Effect of information and communication technologies on agricultural credit in sub-Saharan Africa Open Access

Agriculture remains a cornerstone of economic development in Sub-Saharan Africa and serves as the main source of income for rural populations. However, this strategic sector continues to face major challenges in accessing financial services. According to Brulé-François et al. (2016), less than 10% of the rural population relies on borrowing, and only 1% of bank credit is allocated to agriculture in Africa. The World Bank (2020) confirms that fewer than 10% of smallholder farmers in the region have access to formal financial services. This persistent financial exclusion stems from structural constraints such as the lack of formal collateral, high transaction costs in rural areas, information asymmetry between lenders and borrowers and the perception of high risk associated with the agricultural sector by financial institutions (IFPRI, 2018; Ghosh, 2013). The low level of financial inclusion in rural areas, estimated at 5–6% in several African countries (Brulé et al., 2016), further exacerbates the problem.

In this context, the adoption of information and communication technologies (ICTs) in financial systems offers a promising avenue to enhance financial inclusion, particularly in the agricultural sector. Through mobile money, Internet access and mobile phones, ICTs can help reduce barriers to credit access. Jack and Suri (2014), using longitudinal panel data for Kenya, demonstrate that M-Pesa has significantly fostered financial inclusion and strengthened the resilience of rural households. Similarly, studies employing cross-sectional and panel regression methods (Aker and Mbiti, 2010; Batista and Vicente, 2020; Lwanga and Adong, 2021; Deichmann et al., 2016; Asuming and Frempong, 2023; Osabohien, 2024; Ayimah et al., 2024) have emphasized that the expansion of digital financial services can substantially improve access to credit, particularly in areas underserved by traditional banks. Mobile phones, through Unstructured Supplementary Service Data (USSD) technologies and internet-based applications, enable farmers to conduct transactions, save, receive payments and access microcredit without relying on conventional banking systems (Mbiti and Weil, 2013). Initiatives such as M-Shwari in Kenya, Esoko in Ghana and MoBis in Uganda exemplify the transformative impact of ICTs on agricultural finance (Esoko, 2023; M-Shwari, 2023; MoBis, 2023).

Despite these advances, the literature reveals significant gaps. Most studies focus on national contexts or case-based analyses, which limits the generalizability of findings at the regional level. Few quantitative studies have directly investigated the relationship between different dimensions of ICTs (mobile money, Internet, mobile phones) and access to agricultural credit within a multi-country Sub-Saharan African framework (Demirgüç-Kunt et al., 2021; FAO, 2021; GSMA, 2022; World Bank, 2022). Moreover, existing works highlight the need to incorporate institutional quality, digital infrastructure and socio-economic conditions into the analysis, factors often overlooked in earlier research (Asuming and Frempong, 2023; Blimpo and Cosgrove-Davies, 2019). This situation underscores a paradox: while financial digitalization is rapidly expanding, farmers, particularly smallholders in rural areas, continue to face persistent barriers to financing (Asuming and Frempong, 2023; Lwanga and Adong, 2021). This raises a central question: to what extent can ICTs genuinely reduce structural obstacles and improve access to agricultural credit for farmers in Sub-Saharan Africa? The issue extends beyond mere technological availability. It involves a complex interplay of institutional, infrastructural and social factors that determine the effectiveness of ICTs in transforming rural finance (Blimpo and Cosgrove-Davies, 2019; Deichmann, 2019; Batista and Vicente, 2020). Indeed, weak regulatory frameworks, inadequate digital infrastructure and socio-economic inequalities often constrain the reach of digital financial innovations, undermining their capacity to foster inclusive and sustainable agricultural financing (GSMA, 2023; World Bank, 2023).

The objective of this study is to empirically examine the impact of ICTs on access to agricultural credit in Sub-Saharan Africa over the period 2015–2023. Specifically, the research seeks to: (1) identify the key ICT variables (mobile money, Internet, mobile phones) influencing agricultural credit; (2) assess the short- and long-run effects of ICTs on agricultural credit using a Vector Error Correction Model (VECM) suitable for non-stationary and cointegrated time series and (3) analyze the role of institutional and socio-economic conditions in the relationship between ICTs and agricultural credit.

By addressing these gaps, this study contributes to a deeper understanding of how the digital revolution can promote inclusive agricultural finance and support rural development in Sub-Saharan Africa. It extends the scope of existing research by adopting a multi-country approach and integrating institutional and socio-economic dimensions that are often neglected in previous studies.

The remainder of the paper is structured as follows: Section 2 presents the theoretical underpinnings and literature review. Section 3 outlines the methodology. Section 4 reports and discusses the empirical findings. Section 5 provides robustness checks and sensitivity analyses. Section 6 discusses the empirical results. Section 7 concludes with key policy implications and future research directions.

Studies examining the impact of ICTs on agricultural credit in Sub-Saharan Africa are grounded on interconnected theoretical frameworks: information asymmetry, transaction costs, innovation diffusion and financial inclusion. These theories collectively explain how ICTs transform the structure, costs and performance of rural finance systems.

The information asymmetry theory highlights that lenders often have limited reliable information regarding farmers’ creditworthiness (Tchamyou and Asongu, 2017). ICTs, through mobile money transaction histories, satellite data and digital agricultural information services, help to reduce such asymmetries and facilitate borrower assessment (Asuming and Frempong, 2023; Lwanga and Adong, 2021). In the Sub-Saharan context, these innovations enable a shift from community reputation-based credit systems toward data-driven financing mechanisms. The transaction cost theory emphasizes the inefficiencies associated with information search and loan management, which are particularly high in rural areas – estimated to be 10–20 times higher than in urban zones (Blimpo and Cosgrove-Davies, 2019). Digital tools and mobile platforms reduce these costs by automating payment processes and data collection (Batista and Vicente, 2020; GSMA, 2023; Asongu and Odhiambo, 2024), thereby enhancing credit market efficiency. From an African perspective, where geographic dispersion and weak infrastructure amplify intermediation costs, ICTs act as an institutional substitute compensating for the absence of physical banking branches. The innovation diffusion theory (Rogers, 2003) explains how new technologies are adopted within societies. The adoption of ICTs in rural areas depends on structural factors such as infrastructure availability, sociocultural norms, education levels and institutional support (Deichmann, 2019; FAO, 2021).

In Sub-Saharan Africa, diffusion often follows a “peer network” logic, where imitation, community trust and public–private partnerships are decisive. Digital innovations become embedded in existing social dynamics, gradually transforming financial practices and credit behaviors. The financial inclusion theory underscores the importance of access to appropriate and affordable financial services to reduce poverty and stimulate growth (Demirgüç-Kunt and Klapper, 2021; World Bank, 2022). In the Sub-Saharan context, “appropriate” refers to financial services tailored to agricultural needs, while “affordable” denotes costs proportionate to rural incomes. ICTs facilitate such inclusion by enabling flexible lending mechanisms, deferred payments and climate-indexed insurance products, thus making agricultural finance more resilient and accessible.

These theoretical perspectives are complementary in explaining the catalytic role of ICTs in agricultural finance. Information asymmetry highlights credit constraints, transaction cost theory explains reductions in operational inefficiencies, innovation diffusion justifies the pace of technological adoption and financial inclusion theory elucidates the long-term socio-economic impacts. Similar to the role of technological progress in the Harrod–Domar (1939) and Solow (1956) growth models, ICTs emerge here as an exogenous factor enhancing the efficiency of agricultural credit systems. However, their full potential depends on the interaction between technological innovation, institutional capacity and social acceptance.

In sum, the theoretical framework of this study adopts an integrated perspective in which ICTs are not merely tools for reducing asymmetries or costs, but function as a structural mechanism for institutional transformation within agricultural finance in Sub-Saharan Africa.

The empirical literature highlights the growing importance of ICTs in improving access to agricultural credit in Sub-Saharan Africa, where distance, lack of collateral and high transaction costs have historically constrained rural financing.

Among the pioneering works, Binswanger and Khandker (1995), using stochastic frontier analysis (SFA) models, concluded that the efficiency of agricultural finance depends on reducing intermediation costs and improving the availability of reliable information. In the same line, Zeller and Meyer (2002), employing logit and probit models, demonstrated that access to credit is determined by informational proximity between lenders and borrowers, thereby paving the way for the integration of ICTs to correct these asymmetries. Furthermore, Aker and Mbiti (2010), through a static panel model, emphasized that mobile phones usage reduces information costs and enhances the performance of African agricultural markets. Jack and Suri (2014), using a quasi-experimental approach on the Kenyan case of M-Pesa, found that mobile money strengthens financial resilience and increases access to credit through the digital traceability of transactions.

More recent studies corroborate these findings. Batista and Vicente (2020), in Mozambique, through a randomized controlled trial (RCT), showed that mobile services reduce transaction costs and promote financial inclusion. In Uganda, Lwanga and Adong (2021), employing a dynamic Generalized Method of Moments (GMM) model, found that digital payment platforms expand agricultural credit by providing banks with actionable data on small-scale producers. Similarly, Asuming and Frempong (2023), using a VECM, highlighted a positive cointegration relationship between mobile money development and credit supply across several West African countries. At the same time, the expansion of rural Internet connectivity promotes transparency, competition among financial institutions and the diffusion of innovations such as agricultural fintechs. Platforms like M-Shwari (Kenya), Esoko (Ghana) and MoBis (Uganda) enable banks, fintechs and microfinance institutions to offer credit services tailored to the agricultural sector, thereby mitigating geographic and informational constraints. However, Kaufmann et al. (2010), using static panel models, emphasized that institutional quality, regulation, political stability and governance, conditions trust and the sustainability of digital innovations. Likewise, reports by CGAP (2021), World Bank (2021) and GSMA (2022) stress the necessity of an enabling regulatory framework to ensure data security and the long-term viability of digital financial services. Other recent studies have focused on the importance of electronic wallets in facilitating agricultural finance at the rural level in countries such as Nigeria (Uduji et al., 2019).

In summary, theoretical analyses converge on the positive effects of ICTs through the reduction of information asymmetries and transaction costs, the acceleration of technological adoption and the stimulation of financial inclusion, analogous to the role of technological progress in growth models. Nevertheless, they underscore the critical role of the institutional environment. Empirical studies confirm these positive effects but remain focused on a limited number of countries and static models. This study broadens the scope by covering 30 Sub-Saharan African countries over the period 2015–2023, applying a VEC model combined with a GMM framework to examine the short- and long-term effects of ICTs and institutional quality on agricultural credit.

The method of analysis is based on VEC (Vector Error Correction) to determine the short- and long-term links between ICT and agricultural credit. Building on the studies of Jack and Suri (2014), Aker and Mbiti (2010) and Ghosh (2013), the model considers that ICT stimulates financial inclusion in rural areas. The use of the VEC model assumes: (1) a long-term relationship between ICT and agricultural credit in sub-Saharan Africa; (2) a positive short-term effect of ICT on agricultural credit and (3) the existence of an error correction process towards long-term equilibrium. Unlike a traditional Vector Autoregressive (VAR) model, it allows for distinguishing between transitory and structural effects. The model is presented as follows:

With $ΔYit:$ vector of endogenous variables in first difference (short-term effects); level of lagged variables containing the $ΠYit−1$ cointegration relations; $Γk$⁠: matrices of short-term coefficients; $μi and λt$ are respectively the country i effects and the time effects and $ϵit$⁠: vector of errors.

The functional form of the VEC model in the context of this study is:

Where: $Δ$⁠, represents the variation from one period to another (short term); $it$⁠, country i at date t; $lncredit_agriit$⁠, logarithm of agricultural credit (% of total credit); $ln⁡mmoneyit$⁠, logarithm of mobile money account per adult; $lninternetit$⁠, logarithm of connectivity and information (% population); $lnmobileit$⁠, logarithm of mobile subscriptions per 100 inhabitants; $lnpib_hbtit$⁠, logarithm of GDP per capita (constant USD); $stab_soit$⁠, political stability index; $reguit$ regulatory quality index; $ln⁡pop_ruit$⁠, logarithm of rural population (% of total population). The parameters $αii=1 to 7$⁠, are coefficients that reflect the short-term relationship in our model and the presence of coefficients $βii=1 to 8$ indicates the long-term dynamics. The coefficient $β1$ is the error correction coefficient and $εit$ the error term. Table 2 presents the descriptive statistics of these variables.

This study uses secondary data from 30 Sub-Saharan African countries (see  Appendix 1) over the period 2015–2023. Data were obtained from the Food and Agricultural Organization (FAO) for agricultural credit, the World Bank’s World Development Indicators (WDI) for macroeconomic and technological variables and the Worldwide Governance Indicators (WGI) for institutional measures. The selection of variables follows prior empirical work on ICTs, finance and agricultural development in Africa (Aker and Mbiti, 2010; Asuming and Frempong, 2023; Deichmann et al., 2016; World Bank, 2023).

Agricultural credit (cred_agri) represents the share of total domestic credit allocated to agriculture (Binswanger and Khandker, 1995; IFPRI, 2018). ICTs indicators include mobile subscriptions (mobile), Internet usage (Internet) and mobile money transactions (mmoney), reflecting technological diffusion and digital financial inclusion (Jack and Suri, 2014; GSMA, 2023). GDP per capita (pib_hbt) captures the level of economic development (Solow, 1956), while rural population (pop_ru) represents demographic structure.

Institutional quality is measured by regulatory quality (regu) and political stability (stab_so), both drawn from the WGI (Kaufmann et al., 2010). These indices are transformed using the Inverse Hyperbolic Sine (IHS) function to accommodate zero or negative values and stabilize the data distribution (Cheng et al., 2015). All continuous variables are expressed in natural logarithms to linearize relationships and reduce heteroscedasticity. Table 1 summarizes the variables under study.

Table 1 shows that the average value of agricultural credit (credit_agri) is 10.64% of total credit, with a standard deviation of 5%. This indicates a moderate dispersion of the series, as the mean exceeds the standard deviation. The distribution is slightly positively skewed (0.13) with a kurtosis of 1.82, suggesting a concentration of values around the mean. The Jarque–Bera test rejects the null hypothesis of normality (p-value = 0.00026). The average level of Internet usage is 45.83% of the population, displaying high variability (standard deviation = 23.96%), slight negative skewness (−0.096) and a platykurtic distribution (kurtosis = 1.717), indicating non-normality. Mobile subscriptions (mobile) average 89.33 subscriptions per 100 inhabitants, with a slightly positive skewness (0.457) and substantial dispersion (standard deviation = 32.95). The series also fails the normality test (p-value = 0.00678). Mobile money (mmoney) represents on average 36.33% of GDP, with mild positive skewness (0.085), moderate dispersion (standard deviation = 19.17%) and a flat distribution (kurtosis = 1.78). The normality hypothesis is again rejected (p-value = 0.0002). Overall, macroeconomic variables (such as GDP and rural population) exhibit extreme values and highly asymmetric distributions, whereas ICTs and institutional variables appear more centered and less dispersed. The generalized non-normality of the data justifies the application of logarithmic and IHS transformations in the econometric analysis.

The correlation matrix for correlation coefficients which is available upon request, reveals weak bivariate relationships between agricultural credit and ICTs variables (mobile, mobile money and Internet), ranging from −0.0355 to 0.101, suggesting dynamic rather than contemporaneous effects. The positive correlations between mobile money and Internet with agricultural credit (approximately 10.2 and 0.75%, respectively) support the hypothesis of digital financial inclusion. Conversely, the negative correlation between the rural population and mobile penetration (−9.4%) highlights the digital divide between urban and rural areas. The modest correlations with institutional variables (0.03 and −0.06) further indicate long-term structural effects. Overall, the low level of multicollinearity ensures the empirical validity of the model and reflects the gradual influence of ICTs on agricultural credit.

The unit root tests which is available upon request indicate that the variables are integrated of different orders, specifically I(0), I(1) and I(2), with mobile and GDP per capita being integrated of order one. This combination supports the use of the VECM, which is suitable for capturing short-run dynamics while preserving long-run equilibrium relationships (Li and Bauer, 2020). The rural population variable was excluded because it is integrated of order two [I(2)], which violates the assumptions of the VECM framework. Finally, the Johansen cointegration test was conducted to verify the existence of long-run relationships among the model variables (Table 2). The hypotheses tested are: H₀ of at most r cointegration relationships between variables versus Ha of more than r cointegration relationships.

The results of the Johansen cointegration test applied to the model variables. Both the trace and maximum eigenvalue statistics exceed their respective 5% critical values, confirming the presence of a long-run cointegrating relationship. Seven long-term relationships are thus identified among agricultural credit, ICTs, GDP per capita and institutional variables. These findings justify the use of the VECM. Additionally, the Generalized Method of Moments (GMM) approach was employed as a robustness check to account for potential endogeneity and panel dynamics.

The VEC model estimations in Table 3 presents worthwhile results on the effect of ICTs and institutional variables on agricultural credit in Sub-Saharan Africa in both the long and short terms.

The results of the VEC model reveal statistically and economically consistent elasticities, with a satisfactory overall fit (adjusted R² = 0.35; F = 7.36). The estimates indicate that mobile phones usage exerts a positive and statistically significant effect at the 5% level on agricultural credit in the long run. A 1% increase in the mobile subscription rate leads to a 0.83% rise in agricultural credit. Similarly, the use of mobile money has a positive and highly significant effect at the 1% level, as a 1% increase in mobile money services boosts access to agricultural credit by 2.23%. Although the coefficient of Internet usage (0.075; t = 0.28) is positive, it remains statistically insignificant. Likewise, GDP per capita shows a negative but insignificant effect (−0.279; t = −1.43). The coefficients associated with regulatory quality and political stability are −0.863 (t = −4.49) and −1.000 (t = −5.42), respectively, both negative and significant at the 1% level. These counterintuitive signs are more indicative of an ongoing phase of institutional adjustment rather than a persistently adverse effect of governance on agricultural credit.

The adjustment equation of the VECM highlights the short-term dynamics between ICTs and agricultural credit. The error correction term is negative and statistically significant (−0.098; t = −2.49), indicating that approximately 9.8% of disequilibrium is corrected each period (year) toward the long-run equilibrium relationship. Although this speed of adjustment appears moderate, it remains economically plausible in the rural and agricultural contexts of Africa, particularly Sub-Saharan Africa, where the diffusion of financial innovations tends to be gradual and progressive. The coefficient associated with mobile money (Dlnmmoney (−1) = 0.193; t = 2.53) is positive and significant at the 5% level, suggesting that a 1% increase in mobile money usage immediately raises agricultural credit by about 0.19%. Similarly, the coefficient for Internet use (Dlninternet(–2) = 0.109; t = 2.04) is positive and significant at the 5% level, indicating that the impact of Internet diffusion on agricultural credit occurs with a slight time lag. Regarding political stability (Dihsstab), the first two lagged coefficients (−0.082; t = −1.97) and (−0.066; t = −1.81) are both negative and statistically significant, whereas regulatory quality (Dihsregu) shows no significant effect. Overall, ICTs, particularly mobile money (+2.23% in the long run; +0.19% in the short run) and mobile phones use (+0.83% in the long run), emerge as potential drivers of agricultural financing. The negative institutional effects observed do not contradict theoretical expectations; rather, they reflect an ongoing phase of institutional and financial adjustment. The relatively slow speed of convergence (9.8%) is econometrically plausible in settings where rural credit channels adapt progressively to digitalization.

The diagnostic test estimates (i.e. available upon request) confirm the overall robustness of the VECM model. The heteroskedasticity test (p = 0.1127) indicates that the residuals are homoscedastic, meaning the variance remains constant. The LM test for serial correlation shows no significant autocorrelation beyond the second lag, thereby validating the model’s dynamic specification (Kilian and Lütkepohl, 2017). Furthermore, according to the autoregressive root diagram (AR root diagram, which is available upon request), three roots are equal to one, while all others lie within the unit circle. Therefore, the VECM model satisfies the stability conditions.

In the Jarque-Bera tests which are available upon request, the results of this residual normality table do not accept the normality hypothesis (p-value < 0.01). This implies several reasons. One of them is related to the notable shocks on the technology-related variables (mobile, Internet, mobile money) according to the estimation results. The other plausible reason may be related to the rapid deployment of technologies (3 G/4G), the advent of the Covid-19 health crisis, the multi-SIM phenomenon and the sociopolitical instabilities observed in some sub-Saharan countries (Demirgüç-Kunt et al., 2021). This structural context can substantially modify the distribution of residuals and also bias the estimates if they are not taken into account (Kilian and Lütkepohl, 2017). In this perspective, robustness of the results can be ensured by jointly analyzing impulse response functions and variance decomposition (VD).

In Table 4, the analysis of the dynamics between ICTs and agricultural credit in Sub-Saharan Africa is based on the forecast error variance decomposition (FEVD) derived from the VECM model. This approach allows for the assessment of the proportion of future agricultural credit variability explained by its own shocks and by those of the ICTs variables. It highlights the relative contribution of mobile phones, mobile money and Internet usage to the volatility of agricultural credit, reflecting their capacity to sustainably influence the dynamics of rural financing. In the long run, the share explained by ICTs tends to increase, particularly in countries that have implemented coherent digital and agricultural policies, confirming their structural role in enhancing the stability and predictability of agricultural credit.

The VD evaluates the contribution of own and external shocks to the dynamics of agricultural credit within the VECM framework. It highlights the relative importance of ICTs, institutional factors and economic aggregates in shaping agricultural credit in both the short and long run (Lwanga and Adong, 2021).

In the short term (period 1), the variability of agricultural credit (lncredit_agri) is entirely explained by its own shocks (100%), reflecting strong inertia in agricultural financing behavior. This share gradually decreases to 71.6% in the long run (period 10), as innovations related to ICTs and institutions begin to play a larger role. Shocks from mobile phones usage (lnmobile) account for 12.8% of the variance from the second period and exceed 21% in the long run, underlining its key role in information dissemination and the facilitation of financial transactions (Batista and Vicente, 2020; GSMA, 2022, 2023).

Mobile money (lnmmoney) contributes only marginally in the short term (0.02%) but rises to 2.5% in the long term, indicating a delayed effect of digital finance on agricultural credit (Asuming and Frempong, 2023; Lwanga and Adong, 2021). Internet connectivity (lninternet) remains marginal (0.8%) due to limited rural infrastructure (Deichmann, 2019; FAO, 2021). Institutional variables such as regulatory quality (ihsregu) and political stability (ihsstab) have a moderate but lasting effect (0.5%–2.6%), reinforcing confidence and security in the financial system (Cheng et al., 2015). Finally, GDP per capita (lnpib_hbt) has a marginal impact (less than 0.4%), suggesting that overall economic growth does not automatically translate into increased agricultural credit, due to financial dualism and low rural banking penetration (World Bank, 2021; CGAP, 2021).

Overall, the VD analysis indicates that:

1. Agricultural credit remains largely driven by internal dynamics (share above 70%);

2. Mobile phones are the primary external factor (share exceeding 20% in the long term);

3. Mobile money acts as a gradual lever for financial inclusion (share ranging from 0.02% in the short term to 2.5% in the long term);

4. Institutions support financial stability and confidence;

5. Internet and economic growth still exert a limited structural effect.

These results underscore the complementarity between ICTs diffusion and institutional reforms in deepening rural finance and enhancing the resilience of agricultural credit in Sub-Saharan Africa (FAO, 2021; CGAP, 2021; GSMA, 2023).

To test the robustness of the VECM results, a dynamic system GMM model was applied. This approach addresses potential biases related to endogeneity, unobserved heterogeneity and reverse causality between ICTs, institutions and agricultural credit. The method also incorporates ICTs–institution interactions (tic_regu, tic_stab) to assess whether the impact of digital technologies on agricultural credit depends on the regulatory framework and socio-economic stability. The GMM results ( Appendix 2) confirm the robustness of the dynamic model. The Wald statistic (χ² = 146,523; p = 0.000) indicates overall significance. First-order autocorrelation is present (AR (1): p = 0.000) but absent at the second order (AR (2): p = 0.837), validating the model specification. The Sargan (p = 0.024) and Hansen (p = 0.433) tests confirm the validity of the instruments. Economically, agricultural credit exhibits low inertia (L.lncred_agri = −0.061; p = 0.07). Mobile phones usage has a significant negative effect (−0.281; p = 0.016), highlighting a temporary adjustment toward informal financial channels. Mobile money is not significant in the short term (0.025; p = 0.39), reflecting a delayed impact (Lwanga and Adong, 2021). Internet usage remains marginal (0.008; p = 0.82), whereas GDP per capita positively affects agricultural credit (0.107; p = 0.031). Institutionally, regulation slightly reduces the effectiveness of ICTs (−0.025; p = 0.007), while political stability enhances it (0.025; p = 0.007). These estimates confirm that the impact of ICTs depends on the institutional environment: socio-economic stability strengthens trust and fosters the digitalization of agricultural finance in Sub-Saharan Africa (Asuming and Frempong, 2023; GSMA, 2023).

The VECM results confirm the structural role of ICTs in agricultural financing. In the long term, mobile phones usage (+0.83%) and mobile money (+2.23%) exert a positive and significant effect on agricultural credit, indicating that financial digitalization strengthens access to credit. These findings align with those of Batista and Vicente (2020) in Mozambique and Asuming and Frempong (2023) in West Africa, who report that mobile money enhances banking access and borrowing capacity for smallholder farmers. Similarly, Lwanga and Adong (2021) note that in Uganda, digital transactions facilitate agricultural credit distribution by reducing intermediation costs.

The effect of Internet usage, while positive but not significant, reflects a delayed adoption of complex digital tools in rural areas where access remains uneven (GSMA, 2022). The negative and non-significant coefficient of GDP per capita suggests that average economic growth does not automatically translate into financial inclusion without targeted and coordinated policies (Medase and Savin, 2024). The negative and significant coefficients associated with regulatory quality (−0.863) and political stability (−1) indicate a phase of institutional adjustment. Several countries, such as Cameroon, Chad and Côte d’Ivoire, have recently strengthened financial supervision, which may temporarily constrain the smooth flow of agricultural credit before the regulatory framework stabilizes (FAO, 2021; World Bank, 2022).

In the short term, the error-correction term indicates that approximately 9.8% of disequilibria are corrected annually, reflecting a realistic adjustment speed in agricultural contexts where innovations diffuse slowly (CGAP, 2021). The positive short-term effects of mobile money (0.19%) and Internet connectivity (0.11%) demonstrate that ICTs rapidly stimulate access to agricultural finance, as observed in countries such as Kenya, Ghana and Uganda, where solutions like M-Pesa and MoB is enhance the flow of agricultural liquidity (Aker et al., 2016; GSMA, 2023).

Conversely, political instability negatively affects short-term outcomes, indicating that security tensions in the Sahel region (Mali, Niger, Burkina Faso) compromise the distribution of rural and agricultural credit (World Bank, 2022).

Overall, ICTs appear as essential levers for agricultural financing, provided they are supported by stable governance and flexible regulation (Deichmann, 2019). East and West Africa benefit the most from digitalization, whereas Central Africa is still in the early stages of progressive ICTs integration.

This research aimed to evaluate the impact of ICTs, specifically mobile phones, mobile money and the Internet, on agricultural credit in 30 Sub-Saharan African countries between 2015 and 2023. Relying on panel data from the World Bank (WDI), FAO and the WGI, the study employed both a VECM and a dynamic GMM estimation to assess the short- and long-term relationships among the variables.

The VECM estimates reveal a cointegration relationship between ICTs and agricultural credit. In the long term, mobile phones (+0.83%) and mobile money (+2.23%) exert a positive and significant effect on agricultural credit, highlighting the catalytic role of technologies in reducing information asymmetries and transaction costs. In the short term, the significant error-correction term indicates a gradual convergence toward equilibrium at a rate of 9.8% per year, reflecting the slow adaptation of the agricultural sector to digital financial innovations. Meanwhile, the GMM model confirms the robustness of these dynamic effects, emphasizing the importance of institutional frameworks in the diffusion of digital financial services through various ICTs channels.

Economically, these results suggest that strengthening financial inclusion through digital means contributes to increased agricultural financing, enhanced producer resilience and improved productivity. Monetary authorities and governments in Sub-Saharan Africa (Benin, Burkina Faso, Cameroon, Congo, Gabon, Guinea, Niger, Chad, Mali, Côte d’Ivoire) are thus encouraged to promote the interoperability of mobile money services, enhance financial cybersecurity and support digital literacy among agricultural producers.

In summary, ICTs and digital finance can boost agricultural credit by creating a traceable digital history for farmers that can be used for credit scoring so that mobile money enhances agricultural credit and facilitates farmers’ access to loans. Furthermore, such ICTs platforms will enhance liquidity, lower transaction costs and cash-related risks, enabling direct payments for harvests and promoting the development of additional financial services such as savings and insurance. Corresponding mobile money externalities can empower farmers to invest in their operations and bolsters financial inclusion in rural communities by providing a secure platform for transactions and savings.

Future research could expand the analysis to examine the interactions between ICTs, agricultural insurance and sustainable farming, or explore differentiated effects across crops or agro-ecological zones. These perspectives provide valuable insights into how digitalization can sustainably transform finance and agricultural development in Sub-Saharan Africa.

The authors are indebted to the editor and referees for constructive comments.

South Africa; Angola; Botswana; Malawi; Mozambique; Namibia; Zambia; Zimbabwe; Benin; Burkina Faso; Côte D’Ivoire; Gambia; Ghana; Guinea; Mali; Mauritania; Niger; Nigeria; Senegal; Sierra Leone; Togo; Cameroon; Democratic Republic of Congo; Chad; Ethiopia; Kenya; Uganda; Rwanda; Sudan and Tanzania.