The digital transformation of African public administrations is a major strategic challenge for improving local governance and delivering quality public services. According to the United Nations e-Government Survey ([31]), African countries have an average e-government development index of 0.39, significantly lower than the global average of 0.60. This administrative digital divide reflects not only infrastructural deficits but also difficulties in the appropriation and effective adoption of technologies by local public actors ([38]).

Local authorities, the administrative level closest to citizens, play a crucial role in delivering essential services (civil registration, local taxation, urban planning, and sanitation). Yet, their capacity to adopt and effectively utilize digital technologies remains limited ([20]). Considerable investments in municipal information systems, e-service platforms, and management tools remain underutilized or even abandoned, revealing a failure that is not technological but rather organizational and human ([16]; [24]; [30]).

The literature on technology adoption offers several proven theoretical models: Technology Davis’s Acceptance Model (TAM) (1989), Tornatzky and Fleischer’s Technology-Organization-Environment (TOE) framework (1990), and Venkatesh et al.’s Unified Theory of Acceptance and Use of Technology (UTAUT) (2003) have demonstrated robustness in Western contexts and the private sector ([37]; [13]). However, their direct application to African public administrations raises questions of external validity and contextual relevance ([36]; [25]).

Main research question: How to construct and empirically validate a hybrid model of technology adoption integrating the institutional, socio-cultural and economic specificities of Cameroonian local authorities?

This research pursues three complementary objectives:

1. Theoretical objective: To identify the limitations of classical models (TAM, TOE, UTAUT) when applied to Cameroonian public contexts and to propose an integrative framework enriched with relevant contextual factors.

2. Empirical objective: To quantitatively test and validate the proposed hybrid model with a representative sample of local authorities in three pilot local authorities (Yaoundé Urban Community, Yaoundé 3 District Municipality and Mbankomo Rural Municipality).

3. Managerial objective: To provide public decision-makers and support organizations with a diagnostic framework to identify priority levers for action to promote technology adoption.

This research makes a threefold contribution:

Theoretical contribution: Development of a contextualized theory of technology adoption in the Cameroonian public sector, enriching the corpus of ICT4D and public information systems.

Methodological contribution: Validation of a suitable measurement instrument and demonstration of the explanatory superiority of a hybrid approach over mono-theoretical models.

Practical contribution: Provision of an operational diagnostic tool to guide local digital transformation public policies.

2.1.1. The Technology Acceptance Model (TAM)

The Technology Adoption Model (TAM), developed by [11] and extended by [33], is the dominant model of technology adoption at the individual level. Grounded in Fishbein and Ajzen’s theory of reasoned action, it posits that the behavioral intention to use a technology is primarily determined by two beliefs:

Perceived usefulness: The degree to which a person believes that using a system will improve their job performance. Perceived ease of use: the degree to which a person believes that using a system will be effortless.

The TAM has been validated in multiple contexts and technologies, demonstrating remarkable robustness ([18]; [19]). However, its limitations are documented: exclusive focus on individual perceptions, neglect of organizational and environmental factors, and reduced applicability in mandatory usage contexts such as the public sector ([5]; [8]; [27]) (Table 1).

Table 1. The TAM (Technology acceptance model).

2.1.2. The Technology-Organization-Environment (TOE) Framework

The TOE framework of [29] adopts an organizational perspective by identifying three categories of factors influencing technology adoption by firms:

Technological context: characteristics of the technology (compatibility, complexity, relative advantage).

Organizational context: organizational resources (size, financial slack, managerial support, skills).

Environmental context: industry characteristics, competitive pressures, regulatory framework.

The TOE has the advantage of considering the organization as the unit of analysis and integrating external contextual variables. It has been widely used to study technology adoption in companies ([6]; [4]). However, it neglects the individual and behavioral dimension of adoption, limiting its ability to explain actual use beyond the formal organizational decision ([22]; [2]) (Table 2).

Table 2. The TOE (Technology-organization-environment) framework.

2.1.3. The Unified Theory of Acceptance and Use of Technology (UTAUT)

The UTAUT model, developed by [34] and later extended into UTAUT2 ([35]), synthesizes eight competing models of technology acceptance. It identifies four direct determinants of behavioral intention and use.

Table 3.The UTAUT model (Unified theory of acceptance and use of technology).

Performance expectancy: the degree to which the use of a technology will bring benefits in the accomplishment of certain activities.

Effort expectancy: the degree of ease associated with using the technology.

Social influence: the degree to which an individual perceives that other important people believe they should use the new technology.

Facilitating conditions: the degree to which an individual believes that the organizational and technical infrastructure exists to support the use of the system.

The UTAUT model incorporates moderating variables (age, gender, experience, voluntary use) and has demonstrated superior explanatory power (R² = 70%) compared to previous models ([37]; [28]). However, its development within Western organizational contexts limits its external validity in African public administrations ([13]) (Table 3).

Despite their empirical robustness, the TAM, TOE, and UTAUT models have limitations when applied to African local authorities. The literature identifies five main shortcomings (Table 4).

Table 4. Limitations of classical models in an African context.

The literature review on ICT4D and e-government in Africa identifies several critical contextual factors absent from classical models:

Institutional capacity: The ICT4D literature highlights that the weak institutional capacity of African administrations (limited skills, informal processes, constrained financial resources) constitutes the main bottleneck to technology adoption ([15]; [3]; [32]). This dimension, absent from classical models, nevertheless appears crucial in contexts of low administrative maturity.

Transformational leadership: Research on e-government in Africa identifies political and administrative leadership as a critical success factor ([38]; [21]). Transformational leadership capable of articulating a clear digital vision and mobilizing stakeholders is essential but is not addressed by traditional models.

Institutional pressures: Neo-institutional theory ([12]) suggests that organizations adopt practices through institutional mimicry or under normative pressure. In the African context, donor pressure, imitation of other communities, or government directives strongly influence adoption ([20]).

Cultural factors: The work of [17] and [26] demonstrates that national cultural values (power distance, collectivism, uncertainty aversion) moderate technology adoption. African societies, characterized by high power distance and pronounced collectivism, exhibit specific adoption dynamics (Table 5).

Table 5. Contextual factors specific to African communities.

Faced with the limitations of single-theoretical models, we propose a hybrid model of technology adoption for African local authorities. This model integrates:

Individual variables of the TAM/UTAUT: perceived usefulness, ease of use, social influence.

The organizational variables of TOE: managerial support, resources, skills.

Specific contextual factors: institutional capacity, transformational leadership, institutional pressures, cultural factors (Table 6).

Alignment of the conceptual model:

The final conceptual model selected for the empirical analysis includes the following variables:

Perceived usefulnessEase of useSocial influenceEnabling conditionsInstitutional capacityTransformational LeadershipInstitutional pressures

Table 6. Proposed hybrid model of technology adoption.

The variables initially explored in the literature review (cultural factors, managerial support, organizational resources, technical skills, performance expectancy) were excluded from the final model after exploratory analysis and statistical validation, due to their weak empirical contribution or conceptual redundancies.

Table 7. Assumptions of the hybrid model.

Addition of institutional pressures:

The variable “institutional pressures” was measured using three items inspired by neo-institutional theory ([12]), adapted to the Cameroonian context:

Pressure from central authoritiesImitating other communitiesInfluence of technical and financial partners

The model postulates many research hypotheses structured into three levels (Table 7).

This research adopts a positivist stance and a hypothetico-deductive approach aimed at empirically testing the proposed theoretical model ([10]). Quantitative methodology using questionnaire surveys is particularly well-suited for:

Testing causal relationships between latent variables.

Generalize the results to a larger population.

Compare the explanatory power of competing models.

Psychometrically validate measurement instruments.

Partial least squares structural equation modeling (PLS-SEM) is retained as the main analysis technique, in accordance with methodological recommendations for complex models with formative and reflective constructs ([14]; [23]).

The survey targeted local agents and elected officials from the three pilot decentralized territorial authorities and 07 directorates of the Ministry of Decentralization and Local Development to represent the diversity of French-speaking, English-speaking and digital maturity levels.

The sample for this study is based on a reasoned (non-probabilistic) sampling strategy aimed at capturing the diversity of institutional contexts across Cameroonian decentralized territorial communities.

The three selected local authorities: Yaounde Urban Community (CUY), Yaounde 3 District Municipality (CY3), and Mbankomo Rural Municipality (CRM) were chosen according to three criteria:

Their level of digital maturity (high, intermediate, low)..Their territorial typology (metropolitan urban, intermediate urban, rural).Their accessibility and institutional availability for data collection.

The seven directorates of the ministry in charge of decentralisation were selected because of their direct role in defining, coordinating, and implementing public policies for the digital transformation of local and regional authorities.

The other subgroup (107 respondents) includes technical agents, administrative managers, IT managers and operational staff involved in digital processes.

This sample, while not probabilistic, captures a diversity of profiles and institutional situations, thus supporting the analytical validity of the results for Cameroonian CTDs. The inclusion of ministry officials reflects an analytical rationale aimed at integrating a multi-level perspective on digital governance.

Indeed, ministerial actors play a central role in:

The definition of digital public policies.Resource allocation.Technical support for local authorities.

An exploratory comparative analysis was conducted to compare the perceptions of local stakeholders (CTD) with those of central stakeholders (ministry). However, the conclusions of this study are interpreted with caution and considered representative of a mixed administrative sample (Table 8).

Table 8. Sample characteristics.

The questionnaire, developed using validated scales and adapted to the Cameroonian and African context, comprised 68 items measured on 7-point Likert scales. English-French translation and back-translation were performed. A pre-test with 35 respondents allowed for refinement of the question wording.

The constructs of the model are measured using scales validated in the literature, adapted to the context of African local authorities (Table 9).

Table 9. Operationalization of constructs.

The evaluation of the measurement model follows the recommendations of [14] for PLS-SEM. All reflective constructs demonstrate satisfactory reliability and validity (Table 10).

Table 10. Quality of the measurement model.

Extension of PLS-SEM reporting: The measurement model evaluation was supplemented by:

Analysis of item loading times (>0.70).Discriminant validity (Fornell-Larcker criterion and HTMT ratio < 0.85).Collinearity verification (VIF < 5).

All constructs in the model were specified as reflective, in accordance with the recommendations of [14].

The Fornell-Larcker test indicates that the square root of the AVE of each construct is greater than the inter-construct correlations, confirming discriminant validity.

The HTMT ratio is below the critical threshold of 0.85 for all pairs of variables.

Analysis of the structural model reveals excellent explanatory power:

R² (Adoption Intention) = 0.68 (substantial according to [9]).

R² (Actual Use) = 0.62 (substantial).

Q² (Stone-Geisser) > 0 for all endogenous constructs, confirming predictive relevance (Table 11).

Table 11. Hypothesis test results.

*** means highest value where by ** is high.

To demonstrate the superiority of the hybrid model, we tested the TAM, TOE and UTAUT models in isolation on the same sample (Table 12).

Table 12. Comparison of the explanatory power of the models.

Control of Common Method Bias:

A single-factor Harman test was performed to assess the presence of a common method bias. The results show that no single factor explains the majority of the total variance, suggesting that this bias is not critical.

However, it should be noted that the variables “adoption intention” and “actual use” were measured using the same cross-sectional questionnaire, which could potentially introduce bias. This limitation should be taken into account when interpreting the results.

The results reveal three major theoretical lessons:

1. The primacy of institutional capacity: With a coefficient of 0.42 (p < 0.001), institutional capacity emerges as the primary determinant of technology adoption in African communities. This result confirms the postulates of ICT4D ([15]; [3]) regarding the importance of aligning technological solutions with pre-existing organizational capacities. It suggests that in contexts of low administrative maturity, strengthening institutional capacities is a prerequisite for any digital transformation initiative.

2. The amplifying role of social influence: Social influence has a significantly stronger effect (β = 0.38) than in Western UTAUT studies (β ≈ 0.15 - 0.25). This difference corroborates the work of [17] and [26] on the importance of collectivism and social conformity in African cultures. Technology adoption decisions are not solely rational and individual, but are strongly influenced by peers, hierarchy, and social norms.

3. The catalytic effect of transformational leadership: Transformational leadership exerts a significant direct effect (β = 0.31) and an indirect effect via institutional capacity. This result confirms the critical importance identified by [38] and [21] of political and administrative leadership capable of articulating a clear digital vision, mobilizing resources, and transforming cultural resistance.

The validated model offers public decision-makers a structured diagnostic framework to identify priority levers for action (Table 13).

Table 13. Strategic recommendations are determined.