Do trade and foreign direct investment contribute to growth? A study of BIMSTEC nations

Abstract

Purpose

The study examines the association between trade (exports and imports), foreign direct investment (FDI) and economic growth in the Bay of Bengal Initiative for Multi-Sectoral Technical and Economic Cooperation (BIMSTEC) countries using data from 1991 to 2019.

Design/methodology/approach

Augmented Dickey–Fuller (ADF) and Phillips–Perron (PP) unit root tests were applied to check the stationary of the data while the Johansen cointegration test and Vector Error Correction Model (VECM) was used to analyze long-run and short-run relationships.

Findings

The results indicate a long-run relationship between trade, FDI and economic growth in all selected countries except Bhutan. Additionally, a bidirectional causality exists between gross domestic product (GDP) and FDI in India, Bangladesh, Myanmar, Nepal, Bhutan and Sri Lanka, while unidirectional causality from GDP to FDI is observed in Thailand. Moreover, a one-way causality from exports to GDP exists in Bangladesh, Nepal, Bhutan, Sri Lanka and Myanmar, whereas a bidirectional relationship exists in India and Thailand.

Practical implications

This paper will be highly beneficial for regulators and policymakers in the designated economies, aiding in the formulation of FDI and trade policies that promote economic progress and development.

Originality/value

Most previous studies examining the relationship between macroeconomic variables have focused on developed nations. This study is the first to explore the relationship between trade (exports and imports), FDI and economic growth in the BIMSTEC countries.

Citation

Arya, V., Verma, R. and Saini, V.P. (2024), "Do trade and foreign direct investment contribute to growth? A study of BIMSTEC nations", Journal of Business and Socio-economic Development, Vol. ahead-of-print No. ahead-of-print. https://doi.org/10.1108/JBSED-09-2022-0100

1. Introduction

The economic growth of a country is affected by many factors, with foreign direct investment (FDI) and trade being among the primary drivers. FDI is regarded as a crucial basis for economic progress, augmenting technology, enhancing trade, creating job opportunities and integrating countries into the global economy (Mohanty and Sethi, 2019; Arya and Singh, 2022; Ravinder et al., 2022a; Singh, 2022; Singh et al., 2023). There is growing consensus among economists that FDI significantly benefits both advanced and emerging economies (Hussain et al., 2021; Saini et al., 2021; Vandana and Singh, 2023). According to the Keynesian model, FDI assists in boosting consumption, which is a crucial driver of development. Since the 1970s, FDI has risen substantially worldwide. Rehman (2016) and Singh (2022) stated that although some nations efficiently utilize FDI to accelerate economic growth, others cannot. Chowdhury and Mavrotas (2005) contended that the fundamentals of an economy are more important than the specific investment policies. Sethi et al. (2020) provided that the impact of FDI on economic progress depends on the varying levels of absorption capabilities of different countries. Thus, the repercussions of foreign inflows on growth theoretically depend on several factors, including their motive, type and the host country’s economic structure. Given the differences across nations and the diverse motives and types of inflows, various researchers have reported with varied results (Gunaydin and Tatoglu, 2005; Har et al., 2008; Gürsoy et al., 2013; Rehman, 2016; Bermejo Carbonell and Werner, 2018; Sethi et al., 2020; Hussain et al., 2021; Saini et al., 2021; Zardoub, 2021; Saini and Ravinder, 2022; Ravinder et al., 2022b; Singh et al., 2023; Yadav et al., 2023, 2024).

The importance of trade in enhancing economic growth has also been an area of extensive academic interest. An increase in exports contributes to an economy’s growth by increasing productivity, offering greater economies of scale and thus, providing access to global markets more liberally. Earlier studies have confirmed empirical and econometrical connections between exports and economic growth (Ho and Iyke, 2020; Bruns and Ioannidis, 2020; Saini et al., 2021; Saini and Ravinder, 2022; Ravinder et al., 2022c). Likewise, the import of advanced equipment and techniques enables domestic firms to be more competitive in the international market.

Trade has traditionally been the primary mechanism connecting national boundaries to create a global economy. FDI also creates connectivity across countries, and these two mechanisms reinforce each other. The trade consequences of FDI vary depending on whether the purpose of the investment is to access natural resources and consumers or exploit locational comparative advantage. Robert Solow’s 1956 neoclassical growth model theorizes that FDI inflows assist capital-deficient nations by providing funds that contribute to the accumulation of physical investment and gross savings in nations (Baldé, 2011; Ravinder et al., 2022d; Singh et al., 2023; Vandana and Singh, 2023; Yadav et al., 2024).

FDI also facilitates the import of advanced technology in technologically lagging countries, and through this import and the use of advanced industrial technologies and innovation, trade assists in upgrading skills. Exporters and domestic manufacturers use innovation and sophisticated manufacturing techniques to work as subcontractors for foreign companies or as contestants in the international market (Bruns and Ioannidis, 2020; Ravinder et al., 2022c). FDI grew from US$12.36 bn in 1970 to US$1.54 tri in 2019, and the FDI inflows have increased 124 times compared to 1970. Most South Asian nations have received enormous capital inflows in recent decades due to their rapid economic expansion and consistent export performance (Sahoo and Sethi, 2020; Saini et al., 2021; Ravinder et al., 2022b; Singh et al., 2023; Yadav et al., 2023, 2024). Thus, the nexus between FDI, trade and economic growth and the direction of causation among these variables is critically important, particularly for developing and lower-income nations.

Bay of Bengal Initiative for Multi-Sectoral Technical and Economic Cooperation (BIMSTEC) was formed on June 6, 1997, and includes seven members from South Asia: Bangladesh, Bhutan, India, Nepal, Sri Lanka, Myanmar and Thailand. BIMSTEC member countries are home to around 1.5 bn people and have a total gross domestic product (GDP) of 7.7 tri dollars.

As shown in Table 1, FDI inflows in BIMSTEC countries have increased significantly over the past three decades. For instance, India received $73.53 m FDI in 1991, rising to $50610.7 m in 2019. Similarly, FDI inflows have grown in other BIMSTEC countries: Bangladesh (from $1.30 m in 1991 to $1908.05 m in 2019), Bhutan (from $0.6 m in 1991 to $13.03 m in 2019), Myanmar (from $863 m in 2008 to $2292.33 m in 2019), Nepal (from $2.22 m in 1991 to 185.5 m in 2019) and Thailand (from $2013.9 m in 1991–4816.64 m in 2019). Likewise, per capita GDP has also increased significantly in the BIMSTEC countries. Moreover, international trade, represented by exports and imports, shows a consistent uptrend, indicating a rise in trade activities in the BIMSTEC region.

The structure of the remaining section of this paper is as follows: Section 2 reviews earlier relevant studies. Section 3 provides the conceptual framework and rationale of the study. Section 4 describes the methodology employed and specifies the data. The analytical results are discussed in Section 5. The conclusion and policy suggestions are presented in the final section.

2. Review of literature

Foreign direct investment has poured into emerging nations over the past few decades. The theoretical impact of FDI inflows on growth depends on several factors, including their motive, type and the host country’s economic structure and variations across nations. Empirical research on the growth impacts of foreign capital inflow has produced mixed results. Consequently, extensive research has been conducted on the influence of FDI in boosting economic growth. Although findings on the growth impact of FDI are disputed, some empirical studies suggested that FDI positively influences economic progress (Gunaydin and Tatoglu, 2005; Har et al., 2008; Gürsoy et al., 2013; Sethi et al., 2020; Saini et al., 2021; Ravinder et al., 2022d; Saini and Ravinder, 2022; Yadav et al., 2023). However, other studies did not support this view (Rehman, 2016; Zardoub, 2021). Additionally, some research found no relationship between FDI and economic growth (Bermejo Carbonell and Werner, 2018; Hussain et al., 2021). Thus, no consistent statistical evidence supports the development impact of FDI.

Gunaydin and Tatoglu (2005) studied the directional connection between FDI and economic progress using data from 1968 to 2002. Their Johansen cointegration test indicated a long-term cointegrating relationship between the two variables and confirmed a two-way causality between FDI and GDP. Gürsoy et al. (2013) examined the effect of FDI on economic expansion in Azerbaijan, Kazakhstan, Tajikistan, Turkmenistan and Uzbekistan from 1997 to 2010 using Johansen cointegration and found a beneficial connection between FDI and GDP. Har et al. (2008) explored the nexus between FDI and economic growth in Malaysia utilizing Ordinary Least Squares (OLS) over the 1970–2005 periods and found a substantial relationship between FDI and economic advancement. Das and Sethi (2020) argued that FDI and remittances significantly influenced growth in India from 1980 to 2016. Tahir et al. (2020) highlighted the contributory role of foreign inflows on economic growth in Pakistan from 1976 to 2008 using Autoregressive Distributed Lag (ARDL) and found evidence for the FDI-led growth hypothesis. Rehman (2016) investigated the association between FDI and economic progress in Pakistan from 1970 to 2012 using the Vector Error Correction Model (VECM), concluding that FDI is largely influenced by economic growth. Additionally, FDI, human capital and exports are found to be critical for economic progress.

Conversely, Sahoo and Sethi (2017) used yearly data from 1990–1991 to 2013–2014 to examine the long-run relationship between foreign inflows and economic progress in India and found significantly negative repercussions of FDI on economic development in India. Bhujbal and Sethi (2020) found a one-way causality from growth to FDI in the South Asian Association for Regional Cooperation (SAARC) countries.

The connection between trade and economic growth can be traced back to the periods of Adam Smith and Karl Marx. Grossman and Helpman (1991) argued that openness promotes economic growth in numerous ways: it increases the diversity of available goods and heavy equipment, enhances the productivity of other resources and improves access to techniques from advanced nations, leading to an efficient absorption of capital and a broader market for domestic companies. However, previous analytical works on the influence of trade on economic progress have revealed mixed results. For example, several scholars found that trade has a substantial positive impact on growth (Andraz and Rodrigues, 2010; Mukhtar and Rasheed, 2010; Dritsaki and Dritsaki, 2012; Dritsaki and Stiakakis, 2014; Singh and Kumar, 2020; Saini et al., 2021; Saini and Ravinder, 2022; Singh et al., 2023; Vandana and Singh, 2023; Yadav et al., 2024), while others suggested negative impact (Burange et al., 2019; Tahir et al., 2019). These conflicting findings imply that further research is needed, as the trade–growth nexus remains inconclusive.

Burange et al. (2019) investigated the association among exports, imports and economic progress for the BRICS (Brazil, Russia, India, China and South Africa) countries using the Johansen test and VECM over the period 1989–2013. They found a one-way causality from economic growth to export, supporting the growth-led trade hypothesis in BRICS nations. Andraz and Rodrigues (2010) studied the connection between exports, foreign inflows and economic growth in Portugal using three-stage movement and causality methods with data from 1977 to 2004 and found a positive relationship among exports, FDI and economic growth. Mohanty and Sethi (2021) indicated the insignificant adverse repercussions of FDI on real exports. Singh and Kumar (2020) observed a long-run association between India’s GDP, exports and imports using the Johansen cointegrating method with data from 1995 to 2018, reporting a long-run cointegrating nexus among the variables. They also found a feedback connection between exports and GDP and a one-way relation from growth to imports of the nation. Moreover, Dritsaki and Stiakakis (2014) examined the relationship and association between FDI, exports and economic progress in Croatia using annual data from 1994 to 2012 and employing the ARDL methodology and found a two-way nexus between exports and the economic progress.

Thus, the effects of FDI and trade are positive on economic progress in some developing nations (Vandana and Singh, 2023; Singh et al., 2023; Singh, 2022; Saleem et al., 2020; Andraz and Rodrigues, 2010), while an adverse relationship between FDI, trade and GDP is found in other countries (Burange et al., 2019; Tahir et al., 2019). These mixed empirical findings highlight the need to further investigate the exact nexus among these variables and their contribution to the country’s growth. From the literature, it is concluded that the relationship between trade, FDI and economic growth remains indecisive. Consequently, researchers are keen to understand the relationship and direction of causation among FDI, trade and economic progress. Therefore, the current study aims to explore the effects of FDI and trade on the economic growth of seven BIMSTEC nations from 1991 to 2019.

3. Conceptual framework and rationale of the study

3.1 Conceptual framework

Despite the availability of literature on the growth-enhancing impact of FDI and trade, there are several gaps in existing research. Previous studies are mainly focused on understanding this link between FDI and economic growth or trade and growth in other regional groups. However, the results are inconsistent, with some studies supporting a positive effect of FDI on economic progress (Gunaydin and Tatoglu, 2005; Har et al., 2008; Gürsoy et al., 2013; Sethi et al., 2020) while others did not (Rehman, 2016; Zardoub, 2021; Singh, 2022; Singh et al., 2023; Vandana and Singh, 2023). Similarly, some studies found no relation between FDI and economic growth (Bermejo Carbonell and Werner, 2018; Hussain et al., 2021).

Similarly, the impact of trade on economic growth has also yielded mixed results. While some studies found a substantial influence of trade on growth (Andraz and Rodrigues, 2010; Mukhtar and Rasheed, 2010; Dritsaki and Dritsaki, 2012; Dritsaki and Stiakakis, 2014; Singh and Kumar, 2020), others found negative impact (Burange et al., 2019; Tahir et al., 2019). Thus, further research is needed to fully understand the trade–growth nexus. Moreover, the impact of FDI and trade on economic growth varies across developing nations, with some experiencing the positive effects (Saleem et al., 2020; Andraz and Rodrigues, 2010) and others showing adverse relationships (Burange et al., 2019; Tahir et al., 2019). These mixed empirical findings highlight the need to explore the exact relationship between these variables and their contribution to a country’s growth. Furthermore, there is a lack of research on this topic, specifically for BIMSTEC countries.

To address this gap, our study explores the relationship between FDI, trade and economic growth in all member countries of the BIMSTEC regional group. Table 3 provides a description of the variables studied.

3.2 Justification of the study

The BIMSTEC countries, like many other emerging economies, have been striving to achieve and sustain long-term economic growth since their liberalization (Sahoo and Sethi, 2020). Since then, FDI inflows in BIMSTEC countries have significantly increased, although not all countries have seen the same level of growth due to their inability to attract larger foreign flows. Further, imports and exports have also been on the rise in the BIMSTEC region.

More recently, in the 17th BIMSTEC Ministerial meeting held on April 01, 2021, it was suggested that BIMSTEC will take lead in trade, investment and development in the coming year. This was further emphasized in the 17th BIMSTEC Ministerial Meeting held in Colombo, Sri Lanka, on 01 April 2021. Furthermore [1], the focus was on enhancing the interconnectivity to enhance the trade integration and facilitate the cross-country movement of people and goods. Furthermore, in a virtual meeting organized jointly by [2] BIMSTEC members and the Asian Development Bank (ADB) on September 14, 2021, titled “BIMSTEC Trade Facilitation Strategic Framework 2030,” all members highlighted the need for joint and coordinated strategies to facilitate trade and fund movement. These developments demonstrate that BIMSTEC is becoming a more prominent regional trade bloc, with member countries making prominent progress in foreign inflow and trade. Furthermore, all member countries are making changes to their foreign policies to liberalize their economies.

In light of these developments, this study’s aim is to examine the relationship between trade, FDI and economic growth in the BIMSTEC countries. To the best of the author’s knowledge, there has been no previous study that simultaneously examined the relationship between all four variables in the case of BIMSTEC countries. While many studies explored the association between FDI and economic progress (Gunaydin and Tatoglu, 2005; Har et al., 2008; Gürsoy et al., 2013; Rehman, 2016; Bermejo Carbonell and Werner, 2018; Sethi et al., 2020; Hussain et al., 2021; Zardoub, 2021; Saini et al., 2021; Ravinder et al., 2022b; Singh, 2022; Singh et al., 2023) and between export/imports and GDP (Andraz and Rodrigues, 2010; Dritsaki and Stiakakis, 2014; Burange et al., 2019; Singh and Kumar, 2020; Vandana and Singh, 2023; Yadav et al., 2024) in various regional groups in the world, none of the studies have specifically explored the relationship between trade, FDI and economic growth simultaneously in BIMSTEC countries. Thus, this study fills this gap and provides unique insights.

Furthermore, this study includes all member countries of the BIMSTEC region, making its findings widely applicable. Additionally, the data used in the study is from before the COVID-19 pandemic, making it a useful benchmark for policymakers to compare post-pandemic performance. Moreover, the study utilizes econometric techniques, i.e. Johansen cointegration and VECM methodology, to observe the cointegrating relation between the variables, providing efficient results. The study also employs the ARDL methodology to test the robustness of the results. Finally, this study provides a causal nexus between the studied variables for each country separately, providing valuable insights for policymakers.

4. Data and methodology

4.1 Data

The aim of the current study is to explore the impact of FDI and trade on the economic growth of seven BIMSTEC nations. The analysis focuses on the following variables: Foreign Direct Investment net inflow, exports and imports of goods and services and per capita GDP. Per capita GDP is used as a measure of economic growth because it provides a comparative measure of GDP that accounts for the population size. The study uses annual data from World Development Indictors (WDI) provided by the World Bank covering the period from 1991 to 2019.

Foreign Direct Investment net inflow (Current US$) is used as a proxy for FDI (Gunaydin and Tatoglu, 2005; Gürsoy et al., 2013; Rehman, 2016; Bermejo Carbonell and Werner, 2018; Sethi et al., 2020; Hussain et al., 2021; Zardoub, 2021). Further, Import and Exports (Current US$) is representing the trade (Ho and Iyke, 2020; Bruns and Ioannidis, 2020). Moreover, Per Capita GDP (Current US$) is representing the Economic growth (Rehman, 2016; Bermejo Carbonell and Werner, 2018; Sethi et al., 2020; Hussain et al., 2021; Zardoub, 2021; Ho and Iyke, 2020; Bruns and Ioannidis, 2020) (Table 2).

4.2 Methodology

The study estimates the following model, based on the Solow neoclassical growth model, to examine the association between FDI, trade and economic growth in the BIMSTEC countries:

GDP=f (FDI,Exp,Imp)

where FDI depicts the foreign direct investment, Exp indicates exports and Imp signifies imports. The research uses the following equation for the study:

(1)GDPt=αi+β1FDIt+β2Expt+β3Impt+εt

To examine the long-term equilibrium and short-term dynamic relationship between FDI, trade and economic progress in the BIMSTEC countries, this paper utilizes time series methods, specifically the Johansen cointegration test and the VECM.

Srinivasan et al. (2011) utilized time series techniques, specifically the VECM, to observe the underlying association between FDI and economic growth in SAARC countries from 1970 to 2007. Similarly, Saleem et al. (2021) employed the time-series technique of ARDL to examine the relationship between FDI, GDP and trade liberalization in five South Asian nations from 1975 to 2016. Hossain and Hossain (2012) used time-series techniques to study the co-integrating nexus between FDI and GDP in Bangladesh, Pakistan and India from 1972 to 2008. Likewise, Gürsoy et al. (2013) explored the impact of FDI on output growth in East Asian nations from 1997 to 2010 using Johansen cointegration and Granger causality testing.

Following previous literature, we employed time series techniques to examine the long-term equilibrium and short-term dynamic relationship between FDI, trade and economic growth in the BIMSTEC countries. Time series data often exhibit integrating effects that can lead to “pseudo-regression.” To determine whether the series contains a unit root or not, we used the augmented Dickey–Fuller Test (Dickey and Fuller, 1979) as follows:

(2)ΔYt=α0+βt+α1 Yt−1+Σ γj ΔYt−j+εt

where,

• yt = the time series being considered

• t = a time trend term

• εt = white noise error term.

A data series is considered stationary when its mean and variances do not fluctuate over time, and the covariance between two time periods depends only on the lag between them, not on the specific times at which the covariance is computed. Further, Engle and Granger (1991) suggested that when two variables are stationary at the same order, they may exhibit a long-term stable relationship, known as a cointegrating relation. To examine the long-term connection between the variables integrated at the same order (i.e. stationary at first differencing); we use the Johansen cointegration approach. This approach has several advantages over other cointegration methods, such as the Engle–Granger method. It can be utilized when two or more variables are integrated at the same order (i.e. first order), providing a higher degree of accuracy. Unlike the Engle–Granger method, the Johansen test does not require selecting a dependent variable and carrying forward residuals to the next step. It can detect multiple cointegration relationships, making it more suitable for multivariable analysis. Another benefit of the Johansen test is that it treats all variables as endogenous (Andraz and Rodrigues, 2010; Gunaydin and Tatoglu, 2005; Srinivasan et al., 2011).

To further elaborate on the multiple variable models for short- and long-term equilibrium relations, we apply the VECM method as follows:

(3)ΔGDPt=α1+∑i=1p−1βiΔFDIt−i+∑i=1p−1γiΔGDPt−i+∑i=1p−1δiΔIMPt−i+∑i=1p−1ζiΔEXPt−i+η1ECMt−1+νt1

where GDP is the explained variable while FDI, IMP and EXP are the explanatory variables, i represents the optimal lag for the model, and ECM stands for error correction term. To observe the lead-lag relationship among the variables in the short run, we use the Wald test (Chi-Square), which tests the hypothesis that the parameters of the independent variable are zero (H0: βi=0).

To validate the model’s findings, we conduct diagnostic and robustness testing. We apply the Breusch–Godfrey LM test for autocorrelation, the Jarque–Bera test of normality, and the White test for heteroscedasticity. To further examine the robustness of the model, we employ the ARDL methodology. The ARDL approach is used because it is a cointegrating technique that can be applied regardless of the order of the integration of the variables. Additionally, it takes into account lagged values of the variables, addressing issues for endogeneity and multicollinearity.

The null and the alternative hypotheses under the estimated ARDL model are:

H0.

δ₁ = δ₂ = δ₃ (no cointegrating connection)

Ha.

δ₁≠ δ₂≠ δ₃ (cointegrating connection)

If the combined significance F-statistic exceeds the critical value determined by Narayan (2004), the null hypothesis is rejected, indicating cointegration among the variables. If the F-statistic falls between the upper and lower bounds, the decision is inconclusive. This suggests that if the F-statistics surpasses the critical value, the variables are considered cointegrated.

5. Empirical results

This section presents the empirical findings from summary statistics, unit root tests, the Johansen cointegration test and the VECM.

Table 3 provides the descriptive statistics of all variables of the BIMSTEC countries. As shown in the table, the average per capita GDP is the highest in Thailand (US$3897), followed by Sri Lanka (US$1925), Bhutan (US$1566.6) and India (US$938). The average per capita GDP is the lowest in Nepal among the BIMSTEC countries. India is the largest recipient of FDI (US$50610bn), while Bhutan received the least FDI (US$16.6bn) among all the countries over the period from 1991 to 2019. Thailand’s exports exceed its imports, whereas for the remaining countries, average imports exceed exports. Large variations are observed in the imports and exports of Thailand, followed by India and Myanmar. All macroeconomic variables series, except for Myanmar’s GDP, are positively skewed, implying that these variables are vulnerable to large shocks such as the 2007 Black Swan crisis. All macroeconomic series except for Bangladesh’s GDP, Nepal’s FDI and imports and Bhutan’s FDI are normally distributed as the p-values of Jarque-Bera statistics for these series are larger than 5%.

Unit root testing is conducted to examine the presence of a unit root in the series, with the null hypothesis being no unit root and the alternate hypothesis being the presence of a unit root in the variables. The results of the unit root testing for the series of GDP, FDI and exports and imports series of BIMSTEC countries are shown in the table. All the variables are non-stationary at the level, since the t-values of all the macroeconomic variables are less than the level of significance. Also, all the variables are integrated at the first difference, as all the values are significant at the 5% level of significance (Table 4).

Since all the variables are stationary at first order, there may be cointegrating relationships. Thus, to examine the long-run cointegrating nexus among the studied variables, the Johansen cointegration test is employed in the current study. The elementary and necessary condition for applying this testing is that all the variables are constant in the same order. All the macroeconomic variables of BIMSTEC countries are integrated at the first difference, making the Johansen test suitable for exploring the long-run association among these variables. Table 5 presents Johansen’s maximum eigen and trace statistics for every BIMSTEC nation: Bangladesh, India, Myanmar, Nepal, Bhutan, Sri Lanka and Thailand. Based on eigenvalue and trace statistics, the hypothesis of no cointegrating vector (r = 0) is rejected for all countries, except Bhutan. In Bhutan, no cointegration is found among FDI, trade and economic growth. This is likely due to the lack of long-run data for the variables studied in this country. Dhume (2020) also argued that the non-availability of FDI data for the year before 2002 is the main reason behind the absence of a long-run relationship between FDI, trade and economic growth in Bhutan.

Consequently, the results support the theoretical proposition of a cointegrating nexus between FDI inflows, import–export and economic growth, indicating that there are stable long-run associations among these four variables in all BIMSTEC countries, except Bhutan (Table 5).

Having established the presence of a cointegrating relationship among foreign inflows, imports, exports and economic progress in the BMISTEC nations (except for Bhutan), the study extends to examine the long- and short-run coefficients using the VECM. This model helps to further assess the speed of adjustment of the variable. The VECM model’s estimated results are presented in Table 6. Since the estimated values are sensitive to the selected lag length, the optimal lag length is determined based on the Schwarz Information Criteria (SIC) and is selected to be one for all the chosen countries.

The results illustrate that the error correction term is negative and statistically significant, indicating the robustness of the long-run equilibrium relationship among FDI inflows, trade and GDP in all the countries.

Additionally, the findings reveal a two-way relationship between GDP and FDI in India, Bangladesh, Nepal, Bhutan and Sri Lanka, supporting the findings of Srinivasan et al. (2011). Myanmar also exhibits a two-way causation between FDI and GDP, while in Thailand; there is a one-way causality from GDP to FDI.

FDI is a crucial element for economic growth due to its role in enhancing forex reserves, job prospects and more. Our results regarding the positive association between FDI and economic growth align with Gunaydin and Tatoglu (2005), who confirmed the two-way causality between FDI and GDP.

Furthermore, our results are consistent with Gürsoy et al. (2013), who found a feedback relationship between FDI and GDP, and similar to the findings of Har et al. (2008), Ehigiamusoe and Lean (2019), Tahir et al. (2020) and Sethi et al. (2020). These studies argued that foreign inflows contribute to the development of evolving domestic markets, leading to more efficient utilization of technical know-how and the creation of innovative products. They also suggest that FDI brings advanced technology, ultimately boosting efficiency and effectively meeting consumer demand. Tahir et al. (2020) note that emerging nations often lack sufficient national capital necessary for progress and external capital complement to the domestic capital. Har et al. (2008) argued that external investors can create monopolies, making it difficult for the local firms to survive, a sentiment echoed by Zardoub (2021), who confirmed that FDI inflow might lead to the closure of small firms due to the tough competition from multinationals. Our results differ from Rehman (2016), who found a one-way association between FDI and economic growth, indicating that FDI contributes to economic development but vice versa. Likewise, our results are in contrast with Bermejo Carbonell and Werner (2018), Hussain et al. (2021), Singh (2022), Singh et al. (2023) and Vandana and Singh (2023), who supported the impartial hypothesis, indicating no impact of FDI on economic growth. Moreover, our results differ from Adams and Elassal (2020) and Zardoub (2021), who reported a detrimental impact of FDI on economic growth.

The empirical results also suggest a unidirectional causation from exports to GDP in Bangladesh, Nepal, Myanmar and Sri Lanka, while a two-way causality is found in India and Thailand, the major exporting countries in BIMSTEC. There is no causality between GDP and imports in all the BIMSTEC nations.

Our findings on one-way causality from exports to GDP in Bangladesh, Nepal, Myanmar and Sri Lanka differ from Burange et al.’s (2019), who revealed one-way causality from economic growth to exports. However, our results on the two-way causality between exports and economic progress in India and Thailand are in line with the findings of Saleem et al. (2020), Singh (2022) and Singh et al. (2023) for India, Bangladesh, Nepal and Sri Lanka. Andraz and Rodrigues (2010), Dritsaki and Dritsaki (2012) and Dritsaki and Stiakakis (2014) confirmed the positive and two-way nexus between exports and economic growth. However, Andraz and Rodrigues (2010) argued that increasing exports does not always equate to national progress, as it depends on the quality and diversity of the traded goods. Our results partially align with Mukhtar and Rasheed (2010), who revealed the two-way causality between exports and imports, and Singh and Kumar (2020), who found a feedback relationship between exports and GDP.

In Bhutan, no cointegration is found between FDI, trade and economic growth. The likely cause for this lack of long-term cointegration is the non-availability of FDI data for Bhutan. Dhume (2020) also argued that the absence of FDI data before 2002 is the main reason behind the lack of a long-run relationship between FDI, trade and economic growth. Furthermore, Bhutan’s ranking in Ease of Doing Business has fallen from 71 in 2017 to 89 in 2020, indicating a less favorable investment environment in the country.

The empirical results regarding the relationship between FDI, exports, imports and GDP in each BIMSTEC nation are summarized in Table 7. Overall, our findings support the FDI-led growth hypothesis and Solow’s new classical theory, which states that increased FDI contributes positively to economic progress and enhances exports. Thus, economic growth ultimately creates a conducive environment for further inflows of foreign resources. Finally, trade contributes positively to the economic growth of these countries.

Finally, to confirm the validity of the results, we also conduct diagnostic testing and robustness testing. We apply the Breusch–Godfrey LM test for autocorrelation, the Jarque–Bera test of normality, and the White test for heteroscedasticity. The null hypothesis of no autocorrelation and normality is not rejected, suggesting that the models are free from these biases. Thus, the estimated results are valid (Table 8).

To examine the robustness of the results, we have also applied the ARDL methodology. ARDL is applicable regardless of the integration order of the variables and accounts for the lagged values of both dependent and independent variables, thus addressing endogeneity and multicollinearity issues.

If the joint significance F-statistic exceeds the critical value determined by Narayan (2004), the null hypothesis is rejected; otherwise, it is not. However, if the F-statistic lies between the upper and lower bounds, the conclusion is indecisive. Table 9 presents the results of the ARDL methodology, which confirms the findings of the Johansen cointegration technology. This suggests that FDI, trade and economic growth are cointegrated in BIMSTEC countries.

6. Conclusion

The study examines the association between trade (exports and imports), FDI inflows and economic progress in the BIMSTEC countries. The research is based on annualized data from 1991 to 2019. Stationarity of the annualized data were tested using the ADF and PP unit root test. For analyzing long-term relationships, the Johansen cointegration test and VECM were used. The results suggest that the error correction coefficient is negative and statistically significant, inferring the validity of the long-run equilibrium relationship among FDI, trade and GDP in all countries, except Bhutan. Additionally, the results reveal a two-way relationship between economic progress and FDI inflows in India, Bangladesh, Nepal, Bhutan, and Sri Lanka, supporting the findings of Srinivasan et al. (2011). There is also a two-way causation between FDI inflows and economic progress in Myanmar, representing a new finding. Furthermore, there is a one-way causality from GDP to FDI in Thailand.

On the other hand, there is a unidirectional causality from exports to GDP in Bangladesh, Nepal, Bhutan, Sri Lanka and Myanmar, with a bidirectional causality in India and Thailand. India and Thailand are two major exporting countries in the BIMSTEC region. Furthermore, no causality is found between Imports and GDP in BIMSTEC nations.

These findings carry significant policy implications for the BIMSTEC region and individual nations. Specifically, economic development is found to Granger-cause other macro variables, including FDI, imports and exports in most nations. As a result, BIMSTEC officials should concentrate on policies and tactics that support economic growth. Furthermore, the current analysis shows that liberalizing FDI-oriented policies and improving countries economic development performance are critical for attracting foreign direct investment flows in the BIMSTEC region. Furthermore, BIMSTEC governments, particularly Myanmar, should focus on boosting global trade to increase national economic growth. Countries like Myanmar and Bhutan should focus more on their competitive advantages in sectors like travel and tourism, while other countries should capitalize on trade in specific industries to further benefit from global trade. Some countries in the BIMSTEC region attract less FDI, potentially due to lower infrastructural development, less skilled labor and restrictive foreign policies. Therefore, governments are recommended to efficiently invest in infrastructure, human capital development and financial resources to further maximize foreign inflows. Measures to facilitate businesses, such as streamlining processes for conducting business in India, extending the function of investment promotion agencies (IPAs) and designating single investment portals rather than myriad rules, should be prioritized.

Further studies could consider asymmetric and threshold relationships among variables. Additionally, incorporating more macroeconomic variables such as financial development, technological development and their interactive effects on trade and FDI and their contribution to national economic growth should be investigated.

Table 1

Economic indicators

Particulars	1991	1998	2008	2019
Bangladesh
GDP	293.16	407.429	634.987	1855.74
FDI$	1.39044	190.059	1328.42	1908.05
EXP$	2062.58	5876.85	16,181	46363.7
IMP$	3785.24	8058.88	22873.1	64859.3
Bhutan
GDP	467.697	667.914	1828.15	3316.18
FDI$	0.6		3.144	13.0324
EXP$	80.4103	124.772	613.944	860.578
IMP$	100.325	186.282	729.319	1274.16
India
GDP	303.056	413.299	998.522	2099.6
FDI$	73.5376	2634.65	43406.3	50610.7
EXP$	22943.4	46426.5	288,902	528,298
IMP$	22941.4	53431.6	350,927	606,366
Myanmar
GDP	–	–	638.149	1407.81
FDI$	–	–	863.88	2292.33
EXP$	–	–	37.1378	23147.3
IMP$	–	–	21.2037	23078.3
Nepal
GDP	202.081	210.612	470.456	1071.05
FDI$	2.22	12.0247	0.99512	185.555
EXP$	450.607	1108.3	1602.78	2656.09
IMP$	909.167	1645.67	4172.66	14,174
Thailand
GDP	1716.42	1845.83	4379.66	7806.74
FDI$	2013.99	7314.81	8561.56	4816.64
EXP$	35329.4	65860.6	208,095	324,875
IMP$	41756.5	48088.4	201,114	275,171

Source(s): WDI

Table 2

Description of the variables

Variable name	Variable description	Studies
Foreign direct investment	Foreign Direct Investment net inflow (Current US$)	Gunaydin and Tatoglu (2005), Gürsoy et al. (2013), Rehman (2016), Bermejo Carbonell and Werner (2018), Sethi et al. (2020), Hussain et al. (2021), Zardoub (2021)
Trade	Import, Exports (Current US$)	Ho and Iyke (2020), Bruns and Ioannidis (2020)
Economic growth	Per Capita GDP (Current US$)	Rehman (2016), Bermejo Carbonell and Werner (2018), Sethi et al. (2020), Hussain et al. (2021), Zardoub (2021), Ho and Iyke (2020), Bruns and Ioannidis (2020)

Source(s): The authors

Table 3

Summary statistics

	Var	Mean	Minimum	Maximum	Std. Dev	Skew ness	Kurtosis	Jarque-Bera	Prob
Bangladesh	GDP	710.2	293.2	1855.7	454.7	1.19	3.26	6.98	0.03
	FDI	909.0	1.4	2831.2	953.3	0.73	2.08	3.56	0.17
	EXP	16135.8	2062.6	46363.7	13630.6	0.78	2.21	3.73	0.16
	IMP	22594.4	3785.2	64859.3	19057.9	0.89	2.48	4.12	0.13
India	GDP	938.3	301.2	2099.6	598.4	0.56	1.87	3.05	0.22
	FDI	18660.5	73.5	50610.7	17708.6	0.39	1.52	3.40	0.18
	EXP	226005.9	22943.4	538635.2	191432.1	0.37	1.47	3.49	0.17
	IMP	263434.8	22941.4	639013.3	226681.6	0.36	1.46	3.49	0.17
Nepal	GDP	451.7	170.6	1071.1	284.6	0.78	2.26	3.62	0.16
	FDI	38.5	−6.6	196.3	54.5	1.60	4.87	16.58	0.00
	EXP	1464.1	450.6	2656.1	604.0	0.45	2.36	1.47	0.48
	IMP	4417.2	875.3	14174.0	3813.7	1.17	3.37	6.84	0.03
Bhutan	GDP	1566.6	441.7	3316.2	1007.9	0.44	1.67	3.09	0.21
	FDI	13.8	−16.6	75.3	21.8	1.85	6.01	21.76	0.00
	EXP	399.3	74.3	860.6	290.1	0.14	1.23	3.87	0.14
	IMP	632.4	100.3	1367.5	482.6	0.33	1.43	3.50	0.17
Thailand	GDP	3879.6	1716.4	7806.7	1897.3	0.54	1.89	2.90	0.23
	FDI	6288.1	1366.4	15936.0	4030.1	0.86	2.92	3.59	0.17
	EXP	160769.9	35329.4	328570.0	100538.7	0.34	1.53	3.17	0.20
	IMP	146888.0	41756.5	283801.7	87141.3	0.31	1.49	3.23	0.20
Sri Lanka	GDP	1925.7	513.3	4080.6	1349.4	0.53	1.57	3.85	0.15
	FDI	501.5	48.3	1614.0	409.2	1.00	3.29	4.93	0.09
	EXP	9638.0	2586.8	20265.4	5494.6	0.62	2.02	3.01	0.22
	IMP	12998.2	3497.1	26801.4	7815.5	0.49	1.65	3.38	0.18
Myanmar	GDP	781.5	137.2	1418.2	500.3	−0.11	1.28	2.51	0.29
	FDI	1471.0	150.5	4733.3	1375.3	0.97	2.95	3.12	0.21
	EXP	7039.2	19.4	23147.3	8778.3	0.69	1.91	2.58	0.28
	IMP	7720.8	11.4	23567.4	9895.5	0.72	1.78	2.94	0.23

Source(s): The authors

Table 4

Unit root results

	Level								First difference
	ADF test				PP test				ADF test				PP test
	GDP	FDI	EXP	Imp	GDP	FDI	EXP	Imp	GDP	FDI	EXP	Imp	GDP	FDI	EXP	Imp
India	1.91	−0.32	0.47	0.15	1.69	−0.54	0.64	0.23	−4.38	−5.47	−3.89	−2.05	−5.24	−5.45	−3.88	−2.45
Bangladesh	10.12	0.54	3.44	3.52	9.67	0.70	3.23	3.38	−0.91	−2.16	−3.54	−4.85	−2.71	−2.70	−5.64	−6.06
Nepal	2.66	−0.45	−0.47	3.56	2.58	−0.68	−0.63	−0.56	−4.00	−7.75	−4.92	0.81	−5.04	−7.75	−4.86	−5.35
Bhutan	1.07	−3.61	0.08	−0.25	0.98	−3.02	−3.67	0.19	−4.68	−5.82	−3.60	−4.91	−5.07	−6.21	−3.58	−4.82
Myanmar	−0.24	−2.14	1.85	0.08	−0.39	−1.97	1.59	0.14	−2.97	−2.05	−6.67	−5.01	−2.80	−2.11	−5.67	−5.51
Thailand	0.55	−4.04	0.35	−0.40	0.49	−3.95	0.42	−52.00	−3.18	−7.34	−4.62	−4.73	−3.49	−4.83	−4.62	−4.64
Sri Lanka	−1.87	0.72	1.93	−0.34	−1.57	0.72	2.02	−0.35	−3.02	−4.87	−1.94	−5.69	−2.84	−5.08	−5.72	−5.74

Note(s): * indicates the significance at 5% level of significance

Source(s): The authors

Table 5

Results of Johansen cointegration tests

Countries	Vector	Trace statistics	Critical value	Max-eigen statistics	Critical value	Remarks
Bangladesh	0	88.573	55.24578	50.99703	30.81507	Cointegration
	1	37.57597	35.0109	28.15466	24.25202
India	0	94.97033	55.24578	53.18445	30.81507	Cointegration
	1	41.78588	35.0109	27.80338	24.25202
Nepal	0	65.24821	55.24578	31.38856	30.81507	Cointegration
	1	34.85966	35.0109	14.99801	24.25202
Bhutan	0	38.15839	47.85613	17.29272	27.58434	No Cointegration
	1	20.86568	29.79707	11.46506	21.13162
Thailand	0	57.43045	55.24578	35.70826	30.81507	Cointegration
	1	31.72219	35.0109	21.47818	24.25202
Sri Lanka	0	105.2674	55.24578	43.16943	30.81507	Cointegration
	1	62.098	35.0109	34.3137	24.25202
Myanmar	0	138.6338	55.24578	64.64653	30.81507	Cointegration
	1	73.98729	35.0109	49.22953	24.25202

Source(s): The authors

Table 6

VECM on BIMSTEC countries

	Variables	ECT		D(GDP(−1))		D(FDI(−1))		D(IMP(−1))		D(EMP(−1))		C
Bangladesh	D(GDP)	−0.214266	[−4.02762]	0.917524	[9.05506]	−0.008766	[−2.54161]	−0.012573	[−3.42408]	0.002617	[0.57818]	35.31178	[4.08137]
	D(FDI)	−0.534299	[−0.59300]	−1.435493	[−2.83647]	−0.346028	[−1.26228]	−0.028292	[−0.45494]	0.021628	[0.28212]	208.7713	[1.42473]
	D(IMP)	14.01002	[2.18458]	40.8062	[3.34067]	1.203721	[0.61692]	0.580051	[1.31042]	−0.591918	[−1.08475]	−429.3376	[−0.41164]
	D(EMP)	−0.278326	[−0.08438]	15.36219	[2.44518]	1.721931	[1.71580]	0.176116	[0.77356]	−0.373597	[−1.33114]	812.5358	[1.51465]
India	D(GDP)	0.002734	[0.10032]	0.407372	[1.49008]	0.002066	[0.77162]	−0.001452	[−1.60647]	0.001092	[0.88305]	49.37766	[2.34404]
	D(FDI)	−0.921374	[−0.43331]	39.11542	[2.83358]	0.003815	[0.01826]	−0.12619	[−1.78936]	0.131416	[1.36133]	−239.2814	[−0.14557]
	D(IMP)	31.56728	[3.04933]	313.2996	[3.01657]	−1.450039	[−1.42548]	−0.393283	[−1.14546]	0.819794	[1.74429]	−2750.671	[−0.34372]
	D(EMP)	10.66153	[1.19913]	190.762	[2.13859]	−1.880653	[−2.15264]	−0.082173	[−0.27867]	0.128621	[0.31864]	8938.775	[1.30056]
Nepal	D(GDP)	−0.058935	[−0.76748]	−0.217529	[−0.79188]	0.652476	[3.10358]	0.04161	[1.72398]	−0.112782	[−1.64265]	28.10844	[3.58952]
	D(FDI)	−0.344906	[−5.02429]	0.543982	[2.21518]	−0.176205	[−0.93755]	−0.084814	[−3.93083]	0.004605	[0.12071]	29.54362	[4.22030]
	D(IMP)	−1.302848	[−0.84398]	−7.364033	[−1.33354]	11.52289	[2.72649]	0.789013	[1.62616]	−1.176449	[−1.37125]	419.0402	[2.66195]
	D(EMP)	0.784025	[1.54398]	−1.070578	[−3.58936]	−0.823329	[−0.59223]	0.260111	[1.62971]	−0.502316	[−1.77990]	32.38921	[0.62549]
Thailand	D(GDP)	−0.111353	[−0.91350]	0.854155	[2.66635]	0.016528	[0.67958]	−0.012755	[−1.48739]	0.005637	[0.43630]	87.5175	[0.92654]
	D(FDI)	−0.200441	[−4.15749]	−1.454035	[−3.44232]	0.280039	[1.12208]	−0.07428	[−0.84413]	0.047181	[0.35585]	438.1794	[0.45207]
	D(IMP)	1.744059	[0.20886]	40.52191	[1.84650]	−0.23889	[−0.14338]	−1.01207	[−1.72283]	0.919982	[1.03937]	−693.3019	[−0.10714]
	D(EMP)	−0.363737	[−0.06936]	31.76548	[2.30493]	0.208285	[0.19907]	−0.922312	[−1.50008]	0.766868	[1.37961]	3795.704	[0.93408]
Myanmar	D(GDP)	−0.198979	[1.35866]	0.218207	[1.00437]	0.074296	[2.30094]	−0.052134	[−2.07049]	0.057381	[1.26642]	43.16152	[1.76343]
	D(FDI)	−0.519811	[−0.39632]	3.892794	[2.00069]	−1.04868	[−3.62639]	0.58874	[2.61076]	−0.469991	[−1.07279]	−220.6403	[−1.00655]
	D(IMP)	1.010173	[0.15600]	−4.463516	[−0.46464]	−0.592342	[−0.41488]	0.868756	[0.78030]	−1.147766	[−1.02527]	1991.274	[1.83993]
	D(EMP)	10.66075	[1.59539]	−14.38186	[−1.45083]	1.593414	[1.08154]	−1.213798	[−1.05651]	0.770971	[0.66740]	2708.239	[2.42506]
Sri Lanka	D(GDP)	−0.014902	[−0.29451]	1.049432	[4.55944]	0.034846	[2.17298]	−0.060138	[−1.94806]	−0.07863	[−1.14675]	85.16197	[2.28877]
	D(FDI)	0.226942	[2.44583]	0.914871	[2.16751]	0.283533	[0.76752]	0.014039	[0.24800]	−0.131883	[−1.04884]	−39.54349	[−0.57953]
	D(IMP)	1.924749	[3.72868]	9.605984	[4.09084]	2.28231	[1.11053]	−0.573339	[−1.82045]	0.28212	[0.40330]	−328.839	[−0.86627]
	D(EMP)	0.307752	[1.23154]	4.177127	[3.67464]	1.316608	[1.32336]	−0.47225	[−3.09745]	0.210618	[0.62195]	246.7809	[1.34291]

Source(s): The authors

Table 7

Relation between variables

Table 8

Diagnostic testing

Countries	Serial correlation test	Normality test	Heteroscedasticity test
Bangladesh	0.893 (0.51)	0.598 (0.61)	1.87 (0.35)
India	199.74 (0.89)	35.85 (0.69)	34.44 (0.68)
Myanmar	98.88 (0.96)	50.86 (0.87)	39.58 (0.890)
Nepal	12.68 (0.36)	0.748 (0.46)	8.5870 (0.74)
Sri Lanka	50.75 (0.54)	64.02 (0.78)	32.63 (0.64)

Source(s): The authors

Table 9

Robustness testing results

Country	F-test	Selected model	Remarks
Bangladesh	56.33449	ARDL (1, 0, 1, 0)	Cointegration
India	10.60020	ARDL (1, 0, 0, 1)	Cointegration
Nepal	9.045825	ARDL (1, 0, 1, 1)	Cointegration
Bhutan	3.224734	ARDL (1, 0, 0, 0)	No Cointegration
Thailand	5.696052	ARDL (1, 0, 1, 1)	Cointegration
Myanmar	4.619435	ARDL (1, 0, 0, 1)	Cointegration
Sri Lanka	6.822991	ARDL (1, 0, 1, 0)	Cointegration

Critical value (percent)	Lower bound 1(0)	Upper bound 1(1)
10%	4.428	5.816
5%	3.164	4.194
1%	2.618	3.532

Source(s): The authors

Notes

1. https://www.mea.gov.in/press-releases.htm?dtl/33763/17th_BIMSTEC_Ministerial_Meeting_April_01_2021

2. https://bimstec.org/?event=moving-towards-finalizing-the-bimstec-trade-facilitation-strategic-framework-2030