Abstract
Purpose
This study attempts to recognize the effects of the Pakistan–China free trade agreements (PCFTA) on promoting trade between the two economies.
Design/methodology/approach
Following the concept of revealed comparative advantage (RCA) and free trade agreements, the study first identifies those commodities in which Pakistan and China have a robust RCA and then analyze the effect of PCFTA on the export value of those commodities for the bilateral trade between Pakistan and China. The study used the panel data in which more than the top 150 importers (j) have been selected for each case of Pakistan and China for the period of 2003–2015.
Findings
The study concludes that even with the higher convergence rate, the good RCA does not guarantee a positive effect of the free trade agreement on the commodities.
Originality/value
The study contributes to the existing literature by integrating RCA with the gravity model by adopting a sequential mode for Pakistan–China free trade agreement.
Keywords
Citation
Ahmed, A., Nazeer, N., Zahid, G. and Nawaz, F. (2023), "Does revealed comparative advantage matter in the gravity of FTAs?", Journal of International Logistics and Trade, Vol. 21 No. 2, pp. 84-107. https://doi.org/10.1108/JILT-06-2022-0018
Publisher
:Emerald Publishing Limited
Copyright © 2023, Arsalan Ahmed, Nazia Nazeer, GulRukh Zahid and Faisal Nawaz
License
Published in Journal of International Logistics and Trade. Published by Emerald Publishing Limited. This article is published under the Creative Commons Attribution (CC BY 4.0) licence. Anyone may reproduce, distribute, translate and create derivative works of this article (for both commercial and non-commercial purposes), subject to full attribution to the original publication and authors. The full terms of this licence may be seen at http://creativecommons.org/licences/by/4.0/legalcode
Introduction
In 2021, the total value of trade among the countries was US $22.3tr which was just US $3.2tr at the start of the 21st century [1]. It shows that the trend of trade liberalization spread substantially over the last two decades throughout the world. However, this trade liberalization arises through various channels. Among them, the two major categories are multilateralism and bilateralism. Fundamentally, multilateral trade agreements are between three or more countries, whereas bilateral trade agreements consist of between two countries. These countries engaged in trade agreements while considering the contract of General Agreement on Tariffs and Trade (GATT), which requires member countries to undertake trade liberalization on a most favoured nation (MFN) or nondiscriminatory basis, and permits a subset of World Trade Organization (WTO) members to pursue free trade agreements (FTAs) under which they can grant concessions to each other that they do not have to extend to others. According to the world trade organization, there are around 420 regional trade agreements in force globally. All these statistics highlight the colossal significance of trade liberalization, especially in the modern world. However, the theories about trade liberalization have been introduced in the literature long ago. The concept of trade openness (TOE) has been highlighted by many researchers in the existing literature, such as Adam Smith's (1776) pioneer work in “The Wealth of Nations,” Richardo's (1817) idea of comparative advantage, Factor endowments, the Heckscher-Ohlin theory (1932), Leontief's (1953) patchy conclusion on factor abundant, and Vernon (1966) product life cycle theory. Among them, the majority of the studies focused directly or indirectly on the revealed comparative advantage (RCA). Yet; there is a lacking of amalgamation of the other theories of the free trade agreements with the theory of RCA. For instance, the gravity model has had a very significant impact on the empirical analysis of trade for more than three decades too. It generally explained the trade relationship between the countries based on their economic size, connectivity and distance among them. Although the gravity model holds a high acceptance among economists with some altercations, it ignores the comparative advantage between countries which is still one of the most important concepts in understanding international trade as it includes the trade structure and patterns of the country. Specifically, when the gravity model is used for the policy implication of trade potential and how to reduce the resistance of trade between the country, it should consider the comparative advantage because the countries which have complementary patterns of comparative advantage have high trade potential as compared to the countries which have the similar patterns of comparative advantage. Therefore, by combining the gravity model with the comparative advantage, this limitation of the gravity model can be eliminated. In addition to that, rather than apply the gravity model and comparative advantage together in one equation, this study suggests taking a two-filter approach. The intuition behind this approach is that each of these theories has its own set of assumptions, therefore it is better that firstly we filter out the trade between the countries with the comparative advantage and then quantify the effect of free trade garment on the trade between the countries by gravity model. Based on this idea, the study analyzed Pakistan–China Free Trade Agreement (PCFTA). Such that, this study first identifies the commodities in which the two counties have a RCA separately and then analyzes the effect of PCFTA on the export value of those commodities for the bilateral trade between these two countries.
Pakistan and China approved a free trade agreement to encourage development plus diversification of trade between the two economies at the end of 2006. This agreement was divided into two phases: Phase I, already concluded in 2011, China has lessened tariffs on almost 6,418 product lines, and Pakistan also presented the parallel on 5,686 product lines for five years (Table 1). China and Pakistan agreed to reduce the tariff gradually between 2007 and 2011. The tariff concessions given by China are on finished goods, while Pakistan commented on tariff reduction on raw materials and intermediary goods. Phase II of the PCFTA covering 2019–2024 was confirmed between the two economies at the beginning of 2019 and started execution on January 1, 2020. Though several studies have been done earlier on Pakistan–China FTA and provided the insides related to the topic, these studies were mainly conducted before the complete implementation of the agreement.
Based on the objectives, this study frames the following two hypotheses.
PCFTA positively affects the export value of commodities from Pakistan to China in which Pakistan has a higher RCA.
The export value of commodities from China to Pakistan in which China has a higher RCA increases substantially due to PCFTA.
This study is novel in its idea, as it contributes to the existing literature by combing the RCA and gravity model in the context of PCFTA. The study first identified the commodities in which both nations have higher RCA separately. Afterward, it will apply the gravity model to find the effect of the Pakistan–China FTA on those commodities. The paper is arranged in the manner of introduction, literature review, material and method, results and analysis, discussion and conclusion (ILMARD). The following section will be a literature review, which discusses the previous studies in detail. Then, the section on material and method presents the proposed methodology with the model specification, data description and estimation techniques. Later, the section on analysis and results provides the estimated results and analysis based on the results, and the last section discusses the conclusion, presents a summary of the study, and proposed policy recommendations.
Literature review
In the previous literature, several studies highlighted the notion of RCA such as Maqbool et al. (2020), Elsa and Gianluca (2014), Laursen (2015), Torok and Jambor (2016), Abbas et al. (2017), Mahboob et al. (2017), Deb and Sengupta (2018), Jagdambe (2019), Fakhrudin et al. (2019), Siddique et al. (2020). Maqbool et al. (2020) measured the cereal export competitiveness of Pakistan in the global economy by utilizing different indices of RCA and suggested that Pakistan needs to concentrate on the production and exports of cereals. Elsa and Gianluca (2014) present a new database of RCA measures based on the recent index of RCA proposed by Costinot et al. (2012). This study presents empirical distribution features of the new index in comparison with the traditional Balassa Index. Similarly, Laursen (2015) demonstrates an analysis of Balassa's ‘revealed comparative advantage (RCA), and the proposed adjusted index is called ‘revealed symmetric comparative advantage’ (RSCA). Also, Torok and Jambor (2016) analyzed RCAs in European ham trade from 1999 to 2013 and identify its determinants by using the panel data econometrics. Moreover, Jagdambe (2019) explored the competitiveness of India's agricultural products in world markets through four indices of RCA at the four-digit level of the harmonized system (HS) of classification between 1996 and 2015. Their study proposed an ordinal interpretation of RCAs indices for better policy formulation. Also, Siddique et al. (2020) estimates the effects of devaluation and appreciation of real effective exchange rate (REER) on RCA at HS 2-digit level of exports in Pakistan and concluded that an increase in selected RCA's index values, world aggregated income and REER depreciation is useful to decrease in the deficit trade balance of Pakistan. The main focus of all these studies is to analyze the RCA, but are failed to amalgamate the other theories of the free trade agreements with the theory of RCA.
Moreover, several studies (Wilson and Mann, 2003; Longo and Sekkat, 2004; Paas and Tafenau, 2005; Kang and Fratianni, 2006; Kalirajan, 2007; Nowak-Lehmann et al., 2007; Abedini and Péridy, 2008; Grant and Lambert, 2008; Henderson and Millimet, 2008; Kepaptsoglou et al., 2010; Sheng et al., 2012; Tang et al., 2013; Barattieri, 2014; Narayan and Nguyen, 2016; Timsina and Culas, 2021) have considered the gravity model to explore the relations between the various economic-related variables and international trade. For instance, Wilson and Mann (2003) analyze the relationship between trade facilitation and trade flows in the Asia Pacific region with the gravity model. Similarly, Grant and Lambert (2008) finds that regional trade agreement (RTA) effects on member's trade depend substantially on different sectors and the length of the phase-in period. Also, Henderson and Millimet (2008) estimate gravity models in levels and logs using two data sets via nonparametric methods and stated that concerns in the gravity literature over functional form appear unwarranted, and estimation of the gravity model in levels is recommended. Moreover, Narayan and Nguyen (2016) demonstrate that the influence of trade gravity variables is dependent on trading partners and suggests that the trade policy should not ignore the other importance of the level of development and maintenance of close economic ties. Moreover, Timsina and Culas (2021) estimate the trade creation and export diversion effects of different Australia's free trade agreements (FTAs) using the panel data from 1996 to 2017. Still, all these studies are also short of combining RCA with the gravity model.
In addition, the existing literature exhibits few studies only that ponder to connect the RCA with the gravity model like Abdul Kamal et al., 2021; Ariza Marin, 2020; Ciuriak and D.and Kinjo, 2006; Guglielmo et al., 2015; Shakur and Tsang, 2017. However, there is a gap in the existing literature vis-à-vis the approach to integrate RCA with the gravity model. All these studies interleaved the indicator of RCA into the gravity equation instead of adopting a sequential mode, which will be more suitable to perform and join the RCA and gravity model together.
Furthermore, the works of literature about the Pakistan–China FTA remain inclusive of the effect of this FTA on Pakistan and China. Some studies suggest that Pakistan–China FTA is equally beneficial for both Pakistan and China (Kataria and Naveed, 2014; Farrukh et al., 2013; Junaid and Manzoor, 2017; Kataria and Naveed, 2014; Irshad et al., 2016; Liu, 2010; Kabraji, 2012; Ahmed et al., 2021a, b). According to Kataria and Naveed (2014), Pakistan is the first country located in South Asia that have a free trade agreement with China. Farrukh et al. (2013) found out that PCFTA strengthens the economic relations of China–Pakistan. Junaid and Manzoor (2017) studied the dynamic of Pakistan's trade balance with China and advocate that PCFTA is a good opportunity to exploit the huge Chinese market. Similarly, Hamid and Hayat (2012) compared Pakistan's trade with China and other neighbors and proposed policies concerning China including gaining market access for Pakistan's exports and attracting Chinese investment to the export industries. Furthermore, Irshad et al. (2016) extend the free trade agreement between Pakistan and China with the context of the The Association of Southeast Asian Nations (ASEAN) free trade agreement. The study concluded that Pakistan's integration with China and ASEAN would be beneficial for enhancing the bilateral trade flows of all the members. In contrast, other studies argued that Pakistan–China FTA is more beneficial to China as compared to Pakistan (Musleh-ud-Din et al., 2009; Chaudhry et al., 2017; Qi and Irshad, 2015; Shaista, 2010; Ahmad, 2014). Musleh-ud-Din et al. (2009) argue that the short-term effect of the FTA between Pakistan and China is tilted towards China. Chaudhry et al. (2017) found out that in the sector in which the tariffs are reduced by Pakistan for Chinese commodities, productivity and the value-added per worker have fallen relative to other sectors. Qi and Irshad (2015) argued that from a Pakistan perception, FTA covering only goods trade will not be valuable to Pakistan. Shaista (2010) empirically examines the trade agreements of Pakistan with China, Sri Lanka, Malaysia, Iran, Mauritia and the South Asian Free Trade Area. The study found some evidence for product diversification under the PTAs with Malaysia and Mauritius, however in the case of Sri Lanka and China product diversification declined. Ahmad (2014) shows that FTA has enhanced bilateral trade but it has largely benefitted Chinese exporters. Hence, there is a substantial gap in the literature about the analysis of free trade agreements through RCA with the Gravity model. Also, most of the studies on PCFTA concentrated on finding whether PCFTA was equally advantageous for both nations or not. Therefore, this study will attempt to fill this gap. In the following section, the study will focus on the methods used for the estimation.
Material and method:
Revealed comparative advantage
Rendering to the international trade theory, trade advantages come when countries specialize in producing goods for which they have a comparative advantage. The RCA index can discover the products in which a country has a trade potential. It can also give helpful evidence about potential trade possibilities with new cohorts. The RCA index of a nation for commodity j is measured by the commodity's share in the nation's total exports concerning its share in world trade. The formula is:
Table 3 shows the RCA results of Pakistan and China for different groups of commodities. The results show that Pakistan has the RCA in the commodity group of HS code 06-15, 41-43, 50-63. While, China has the RCA in the commodity group of HS codes 41-43, 50-63, 64-67, and 84-85. Among all the measured commodities, Pakistan's RCA for the commodity group of HS codes 50-63 and 41-43 have comparatively high values compared to China's RCA.
Before going further to perform the empirical test, it is imperative to ascertain whether these commodity groups mentioned in Table 3 are included in the Pakistan–China FTA or not. Therefore, to identify the commodities included in Pakistan–China FTA, this study summarizes the Pakistan–China FTA convergence rate.
Convergence rate
This study quantifies the convergence rate (CR) for those commodities, which showed a higher RCA for Pakistan and China separately.
Table 4 demonstrates the convergence rate results for China and Pakistan separately for different HS commodity sections. While considering Pakistan, the results indicate that there are 7,381 different kinds of commodities at HS 8 digits imported by Pakistan from China, in contrast only 4,808 various commodities at HS 8 digits lines are added in the PCFTA, so the convergence rate of Pakistan–China free trade agreement for imports by Pakistan from China during last year is 0.65. It shows that Pakistan cannot utilize at least one-third of the items (included in the FTA) as Pakistan does not export these commodities to China.
Besides, the results of the convergence rate for China illustrate that there are 6,219 different kinds of commodities at HS 8 digits lines imported by China from Pakistan, and almost 4,585 different kinds of commodities at HS 8 digits lines are added in PCFTA. Hence, the convergence rate for the exports of China to Pakistan is nearly 0.74. Thus, around one-fourth of the commodities for the case of China, and one-third of the commodities for the case of Pakistan have not been included in the tariff relaxation in the PCFTA for bilateral trade.
In addition, for the commodity sections 01-05, 06-15, 41-43 and 50-63 in which Pakistan has RCA, the convergence rates are 0.45, 0.68, 0.83 and 0.70 respectively. Similarly, for the commodity sections 41-43, 50-63, 64-67, 72-83 and 84-85 in which China has RCA, the convergence rates are 0.62, 0.80, 0.32, 0.72 and 0.72 respectively. It shows that these commodities sections do not have a very high convergence rate in RCA commodities, which could possibly dilute the positive effect of Pakistan–China free trade agreement on the bilateral trade between Pakistan and China.
Gravity model and FTA
The gravity model of trade proposes trade as a positive function of gross domestic product (GDP) and as a negative function of distance for the two economies. Such that,
Given the hypothesized relationships in the gravity model,
In the literature, Tinbergen initially used the gravity model in 1962 to describe international trade configurations, and economists have constantly instituted it to justify a considerable proportion of the variant in global trade flows, constructing the model preferring to test the marginal influence of other hypothesized variables on international trade. Later, the work of Linnemann (1966), Anderson (1979), and Deardorff and Stern (1998) have been given the theoretical rationalization for the gravity model in the literature.
Gravity models can also be applied to measure FTAs by introducing a binary variable in the baseline specification. It can determine whether or not an FTA has had a statistically significant effect on trade flows using this variable. If it is found positive and significant, one can assume that FTA has had a positive impact on trade flows, along with a magnitude associated with the size of its coefficients. The study used the following two gravity equations for the estimation. Equation (1) illustrates the exports of China, while equation (2) depicts the exports of Pakistan. The study used panel data in which more than the top 150 importers (j) have been selected for each case (Pakistan and China), and the period is 2003–2015. These equations are further used in each commodity section for detailed analysis.
Speciation in the gravity model
Cross-sectional data is often considered to calculate bilateral trade flows via the gravity model for a specific time. However, it is not easy to justify it over panel data; Baltagi and Kao (2001), Arellano (2003), Hsiao (2005), Nerlove (2005) and M'aty'as and Sevestre (2008). In contrast, panel data evaluation indicates more advantages over time series and cross-sectional data due to its control nature for discrete heterogeneity, data availability, increased efficiency of econometric estimations by minimizing collinearity among independent variables through a significant degree of freedom, and challenging methodology.
Given the scenario, where the assumptions of autocorrelation and homoscedasticity are rarely converged in analysis, the expected error component over individuals causes correlation across the composite error terms, inducing odinary least square (OLS) estimation ineffective. It calls for considering feasible generalized least squares (GLS) measures for fixed and random effects (Martinez-Zarzoso, 2013). In the same way, while estimating the trade data with the gravity model, the two fundamental difficulties that need to deal with are log modeling and zero trade flows. To resolve these issues Poisson pseudo-maximum likelihood (PPML) estimator could be considered the best fit. In the presence of heteroscedasticity, the PPML approach is a robust tactic, and this approach can be used for the different levels of trade, hence estimating the nonlinear form of the gravity model directly and avoiding falling zero trade. The dependent variable is “exports,” but not the log function of “exports”) whereas, the explanatory variables may still be in log forms (Santos Silva and Tenreyro, 2006; Soren and Bruemmer, 2012; Head and Mayer, 2014; Eaton et al., 2012). We used panel data from 2003 till 2015 and used both the GLS method and PPML for the analysis. Also, to choose between random and fixed effect model the study adopted the Hausman test, basis on the previous studies (Behera and Sethi, 2022; Ali et al., 2017; Rahman and Miah, 2017; Ayoub, 2019; Uddin et al., 2020). The study set the period from 2003–2015 because it covers the three major subperiod, before PCFTA, during PCFTA implementation and After PCFTA implementation was completed. Correspondingly, it did not exceed the period after 2015 to avoid the effects of other factors. Moreover, the PPML method constructs the baseline results, and the GLS model will be employed to run the robustness test.
Result and Analysis
Empirical results for the case of China
Table 5 reports the aggregate results of the gravity equation for exports of China. The R-squared value for the PPML model is 0.96, indicating the model is highly significant. Moreover, the findings of the Hausman test indicate that the fixed effect model is preferred over random effect models whenever the result of both fixed and random effect models are statistically significant. However, when the coefficient of fixed effect is insignificant then the study considers the coefficient of random effect. Keeping all the other variables constant, the gross domestic product GDP is statistically highly significant, and the value is nearly 0.69. This result implies that a one per cent growth in the GDP of trading countries will increase China's exports by 0.69%. The finding is in line with the primary hypotheses of the gravity model, which states that trade volume will rise with an increase in economic volume.
Moreover, the findings of the Hausman test indicate that the fixed effect model is preferred over random effect models. Likewise, the variable distance is also statistically highly significant and showed a negative sign. The movement of the distance variable is consistent with the gravity model and implies that the result is in line with the primary hypotheses of the gravity model, which states that geographical distance is an imperative resistance factor in trade flow. The absolute difference in per capita GDP (GDPPCDIFijt) exhibits the technological inequalities between China and other trading countries. This variable is statistically significant in the PPML (random effect model (REM)) model and positive. The positive sign indicates that China's exports to its trading countries do not follow the Linder hypothesis, which states that the country exports more to those economies that have the same level of income per capita as China. The variable TOE is statistically highly significant, and the positive sign implies that the TOE significantly contributes to the exports of China. Similarly, the variable exchange rate is statistically significant and positive, this finding translates that the deprecation of currency benefits the total exports of China. Also, the dummy variable of COL, CON and landlock is highly significant in the PPML model. The expected signs of these dummy variables are consistent with the theory. The coefficients of dummy variables used for ASEAN-FTA and Pakistan–China FTA are found to be insignificant.
Subsequently, the study utilized the gravity model for different sectors at disaggregate levels to examine the results of each HS commodity separately. Table 6, reports the disaggregate results of exports of China for HS commodity sections. The findings of R-square value are higher in all PPML models for HS commodity sections 41-43, 50-63, 64-67, 72-83, 84-85, which demonstrates that the overall PPML model is highly significant to explain the results accurately. The result for the variable of gross domestic product and distance is statistically significant and exhibits positive and negative signs, respectively. This outcome implies that the sign of these variables is consistent with the gravity model.
Moreover, the absolute difference in the per capita GDP (GDPPCDIFijt) variable demonstrates the technological inequalities between China and trading countries. This variable is statistically significant in 50-63, 64-67, 72-83 and 84-85 but insignificant for 41-43. Also, the positive sign of this variable implies that China's exports to its trading countries do not follow the Linder hypothesis. Similarly, the variable of TOE for China is statistically highly significant in the PPML fixed effect model with a positive sign in all the HS commodity sections except 41-43, which suggests that the TOE of China's economy is significantly contributing to the exports of China.
Furthermore, the dummy variables for Pakistan–China FTA are statistically significant in the PPML fixed-effect model for HS commodity sections 40-41 and 50-63 with a positive sign. Also, the results suggest that due to PCFTA the exports of China to Pakistan are 43.23% in the HS commodity section 40-41 and 32.08% in the HS commodity section 50-63. In addition, the dummy variables for Pakistan–China FTA are found to be statistically insignificant for the HS commodity sections 64-67, 72-83 and 84-85. Interestingly, the coefficient of the dummy variable for ASEAN-FTA is not only found significant in these HS commodities, but the values of the coefficients are very high. It indicates that although China has the comparative advantage in these HS commodity sections and Pakistan and China reduced the tariff in these commodity sections significantly through the Pakistan–China FTA, the exports of these commodities are more toward ASEAN countries rather than Pakistan. Therefore, even if a country has the RCA and signs a free trade agreement, the exporters will be more focused on the countries that have a more developed free trade agreement, as the China-ASEAN FTA is more mature than the Pakistan–China FTA in term of the implementation period and coverage of commodities. Also, this direction of research can be explored more in future studies.
Empirical results for the case of Pakistan
Table 7 reports the aggregate results of the gravity equation for exports of Pakistan. The R-Square value for the PPML model was observed at 0.96, which specifies that the overall model is highly significant and fit for further analysis. Also, the Hausman test indicates that the fixed-effect model is desirable over random-effect models. The variable gross domestic product, keeping other variables constant, is a statistically highly significant variable. It suggests that if the GDP of Pakistan rises by 1%, the exports of Pakistan will upsurge by 0.68%. The results demonstrate the adequacy of the method for the gravity model, which states that trade volume will grow with a surge in economic volume. Similarly, the coefficient of the distance variable is statistically significant and exhibits an expected negative sign, which illustrates that this result of the distance variable is also consistent with the gravity model. The results of the methods appear to tally with our expectations of the primary hypotheses of the gravity model, which states that geographical distance is an imperative factor in the resistance of trade flow. Additionally, The absolute difference in per capita GDP (GDPPCDIFijt) exhibits the technological inequalities between Pakistan and the trading country. This variable is statistically insignificant in the estimated model of PPML (fixed effect). It indicates that this study cannot conclude anything about whether Pakistan's exports to its trading countries follow the Linder hypothesis or not. In contrast, the variable TOE is statistically highly significant, and the positive sign implies that the TOE significantly contributes to the exports of Pakistan. Moreover, the estimates of the COL, CON, and LL are not significant for HS 01-05, and significant for 06-15, 41-43, and 50-63. Also, the signs of the COL and LL are different across HS code groups, such that it is positive for 06-15 but negative for 41-43 & 50-63. Hence, not all signs of these variables are consistent with the theory. The dummy variable Pakistan–China FTA is statistically significant and demonstrates a positive sign. The coefficient of the variable suggested that Pakistan–China FTA increases the exports from Pakistan to China by 61.60%.
Subsequently, the study also applies the gravity model at different sectoral-disaggregated levels to examine the results of each HS commodity separately. Table 8 reports the disaggregated results of exports of Pakistan for HS commodity sections 01-05, 06-15, 41-43 and 50-63, respectively. The result of the R-square is high in the PPML model for all HS commodity sections. This means that the overall PPML model is highly significant to explain the results accurately. First of all, the coefficient of GDP is found to be statistically significant and positive in all four HS commodity sections, which is consistent with the theory of gravity model. In comparison among these HS commodity sections; the effect of GDP is highest in 06-15. Similarly, the results indicate a negative sign of distance's variable statistically significant coefficient in all four HS commodity sections, which is consistent with the standard hypothesis of the gravity model. Meanwhile, the negative effect of distance is highest in 01-05 as compared to the other three HS commodity sections. The reason for the highest coefficient is that 01-05 are highly sensitive agriculture products that have a low shelf life. Therefore, longer distance creates higher exports cost for these products.
The absolute difference in the per capita GDP (GDPPCDIFijt) variable demonstrates Pakistan's technological inequalities concerning its trading countries. The coefficient of this variable is positive and statistically significant in all the HS commodity sections except 06-15. The positive sign of this variable implies that Pakistan exports more to those economies which do not have a similar level of income per capita as Pakistan for HS commodity sections or it translate that Pakistan export to its trading countries but does not follow the Linder hypotheses. Moreover, the variable of the population is also statistically significant in PPML fixed-effect models in all four HS commodity sections. However, the sign of this coefficient is only positive in 01-05. It indicates that the higher population of the trading partner did not benefit the Pakistani exports except in the agricultural products. The reason behind this is that agricultural products are the staple products for any country, therefore a higher number of people could lead to higher demand for agricultural products, hence the high population benefits the exporters only in this HS commodity section. Moreover, for Pakistan, the dummy variable Pakistan–China FTA displays a positive sign for HS commodity code sections 06-15, 50-63, implying export growth through Pakistan–China FTA by Pakistan in these commodities. The coefficient of the variable suggested that Pakistan–China FTA increases the exports from Pakistan to China by 366.45% in 06-15 and 90% in 50-63. However, the dummy variable for Pakistan–China FTA confirms a negative sign for HS commodity sections 41-43 and 01-05. These results demonstrate that even with the higher RCA it could not guarantee an increase in the exports of trade between the two countries after signing the free trade agreement. This result is very interesting as it indicates that free trade agreements should look beyond the fundamental trade indicator like comparative advantage between the countries. In the future, dual margins of exports can also be explored to extend the results of this study.
Discussion and conclusion
The free trade agreements allow economies to provide a favored mode for trade between the partner economies within a specific period. This study intends to fill the gap in the literature on trade liberalization by integrating RCA with the gravity model. As mentioned earlier in the section literature review, previous studies (Abdul Kamal et al., 2021; Ariza Marin, 2020; Ciuriak and D.and Kinjo, 2006; Guglielmo et al., 2015; Shakur and Tsang, 2017) interleaved the indicator of RCA into the gravity equation instead of adopting a sequential model, which will be more suitable to perform and join the RCA and gravity model together for the FTA agreement. The objective of this study was to identify the commodities in which the two countries have a robust RCA separately and then analyze the effect of FTA on the export value of those commodities for the bilateral trade between the two countries.
The study concludes that the tariff arrangement between Pakistan and China in PCFTA shows a visible improvement in the bilateral trade between the two countries. However, the empirical analysis and its findings did not support the study's hypothesis. Although PCFTA substantially increases commodities export with higher RCA. This statement is valid only for some commodity groups. There is a commodity group in which Pakistan holds an RCA, but they did not positively affect the PCFTA. Similarly, the study observed the same conclusion for China; therefore, this study rejects the null hypothesis.
Interestingly, these conclusions open up a new dimension of research for free trade agreements, suggesting that free trade agreements are more complicated than export between two countries. Moreover, the relevance of RCA while establishing a free trade agreement in the context of the current economic system is also questionable. One can argue that the effect of RCA could be invalidated because the commodities included in the FTA do not have a decent convergence rate in the agreements. Nevertheless, this study has analyzed it and concluded that even with the higher convergence rate, the good RCA does not guarantee a positive effect of the free trade agreement on the commodities. One plausible argument for these results could be the asymmetries of countries participating in the FTA. As the exports and import value of one country is much bigger than another country, the effect of FTA on RCA commodities could not be transferred properly for each country in the free trade agreement. Therefore, this could be used as a direction for future studies, such that to explore the same study for symmetric countries.
In light of the findings of this study, it is recommended that countries should look deeper into their trading agreements for increasing bilateral trade. The exports between the countries can be amplified if they go beyond the traditional method of including the commodities in the free trade agreements. Furthermore, to accelerate export by using improved tariff reductions, exporters should now have to focus on tackling nontariff barriers to export; particularly capability issues in the business community. For that purpose, the government has to improve the regulating system and provide a conducive business environment for local businesses too, as these factors affect the capability of delivering orders of scale on time to exporting market. Finally, this study focuses on commodity-level data; future studies can consider firm-level data and test the theoretical assumption of this study and extend the findings in the context of dual margins of trade.
Tariff reduction modality of China and Pakistan
Category no. | Track | Tariff reduction modality of China | Tariff reduction modality of Pakistan | ||
---|---|---|---|---|---|
No. of tariff lines | % of tariff lines at eight-digit | No. of tariff lines | % of tariff lines at eight-digit | ||
I | Elimination of tariff (Three years) | 2,681 | 35.5% | 2,423 | 35.5% |
II | 0–5% (five years) | 2,604 | 34.5% | 1,338 | 19.9% |
III | Reduction in margin of preference of 50% (five years) | 604 | 8% | 157 | 2% |
IV | Reduction in margin of preference from 20% (five years) | 529 | 7% | 1,768 | 26.1% |
V | No concession | 1,132 | 15% | 1,025 | 15% |
VI | Banned item for trade | 92 | 1.4% |
Source(s): Authors' estimations based on the data at the website of the Ministry of commerce, China
Description of HS codes by sections
Group | Sections | HS codes | Description |
---|---|---|---|
1 | SECTION I | 01–05 | Live animals |
2 | SECTIONS II AND III | 06–15 | Vegetable products |
3 | SECTION IV | 41–43 | Animal or vegetable fats and oils |
4 | SECTION XI | 50–63 | Textile and textile articles |
5 | SECTION XII | 64–67 | Footwear, headgear, umbrellas, sticks, whips, riding-crops |
6 | SECTION XV | 72–83 | Base metals and articles of base metal |
7 | SECTION XVI | 84–85 | Mechanical appliances and electrical equipment |
Source(s): World Customs Organization (https://www.wcotradetools.org/en/harmonized-system)
Revealed comparative advantages of Pakistan and China
Years | Pakistan's RCA | China's RCA | |||||||
---|---|---|---|---|---|---|---|---|---|
01–05 | 06–15 | 41–43 | 50–63 | 41–43 | 50–63 | 64–67 | 72–83 | 84–85 | |
2003 | 0.80 | 3.06 | 7.35 | 11.78 | 2.73 | 2.32 | 3.45 | 0.72 | 1.13 |
2004 | 0.66 | 2.99 | 7.91 | 12.25 | 3.12 | 2.75 | 4.02 | 0.99 | 1.47 |
2005 | 0.71 | 4.16 | 9.00 | 12.66 | 2.96 | 2.80 | 4.02 | 1.00 | 1.53 |
2006 | 0.84 | 4.16 | 9.19 | 13.42 | 2.48 | 2.99 | 3.80 | 1.07 | 1.56 |
2007 | 0.85 | 3.86 | 9.87 | 12.98 | 2.20 | 2.94 | 3.59 | 1.07 | 1.61 |
2008 | 0.97 | 5.13 | 10.27 | 12.56 | 2.32 | 3.01 | 3.75 | 1.18 | 1.72 |
2009 | 1.03 | 4.20 | 8.37 | 11.86 | 2.40 | 2.90 | 3.69 | 0.91 | 1.73 |
2010 | 1.15 | 4.53 | 7.89 | 12.42 | 2.41 | 2.90 | 3.71 | 0.93 | 1.74 |
2011 | 1.25 | 5.18 | 7.41 | 12.57 | 2.58 | 2.98 | 3.74 | 1.01 | 1.79 |
2012 | 1.52 | 4.42 | 7.38 | 12.82 | 2.48 | 2.92 | 3.81 | 1.04 | 1.80 |
2013 | 1.49 | 4.51 | 7.50 | 12.63 | 2.33 | 2.88 | 3.61 | 1.06 | 1.80 |
2014 | 1.41 | 4.36 | 7.70 | 12.35 | 2.21 | 2.72 | 3.41 | 1.14 | 1.70 |
2015 | 1.60 | 4.31 | 6.98 | 12.13 | 2.13 | 2.49 | 3.09 | 1.14 | 1.60 |
2016 | 1.57 | 3.84 | 7.00 | 12.41 | 2.17 | 2.49 | 2.85 | 1.13 | 1.58 |
Source(s): Author's calculations based on data from trade map
Convergence rate for Pakistan and China
HS code | Pakistan | China | ||||
---|---|---|---|---|---|---|
FTA | MFN | C.R | MFN | FTA | C.R | |
01-05 | 141 | 353 | 0.40 | 155 | 69 | 0.45 |
06-15 | 226 | 488 | 0.46 | 304 | 207 | 0.68 |
16-24 | 135 | 304 | 0.44 | 231 | 192 | 0.83 |
25-27 | 138 | 157 | 0.88 | 151 | 135 | 0.89 |
28-38 | 882 | 1,210 | 0.73 | 1,067 | 935 | 0.88 |
39-40 | 214 | 273 | 0.78 | 278 | 139 | 0.50 |
41-43 | 65 | 105 | 0.62 | 52 | 43 | 0.83 |
44-49 | 39 | 259 | 0.15 | 218 | 152 | 0.70 |
50-63 | 909 | 1,141 | 0.80 | 886 | 622 | 0.70 |
64-67 | 23 | 71 | 0.32 | 51 | 39 | 0.76 |
68-71 | 152 | 250 | 0.61 | 238 | 172 | 0.72 |
72-83 | 547 | 764 | 0.72 | 714 | 524 | 0.73 |
84-85 | 908 | 1,228 | 0.74 | 1,150 | 854 | 0.74 |
86-89 | 169 | 336 | 0.50 | 228 | 57 | 0.25 |
90-97 | 260 | 442 | 0.59 | 496 | 445 | 0.90 |
All | 4,808 | 7,381 | 0.65 | 6,219 | 4,585 | 0.74 |
Source(s): Author's calculations based on data from trade map
Exports of China (aggregate level)
Variables | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM |
---|---|---|---|---|
Ln | 0.7771*** | 0.7947*** | 0.7366*** | 0.6968*** |
(0.0164) | (0.0150) | (0.0271) | (0.0165) | |
0.0328** | 0.0270 * | 0.1639*** | 0.0184 | |
(0.01543) | (0.0152) | (0.0257) | (0.0155) | |
0.6964*** | 0.6635*** | 0.9462 *** | 0.7694*** | |
(0.07683) | (0.0732) | (0.1585) | (0.0595) | |
0.2934*** | 0.1975*** | 0.3722*** | 0.0125 | |
(0.0446) | (0.0334) | (0.0655) | (0.0461) | |
(omitted) | −0.1497 | 0.2137*** | −0.4294*** | |
(0.1625) | (0.0352) | (0.0626) | ||
−0.1873 *** | −0.0187 | 0.0297*** | 0.1370*** | |
(0.0475) | (0.0243) | (0.0103) | (0.0487) | |
(omitted) | 0.4132 | 0.4318 *** | −0.7229 | |
(0.3092) | (0.0874) | (0.4393) | ||
−0.2554 | −0.2951* | 0.0678 | −0.0981 | |
(0.1767) | (0.17734) | (0.1006) | (0.0651) | |
(omitted) | 0.9107 | 0.6022*** | 3.1245*** | |
(0.5734) | (0.1015) | (0.7237) | ||
(omitted) | 0.7974*** | −0.0912 | 0.9428 *** | |
(0.2912) | (0.0778) | (0.3884) | ||
(omitted) | −0.1725 | −0.1588 | −0.5002*** | |
(0.2401) | (0.1268) | (0.1656) | ||
0.5043*** | 0.1385*** | 0.2289*** | 0.2386** | |
(0.1190) | (0.0520) | (0.0351) | (0.1029) | |
constant | −38.0443 *** | −24.6884*** | −25.0570*** | |
(4.0265) | (2.2461) | (1.1956) | ||
No. of obs. | 1,146 | 1,146 | 1,146 | 1,146 |
R-sq: within: | 0.8815 | 0.8789 | ||
R-sq: between: | 0.7051 | 0.8052 | ||
R-sq: Overall: | 0.7106 | 0.8107 | 0.9436 | 0.9959 |
Pseudo-log-likelihood: | −1170000000000 | −120000000000 | ||
F-Stats/Wald Chi | F (71,045) = 1,110.99 | Wald χ2 (12) = 7,821.31 | ||
corr (u_i, Xb) | corr (u_i, Xb) = −0.6357 | |||
Prob > F | Prob > F = 0.0000 |
Source(s): Author's own estimations (Here *** = 1%, ** = 5% and * = 10% significance level)
Exports of China (disaggregate level)
HS commodity 41-43 | HS commodity 50-63 | |||||||
---|---|---|---|---|---|---|---|---|
Variables | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM |
Ln | 0.6357*** | 0.6734*** | 0.9417*** | 0.3314*** | 0.6601*** | 0.6614*** | 0.4494*** | 0.6047*** |
(0.0297) | (0.0257) | (0.0771) | (0.0962) | (0.0334) | (0.0292) | (0.0391) | (0.0441) | |
0.0778*** | 0.0851*** | 0.0398 | 0.0128 | 0.0063 | 0.0288*** | 0.2912*** | 0.0681** | |
(0.0268) | (0.0259) | (0.0590) | (0.0352) | (0.0314) | (0.0301) | (0.0415) | (0.0327) | |
0.0177 | −0.1001 | 1.4534*** | 0.0401 | 0.8613*** | 0.6337*** | 0.2533 | 0.7912*** | |
(0.1369) | (0.1275) | (0.3041) | (0.1933) | (0.1519) | (0.1414) | (0.2239) | (0.1461) | |
0.2627*** | 0.2745*** | −0.0192 | 0.0654 | −0.1023 | 0.0265 | 0.0965 | −0.0841 | |
(0.0817) | (0.0537) | (0.0732) | (0.1474) | (0.0889) | (0.0591) | (0.0610) | (0.0820) | |
(omitted) | −0.0090 | −0.0849** | −0.4131** | (omitted) | −0.4187 | −0.3112*** | −0.1606 | |
(0.2139) | (0.0497) | (0.1869) | (0.2285) | (0.0545) | (0.0987) | |||
−0.0584 | −0.0420 | −0.0091 | 0.1614 | 0.1590* | 0.0477 | 0.0109 | 0.1651** | |
(0.0844) | (0.0370) | (0.0162) | (0.1570) | (0.0921) | (0.0414) | (0.0160) | (0.0824) | |
(omitted) | −0.5591 | −0.0158 | 9.7602*** | (omitted) | −0.7211* | 0.2389* | −1.0154 | |
(0.3769) | (0.1231) | (2.6039) | (0.3973) | (0.1207) | (0.5176) | |||
0.0897 | −0.0025 | −1.2550*** | 0.4323* | 0.3465 | 0.3202 | −0.2939** | 0.3208*** | |
(0.2675) | (0.2630) | (0.2510) | (0.2308) | (0.2953) | (0.2917) | (0.1318) | (0.1131) | |
(omitted) | 1.2341* | 1.0086*** | −7.2927*** | (omitted) | 1.0237 | −0.0996 | 2.5888*** | |
(0.6768) | (0.1792) | (2.7310) | (0.7307) | (0.1937) | (0.8864) | |||
(omitted) | 0.7021* | 0.9218*** | −7.0664*** | (omitted) | 0.7468* | 0.2993*** | −1.0917 | |
(0.3762) | (0.1529) | (2.3389) | (0.3976) | (0.1138) | (0.7212) | |||
(omitted) | −0.5837 | −0.5679*** | 3.3554*** | (omitted) | −0.7922* | −0.1780 | 0.1852 | |
(0.3829) | (0.17311) | (0.9332) | (0.4140) | (0.1647) | (0.5932) | |||
0.8970*** | 0.1840** | −0.0904 | 2.1437*** | 0.5135* | −0.0483 | 0.1845*** | 0.2846 | |
(0.2443) | (0.0823) | (0.0872) | (0.6905) | (0.2860) | (0.0953) | (0.0511) | (0.2192) | |
constant | −58.2586*** | −33.4325*** | −35.1194*** | −41.7130*** | −16.6923*** | −17.2597*** | ||
(8.3963) | (3.4478) | (1.7041) | (9.8763) | (3.7910) | (1.7733) | |||
No. of obs. | 836 | 836 | 837 | 837 | 841 | 841 | 841 | 841 |
R-sq: with in: | 0.7579 | 0.755 | 0.6395 | 0.6365 | ||||
R-sq: between: | 0.6098 | 0.7543 | 0.4099 | 0.5509 | ||||
R-sq: Overall: | 0.6074 | 0.7498 | 0.8905 | 0.9685 | 0.4249 | 0.5622 | 0.8199 | 0.9719 |
Pseudo log-likelihood: | −41598268 | −8033518.2 | −398000000 | −64855577 | ||||
F-Stats/Wald Chi | F (7,763 = 341.32 | Wald χ2 (12) = 2,533.53 | F (7,768) = 194.61 | Wald χ2 (12) = 1,397.61 | ||||
corr (u_i, Xb) | corr (u_i, Xb) = −0.6187 | corr (u_i, Xb) = −0.6312 | ||||||
Prob > F = 0.0000 | Prob > F = 0.0000 | Prob > F = 0.0000 |
HS commodity 64-67 | HS commodity 72-83 | |||||||
---|---|---|---|---|---|---|---|---|
Variables | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM |
Ln | 0.6292*** | 0.6984*** | 0.7003*** | 0.5056*** | 1.0601*** | 1.0648** | 0.8114*** | 0.8941*** |
(0.0352) | (0.0309) | (0.0659) | (0.0340) | (0.0263) | (0.0242) | (0.0313) | (0.0325) | |
0.0619** | 0.0757** | 0.1601** | 0.0492 | −0.1091*** | −0.1165*** | 0.0525** | 0.0022 | |
(0.0327) | (0.0317) | (0.0620) | (0.0351) | (0.0238) | (0.0233) | (0.0262) | (0.0251) | |
0.0567*** | −0.2033 | 0.5695 | 0.2201 * | 1.6953*** | 1.7028*** | 1.6172*** | 2.0630*** | |
(0.1593) | (0.1501) | (0.3375) | (0.1226) | (0.1231) | (0.1175) | (0.2480) | (0.1278) | |
0.1358 | 0.1120 | 0.0281 | 0.0420 | 0.1848*** | 0.1064* | 0.6289*** | −0.1317 | |
(0.0929) | (0.0638) | (0.0803) | (0.0897) | (0.0730) | (0.0571) | (0.1138) | (0.0742) | |
(omitted) | 0.0093 | 0.0799 | −0.2123 | (omitted) | −0.7371** | −0.3197*** | −0.2131 ** | |
(0.2681) | (0.0629) | (0.1501) | (0.3471) | (0.0558) | (0.1021) | |||
−0.1467 | −0.0567 | −0.0393** | 0.0513 | −0.1567** | −0.0494 | 0.0631*** | 0.2647*** | |
(0.0956) | (0.0453) | (0.0197) | (0.1018) | (0.0763) | (0.0508) | (0.0130) | (0.0717) | |
(omitted) | −0.6920 | 0.1586 | 5.5513*** | (omitted) | 0.5541 | 0.4381*** | 2.5875*** | |
(0.4705) | (0.1275) | (1.2390) | (0.6041) | (0.1046) | (0.9331) | |||
−0.4934 | −0.5137 | −0.8883*** | −0.1736 | 0.0714 | 0.0505 | 0.2548 | 0.2437 | |
(0.3097) | (0.3096) | (0.2202) | (0.1793) | (0.2414) | (0.2401) | (0.1601) | (0.1454) | |
(omitted) | 1.0132 | 0.5295** | −2.3813*** | (omitted) | 0.3961 | 0.1032 | −0.5998 | |
(0.8450) | (0.2162) | (0.7110) | (1.1079) | (0.1384) | (0.8198) | |||
(omitted) | 0.8437* | 0.5439*** | −4.1696*** | (omitted) | 0.6014 | 0.0384 | 1.6163** | |
(0.4602) | (0.1174) | (0.8779) | (0.6259) | (0.0912) | (0.7598) | |||
(omitted) | 0.0220 | 0.4653** | 1.9781*** | (omitted) | 0.1885 | −0.2572 | −0.1834 | |
(0.4731) | (0.1912) | (0.6294) | (0.6856) | (0.1659) | (0.3511) | |||
1.5535*** | 0.0412 | 0.1201 | 1.1534*** | 0.0198 | −0.0671 | 0.0804** | 0.3577* | |
(0.3097) | (0.1021) | (0.0771) | (0.2530) | (0.1799) | (0.1067) | (0.0382) | (0.1981) | |
constant | −81.6252*** | −28.8109*** | −32.3885*** | −42.4133*** | −33.4472*** | −29.7631*** | ||
(10.6868) | (4.3658) | (1.9499) | (6.0858) | (4.6215) | (1.6909) | |||
No. of obs. | 844 | 844 | 845 | 845 | 829 | 829 | 830 | 830 |
R-sq: with in: | 0.7128 | 0.7032 | – | – | 0.84 | 0.8395 | ||
R-sq: between: | 0.407 | 0.5807 | – | – | 0.3663 | 0.5272 | ||
R-sq: Overall: | 0.3799 | 0.5956 | 0.8906 | 0.9891 | 0.4502 | 0.5953 | 0.8307 | 0.9594 |
Pseudo log-likelihood: | −128000000 | −12535662 | −132000000 | −27367129 | ||||
F-Stats/Wald Chi | F (7,770) = 273.07 | Wald χ2 (12) = 1878.76 | – | – | F (7,756) = 567.09 | Wald χ2 (12) = 4,007.24 | ||
corr (u_i, Xb) | corr (u_i, Xb) = −0.8495 | – | – | – | corr (u_i, Xb) = −0.4273 | |||
Prob > F = 0.0000 | Prob > F = 0.0000 | Prob > F = 0.0000 |
Exports of China HS Commodity 84-85 | ||||
---|---|---|---|---|
Variables | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM |
Ln | 0.8778*** | 0.8513*** | 0.8368*** | 0.7347*** |
(0.0180) | (0.0160) | (0.0358) | (0.0236) | |
0.0010 | −0.0206 | 0.1600*** | 0.0220 | |
(0.0169) | (0.0165) | (0.0329) | (0.0194) | |
0.6050*** | 0.7579*** | 1.2996*** | 0.6369*** | |
(0.0837) | (0.0795) | (0.1747) | (0.0924) | |
0.2822*** | 0.1417*** | 0.6360*** | 0.1086* | |
(0.0489) | (0.0354) | (0.0770) | (0.0588) | |
(omitted) | −0.3513** | −0.1800*** | −0.6201*** | |
(0.1661) | (0.0412) | (0.0821) | ||
−0.2342*** | −0.0582** | 0.0392*** | 0.1327** | |
(0.0504) | (0.0286) | (0.0134) | (0.0553) | |
(omitted) | 0.5030* | 0.3095*** | 0.4548 | |
(0.2958) | (0.1058) | (0.3569) | ||
−0.2102 | −0.2421 | 0.4238*** | −0.1095 | |
(0.1559) | (0.1564) | (0.1098) | (0.0757) | |
(omitted) | 1.1562** | 0.8594*** | 0.7475 | |
(0.5215) | (0.1233) | (0.8059) | ||
(omitted) | 0.3490 | −0.4542*** | 1.8279*** | |
(0.2949) | (0.0913) | (0.5523) | ||
(omitted) | 0.0650 | 0.0684 | −3.6180*** | |
(0.2655) | (0.1655) | (1.1238) | ||
−0.4634*** | 0.1122* | 0.3354*** | −0.4108*** | |
(0.1551) | (0.0624) | (0.0424) | (0.1599) | |
constant | −14.7324*** | −32.6035*** | −42.3291*** | |
(5.3697) | (2.6853) | (1.2330) | ||
No. of obs | 762 | 762 | 762 | 762 |
R-sq: within | 0.9087 | 0.9052 | ||
R-sq: between | 0.2668 | 0.8546 | ||
R-sq: Overall | 0.359 | 0.8582 | 0.9539 | 0.9951 |
Pseudo-log-likelihood | −360000000 | −46347239 | ||
F-Stats/Wald Chi | F (7,693) = 985.10 | Wald χ2 (12) = 6,949.83 | ||
corr (u_i, Xb) | corr (u_i, Xb) = −0.4686 | |||
Prob > F = 0.0000 | Prob > F = 0.0000 |
Source(s): Author's own estimations (Here *** = 1%, ** = 5% and * = 10% significance level)
Exports of Pakistan (aggregate level)
Variables | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM |
---|---|---|---|---|
0.6555*** | 0.4834 *** | −0.7452*** | 0.6845*** | |
(0.0591631) | (0.0503227) | (0.1237692) | (0.0674967) | |
0.1021999 ** | 0.0653038 * | 1.186263 *** | 0.0362 | |
(0.0369704) | (0.035269) | (0.1259552) | (0.0504615) | |
0.3957 | 0.3687 | −1.0539 ** | 1.0312*** | |
(0.2520477) | (0.2260447) | (0.4806069) | (0.2173854) | |
0.1976*** | 0.1674 *** | 0.3996 *** | 0.0149 | |
(0.0626892) | (0.0582908) | (0.1297785) | (0.0940363) | |
(omitted) | −0.8022*** | −1.3361*** | −1.0318 *** | |
(0.1786207) | (0.1059093) | (0.1157194) | ||
−0.0431 | −0.0086 | −0.0896*** | −0.0431 | |
(0.0301425) | (0.0243346) | (0.0164324) | (0.0933156) | |
0.324324 | 0.5256426 | 0.3127157 | 0.4813*** | |
(0.3552652) | (0.3506386) | (0.2587975) | (0.1174585) | |
(omitted) | 0.9473246 *** | 1.132369 *** | 4.1871*** | |
(0.2626063) | (0.0834482) | (0.6128612) | ||
(omitted) | 0.7507114 | −0.698053 ** | 3.390632 *** | |
(0.6306613) | (0.2906049) | (0.697193) | ||
(omitted) | −1.325915*** | 0.2800368 | 0.5115161 | |
(0.3023017) | (0.1851578) | (0.5925246) | ||
−0.8298365*** | 0.0502976 | 1.586282 *** | −0.6315412 *** | |
(0.2100047) | (0.0895825) | (0.1372828) | (0.2093814) | |
constant | 13.75627 ** | −1.715604 | −0.4924939 | |
(5.758952) | (2.534368) | (1.744573) | ||
No. of obs. | 1,253 | 1,253 | 1,253 | 1,253 |
R-sq: within: | 0.3073 | 0.2939 | ||
R-sq: between: | 0.0285 | 0.6346 | ||
R-sq: Overall: | 0.0484 | 0.5893 | 0.63648652 | 0.96363105 |
Pseudo-log-likelihood: | −84200000000 | −6380000000 | ||
F-Stats/Wald Chi | F (7,1144) = 72.50 | Wald χ2 (12) = 648.57 | ||
corr (u_i, Xb) | corr (u_i, Xb) = −0.5097 | |||
Prob > F = 0.0000 | Prob > F = 0.0000 |
Source(s): Author's own estimations (Here *** = 1%, ** = 5% and * = 10% significance level)
Exports of Pakistan (disaggregate level)
HS commodity 01-05 | HS commodity 06-15 | |||||||
---|---|---|---|---|---|---|---|---|
Variables | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM |
1.189866 *** | 1.000751 *** | 0.410461 *** | 0.4649574 *** | 1.0791*** | 0.6126*** | 0.3158*** | 1.0860*** | |
(0.1867832) | (0.1683414) | (0.1486673) | (0.1429467) | (0.1483) | (0.1174) | (0.0870) | (0.1007) | |
0.1213865 | 0.0839763 *** | 0.2087655 ** | 0.5024713 *** | 0.1376* | 0.0697 | −0.0527 | −0.0151 | |
(0.1468507) | (0.1380832) | (0.1253155) | (0.1450184) | (0.0750) | (0.0714) | (0.0603) | (0.0614) | |
1.855195 ** | 0.9274192 | −0.5582319 | 0.7073578 * | 0.1987 | −1.0896* | −0.9697** | 0.4975 | |
(0.7488216) | (0.6827863) | (0.615009) | (0.3709777) | (0.6588) | (0.5646) | (0.4556) | (0.4196) | |
0.6544716 ** | 0.1558249 | 1.399348 *** | 0.2305714 | 0.1863 | −0.0961 | 0.4787*** | 0.1905 | |
(0.2631528) | (0.14366) | (0.1649034) | (0.1899002) | (0.2206) | (0.1182) | (0.0950) | (0.1542) | |
(omitted) | −1.335139 *** | −1.683194 *** | −6.596203 *** | (omitted) | −0.8332** | −0.9654*** | −0.7707*** | |
(0.5335903) | (0.1104267) | (2.355887) | (0.3938) | (0.1204) | (0.2578) | |||
−0.7321223 *** | −0.0347542 | 0.0324668 | 0.3807828 | −0.3138 | 0.2163*** | 0.0939*** | −0.4954 | |
(0.28583) | (0.09873) | (0.0321012) | (0.3484319) | (0.2372) | (0.0786) | (0.0254) | (0.3066) | |
−1.315534 * | −0.8289009 | −0.0936152 *** | −0.9090734 *** | 0.6988 | 1.1626* | 1.1580*** | 1.5432*** | |
(0.688339) | (0.6776776) | (0.2772971) | (0.2841848) | (0.6685) | (0.6528) | (0.4175) | (0.3276) | |
(omitted) | −1.276098 * | −3.243248 *** | 5.511021 | (omitted) | 0.4058 | 0.0813 | 2.4643*** | |
(0.7361531) | (0.3958881) | (5.240572) | (0.4724) | (0.1415) | (0.6950) | |||
(omitted) | 0.6746695 | −0.2080078 *** | 2.081689 | (omitted) | 0.2523 | −0.0474 | 4.6824*** | |
(1.128637) | (0.2505976) | (2.144077) | (0.9208) | (0.2693) | (1.8016) | |||
(omitted) | −0.2538307 | 0.1060007 | −6.400892 | (omitted) | 0.9529 | 1.1022*** | 3.5810*** | |
(1.07258) | (0.2469407) | (4.955741) | (1.0524) | (0.2367) | (0.4433) | |||
−0.3001257 | −0.4675261 ** | 0.4270096 *** | 0.9348708 *** | −1.2653*** | −0.4561*** | −0.2058** | −1.4832*** | |
(0.5393945) | (0.2325515) | (0.1581621) | (0.3104253) | (0.4890) | (0.1781) | (0.0950) | (0.2772) | |
constant | −37.99121 *** | −15.01107 ** | −15.68926 *** | 7.6397 | 4.9297 | 15.2036*** | ||
(14.38023) | (6.08166) | (2.039659) | (13.3164) | (4.9498) | (1.5213) | |||
No. of obs. | 470 | 470 | 532 | 532 | 595 | 595 | 595 | 595 |
R-sq: with in: | 0.2939 | 0.2799 | 0.2737 | 0.25 | ||||
R-sq: between: | 0.0493 | 0.3136 | 0.0123 | 0.1694 | ||||
R-sq: Overall: | 0.0515 | 0.3558 | 0.60689704 | 0.92749732 | 0.0232 | 0.1782 | 0.5904 | 0.8857 |
Pseudo log-likelihood: | −2028798.2 | −419951.76 | −1.25E+10 | −3.48E+09 | ||||
F-Stats/Wald Chi | F (7,422) = 25.10 | Wald χ2 (11) = 175.50 | F (7,542) = 29.18 | Wald χ2 (11) = 176.30 | ||||
corr (u_i, Xb) | corr (u_i, Xb) = −0.7903 | corr (u_i, Xb) = −0.8737 | ||||||
Prob > F = 0.0000 | Prob > F = 0.0000 | Prob > F = 0.0000 |
HS commodity 41-43 | HS commodity 50-63 | |||||||
---|---|---|---|---|---|---|---|---|
Variables | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM | GLS-FEM | GLS-REM | PPML-REM | PPML-FEM |
0.8550*** | 0.5824*** | 0.1556 | 0.5601*** | 0.3111*** | 0.1298* | 0.1606 | 0.5533*** | |
(0.1132) | (0.1076) | (0.2204) | (0.1327) | (0.0777) | (0.0712) | (0.1729) | (0.0964) | |
−0.0041 | 0.0690 | 0.4912*** | 0.1815** | 0.4062*** | 0.3981*** | 0.3408** | 0.2087** | |
(0.1024) | (0.0989) | (0.1781) | (0.0897) | (0.0624) | (0.0611) | (0.1624) | (0.0972) | |
0.5593 | 1.7003*** | 1.6049*** | −0.0712 | 0.4782 | 0.8218*** | 1.3718*** | 0.8008*** | |
(0.4335) | (0.4132) | (0.5959) | (0.2840) | (0.3188) | (0.3009) | (0.4412) | (0.2816) | |
0.7060 *** | 0.3355*** | 0.3605*** | 0.2816 | 0.5863*** | 0.3806*** | 0.1160 | 0.1730 | |
(0.1477) | (0.0918) | (0.1097) | (0.2128) | (0.1090) | (0.0743) | (0.0711) | (0.1501) | |
(omitted) | −0.6253** | −0.8367*** | −1.9955*** | (omitted) | −0.4468 | −0.9990 *** | −0.3772*** | |
(0.2786) | (0.1499) | (0.2482) | (0.3011) | (0.1354 | (0.1330) | |||
−0.4521*** | 0.0022 | 0.0838*** | −0.2097 | −0.3832*** | −0.1232** | −0.1483*** | 0.0688 | |
(0.1516) | (0.0582) | (0.0235) | (0.2223) | (0.1125) | (0.0597) | (0.0233) | (0.1431) | |
−0.3700 | 0.0996 | 1.5182*** | −0.2916** | 0.5022 | 0.7355** | 2.6839*** | 0.6024*** | |
(0.4739) | (0.4784) | (0.5024) | (0.1264) | (0.3339) | (0.3343) | (0.6153) | (0.1878) | |
(omitted) | 0.7843* | 1.1595*** | −2.7062*** | (omitted) | 0.3633 | 1.0787 | −1.5354*** | |
(0.4289) | (0.1899) | (0.5955) | (0.4722) | (0.1196) | (0.2297) | |||
(omitted) | −1.0342 | −1.6860*** | 12.9102*** | (omitted) | −0.3010 | −3.5730*** | 6.1287*** | |
(0.7599) | (0.4425) | (1.9793) | (0.8373) | (0.6624) | (1.2281) | |||
(omitted) | −1.0946** | −1.0926*** | −12.6983*** | (omitted) | −2.1877*** | −0.0569 | −5.4412*** | |
(0.5160) | (0.1633) | (1.5661) | (0.7968) | (0.2178) | (0.8852) | |||
−2.1488*** | 0.1439 | 0.6046*** | −2.1045*** | −0.7115** | 0.5381*** | 0.8943*** | −1.1868*** | |
(0.3397) | (0.1554) | (0.2214) | (0.4095) | (0.2898) | (0.1414) | (0.1781) | (0.2689) | |
constant | 43.4698*** | −19.2668*** | −15.2672*** | 19.2738** | −12.7607*** | −21.0623*** | ||
(9.3642) | (3.6059) | (2.6745) | (8.2224) | (4.2110) | (1.8045) | |||
No. of obs. | 705 | 705 | 715 | 715 | 660 | 660 | 661 | 661 |
R-sq: within: | 0.2018 | 0.1306 | 0.2603 | 0.2279 | ||||
R-sq: between: | 0.0519 | 0.6981 | 0.0165 | 0.6092 | ||||
R-sq: Overall: | 0.0335 | 0.5767 | 0.7017 | 0.9092 | 0.0218 | 0.5642 | 0.9163 | 0.9732 |
Pseudo-log-likelihood: | −2357606.6 | −481005.36 | −24741019 | −3073256.1 | ||||
F-Stats/Wald Chi | F (7,641) = 23.14 | Wald χ2 (11) = 197.90 | F (7,601) = 30.21 | Wald χ2 (11) = 249.66 | ||||
corr (u_i, Xb) | corr (u_i, Xb) = −0.9175 | corr (u_i, Xb) = −0.7637 | ||||||
Prob > F = 0.0000 | Prob > F = 0.0000 | Prob > F = 0.0000 |
Source(s): Author's own estimations (Here *** = 1%, ** = 5% and * = 10% significance level)
Notes
For the independent variable not in logs (such as PCFTA dummy), the semielasticity is concluded by 100*(exp(coefficient) − 1) %.
References
Abbas, S. and Waheed, A. (2017), “Trade competitiveness of Pakistan: evidence from the revealed comparative advantage approach”, Competitiveness Review: An International Business Journal, Vol. 27 No. 5, pp. 462-475.
Abdul Kamal, M., Qayyum, U., Khan, S. and Ngozi Adeleye, B. (2021), “Who is trading well with China? A gravity and constant market share analysis of exports of Pakistan and ASEAN in the Chinese market”, Journal of Asian and African Studies, Vol. 57 No. 6, 00219096211045098.
Abedini, J. and Péridy, N. (2008), “The greater arab free trade area (GAFTA): an estimation of its trade effects”, Journal of Economic Integration, Vol. 23 No. 4, pp. 848-872.
Ahmad, R.M. (2014), “Pakistan-China bilateral trade: the future trajectory”, Institute of Strategic Studies Islamabad. Strategic Studies, Vol. 37 No. 1, pp. 66-89.
Ahmed, A., Qi, H.J. and Tahir, H. (2021a), “Analysis of Pakistan–China FTA by propensity score matching with difference in differences”, SN Business and Economics, Vol. 1 No. 7, pp. 1-29.
Ahmed, A., Qi, H.J. and Tahir, H. (2021b), “Effect of trade liberalization on asymmetric countries for homogenous and differentiate product: case of Pakistan-China free trade agreement”, Journal of International Logistics and Trade, Vol. 19 No. 1, pp. 33-48, doi: 10.24006/jilt.2021.19.1.033.
Ali, A.M., Ullah, I.M. and Asghar, A.M. (2017), “Effect of health expenditure on GDP, a panel study based on Pakistan, China, India and Bangladesh”, International Journal of Health Economics and Policy, Vol. 2 No. 2, pp. 57-62.
Anderson, J.E. (1979), “A theoretical foundation for the gravity equation”, The American Economic Review, Vol. 69 No. 1, pp. 106-116.
Arellano, M. (2003), Panel Data Econometrics, Oxford University Press, doi: 10.1093/0199245282.001.0001.
Ariza Marin, S.D. (2020), “Determinants of trade flows between Colombia and South Korea”, Apuntes del CENES, Vol. 39 No. 70, pp. 75-105.
Ayoub, Z. (2019), “Does the oil revenues affect the macro economic variables performance in Africa? A panel data analysis”, International Journal of Economics, Business and Management Studies, Vol. 6 No. 2, pp. 298-306.
Baltagi, B.H. and Kao, C. (2001), “Nonstationary panels, cointegration in panels and dynamic panels: a survey”, Nonstationary Panels, Panel Cointegration, and Dynamic Panels, Emerald Group Publishing, doi: 10.1016/S0731-9053(2001)15.
Barattieri, A. (2014), “Comparative advantage, service trade, and global imbalances”, Journal of International Economics, Vol. 92 No. 1, pp. 1-13.
Behera, P. and Sethi, N. (2022), “Nexus between environment regulation, FDI, and green technology innovation in OECD countries”, Environmental Science and Pollution Research International, Vol. 29 No. 35, pp. 52940-52953, doi: 10.1007/s11356-022-19458-7, PMID: 35275367.
Chaudhry, T., Jamil, N. and Chaudhry, A. (2017), “Pakistan's experience with the Pakistan–China FTA: lessons for CPEC”, The Lahore Journal of Economics, Vol. 22 No. SE, pp. 1-24.
Ciuriak, D. and Kinjo, S. (2006), “Trade specialization in the gravity model of international trade”, in Curtis, J.M. and Ciuriak, D. (Eds), Trade Policy Research 2005, Department of Foreign Affairs and International Trade, pp. 189-197, available at SSRN: https://ssrn.com/abstract=1549323
Costinot, A., Donaldson, D. and Komunjer, I. (2012), “What goods do countries trade?”, A Quantitative Exploration of Ricardo's Ideas, Review of Economic Studies, Oxford University Press, Vol. 79 No. 2, pp. 581-608.
Deardorff, A.V. and Stern, R.M. (1998), Measurement of Non-tariff Barriers, Vol. 179, University of Michigan Press, doi: 10.3998/mpub.15472.
Deb, K. and Sengupta, B. (2018), “Value-added trade and empirical distributions of RCA indices”, Journal of Quantitative Economics, Vol. 16 No. 1, pp. 235-264.
Eaton, J., Kortum, S.S. and Sotelo, S. (2012), International Trade: linking Micro and Macro (No. w17864), National Bureau of Economic Research.
Elsa, L. and Gianluca, O. (2014), “New revealed comparative advantage index: dataset and empirical distribution”, International Economics, Vol. 139, pp. 48-70.
Fakhrudin, U., Hastiadi, F.F. and Haidir, B.M. (2019), “Impact analysis of normalized revealed comparative advantage on ASEAN's non-oil and gas export pattern using a gravity model approach”, Globalisation, Productivity and Production Networks in ASEAN, Palgrave Macmillan, Cham, pp. 105-145.
Farrukh, N.K., Ahmed, M. and Kayani, U.N. (2013), “China-Pakistan economic relations: lessons for Pakistan”, Pakistan Journal of Commerce and Social Sciences, Vol. 7 No. 3, pp. 454-462.
Grant, J.H. and Lambert, D.M. (2008), “Do regional trade agreements increase members' agricultural trade?”, American Journal of Agricultural Economics, Vol. 90 No. 3, pp. 765-782.
Guglielmo, M.C., Anamaria, S. and Robert, S. (2015), “Trade flows and trade specialisation: the case of China”, China Economic Review, Vol. 34, pp. 261-273.
Hamid, N. and Hayat, S. (2012), “The opportunities and pitfalls of Pakistan’s trade with china and other neighbors”, Lahore Journal of Economics, Department of Economics, The Lahore School of Economics, Vol. 17, Special E, pp. 271-292, September.
Head, K. and Mayer, T. (2014), “Gravity equations: workhorse, toolkit, and cookbook”, Handbook of International Economics, Vol. 4, Elsevier, pp. 131-195, doi: 10.1016/B978-0-444-54314-1.00003-3.
Henderson, D.J. and Millimet, D.L. (2008), “Is gravity linear?”, Journal of Applied Econometrics, Vol. 23 No. 2, pp. 137-172.
Hsiao, C. (2005), “Why panel data?”, The Singapore Economic Review, Vol. 50 No. 02, pp. 143-154.
Irshad, M.S., Qi, X., Li, X. and Arshad, H. (2016), “Deltoid analysis of Pakistan-ASEAN-China free trade agreements and opportunities for Pakistan”, Asian Economic and Financial Review, Vol. 6 No. 5, pp. 228-237.
Jagdambe, S. (2019), “Consistency test of revealed comparative advantage index: evidence from India's agricultural export”, Foreign Trade Review, Vol. 54 No. 1, pp. 16-28.
Junaid, K. and Manzoor, H.M. (2017), “Dynamics of Pakistan’s trade balance with China”, SBP Staff Notes, Vol. 04 No. 17.
Kabraji, R. (2012), The China-Pakistan Alliance: Rhetoric and Limitations, Chatham House Asia Program, CID: 20.500.12592/q2dgzv.
Kalirajan, K. (2007), “Regional cooperation and bilateral trade flows: an empirical measurement of resistance”, The International Trade Journal, Vol. 21 No. 2, pp. 85-107.
Kang, H. and Fratianni, M.U. (2006), “International trade efficiency, the gravity equation, and the stochastic frontier”, Available at: SSRN 952848.
Kataria, R.J. and Naveed, A. (2014), “Pakistan-China social and economic relations”, South Asian Studies, A Research Journal of South Asian Studies, Vol. 29 No. 2, pp. 395-410.
Kepaptsoglou, K., Karlaftis, M.G. and Tsamboulas, D. (2010), “The gravity model specification for modeling international trade flows and free trade agreement effects: a 10-year review of empirical studies”, The Open Economics Journal, Vol. 3, pp. 1-13, doi: 10.2174/1874919401003010001.
Laursen, K. (2015), “Revealed comparative advantage and the alternatives as measures of international specialisation”, Eurasian Business Review, Vol. 5 No. 1, pp. 99-115.
Leontief, W. (1953), “Domestic production and foreign trade; the American capital position re-examined”, Proceedings of the American Philosophical Society, Vol. 97 No. 4, pp. 332-349.
Linnemann, H. (1966), An Econometric Study of International Trade Flows (No. 42), Amsterdam, North-Holland.
Liu, B. (2010), Several Legal Issues on ECFA Revisited, Present Day Law Science, 2010-05.
Longo, R. and Sekkat, K. (2004), “Economic obstacles to expanding intra-African trade”, World Development, Vol. 32 No. 8, pp. 1309-1321.
Mahboob, A.M., Vichayanan, R. and Abusayed, K.A. (2017), “Comparative positioning of small and medium enterprises in Bangladesh, Thailand, and the Philippines”, Экономика региона, Vol. 13 No. 2, pp. 381-395, doi: 10.17059/2017-2-5.
Mátyás, L. and Sevestre, P. (Eds) (2008), The Econometrics of Panel Data: Fundamentals and Recent Developments in Theory and Practice, Vol. 46, Springer Science & Business Media, ISBN-10: 3540758895 ISBN-13: 978-3540758891.
Maqbool, M.S., Mahmood, T., Hussain, S. and Ashraf, M. (2020), “Analysis of trade competitiveness of Pakistan cereal products in global perspective”, Review of Economics and Development Studies, Vol. 6 No. 1, pp. 97-106.
Martínez-Zarzoso, I. (2013), “The log of gravity revisited”, Applied Economics, Vol. 45 No. 3, pp. 311-327.
Musleh, D., Ghani, I. and Qadir, U. (2009), “Recent experience and future prospects of Pakistan's trade with China”, Lahore Journal of Economics, Vol. 14 No. E, pp. 87-110.
Narayan, S. and Nguyen, T.T. (2016), “Does the trade gravity model depend on trading partners? Some evidence from Vietnam and her 54 trading partners”, International Review of Economics & Finance, Vol. 41, pp. 220-237.
Nerlove, M. (2005), Essays in Panel Data Econometrics, Cambridge University Press, ISBN: 9780521022460.
Nowak‐Lehmann, F., Herzer, D., Martinez‐Zarzoso, I. and Vollmer, S. (2007), “The impact of a Customs union between Turkey and the EU on Turkey's exports to the EU”, JCMS: Journal of Common Market Studies, Vol. 45 No. 3, pp. 719-743.
Paas, T. and Tafenau, E. (2005), European trade integration in the Baltic Sea Region-A gravity model-based analysis.
Qi, X. and Irshad, M.S. (2015), “Pakistan-China free trade agreement (PCFTA) treaty model: capability, prospects, and disputes”, Academic Research International, Vol. 6 No. 3, pp. 53-60.
Rahman, M.S. and Miah, D.M. (2017), “The impact of sources of energy production on globalization: evidence from panel data analysis”, Renewable and Sustainable Energy Reviews, Vol. 74, pp. 110-115.
Ricardo, David. ([1817] 1951), On the Principles of Political Economy and Taxation. Vol. 1 of The Works and Correspondence of David Ricardo, edited by Piero Sraffa with the collaboration of M. H. Dobb, Cambridge University Press, Cambridge.
Shaista, A. (2010), The Effect of Preferential Trade Agreements on Pakistan's Export Performance, Centre for Research in Economic Development and International Trade, University of Nottingham, Credit Research Paper No. 15/10.
Shakur, S. and Tsang, C.Y. (2017), New Zealand's trade prospects in an uncertain trans-pacific partnership (TPP) environment: results from gravity model.
Sheng, Y., Tang, H.C. and Xu, X. (2012), The Impact of ACFTA on People's Republic of China-ASEAN Trade: Estimates Based on an Extended Gravity Model for Component trade (No. 99), ADB Working Paper Series on Regional Economic Integration.
Siddique, M., Anwar, A. and Quddus, M.A. (2020), “The impact of real effective exchange rate on revealed comparative advantage and trade balance of Pakistan”, Economic Journal of Emerging Markets, Vol. 12 No. 2, pp. 193-207.
Silva, J.S. and Tenreyro, S. (2006), “The log of gravity”, The Review of Economics and Statistics, Vol. 88 No. 4, pp. 641-658.
Smith, A. (1776), An Inquiry into the Nature and Causes of the Wealth of Nations: Volume One, Printed for W. Strahan; and T. Cadell, London.
Sören, P. and Bruemmer, B. (2012), Bimodality & the Performance of PPML, Institute for Agriceconomics Discussion Paper, 1202.
Tang, Y., Zhang, Y. and Findlay, C. (2013), “What explains China's rising trade in services? Empirical analysis with a modified gravity model and panel data”, Chinese Economy, Vol. 46 No. 6, pp. 7-31.
Tinbergen, J. and Bos, H. (1962), Mathematical models of economic growth.
Timsina, K.P. and Culas, R. (2021), “Australia’s free trade agreements (FTAs) and potentiality of wheat exports: a panel gravity model approach”, Journal of East-West Business, Vol. 28 No. 2, pp. 1-28.
Torok, A. and Jambor, A. (2016), “Determinants of the revealed comparative advantages: the case of the European ham trade”, Agricultural Economics, Vol. 62 No. 10, pp. 471-482.
Uddin, N.M., Uddin, M.J., Uddin, J. and Ahmmed, M. (2020), “Remittances and economic growth tie in selected South Asian countries: a panel data analysis”, International Journal of Financial Research, Vol. 11 No. 1, pp. 197-205.
Vernon, R. (1966), “International investment and international trade in the product cycle”, The Quarterly Journal of Economics, Vol. 80 No. 2, pp. 190-207, doi: 10.2307/1880689.
Wilson, J.S. and Mann, C.L. (2003), Trade Facilitation and Economic Development: Measuring the Impact, World Bank Policy Research Working Paper, (2988).
Further reading
Asian Development Bank (2010), available at: https://www.adb.org/sites/default/files/institutional-document/31388/adb-ar2010-v1.pdf
French, S. (2017), A Gravity-Based Revealed Comparative Advantage Estimator, UNSW Business School Research Paper, (2017-05).