The Nexus Between Infrastructure and Export: An Empirical Evidence from Pakistan

Abstract

This research, by using the autoregressive distributive lag method, examines the long- and short-term causal relationship between infrastructure and exports in Pakistan over the period 1990–2017. The empirical results revealed the existence of short- and long-term bi-directional causality concerning infrastructure and export in Pakistan. The results demonstrated that infrastructure strongly improves export in the short and long run. Conversely, export encourages the quality and availability of infrastructure in Pakistan in the long run. Furthermore, this study also uses sub-indices of infrastructure individually as dependent and independent variables. The study result demonstrated that the long- and short-term effects of infrastructure and its sub-indices (transport, electricity, communication, finance) on export is positive and significant. Also, when infrastructure sub-indices are used as dependent variables, the results indicate that the effect of export on sub-indices is positive and significant in the long run; however, in the short run, it is insignificant. The bi-directional linkage between infrastructure and export suggests that improving the quality and increasing the availability of infrastructure would enable Pakistan’s economy to catch up with the advanced economies, specifically in export. Furthermore, control variables of per-capita GDP, exchange rates, human capital, and domestic spending also expand the bi-directional causal relationship between Pakistan’s infrastructure and exports.

Keywords

Infrastructure export ARDL Pakistan

Introduction

The availability of infrastructure in determining trade, especially export, has gained abundant attention in the modern era where numerous empirical research studies encourage the opinion that availability of infrastructure enhances exports (Ahmad et al., 2015; Donaubauer et al., 2018). The current literature stock, however, hardly explores the role exports play in improving the efficiency and availability of the country’s infrastructure. This study examines that the availability of infrastructure significantly encourages the export (Andrés et al., 2014; Ismail & Mahyideen, 2015) while the lack of infrastructures, such as transport, energy, and telecommunication can adversely impact the export (Portugal-Perez & Wilson, 2012; Straub, 2011; Yeaple & Golub, 2002). Countries, such as Singapore and Hong Kong, with improved quality and abundance of infrastructure, perform well in international trade, whereas poor-infrastructure economies, such as Pakistan and Nepal, face vulnerabilities in the outward sector (Andrés et al., 2014; Asif et al., 2019). This presents that infrastructure is vital for the encouragement of trade, especially export (Anderson & Wincoop, 2003; Brooks & Menon, 2008).

International trade, especially export, is often suggested as a strategy for enhancing economic growth and development (Donaldson & Hornbeck, 2016; Maparu & Mazumder, 2017). Intuitively, once the economy opens-up to trade, the government has a strong incentive to invest in infrastructure (trade related) to boost firms’ productivity and thus international competitiveness.

The literature on the role of infrastructure mainly emphasizes on its role in providing physical connectivity and information (Vijil & Wagner, 2012; Volpe et al., 2017). The infrastructure supports a firm in expanding catchment areas (Francois & Manchin, 2013; Limao & Venables, 2001), access larger labor markets (Duranton & Turner, 2012) and reduces logistic costs (Donaldson, 2018; Roller & Waverman, 2001; Wan & Zhang, 2016). Roller and Waverman (2001) concluded that a 10 percent reduction in transport costs increases trade by about 6 percent, and a 1 percent rise in total investment in infrastructure raises exports by about 0.6 percent and imports by 0.3 percent. These advantages accrue from the direct usage of infrastructure by firms and households which can create a direct impact. On the other hand, the proximity of businesses and industries contributes to cost diminution and productivity enhancement by permitting firms to share a common labor pool and enjoy knowledge spillovers (Fujita et al., 2001; Fujita & Thisse, 2013). Chengete and Alagidede (2017) demonstrate that a 1 percent rise in the availability of infrastructure would increase productivity by 0.24 percent each year in developing African countries. This impact arising from agglomeration can be defined as the indirect infrastructure effect. Therefore, this research article examines whether infrastructure plays any role in the increase of international trade, particularly exports and, does increased exports to improve infrastructure in Pakistan? Put more simply, this article examines the bi-directional causal relationship between infrastructure and export in Pakistan’s economy.

Previous literature indicates that many studies have been conducted to establish the causal relationship between infrastructure and export. However, most of them are based on specific aspects of infrastructure. Roller and Waverman (2001) and Ismail and Mahyideen (2015) evaluated the effect of telecommunications on economic development and trade. Hoffmann (2003) examined single indicator international telephone lines, the total length of paved and unpaved roads, and the number of aircraft departures to find the association between infrastructure related to public and international capital flows. Brun et al. (2005) and Limao and Venables (2001) considered a wider scope and studied several features of infrastructure to investigate the relationship between transport cost and infrastructure. They constructed an average of specific variables, considering all aspects of infrastructure by assigning them equal weights. Some research studies reduced this challenging assumption by using principal components analysis (PCA); Francois and Μanchin (2013) and Kumar (2006) applied PCA in a panel data. Conversely, employing PCA in data tends to unduly restrict the set of economies and the data series that can be involved in the study. Any gaps in the data series would have the influence that the devised indices are no longer comparable over time (Donaubauer et al., 2015). Several of the latest studies still face the very same problem of error in measurement of infrastructure issue as Donaldson (2018) examined the development of rail and road networks in colonial India during 1853–1930 and others like Volpe Martincus et al. (2017) for Peru, Duranton et al. (2013) for the USA, and Ghani et al. (2014) and Datta (2012) for India used the total length of highway networks for their studies.

This study applies a new global infrastructure index, used for the first time by Donaubauer et al. (2016), to the annual dataset of thirty indicators of the efficiency and quantity of the network and its sub-indices such as transport, electricity, telecommunications, and finance (see Appendix 1) by applying Unobserved Component Model (UCM) to be used to resolve these limitations. Complete information about this index is available in Donaubauer et al., (2015). The work of this study is consistent with recent studies (Khan et al., 2018; Khan et al., 2019; Rehman et al., 2019) and uses the autoregressive distributive lag (ARDL) model to test the long- and short-term relationships of infrastructure and export. Over the other econometric techniques, the advantage of the ARDL method is that it provides both the short- and the long-term results. In addition, it also suggests long-term speed of adjustment (Pesaran et al., 1997; 1999).

The remainder of the article is structured as follows: The following section discusses the trend of Pakistan’s infrastructure and export. The next section describes the data collection and source of the current study, followed by the research methodology. The next section then explains the results and discussions of the study and the final section elaborates on the conclusions and recommendations of the study.

The Trend of Infrastructure and Export in Pakistan

Compared to international standards, Pakistan’s infrastructure is weaker which has a detrimental effect on overall trade and investment, both FDI as well as domestic investment. Businesses suffer from the shortage of electricity and lack of sufficient water and sanitation regulations. All public and private manufacturing companies battle uphill against the poor infrastructure. Improving the quality and enhancing the availability of trade-related infrastructure is a prerequisite for maintaining and increasing trade. It is important for the economy as a whole to strengthen electricity, transport, communications, and logistics and to promote exports, especially trade.

Assessments indicate that owing to inadequate infrastructure, Pakistan loses about 4–6 percent of its gross domestic product (GDP), amounting to almost $6 billion. Logistical bottlenecks upsurge the production cost of the products by around 30 percent. This has a substantial influence as Pakistan is facing tough competition from countries such as India and China in the export market. To improve the quality and availability of infrastructure, the needs are immense and its resources scarce. Not only is there a reduced fiscal space but also enormous gaps in public sector capability to build and maneuver infrastructure. The tight fiscal variables such as a fiscal deficit of 8.7 percent, a trade deficit of approximately $22 billion, and a current account deficit of 4.4 percent of GDP in the FY2018 do not allow to spare public sector resources for infrastructure development in Pakistan. During the last two decades, Pakistan is suffering from enormous trade deficits (imports are more than exports). The trade deficit for the FY2016–2017 was $28,413.12 million, exports being $25,114.12 million, and import $53,527.24 million (Asif & Rehman, 2019; Li et al., 2019). This analysis is focused on both aggregate infrastructure as well as its sub-indices, including transportation, telecommunications, energy, and finance, to examine the effect of infrastructure on exports in Pakistan. Figure 1 illustrates the infrastructure deficit for the undermentioned sectors of Pakistan’s economy. See Appendix 1 for more specific details about the selected metrics used in the construction of aggregates and infrastructure sub-indices.

Figure 1.

Source: The authors.

The condition of the roads in Pakistan is no different compared with other South Asian economies such as India and Sri Lanka (Figure 2). In FY2016, the total length of roads in Pakistan was 269,618 km, of which 63 percent were paved. Around 60 percent of the road network is in poor condition due to insufficient maintenance and traffic congestion. The length of motorways and national highways in Pakistan’s total road network is just 4.2 percent, but together they cater to over 85 percent of Pakistan’s overall traffic (Pakistan Economic Survey, 2017). The low quality of the roads not only contributes to the cost of production but also greatly hampers mobility of people and goods. Poor transport infrastructure thus holds both domestic and foreign trade on pause (Andrés, et al., 2014). Most of the previous literature indicates a positive and significant impact of infrastructure on trade, particularly export. In their studies, Coşar and Demir (2016), Li and Qi (2016), and Yeaple et al. (2002) examined that better developed ports, roads, transport facilities, and airports lead to expanding national and international trade. Moreover, Yeaple and Golub (2002) and Njinkeu et al. (2008) explored that the quality and availability of road infrastructure enhance productivity, especially in the manufacturing and textile sector.

In comparison with other Asian countries, the number of Internet users in Pakistan has expanded since the early 2000s. Although cell phone subscribers began to increase in Pakistan one decade later, its prompt growth in the last 10 years is fairly significant. In FY2001, Pakistan had the least number of cell phone subscribers—5 per 1,000 population—but it boosted significantly to 573 per 1,000 populations in FY2010. Figure 1 demonstrates the overall telecommunication infrastructure progress from 1990 to 2017, which indicates that immense investment is needed to expand the communication sector in Pakistan. Fink et al. (2005), Giuseppe et al. (2003), Francois and Manchin (2007), Portugal-Perez and Wilson (2012), and Roller and Waverman (2001) argued that the trade flow is significantly related to infrastructure, especially telecommunication sector. They used different aggregate indicators individually as a proxy of telecommunication sectors such as per capita broadband, mobile phone subscribers and telephone subscribers, but the present study uses different variables of telecommunication by constructing one composite sub-index with the method of UCM (Donaubauer et al., 2016).

Figure 2.

Source: The authors.

In addition to telecommunication and transportation, the energy deficit and its related infrastructure presents a major constraint to Pakistan’s domestic and foreign trade. Rehman et al. (2020), Escibano et al. (2009), Ayogu (1999), Kirubi et al. (2009), and Blalock and Veoso (2007) reported that the energy industry had a more important impact on foreign trade than any other infrastructure field. Various indicators were used as a proxy for the energy sector such as per capita access to electricity, natural gas, and fuel oil to access the relationship between infrastructure and trade but did not offer a true picture of the energy infrastructure. This study considers various indicators of the energy sector devised by Donaubauer et al. (2016) to present comprehensive energy infrastructure. Also, spending in the energy sector is essential for growth and for maintaining a high trade balance in the case of Pakistan where per capita electricity access is about 98 percent. Despite wide access to electricity, excessive load-shedding has adversely affected the economy of Pakistan (Asian Development Bank, 2017). The deteriorating condition of the energy sector in Pakistan (Figure 1) advocates for a huge investment needed in energy infrastructure.

Data Source

To assess the relationship between infrastructure and exports over the period 1990–2017, by applying UCM, we based our study on a composite index of global infrastructure developed from data collected from various data sources (see Table 1). Complete detail on the development of this index can be found in Donaubauer et al. (2015). Besides, the devised index includes four infrastructure sub-indices, that is, transportation (TINF), communication (CINF), energy (EINF), and financial (FINF) to understand the function of physical infrastructures better. Nevertheless, the aggregate infrastructure and its sub-indices are composed of 30 indicators of the quality and quantity of infrastructure to ensure that all main factors of infrastructure are included. This research also uses control variables such as exchange rate (EXR), per capita GDP (PGDP), domestic investment (DI), and human capital (HC) in addition to the main variables (i.e., infrastructure and export). Data for all of these control variables are obtained from world development indicators (WDI). Besides this, the data for Pakistan’s total export is also taken from WDI.

Global infrastructure aggregates and sub-indices contain negative numbers that were transformed to positive values by adding 1 before getting a natural log.

Table 1.

List of 30 Indicators of Infrastructure Along with Data Sources

S. No.	Variables	Data Source
1	Transport infrastructure A. Land transport: i. Total length of road network ii. Total length of paved or cemented road iii. Percentage of motorways to total road network iv. Numbers of registered passenger vehicles	World Road Statistic (WRS), WDI and International Road Federation (IRF)
	v. Numbers of registered commercial vehicles	Facts and figures of VDA and the German Association of Automotive Industry (GAAI)
	vi. Total length of railway tracks	WDI
	vii. Capacity of goods transportation
	viii. Total numbers of passengers travelled through railway
	B. Sea transport: Aggregate shipment capacity of the economy’s ships. i. Comparative to its aggregate capacity ii. Comparative to the world aggregate capacity	The United Nations Conference on Trade and Development (UNCTAD) database
	C. Air transport i. Total numbers of registered cargo in Pakistan ii. Total numbers of registered passenger airplanes in Pakistan iii. Total numbers of passengers travelled through air transport iv. Cargo carrying capacity of the economy’s air transport	WDI
2	A. Telecommunication infrastructure i. Total numbers of landline subscribers ii. Total numbers of cell phone subscribers iii. Total numbers of integrated service digital network (ISDN) Subscribers	WDI
	B. Quality measures i. Rate of fault per 100 landlines (expressed in per capita terms)
3	A. Energy infrastructure i. Total consumption of electricity (MgW) ii. Total output of electricity (MgW)Note: Both are measured in per capita terms	WDI
	B. Quality measures i. Transmissions and distribution losses as a percentage of total output
4	A. Financial Infrastructure i. Stock market’s turnover ratio ii. Per capita bank accounts iii. Total numbers of public companies in per capita term iv. Credit from deposits of financial institutions as a percentage of GDP v. Value of the traded shares as a percentage of GDP vi. M2 money supply as a percentage of GDP	World Bank Global Financial Development Database (WBGFDD)

Source: Author’s own calculations.

Econometric Methodology

Using the ARDL method, this research study examines bi-directional long- and short-term dynamic causal linkages between infrastructure and exports in Pakistan. Pesaran et al. (1999) and Pesaran et al. (2001) created the ARDL cointegration estimator. Whether the selected variables of our interest are integrated at level, first difference {I(I)}, or combination of both, this approach is successful. Furthermore, the term error-correction can simply be derived from a simple linear transformation (Banerjee, 1993). Therefore, the additional advantages of the ARDL technique are as follows: (a) It can be implemented in a time series dataset having comparatively limited sample size and (b) this approach is additionally advantageous compared to the Juselius and Johansen cointegration estimators for small samples and also offers short-term adjustment without eluding long-term information (Pesaran et al., 1999). The ARDL approach estimates the following unrestricted ECMs to investigate the relationship between infrastructure and exports.

\begin{array}{l} Δ \log EX = α_{0 EX} + \sum_{i = 1}^{p} ψ_{i EX} Δ \log EX + \sum_{i = 1}^{p} ϕ_{i EX} Δ {GINF}_{t - i} + \sum_{i = 1}^{p} δ_{i EX} Δ {PGDP}_{t - i} + \sum_{i = 1}^{p} ϖ_{i EX} Δ {DI}_{t - i} \\ + \sum_{i = 1}^{p} θ_{EX} Δ TO + \sum_{i = 1}^{p} π_{EX} Δ EXR + λ_{1 EX} \log {EX}_{t - 1} + λ_{2 EX} {GINF}_{t - 1} + λ_{3} {PGDP}_{t - 1} \\ + λ_{4EX} {DI}_{t - 1} + λ_{5EX} \log {TO}_{t - 1} + λ_{6EX} {EXR}_{t - 1} + μ_{1 t} \end{array}

(1)

\begin{array}{l} Δ {GINF}_{t} = α_{0 GINF} + \sum_{i = 1}^{p} ψ_{i GINF} Δ {GINF}_{t - i} + \sum_{i = 1}^{0 p} ϖ_{i GINF} Δ {EX}_{t - i} + \sum_{i = 1}^{p} ϕ_{i GINF} Δ {PGDP}_{t - i} \\ + \sum_{i = 1}^{p} δ_{i GINFRA} Δ {DI}_{t - i} + \sum_{i = 1}^{p} θ_{i GINF} Δ {TO}_{t - i} + \sum_{i = 1}^{p} π_{i GINF} Δ {EXR}_{t - i} + γ_{1 GINF} {GINF}_{t - 1} \\ + \sum_{i = 1}^{p} γ_{2 GINF} Δ {EX}_{t - i} + γ_{3 GINF} {PGDP}_{t - 1} + γ_{4 INFRA} {DI}_{t - 1} + γ_{5 GINF} {TO}_{t - 1} + γ_{6 GINF} {EXR}_{t - 1} + μ_{2 t} \end{array}

(2)

While Δ shows operator difference, μ_1t and μ_2t are the disturbance term, and p signifies the length of the lags of the chosen indicators. γ indicates the particular estimates of the long run, and finally ψ, θ, φ, ϖ, δ, and π are the coefficients of short-term dynamics.

The present research uses bounds testing technique to assess the long-term relation between infrastructure and export with the null hypothesis of no cointegration against the alternative of cointegration existence. For Equation 1, the dependent variable is ΔEX and the following is the long-term hypothesis regarding the relationship.

H₀: There is no cointegration between selected variables.

H₁: There is cointegration between selected variables.

Pesaran et al. (2001) postulates two lower and upper critical values for the cointegration test. If the calculated value of F statistics is greater than the upper bound critical value, then the null hypothesis of no cointegration among the selected variables is rejected and the null hypothesis cannot be rejected if it lies below the lower bound critical value. Finally, the test is inconclusive if the calculated F statistics value lies in between the lower and upper bound values. Regardless of the integration order, the Granger causality test is successful if the relation of long-term equilibrium among the selected variables is formed (Groenewold & Tang, 2007; Rehman & Khan, 2015). The Granger causality check, however, requires us to rely on the vector error correction mechanism (VECM) to regulate the system’s short-term dynamics among the co-integrated indicators (Narayan & Smyth, 2005). To analyze the long- and short-term analysis of Granger causality, we examine the following VECM example.

\begin{array}{l} Δ {EX}_{t} = α_{1} + ω_{11}^{p} (L) {EX}_{t} + ω_{11}^{q} (L) {GINF}_{t} + ω_{11}^{q} (L) {PGDP}_{t} + ω_{11}^{q} (L) {DI}_{t} + ω_{11}^{q} (L) {TO}_{t} + δ_{1} {EXR}_{t - 1} \\ + δ_{1} {ECT}_{t - 1} + μ_{3 t} \end{array}

(3)

\begin{array}{l} Δ {GINF}_{t} = α_{2} + ω_{21}^{p} (L) {GINF}_{t} + ω_{22}^{q} (L) {EX}_{t} + ω_{22}^{q} (L) {PGDP}_{t} + ω_{22}^{q} (L) {DI}_{t} + ω_{22}^{q} ({L)TO}_{t} \\ + δ_{2} {EXR}_{t - 1} + δ_{2} {ECT}_{t - 1} + μ_{4 t} \end{array}

(4)

While L presents the “lag-operator”, which denotes that “L” ΔY = ΔYt_t–1

In Equations 4 and 5, the term (ECT_t−1) presents the speed of selected indicators to achieve their long-term equilibrium direction after a short-term shock. The significance of the calculated F-statistics from Equations 4 and 5 reveals the existence of a short-term causal relationship between the selected indicators while the significant value of the t-statistics-based coefficient of (ECT_t−1) indicates the existence of a long-term causal relationship between the selected indicators.

Results and Discussions

Before evaluating the long-term relationship between infrastructure and export, it is important to decide the order of integration of all the selected variables, because if the variables are incorporated of order I(2) or higher than the estimated, F-statistics become ineffective (Ouattara, 2004). To investigate the order of integration among these selected variables, Augmented Dickey–Fuller (ADF) and the DF-GLS tests are thus employed.

The findings in Table 1 suggests that all variables are integrated of order I(1) (first difference) and that none of the variables are integrated of order I(2) (second difference) or higher, specifically promoting the use of ARDL strategies rather than any other alternative cointegration procedures.

Table 2.

Unit Root Test Results

Variables	ADF Test with Intercept	ADF Test with Intercept and Trend	DF-GLS Test with Intercept	DF-GLS Test with Intercept and Trend
Ln_EX	–1.606	–1.166	0.369	–1.484
Ln_GINF	–2.53	–1.98	–1.456	–1.542
Ln_PGDP	1.562	–1.232	1.603	–1.171
Ln_HC	–1.544	–2.40	–1.230	–2.509
Ln_DI	1.467	–0.946	1.510	–1.180
Ln_EXR	–0.285	–2.66	–0.152	–2.705
∆ln_EX	–4.393***	–4.362***	–3.637***	–4.257***
∆ln_GINF	–1.863	–5.358***	–1.358	–5.547***
∆ln_PGDP	–4.211***	–4.927***	–4.296***	–5.09***
∆ln_TO	–6.354***	–6.344***	–5.695***	–6.145***
∆ln_DI	–3.762***	–4.25***	–3.849***	–4.437***
∆ln_EXR	–5.257***	–5.203***	–4.62***	–5.811***

Source: Author’s own calculations.

Notes: *** signify significance level at 1% respectively. ∆ denotes first difference. the alternative hypothesis of stationarity is accepted, for both intercept as well as intercept and trend, at 1%, 5%, and 10%, respectively.

Table 3.

Long-Term Results

Export Is a Dependent Variable		Infrastructure Is a Dependent Variable
Variables	F-Stat	Variables	F-Stat
F(EX/GINF)	3.68	F(GINF/EX)	0.23
F(EX/GINF/PGDP)	3.89*	F(GINF/EX/PGDP)	0.29
F(EX/GINF/PGDP/DI)	7.53***	F(GINF/EX/PGDP/DI)	3.92*
F(EX/GINF/PGDP/DI/HC)	5.48***	F(GINF/EX/PGDP/DI/HC)	5.17***
F(EX/GINF/PGDP/DI/HC/EXR)	6.39***	F(GINF/EX/PGDP/DI/HC/EXR)	4.82***

Source: Author’s own calculations.

Notes: *** and * indicate significance at 1% and 10% level, respectively. All variables are converted into a natural log (ln).

By applying the Wald test, the results in Table 2 show a likely long-term relationship between infrastructure and exports. AIC is the basis for selecting the optimal lag period for testing the models. The F-statistics calculated values for Equations 1 and 2 are higher than the critical upper-bound values at 1 percent and 5 percent respectively; hence, the null hypothesis of no cointegration among the variables selected is rejected. The empirical findings of the ARDL-bound test suggest that there exists a stable long-term relationship between infrastructure and export as well as the PGDP, domestic investment (DI), human capital (HC), and exchange rate (EXR) which are the control variables.

The long-term association is only a necessary condition, which is not enough for causal relationships among the selected variables (Morley, 2006). Confirmation of the long-term relationship between indicators indicates that there must be at least unidirectional causality between the variables described without specifying the causal direction (Narayan & Smyth, 2005). Thus, the Vector Error Correction Model (VECM) is calculated to evaluate the direction of the short- and long-term causal relation between infrastructure and export (Rehman & Ding, 2019).

Table 4.

Results of Granger Causality (Export Is the Dependent Variable)

	In the Short Run					In the Long Run
	∆GINF	∆PGDP	∆HC	∆DI	∆EXR	ECT_(t–1)
∆EX	0.87**					–0.43***
∆EX	0.33*	0.44**				–0.46***
∆EX	.019*	0.95**	0.23**			–0.32***
∆EX	0.43***	0.49**	0.61**	0.18*		–0.28**
∆EX	0.16*	0.43**	0.78**	0.59**	0.50***	–0.34***

Source: Author’s own calculations.

Notes. ***, **, and * indicate significance at 1%, 5%, and 10% level, respectively. All variables are converted into a natural log (ln).

VECM’s reported empirical finding in Table 3 highlighted the long- and short-term causality from infrastructure (GINF) to export (EX). The findings show that in both cases the error correction term (ECT) coefficient is important, that is, with and without control variables. Such empirical findings indicate the existence of short- and long-term causality from infrastructure to export, and also control variables of PGDP, DI, EXR, and HC, suggesting that infrastructure matters in promoting Pakistan’s aggregate export. The robustness of the causal effect from infrastructure to export is evident from reliable results in Table 4, both with and without selected control variables.

On the other hand, in the short run, the casual impact of export on infrastructure (GINF) is high, while in the long run, it is weakly significant. The values of (ECT – 1) in Table 3 express rapid export equilibrium adjustment while in Table 4 indicate a gradual return to the long-term equilibrium position of the infrastructure. From the empirical findings in Table 4, it is reported that the causal effect of exports on infrastructure is not so strong in the long run because of the indirect influence of exports on infrastructures. But, the values of ECT_(t–1) are significant. More specifically, Pakistan is a developing country and its aggregate export is low as compared to other developing economies such as India and Iran. So, the spillover effect is weak through which the export of Pakistan cannot enhance the macro-variables including infrastructure in the long run.

Short- and Long-Term Coefficients

Appendix 2 shows that the impact (i.e., coefficient) of aggregate and sub-indices of infrastructure like transport (TINF), communication (CINF), energy (EINF), and financial (FINF) on export is positive and significant in the long run and in the short run. Empirical findings showed that improved quality and availability of infrastructure encourages export within the home country. The significant positive role of infrastructure in export is due to the crucial role it plays in trade cost. The present finding is consistent with the idea that the lack of infrastructure deteriorates market connectivity, creates hindrance in achieving trade potential, causes frictions in the market, and imposes unnecessary delays, and thus increases the overall cost of production and vice versa (Andrés et al., 2014). The results are in line with findings of Ahmad et al. (2015). In addition to the key variables, the control variables of PGDP, human capital (HC), and exchange rate (EXR) are also significant in both the short run and the long run. To check robustness, we examined the key variables with and without control variables that give consistent coefficients.

At the other hand, the long- and short-term effects of exports on infrastructure (i.e., transport, communication, electricity, and finance) is positive (see Appendix 2). In the long run, this impact is important but not relevant in the short run. The empirical results showed that the export coefficient and all other control variables are positive in the long run. The findings contended that export improves infrastructure availability and efficiency. The reason being that when a country opens the economy to trade, the government has a strong motive to invest in the infrastructure sector (trade related), for the promotion of a firm’s productivity (Hochman et al., 2013) and to increase export. The findings of the present research study demonstrated that the reverse (bi-directional) causality also holds (accurate) that higher export positively influences the availability and quality of infrastructure in the long run.

Diagnostic Tests Results

Table 5 indicates that the selected variables are not serially correlated as indicated by the insignificant F-statistic values of the “Breusch-Godfrey Serial LM Test.” It demonstrates that we can accept the “null hypothesis” (H₀: The selected model is not serially correlated) and reject the alternative hypothesis of the serial correlation problem (i.e., H₁: The chosen model effected by serial correlation problem) (Baharumahah et al., 2009). Correspondingly, this study performs ARCH to test for heteroskedasticity. The insignificant F-statistic value of the ARCH test illustrated that we can reject the alternative hypothesis (H₁: The existence of heteroskedasticity) in the model of our interest, and we accept the null hypothesis of no heteroskedasticity in the current model (Baharumahah et al., 2009). Jarque–Bera test was conducted to test the normality of the present model. The Jarque–Bera test’s insignificant values showed that the data for our selected variables are normally distributed. Ramsey RESET Test implies that the model is correctly specified, which indicates that there is no variable beyond the boundary (Baharumah et al., 2009). The study also performed a CUSUM and CUSUM square test to check whether the model is stable or not (Appendix 3). In table 6, the results of BG and ARCH test confirm that result is not suffering from the problem of heteroscadasticity, JB test suggests error is normally distributed and RR specify that the correct functional form is chosen.

Table 5.

Results of Granger Causality (Infrastructure Is the Dependent Variable)

	In the Short Run					In the Long Run
	∆EX	∆PGDP	∆HC	∆DI	∆EXR	ECT_(t–1)
∆GINF	0.13*					–0.005
∆GINF	0.09	0.25**				–0.02
∆GINF	0.44**	0.61**	0.40*			0.14*
∆GINF	0.53**	0.25**	0.89**	0.59**		–0.18**
∆GINF	0.85**	0.15*	0.79**	0.45**	0.82**	–0.22**

Source: Author’s own calculations.

Notes: **, and * indicate significance at 5%, and 10% level, respectively. All variables are converted into a natural log (ln).

Table 6.

Diagnostic Tests

Tests	Export Is a Dependent Variable		Infrastructure Is a Dependent Variable
Tests	F-statistics	Prob. Value	F-statistics	Prob. Value
BG	0.66	0.54	2.45	0.112
ARCH	1.13	0.48	0.74	0.65
JB	0.51	0.78	0.23	0.85
RR	0.35	0.75	0.95	0.36

Source: Author’s own calculations.

Notes: BG stands for Breusch–Godfrey Serial LM Test, JB presents Jarque–Bera Normality Test, and RR is Ramsey RESET Test. The results of diagnostic tests are based on F-statistics.

Conclusion and Policy Implications

The empirical results demonstrated that there is a bi-directional relationship between infrastructure and export that infrastructure encourages export in the long and the short run. On the other hand, export in Pakistan enhances the quality and availability of infrastructure in the long run, but the impact is weak in the short run. These findings are consistent with the idea that if Pakistan increases its export, it may increase economic development including availability and quality of home infrastructure. Furthermore, this study also uses sub-indices of infrastructure (like transport, communication, energy, and financial) one by one as dependent variables. The results revealed that export significantly enhances the quality and availability of transport, communication, energy, and financial infrastructure in the long run, but it is insignificant in the short run. The results are in line with Hochman et al., (2013). The results are effective regarding policy towards export promotion.

In addition to the key variables (infrastructure and export), the control variables PGDP, DI, HC, and EXR also substantially increase the production and availability and efficiency of the overall and selected infrastructure sub-indices.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

This research work is supported by the National Social Science Fund of China under Grant number (17JJD790007).

Acknowledgment

We are grateful to Dr Julian Donaubauer for providing the full global infrastructure index dataset and enabling us to use it for estimation purposes.

Appendices

Appendix 2. ARDL Results (Aggregate and Sub-indices of Infrastructure Is Dependent Variables)

Variables	lnTINF	lnCINF	lnEINF	lnFINF	lnGINF
lnPGDP	0.290**	–0.149	1.197***	0.332**	–0.179
S.E	0.13	0.15	0.14	0.17	0.15
lnHC	1.129***	0.975***	0.008	0.951***	0.993***
S.E	0.05	0.08	0.14	0.11	0.06
lnEXR	–0.180**	–0.337***	–0.476***	–0.250***	–0.26***
S.E	0.08	0.11	0.17	0.12	0.16
lnDI	0.394***	0.644***	0.048***	0.525***	0.72***
S.E	0.15	0.17	0.14	0.16	0.15
lnEX	0.578***	0.327***	1.820***	0.154**	0.370***
S.E	0.06	0.11	0.41	0.05	0.11
In the Short Run
∆lnPGDP	–1.308	–2.556**	–1.570*	1.638	–1.951*
	1.521	1.395)	1.151	1.458	1.565
∆lnHC	–0.470	–0.020	0.519***	0.497	0.079
	0.469	0.221	0.099	0.068	0.136
∆lnDI	0.448	0.491	0.135	0.584	0.221
	0.744	0.682	0.394	0.915	0.750
∆lnEXR	–0.470	–0.020	0.519***	0.497	0.079
	0.469	0.221	0.099	0.068	0.136
∆lnEX	0.572*	–0.938	0.252***	0.457	0.445
	0.350	1.666	0.105	0.385	0.744
Constant	–2.033**	1.569***	2.198*	0.131	0.677
S.E	1.109	0.446	1.003	0.324	(0.089)

Source: Author’s own calculations.

Notes. ***, ** and * indicate significance at 1% , 5%, and 10% level, respectively. All variables are converted into natural log (ln). S.E presents standard error.

Appendix 3. CUSUM and CUSUM of Square

Plot of CUSUM and CUSUM of square of recursive residuals, (Dependent = Export)

Plot of CUSUM and CUSUM of square of recursive residuals, (Dependent = Infrastructure)

References

Anderson

J. E.

, & Wincoop

E. V.

(2003). Gravity with gravitas: A solution to the border puzzle. The American Economic Review , 93, 170–192. doi:10.1257/00028280332145521.

Ayogu

(1999). Before prebendalism: A positive analysis of core infrastructure investment in a developing fiscal federalism. African Development Review , 11(2), 169–98

Ahmad

, Jaini

A. N.

, & Zamzamir

B. Z.

(2015). The impact of infrastructure on trade in Malaysia. Journal of Entrepreneurship and Business , 3, 42–49. doi:10.17687/jeb.0302.0.

Andrés

, Biller

, & Dappe

M. H.

(2014). Infrastructure gap in South Asia: Infrastructure needs, prioritization and financing (Policy Research Working Paper 7032). South Asia Region Sustainable Development Department (World Bank Group). doi: 10.1596/1813-9450-7032

Asian Development Bank. (2017). Detailed record of ADB’s operations, activities and financial results . https://www.adb.org/documents/adb-annual-report-2017

Asif

, & Rehman

(2019). The nexus among trade volume, trade deficit and real exchange rate: An empirical evidence from Pakistan. Journal of Economics and Sustainable Development , 10(4). doi:10.7176/JES.

Asif

, Rehman

F. U.

, Zheng

, & Shah

H. S.

(2019). Does trade with China can make growth in Pakistan more inclusive? Pre and Post Empirical Impact of China-Pakistan Economic Corridor. Developing Country Studies , 9(3), 57–67.

Brun

J. F.

, Carri

, Guillaumont

, & Melo

J. D.

(2005). Has distance died? Evidence from a panel gravity model. World Bank Economic Review , 19(1), 99–120.

Blalock

, & Veloso

F. M.

(2007). Imports, productivity growth, and supply chain learning. World Development , 35(7), 1134–1151.

10.

Brooks

H. D.

, Menon

Eds. (2008). Infrastructure and trade in Asia . Edward Elgar Publishing.

11.

Banerjee

(1993). Co-integration, error-correction, and the econometric analysis of non-stationary data . Oxford University Press.

12.

Baharumahah

Z. A.

, Mohd

H. S.

, & Masih

M. A.

(2009). The stability of money demand in China: Evidence from the ARDL model. Economic System , 33, 231–244. doi:10.1016/j.ecosys.2009.06.00.

13.

Chengete

, & Alagidede

(2018). The nexus between infrastructure (quantity and quality) and economic growth in Sub Saharan Africa. International Review of Applied Economics , 32(5), 641-672. https://doi.org/10.1080/02692171.2017.1355356

14.

Coşar

A. K.

, & Demir

(2016). Domestic road infrastructure and international trade: Evidence from Turkey. Journal of Development Economics , 118, 232–244. doi:10.1016/j.jdeveco.2015.10.00.

15.

Donaubauer

, Glas

, Meyer

, & Nunnenkamp

(2018). Disentangling the impact of infrastructure on trade using a new index of infrastructure. Review of World Economics , 154, 745–784. doi:10.1007/s10290-018-0322-.

16.

Donaubauer

, Meyer

, & Nunnenkamp

(2015). A new global index of infrastructure: Construction, rankings and applications. World Economy , 39: 236–259. doi:10.1111/twec.1229.

17.

Duranton

, & Turner

M. A.

(2012). Urban growth and transportation. Review of Economic Studies , 79(4), 1407–1440. https://dx-doi-org.web.bisu.edu.cn/10.1093/restud/rds010

18.

Donaldson

(2018). Railroads of the raj: Estimating the impact of transportation infrastructure. American Economic Review , 108(4–5), 899–934.

19.

Donaldson

, & Hornbeck

(2016). Railroads and American economic growth: A market access approach. The Quarterly Journal of Economics , 131(2), 799–858.

20.

Donaubauer

, Meyer

, & Nunnenkamp

(2016). Aid, Infrastructure, and FDI: Assessing the transmission channel with a new index of infrastructure. World Development , 78, 230–245. https://dx-doi-org.web.bisu.edu.cn/10.1016/j.worlddev.2015.10.015

21.

Duranton

, Morrow

P. M.

, & Turner

M. A.

(2013). Roads and trade: Evidence from the US. The Review of Economic Studies , 81(2), 681–724.

22.

Datta

(2012). The impact of improved highways on Indian RMS. Journal of Development Economics , 99(1), 46–57.

23.

Escribano

, Guasch

J. L.

, & Pena

(2009). Assessing the impact of infrastructure constraints on firm productivity in Africa (Working Paper No. 9). Africa infrastructure sector diagnostic, World Bank.

24.

Francois

, & Manchin

(2013). Institutions, infrastructure and trade. World Development , 46, 165–175.

25.

Fink

, Mattoo

, & Neagu

I. C.

(2005). Assessing the impact of communication costs on international trade. Journal of International Economics , 67(2), 428–445.

26.

Fujita

, Krugman

, & Venables A.

(2001). The spatial economy: Cities, regions and international trade (First ed.). The MIT Press.

27.

Fujita

, & Thisse

(2013). Economics of agglomeration . Cambridge University Press. https://EconPapers.repec.org/RePEc:cup:cbooks:9780521171960.

28.

Ghani

, Goswami

A. G.

, & Kerr

W. R.

(2014). Highway to success: The impact of the golden quadrilateral project for the location and performance of Indian manufacturing. The Economic Journal , 126(591), 317–357.

29.

Giuseppe

, & Stefano

(2003, April 1). Regulation, productivity and growth: OECD evidence. Economic Policy , 18(36), 9–72. https://doi.org/10.1111/1468-0327.00102

30.

Groenewold

, & Tang

S. K. H.

(2007). Killing the goose that lays the golden egg: Institutional change and economic growth in Hong Kong. Economic Inquiry , 45, 787–799. doi:10.1111/j.1465-7295.2007.00044..

31.

Hochman

, Tabakis

, & Zilberman

(2013). The impact of international trade on institutions and infrastructure. Journal of Comparative Economics , 41(2013), 126–140. https://dx-doi-org.web.bisu.edu.cn/10.1016/j.jce.2012.07.001

32.

Hoffmann

(2003). Cross-country evidence on the link between the level of infrastructure and capital inflows. Applied Economics , 35(5), 515–526.

33.

Ismail

W. N.

, & Mahyideen

M. J.

(2015). The impact of infrastructure on trade and economic growth in selected economies in Asia (ADBI Working Paper 553). Tokyo: Asian Development Bank Institute.

34.

Khan

M. A.

, Ilyas

R. M. A.

, & Hashmi

S. H.

(2018). cointegration between institutional quality and stock market development. NUML International Journal of Business & Management , 13(2), 90–103.

35.

Khan

M. A.

, Khan

M. A.

, Abdulahi

M. E.

, Liaqat

, & Shah

S. S. H.

(2019). Institutional quality and financial development: The United States perspective. Journal of Multinational Financial Management , 49, 67–80. https://doi.org/10.1016/j.mulfin.2019.01.001

36.

Kumar

(2006). Infrastructure availability, foreign direct investment inflows and their export orientation: A cross-country exploration. Indian Economic Journal , 54, 125–44. doi:10.1177/001946622006010.

37.

Kirubi

, Jacobson

, Kammen

D. M.

, & Mills

(2009). Community based electric micro-grids can contribute to rural development: Evidence from Kenya. World Development , 37(7), 1208–1221. https://doi.org/10.1016/j.worlddev.2008.11.005

38.

Limao

, & Venables

A. J.

(2001). Infrastructure, geographical disadvantage, transport costs and trade. World Bank Economic Review , 15(3), 451–479.

39.

Vijil

, & Wagner

(2012). Does aid for trade enhance export performance? Investigating the infrastructure channel. The World Economy , 35, 838–868. doi:10.1111/j.1467-9701.2012.01437..

40.

Maparu

S. T.

, & Mazumder

N. T.

(2017). Transport infrastructure, economic development and urbanization in India (1990–2011): Is there any causal relationship. Journal of Comparative Economics , 41, 126–140.

41.

Morley

(2006). Causality between economic growth and immigration: An ARDL bounds testing approach. Economics Letters , 90(1), 72–76.

42.

Narayan

P. K.

, & Smyth

(2005). Electricity consumption, employment and real income in Australia: Evidence from multivariate granger causality tests. Energy Policy , 33, 1109–1116. doi:10.1016/j.enpol.2003.11.010

43.

Ouattara

(2004). Foreign aid and fiscal policy in Senegal (Mimeo). University of Manchester.

44.

Pakistan Economic Survey. (2017). Government of Pakistan. Finance Division . http://www.finance.gov.pk/survey_1718.html

45.

Pesaran

M. H.

, Shin

, & Smith

R. P.

(1997). Pooled estimation of long-run relationships in dynamic heterogeneous panels (DAE Working Papers Amalgamated Series, No 9721). University of Cambridge.

46.

Pesaran

M. H.

, Shin

, & Smith

R. P.

(1999). Pooled mean group estimation of dynamic heterogeneous panels. Journal of the American Statistical Association , 94, 621–634. doi:10.1080/01621459.1999.1047415.

47.

Pesaran

M. H.

, Shin

, & Smith

R. J.

(2001). Bounds testing approaches to the analysis of level relationships. Journal of applied econometrics , 16(3), 289–326.

48.

Portugal-Perez

, & Wilson

J. S.

(2012). Export performance and trade facilitation reform: Hard and soft infrastructure. World Development , 40(7), 1295–1307. doi:10.1016/j.worlddev.2011.12.00.

49.

Rehman

F. U.

, Khan

M. A.

, Khan

M. A.

, Pervaiz

, & Liaqat

(2019). The causal, linear and nonlinear nexus between sectoral FDI and infrastructure in Pakistan: Using a new global infrastructure index. Research in International Business and Finance, Volume , 52, 1–15. https://doi.org/10.1016/j.ribaf.2019.101129

50.

Rehman

F. U.

, & Ding

(2019). The nexus between outward foreign direct investment and export sophistication: New evidence from China. Applied Economics Letters , 27(5), 357–365. doi:10.1080/13504851.2019.161605.

51.

Rehman

F. U.

, Noman

A. A.

, & Yibing

(2020). Does infrastructure increase exports and reduce trade deficit in selected South Asian countries? By using a new global infrastructure index. Journal of Economic Structures , 9(1), 10. https://doi.org/10.1186/s40008-020-0183-x

52.

Rehman F.

, & Khan

(2015). Determinants of food price inflation in Pakistan: An econometric analysis. Advances in Economics and Business , 3(12): 571–576. doi:10.13189/aeb.2015.03120.

53.

Roller

L. H.

, & Waverman

(2001). Telecommunications infrastructure and economic development: A simultaneous approach. American Economic Review , 91: 909–923. doi:10.1257/aer.91.4.90.

54.

Straub

(2011). Infrastructure and development: A critical appraisal of the macro-level literature. Journal of Development Studies , 47, 683–708. doi:10.1080/00220388.2010.50978.

55.

, & Qi

(2016). Transport connectivity and regional development in China. Journal of International Logistics Trade , 14, 142–155.

56.

, Pervaiz

, Khan

M. A.

, Rehman

F. U.

, & Oláh

(2019). On the asymmetries of sovereign credit rating announcements and financial market development in the European region. Sustainability , 11, 6636. doi:10.3390/su1123663.

57.

Njinkeu

, John Wilson

, & Fosso

B. P.

(2008). Expanding trade within Africa: The impact of trade facilitation. Policy Research Working Paper No. 4790. World Bank. https://openknowledge.worldbank.org/handle/10986/6297

58.

Volpe

M. C.

, Carballo

, & Cusolito

(2017). Roads, exports and employment: Evidence from a developing country. Journal of Development Economics , 125, 21–39. doi:10.1016/j.jdeveco.2016.10.00.

59.

Wan

, & Zhang

(2017). The direct and indirect effects of infrastructure on firm productivity: Evidence from Chinese manufacturing. China Economic Review , 49, 143–153. doi: 10.1016/j.chieco.2017.04.010

60.

Yeaple

, & Golub

S. S.

(2002). International productivity differences, infrastructure and comparative advantage (mimeo). University of Pennsylvania.