Community-based ecotourism is frequently presented as an alternative to state-centred conservation because it seeks to connect biodiversity protection with local livelihoods and participation. Yet participation is not equivalent to empowerment. A community may attend meetings, serve on committees, or receive a share of tourism revenue while remaining unable to influence the decisions that determine land access, tourism development, revenue allocation, or conservation rules [1]-[4].

Ghana provides a useful comparative setting because ecotourism and community-based conservation operate through different institutional configurations. Kakum is predominantly state-managed; Boabeng-Fiema relies heavily on customary institutions; Wechiau developed through an NGO-facilitated community trust; and Mole remains strongly state-dominated. The cases, therefore, provide variation in governance arrangements within a common national legal and policy environment [5]-[8].

The paper asks how different forms of power shape participation across the four sites, how governance arrangements affect the distribution and accountability of tourism benefits, and what institutional conditions are associated with stronger community influence.

The paper makes three contributions. First, it provides a reproducible account of how the reviewed literature was assembled and coded at the site level. Second, it separates evidence about governance practice from claims about formal legal rights and policy. Third, it refines the comparison of the four sites by treating traditional authority as a form of institutional power rather than automatically equating it with community-wide empowerment.

Luke’s three-dimensional framework distinguishes visible decision-making power, agenda-setting power, and ideological power [2]. Arnstein’s ladder treats participation as a continuum ranging from manipulation and tokenism to partnership, delegated power, and citizen control [1]. These frameworks are used as interpretive devices rather than as claims that every empirical case can be assigned to one uncontested category. Table 1 maps the rungs of Arnstein’s ladder onto the forms of participation observed across the Ghanaian sites.

Table 1. Participation ladder applied to Ghanaian ecotourism governance.

Note: Adapted from Arnstein [1].

Governance arrangements distribute authority among state agencies, traditional authorities, community organizations, NGOs, private operators, and other actors. Benefit sharing is therefore treated as both an economic and an institutional process. The relevant question is not only how much revenue is nominally allocated to local actors, but who controls collection, allocation, disclosure, timing, and accountability.

The Wildlife Resources Management Act, 2023 (Act 1115), establishes the Forestry Commission as the implementing agency and provides for Protected Area Management Advisory Committees with community, traditional-authority, district, private-sector, and women’s representation. It also requires consultation with local authorities and landowners before the establishment of a protected area. These provisions establish participation and consultation requirements but do not amount to a general statutory right of community control [9]. Table 2 summarises the principal stakeholder categories and the resource or power base each brings to site governance.

The paper proposes that governance outcomes are better explained by the interaction of decision rights, resource control, information, accountability, and internal social differentiation than by participation labels alone. Ecotourism itself is understood here, consistent with standard usage in the field, as tourism to natural areas that aims to conserve the environment and sustain the well-being of local people [10]. This framing also draws on institutional bricolage perspectives, which

Table 2. Principal stakeholders in Ghanaian ecotourism governance.

Note: The table directly supports the reviewer-requested power analysis.

emphasise that natural-resource governance arrangements are assembled from overlapping formal and customary institutions rather than designed from a single coherent blueprint, so authority in practice can diverge substantially from authority on paper [11].

The study uses a comparative secondary-data case-study design covering Kakum National Park, Boabeng-Fiema Monkey Sanctuary, Wechiau Hippo Sanctuary, and Mole National Park. The sites were purposively selected because they represent contrasting governance arrangements: state management with formal benefit-sharing at Kakum; customary/traditional management at Boabeng-Fiema; NGO-facilitated community governance at Wechiau; and highly state-dominated management at Mole. Table 3 summarises the key characteristics of the four sites.

The review covers 2000-2025 and uses the reported corpus of 327 peer-reviewed publications indexed in Scopus and Web of Science. The available research record does not preserve the original database export or complete database-specific search histories; consequently, this revision does not invent a retrospectively exact Boolean query. The reproducible conceptual search strategy combines ecotourism/community-based tourism/nature-based tourism with community participation/community engagement/governance/power/benefit sharing/empowerment/ sustainability and Ghana, supplemented by searches for Kakum, Boabeng-Fiema, Wechiau, and Mole.

Table 3. Site characteristics.

Note: Visitor, revenue, and revenue-share figures are author-reported estimates from the underlying literature and must be paired with the original source/year in the final submission. They are not harmonized cross-site survey statistics.

Eligibility required English-language scholarly publications within 2000-2025 that addressed Ghanaian ecotourism or closely related community-based natural-resource/tourism governance and contained evidence relevant to participation, governance, power, benefit sharing, livelihoods, or sustainability. The screening sequence consisted of identification, duplicate removal, title/abstract relevance screening, substantive assessment, and thematic synthesis. Because the original screening log is not preserved, exact stage-by-stage counts are not reported as independently verified.

Site-level coding extracted governance structure, decision authority, participation mechanisms, benefit mechanisms, accountability, employment, inclusion/exclusion, conflict, and documented outcomes. Numerical claims were retained only when an identifiable source was available; otherwise, they were converted to qualitative findings or explicitly labelled as unverified estimates.

An earlier draft of this paper listed bibliometric analysis as a fourth data source alongside the literature synthesis, policy documents, and site-level case evidence, but reported no bibliometric findings anywhere in the text. That inconsistency is corrected here. A separate bibliometric mapping exercise was conducted on a larger, independently assembled corpus of 498 Scopus- and Web of Science-indexed records (2000-2025) on community engagement, power dynamics, and benefit sharing in Ghanaian ecotourism, using Bibliometrix/R for performance analysis, VOSviewer for co-occurrence and collaboration mapping, and CitNetExplorer for citation-network analysis [12][13]. Because that corpus (498 records) is distinct from the 327-publication set synthesised qualitatively for the present site-level comparison, the two should not be conflated or treated as a single harmonised dataset.

Bibliometric analysis is therefore not listed as a data source for the site-level findings and comparative analysis reported in Sections 4 and 5 of this paper, since no bibliometric performance indicators, cluster maps, or network statistics are presented as evidence here. Readers seeking the quantitative mapping of the wider research field—publication trends, thematic clusters, leading institutions, and collaboration networks—should consult the companion bibliometric study, which reports those results in full. This paper instead relies on the qualitative, source-verified synthesis and site-level power coding described in Sections 3.1, 3.2, and 3.4.

Six analytical lenses were applied to the site evidence. Decision-making examined who participates in formal decisions, who holds binding authority, and whose preferences prevail. Agenda setting examined who defines priorities and which issues can be excluded. Resource control examined control over land, capital, revenue, information, and expertise. Discourse examined whose knowledge and framing of conservation and development are treated as authoritative. Network analysis examined access to donors, government, private operators, technical institutions, and markets. Outcome analysis examined who receives benefits and who bears conservation and tourism costs.

These lenses were translated into seven observable forms of power: coercive, economic, expert, legitimate, informational, network, and discursive power. Each form was coded qualitatively as low, medium, or high when evidence permitted. Low means limited documented control; medium means shared, conditional, or contested control; high means concentrated or decisive control. The coding is interpretive and evidence-based, not a statistically estimated scale. The full coding matrix, including the coding question and the low/medium/high criteria applied to each lens, is provided in Appendix A.

The study relies on secondary evidence. Underlying studies differ in sampling, dates, measures, and definitions. Site-level percentages, therefore, cannot be treated as directly comparable statistics unless derived from harmonized instruments. The comparative design also cannot fully eliminate ecological, regional, historical, and market differences among sites.

Kakum is characterized by strong state authority and formal participatory structures. Existing research documents community expectations and participation constraints and indicates that formal consultation does not necessarily translate into substantive decision-making influence [5][6].

The previous categorical statement that 40% of revenue is legally earmarked for communities is removed. Historical evidence shows that Kakum revenue-sharing formulas have changed. A 2009 parliamentary report described a then-applicable formula allocating 20% to traditional authorities and communities, 20% to district assemblies, 10% to the central government, and 50% to a maintenance fund, and also reported implementation difficulties [14]. The defensible conclusion is therefore that Kakum has operated with formal revenue-sharing arrangements whose effects depend on the applicable formula, implementation, timing, transparency, and community influence.

The former manuscript’s exact claims concerning universal six-to-nine-month disbursement delays, 42% elite capture, and 34% satisfaction are not presented as harmonized facts unless the underlying original empirical sources are supplied (see Appendix B for the verification status of each claim).

Boabeng-Fiema differs because customary institutions play a central role in conservation and tourism governance. Historical research documents traditional conservation associated with the sanctuary [15]. However, strong customary authority should not be equated with community-wide citizen control.

The revised classification is ‘customary/traditional authority with substantial local autonomy.’ Key decisions are concentrated among chiefs, elders, spiritual authorities, and related customary institutions. This can produce strong local ownership while simultaneously excluding women, youth, migrants, or less influential lineages.

Empirical fieldwork at the sanctuary documents that tourism revenue is divided between customary beneficiaries (including chiefs, landowners, and fetish priests) and other institutional stakeholders (including the District Assembly, the Traditional Council, and the Wildlife Division), rather than being retained entirely by the community as a single, undifferentiated body [16]. The more defensible conclusion is that a substantial share of benefits has historically been retained or allocated locally through customary and sanctuary institutions, with the exact distribution depending on the arrangement and period (see Appendix B).

Wechiau represents a hybrid governance model involving community institutions and NGO facilitation. The NGO contributed financial, technical, organizational, and external-network resources, while community structures were established to participate in management and benefit allocation.

The central issue is dependency. NGO support can strengthen local organizational capacity while also creating asymmetry because external actors may control funding, expertise, donor relationships, and agenda-setting. Wechiau is therefore coded as a medium community influence with shared or contested authority rather than fully delegated community power.

Mole represents the most state-dominated of the four cases in the original comparison. The Wildlife Division retains primary management authority, while surrounding communities face restrictions on access to park resources. Existing literature identifies tensions associated with livelihoods, conservation restrictions, human-wildlife interactions, and limited local benefits [7][8].

The previous exact claims of a 20% community revenue share, a 10% employment quota, and 18% benefit satisfaction are not retained as established facts unless their original empirical sources are verified. The robust finding is qualitative: Mole exhibits weak formal community influence and strong state control. A summary of the verification status of every site-level numerical claim discussed above is provided later in the paper (See Appendix B).

Building on the six analytical lenses introduced in Section 3.4, Table 4 compares five core forms of power side by side across the four sites, applying the coding scheme set out in Appendix A. The complete seven-form comparison, including network and discursive power, is provided in Appendix B (Table B1).

Table 4. Forms of power across sites.

The comparison shows that no governance model automatically produces inclusive empowerment. State systems may institutionalize consultation without transfer of authority; customary systems may create local autonomy while concentrating authority internally; NGO systems may strengthen participation while generating dependency; and highly centralized systems may leave communities with minimal influence. Table 5 extends this comparison to decision-making authority, revenue control, and inclusion outcomes.

Table 5. Comparative power analysis.

Note: Satisfaction and elite-capture values are study-specific reported estimates and should not be interpreted as directly comparable statistics without harmonized instruments. Boabeng-Fiema’s former “High (traditional)” empowerment label has been revised to recognize substantial local autonomy alongside an internal concentration of authority.

Participation quality depends on whether communities can influence agendas, control or co-control resources, obtain relevant information, and hold decision makers accountable. A meeting, committee seat, or revenue percentage is therefore insufficient evidence of empowerment by itself. Table 6 outlines the short- and long-term effects of this gap for each stakeholder domain.

Table 6. Effects of power imbalance by stakeholder domain.

Benefit sharing should be assessed as a governance chain: collection, recording, allocation, disbursement, disclosure, monitoring, and grievance procedures. This explains why nominal percentages alone may be poor indicators of equity. The proposition remains to be tested with harmonized financial and household-level data.

The term community conceals internal differences. Traditional authority can redistribute power away from the state without distributing it equally among all residents. Women, youth, migrants, poorer households, and less influential lineages may remain excluded.

The Wildlife Resources Management Act, 2023 (Act 1115), establishes the Forestry Commission as the implementing agency and the highest administrative authority over protected areas. It provides for Protected Area Management Advisory Committees and specifies representation from the Wildlife Division, traditional authority, District Assembly, community-based structures, private sector, and a women’s group. It also requires consultation with local authorities and landowners before the establishment of protected areas [9].

These provisions should not be paraphrased as a statutory grant of general community veto power or citizen control. The legal architecture is better described as participatory advisory governance within a system retaining central administrative authority [9].

The previous categorical statement that Ghana has no FPIC requirement is revised. Ghanaian law does not generally use FPIC as a universal protected-area requirement. However, consultation, land, compensation, and participation provisions may overlap with some FPIC principles. The appropriate claim is therefore that no general statutory FPIC requirement has been identified for all protected-area decisions, rather than asserting an absence of all consent-related protections. The Wildlife Resources Management Act expressly requires consultation with local authorities and landowners before the establishment of a protected area [9].

Revenue-sharing percentages are treated as site- and instrument-specific. Each numerical percentage retained in a final submission should be accompanied by the exact legal instrument, management agreement, administrative record, parliamentary statement, or empirical study from which it is drawn.

The revision of Boabeng-Fiema is theoretically important. Traditional authority can generate local legitimacy and conservation commitment without producing inclusive democratic control. Because authority is concentrated among chiefs and elders, describing the arrangement as citizen control overstates the evidence. The case is better understood as customary institutional autonomy with internal hierarchy.

The six analytical lenses show why visible committee structures are insufficient. Decision-making may be shared while agenda setting remains concentrated. Communities may possess local knowledge while external actors control technical expertise and donor networks. Revenue may be nominally shared while information about collection and allocation remains inaccessible.

Future evaluations should record the legal or contractual share; actual amounts received; timing of transfers; decision rights over allocation; public disclosure; beneficiary coverage; grievance procedures; and distribution within the community.

Policy reform should prioritize enforceable participation rights, transparent revenue information, representative local institutions, and mechanisms for women and youth to participate independently of elite gatekeepers. The current wildlife law already provides a basis for community representation and community-based wildlife governance; the challenge is translating representation into meaningful influence and accountability [9]. This challenge is consistent with a broader pattern in Ghanaian tourism policy, where national plans articulate ambitious community-development goals [17] but implementation reviews report persistent gaps between stated objectives and enforceable practice [18]; closing that gap is as much a matter of institutional accountability as of policy design. Table 7 sets out priority reform areas, proposed changes, and their expected impact.

Table 7. Policy reform priorities.

Complementary capacity-building needs for each stakeholder group are summarised in Table 8.

Finally, Table 9 summarises the verification status of every site-level numerical claim discussed throughout the paper, consolidating the evidence flags raised in Sections 4 and 6 together with the claim-by-claim revision record set out in Appendix B.

Table 8. Capacity-building priorities by group.

Note: The 40% women’s leadership indicator is a proposed success benchmark, not a reported current condition.

Table 9. Evidence-status matrix for site-level numerical claims.