Socio-Economic and Environmental Trade-Offs of Sustainable Energy Transition in Kentucky

Abstract

A just and sustainable energy transition in historically coal-dependent regions like Kentucky requires more than the adoption of new technologies and market-based solutions. This study uses a stated preferences approach to evaluate public support for various attributes of energy transition programs, revealing broad backing for moving away from coal, as indicated by a negative willingness to pay (WTP) for the status quo (–USD 4.63). Key findings show strong bipartisan support for solar energy, with Democrats showing the highest WTP at USD 8.29, followed closely by Independents/Others at USD 8.22, and Republicans at USD 8.08. Wind energy also garnered support, particularly among Republicans (USD 4.04), who may view it as more industry-compatible and less ideologically polarizing. Job creation was a dominant priority across political affiliations, especially for Independents (USD 9.07), indicating a preference for tangible, near-term economic benefits. Similarly, preserving cultural values tied to coal received support among Independents/Others (USD 4.98), emphasizing the importance of place-based identity in shaping preferences. In contrast, social support programs (e.g., job retraining) and certain post-mining land uses (e.g., recreation and conservation) were less favored, possibly due to their abstract nature, delayed benefits, and political framing. Findings from Kentucky offer insights for other coal-reliant states like Wyoming, West Virginia, Pennsylvania, Indiana, and Illinois. Ultimately, equitable transitions must integrate local voices, address cultural and economic realities, and ensure community-driven planning and investment.

1. Introduction

In the last two decades, US coal production has declined by 45.7% to 594,155 thousand short tons in 2022 [1]. During this same period, Kentucky experienced a sharper decline of 77.0%, with production dropping to 28,527 thousand short tons [1,2] (Figure 1). While environmental regulations by the Environmental Protection Agency (EPA) are often cited as the primary cause of this downturn, they account for only 3.5% of the 33% decline in US coal production [3]. Other factors, such as diminishing coal reserves, falling prices of market substitutes (such as shale gas and natural gas), technological advancements in wind and solar, coal-plant retirements, evolving technologies in coal mining and energy generation, and relatively stable demand for electricity, have also contributed to the decline of the coal industry [2,3].

Declining coal reserves have made mining increasingly labor-intensive and costly, while factors such as coal mine productivity, heat content, higher sulfur content, spot pricing, and transportation costs have further reduced coal market demand [2,3]. The advent of hydraulic fracturing (fracking) has led to the surge in cheap natural gas, positioning it as a more economically attractive alternative to coal. Simultaneously, market volatility and shrinking profit margins have discouraged investments in coal, prompting a financial shift toward renewable energy sources [3,4,5].

Global demand for coal, particularly from major importers such as China and India, has negatively impacted US coal exports. At the same time, technological advancements (including automation and mechanization in mining) have reduced the need for labor, contributing to widespread job losses in coal-dependent regions [3,5,6]. Many existing coal-fired power plants are aging and inefficient, making them expensive to retrofit and more likely to retire in favor of modern energy sources. The modernization of the energy grid, including smart grid adoption with the adoption of decentralized power generation, has further reduced reliance on centralized coal-fired plants [3,7,8].

Environmental and regulatory factors have also played a role in coal’s decline. The management of coal combustion residuals (e.g., coal ash) imposes additional costs and oversight. Air quality standards and climate policies aimed at reducing greenhouse gas emissions have made coal plants more expensive to operate. Furthermore, coal extraction and processing require significant water resources, raising environmental concerns and intensifying conflicts over water use [3,9,10,11].

A growing impetus to address the effects of global warming has led to a significant shift in climate mitigation strategies, with an increasing focus on the adoption of renewable energy sources [12]. However, Cha [12] acknowledges that in regions historically dependent on coal extraction, such as Kentucky, the transition from fossil fuel-based economies to renewable energy systems has encountered considerable challenges. These challenges may span social, political, cultural, and economic dimensions. Socially, the decline of the coal industry and the rise in renewable alternatives have profound implications for workers and communities. Job displacement involves not only the loss of employment but also the need to adapt to new industries with lower wages, longer commutes, and the erosion of community bonds rooted in coal mining traditions [3,13,14,15]. These traditions represent a strong cultural identity forged around shared experiences in mining communities. As Mayer [14], Cha [12], and Lewin [16] emphasize, this cultural attachment makes the idea of transition deeply unsettling, threatening not just livelihoods but long-held values and ways of life. Consequently, preserving cultural heritage becomes an essential, yet often overlooked, aspect of just transition strategies.

Culturally, the challenge is more nuanced. In coal-mining communities, coal is not merely an industry; it is a symbol of identity, pride, and self-reliance [16]. Generations have worked in the mines, with entire towns shaped by coal’s economic footprint. Bell and York [17] describe this as a “community economic identity,” where mining defines collective self-worth. Mayer [14] notes that coal communities often reminisce about the peak of coal production, with organizations like “Friends of Coal” reinforcing nostalgic narratives that promote environmental deregulation in the name of cultural preservation. These deeply embedded values can generate strong resistance to renewable energy, even when coal is no longer viable. Hence, any transition must respect local culture and avoid imposing new systems that ignore community identity [12,14]. From a political perspective, energy transition debates remain polarized. Cha [12] and Lockwood [18] see that liberal politicians champion renewable energy as a climate solution, while conservative politicians are often aligned with fossil fuel interests. In coal-reliant areas, political resistance is amplified by the close alignment of economic and political power with coal interests. These dynamics create political inertia, making the implementation of energy transition policies more difficult. Economically, the effects of coal decline are substantial. Coal once provided a huge tax revenue for public services like education, healthcare, infrastructure, and waste management [15]. The closure of mines often leads to fiscal shortfalls, exacerbating socio-economic vulnerabilities. While renewable energy holds long-term promise, it may not immediately replicate coal’s economic role. The timing mismatch between the loss of fossil fuel revenues and the growth of renewable industries creates a period of economic fragility for transition efforts [3,12,14,15,19,20].

Despite the downturn, the coal industry continues to play a major role in Kentucky’s economic and energy policy [21]. Stakeholders increasingly recognize the need to manage this decline while promoting viable alternatives. Benefits of the transition include reduced air and water pollution and new revenues from renewable investments. However, these are accompanied by challenges such as job losses and expensive mine reclamation efforts [22]. In Kentucky, the shift from coal to renewables must be guided by a sustainable policy framework that reflects the political, cultural, and economic realities of coal-reliant communities [3,12,13]. Rising social inequalities due to transition-related disruptions highlight the need for equitable policies that create a level playing field for affected populations [23]. An integrated approach that incorporates political, cultural, and economic values will be critical to delivering a just and sustainable energy transition.

Many scholars [12,14,19,20,22,23,24] admit that energy transitions must not only address the technical and environmental considerations but also the socio-economic impacts that arise. A just energy transition must reduce existing inequalities in the current energy system while preventing the creation of new ones [12,24]. Thus, energy policy must account for these dimensions to secure broad public support [23,25]. A central concern of a just transition framework is identifying how different populations perceive the benefits and burdens of energy transitions. These perceptions shape the social acceptability of new policies and can either bridge or widen the gap in public engagement. Bridging this gap requires policies that mitigate adverse economic effects while ensuring social equity. These complex challenges also create a unique opportunity to develop hybrid policies that integrate environmental, economic, and social priorities. Rather than simply replacing coal with renewables, the transition must also redefine how energy is provided through inclusive, community-benefiting programs. This underscores the need to center the voices of those most affected by the transition [25].

This study proposes a framework to evaluate the socio-economic factors influencing Kentucky’s energy transition. Using a stated preference approach, we aim to assess public preferences for sustainable energy transition programs. The scientific novelty of this article lies in its comprehensive integration of diverse attributes within a discrete choice experiment (DCE) framework to evaluate the multi-faceted public preferences for energy transition programs in a historically coal-dependent region. Unlike previous studies that often focus on isolated components or single variables, this research specifically evaluates the trade-offs between social, economic, environmental, and cultural dimensions. Furthermore, it defines cultural values through a more diverse and context-specific set of attributes, directly reflecting the lived realities and priorities of Kentucky residents. This holistic approach provides a more robust understanding of public preferences and perceived fairness, offering insights crucial for developing socially acceptable and economically equitable policies. This engagement process is designed to support the development of policies that are socially acceptable and economically equitable, reducing resistance and mitigating the adverse impacts of a poorly managed energy transition. Ultimately, a successful transition must go beyond technological and market-based solutions. It must address political resistance, offer targeted support, and respect the heritage of the communities most affected.

Review of Studies

Previous studies on energy transitions often focus on isolated components such as social support programs (weatherization, retraining, medical compensation), types of renewable energy sources, job creation, post-mining land use, preserving cultural values, and willingness to pay. In contrast, our DCE integrates a broader set of attributes that collectively reflect the social, economic, and environmental dimensions of the clean energy transition. For instance, Scheer et al. [26] disaggregated renewable energy sources such as onshore and offshore wind projects, solar photovoltaic (PV), and solar thermal technologies. While their study emphasizes energy portfolios, ours expands the scope to include potential clean energy alternatives such as biomass, wind, and solar energy that can replace coal in Kentucky. Singh et al. [27] explored how post-mining land use, such as landfill development or rehabilitation for agriculture, can contribute to job creation and ecological recovery. These attributes emphasize regional regeneration and are particularly important for coal-dependent communities.

Other studies, such as Awuah-Offei [28] and Que [29], tend to focus narrowly on single variables like utility, cost, the net present value (NPV) of social risks as it relates to mine planning, and new job creation as the sole drivers of public support for energy transition. In a separate study, Duran et al. [30] conducted a DCE on maritime heritage using cultural attributes such as fishermen’s knowledge, fishing architecture, traditional boats, folklore, and other traditional knowledge to explore how cultural values can be preserved through festivals, workshops, and rebuilding architecture. While informative, their study focused exclusively on cultural preservation using primarily quantitative indicators. In contrast, our study defines cultural values through more diverse and context-specific attributes, as outlined in

Table 1. Attributes and attribute levels for the sustainable energy transition program.

Attribute	Description	Levels
Possible clean energy alternatives (PCEAs)	Energy source that will replace coal in energy generation.	Biomass (wood and wood processing wastes) * Wind Solar
Post-mining land use (PLU)	Benefits of post-mining land use, both surface and closed mines. This will involve restoration and an additional benefit that will be voted for by the respondents.	Restoration and timber pro-duction * Restoration and conservation (reintroduction of endangered species and improving biodiversity) Restoration and recreation (developing nature trails and camping sites)
New job creation (NJC)	New job opportunities developed in the clean energy sector, recreation, museums, social support programs, and auxiliary services (transport, maintenance of energy infrastructure, etc.).	320 new jobs/year * 640 new jobs/year 1280 new jobs/year
Preserving cultural values (PCV)	Preserving cultural values of coal mining communities through coal museums and heritage sites, archival projects, education, and public engagement (curriculum integration, public exhibits and festivals, and documentaries and festivals), heritage tourism, crafts and local art, coal town revitalization (through museums and art), and sustainable coal mining practices.	No preservation of cultural values * Preservation of cultural values
Social support programs (SSPs)	Support through retraining of workers/communities who have lost their jobs and medical compensation for workers/communities affected by environmental impacts of coal mining. Energy efficiency programs to reduce electricity bills.	Provision of medical compensation * Energy efficiency programs Retraining programs
Cost (payment vehicle)	The energy transition program will be supported by an additional tax to the monthly utility bill administered by the local utility company (gas and electricity) and paid by all households in the state.	USD 1/household/month * USD 4/household/month USD 7/household/month USD 10/household/month

Note: * indicates that the attribute levels biomass, restoration, and timber production, no preservation of cultural values, provision of medical compensation, and USD 1/household/month are baseline variables.

. These include coal museums and heritage sites, archival projects, curriculum integration, public exhibits and festivals, heritage tourism, crafts and local arts, coal town revitalization, and sustainable coal mining practices. Building on insights from Baute [31], who emphasizes the importance of unemployment benefits and low-income household subsidies in the absence of weatherization programs, our DCE includes retraining programs, medical compensation, and energy efficiency programs as key social support programs (SSPs) that reflect Kentucky’s socio-economic needs. Finally, the cost or WTP attribute captures the economic trade-offs involved in switching from coal-based energy systems to renewable alternatives. Unlike Danne et al. [32], who directly estimate WTP values as parameters, we derive these values from marginal utility coefficients, which allows us to more clearly communicate the trade-offs faced by respondents.

Overall, our study incorporates a comprehensive and context-sensitive set of attributes that more accurately reflect the lived realities and priorities of Kentucky residents. Our studies also differ by evaluating not only these individual attributes but also the trade-offs between them. Hence, we provide a more holistic understanding of public preferences and the perceived fairness of energy transition, thereby contributing to the broader discourse on just transitions. A detailed review of previous studies and their relation to our attributes is further described in Section 2.3.

This paper is structured as follows: Section 2 explains the methodological and theoretical approach, including a detailed discussion of our six key attributes. Section 3 presents our results and discusses the Random Parameter Logit (RPL) estimates for the attributes and marginal WTP measures; Section 4 concludes this study and provides policy implications.

2. Materials and Methods

2.1. Theoretical Framework

This study applies the DCE method, estimated using an RPL model, to assess individual preferences for hypothetical energy transition programs in Kentucky. The DCE framework is based on Lancaster’s theory of consumer choice [33] and Random Utility Theory (RUT) [34]. DCEs, informed by RUT, allow researchers to estimate the trade-offs individuals are willing to make between different policy attributes by observing their selections among hypothetical alternatives. These foundational theories model utility as a function of observed attributes, individual characteristics, and unobserved stochastic components.

The utility (Ujtq) that respondent q (q = 1,Q) derives from selecting alternative j (Option A, B, or C) in choice situation t (t = 1,6) is given by

U_jtq = V_jtq + ε_jtq ≡ β′_qkX_jtqk + δ′_kz_qZ_jtqk + ε_jtq

(1)

The utility consists of a deterministic component (Vjtq) and a stochastic term (εjtq). The deterministic component is modeled as a function of a vector (k) of choice-specific attributes (Xjtqk), with associated parameters (ßqk). These parameters vary randomly across respondents due to preference heterogeneity, following a distribution with a mean (ßk) and standard deviation (δk).

The utility function of the model, without covariates apart from the error term (εjtq), is expressed as a linear function of an attribute vector (X1, X2, X3, X4, X5, X6), which represents the following attributes: possible clean energy alternatives, post-mining land use, new job creation, preserving cultural values, social support programs, and cost. The model also includes an alternative-specific constant (ASC) to capture the utility associated with selecting the status quo option, which accounts for factors not explicitly included in the attribute vector (Xjtqk). The RPL framework accounts for preference heterogeneity by incorporating individual characteristics such as the socio-demographic variables (gender, age, education, employment rate, and political party affiliation) (Zq). These characteristics interact with choice-specific attributes (Zjtqk), enabling the identification of how individual specific characteristics influence preferences for different alternatives [35,36].

Hence, Equation (1) becomes

V_jq = ASC_q + ß₁X_1qj + ß₂X_2,qj + ß₃X₃,_qj + β₄X_4,qj + β₅X_5,qj + β₆X_6,qj

(2)

The probability that an individual q selects option i over another option j within a given choice set t is determined by

Piq = Pr (Viq + εiq > Vjq + εjq)   ∀ j ∈ t

This can also be expressed as

P_iq {(V_iq − V_jq) > (ε_jq − ε_iq)}

(3)

Assuming the error terms are distributed independently and normally distributed according to a type I extreme value distribution, the choice probability becomes

P_i = exp (µV_iq)/Ʃjexp(µV_jq) ∀ j ∈ t

(4)

Here, µ is a scale parameter inversely related to the variance of the error term and influences the degree of randomness in the choice behavior. The deterministic utility component Vjq can be expressed more generally as

V_jq = Ʃ_kβ_jkX_jk

(5)

The implication of Equation (5) is that the estimated β values cannot be directly interpreted, as they are confounded with the scale parameter µ. However, the marginal rate of substitution (MRS) between any pair of attributes can still be determined because the scale parameter cancels out in the ratio, as shown:

MRS = −(µ.β_{attribute a}/µ.β_{attribute b}) = −(β_{attribute a}/β_{attribute b})

(6)

When cost is included as an attribute, Equation (6) can be used to estimate the “implicit price” (p*a) or marginal WTP for an attribute.

P*_a = −(β_a/β_cost)

(7)

The implicit prices (p*a) represent the trade-off that respondents are willing to make between a given attribute and cost, offering insights into how much respondents value changes in that attribute. This approach allows us to quantify preferences and examine how individuals prioritize the various attributes of hypothetical energy transition programs. The RPL model was preferred for this study because of the complex and polarized nature of energy transition debates in regions like Kentucky, where attitudes towards clean energy, post-coal economic development, and environmental values are shaped by distinct political and cultural identities. The attributes analyzed in this study are not likely to be valued uniformly across respondents, making it imperative to account for unobserved heterogeneity in preferences. Moreover, the RPL model aligns with theoretical approaches in environmental sociology and political economy, which emphasize how individual choices related to environmental policy are shaped by broader ideological, cultural, and institutional structures [17].

2.2. Sampling Framework and Data Collection Protocol

The study’s sampling strategy is based on recent U.S. Census data to ensure that the sample accurately reflects Kentucky’s socio-demographics (Table 2). The target respondents are residents of Kentucky aged 18 and older. Sampling will consider both urban and rural populations and distinguish between counties with and without coal mining activity.

2.3. Identifying Attributes

To assess public preferences for energy transition programs in Kentucky, six attributes were identified through a review of the literature (Table 1). These attributes include possible clean energy alternatives (PCEAs) [37,38,39], new job creation (NJC) [37,38,40,41], preserving cultural values (PCV) [16,42,43,44], post-mining land uses (PLUs) [45,46,47,48,49], social support programs (SSPs) [50,51,52], and cost [37,38,53,54].

Also, three of these attributes, PCEAs, NJC, and PCV, were based on key beliefs from Appalachian coal culture as described by Lewin [16]. Lewin [16] summarizes coal heritage to be rooted in four main beliefs: (1) replacing coal with renewable energy sources would raise electricity costs; (2) coal provides good jobs in an area with limited economic opportunities; (3) mining symbolized community value and (4) miners upheld the dignity of rural life in the face of urban onslaught. We characterize these beliefs to represent hypothetical yet plausible attributes to assess the relative preferences for attributes and inherent trade-offs against cultural values. From a policy perspective, the attributes were designed to be both feasible and actionable for adoption by decision-makers. To enhance respondent comprehension, we used clear, concise, and accessible language to describe and explain the attributes and corresponding levels, incorporating a cheap talk script to mitigate hypothetical bias. For simplicity in the survey design, the attribute levels expected to have the lowest utility were designated as the omitted variables across all attributes (Table 1).

The first attribute was PCEAs, represented by the levels of biomass, solar, and wind as three viable renewable energy options to replace coal in energy transition. Notably, geothermal and hydropower were excluded from this study due to specific limitations. Despite state-level initiatives to expand hydropower, challenges such as aging infrastructure, flood control issues, and navigation risks diminish its suitability [55,56]. Correspondingly, Kentucky’s geology limits large-scale geothermal generation, with its applications restricted to heating and cooling through geothermal heat pumps [48,57].

Kentucky’s wind energy potential has historically been constrained by low wind speeds, complex topography, competing land uses (such as forest and farmland), and the absence of enabling policy frameworks [58,59]. However, advancements in wind technology improved efficiency in capturing lower wind speeds, thereby increasing wind energy viability. Furthermore, federal support through the Inflation Reduction Act (IRA) provides tax credits covering up to 50% of capital costs in designated energy communities [60]. The Louisville Gas and Electric (LG&E) company in Louisville, Kentucky is piloting utility-scale wind projects to assess feasibility [61]. Furthermore, wind energy and the manufacturing sector have the potential to create up to 10,000 jobs by 2030 [62]. Integrating wind energy with solar power can enhance grid reliability during periods of low solar output [60].

Kentucky has a moderate solar energy potential, with an average solar radiation of 4.94 kWh/m²/day, comparable to Germany, a leader in solar adoption, implementing an installed solar capacity of 104.8% GW [63,64]. As of 2024, Kentucky’s total installed solar capacity was 250 MW, indicating an over 250% increase since 2020 [65,66]. Although Kentucky lacks Renewable Portfolio Standards (RPSs), it has implemented net-metering policies that allow the installation of up to 45 kW for residential and commercial users [63]. An emerging opportunity includes the conversion of former mine lands into large-scale solar farms, offering dual benefits of land reclamation and economic revitalization [64].

Despite substantial potential, biomass contributes just 0.6% to electricity generation in Kentucky. Kentucky has 12 million acres of forest, generating up to 2.6 million tons of timber residues annually, which could be harnessed for co-firing with coal [67,68]. Moreover, 20% of Kentucky’s land is underutilized, including 4.69 million acres of agricultural land and 740,000 acres of post-mining marginal land [69]. These lands could be used to cultivate native grasses like switchgrass (Panicum virgatum), which alone could yield up to 22.8 million tons, capable of meeting 17.2% of Kentucky’s energy needs [70].

The second attribute in this study, PLU, relates to potential land uses for former mining sites and includes three levels: restoration and conservation, restoration and recreation, and restoration and timber production. The restoration and conservation level focuses on improving environmental quality through soil stabilization, water resource enhancement, biodiversity conservation, and ecosystem restoration [45]. The restoration and recreation level reflects the conversion of post-mining landscapes into areas for tourism and outdoor activities. Given the unique cultural and aesthetic appeal of abandoned mines, such sites offer potential for adventure tourism (e.g., hiking trails, all-terrain vehicle (ATV) parks, camping sites, theme parks, zip-lining), underground museums, and wildlife reserves [45,48]. The third level, restoration, and timber production, highlights the potential of reforestation and timber production of restored land, for commercial timber and carbon sequestration [46].

The third attribute, NJC, captures the number of jobs generated from the hypothetical energy transition program. The jobs considered include direct employment in renewable energy as well as indirect opportunities in the tourism, restoration, and agroforestry sectors. NJC was categorized into three levels: low (320 new jobs), medium (640 new jobs), and high (1280 new jobs). These levels were selected based on employment data and projected growth trends. According to the U.S. Energy and Employment Report [71], Kentucky’s electric power generation sector added 356 jobs in renewable energy, representing a 7.7% increase from the prior year. To reflect a plausible range for a choice experiment, the base level of 320 jobs was determined, while higher levels were generated using a doubling structure informed by national projections, indicating a potential 47.9% job increase in the renewable energy sector [71,72].

The fourth attribute, PCV, addresses the preservation of cultural identity and heritage linked to coal mining communities in Kentucky. The attribute has two levels: preservation and no preservation. Preserving cultural values involves repurposing coal-related infrastructure and narratives in ways that honor the historical significance of mining [42]. Examples include the creation of mining museums, cultural heritage trails, and community events that celebrate mining traditions, promoting social cohesion and cultural continuity in the wake of energy transition. It further supports the gradual public acceptance of renewable energy by respecting deeply rooted regional values associated with labor, identity, and rural dignity.

The fifth attribute, SSPs, is central to ensuring community resilience during energy transition, creating safety nets, and building adaptive capacity for impacted households. Informed by adaptive capacity frameworks [50,51,52], the SSP attribute includes three levels: energy efficiency programs, job retraining, and medical compensation. These support systems target vulnerable populations affected by mine closures, unemployment, or coal-related health issues. Medical assistance was selected over traditional weatherization programs to better reflect public concerns about long-term health impacts from mining.

The final attribute is a cost variable designed as a payment vehicle, capturing respondents’ WTP through additional monthly utility bill charges to support the energy transition program. Cost levels were set at USD 1, USD 4, USD 7, and USD 10 per household per month, drawing from prior stated preference studies on renewable energy preferences [37,38,53,54].

2.4. The Questionnaire and Sampling Framework

The survey instrument was divided into four main sections, each designed to guide respondents through the context, task, and collection of relevant data. The first section provided the research context, ensuring that respondents understood the broader relevance of their participation and reaffirming confidentiality and voluntary participation.

The second section provided a concise introduction to the purpose of the survey, including background information on coal mining, energy transitions, renewable energy technologies, environmental implications, and relevant government policies. This section ensured that respondents had a consistent understanding of key concepts and terminology, serving as a ‘cheap talk’ script to reduce hypothetical bias. Additionally, it included a set of questions to assess respondents’ baseline awareness and understanding of energy transition-related terms.

The third section of the questionnaire collected socio-demographic information, including gender, age, educational attainment, occupation, household income, and place of residence (mining/non-mining counties). These variables enabled subgroup analysis and heterogeneity assessment within the DCE.

In the fourth and final section, respondents were introduced to the DCE. Participants were asked to evaluate six hypothetical energy transition programs, each defined by a unique combination of attributes and levels as outlined in Table 1. The full factorial design, based on the combination of six attributes with varying levels (3 × 3 × 3 × 2 × 3 × 4), would have produced 648 unique profiles. To ensure feasibility and minimize respondent fatigue, a D-efficient fractional factorial design was generated using the R statistical software, version 4.4.3. This design optimized orthogonality, level balance, and minimum attribute overlap, yielding 72 unique profiles. These profiles were randomly paired to form 36 choice sets, divided into 6 blocks, with each respondent randomly assigned 1 block consisting of 6 choice cards. Each choice card presented respondents with a discrete task involving the selection of a preferred energy transition program from two alternatives or the option to select the status quo described as “no energy transition program.” The DCE section of the questionnaire comprised three parts: (1) an instructional guide on how to complete the choice tasks, (2) six individual DCE tasks, and (3) a follow-up question for each task asking respondents to indicate their level of certainty about their choice (Figure 2).

The survey questionnaire was pre-tested with a pilot sample (n = 60) to evaluate the clarity, realism, and relevance of the DCE task, as well as the appropriateness of the attributes and their corresponding levels [73,74]. Feedback from the pilot study informed minor revisions to the questionnaire wording and layout. The revised survey was deemed easy to understand and appropriate for the Kentucky population. The finalized survey was conducted in December 2023 via the Qualtrics online platform. The target sample included adult residents (18 years and older) of Kentucky from both coal mining and non-mining counties. Respondents were screened based on key eligibility criteria based on the socio-demographic characteristics defined in Table 2. A total of 960 survey invitations were distributed, yielding 675 fully completed responses, corresponding to a response rate of 70.4%. This is consistent with prior studies using modest incentives and web-based channels [37,53,54]. Data analysis was performed using Random Parameters Logit (RPL) modeling to account for preference heterogeneity across respondents. All statistical analyses were conducted in Stata 18.0 SE.

2.5. State Socio-Demographic Variables

The sampling approach was guided by key demographic characteristics, including age, household size, education level, gender, proportion of coal miners, urban/rural residence, county classification (mining vs. non-mining), and political party affiliation, in order to ensure that the sample was broadly representative of the target population and to minimize sampling bias. To assess the representativeness of the survey sample to the population and reduce the potential sampling bias, we used Pearson’s chi-square (χ²) independence test on the socio-demographic variables for balance with US Census 2021 data for Kentucky (Table 2). The results showed no statistical difference in most socio-demographic variables, with the exception of the percentage of coal miners in Kentucky and the area of residence. This was due to pre-stratification weighting procedures to meaningfully incorporate the perspectives of coal miners [14].

Table 2. Socio-demographic characteristics of survey respondents vs. the Census for Kentucky (2020) (US Census Data Bureau [75]).

Categories	Sample (n = 675)	Kentucky	Pearson χ2 Test
Median age (years)	42	39.2	***
Household size (people)	2.83	2.52	***
Education (bachelor’s degree or higher)	20.89%	19.1%	***
Female (%)	50.37%	50.58%	***
Employment rate (%)	62.2%	56.8%	***
Coal miners (%)	1.62%	0.078%	-
Area of residence (rural)	58.51%	41.38	-
Political party affiliation (Republican)	40.00%	42.89%	***
Population from mining county	26.22%	19.55%	***

The data represents significance at 1% according to the Pearson χ² test. *** indicates significance at 1% level, ** indicates significance at 5% level, and * indicates significance at 10%.

Table 2. Socio-demographic characteristics of survey respondents vs. the Census for Kentucky (2020) (US Census Data Bureau [75]).

Categories	Sample (n = 675)	Kentucky	Pearson χ2 Test
Median age (years)	42	39.2	***
Household size (people)	2.83	2.52	***
Education (bachelor’s degree or higher)	20.89%	19.1%	***
Female (%)	50.37%	50.58%	***
Employment rate (%)	62.2%	56.8%	***
Coal miners (%)	1.62%	0.078%	-
Area of residence (rural)	58.51%	41.38	-
Political party affiliation (Republican)	40.00%	42.89%	***
Population from mining county	26.22%	19.55%	***

The data represents significance at 1% according to the Pearson χ² test. *** indicates significance at 1% level, ** indicates significance at 5% level, and * indicates significance at 10%.

3. Results and Discussion

3.1. Parameter Estimation Results

The estimated coefficients from the RPL model are summarized in

Table 3. RPL parameter estimates for energy transition programs in Kentucky.

Attribute Levels		Random Parameter Logit (RPL)
		Mean	Std Dev
Possible Clean Energy Alternatives (PCEAs)	Wind	0.309 (0.092) ***	0.679 (0.142) ***
	Solar	0.856 (0.083) ***	0.641 (0.148) ***
Post-mining land use (PLU)	Recreation	−0.381 (0.131) **	−0.636 (0.176) ***
	Conservation	−0.242 (0.123)	−0.180 (0.375)
New Job Creation (NJC)	640 jobs/year	0.362 (0.083) ***	−0.025 (0.145)
	1280 jobs/year	0.771 (0.089) ***	0.644 (0.133) ***
Preserving cultural values (PCV)	Preservation	0.382 (0.057) ***	0.544 (0.102) ***
Social Support Programs (SSPs)	Job retraining	−0.213 (0.078) *	0.381 (0.199)
	Medical compensation	−0.132 (0.079)	0.145 (0.236)
Cost (USD/household/month)		−0.116 (0.012) ***	0.199 (0.017) ***
Alternative Specific Constant (ASC)		−4.633 (0.398) ***	4.216 (0.356) ***
Pseudo R2		0.1861
Loglikelihood		−3621.233
No. of observations		12,150
No of respondents		675

*** indicates significance at 1% level, ** indicates significance at 5% level, and * indicates significance at 10%. Note that the attribute levels (biomass, timber production, 320 jobs/year, no preservation, energy efficiency programs, 1 USD/household/month) were omitted/baseline variables.

. The utility function coefficients for the attribute levels had both expected and unexpected signs. Overall, the model demonstrated acceptable goodness-of-fit (pseudo-R² = 0.1861). The loglikelihood at convergence (−3621.233) indicates a reasonably good model fit, reflecting the model’s ability to explain observed choices. The RPL model was estimated using 400 Halton draws to simulate the integrals over the distributions of random parameters. Consistent with economic theory, the coefficient for the COST attribute was negative and statistically significant at the 1% level (Table 3), indicating that utility decreases as the price of sustainable energy program attributes increases. Furthermore, most of the coefficients for standard deviations were statistically significant, revealing considerable preference heterogeneity among respondents. This heterogeneity may stem from variations in perceptions, cultural values, and socio-demographic characteristics, which influence how individuals evaluate trade-offs among the attribute levels.

For the PCEA attribute, both solar and wind had positive and statistically significant coefficients at the 1% level for means and standard deviations. This finding suggests strong support for renewables as alternatives to coal among Kentucky residents. The ASC had a negative and statistically significant mean coefficient at the 1% level, reinforcing public preference for transitioning away from the current energy system. Interestingly, the standard deviation of the ASC was also positive and statistically significant at the 1% level, showing a considerable segment of the population supporting the status quo. This mixed result was somewhat unexpected given Kentucky’s historical and cultural attachment to coal. However, it may reflect a growing recognition of the role renewables can play in a just energy transition, despite persistent political rhetoric favoring coal [76,77]. Several factors may explain this shift in public perception: (1) declining coal employment and increased awareness of renewable-related job opportunities [78,79]; (2) federal incentives via the Inflation Reduction Act (IRA) [80]; (3) widespread health impacts of coal pollution [78]; and (4) technological advances making solar more affordable. Moreover, lived experiences with extreme weather events (e.g., Eastern Kentucky flooding) have made residents more receptive to community-based, resilient energy systems, even when “climate change” remains a polarizing term [81].

For PLU, the “restoration and conservation” level had a negative but statistically insignificant coefficient. The “restoration and recreation” level also had a negative coefficient, significant at the 5% level, with its standard deviation significant and negative at the 1% level, indicating preference heterogeneity. These results suggest that respondents prioritized energy alternatives, job creation, and cultural values over recreational or conservational land uses. While recreation and conservation provide long-term ecological and economic benefits, they may be perceived as less urgent in economically distressed areas. Renewable energy and cultural attributes offer more tangible benefits like job and health improvements [82]. Conservation initiatives are sometimes viewed as restrictive or top-down, especially when they clash with traditional practices like mining [83], reducing their appeal in energy transition contexts.

The NJC attributes showed a clear preference for increased employment; support increases as the number of jobs increases from 640 to 1280 per year, both significant at the 1% level. Wind and solar sectors promise new jobs, especially in rural areas needing economic revitalization [80]. Oluoch et al. [38], Maxim et al. [40], and Amar et al. [41] found that job creation increases individuals’ willingness to pay for renewables. The “preserving cultural values” level had a positive and statistically significant mean and standard deviation at the 1% level. Cultural identity plays a crucial role in public preferences. Studies in Europe and Australia [30,84,85] show strong WTP for preserving heritage, particularly traditional knowledge, and sacred sites. In the U.S., Lewin [16] and others note that Appalachian coal communities often feel alienated by federal climate policies and environmental groups. These findings underscore how cultural dimensions such as land stewardship, self-reliance, pride, and heritage are not just symbolic but foundational to support just energy transition.

For the SSP attribute, both “medical compensation” and “job retraining” levels had negative coefficients. However, only the latter was statistically significant at the 10% level. This shows a small preference for SSPs within this sample, despite their importance in literature. Trandafir et al. [86] found that support for energy policies was higher when SSPs were perceived to directly benefit the community. Dell et al. [87] also note that there is widespread support for degrowth policies, such as universal healthcare, work-time reductions, fossil fuel production caps, and advertising restrictions. However, SSPs can be viewed as bureaucratic or inaccessible [88]. Their benefits are delayed and less visible than jobs, and they may lack the symbolic resonance of cultural preservation.

3.2. Marginal WTP Measures by Political Affiliation

To highlight preference heterogeneity, we compared WTP by political affiliation (Figure 3). The most valued attributes across all political affiliations were PCEAs, NJC, and PCV, respectively, whereas PLU and SSPs were the least favored.

The WTP values for the attribute levels under PCEAs reveal that support for solar energy was consistent across political affiliations. Democrats showed the highest WTP at USD 8.29, followed closely by Independents/Others at USD 8.22, and Republicans at $8.08. This uniform support reflects solar’s appeal when framed around economic growth, local control, and public health rather than solely as a climate mitigation strategy [89,90]. In contrast, the WTP for wind energy was lower than for solar, with Republicans surprisingly reporting the highest WTP (USD 4.04), followed by Democrats (USD 3.52) and Independents/Others (USD 1.87). This trend may reflect the perceptions of wind as industry-friendly and non-ideological, less culturally stigmatized than solar, financially beneficial for rural landowners, and shaped by localized framing and messaging strategies. Furthermore, Kentucky’s political identity may respond differently to symbolic associations of solar energy, which are often associated with progressive environmentalism, whereas wind may be seen as more neutral or pragmatic [77,78,79].

The WTP values for the restoration and recreation level were negative across all political affiliations, indicating a general lack of public support for this land use option. However, variation exists between groups: Democrats showed a higher WTP exceeding that of Republicans by USD 1.32 and Independents/Others by USD 2.58. This aligns with national trends, with Democrats favoring environmental restoration, while Republicans prioritize fiscal restraint and local autonomy [91,92,93]. The WTP for NJC increased with job numbers, with Independents showing the highest support (USD 9.07), followed by Republicans (USD 7.45) and Democrats (USD 6.82). This reflects differing ideological frames: Republicans support job-centered transition strategies tied to economic development; Democrats emphasize environmental and justice-oriented goals; Independents focus on tangible, nonpartisan outcomes. For the PCV attribute, Independents showed the highest WTP (USD 4.98), followed by Republicans (USD 3.59) and Democrats (USD 3.20). Independents’ strong support likely reflects a desire to preserve local identity and community heritage without aligning with partisan narratives [16,94,95]. For the SSP attribute, the WTP for job retraining was −USD 2.42, with Republicans being the least supportive, falling USD 2.33 below Democrats and USD 2.80 below Independents, respectively. Independents may favor SSPs for their practical and local benefits, while Republicans may distrust federally administered programs, and Democrats support them within broader equity frameworks [96,97].

Overall, the WTP results from the PCEAs reveal broad, cross-partisan support for solar energy, driven by its economic and health-related framing. Wind energy received less support overall, but surprisingly, Republicans showed the highest WTP, suggesting that wind may be viewed as more pragmatic and less politically charged. Restoration and recreation programs were generally unpopular, though Democrats showed relatively higher support, reflecting national environmental attitudes. NJC had the highest WTP overall, especially among Independents, highlighting a shared priority for economic development. PCV was most valued by Independents, and SSPs were least supported by Republicans, likely due to distrust in federally run programs.

4. Conclusions

A successful energy transition in historically coal-dependent regions like Kentucky must go beyond technological adoption and market efficiency. It requires a comprehensive, justice-oriented approach that addresses political resistance, supports economic recovery, and respects community heritage. We apply a stated preferences approach to gather public preferences on sustainable energy transition programs in Kentucky. This study finds that Kentucky residents broadly support clean energy alternatives across all political affiliations, with solar energy receiving robust bipartisan backing (WTP: Democrats USD 8.29, Independents/Others USD 8.22, Republicans USD 8.08) and wind energy garnering notable support, particularly among Republicans (WTP: USD 4.04). Job creation emerged as the strongest driver of public support, especially among Independents/Others (WTP for 1280 jobs: USD 9.07), followed closely by cultural preservation (Independents WTP: USD 4.98). These results underscore the significance of both economic opportunity and symbolic attachment to local identity in shaping preferences for energy transition policies.

In contrast, SSPs and some PLU options, such as recreation and conservation, received negative WTP values (e.g., WTP for job retraining: −USD 2.42). These findings suggest that policies perceived as abstract, redistributive, or disconnected from immediate economic benefit are less compelling, particularly to conservative respondents. Despite persistent barriers such as cultural attachments to coal, misinformation, limited internet access in rural areas, and the absence of state-level RPS, public interest in renewables is growing. Initiatives like community solar projects and solar land-leasing for farmers offer pragmatic, income-generating pathways that complement broader policy efforts.

While this study provides valuable insights into public preferences for an energy transition program in Kentucky, several limitations warrant attention. First, although the additional monthly utility bill was used as a realistic and policy-relevant payment vehicle to estimate WTP, the design did not explicitly incorporate risk or uncertainty such as implementation delays, fluctuating energy costs, or policy failure into the choice tasks. These factors can influence how respondents perceive and value different programs. Although the RPL model captures some unobserved heterogeneity, including preferences potentially shaped by trust, risk aversion, or political skepticism, we acknowledge that this does not fully substitute for the explicit modeling of risk scenarios. Methodologically, the study acknowledges limitations common in choice experiments [98]. Respondents demonstrated stronger engagement with tangible concepts such as job creation or cultural identity, while abstract or unfamiliar attributes, like SSPs and PLU, generated weaker or inconsistent signals. These insights highlight the need for clearer framing and community education to improve salience and understanding. Kentucky’s energy transition experience offers a valuable model for similar coal-reliant states such as Wyoming, West Virginia, Indiana, Pennsylvania, and Illinois. Further research could benefit from assessing the interaction effects of other socio-demographic variables such as gender, age, and education to provide a more comprehensive understanding of heterogeneity in preferences. In addition, the integration of comparative modeling tools, such as computable general equilibrium (CGE) models, to explore region-specific policy effects and identify equitable pathways for renewable energy adoption will be critical for achieving sustainable energy transition in the region.

Overall, these findings underscore the importance of successful energy transition policies that not only address economic and environmental concerns but also cultural identity, political trust, and perceived fairness. Moreover, bipartisan messaging that frames clean energy around jobs, health, and resilience can help bridge ideological divides and enhance policy acceptance. Ultimately, this study reinforces that equitable energy transition cannot rely solely on market mechanisms. Achieving a sustainable energy transition requires policy reform, inclusive community engagement, and the recognition of local social, cultural, and economic contexts. Integrating these dimensions will be critical to building durable support and avoiding the replication of historical injustices in the clean energy future.