** 0.364 Funding Option Reduce the DNR ’ s operating costs 0.9244*** 0.325 Raise license fees on hunters, anglers and trappers 0.123*** 0.348 Increase the number of hunters, anglers and trappers 0.4417*** 0.394 Require hikers, birdwatchers and other users to pay − 0.3237*** 0.355 Create a mechanism whereby all residents pay − 1.1654*** 0.379 Characteristics Has moderate to great interest in natural resource issues − 0.290*** 0.097 − 0.276 Agrees that public lands bene fi t everyone − 0.479*** 0.131 − 0.476 Adjusted Income (natural log) − 0.007 0.030 − 0.008 Anglers and hunters 0.113 0.071 0.081 Male 0.048 0.071 0.051 Over 35 years old 0.003 0.002 0.003 Lives in an urban area − 0.176*** 0.068 − 0.168 White − 0.120 0.145 − 0.120 Interactions with “ Raise license fees on people who hunt, fi sh and trap ” Has moderate to great interest in natural resource issues 0.279** 0.138 − 0.011 Agrees that public lands bene fi t everyone 0.526*** 0.187 0.068 Adjusted Income (natural log) 0.006 0.045 − 0.002 Anglers and hunters − 0.433*** 0.099 − 0.320 Male − 0.122 0.099 − 0.074 Over 35 years old 0.000 0.003 0.003 Lives in an urban area 0.437*** 0.095 0.260 White 0.340 0.204 0.219 Interactions with “ Increase the number of hunters, anglers and trappers ” Has moderate to great interest in natural resource issues 0.345** 0.147 0.055 Agrees that public lands bene fi t everyone 0.879*** 0.191 0.420 Adjusted Income (natural log) − 0.006 0.051 − 0.013 Anglers and hunters 0.0657 0.108 0.178 Male − 0.0318 0.107 0.017 Over 35 years old 0.0011 0.003 0.004 Lives in an urban area 0.1271 0.103 − 0.049 White 0.1871 0.219 0.067 Interactions with “ Require hikers, birdwatchers and other users to pay ” Has moderate to great interest in natural resource issues 0.734*** 0.140 0.444 Agrees that public lands bene fi t everyone 0.526*** 0.187 0.067 Adjusted Income (natural log) − 0.030 0.047 − 0.037 Anglers and hunters − 0.009 0.099 0.104 Male 0.035 0.100 0.084 Over 35 years old 0.006** 0.003 0.009 Lives in an urban area 0.279*** 0.096 0.102 White 0.424** 0.204 0.304 Interactions with “ Create a mechanism whereby all residents pay ” Has moderate to great interest in natural resource issues 0.874*** 0.142 0.584 Agrees that public lands bene fi t everyone 0.914*** 0.192 0.455 Adjusted Income (natural log) 0.057 0.049 0.050 Angler and hunters − 0.168 0.099 − 0.055 Male 0.095 0.100 0.143 Over 35 years old − 0.003 0.003 0.000 Lives in an urban area 0.242** 0.203 0.066 White 0.406** 0.096 0.285 **, *** represent parameters signi fi cant at P < .05, and P < .01, respectively. Nkansah et al. • Conservation funding 469 23285540, 2021, 3, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1220 by University Of Florida, Wiley Online Library on [29/11/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Respondents ’ interest in Wisconsin ’ s fi sh and wildlife signi fi cantly increased support for all funding options rela - tive to its e ff ect on the decreasing WDNR operating costs (Table 3). Similarly, agreement that Wisconsin ’ s public land bene fi ts all citizens also contributed to increased support of all the alternative funding options presented relative to the reducing operating costs. The greatest positive e ff ects of these characteristics were on preferences for the funding option that required all residents to pay for fi sh and wildlife conservation. The summative contributions from respondents who had an interest in Wisconsin ’ s fi sh and wildlife was 0.46 and from those who perceived that Wisconsin public land bene fi ts all citizens was 0.58. The interaction e ff ect of respondents ’ participation in fi shing or hunting was only statistically signi fi cant for the funding option of raising license fees for hunting, fi shing and trapping (Table 3). In this case, respondents ’ partic - ipation in angling or hunting was associated with increased opposition relative to reducing operating costs. The inter - action e ff ects with the remaining 3 funding options were not statistically signi fi cant, indicating no additional in fl u - ence relative to reducing operating costs. The interaction e ff ect of living in an urban area was as - sociated with statistically signi fi cant increases in support for all options designed to increase funding, except for an in - crease in the number of hunters, anglers, and trappers (Table 3). The nonsigni fi cant parameter for the interaction of this characteristic on increasing the number of hunters, fi shers and trappers indicates no di ff erence from its e ff ect on reducing operating costs. There was a signi fi cant interaction e ff ect of race for re - quiring state property access fees and creating mechanisms whereby all residents pay (Table 3). In both cases, white Wisconsinites were more likely to support these options (with summative e ff ects of 0.30 and 0.29 respectively) than were other ethnic groups. For all but one funding option, respondents ’ age did not di ff er in its e ff ect on support rel - ative to its e ff ect on the base funding option (Table 3). Older respondents (those > 35 years of age) were sig - ni fi cantly more likely to show greater support (0.009) for requiring property access fees than were younger re - spondents. Respondents ’ gender and income level showed no statistically signi fi cant di ff erences in their e ff ect on support for all the alternative funding options presented relative to the base funding option. Together with data taken from the U.S. Census and U.S. Fish and Wildlife Service characterizing the population of Wisconsin (U.S. Dept of Interior et al. 2016), this model was used to predict population level support for each of the 5 funding options (Fig. 2.) The option that would receive the least support from the public is reducing the WDNR ’ s operating costs, with two ‐ thirds (67%) of Wisconsin resi - dents probably or de fi nitely opposing it. This option is followed by imposing access fees for other users, which would receive some level of opposition from 58% of resi - dents. Approximately half (52%) of Wisconsin residents would oppose raising license fees. Two options, however, would receive majority support by the public. Our model predicts that increasing the number of hunters, fi shers and trappers to fund conservation e ff orts would be probably or de fi nitely supported by 61% of residents, with opposition by only 22%. Interestingly, support for creating a mechanism whereby everyone pays is also predicted to be high, with 58% of the public supporting this option to some degree, while 24% is likely to oppose it. DISCUSSION Results from our study showed relative preferences among the public for increasing the number of hunters, anglers, and trappers, and creating a general tax mechanism based on known population parameters and mean scores on fi sh and wildlife interest and bene fi ts of public land survey questions. Figure 2. Model predicted distributions of support and opposition among 5 options for conservation funding presented to Wisconsin, USA, residents in a mail survey conducted in 2016. 470 Wildlife Society Bulletin • 45(3) 23285540, 2021, 3, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1220 by University Of Florida, Wiley Online Library on [29/11/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Our fi ndings imply that, once population demographics and opinions are taken into consideration, requiring all citizens to pay for fi sh and wildlife conservation no longer has the strongest opposition, but the second highest level of support. Our results show that support for conservation funding options was heavily driven by an individual ’ s characteristics and self ‐ interest. Our fi ndings also demonstrate the value of modeling interactions of policy options and respondent characteristics to understand support for fi sh and wildlife conservation funding. Our predictive model indicated that funding cuts were the least supported policy option. Given that a strong majority of respondents both expressed an interest in wildlife and a ffi rmed the bene fi ts of public lands, it should not be surprising that a reduction in the WDNR ’ s budget had the highest level of opposition. Without considering factors (interaction between re - spondents ’ characteristics, opinion variables, and the funding options) that in fl uence respondents ’ level of support for funding options, respondents ’ tendency to oppose options that required access fees for non ‐ consumptive users or require a tax on all residents was not surprising. Demand theory dictates that as the price of a good increases, the net bene fi t that an individual derives from the good decreases (Ferris 1983). Ex - cept for the option that raised fees on consumptive users, the consistent and statistically positive e ff ect of the 2 self ‐ interest variables on all the funding options in the model supports our hypothesis that fi sh and wildlife conservation confers some level of bene fi t to respondents ’ self ‐ interest. Respondents were informed in the survey that a reduction in the WDNR ’ s op - erating cost implied a reduction in services provided along with the reduced number of employees. Within a respondent ’ s cost ‐ bene fi t calculus, a potential reduction in WDNR services would not be in the self ‐ interest of those who perceived that fi sh and wildlife conservation confers a personal bene fi t to them. As a result, these respondents were more likely to report lower support ratings for the funding option that reduced the WDNR ’ s operating cost. Respondents who hunted or fi shed tended to oppose in - creases in license fees on hunting, angling and trapping activities. According to the utility maximization framework, increasing the price of a good negatively impacts the cost ‐ bene fi t calculus of a respondent ’ s choice among alternatives. Given that an increase in license fees will impact anglers and hunters directly, it is not surprising that consumptive users were more likely to oppose the conservation funding option that required an increase in license fees. Our results also corroborate other studies that have found a negative rela - tionship between fi shing and hunting license demand and price (Teisl et al. 1999, Poudyal et al. 2008). Urban respondents ’ opposition to WDNR budget cuts is perhaps indicative of perceptions that reductions would lead to diminished bene fi ts derived from fi sh and wildlife con - servation. Support for having all residents pay for con - servation among urban residents may stem from conformity of this option to value orientation of protecting Wisconsin ’ s natural resources for the mutual bene fi t of all residents (user and non ‐ user). Although research that has established connections between urbanization and mutualistic wildlife values (e.g., Teel and Manfredo 2010) may o ff er a partial explanation, it is also likely that distinct political ideologies in urban versus rural areas also drive di ff ering opinions about raising taxes to fund government services (Pew Re - search Center 2018). Given that wildlife ‐ related activities across the country are predominantly dominated by white individuals (U.S. Dept of Interior et al. 2016), higher support among these individuals for the options that require access fee for users of state property (hikers, birdwatchers and other users) and creating a mecha - nism where all users pay was unexpected. Because price as an attribute of a good negatively impacts the cost ‐ bene fi t calculus of demand, one would expect white respondents to be less likely to support the conservation funding options that require access fees for non ‐ consumptive users of state property or re - quire new fees on non ‐ users of state properties. Even though income as a unique variable was not statistically signi fi cant in our model, a positive relationship between income and de - mand for recreational activities has been established in several studies (Teisl et al. 1999, Poudyal et al. 2008). Income dis - parity between whites and other races has been extensively documented across the country and the trend in Wisconsin is not di ff erent (Dresser and Rodriguez 2017, Smeeding and Thornton 2018). Perhaps the higher level of support among white respondents for mechanisms that require increased payment may re fl ect the presence of an income disparity and thus ability to pay between white and non ‐ whites Our research developed a predictive model of respondents ’ level of support for conservation funding options among 5 choices, with an underlying assumption that the transitive principle of utility maximization applies when comparing options. In other words, we did not ask respondents to directly compare one funding option to another, but rather assumed respondent rationality, such that a probably sup - port option would be preferred to an option the respondent would probably oppose. This assumption is inherent in our use of an ordinal logit model. An alternative approach that does not maintain this assumption would be to ask re - spondents to make numerous discrete choices among funding options and then model these choices. Future research employing discrete choice methods is encouraged to test the robustness of our fi ndings. MANAGEMENT IMPLICATIONS As state agencies explore options to fund fi sh and wildlife management, it will be crucial to understand policy support among various segments of the public. Understanding the impact of sociodemographics as well as attitude measures that re fl ect consumer bene fi ts from public lands, fi sh, and wildlife will help agencies tailor their message on the need for alternative funding options to ensure long ‐ term sus - tainability. Our results suggested that, in Wisconsin, there may be more public support for general tax mechanisms and increasing the number of hunters, anglers, and trappers than for increased or additional user fees. The results have im - plications for considering how to adopt the user ‐ pay model of wildlife conservation as societal preferences and wildlife Nkansah et al. • Conservation funding 471 23285540, 2021, 3, Downloaded from https://wildlife.onlinelibrary.wiley.com/doi/10.1002/wsb.1220 by University Of Florida, Wiley Online Library on [29/11/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License needs have changed since the current funding system was established. If wildlife mana