The impact of trade on African welfare: Does seaport efficiency channel matter?

Ayesu, Enock Kojo; Sakyi, Daniel; Arthur, Eric; Osei-Fosu, Anthony Kofi Article The impact of trade on African welfare: Does seaport efficiency channel matter? Research in Globalization Provided in Cooperation with: Elsevier Suggested Citation: Ayesu, Enock Kojo; Sakyi, Daniel; Arthur, Eric; Osei-Fosu, Anthony Kofi (2022) : The impact of trade on African welfare: Does seaport efficiency channel matter?, Research in Globalization, ISSN 2590-051X, Elsevier, Amsterdam, Vol. 5, pp. 1-13, https://doi.org/10.1016/j.resglo.2022.100098 This Version is available at: https://hdl.handle.net/10419/331030 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ Research in Globalization 5 (2022) 100098 Available online 27 September 2022 2590-051X/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/bync-nd/4.0/). The impact of trade on African welfare: Does seaport efficiency channel matter? Enock Kojo Ayesu * , Daniel Sakyi, Eric Arthur, Anthony Kofi Osei-Fosu Department of Economics, Kwame Nkrumah University of Science and Technology, Private Mail Bag, Kumasi, Ghana ARTICLE INFO JEL Codes: C33 F10 I30 O55 Keywords: Seaport efficiency Trade Welfare system-GMM Africa ABSTRACT This paper examines the role of seaport efficiency in the relationship between trade and welfare outcomes in Africa. Data on 28 African countries for the period 2006 to 2018 is employed while the system generalized method of moments estimation technique is used for the empirical analysis. The paper reveals that seaport efficiency and trade enhances welfare outcomes directly in Africa. More importantly, the paper finds that seaport efficiency complements trade to have its largest positive impact on welfare outcomes in Africa in both the short and long run periods. Based on these findings, we suggest that policies aimed at using trade to enhance welfare should pay a critical attention to efficiency of African seaports by increasing investment in seaport infrastructure, space for container vessels, container terminal capacity to reduce time delays at seaports, the number of total ship services, and port surfaces and depths of water areas to serve ships at seaports, is what officials of ports and harbour authorities, regional economic communities, the African continental free trade area, and other organizations that champions international trade should consider in their policy reforms. 1. Introduction Increasingly, various countries in the world are searching for ways to enhance the welfare of their citizens. Following this, an important driver of economic prosperity and hence welfare worth mentioning is international trade (UNCTAD, 2018). The role played by trade as a growth enhancing policy variable has the benefits of enhancing welfare by reducing poverty, boosting employment, education, and health status (see Dollar & Kraay, 2004; Winters, McCulloch, & McKay, 2004; Nguyen Viet, 2015; Narayanan, Sharma, & Razzaque, 2016; Herzer, 2017; Sakyi, Bonuedi, & Opoku, 2018; Byaro, Nkonoki, & Mayaya, 2021; Hamdi & Hakimi, 2021; Zhao, Geng, Liu, Liu, Zheng, Xue, & Zhang, 2022). The afore benefits notwithstanding, the effectiveness of trade in improving welfare depends highly on how trade is facilitated. As regards trade facilitation, globally, seaports play a very significant role because about 80 percent of merchandise trade (in volume) occurs through the sea (African Development Bank [AfDB], 2015; Hlali & Hammami, 2017; UNCTAD, 2016). Interestingly, in the case of Africa, roughly 90 percent of total trade is done via the sea (UNCEA, 2016). Therefore, ensuring that seaports are efficient remain a top priority concern for most economies all over the world, especially those in Africa with the view to improving welfare outcomes (Blonigen & Wilson, 2008; African Development Bank [AfDB], 2015; Hoffmann & Kumar, 2013). The performance of welfare indicators in Africa remains generally low compared with the rest of the world. Life expectancy at birth and secondary school enrollment (Gross) in Africa remains generally low compared to developed economies such as Europe and North America. For instance, from 2006 to 2018, life expectancy at birth in Africa averaged 58.07, while that of Europe and North America averaged 80.13 and 78.73, respectively (World Bank, 2020). Similarly, according to the World Bank reports, secondary school enrollment (Gross) for Africa from 2012 to 2018 averaged 53.78, while Europe and North America recorded averages of 106.47 and 90.07, respectively (World Bank, 2020). The Development Aid Report (2021) also shows that as of February 2021, 36 percent of the total population in Africa, representing 490 million people are living in extreme poverty. This threatens the ability of the continent towards achieving the Sustainable Development Goal of eradicating poverty by 2030. It has been argued that seaport efficiency could lead to improvement in welfare indicators because of its potential of increasing trade revenue and creating opportunities for trading activities through trade facilitation. This leads to improvement in welfare as for instance the availability of varied goods and services, which otherwise would not be available to the citizens of a country. These can be argued to contribute * Corresponding author. E-mail address: [email protected] (E. Kojo Ayesu). Contents lists available at ScienceDirect Research in Globalization journal homepage: www.sciencedirect.com/journal/research-in-globalization https://doi.org/10.1016/j.resglo.2022.100098 Received 10 July 2022; Received in revised form 22 September 2022; Accepted 26 September 2022 Research in Globalization 5 (2022) 100098 2 to improved health and educational outcomes through the improvement in life expectancy rates. According to the literature (see, for example Winters et al., 2004; Narayanan et al., 2016; Sakyi et al., 2018), increased trade creates more jobs for the populace, and this potentially translates into positive welfare outcomes. Notwithstanding the positive benefits of an efficient seaport, African countries underperform in terms of seaport efficiency, which is largely attributed to long duration of ships stay at seaports, inadequate quality of container handling capacity, poor service quality, poor liner shipping connectivity, increasing cost of trade, and poor infrastructural facilities, among others (Raballand, Refas, Beuran, & Isik, 2012). For instance, on average, compared with most international seaports where cargos spend three to four days in seaports before departure, cargos spend nearly 20 days in seaports before departure in most African countries (African Development Bank[AfDB], 2015; Raballand et al., 2012). It therefore becomes imperative to understand the positive influence of seaport efficiency in the improvement of welfare outcomes to encourage African countries to improve their seaports. Empirically, studies have been devoted to investigating the correlation between trade and/or trade facilitation and welfare outcomes (Cabote & Hiwatari, 2014; Nguyen Viet, 2015; Narayanan et al., 2016; Herzer, 2017; Novignon, Atakorah, & Djossou, 2018; Sakyi et al., 2018; Byaro et al., 2021; Hamdi & Hakimi, 2021; Zhao et al., 2022). However, while trade facilitation and trade are important for welfare, the role of seaport efficiency in the relationship between trade and welfare in the context of African countries has not been studied. Therefore, it becomes imperative to understand how the interaction between seaport efficiency and trade affect welfare outcomes as well as the direct effect of seaport efficiency on welfare outcomes in Africa. To this end, the study provides a novel evidence on how seaport efficiency serves as an important channel via which trade can influence welfare outcomes in African countries with seaports. In studies as this, how seaport efficiency, trade and welfare are measured is very vital as the use of an adequate measure will better inform policy. As regards the seaport efficiency measure, we use the UNCTAD (2019) liner shipping connectivity index to capture seaport efficiency. This is because this measure is not only directly linked to the physical infrastructure, service dimensions, and cargo handling capacity aspect of seaport efficiency but it is also based on how countries are connected in terms of shipment of goods and services. Using this measure would provide more insight to policy makers on how to improve the efficiency of seaports in Africa. In addition, we measure both trade and welfare in a broader sense, by considering a number of significant primary indicators. We use exports, imports, and overall trade as percentage of GDP as measures of trade also to ensure the robustness of our estimates. Additionally, broader indicators of welfare namely gross secondary education enrolment, life expectancy, household consumption expenditure per capita and human development index were employed. From estimation point of view, the two-step system generalized method of moments panel data estimation technique that deals adequately with potential endogeneity concerns is employed. The study is therefore important, if African countries are to reap the full welfare benefits of trade especially with the existence of the African Continental Free Trade Area (AfCFTA) agreement. The remaining of the study is structured as follows. Sections 2 and 3 cover literature review on seaport efficiency, trade and welfare and the methodology. respectively. Sections 4 and 5 cover the empirical results and discussion and the conclusion and policy suggestions, respectively. 2. Literature review 2.1. Theoretical review Theoretically, the effect of seaport efficiency and trade on welfare can be traced to the Heckscher and Ohlin (1991) model of international trade. According to the model, trade will increase when countries are allowed to trade freelyand if facilitated efficiently, it will drive economic growth and improvement in the welfare of the populace. Thus, with trade, countries can focus on the production of commodities in which they have the greatest factor intensity which leads to improved trade revenue, higher incomes and welfare. The increased trade revenue and higher income increase access to basic requirements of life such as food and nutrition, clean water and improved sanitation, quality education, and increase expenditure on health leading to a rise in standard of living (Owen & Wu, 2007; Levine & Rothman, 2006; Novignon et al., 2018; Byaro et al., 2021). The implication of this is that a rise in living standard is associated with increased enrollment in education and consumption, increase life expectancy, and improvement in overall human development (see: Novignon et al., 2018; Byaro et al., 2021). Also, trade enables the importation of new technology to improve our healthcare system. This means that increased trade can positively affect life expectancy and human development by increasing access to pharmaceutical services and medical treatment (Owen & Wu, 2007; Novignon et al., 2018; Byaro et al., 2021; Hamdi & Hakimi, 2021). Furthermore, increased trade can positively affect education by increasing access to quality educational materials used for learning leading to improvement in human development. The afore is particularly important for African countries beleaguered with vibrant health facilities and pharmaceutical firms as well as educational input materials. In spite of the above, the trade effects on welfare outcomes is only possible if there is an improved seaport efficiency as postulated in the literature (Blonigen & Wilson, 2008). The possible linkage between seaport efficiency and measures of welfare is gleamed from the literature on the effects of seaport efficiency on economic performance and on how seaport efficiency affects trade costs and its associated effects on trade and welfare outcomes. Thus, seaport efficiency enhances economic performance as it generates resources for a country due to its effects on increasing trade revenue. The literature has made clear that improved economic growth is an important source of fiscal space for the improvement of welfare through the health sector, education sector, household consumption, human development, population income, rising standard of living, and job creation, among others (Dollar & Kraay, 2004; Nguyen Viet, 2015; Sakyi et al., 2018; Novignon et al., 2018). This suggest that seaport efficiency could indirectly affect the measures of welfare through its impact on economic performance. Regarding how seaport efficiency affect trade costs, it has been well acknowledged in the literature that an efficient seaport lowers transportation (shipping or trade) costs and thus facilitate trade (import and export) activities (Clark, Dollar, & Micco, 2004; Blonigen & Wilson, 2008; Merk & Deng, 2012; Raballand et al., 2012; Lei & Bachmann, 2019). The implication is that, since efficient seaports decreases long duration of ship stay at seaports, limits seaport congestions, lowers tariffs, increases access to global trade, and increase the participation of a seaport in the carriage of its seaborne trade among several others, it reduces the transportation (shipping or trade) costs that seaports users (exporters, importers, shippers, and shipping lines) would have incurred on freight, shipping and port charges in the event of dealing with an inefficient seaports. The consequence of a reduction in the transportation cost are that, for seaport users such as shippers, importers, and shipping lines, they are in better position to increase access to global trade and the aftermath is an increase in trade/shipping due to a reduction in transportation costs due to improved seaport efficiency. In addition, efficient seaports are likely to steal container traffic from other neighboring seaports countries and this increases the flow of trade (Clark et al., 2004; Raballand et al., 2012; Mlambo, 2021). From the aforementioned discussions, with increase trade/ shipping, being the consequence of a reduction in transportation costs, there will be an increased in trade revenue, foreign exchange, increase access to foreign technology, capital goods and a variety of healthier consumable products among others that has significant impact on the economic growth of countries involved in trade which in turn supports the E. Kojo Ayesu et al. Research in Globalization 5 (2022) 100098 3 improvement in welfare outcomes (Wilson, Mann, & Otsuki, 2003; Levinson, 2008). 2.2. Empirical review Empirically, studies on trade facilitation and trade on welfare outcomes have been examined. A number of indicators have been used in these studies to measure welfare from health, education to poverty. For example, Herzer (2017), Sakyi et al. (2018), Novignon et al. (2018), Byaro et al. (2021), and Zhao et al. (2022) used life expectancy at birth as the measure of welfare and concluded that trade and trade facilitation contributes to increases in life expectancy at birth. These studies have therefore suggested that a potential channel to improve population health is through improvement in trade and trade facilitation. In addition, Sakyi et al. (2018) concluded that trade facilitation through improved institutions, infrastructure, and market efficiency improves educational outcomes and human development. Regarding studies on trade and human development, Hamdi and Hakimi (2021) employed different estimation techniques and found that trade has positive impact on human development for Sub-Saharan Africa countries. Another strand of the literature has considered the relationship between trade facilitation and a number of macroeconomic indicators including GDP, employment and the trade balance. For instance, Narayanan et al. (2016) explored the relationship between trade facilitation and welfare using GDP, employment, and the trade balance as indicators of welfare. The authors concluded that trade facilitation improves these economic indicators. Cabote and Hiwatari (2014) and Nguyen Viet (2015) also considered the association between trade facilitation and income inequality as well as poverty and concluded that trade facilitation contributes to the reduction in income inequality and poverty. From the empirical review above, though studies on the impact of trade facilitation and trade on welfare outcomes has been examined (see: Cabote & Hiwatari, 2014; Nguyen Viet, 2015; Narayanan et al., 2016; Herzer, 2017; Sakyi et al., 2018; Novignon et al., 2018; Byaro et al., 2021; Hamdi & Hakimi, 2021; Zhao et al., 2022), to the best of authors knowledge, there is total death of knowledge on the direct effects of seaport efficiency as well as the interaction effects of seaport efficiency and trade on welfare outcomes in Africa. Thus, this study fills an important research gap by examining the role of seaport efficiency in the relationship between trade and welfare outcomes by focusing on African countries with seaports. 3. Data sources and methodology 3.1. Data sources and measurement Panel data on 28 African countries for the period 2006 to 2018 is employed for analysis. 1 The list of countries used for the paper can be found in Table A1 in the appendix section. A brief summary and definition of how the variables are measured is presented in Table 1. 3.1.1. Seaport efficiency, trade and welfare measures Data on seaport efficiency is proxied by the UNCTAD liner shipping connectivity index (LSCI) (UNCTAD, 2019). The LSCI measures how a country’s seaport is efficient and integrated into the worlds liner shipping connections in terms of maritime connectivity and trade facilitation. The minimum value of the LSCI is 1 while values of 100 and above indicate efficient seaports. Using the LSCI is based on the fact that only efficient seaport can access high capacity and regular international maritime freight transport system (UNCTAD, 2019), implying that such countries can participate effectively in international trade. Regarding trade, we measure it in three ways; exports as share of GDP, imports as share of GDP and overall trade as a share of GDP. With respect to welfare measures, four measures are employed; gross secondary education enrollments, life expectancy, household consumption, and human development. With the exception of data on human development that is obtained from the (UNDP, 2019), data on all other variables are sourced from World Bank (2020b). We present the change in seaport efficiency and measures of trade in Fig. 1 over the period 2006 to 2018. From Fig. 1 it can be seen that over the period 2006 to 2018, substantial number of African countries have experienced larger improvements in seaport efficiency whereas majority of them have seen a smaller change in the level of seaport efficiency and the trade indicators. On the other hand, Fig. 2 which present the scatterplot of seaport efficiency and trade indicators (exports as share of GDP, imports as share of GDP, and overall trade as a share of GDP) on the welfare outcomes (gross secondary education enrollments, life expectancy, household consumption, and human development) clearly shows that improvement in seaport efficiency and trade indicators are strongly correlated with better welfare outcomes. 3.1.2. Control variables Inflation, population density, income, and education were used as control variables. Data on all the control variables were obtained from Table 1 Variable definition and source of data. Variables Notation Definition Source Life expectancy at birth LEXP Life expectancy years (total) at birth World Bank. (2020b) (2020b) Secondary education SEDU Secondary education enrollment, gross World Bank. (2020b) (2020b) Household consumption expenditure HCEP Measured as household consumption expenditure per capita using private consumption, in constant 2010 US$ World Bank. (2020b) (2020b) Human development HDEV Human development index (comprising per capita income, education, and life expectancy measures) HDR UNDP, 2020 Seaport efficiency SEF Measured using liner shipping connectivity index (based on ship calls, physical capacity of seaports, shipping companies, daily total ship services, maximum and average vessel size, ship connectivity to a countries seaport, and direct seaborne shipping connectivity) UNCTAD, 2019 Exports EXP Exports (annual), as a % GDP World Bank. (2020b) (2020b) Imports IMP Imports (annual), as a % GDP World Bank. (2020b) (2020b) Trade TRD Trade (annual), as a % GDP World Bank. (2020b) (2020b) Population density PDEN Number of people leaving per sq.km of land area World Bank. (2020b) (2020b) Income Inc Measured as real GDP per capita in constant 2010 US$ World Bank. (2020b) (2020b) Inflation INF Measured using changes in annual consumer price index World Bank. (2020b) (2020b) Source: Authors. 1 There are 38 African countries with seaports and UNCTAD has data for 33 of these countries for the period 2006 to 2018. However, out of the 33 countries, 28 African countries were used for the analysis. The number of countries included as well as the sample period are influenced by data availability. E. Kojo Ayesu et al. Research in Globalization 5 (2022) 100098 4 the World Bank. (2020b) (2020b). Inflation is expected to have a positive or negative effects on welfare measures since lower or higher inflation can positively or negatively affect the welfare of the populace by increasing or decreasing their purchasing power. We also expect population density to have a positive or negative impact on welfare measures. The reason is that where increase in population density is Fig. 1. Change in seaport efficiency and trade indicators between 2006 and 2018. Note: Countries with bars above (below) zero have experienced improvement (deterioration) in their seaport efficiency and trade indicators between 2006 and 2018. Source: Authors’ construction. Fig. 2. Scatterplots of seaport efficiency, exports (% of GDP), imports (% of GDP), trade (% of GDP) against welfare measures. Note: Line trends are fitted values. Seaports efficiency and trade indicators are normalized between 0 and 1, with higher values implying better seaport efficiency and trade outcomes. Source: Authors’ construction. E. Kojo Ayesu et al. Research in Globalization 5 (2022) 100098 5 accompanied by increment in infrastructure and basic social amenities it may lead to an improvement in welfare outcome of the populace and vice versa. As regard education, it is found to be associated with improvement in measures of welfare (Grossman, 2000; Immurana, Iddrisu, & Boachie, 2021). Thus, education provides individuals with knowledge about health-related issues as well as employment that boost their income and this may enhance indicators of welfare. For instance, if higher levels of education are associated with secured and desirable jobs which are accompanied by higher salaries (income), job satisfaction, increased consumption (human development), and health benefits then education’s effect on welfare outcomes will be positive (Immurana et al., 2021). On the other hand, where higher/increased education is accompanied by no jobs (unemployment) which leads to lower income thereby making people more vulnerable to the rising cost of living then education’s effect on welfare may be negative (Immurana et al., 2021). Also, a positive change in individual’s income increases their consumption on goods and services that improves inputs of welfare (such as education, improve hygiene and sanitation, and medical health care services) and vice versa, hence we expect income to have a positive or negative effect on welfare measures. 3.2. Empirical model The theoretical basis of the study stems from the Heckscher-Ohlin model of international trade. For empirical purposes, the paper specifies an estimable model that relates welfare indicators to seaport efficiency and trade as follows: WFit =θ+δWFit−1+βSEFit +γTRDit + ω (SEFit *TRDit)+∅Zit + μ i+γt+ ε it (1) From Equation (1), i and t represent the cross-section and time dimensions, WFit is a measure of welfare (i.e., life expectancy, gross secondary school enrollment, household consumption, and human development) which is the dependent variable while WFit−1 represents the initial level of welfare - the level of its persistence over time. SEFit and TRDit represent seaport efficiency and trade, respectively. The interaction term, SEFit*TRDit, seeks to measure the extent to which the impact of trade on welfare is enhanced by seaport efficiency. Zit represents a vector of other explanatory variables which include inflation, population density, income, and education. It must be noted that in the case where the dependent variable used is education, we do not control for education. Similarly, we do not control for education in the model where we use human development as the dependent variable as education is included in the human development measure used. The country and time specific effects are denoted by μ i and γt while ε it and δ, β,γ, ω and∅ represent the idiosyncratic stochastic term and the coefficients to be estimated, respectively. In our estimations, we use the logarithm (logs) of variables with the exception of human development because the logarithms help in dealing with outliers in the measurement of variables as well as helps interprets the coefficients as elasticities where appropriate. It is worthy to emphasize that, in Equation (1), the addition of the interactive term shows that the effects of trade on welfare is dependent on some reasonable values of seaport efficiency. This implies that, a statistically significant positive (negative) value suggests that an improvement in seaport efficiency boosts the positive (negative) impact of trade on welfare. Therefore, in Equation (1), ω represent the direct impact of a unit increase in trade on welfare when seaport efficiency is at its mean value. It however remains important to note that, irrespective of how seaport efficiency is measured, it is meaningless to assume seaport efficiency to be equal to zero (absent). This is so because seaport efficiency has non-zero values within the sampled period used in this study. Hence, following the procedure outlined by Brambor, Clark, and Golder (2006) we have mean centered seaport efficiency and trade indicators in other to make the interpretation of the coefficient meaningful and clear. For this reason, when seaport efficiency is at its mean value, the direct impact of a unit increase in trade on welfare is captured by the parameter γ. Based on this outcome, the total conditional marginal effect of trade on welfare as seaport efficiency improves (increases) can be evaluated from Equation (1) by taking the partial differential of welfare with respect to trade to arrive at Equation (2); ∂ WFit ∂ TRDit =γ+ ω *SEFit (2) 3.3. Estimation procedure The study relies on the dynamic system generalized method of moments (system-GMM) estimator in estimating Equation (1) to help in deal with the endogeneity problem (Arellano & Bond, 1991; Roodman, 2009). In addition, we postulate that there is likely to be a reverse causality issue which needs addressing. For example, seaport efficiency and trade is likely to influence welfare measures, in the same way there is also the possibility that measures of welfare can influence trade and seaport efficiency. This is because, as welfare (in terms of good health of the populace) improves, they will be able to trade more, all other things being equal. In addition, as trade increases they are able to earn more income (revenue) hence able to improve on education, consumption, population health and overall human development. Moreover, there is the possibility that some of the independent variables will be endogenous. For instance, income affects trade through an increase in imports of goods and services. Similarly, trade can affect income via higher export of goods and services. Concerning, possible source(s) of endogeneity, the initial level of welfare (WFit−1) introduced in the Equation (1) as explanatory variable leads to an endogeneity issue because it is likely to correlates with the stochastic error term. According to Nickell (1981) using the ordinary least squares estimator in such instances will produce inconsistent and spurious estimates. In this regard, one possible way of addressing the endogeneity issue is the use of reliable instruments. However, finding valid and reliable instruments possess a challenge. According to Arellano and Bond (1991), a valid instrument could be the lag of either the dependent, endogenous or explanatory variables. This paper therefore adopts the twostep system Generalized Method of Moments (system GMM) estimation strategy as it is able to deal adequately with the issues of reverse causality and potential endogeneity. Again, we use the ‘‘collapse’’ routine in Stata to avoid instruments proliferation as the system-GMM is prone to instruments proliferation. This is done to make certain that the number of instruments is less than the number of groups and also ensure that the power of the Hanssen J test is not weakened (Arellano & Bond, 1991; Roodman, 2009). Finally, to verify the validity of our instrument and non-existence of second-order autocorrelation, the study employs the Hansen J test for over-identification restriction (Arellano & Bond, 1991; Roodman, 2009). Concerning the analysis, the long run and short run effects of the explanatory variables are estimated. It is worth stating that with the exception of the coefficient of the initial level of welfare indicators that measures persistence over time, all the parameter estimates measure the short run impacts of the variables used in the analysis. The purpose of estimating the long run results of our explanatory variables is to enable policy recommendations. To achieve this, we divide each coefficient in the short run results by one minus the coefficient of the initial level of welfare indicators (see Papke & Wooldridge, 2005; Egyir, Sakyi, & Baidoo, 2020) with the exception of the initial level of welfare. 4. Results and discussion The study presents and discusses the estimated results in this section. First, we present and discuss the descriptive statistics as well as the long run and short run coefficients of our estimations. Additionally, the E. Kojo Ayesu et al. Research in Globalization 5 (2022) 100098 6 section provides the interpretation of the marginal effect estimates. 4.1. Descriptive statistics Table 2 summarizes the descriptive statistics of all variables. The results show that for the period 2006 to 2018 the averages of life expectancy at birth, secondary education enrollment, household consumption expenditure, and human development for the period 2006 to 2018 was, 61.477, 54.519, 1385.518, and 0.535, respectively. Seaport efficiency averaged 16.358 with the maximum and minimum seaport efficiency found to be 65.041, and 1.588 respectively, indicating extremely low levels of seaport efficiency for some countries. Also, the averages of exports, imports, and trade was 31.167, 42.454, and 73.621, accordingly. Population density averaged 96.146 with a minimum and maximum of 2.394 and 623.302. Regarding inflation, the average was 11.6038 percent whereas the maximum was 264.3751 percent. Finally, the results in Table 2 reveal that income per capita averaged US$ 127.887 dollars with the minimum and maximum income per capita found to be US$ 1.075 dollars US$4000.188 cedis Table 2 Summary statistics. Variables Obs. Mean Std. Dev. Min. Max. LEXP 364 61.477 7.101 45.517 76.693 SEDU 226 54.519 23.474 15.726 109.444 HDEV 364 0.535 0.103 0.354 0.796 HCEP 358 1385.518 1368.636 219.631 7244.181 SEF 364 16.358 [0.000] 11.940 [0.319] 1.588 [-0.527] 65.041 [0.472] EXP 364 31.167 [0.000] 13.885 [0.319] 8.221 [-0.501] 82.446 [0.499] IMP 364 42.454 [0.000] 22.017 [0.318] 10.666 [-0.492] 236.391 [0.508] TRD 364 73.621 [0.000] 31.340 [0.311] 20.722 [-0.492] 311.354 [0.502] PDEN 364 96.146 126.149 2.394 623.302 Inc 364 127.887 415.385 1.075 4000.188 INF 356 116.038 34.401 56.887 264.3751 Note: Centered values in parenthesis. Source: Authors. Fig. 3. Averages of seaport efficiency, trade and human development index performance of African countries for 2006 to 2018. Source: Authors’ construction based on data from UNCTAD, WDI WB, and HDR UNDP. E. Kojo Ayesu et al. Research in Globalization 5 (2022) 100098 7 respectively. In addition, we perform a country-level analysis of average seaport efficiency, trade as a percentage of GDP, and human development index performance for the best (worst) countries for the sampled 28 African countries for the period 2006 to 2018. This analysis can be found in Fig. 3. Concerning seaport efficiency, it can be observed that the five best (worst) performers are Egypt, Morocco, South Africa, Mauritius, and Nigeria (Sierra Leone, Liberia, Congo Democratic Republic, Comoros, and GuineaBissau). This outcome for the best-performing (relatively efficient) countries are not startling given that these regions have larger seaports, high volume of trade, limited navigational challenges, and high direct large vessel calls that serve large catchments (small regional seaports) by transhipping containers and general cargo in smaller vessels (UNCTAD, 2019. Mlambo, 2021) compared to their counterparts (worst performers). For this reasons, they have transformed seaports that relatively reflect the factors that characterized an efficient seaport (Mlambo, 2021). Regarding trade as a percentage of GDP, we find that the relativelyfive best performers within the sampled countries are Liberia, Congo Republic, Mauritius, Tunisia, and Namibia whiles the least performers are The Gambia, Egypt, Tanzania, Comoros, and Nigeria. It is worth noting that, since overall trade as a percentage of GDP is the summation of exports and imports, it is more likely that the countries found in the top five may have higher volumes of imports than exports and hence accounting for their increased volume of trade as % of GDP and vice versa. With respect to human development the overall measure of welfare, we find that the top (least) five countries were Mauritius, Algeria, Tunisia, Egypt and South Africa (Guinea-Bissau, Guinea, Congo Democratic Republic, Mozambique, and Sierra Leone). The reason for the top five performers in the sample is not surprising because according to the World Health Organization (2018) reports these economies have relatively better health care and educational facilities, access to basic lifesaving interventions such as treatment for common childhood diseases that affects young ones, improved standard of living, and high-ratio of doctors to patients among several others comparative to the least performing counterparts. 4.2. The short-and long run results We report in Tables 3 to 6 the short and long run results. In each Table, we report the short and long run estimates as well as three estimation results (models 1 to 3) distinguished by the trade measure (exports, imports or overall trade) used, respectively. These specifications are employed to ensure our results are robustness to alternative measures of trade. Also, and as earlier indicated, the measures of welfare used are secondary education enrollment, life expectancy, household consumption, and human development. Before proceeding to the findings, as indicated earlier, the validity of our estimates depends on model diagnostics. Particularly, it is required that there is the absence of second-order autocorrelation in the error term (see Arellano & Bond, 1991, Roodman, 2009). In doing so, we report (see last part of Tables 3 to 6) the test for second-order autocorrelation by providing the associated p-values. As evident, there is no second-order autocorrelation in all the estimated models. Also, as reported on the last part of Tables 3 to 6, the instruments used are valid and this is established by the outcome of the Hansen J test. The implication is that; policy suggestions can be proposed on the study’s findings since the estimates are valid. The estimated coefficients of the direct effect of seaport efficiency on welfare suggests that seaport efficiency has positive and significant effects on welfare outcomes. Thus, an increase in seaport efficiency triggers positive effects on education, life expectancy, household consumption and human development. The coefficients obtained implies that for every 1 percent increase in seaport efficiency, the various indicators of welfare improve by percentages between 0.2 and 2.4 as reported in Tables 3 to 6. By implication, the beneficial effect of seaport efficiency on welfare, can be seen from the fact that improved seaport Table 3 The effect of seaport efficiency and trade on education in Africa. Short run estimates Long run estimates Model-1 Model-2 Model-3 Model-1 Model-2 Model-3 lnSEDU(-1) 0.241 *** 0.343 *** 0.266 *** – – – (0.036) (0.049) (0.044) lnINF 0.873 *** 0.890 ** 1.003 ** 1.151 *** 1.355 *** 1.367 ** (0.259) (0.331) (0.423) (0.349) (0.514) (0.626) lnINC 0.458 *** 0.485 ** 0.378 0.604 *** 0.738 *** 0.515 (0.130) (0.190) (0.398) (0.191) (0.284) (0.569) lnPDEN 0.323 0.289 0.195 0.426 0.440 0.266 (0.328) (0.322) (0.647) (0.419) (0.481) (0.867) lnSEF 0.767 *** 1.519 *** 1.779 *** 1.011 *** 2.313 *** 2.424 *** (0.231) (0.442) (0.491) (0.310) (0.807) (0.643) lnEXP 0.753 *** 0.993 *** (0.210) (0.275) lnSEF*EXP 0.360 0.475 (0.663) (0.876) lnIMP −0.370 −0.563 (0.585) (0.879) lnSEF*IMP 0.435 0.663 (1.291) (1.944) lnTRD 0.788 ** 1.074 ** (0.317) (0.441) lnSEF*TRD 0.003 0.004 (1.262) (1.719) Constant −4.648 *** −4.640 ** −4.728 ** −6.126 *** −7.0653 ** −6.441 *** (1.214) (2.118) (1.780) (1.507) (3.155) (2.455) No. of observations 336 336 336 No. of groups 28 28 28 No. of instruments 27 25 25 AR(2)[prob] 0.989 0.326 0.995 Hansen[prob] 0.145 0.160 0.0783 Note: Standard errors in parentheses; *** and ** denote 1%, and 5% significance levels, respectively. Source: Authors. E. Kojo Ayesu et al. Research in Globalization 5 (2022) 100098 8 efficiency is an indication of reduction in trade cost, accessibility to global trade, and trade competitiveness (Clark et al., 2004; Blonigen & Wilson, 2008), which would result in increased economic growth, job creation, and reduction in poverty (Clark et al., 2004; Blonigen & Wilson, 2008; Raballand et al., 2012). These therefore bolster people’s ability to afford education, improved human development through the satisfaction of individual needs, spend more on consumption goods as well as affording quality healthcare and other health inputs, that leads to better welfare outcomes. In essence, this ensures seaport efficiency provides an avenue for the government to raise more financial resources that could be used to improve the welfare of the populace. Concerning the direct impact of trade on welfare measures used, the results reveal that trade enhances the various welfare outcomes and this is statistically significant. It is also evident that an increase in exports leads to a positive significant effect on education, household consumption, life expectancy and human development. Also, from the findings, imports contribute to enhancements in life expectancy. As reported on Tables 3 to 6, the magnitude of the coefficients obtained indicate that for every 1 percent increase in trade, the various indicators of welfare improve by percentages between 0.002 and 1.18. The implication of this results is that an increase in trade may boost economic activities, employment and hence income which would make it very easy for people to afford healthcare as well as other health enhancing products, increased investment in quality of education, household consumption, and human development. Also, trade may lead to the transfer of medical equipment and drugs to less developed regions like Africa which may enhance access to quality healthcare. These results confirm the positive significant effect of trade on life expectancy by Levine and Rothman (2006), Owen and Wu (2007), Herzer (2017), and Novignon et al. (2018). Regarding the interaction effects, the results indicate a significant positive effects signifying that seaports efficiency improves the effect of trade indicators on welfare measures. The elasticities range between 0.28 percent and 1.80 percent. This indicates that to enhance the positive effect of trade on welfare, there is the need to improve the efficiency with which trade occurs by improving seaport efficiency. The results imply that seaport efficiency improves the impact of trade on education, life expectancy, household consumption, and human development. Generally, the results from the interaction term mean that seaport efficiency complements the effect of exports, imports, and trade on welfare outcomes in Africa. For this reason and given that there are macroeconomic benefits from trade for African countries as already outlined in this paper, policies aimed directly at enhancing the efficiency of seaports system in Africa will be a step in the right direction. Doing so will help take full complementarity of seaport efficiency on exports, imports, and trade on welfare outcomes in Africa. More importantly, the interaction term results indicate that the coefficients are much larger than the direct impact of trade indicators on welfare measures. This point to the fact that trade impact on welfare measures requires an efficient seaport system. Our results further imply that income appears to be a robust welfare driver, as it ensures improvement in education, life expectancy, household consumption, and human development in both the short run and long run. Second, the effect of population density is found to be inconclusive depending on how welfare is measured. For example, considering gross secondary education enrollment as a measure of welfare, we find population density to have a positive insignificant effects on education. In addition, the results reveal a significant negative effects of population density on human development. This result is not startling, implying that when increasing population density is not accompanied by increase in resources such as health and infrastructure, and social amenities among several others that are welfare improving, this put Table 4 The effect of seaport efficiency, and trade on life expectancy in Africa. Short run estimates Long run estimates Model-1 Model-2 Model-3 Model-1 Model-2 Model-3 lnLEXP(-1) 0.362 *** 0.609 *** 0.376 *** (0.085) (0.084) (0.083) lnINF 0.243 *** 0.144 ** 0.156 *** 0.381 *** 0.369 ** 0.250 *** (0.035) (0.058) (0.033) (0.074) (0.165) (0.082) lnINC 0.186 *** 0.170 *** 0.160 *** 0.291 *** 0.434 *** 0.257 *** (0.037) (0.036) (0.032) (0.068) (0.134) (0.082) lnPDEN −0.112 ** −0.113 ** −0.130 ** −0.175 *** −0.288 ** −0.209 ** (0.041) (0.053) (0.057) (0.063) (0.141) (0.089) lnSEDU −0.002 −0.007 0.012 ** −0.003 −0.019 0.019 ** (0.005) (0.007) (0.005) (0.008) (0.019) (0.008) lnSEF 0.078 ** 0.042 0.125 *** 0.122 ** 0.108 0.201 *** (0.033) (0.054) (0.043) (0.052) (0.144) (0.067) lnEXP 0.374 *** 0.586 *** (0.048) (0.061) lnSEF*EXP 0.335 ** 0.525 ** (0.130) (0.250) lnIMP 0.166 *** 0.425 *** (0.041) (0.094) lnSEF*IMP 0.268 ** 0.686 ** (0.126) (0.342) lnTRD 0.390 *** 0.626 *** (0.071) (0.063) lnSEF*TRD 0.093 0.148 (0.111) (0.182) Constant 0.975 ** 0.668 1.390 ** 1.528 *** 1.706 *** 2.228 *** (0.373) (0.567) (0.600) (0.429) (1.292) (0.727) No. of observations 336 336 336 No. of groups 28 28 28 No. of instruments 26 23 27 AR(2)[prob] 0.497 0.477 0.675 Hansen[prob] 0.360 0.186 0.251 Note: Standard errors in parentheses; ***, **, and * denote 1%, 5%, and 10% significance levels, respectively. Source: Authors. E. Kojo Ayesu et al.