Digital dynamics: exploring ICT's role in revolutionising Nigeria's trade sectors Open Access

Information and Communication Technology (ICT) is a critical factor in promoting inclusive development at both the micro and macroeconomic levels, as it stimulates various economic activities. It improves Total Factor Productivity (TFP) through industrial research, development, and innovation (Efobi, Tanankem, & Asongu, 2018; Franck & Galor, 2017). ICT has enhanced production capacity, which is crucial for overall economic growth and has improved economic welfare, as noted by Heshmati and Rashidghalam (2016) Adeleye and Eboagu (2019), Asongu and Le Roux (2017), Ejemeyovwi and Osabuohien (2018), Makun and Jayaraman (2020). In addition, it increased household income (Comin & Mestieri, 2013) and enhanced employment opportunities (Azu et al., 2021; Sovbetov, 2018). ICT has been shown to contribute to educational development (Uibu & Kikas, 2008), stimulate entrepreneurs' engagement (Shamaki et al., 2022), reduce female discrimination, and revitalise health delivery (Asongu & Odhiambo, 2018, 2020; Makun, Singh, Lal, & Chand, 2022; Kliner, Knight, Mamvura, Wright, & Walley, 2013).

As a driving force behind globalisation, ICT is increasingly transforming the global economy into a digital one (Azu et al., 2021). The ability of nations and businesses to efficiently process information is becoming more critical for maintaining competitiveness in international trade. This research seeks to contribute to the literature by examining the impact of ICT on Nigeria's trade sectors. Specifically, it will analyse how ICT influences exports and imports across various sectors, as classified under the Standard International Trade Classification (SITC). The study will also explore the consistency of ICT's impact across different sectors in Nigeria while assessing the short-term and long-term effects of ICT on trade. Rauch (1996), Belderbos & Sleuwaegen (1998), and Rauch & Casella (2003) identified information costs as barriers to international trade, leading to higher transaction costs. Lin (2014) and Wang and Choi (2018) further noted that Internet adoption enhances suppliers' access to customer and market data. ICT refers to the use of digital tools and systems like the Internet, mobile phones, and telecommunications infrastructure to facilitate communication, information sharing, and connectivity. It provides an avenue to enhance market access and reduce trade costs.

The literature offers diverse perspectives on the impact of ICT on bilateral and international trade, with many studies employing gravity models to explore this relationship. Freunda and Weinhold (2004) discovered that a 10% increase in web hosts led to a 1% rise in trade, particularly benefiting developing nations. Bojnec and Fertö (2009) highlighted that internet usage positively affected the export of manufactured goods in OECD nations, while Yadav (2014) found that internet use positively influenced the extensive margin of trade, though not the intensive margin. Lin (2014) noted that even small increases in internet use could boost bilateral trade, and Wang and Choi (2018) observed that ICT had a more significant positive impact on exports than imports in BRICS nations, particularly benefiting labour-intensive countries. Tay (2018) identified internet connectivity as having a significant effect on service trade and exports, with broadband and telephone subscriptions being especially impactful across various forms of trade.

Recent studies continue to support the positive role of ICT in international commerce. Dumor et al. (2023) found that ICT improves bilateral exports and economic growth in Eastern African BRI countries. Islam, Haque, Islam, Hassan, and Alam (2024) emphasised the importance of education human capital in enhancing ICT-driven trade flows. Azu and Nwauko (2021) and Rodriguez-Crespo, Marco, and Billon (2021) further confirmed the favourable impact of ICT on service exports and bilateral trade, with mobile phone usage playing a pivotal role. Özsoy, Ergüzel, Ersoy, and Saygılı (2022) demonstrated that ICT development promotes high-tech manufacturing exports in developing countries. Abendin, Duan, and Nkukpornu (2022) identified a positive impact of ICT on West African trade, and Kere and Zongo (2023) revealed that ICT usage boosts intra-African trade, though it may reduce imports of primary and total commodities. Overall, the literature underscores ICT's significant and wide-ranging effects on trade, particularly for exports.

Despite the extensive body of research, there remains a significant gap in understanding the specific impact of ICT on Nigeria's various trade sectors. Many studies have focused on broader regional contexts or comparative analyses across multiple countries, often neglecting Nigeria's unique trade dynamics. Moreover, the distinct effects of ICT across different sectors and the potential variations in ICT's influence on each have yet to be fully explored. This study aims to fill these gaps by focussing on Nigeria and employing a dynamic panel ARDL approach emphasising Pooled Mean Group (PMG) estimators. This method is chosen for its ability to distinguish between short-term and long-term effects, allowing for a more nuanced understanding of ICT's lasting impact on trade. According to the SITC classification, Nigeria's trade data is categorised into ten sectors: 0-Food and Animals; 1-Beverage and Tobacco; 2-Crude Materials, Inedible, Except Fuel; 3-Mineral Fuel; 4-Animal and Vegetable Oils, Fats, and Waxes; 5-Chemicals and Related Products; 6-Manufactured Goods; 7-Transport and Machinery; 8-Miscellaneous Manufactured Articles; and 9-Transactions and Commodities Not Classified Elsewhere.

A detailed and focused examination of Nigeria's trade sectors is essential to uncover the specific patterns and nuances that broader regional studies may overlook, using data from 1995 to 2022. Unlike the widely used gravity model in previous studies (Abendin et al., 2022; Rodriguez-Crespo et al., 2021; Azu & Muhammad, 2020; Azu, 2022; Muhammad, Diyoke, & Azu, 2020; Azu, 2019; Julius, Azu, & Muhammad, 2019), the dynamic panel ARDL approach with Pooled Mean Group (PMG) estimators provides a deeper analysis of both short- and long-term effects, adding a fresh perspective to the literature. This study focuses on Nigeria's trade data from 1995 to 2022 and categorised according to the Standard International Trade Classification (SITC). By incorporating ICT indicators such as internet penetration rate, mobile telephone subscriptions, and fixed telephone subscriptions, this research offers a comprehensive view of ICT's impact on trade, contributing to the ongoing discourse on global commerce and digital transformation. Choosing Nigeria as the focus of this study is justified by the country's unique position as one of Africa's largest economies and a key player in regional and global trade. Nigeria's diverse and complex trade sectors and significant ICT infrastructure growth provide a compelling case for examining the interplay between ICT and trade.

The significance of this study is its capacity to enlighten strategic decisions and policies that can improve Nigeria's trade performance by leveraging ICT advancements. Identifying the differential impacts of ICT across various sectors enables policymakers to customise interventions to optimise benefits in the most critical areas. Furthermore, comprehending the long-term and short-term impacts of ICT on trade can assist in developing sustainable trade policies that capitalise on digital technologies. This study addresses a critical lacuna in the literature and offers actionable insights to improve Nigeria's competitiveness in the global market, thereby contributing to economic growth and development.

The assimilation of new technology in any sector is crucially influenced by the Technology Acceptance Model (TAM) proposed by Davis (1989). TAM posits that the acceptance of a given technology is determined by the user's voluntary and willing intention to adopt and utilise the technology. This model, along with the Theory of Reasoned Action (TRA), as discussed by Asongu (2018) and Efobi et al. (2018), forms the theoretical foundation for understanding technology adoption in this research. The TRA, popularised by Fishbein and Ajzen (1975) and Ajzen and Fishbein (1980) and further reinforced by Bagozzi (1982), assumes that consumers make rational decisions by carefully considering the potential outcomes of their actions before forming an attitude. This model is intuitive and effective in clarifying attitudes and identifying factors that motivate intentionally adopted behaviours.

Davis (1989) extended the TRA framework with the TAM, which suggests that an individual's acceptance of technology can be explained by their enthusiastic intention to adopt and use it. Asongu (2018) elaborates that in this context, “intention” implies an individual's perception of the technology's usefulness or attitude towards its use. By leveraging these theoretical models, this research explores how Information and Communication Technology (ICT) revolutionises Nigeria's trade sectors, focussing on technology acceptance and utilisation dynamics.

The literature has provided varying assessments of the effects of ICT on bilateral and international commerce (Freunda & Weinhold, 2004; Bojnec & Fertö, 2009; Yadav, 2014; Lin, 2014; Wang & Li, 2017; Nath & Liu, 2017; Wang & Choi, 2018; Tay, 2018; Azu & Nwauko, 2021; Rodriguez-Crespo et al., 2021; Özsoy et al., 2022; Abendin et al., 2022; Kere & Zongo, 2023; Billon, Rodríguez-Andrés & Rodríguez-Crespo, 2023; Dumor et al., 2023; Shanmugalingam, Shanmuganeshan, Manorajan, Kugathasan, & Pathirana, 2023; Islam et al., 2024). For example, Freunda and Weinhold (2004) employed a gravity model to analyse data from 56 nations between 1995 and 1999. They discovered that a 10% increase in web hosts resulted in a 1% increase in trade. Their research emphasised that the Internet substantially impacted trade flows in developing countries, particularly the most impoverished.

Similarly, Bojnec and Fertö (2009) implemented the gravity model method to evaluate the influence of the Internet on the export of manufactured products across OECD nations from 1995 to 2003. The results of their panel regression analysis demonstrated that the utilisation of the Internet has a substantial positive impact on industrial exports and alleviates the negative consequences of distance. Yadav (2014) discovered that the extensive margin of export and import behaviour for enterprises was positively influenced by Internet use, while the intensive margin was not significantly affected. This was based on data from 52 Asian and African nations between 2006 and 2010. Lin (2014) demonstrated that bilateral trade would increase by 0.02-0.04% for every percentage point increase in internet use by utilising a gravity model to analyse data from approximately 200 nations between 1990 and 2006. Using panel data from 2000 to 2016, Wang and Choi (2018) investigated the influence of ICT on the trade volumes of the BRICS nations. Their research revealed that ICT had a more favourable impact on exports than on imports, with the effect diminishing as it progressed through the value chain. Furthermore, they discovered that the influence of ICT on trade increases over time, with labour-intensive nations benefiting more than resource-intensive ones. Tay (2018) found that internet connectivity significantly affects service trade and export but not service import, while fixed broadband and telephone subscriptions had the most substantial impact on all three forms of service trade.

The study by Dumor et al. (2023) explores the influence of ICT on bilateral trade and economic growth in Eastern African Belt and Road Initiative (BRI) countries, using a panel data structural gravity approach. It finds that greater access to ICT enhances bilateral exports and economic growth, especially within the East African Community (EAC), though more investment in ICT infrastructure is necessary to sustain this progress. Islam et al. (2024) examine how education human capital impacts ICT-trade relationships, showing that ICT usage significantly boosts merchandise export flows, particularly for countries with higher education levels. Wang and Li (2017) identify ICT as a source of comparative advantage in international trade, where a country's export in an industry increases with improved ICT development and research intensity. Nath and Liu (2017) highlight that ICT development positively influences the trade of services, particularly ICT-enabled services like financial and business services. Shanmugalingam et al. (2023) find that e-commerce is crucial in boosting international trade among Asian countries, emphasising the importance of technology and telecommunications policies for enhancing trade.

Azu and Nwauko (2021) evaluated the impact of digital technology on service commerce in West Africa from 1995 to 2020 by employing a panel ARDL model. Their results were consistent with those of Wang and Choi (2018), suggesting that digital transformation has a substantial positive impact on service exports in the medium and long term, but has a less significant effect on imports. Rodriguez-Crespo et al. (2021) employed a gravity model to analyse data from 55 countries between 2004 and 2013, demonstrating that ICT has a significant and favourable impact on bilateral trade. The most significant effect was shown by mobile phone usage. Özsoy et al. (2022) found that ICT can accelerate the manufacturing of high-tech goods in developing nations. Their study, using panel data from 2007 to 2017, indicated that the ICT development index significantly impacts the export of high-tech items. Abendin et al. (2022) also discovered a beneficial impact of ICT on bilateral commerce in West Africa, utilising data from 2000 to 2018. Kere and Zongo (2023) demonstrated that intra-African commerce, particularly internet use, is positively influenced by ICT usage, while imports of primary items and total commodities are negatively impacted.

Figure 1 presents a comprehensive overview of Nigeria's trade growth and ICT development from 1995 to 2022. It showcases the import and export growth rates alongside the proliferation of ICT, indicated by the percentage of individuals using the Internet, mobile cellular subscriptions per 100 people, and fixed telephone subscriptions per 100 people. In the mid-1990s, Nigeria had minimal ICT infrastructure, with virtually no Internet usage and extremely low mobile and fixed telephone subscriptions (Azu et al., 2021; Azu & Nwauko, 2021). During this period, trade growth exhibited significant volatility, with substantial fluctuations in both import and export growth rates (See  Figure 1).

As we move into the early 2000s, the data shows a marked increase in Internet usage and mobile cellular subscriptions. By 2002, there was a notable jump in Internet usage to 0.32% of the Population and mobile subscriptions to 1.21 per 100 people, reflecting the beginning of a rapid ICT adoption phase. This period coincides with fluctuating trade growth, indicating that while ICT infrastructure was being established, its direct impact on stabilising trade growth was not immediately apparent. Import and export growth rates continued to show substantial variability, with years of significant positive and negative development.

In the latter years, particularly from 2010 onwards, there is a clear correlation between the expansion of ICT and trade growth. Internet usage surged from 11.5% in 2010 to 61.74% in 2022, and mobile cellular subscriptions dramatically increased from 54.24 to 101.69 per 100 people. This period also saw a general trend of more stable trade growth despite some years of decline, such as in 2015 and 2020. The increased connectivity likely facilitated better communication, efficient market transactions, and access to global markets, contributing to more consistent trade performance. However, despite these advancements, fixed telephone subscriptions declined over time, highlighting a shift towards mobile and internet-based communication technologies as primary tools for trade and economic activities.

In evaluating the influence of digitalisation on service trade in West African countries, Azu and Nwauko (2021) augmented a gravity model presented by Choi (2010) as follows:

Where $X/Mit$ represented service export and import; $Tit$ is a vector of digitisation that is proxied with the internet penetration rate $(T1)$ and mobile telephone subscription (⁠$T2)$⁠; $Yit$ is a vector of Market size (⁠$Git)$ is proxied with Real GDP (⁠$G1,it)$ and Population (⁠$G2,it)$ and financial depth measured with broad money supply (M2). In the interest of this research, Equation (1) is modified to realise the objectives of this research as follows:

The dependent variable $X/Mit$ stands for export $Xit$ and import $Mit$ in various traded sectors in Nigeria. In this research, trade is captured in 10 different sectors based on SITC classification: 0-Food and Animals; 1-Beverage and Tobacco; 2-Crude Materials, inedible, except Fuel; 3-Mineral Fuel; 4-Animal and Vegetable oil, fat and Waxes; 5-Chemical and related products; 6-Manufactured Goods; 7-Machinery and Transport; 8-Miscellaneous Manufactured articles; 9-Commodities and Transactions not classified elsewhere in the SITC. $Tit$ represents a vector of ICT captured in three perspectives: internet penetration rate (⁠$T1,it$⁠), mobile telephone subscriptions (⁠$T2,it$⁠) and fixed telephone subscriptions (⁠$T3,it$⁠). All the ICT variables are taken as a percentage of the Population. It is expected that ICT will be instrumental to an increase in trade since available literature has affirmed that it reduces trade costs (Yadav, 2014; Lin, 2014; Wang & Choi, 2018; Freunda & Weinhold, 2004).

On the other hand, $Yit$ stand for current GDP, which provides a real-time, accurate representation of economic activity at market prices, making it particularly relevant for examining the immediate effects of ICT advancements on trade. It includes inflation and price changes, offering a comprehensive view of the economy and the impact of ICT investments on trade sectors under current conditions. $Fit$ is the labour participation rate that captures the percentage of the labour force employed and can reflect the economic activity and productivity levels in different sectors. Finally, $Rit$ is the nominal exchange rate, which is crucial as it directly affects the cost of imports and exports, influencing trade balances and the competitiveness of domestic industries in the global market. Table 1 presents the source of data and a priori expectation.

This study examines ten traded sectors based on SITC single digits over 28 years from 1995 to 2022. It employs the panel-ARDL model, proposed by Pesaran and Smith (1995) and Pesaran, Shin, and Smith (1999), contingent on the stationarity of the variables, whether integrated at I(1) or I(0). Therefore, stationarity tests such as the Im-Pesaran-Shin (IPS) panel unit-root test (Im et al., 2003), which accounts for dependence between individuals and heterogeneity across cross-sections, are essential. The panel ARDL technique is reliable for estimating panel data that meet its criteria, allowing for the estimation of both long- and short-run coefficients using the Mean Group (MG) estimator to address bias from heterogeneous slopes in dynamic panels.

The MG estimator averages the long-run ARDL model parameters across individual countries, providing consistent estimates but potentially inefficient with homogeneous slopes. Conversely, Pesaran et al. (1999, 2001) suggest the Pooled Mean Group (PMG) estimator for more efficient estimation when long-run coefficients are homogeneous across groups, allowing short-run parameters to be heterogeneous. The choice between MG and PMG depends on the Hausman (1978) test, which evaluates whether MG and PMG estimates differ significantly. A p-value below 0.05 rejects the null hypothesis, favouring MG, while a p-value above 0.05 supports using the more efficient PMG.

Generally, the Panel ARDL (p, q, q…, q) model can be specified as follows.

Where $Zit$ is the dependent variable, $(Vit′)′$ is K*1 vector that allowed to be purely I(0) or I(1) or cointegrated; $γij$ is the coefficient of the lagged dependent variable called scalar; $φij$ are k*1 coefficient vectors; $σi$ is the unit-specific fixed effects; i = 1,…,N; t = 1,2,…,T; p, q are optimal lag orders; $εit$ is the error term. In this study, the re-parameterised Panel ARDL (p, q, q…, q) error correction model is formulated as follows:

Notes: $βi=$ group-specific speed of adjustment coefficient (expected that $θi<0$⁠)

$γi′$ = vector of long-run relationships

$ECT=[lnX/Mi,t−1−γi′(lnTi,t+lnYi,t+lnFi,t+lnRi,t)]$ the error correction term

$ωij$⁠, $δij′$ represent the short-run dynamic coefficients. All variables are in natural logarithm.

Azu et al. (2021) and Azu and Nwauko (2021) chose the panel ARDL estimation technique because it accounts for the lagged forms of both dependent and independent variables. This approach effectively captures the true impact of ICT on trade, as the implementation of ICT requires time for adoption and integration, thus reflecting its effects on exports and imports over time.

The descriptive statistics in Table 2, Panel A summarise the key characteristics of the variables used in the study. With 280 observations for each variable, the means, standard deviations, and ranges (minimum and maximum values) offer insight into the data distribution. For instance, the mean of exports (⁠$lnXi,t$⁠) is 12.2524 with a standard deviation of 2.4707, indicating variability in export values. Imports (⁠$lnMi,t$⁠) have a higher mean of 13.8265, reflecting larger average import values, while the mean for internet penetration rate (⁠$lnT1,i,t$⁠) is 1.1239, suggesting relatively lower ICT adoption. The negative mean value of the fixed telephone subscriptions (⁠$lnT3,i,t$⁠) at −1.3034 indicates lower baseline fixed telephone usage. The GDP (⁠$lnYi,t$⁠) has a mean of 26.2487, showing relatively high economic output, while the labour force (⁠$lnFi,t$⁠) and exchange rate (⁠$lnRi,t$⁠) show moderate means of 4.0911 and 4.8652, respectively. The standard deviations highlight the data spread, with mobile telephone subscriptions (⁠$lnT2,i,t$⁠) and internet penetration rate (⁠$lnT1,i,t$⁠) showing the highest variability.

The correlation matrix in Table 2, Panel B illustrates the relationships between the variables. Notably, exports (⁠$lnXi,t$⁠) and imports (⁠$lnMi,t$⁠) have a moderate positive correlation of 0.4737, suggesting that exports increase as imports increase. ICT-related variables internet penetration rate (⁠$lnT1,i,t$⁠) and mobile telephone subscription (⁠$lnT2,i,t)$ show strong positive correlations with each other (0.9831), indicating that internet and mobile subscriptions are closely linked. However, fixed telephone subscriptions (⁠$lnT3,i,t$⁠) negatively correlates with most variables, especially internet penetration rate (⁠$lnT1,i,t$⁠) (−0.5086), suggesting that higher internet usage is associated with lower fixed telephone subscriptions. GDP (⁠$lnYi,t$⁠) positively correlates with most variables, particularly internet penetration rate (⁠$lnT1,i,t$⁠) (0.799), indicating that higher GDP is associated with greater ICT adoption. Interestingly, the exchange rate (⁠$lnRi,t$⁠) shows positive correlations with most variables, especially with internet penetration rate (⁠$lnT1,i,t$⁠) (0.8778), suggesting a relationship between the exchange rate and the ICT variable. These correlations provide a preliminary understanding of how these variables interact, guiding further analysis of the impact of ICT on trade.

The IPS unit root test results in Table 3 indicate that most variables are integrated at order I(1), becoming stationary only after first differencing. Specifically, exports (⁠$lnXi,t$⁠), imports (⁠$lnMi,t$⁠), fixed telephone subscriptions (⁠$lnT3,i,t$⁠), nominal GDP (⁠$lnYi,t$⁠), and labour force participation rate (⁠$lnFi,t$⁠) are non-stationary at level but stationary at first Difference, implying they are I(1). On the contrary, the internet penetration rate (⁠$lnT1,i,t$⁠), mobile telephone subscriptions (⁠$lnT2,i,t$⁠), and exchange rate (⁠$lnRi,t$⁠) are stationary at the level, indicating they are I(0). Therefore, the mixed integration orders of the variables support the suitability of the panel-ARDL model for further analysis.

The export model was estimated thrice for the internet penetration rate model (⁠$XitT1)$⁠, mobile telephone subscriptions model (⁠$XitT2)$⁠, and fixed telephone subscriptions model (⁠$XitT3)$ respectively to avoid the issue of multicollinearity (Azu & Nwauko, 2021; Azu et al., 2021). Before estimating the long-run and short-run impact of ICT variables on sectorial export in Nigeria, it's essential to establish a long-run relationship. In the Panel ARDL technique, a bound test for cointegration is unnecessary. Instead, the long-run relationship is determined using the error correction term (ECT), as shown in Table 4. According to Banerjee et al. (1998), the export models meet the criteria with a negative ECT (−1) of −0.241, −0.659, and −0.299, for internet penetration rate, mobile telephone subscriptions and fixed telephone subscriptions, respectively, and all significant at 1% level. These indicate an adjustment speed of 24.1, 65.9, and 29.9% toward long-run equilibrium. The negative value of the ECT, typically ranging between −1 and 0, implies the absence of serial correlation and instability issues commonly caused by structural breaks in panel data, as suggested by Sovbetov (2018) and Sovbetov and Saka (2018).

In the long run, the estimated coefficient for the internet penetration rate is negative (−0.587). Still, it is not statistically significant, indicating that changes in internet penetration do not have a reliable impact on exports in this model. On the other hand, the coefficient for mobile telephone subscriptions is also negative (−0.352) and statistically significant at the 1% level. This significant negative coefficient implies that a 1% increase in mobile telephone subscriptions would result in a 35.2% decrease in exports, assuming all other factors remain constant. This suggests a potentially adverse effect of mobile phone proliferation on export activities, possibly due to increased consumer focus on domestic consumption or other factors not captured in this analysis. Meanwhile, the estimated impact of fixed telephone subscriptions is positive but not statistically significant, indicating that changes in fixed telephone subscriptions do not have a reliable or substantial effect on export levels in the long run. This lack of significance may suggest that fixed telephone lines are less relevant in influencing trade activities compared to mobile telephony and internet penetration in the context of this study.

In the short run, the results show more consistency compared to the long run. The internet penetration rate's estimated coefficient is −0.724, statistically significant at the 5% level, indicating a 1% increase in internet penetration results in a 72.4% decrease in exports. This suggests that increased internet access might shift focus towards domestic markets. Similarly, the coefficient for mobile telephone subscriptions is −0.219 and significant at the 5% level, meaning a 1% increase leads to a 21.9% decrease in exports. For fixed telephone subscriptions, the coefficient is −0.530 and statistically significant, implying a 1% increase in fixed telephone subscriptions results in a 53.0% decrease in exports. These findings indicate that increases in internet penetration, mobile phone subscriptions, and fixed telephone subscriptions negatively affect export levels in the short run, possibly due to increased domestic consumption or resource reallocation away from export activities.

Concerning individual sectors and considering internet penetration rate (See  Appendix I), only Beverage and Tobacco (1); Crude Materials, inedible, except Fuel (2); Mineral Fuel (3) and Commodities and Transactions not classified elsewhere in the SITC (9) reported negative coefficient of −2.712, −1.034, −0.945 and −1.809 respectively and statistically significant. Concerning mobile telephone subscriptions ( Appendix II), Crude Materials (2), inedible, except Fuel; 3-Mineral Fuel (3), and -Machinery and Transport (7) also reported negative coefficients of −0.295, −0.287 and −0.599, respectively and statistically significant. Finally, fixed telephone subscriptions ( Appendix III) reported negative coefficients of −0.686, −0.661, −0.654, and −1.427 for Crude Materials, inedible, except Fuel (2) 3-Mineral Fuel (3), Manufactured Goods (6) and Miscellaneous Manufactured articles (8) respectively. This aligns with the overall short-run result and reflects the findings in Azu and Nwauko (2021) and Wang and Choi (2018) that these digitalisation elements usually negatively influence exports.

Again, the import model was estimated thrice for internet penetration rate (⁠$MitT1)$⁠, mobile telephone subscriptions (⁠$MitT2)$⁠, and fixed telephone subscriptions (⁠$MitT3)$ respectively to avoid the issue of multicollinearity (Azu & Nwauko, 2021; Azu et al., 2021). The estimated error correction term (ECT) in Table 5 meets the criteria for establishing a long-run effect. With negative ECT values of −0.0873, −0.608, and −0.744 for internet penetration rate, mobile telephone subscriptions, and fixed telephone subscriptions, respectively, all statistically significant, there is an adjustment speed of 8.7, 60.8, and 74.4% towards long-run equilibrium. The ECT's negative value, typically between −1 and 0, indicates the absence of serial correlation and instability issues commonly caused by structural breaks in panel data, as suggested by Sovbetov (2018) and Sovbetov and Saka (2018) (see Table 5).

From Table 5, the import direction seems to have benefited more from ICT variables in the long run. The internet penetration rate reported a positive but not statistically significant coefficient. However, mobile telephone subscriptions showed a positive coefficient of 0.179, statistically significant at 1%, indicating that a 1% increase in mobile subscriptions results in a 17.9% increase in imports. Fixed telephone subscriptions had an even higher positive coefficient of 0.415, also significant at 1%, suggesting that a 1% increase in fixed telephone subscriptions leads to a 41.5% increase in imports. These findings imply that in the long run, ICT advancements, particularly in mobile and fixed telephone subscriptions, significantly boost import activities in Nigeria, highlighting the role of communication technology in enhancing trade dynamics.

In the short run, internet penetration presents a negative coefficient of −0.915, statistically significant at 5%, suggesting a 91.5% decrease in imports. This indicates that increased internet access might divert resources away from import activities. Similarly, fixed telephone subscriptions report a negative coefficient of −0.507, significant at 1%, implying a 50.7% reduction in imports with a 1% increase in fixed telephone subscriptions. This might be due to a shift in focus towards more traditional communication methods over trade activities. Although mobile telephone subscriptions also report a negative coefficient in the short run, it is not statistically significant, indicating that its impact on imports is unclear. These findings highlight ICT's complex and potentially adverse effects on short-run import activities in Nigeria.

In analysing individual sectors, internet penetration negatively impacted the Mineral Fuel (3) sector with a significant coefficient of −1.955, indicating a substantial decrease in exports ( Appendix IV). Mobile telephone subscriptions did not show statistically significant results in the short run for any sector, suggesting their impact on sector-specific exports is minimal ( Appendix V). Conversely, fixed telephone subscriptions ( Appendix VI) had significant negative effects across multiple sectors: −0.564 for Food and Animals (0), −0.525 for Beverage and Tobacco (1), −0.857 for Crude Materials, Inedible, except Fuel (2), −1.135 for Animal and Vegetable Oil, Fat and Waxes (4), −0.430 for Manufactured Goods (6), and −0.572 for Machinery and Transport (7). These results highlight that traditional communication methods, like fixed telephones, are associated with decreased exports in these sectors.

The findings of this study align with previous literature on the varied effects of ICT on trade, particularly the contrasting impact on exports and imports. Freunda and Weinhold (2004) noted that internet usage positively affects trade in developing countries, though resources can be diverted toward domestic markets, similar to the negative long-term relationship between mobile phone subscriptions and exports found here. This is consistent with Wang and Choi (2018), who showed that ICT enhances domestic consumption more than international trade, particularly in the short run. Similarly, Nath and Liu (2017) highlighted that ICT's influence on services trade is stronger for ICT-enabled services like financial and business services, implying that certain trade sectors may benefit more than others. The short-term disruptions in export activities in this study echo Azu and Nwauko's (2021) findings, where digital transformation in West Africa caused short-term trade challenges due to the reallocation of resources toward domestic markets.

Furthermore, the long-term positive impact of ICT on imports aligns with Bojnec and Fertö's (2009) research, which demonstrated that traditional telecommunication infrastructure, such as fixed telephone subscriptions, continues to play a role in trade. This study's finding that mobile phone subscriptions foster imports is also supported by Wang and Li (2017), who emphasised that mobile technology enhances access to international markets by improving transaction efficiency and communication. Özsoy et al. (2022) further noted that ICT can stimulate high-tech goods manufacturing in developing nations, which may boost import demand for advanced production technologies. Similarly, Abendin et al. (2022) found that ICT positively influences West African bilateral commerce, which underscores the findings in this study that imports benefit from ICT-driven improvements in communication and market access.

The implications of this study build on a broader understanding of ICT's sectoral and temporal impacts on Nigeria's trade. Long-term ICT benefits, especially for imports, affirm the importance of mobile and fixed telephony in fostering international trade, as noted by Rodriguez-Crespo et al. (2021). However, the short-term disruptions reflect the transitional nature of ICT's role, with temporary challenges to imports and exports, as noted by Tay (2018) and Kere and Zongo (2023). Studies like Dumor et al. (2023) and Islam et al. (2024) suggest that education and infrastructure investments are critical to maximising ICT's trade benefits. These findings highlight the need for policies that mitigate short-term disruptions while enhancing long-term digital transformation benefits, particularly in a developing economy like Nigeria, as echoed by Shanmugalingam et al. (2023).

The findings of this study, when linked to the Technology Acceptance Model (TAM) and Theory of Reasoned Action (TRA), emphasise how perceived usefulness and rational decision-making affect ICT adoption and trade. The negative long-term impact of mobile phone subscriptions on exports aligns with TAM, suggesting that Nigerian traders may not fully recognise mobile technology's benefits for international trade, echoing Freunda and Weinhold (2004) and Wang and Choi (2018). Nath and Liu (2017) also noted that ICT's sectoral influence varies, with domestic markets often benefitting more than exports. Conversely, the positive long-term effect of ICT on imports supports TRA, as businesses adopt ICT to improve market access and logistics, consistent with Wang and Li (2017) and Bojnec and Fertö (2009). The study's results, which mirror Özsoy et al. (2022) and Abendin et al. (2022), show that ICT enhances communication and transaction efficiency for imports. However, short-term disruptions in both imports and exports reflect transitional ICT adoption challenges, as noted by Azu and Nwauko (2021), Tay (2018), and Kere and Zongo (2023), requiring focused policy efforts to maximise long-term benefits.

The study investigates the short- and long-run impacts of ICT on sectoral exports and imports in Nigeria. In the long run, internet penetration shows a negative coefficient (−0.587) but is not statistically significant, suggesting no reliable impact on exports. Mobile telephone subscriptions have a significant negative coefficient (−0.352), indicating that a 1% increase in mobile subscriptions leads to a 35.2% decrease in exports, possibly due to increased domestic consumption. Fixed telephone subscriptions have a positive but insignificant impact on exports. In the short run, internet penetration (−0.724), mobile phone subscriptions (−0.219), and fixed phone subscriptions (−0.530) all negatively affect exports significantly, highlighting a potential shift towards domestic markets or resource reallocation.

For imports, the long-run analysis shows a positive but insignificant effect of internet penetration. Mobile telephone subscriptions have a significant positive impact (0.179), and fixed telephone subscriptions have an even higher positive effect (0.415), indicating that ICT advancements boost import activities. In the short run, internet penetration (−0.915) and fixed telephone subscriptions (−0.507) significantly reduce imports, suggesting a diversion of resources from import activities. Mobile telephone subscriptions also negatively impact the short run, but this is not statistically significant. These findings underscore ICT's complex and varying implications on Nigeria's trade dynamics, with ICT promoting imports while potentially hindering exports in the short term. To maximise the benefits of ICT for trade in Nigeria, policymakers should invest in targeted training programs that enhance traders' understanding of the utility of mobile technology for international trade, thus addressing the negative perceptions of its impact on exports. Developing robust infrastructure that integrates ICT across various trade sectors can facilitate smoother transitions during digital transformation, ultimately supporting domestic and international market access.

The supplementary material for this article can be found online.