Is there reciprocity between India and RCEP member countries' goods trade?

Abstract

In early November 2019, India announced its withdrawal from the Regional Comprehensive Economic Partnership (RCEP) Agreement, and stated that there is a non-reciprocal trade relationship with RCEP member states, this incident has aroused extensive discussions between India and the international community. We take reciprocity as the perspective and discuss the following three theoretical propositions: (1) Can the reciprocity coefficient be used as a basis for decision-making? (2) What is the influence mechanism of the reciprocity coefficient? (3) How will the reciprocity of RCEP changes in the future? We find that the reciprocity index is a result indicator of trade policy and trade network status. It is affected by the trade policy and trade status of related countries in the short term. It can be regarded as a warning and decision-making reference directional indicator. However, it cannot be used as a basic index for trade policy adjustments, India should reconsider this decision or choose other ways to participate in RCEP's regional cooperation.

Keywords:

• India
• RCEP
• social networks
• reciprocity
• cooperation

1. Introduction

Slow progress of World Trade Organization negotiations has created a need for large regional trade agreements over the past decade. However, the new wave of nationalism has led to an increase in trade barriers, trade wars, and the break away from Mega-regionals of various economic powers in the world (Gaur 2020). Similar to the practices of the United States in the Trans-Pacific Partnership (TPP) and the United Kingdom in the EU, as one of the founding members of the Regional Comprehensive Economic Partnership (RCEP), India has also withdrawn from the RCEP on the grounds that trade import and export deficit and RCEP negotiations failed to address key concerns.

Although India's move was surprising, it was also not unexpected. On the one hand, there is considerable controversy over whether India should join the RCEP. At the same time, due to the huge trade deficit with the RCEP member countries, India has to make a cautious choice. On the other hand, under great pressure from relevant domestic interest groups (such as the agriculture and dairy industry) (Verma 2020), to stabilize the voter base, the Modi administration was forced to change its policy preferences. During the negotiations, the Modi government also authorized Indian think tanks, universities, and relevant research institutions to conduct a more in-depth and comprehensive study of the pros and cons of RCEP (Jiang 2020). The results of the study are not consistent, but ultimately India's government believes that RCEP is not in line with India's current interests and trade lacks reciprocity, thus making the decision to withdraw. This incident has caused widespread debate between India and the international community. Against this backdrop, it is more important than ever to examine the motivations behind India's withdrawal and the factors that influence the decision making.

Regarding India's withdrawal from RCEP, the academic community has fully discussed it from different angles. Sharma et al. (2020) used the Global Trade Analysis Project (GTAP) model to analyze the impact of various macroeconomic variables in India after joining RCEP. The evaluation results indicate that India's re-entry to RCEP may not be beneficial. Contrary to the view of Sharma et al. (2020), Gaur (2020) analyzed in detail the reasons for India's withdrawal from various aspects and finally concluded that this withdrawal from RCEP will make India lose its export potential and lose the opportunity to become a member of the region's value chain, joining RCEP will be more conducive to the future development of India. In addition, Chakraborty and Chaisse (2021) analyzed the reasons for India's withdrawal from RCEP from the perspective of the participation of RCEP countries' regional value chains and international production networks. The observations indicate that, relatively modest participation in the regional value chains (RVCs), declining domestic value added content of exports and the associated adverse trade balance scenario have critically shaped the Indian standpoint. From what has been discussed above, most of the previous literature used gravity models and other measurement methods to analyze the impact of joining or withdrawing from RCEP on India, or the perspective of the country's short-term and long-term interests, to analyze the reasons and rationality of India's withdrawal from RCEP, the final conclusion are also very different.

As an important form of human motivation (Dufwenberg and Patel 2017), reciprocity has become a consideration factor for most economic actors to make decisions and actions (Fehr and Gächter 2000). However, many social relationships are not often reciprocal, and reciprocity may be the exception rather than the rule. As an important factor affecting the willingness of international exchanges and cooperation, reciprocity is not only an important factor in determining the level of tariffs (Baldwin, Kawai, and Wignaraja 2013), but also an important motivation for countries to seek new free trade agreements or expand existing free trade agreements through unilateral actions (Menon 2013).

In recent years, the withdrawal behavior of countries in the world is not uncommon, such as Brexit, the US withdraw from various international organizations. Under an environment where trade protectionism is on the rise to achieve trade surplus, research on reciprocity is not only an effective method to tap the intrinsic motivation of the state's behavior change, but also an important tool to promote the open trade of partner countries and maintain long term stable cooperation (González, Esposto, and Viego 2015; Frank et al. 2018). Therefore, it is necessary to study whether there is reciprocity in trade between countries, especially what are the influencing factors of reciprocity, and whether it can be used as a basis for decision-making. Second, reciprocity is still an area that has not been fully explored in the network. Most of the researches using network science and social networks to analyze trade between countries focuses on node network indicators (Fagiolo, Reyes, and Schiavo 2010), centrality (Hakeem and Suzuki 2019), clustering (Barigozzi, Fagiolo, and Garlaschelli 2010), and other aspects. Analyzing reciprocity, especially using empirical methods to study the factors affecting reciprocity is relatively scarce, and our research aims to enrich the theoretical basis of this field. To conclude, this article attempts to answer the following three questions: (1) Can the reciprocity coefficient be used as a basis for decision-making? (2) What is the influence mechanism of the reciprocity coefficient? (3) How will the reciprocity coefficient of RCEP change in the future?

Country’s decision-making is not only independent of several factors but is affected by the interweaving of multiple factors. Network analysis has solved the problem that the dynamic changes of multiple factors are difficult to capture, it has been applied in many fields and has provided explanations for various social phenomena (Borgatti et al. 2009). Thus, this article uses the methods of social network analysis to calculate the trade reciprocity between India and RCEP member states and the trade reciprocity between RCEP member states in order to answer the research questions. We find that India’s integration into the goods trade network of RCEP member states has reduced the trade network reciprocity. This is contrary to the perception of traditional economists. In response to this phenomenon, this article continues to introduce the degree centrality, closeness centrality, betweenness centrality, and the external variables (number of trade agreements) in the goods trade network between India and RCEP member countries from the perspective of social networks to explore whether these variables will affect trade reciprocity. The final result shows that the improvement of India's centrality (degree centrality, betweenness centrality) in the goods trade network with RCEP member states and the signing of more trade agreements with RCEP member states will enhance India's reciprocity in the trade network with RCEP.

The rest of the paper is organized as follows. Section 2 discusses the concept of reciprocity and its influencing factors. Section 3 explains the methods and data sources used in this article. Section 4 provides the results of this paper. Section 5 conducts empirical test and robustness test. Section 6 discusses and summarizes.

2. Concept of reciprocity in social networks

2.1. Reciprocity

Reciprocity is a basic terminology for social network which refers to the tendency towards paired vertices to be connected by two interconnected chains pointing in opposite directions (Garlaschelli and Loffredo 2006). This is a special type of correlation found in directed networks and a common attribute of many networks. In many fields, economists believe that economic agents’ decision-making is based on pure self-interest, but the progress of interaction will not come from further adjustments to the model of pure egoism, and a significant proportion of actors adjusts their decisions based on reciprocity (Fehr and Gächter 2000). In highly reciprocal relationships, both parties have the same interest in maintaining the relationship, while simultaneously exploitation of the participating individuals is prevented, and in low reciprocal relationships, one partner is much more active than the other (Akoglu, Vaz de Melo, and Faloutsos 2012; Stojkoski et al. 2018). The contemporary world trade pattern and the multilateral trading system are facing new challenges, this also makes the study of reciprocity more important than ever before.

In order to better express the reciprocity in the network, Garlaschelli and Loffredo (2004) propose a new measure of reciprocity that allows the ordering of networks according to their actual degree of correlation between mutual links, and that networks of the same type (economic, social, cellular, financial, ecological, etc.) display similar values of the reciprocity. The application of reciprocity has become more widespread, and some scholars have also used reciprocity to analyze and explain the evolution of relations between countries. Frank et al. (2018) conducted research on the cooperation, influence, and reciprocity between the EU and the world, and the results showed that many countries in the international system have reciprocal relations, and found that compared with non-reciprocal countries, these reciprocal countries shows essentially different Cooperation dynamics. The research also proves that reciprocity is a universal mechanism for realizing international cooperation. Ruzzenenti, Garlaschelli, and Basosi (2010) studied the various symmetries of the entire world trade network, and the observed evolution of reciprocity is consistent with the symmetry disruption occurring in the production space. Besides, scholars continue to apply reciprocity to international economic sanctions (Cranmer, Heinrich, and Desmarais 2014), international plastic resin trade (Ren et al. 2020), international mergers and acquisitions (Dueñas et al. 2017), global value chains (Amador et al. 2018), and other fields. It can be seen that more and more scholars apply reciprocity to the study of international trade and international affairs, which greatly enriches the theoretical basis and practicality of reciprocity research.

2.2. Factors affecting reciprocity in international trade

To study the influencing factors of the change in India's reciprocity balance, we need to clarify which factors affect or determine the change in India's reciprocity balance, and how these factors affect the change in reciprocity. Although academic research on the influencing factors of reciprocity is scarce, many studies have shown that the structure of the network will have an important impact on network reciprocity(Van Doorn and Taborsky 2012; Akoglu, Vaz de Melo, and Faloutsos 2012), and the establishment of clusters of cooperators will also support the evolution of generalized reciprocity in networks with community structure (Van Doorn and Taborsky 2012). From an economic point of view, the relevance of trade networks stems from the fact that trade relations determine the degree of dependence of other countries on a particular country or the degree of influence of one country on another country (Fagiolo, Reyes, and Schiavo 2010). India's reciprocity is also influenced by a series of factors such as influence, status, and dependence of India and RCEP trade network. In social networks, these factors can be reflected a certain extent in the centrality of nodes in the network (Wasserman and Faust 1994). As Rivera-Batiz and Romer (1991) suggested, the benefits of trade can be verified and fully characterized by using network analysis. Arora and Vamvakidis (2004) also stated that trade gains not only depend on the degree of trade openness but also the number of trading partners and their characteristics. Therefore, we selected representative indicators in the network: centrality, and external dependent variables (number of trade agreements) as the influencing factors that cause changes in reciprocity, the following assumptions are made (Table 1):

Table 1. Hypotheses on the effect of India centrality, the number of trade agreements between India and RCEP members, and the difference in the number of trade agreements on reciprocity difference.

Research study hypothesis	Variable	Reciprocity difference	Indian reciprocity
Hypothesis 1	The difference in the number of trade agreements (TAD)	Positive correlation	Negative correlation
Hypothesis 2	Number of trade agreements between India and RCEP member countries (IRCEP TA)	Negative correlation	Positive correlation
Hypothesis 3	Indian degree centrality (DC)	Negative correlation	Positive correlation
Hypothesis 4	Indian closeness centrality (CC)	Negative correlation	Positive correlation
Hypothesis 5	Indian betweenness centrality (BC)	Negative correlation	Positive correlation

Note: The difference in the number of trade agreements is calculated by subtracting the number of trade agreements between India and RCEP member countries from the number of trade agreements between RCEP member countries.

3. Methods and data

3.1. Network analysis

The simplest form of the network is a combination of nodes with edges as connections, and there may be only a few nodes with limited connectivity, or there may be many nodes with complex connectivity patterns (Hakeem and Suzuki 2019). When applying network analysis to trade flows, we can depict the web of trade relations as a network where countries play the role of nodes and a link describes the presence of an import/export relation between any two countries (and possibly the intensity of that flow). Compared with other research methods, social networks increase our economic understanding of the dynamics of international trade (Fagiolo, Reyes, and Schiavo 2010), and can better capture changes in relationships between economies. At the same time, as an interdisciplinary research method, social network analysis methods have been increasingly favored by academia.

In the existing literature, scholars have made profound studies on international trade networks, such as the structure, evolution and topological characteristics of trade networks. De Benedictis and Tajoli (2011)analyzed the structure and evolution of the world trade network. The results showed that the degree of world trade integration is increasing, but it has not yet been completed. Except for some areas, the selection of trading partners varies greatly among countries and the organization members are more closely connected than other members of the world. He and Deem (2010) analyzed how globalization and economic recession have shaped the structure of the world trade network, and found that due to the structural changes in the world trade network caused by globalization, today's trade networks are more sensitive to recessionary shocks and recovery is also slower. Fan et al. (2014)discussed the role and status of countries in international trade. When viewed from the perspective of network centrality, world trade presents a closed, unbalanced, diversified, and multi-polarized development trend.

These articles not only analyze the attributes of international trade networks, but also analyze the dynamic changes of nodal point attribute and relationships in detail. Therefore, the motivation for choosing the reciprocity indicator in the social network for analysis is to study the reciprocal relationship of countries’ trade within the region, not to analyze the relationship between the two economies in isolation, the interrelationships between economies within the region must be taken into consideration. Besides, in today's global trade network, reciprocity can provide an important explanation for bilateral cooperation in a wide range of international fields. Bilateral reciprocity may even become the basis of multilateral cooperation. Although powerful countries exert a high degree of influence on cooperation in the international trading system, this influence is mutual, and cooperation still partly depends on actual reciprocity rather than unilateral actions of specific countries (Frank et al. 2018). This also makes the study of trade reciprocity realistic and effective for analyzing trade relations and changing trends among countries.

3.2. Indicators of reciprocity coefficient

In the topological structure of the international trade network, not all trade relations are two-way, that is, the export relationship between country i and country j does not mean that country j also has an export relationship with country i. The degree of the two-way trade relationship between two countries, namely reciprocity, has become an important measure of the complex network of international trade. Reciprocity can measure the degree of participation of each country in the international trading system, stability of trade cooperation, degree of trade cooperation (Frank et al. 2018), degree of bilateral trade relations (Ren et al. 2020), symmetry and balance of trade (Ruzzenenti, Garlaschelli, and Basosi 2010). The larger the value, the more equal the value of trade to both parties. Here we use Garlaschelli and Loffredo (2004) definition of reciprocity coefficient, and its calculation formula is: (1) $$\rho =\frac{\sum _{i\ne j}(a_{\mathit{\text{ij}}}-\overline{a})(a_{\mathit{\text{ji}}}-\overline{a})}{\sum _{i\ne j}(a_{\mathit{\text{ij}}}-\overline{a}{)}^2}$$(1) and, $\overline{a}=\frac{\sum _{i\ne j}{a}_{\mathit{\text{ij}}}}{N(N-1)}$

3.3. Research scope

This article focuses on the goods network between India and RCEP member states. The member states are shown in Table 2:

Table 2. List of India and RCEP member states.

Area	Country
Southeast Asia	Malaysia, Singapore, Thailand, Vietnam, Brunei, Myanmar, Laos, Philippines, Cambodia, Indonesia
South Asia	India
North-east Asia	China, Japan, Republic of Korea
Oceania	Australia, New Zealand

3.4. Data source and explanation

We use two sets of databases: one is the Commodity Trade Statistics Database (UN Comtrade), and the other is the World Trade Organization Database, the sample covers the period from 2006 to 2019. It is worth noting that to solve the problem that the import value of a country A and a country B in the database is sometimes different from the export value from a country B to a country A, we define the average of these values as the approximate value of these data. In this way, the result error caused by the data can be reduced as much as possible.

4. Reciprocity changes and influencing factors between India and RCEP member countries

4.1. Evolution analysis of reciprocity coefficient of RCEP member countries' goods trade network

Use formula (1) to calculate the network reciprocity coefficient from 2006 to 2019, and plot the calculation result into a graph, the result is shown in Figure 1:

Figure 1. Changes of reciprocity index of goods trade between India and RCEP member countries.

It can be seen from Figure 1 that: (1) The reciprocity of trade between India and RCEP and the intra-RCEP member states fluctuates, and the trends between the two are similar. There was a slow upward trend from 2006 to 2008, a sharp decline from 2008 to 2010, a shock upward trend from 2010 to 2016, a decline from 2016 to 2018, and an upward trend after 2018. (2) Overall, the reciprocity coefficient between RCEP and India fluctuates at a high level, the lowest value in 2010 is also above 0.86. (3) The reciprocity coefficient within RCEP member states is higher than that between India and RCEP member states in the same period. (4) The reciprocity coefficient between the two shows a fluctuating expanding trend with time.

4.2. Analysis of the factors affecting India's reciprocity difference

This section will analyze the possible impact of these factors on India’s reciprocity gap from the perspective of India’s centrality in the goods trade network with RCEP member countries and the changes in the number of trade agreements between India and RCEP member countries.

4.2.1. Centrality analysis of India in the RCEP-Indian goods trade network

The network centrality measure defines the importance of a single node. In different types of networks, different centrality measures are used to identify important nodes (Hakeem and Suzuki 2017). In the goods trade network, node centrality refers to the possibility of a particular country appearing along a randomly selected trade chain in the trade network. The higher the possibility, the greater the influence of the country in the network (Fagiolo, Reyes, and Schiavo 2010). Among them, the more widely used centrality calculation methods are Freeman's method and Bonacich's method. Freeman believes that the central nodes have three advantages over other nodes: more connections, faster reach to other nodes, and control of the flow between other nodes (Freeman 1978). Based on the above characteristics, three methods of centrality measurement are proposed from three aspects of power, resource acquisition and control ability. Bonacich’s measurement method mainly considers the mutual influence degree of the connection of two nodes, so as to calculate the influence of each node on other nodes, and fully considers the influence of path length (Bonacich 1987). However, because the analysis of centrality in this article focuses more on analyzing the transaction capacity of node countries, the ability to acquire network resources and the ability to control trading partners from different perspectives, and the network scale and path length of this article are relatively small. Therefore, the Freeman’s method is chosen to calculate the degree centrality, closeness centrality and betweenness centrality. The above three centrality calculations can be obtained along the path of Network-Centrality-Multiple Measures in UCINET.

Among them, the degree centrality measures the degree of the direct connection between a point and other points, and if a node is directly connected to many nodes in the network, then the node is in the center of the network and has greater power (Wasserman and Faust 1994). In the trade network, countries connected to all other nodes can have a greater impact on other nodes and the entire network. A node in a non-core position must pass through other nodes to transmit information, the stronger the dependence on other nodes, the weaker the power (Hakeem and Suzuki 2017). Closeness centrality measures the average distance of a node to other nodes (Wasserman and Faust 1994). In the trade network, it can be expressed as the ‘economic distance’ between one economy and another economy (Andal 2016). Countries with a lower average distance from other countries may be easier to transmit transaction information and are not subject to other network controls (Hakeem and Suzuki 2017). Betweenness centrality measures how much a point is located in the ‘middle’ of other points in the network. A node with high middle centrality can maintain high power in the network by controlling resources such as information flow (Wasserman and Faust 1994). In the goods trade network, if the flow of goods needs to reach the destination through a specific node, then this specific node will play a ‘hub’ role in inter-country trade (De Benedictis and Tajoli 2011), and correspondingly has greater power.

Tables A1, A2, and A3 in Appendix 1 respectively summarize the degree centrality, closeness centrality, and betweenness centrality of each country in the goods trade network between India and RCEP member countries calculated by UCINEET in this research. Based on the results, we have following conclusions:.

1. India ranks 10th in average degree centrality and seventh in growth, with growth rates faster than half of the countries. This shows that at this stage, India’s participation in the field of goods trade with RCEP member countries is limited, and its influence and power in the network are relatively weak. However, because of the development potential of India’s domestic market and the development potential of the manufacturing industry, the degree centrality and status of India will show an upward trend in the future.

2. From the perspective of closeness centrality, India’s closeness centrality is slightly low, and its growth rate is positive. It shows that India’s independence and autonomy in the trade network are not strong compared with most countries, and it is weaker in terms of power, prestige and influence. However, in recent years, India's closeness centrality has been increasing, and its trading capacity has been gradually enhanced to reduce its dependence on RCEP members. This also explains from the network perspective why India has frequently increased tariffs on China in recent years.

3. Finally, it can be seen that India's betweenness centrality is low, and it has generally increased in 2009, 2011 and 2012, 2015 and 2017, but then began to decline. It shows that India has little influence on the development of its trade with RCEP. Although it has increased its influence in some years and had become a ‘hub’ country, thereby improving its ability to control the flow of resources and transform resources into economic benefits, but was replaced by other countries with faster trade development, such as Vietnam.

In general, in comparison with the three centralities in India, the closeness centrality is relatively high, while the degree and betweenness centrality are generally low. In addition, from these three sets of data can be analyzed that in terms of its goods trade network with RCEP member countries, India is closely linked to important countries in the network, such as China, Singapore, and Japan, and therefore belongs to key countries, and has certain independence and autonomy. However, its lower degree centrality and betweenness centrality represent India's lower rights and control over the network, and this phenomenon only improved in certain years and then declined.

4.2.2. The fluctuation of Indian centrality and reciprocity difference

Note: The degree centrality, closeness centrality, and betweenness centrality of India are calculated from the goods trade network between India and RCEP member states. The difference of reciprocity is calculated by subtracting the reciprocity index of the goods trade network of India and RCEP member countries from the reciprocity index of the goods trade network of RCEP member countries.

Figure 2 shows the changing trend of the difference between degree centrality, closeness centrality, betweenness centrality, and reciprocity in India. It can be seen that the pattern of the three centrality of India are consistent, which increase in 2008-2009, 2010-2012, and 2015-2017, and then followed by a certain degree of decline. We are trying to find the reasons for this phenomenon from the perspective of changes in India's foreign trade policy. In 2009, India adjusted its ‘Foreign Trade Policy’ and ‘Import and Export Policy’ to reduce export tariffs, and launched a three-year ‘zero-tariff’ export promotion capital goods program; In 2010, India's free trade agreements with South Korea and ASEAN came into force; In 2011, India's free trade agreements with Malaysia and Japan came into force; In 2014, the ‘Act-East’ policy was implemented; In line with Modi's ‘Make in India’ initiative, India's New Foreign Trade Policy (2015-2020) was launched in 2015. These policies have promoted the rise of India’s centrality, but the changes brought about by reciprocity are not the same. The centrality increased from 2008 to 2009 and from 2011 to 2012, while India's reciprocity gap declined, the performance from 2015 to 2017 was the opposite.

Figure 2. Trend of difference between centrality and reciprocity.

Note: Due to the large numerical difference between data, this paper reduces the value of Indian degree centrality and closeness centrality by 10 times and enlarges the difference of reciprocity by 100 times.

Judging from the actual impact of these trade policies on India’s imports and exports, these policies have effectively promoted the increase in the total amount of imports and exports between India and RCEP member countries, and the growth rate of exports was significantly higher than that of imports. The period from 2011 to 2012 was particularly obvious, India’s rapid increase in exports has promoted mutually beneficial trade relations between India and RCEP member states, this may be the reason for the decrease in the difference in reciprocity between India. However, between 2015 and 2017, India implemented a series of trade policies to enhance the competitiveness of the country’s manufacturing industry and promote export, which eventually did enhance India’s position in the network, and the total export volume increased substantially. However, during this period, India's dependence on imports of products such as machinery and equipment further increased, which, on the basis of promoting India's rising status in the network, also expanded the reciprocity difference. In addition, the reason for the decline in India's centrality after each rise may be that after the rise in India's position in the network, it has been replaced by other countries (such as Vietnam) that trade in goods has developed more rapidly.

4.2.3. India and RCEP member states signed trade agreements

Research shows that most countries are more willing to trade with their partners (De Benedictis and Tajoli 2011). Figure 3 also explains to a certain extent why the reciprocity of goods trade within RCEP member countries is higher than that of India and RCEP countries. First, from 2006 to 2019, the member states of RCEP signed 24 trade agreements with each other, the number of bilateral and multilateral trade agreements between RCEP member states still in a state of steady growth. In contrast, India has been stagnant since 2011. Second, among RCEP member countries, China, Japan, Republic of Korea, and other countries have not only signed free trade agreements with ASEAN but also signed bilateral free trade agreements with some countries in ASEAN, forming a closer trade network. These trade agreements within RCEP have greatly promoted trade exchanges among RCEP member states, conducted fair and mutually beneficial trade cooperation, deepened production networks, and at the same time, the scale of imports and exports between countries has also increased significantly. In contrast, India has only signed free trade agreements with ASEAN, Japan, South Korea, and Malaysia, and the share both of intraregional trade and world trade covered by these FTAs is relatively small. It can be assumed that the gap in the number of trade agreements signed between India and the RCEP segment is also an important reason for the change in India's reciprocity difference.

Figure 3. Number of trade agreements between India and RCEP member states and number of trade agreements between RCEP member states.

Note: The statistical value is the number of valid trade agreements in the year in which trade agreements were signed after 2006, and it is not counted before 2006.

5. Ols regression results analysis and robustness test

We further use ordinary least squares (OLS) regression analysis to check whether the assumptions in Table 1 are correct, Table 3 shows the regression results of OLS.

Table 3. OLS and GMM regression estimation results.

Hypothesis	Variable	OLS	GMM
1	TAD	.0054907***(4.17)	.0054907***(5.40)
2	IRCEP TA	-.0019018**(-2.58)	-.0019018***(-3.38)
3	DC	-.0047017*(-2.29)	-.0047017***(-3.70)
4	CC	.0096645**(2.61)	.0096644***(4.15)
5	BC	-.003802 -1.44	-.003802***(-2.80)
	R-squared	0.8792	0.8792
	Prob > F	0.0016	0.0000

Note: The number in brackets under the OLS estimation coefficient is the t statistic of the coefficient estimation, and the number in brackets under the GMM estimation coefficient is the z statistic of the coefficient estimation;

*, ** and *** represent the significance levels of 10%, 5%, and 1%, respectively.

Table A1. The degree centrality of India and RCEP goods trade network from 2006 to 2019.

	Australia	Brunei	Cambodia	China	Indonesia	Japan	Myanmar	Korea, Rep	Lao PDR	Malaysia	New Zealand	Philippines	Singapore	Vietnam	Thailand	India
2006	67	20	13	87	73	80	13	73	7	67	27	47	73	53	87	53
2007	73	20	20	87	73	80	13	67	7	67	27	47	73	67	87	60
2008	80	27	20	87	73	80	20	73	7	73	33	60	80	73	87	60
2009	67	7	27	87	67	80	20	67	7	67	27	53	80	73	87	67
2010	73	13	27	87	67	80	20	80	7	67	33	60	87	73	87	60
2011	80	33	27	87	67	80	20	80	7	67	33	67	93	73	87	73
2012	73	33	27	93	67	87	40	87	7	73	47	67	93	73	93	80
2013	73	27	27	100	67	87	27	80	13	73	47	67	93	73	93	67
2014	73	20	27	100	67	87	27	80	13	73	47	67	87	73	93	67
2015	73	13	27	100	67	87	27	73	13	73	47	67	87	73	93	67
2016	73	7	27	93	67	87	33	73	13	67	40	67	87	73	93	73
2017	73	7	27	93	67	80	27	73	13	73	53	67	87	73	93	80
2018	73	13	33	100	67	93	33	73	13	73	47	67	87	73	93	73
2019	73	13	33	93	67	93	27	73	20	73	47	67	93	80	93	67
Mena	73	18	26	92	68	84	25	75	11	70	40	62	86	72	90	68
Growth rate	1.09	0.9	2.00	1.06	0.93	1.05	1.92	1.03	1.57	1.04	1.48	1.32	1.18	1.36	1.03	1.28

Table A2. The closeness centrality of India and RCEP goods trade network from 2006 to 2019.

	Australia	Brunei	Cambodia	China	Indonesia	Japan	Myanmar	Korea, Rep	Lao PDR	Malaysia	New Zealand	Philippines	Singapore	Vietnam	Thailand	India
2006	75	50	52	88	79	83	52	79	48	75	56	65	79	68	88	68
2007	79	50	54	88	79	83	52	75	48	75	56	65	79	75	88	71
2008	83	52	54	88	79	83	54	79	48	79	58	71	83	79	88	71
2009	75	47	56	88	75	83	54	75	48	75	56	68	83	79	88	75
2010	79	48	56	88	75	83	54	83	48	75	58	71	88	79	88	71
2011	83	58	58	88	75	83	56	83	48	75	58	75	94	79	88	79
2012	79	58	58	94	75	88	63	88	50	79	65	75	94	79	94	83
2013	79	56	58	100	75	88	58	83	54	79	65	75	88	79	94	75
2014	79	56	58	100	75	88	58	83	54	79	65	75	88	79	94	75
2015	79	54	58	100	75	88	58	79	54	79	65	75	88	79	94	75
2016	79	48	56	94	75	88	60	79	52	75	63	75	88	79	94	79
2017	79	47	56	94	75	83	56	79	52	79	68	75	88	79	94	83
2018	79	54	60	100	75	94	60	79	54	79	65	75	88	79	94	79
2019	79	52	60	94	75	94	58	79	54	79	65	75	94	83	94	75
Mena	79	52	56	92	76	86	57	80	51	77	62	76	87	78	91	76
Growth rate	1.05	1.04	1.15	1.07	0.95	1.13	1.12	1	1.13	1.05	1.16	1.15	1.19	1.22	1.07	1.10

Table A3. The betweenness centrality of India and RCEP goods trade network from 2006 to 2019.

	Australia	Brunei	Cambodia	China	Indonesia	Japan	Myanmar	Korea, Rep	Lao PDR	Malaysia	New Zealand	Philippines	Singapore	Vietnam	Thailand	India
2006	1.786	0	0	14.416	4.019	6.956	0	4.019	0	0.527	0	0	2.511	0	23.861	0
2007	5.317	0	0	12.551	4.059	6.519	0	0.408	0	0.408	0	0	2.392	2.540	21.995	0
2008	4.296	0	0	9.058	2.487	4.296	0	2.487	0	1.439	0	0	4.772	2.963	19.630	0
2009	1.667	0	0	9.048	0.238	15.556	0	0.238	0	0.238	0	0	4.603	2.143	18.968	2.540
2010	1.439	0	0	7.471	0.106	7.630	0	7.630	0	0.106	0	0	7.471	2.011	18.519	0
2011	2.905	0	0	6.810	0	2.905	0	2.905	0	0	0	0	10.286	1.667	17.857	1.333
2012	0.544	0	0	6.068	0	3.290	0	3.290	0	0.544	0	0	6.068	1.667	17.163	2.317
2013	0.544	0	0	14.116	0	4.671	0	2.449	0	0.544	0	0	7.449	1.667	10.465	0
2014	0.544	0	0	15.703	0	5.782	0	3.401	0	0.544	0	0	4.751	1.667	10.465	0
2015	0.544	0	0	17.608	0	7.687	0	0.544	0	0.544	0	0	4.751	1.667	10.465	0
2016	0.794	0	0	10.905	0	15.508	0	0.794	0	0	0	0	4.714	1.905	10.905	1.143
2017	0.357	0	0	11.230	0	13.690	0	0.357	0	0.357	0	0	5.040	1.905	11.230	2.500
2018	0.544	0	0	15.925	0	9.259	0	0.544	0	0.544	0	0	3.544	1.143	9.259	1.143
2019	0.544	0	0	8.195	0	10.259	0	0.544	0	0.544	0	0	10.259	4.317	8.195	0
Mena	1.56	0	0	11.36	0.78	8.14	0	2.12	0	0.45	0	0	5.62	1.95	14.93	0.78

Through the regression analysis results in Table 3, we draw the following conclusions:

First, it can be seen that Hypotheses 1, 2, and 3 pass the test. TAD is significantly positive at the level of 1%, indicating that the difference in trade agreements between India - RCEP member states and RCEP member states will expand the difference in India's reciprocity. This means that with the formal entry into force of RCEP, the difference in the number of the above two trade agreements will be enlarged, and the reciprocity gap between India and RCEP countries will be further enlarged. IRCEP TA is significantly negative at the 5% level, indicating that the increase in the number of trade agreements between India and RCEP member states will reduce the difference in reciprocity. From this point of view, India’s rejoins RCEP and increasing the number of trade agreements with RCEP member states may become an effective means to reduce the difference in reciprocity. DC is significantly negative at the level of 10%, which means that when India and RCEP member states engage in closer cooperation in goods trade and obtain a higher degree of centrality, the difference in reciprocity will be reduced. Secondly, the result of Hypothesis 4 is beyond our expectation. CC is significantly positive at the 5% level, contrary to our hypothesis, but this may be explained from the perspective of Indian export competitiveness, namely when trade with other countries in all industries too closely, will deepen India's reliance on foreign imports, and the weak domestic industry impact, these will be a negative impact on India's trade reciprocity. However, the specific principles need to be further discussed in the future. Finally, the empirical test results of Hypothesis 5 are not ideal, and BC is negatively correlated with the difference in reciprocity but not significant, this will be further analyzed in the robustness test. In general, the results confirm that an increase in the difference in the number of trade agreements between India and RCEP member countries will increase the gap between India and its reciprocity. India's promotion of trade liberalization with RCEP member states and the establishment of fair and mutually beneficial trade relations with more countries will effectively reduce the difference in reciprocity.

To test the robustness of our results, we chose another model to test our data to improve estimation efficiency. The fourth column of Table 3 shows the estimated results of the GMM model. In addition to the betweenness centrality result becoming important, other results also confirmed our hypothesis that the in-depth cooperation between India and RCEP orientation has a positive impact on transforming trade reciprocity and will improve what the Indian government calls ‘inequality’. The possible reason for the significant change in the results of betweenness centrality is that most of India's centrality index is 0, which makes the results insignificant under the OLS model, while the GMM model better recognizes the impact of subtle changes in this indicator on the reciprocity gap. Under the GMM model, BC is significantly negative at the level of 1%, indicating that India's promotion of betweenness centrality in the network and becoming a ‘hub’ role, will help improve the reciprocity between India and the RCEP member countries' goods trade network.

Our results are consistent with (Frank et al. 2018), that reciprocity is a universal mechanism for achieving international cooperation, and mutually beneficial countries show a higher level of stable cooperation. Just as the RCEP member states have shown, although the negotiation process has gone through several twists and turns, the high reciprocity among the countries ultimately prompted the member states to reach broader cooperation in the field of trade.

6. Conclusion

The reciprocity index is a result indicator of trade policy and the status of the trade network. The reciprocity of regional trade is affected by the national trade policies and trade status. It should be regarded as the goal of promoting collective action among its member economies and as a means to achieve this goal. Although it can be regarded as a directional indicator for warning and decision-making reference, providing suggestions for policy coordination among countries, it cannot be used as a basic index for trade policy adjustments.

The difference between the reciprocity coefficient between India and RCEP member countries and the reciprocity coefficient within RCEP member countries has been expanding, and it is directly related to the trade policy orientation of India and RCEP member countries. RCEP member countries have adopted more trade promotion policies and signed more free trade agreements. Although India has also signed some trade agreements with RCEP member states, there is a gap with RCEP member states in terms of the number and the strength of promoting bilateral trade development. From the empirical results, the RCEP member states have effectively increased the trade reciprocity within the member states by signing various free trade agreements and building closer, stable, and fair network relations. This is a main reason for the temporary difference in reciprocity between India and RCEP. Similarly, during the signing of the New Free Trade Agreement between India and RCEP member countries, India's trade reciprocity has also been effectively improved. Therefore, it is not a rational choice for India to withdraw from RCEP on the grounds of reciprocity and protection of domestic industries.

Since India implemented economic reforms and trade liberalization in the 1990s, it has had a structural impact on the Indian manufacturing industry. Among them, industrial specialization, where import tariffs are reduced the most, has grown the fastest. In addition, trade liberalization has effectively helped India improve the technological content and comparative advantages of some domestic industries (Alessandrini et al. 2011). However, in the past development process, India’s failure to cash in on the emerging opportunities for specialization within international exchange within global production network (GPNs), and the country is also facing infrastructure bottlenecks and labor market rigidity, which has caused India to be in a passive situation for a long time (Bhat 2011). The broader market and new opportunities brought about by RCEP are expected to help India improve this situation. From a long-term perspective, RCEP member states have an open attitude, and under the influence of global economic development trends, the Indian government may change its cognitive model and will join RCEP in other forms or apply to join again in the future.

In addition, India's position in the RCEP trade network including India has been continuously improved, and the reciprocity coefficient has changed in the same direction. It shows that the higher the status of a country in the network, the greater the difference in policy orientation adopted, and the greater the impact on the trade reciprocity of the network. The status of a country in the trade network (degree centrality, betweenness centrality) is a regulatory intermediary variable, which amplifies the impact of differences in trade policies between subjects.

The last, it can be inferred from the empirical results that after the signing of the RCEP agreement, member states will follow the agreement to reduce tariffs, which will enhance the mobility of commodities, investment, and economic factors, and the trade reciprocity between member states will be further enhanced. The influence of RCEP in the global economy will be further expanded, and more regional economies will be attracted to carry out economic and trade cooperation with RCEP, and the reciprocal spillover effect of RCEP will be more significant.

In summary, although the difference between the reciprocity coefficient between India and RCEP member states and the reciprocity coefficient within RCEP member states has been increasing in recent years, India and the RCEP member states have a good foundation for economic and trade cooperation and huge market potential, it is difficult to completely ‘decouple’ the economic and trade relationship between the two (Wardani and Cooray 2019; Wignaraja 2018). Judging from the results, India will improve its competitiveness in key domestic areas in the future, correct weak areas, and implement a stronger export strategy to promote the development of intra-industry trade and inter-industry trade. At the same time, it accelerates free trade negotiations with Australia and New Zealand, and comprehensively enhances its position and influence in the India and RCEP goods trade network. At the corporate level, the gap in reciprocity in the comprehensive field of trade in goods does not mean that India lacks reciprocity in all industries. The reciprocity index within a certain period of time is only a comprehensive manifestation of the phased domestic industrial structure, national competitiveness and trade policy orientation. Low reciprocity does not mean that the company is unprofitable in the international market. Under the guidance of government policies and the promotion of existing free trade agreements, companies will continue to explore trade and industrial cooperation with RCEP member countries to tap a broader market. To conclude, the formation of government policy depends on various factors such as bilateral relations and critical domestic industry growth, but not solely on any single factors such as trade reciprocity. We hope our analysis on trade reciprocity can be used a tool for policy formulation.

The limitations of this paper can pave the way for future research. The data of this study is to determine and build a network based on a benchmark value of bilateral trade volume, and to calculate related indicators based on this. The actual trade network is a network with weights, countries with different economic scales and different economic strengths obtain social capital through the network quite differently. Defining reciprocity in a non-weighted way will hinder the understanding of the degree of reciprocity between mutual nodes. The limitations of data processing methods limit the depth of the research, and the method of weighting and directed networks will be adopted in the future. At the same time, it should be emphasized that trade reciprocity between countries is affected by a variety of factors, which are ultimately mapped to the flow of trade between countries to calculate reciprocity. The indicators selected in this article can only show the final impact of these factors to a certain extent.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Changping Zhao

Dr. Changping Zhao graduated from Shanghai Jiaotong University with a PhD in management. He has trained more than 50 postgraduate students. He is currently a professor at the Business School of Changshu University of Technology. His research interests include social network analysis, international trade.

Xinli Qi

Miss Xinli Qi is currently pursuing her master's degree in International Business from the School of Shipping Economics and Management, Dalian Maritime University. Her research interests include social network analysis, RCEP, international trade.

Yu Gong

Dr. Yu Gong obtained his PhD at the University of Exeter, he is an Assistant Professor at the University of Southampton, Director CORMSIS Business Liaison (China) at Southampton Business School, University of Southampton, UK. His research interests are sustainable supply chain management and supply chain innovation.

Xiaoling Feng

Dr. Xiaoling Feng, PhD in economics, is a Professor and doctoral supervisor of Shipping Economics and Management School, Dalian Maritime University. Her research interests include RCEP, international trade.

Xuping Cao

Dr. Xuping Cao, obtained his PhD in Economics from Nanjing Forestry University, and is a Professor and Deputy Dean of the Business School of Changshu Institute of Technology. His research interests include regional economy and international trade.

Yun Zhang

Dr. Yun Zhang, obtained his PhD in Management at Wuhan University of Technology. He is Head of Strategy and IT Transformation at China Triumph International Engineering.

Appendix 1.

Calculation results of degree centrality, closeness centrality, and betweenness centrality

Appendix 2.

Form of trade agreements signed between India and RCEP member states from 2006 to 2019

Table

Agreement name	effective date	Expiration date	Australia	Brunei	Cambodia	China	Indonesia	Japan	Myanmar	Korea, Rep	Lao PDR	Malaysia	New Zealand	Philippines	Singapore	Vietnam	Thailand	India
Japan - Malaysia	2006 /07 / 13	2021						√				√
Korea, Rep-Singapore	2006 /03 /0 2 /	2016								√					√
Japan - Thailand	2007 / 11 / 01 /	2022						√									√
Brunei-Japan	2008/ 07 / 31 /	2023		√				√
ASEAN-Japan	2008/12/01	2026		√	√		√	√	√		√	√		√	√	√	√
Japan-Indonesia	2008/07/01	2023					√	√
Japan-Philippines	2008/12/11	2023						√						√
China - New Zealand	2008/10/01	2019				√							√
China - Singapore	2009/01/01	2010				√									√
Japan-Vietnam	2009/10/01	2026						√								√
ASEAN -Australia- New Zealand	2010/01 /01	2025	√	√	√		√		√		√	√	√	√	√	√	√
New Zealand-Malaysia	2010/08/01	2025										√	√
ASEAN-Korea, Rep	2010/01/01	2024		√	√		√		√	√	√	√		√	√	√	√
Korea, Rep-India	2010/01/01	2019								√								√
ASEAN-India	2010/01/01	2024		√			√		√		√	√		√	√	√	√	√
Hong Kong, China-New Zealand	2011/01/0 1	2016				√							√
India-Malaysia	201 1/02/18	2019										√						√
India-Japan	2011/08/01	2026						√										√
New Zealand-Chinese Taipei	2013/12/01	2025				√							√
Malaysia-Australia	2013/0 1/01	2026	√									√
Singapore-Chinese Taipei	2014/4/19	2028				√									√
Korea, Rep-Australia	2014/1 2/12	2033	√							√
Australia- China	2015/12/20	2029	√			√
China-Korea, Rep	2015/12/20	2034				√				√
Japan-Australia	2015/01/15	2031	√					√
Korea, Rep-Vietnam	2015/12/ 20	2029								√						√
Korea, Rep-New Zealand	2015/12/30	2034								√			√
CPTPP	2018/12/30	2039	√	√				√	√				√		√	√