From gravity models to supply chain resilience: a bibliometric analysis of free trade agreement research

International trade studies have, over the past decades, concentrated on trade liberalization and free trade agreements (FTAs) in a quasi-structural way; the gravity model has become the prevailing empirical model of measuring bilateral trade flows and policy implications. Since its conceptualization in international economics, the gravity model has been widely used in measuring the effects of FTAs, regional trade agreements, and multilateral trade arrangements on trade volumes, trade creation, and trade diversion. Consequently, a large and methodologically advanced literature has been created based on efficiency gains, market access, and welfare outcomes related to trade openness (Baier and Bergstrand, 2001).

But in more recent years, world economic shocks such as the COVID-19 pandemic, geopolitical tension, trade wars, and climate-related shocks have exposed structural weaknesses in worldwide supply chains in production and trade. These happenings have diverted scholarly and policy focus on the matters of fragility, resilience, and export concentration in the supply chain. It is in such a dynamic environment that the criterion for judging international trade has not been just efficiency and volume growth anymore, but as a system that is resilient to shocks, persistent, and that is able to curb systemic risks. Traditional trade policy analysis is therefore starting to overlap with concepts like supply chain resilience, global value chain reconfiguration, trade network robustness, and others.

Even though there is an increasing convergence, the literature is still disjointed. FTAs and gravity models research are traditionally based in international economics, while the supply chain resilience research is based primarily in operations management, logistics, and industrial organization. Although new works are trying to connect these areas, the degree to which the aspects of resilience have been incorporated into the literature of gravity-based trade is yet to be understood. Furthermore, the increasing growth and spread of publications covering trade agreements, supply chains, and resilience have also rendered the process of gaining a consistent and clear picture of the field by using only traditional narrative reviews more challenging.

Bibliometric analysis in this sense provides a clear and rigorous method of mapping the intellectual, conceptual and social organization of a research domain. Through quantitative treatment of bibliographic information, bibliometric techniques enable researchers to assess the scientific output, recognize the contributions that have an impact, unravel thematic patterns and reveal new research fronts. Bibliometric analysis based on the framework suggested by Donthu et al. (2021) consists of a combination of performance analysis in terms of productivity and impact indicators and science mapping methods that look at relations between the publications, the authors, the keywords and the citations.

In line with this, this article is aimed at offering a clear-cut bibliometric review of the literature on the intersection of FTAs, gravity models and supply chain resilience. To be more precise, this research is expected to gauge the scientific excellence of the area by determining the most impactful authors, journals, states and documents, as well as pinpoint the intellectual roots and topic development of the literature utilizing citation-based and keyword-based searches, focus on the collaboration patterns and networking structures within the research community and, lastly, discover significant research groups and upcoming trends, including how the concept of resilience and attention to the supply chain can be integrated into the conventional gravity FTA model.

By systematically organizing and synthesizing nearly four decades of research, this article contributes to the literature in three ways. First, it clarifies the dominant knowledge bases and methodological traditions shaping the field. Second, it highlights how resilience-oriented research has evolved relative to established trade theories. Third, it provides a structured research agenda for future studies seeking to integrate trade policy analysis with supply chain resilience and economic robustness.

This research is based on a bibliometric research design where the intellectual framework, theme development and scientific output of the literature on FTAs, gravity models and supply chain resilience are systematically researched. The bibliometric analysis is especially applicable to this end, since the size of a body of scientific literature can be quantitatively and reproducibly evaluated, and the subjectivity and limitations of the traditional narrative review are prevented. In accordance with the methodological construct by Donthu et al. (2021), the analysis merges the analysis of performance and science mapping, in addition to the network analysis and clustering methods that help to identify the results of productivity trends along with the knowledge structures.

Bibliometric analysis is based on two pillars. To begin with, the performance analysis compares the productivity and impact of the research elements, e.g. the authors, journals, countries and documents, based on the publication and citation indicators. Second, science mapping determines the connections between these constituents to display conceptual, intellectual and social order in the area. The combination of these methods makes it possible to have the whole picture of the development as well as the present state of the literature.

Only the Scopus database was used to retrieve bibliographic data, as it is a broad-based database to access peer-reviewed literature in the field of economics, international trade and supply chain research. The query was formulated using a structured query applied to TITLE-ABS-KEY fields to ensure that only publications addressing the intersection of trade agreements and gravity-based trade modeling and supply chain or resilience-related issues are retrieved. The last search query was stated as follows:

TITLE-ABS-KEY (((“FTA” OR “FTAs” OR “free trade agreement*” OR “trade liberalization”) AND (“gravity model” OR “gravity equation”)) OR ((“supply chain*” OR resilience) AND (“trade agreement*” OR “free trade agreement*” OR “FTA” OR “FTAs”))).

This search query strategy was structured to make sure that two complementary bodies of literature are ultimately covered. The former stream includes the fundamental trade economics literature in which the impacts of FTAs and trade liberalization have been studied in the form of gravity models or gravity equations. The latter stream reflects added value on more recent and interdisciplinary contributions in terms of connecting trade agreements to supply chain dynamics and resilience considerations. A combination of the two elements in a single query will allow the search strategy to identify both current studies and future studies on the subject, out of which the gravity-based research has been conducted, and which are currently undergoing various developments in the context of trade policy. The resulting data range is 1985–2026 and consists of 988 papers, such as journal articles, conference papers, book chapters and review articles, covered by Scopus. This time span indicates the initial growth of the gravity-based approaches to trade dealings and the recent rise in research on supply chain resilience after economic shocks at the global scale. Peer-reviewed articles in English were only retained to make the bibliometric analysis consistent, of high quality and comparable.

The key descriptive statistics of the data set are summarized in Table 1, which contains the time frame, publications per number, authors of publications, the sources and the global citation indicators, giving a general idea of the volume of the literature under analysis.

Table 2) examines the annual citation pattern, which indicates that there is an evident inverse relation between the volume of publication and the intensity of citation with time. The average citation rates are very high in the early years of the literature, though there are few publications. This is indicative of the existence of foundational literature on gravity-based studies on trade agreements, which gain massive citations during long durations and still carry on with the intellectual basis of the field. The more publications, the more average citations per year stabilize at the lower levels. This loss cannot be viewed as a drop in scientific relevance. Rather, it is indicative of two structural impacts, namely, the reduction in citation window that recent publications have, and second, the diversification of the research topics, with the emergence of the supply chain and resilience-oriented studies, yet to be granted the recognition they merit. It is interesting to note, though, that the post-2020 years are comparatively calm in terms of citation rates despite a significant rise in the volume of publications, which means that the relatively recent contributions are presented at an equal level of academic visibility. This trend is parallel to the life-cycle analysis that puts the field in the late growth to early maturity phase. On the whole, the citation dynamic analysis of the yearly results supports the shift from a field of one or two highly impactful studies to a larger and more varied research domain with continued, but more spread-out, scientific influence of the extent of the size and scope of the literature under study.

Life-cycle analysis of scientific production: In bibliometric studies, life-cycle analysis is applied to identify in which stage of the scientific production life-cycle a field is, its sustainability as a research activity, as well as to determine whether it is in an emergent, growth, maturity or saturation stage. This especially applies to interdisciplinary areas like FTAs and the gravity model, and supply chain resilience, where older theoretical traditions have been coexisting with more recent research directions.

The life-cycle analysis indicates that the literature on FTAs, gravity models and supply chain resilience has a clear logistic growth pattern. In Table 3, the approximate carrying capacity (K = 1,422 publications) results in a limited yet significant potential for research in the long term. Based on Figures 1 and 2, the inflection point (tm = 2023) is the highest growth period in the number of publications each year, and the jump of the curve indicates the interest in the academic field after the serious disruptions in the global supply chains. The transition period (Δt = 18 years) implies that there should be an expansion phase that is rather extended because of the coexistence of a developed literature on gravity-based trade and a fast-growing stream of resilience-oriented research. Although the model shows a slowing down of the number of annual publications as time progresses and reaches the peak, the cumulative curve shows that the field will not reach saturation before the early 2040s. In general, these findings bring the research field to a late growth/early maturity phase, with future contributions likely being in the area of consolidation, integration and applied extensions, and not necessarily in the realm of exploration.

In (Figure 3), most of the relevant sources and authors are analyzed to define the main outlets and the main contributors that develop knowledge production in the research field. Source productivity refers to the journals that are at the heart of spreading research, whereas author productivity indicates those scholars who have been involved in the evolution of the literature constantly. These indicators, in combination, give hints on the institutional and intellectual concentration of the field. This is vital in bibliometric research so as to be able to differentiate between central and marginal outlets and between core and emerging authors, especially in an interdisciplinary field like that of FTAs, gravity models and supply chain resilience. Most of the sources analyzed have a high level of concentration of publications in a small number of specialized journals in international economics and trade. The World Economy has the most outlets, the largest documents and thus, it plays the central role in hosting research on trade agreements and gravity-based analysis of the documents. A high density of specialized outlets organizes the knowledge dissemination in the discipline, with other leading journals like International Trade, Economic Integration and Applied Economics serving as a core of this topic. The existence of interdisciplinary journals and even journals that focus on sustainability implies a slow expansion of the literature with wider subject matters such as resilience and supply chains. The spread of publications by source indicates that there is a moderately concentrated and diversified journal market. Most of the research is published in a wider variety of outlets, although a minor proportion of the publication activity is concentrated in a small pool of journals, indicating the interdisciplinary nature of the subject and growing overlap of the field of trade economics with logistics, sustainability and resilience research. The same tendency of concentration is shown in the analysis of the majority of the most relevant authors. The number of scholars with relatively high publication rates is rather limited, which means that these scholars have been working in the field over time. These two writers have had a major role to play in the formulation of gravity-based trade models and the overall evaluation of trade agreements, which are empirical. Meanwhile, the distribution can be characterized by the long tail of authors with a smaller number of publications, which indicates the open and developing research area that is attractive to new participants, especially in the recent past. Notably, the fact that there is no domineering presence of one author or journal implies that the field is not controlled by the small academic elite. Rather, it is a pluralistic configuration, with older contributors alongside newer scholars who examine new aspects of the supply chain like supply chain resilience. This structure is in line with the life-cycle and citation dynamics that were found above, and further supports the notion of the field being in a late growth to early maturity phase, characterized by both consolidation and thematic diversification.

Figure 4 investigates the most appropriate words in the literature on the topics of FTAs, gravity models and supply chain resilience, presenting the frequency of keywords in the reviewed publications. It is remarkable that the phrase “gravity model” was predominant in the figure, which proves its centrality as the most significant analysis tool in the area. This finding is indicative of the underlying theoretical basis of the literature, which mainly uses gravity-based specifications as a means to determine the impact of trade agreements on the movement of trade and integration into the economy. Next to this basic idea are the high-frequency keywords of a nature like international trade, trade agreement, free trade and trade liberalization, and these reveal that the primary empirical interest should be on how preferential trade arrangements and policy liberalization affect bilateral trade performance. All these terms are characteristics of a mature research core that focuses on the classical international trade theory. Simultaneously, Figure 4 also shows the increasing popularity of the keywords concerning the use of the terms supply chains and supply chain management, which implies a gradual movement toward the supply chains' resiliency outlook. Since these terms are not active as often as the keywords related to gravity and trade, their popularity points to a thematic expansion that incorporates the global value chains and supply chain vulnerabilities in the framework of trade agreement analysis. The use of geographically specific keywords, which include China and the European Union, also evidences the applied and policy-relevant nature of the literature. On the whole, Figure 4 demonstrates the presence of a dual thematic framework, with an established center of gravity-based trade and an emerging field of study–supply chains and resilience. This structure favors the description of the field as being late growth to early maturity stage, characterized by theoretical consolidation and conceptual diversification.

Figure 5 and Table 4 indicate a distinctively modular co-word arrangement with four thematic clusters, as well as a set of keywords serving to provide conceptual bridges between the field. The international trade with betweenness 120.28, the gravity model with betweenness 117.43 and the trade agreement with betweenness 75.28 are the most central connectors within the network, and they confirm that the literature is still rooted in the traditional trade economics core, with the specifications of the gravity modeling being the dominant one of analysis, as the PageRank of the gravity model is high at 0.0696. Cluster 1 indicates this methodology's traditional and regional applicability extension in terms of keywords like China, ASEAN and industry-based applications like agricultural trade. Cluster 2 focuses on the institutional and policy structure of trade agreements with terms such as trade agreement, free trade, European Union, NAFTA and regional trade agreements, and it depicts the most closeness values around 0.018 to 0.0185, meaning that these notions are at the conceptual center of the network. Significantly, supply chain management belongs to Cluster 2 as well, with a betweenness of 28.95, which confirms that it is a gateway connecting mainstream FTA and gravity research to supply chain viewpoints. Cluster 3 includes the outcome assessment layer of the literature in which policy instruments are converted into quantifiable trade impacts, as the policy instruments of trade liberalization of betweenness 38.81 and PageRank 0.0456 are predominant, with trade liberalization, trade flow, export and econometric terms. Cluster 4 is the resilience and risk-oriented stream, which has betweenness 82.56 in supply chains and betweenness 51.22 in the FTA, along with the risk tree analysis, risk analysis and sustainability-related terms; Bridging roles of the supply chains and FTA are high, which means that this stream is becoming a part of the core of the trade policy instead of an isolated phenomenon. On the whole, the co-word net testifies to a dual structure, with a mature gravity and FTA base and a broadening resilience-oriented research frontier that is linked by mediating concepts that are indicative of continued thematic and theoretical incorporation.

The most relevant authors in the Scopus corpus are displayed in Table 5, and they are sorted by the number of publications and the fraction of authorship contributions. Authorship distribution of the articles demonstrates a moderately concentrated pattern, with the existence of a small sect of the most prolific scholars and a broader category of periodically contributing scholars. Regarding the number of articles published, Baier Scott L, Bergstrand Jeffrey H and Martinez-Zarzoso Inmaculada become the most active contributors with nine articles each to the corpus. Their eminence indicates the longstanding presence of deep research priorities in gravity-based research in trade and explicit trade policy assessment. Other key contributors like Yotov Yoto V, Larch Mario and Anderson James E, among others, extend this base as follows, signaling consistency with the early strands of international trade theory and applied econometric modelling. Comparison of the raw publication counts with fractionalized article measures demonstrates that the fractionalized contribution is consistently reduced in the most prolific authors, implying that most of the research production in this area is produced on the basis of collaborative authorship modes. This trend raises the significance of co-authorship networks and collective research projects, which are increasingly required by the methodological complexity and informational intensity of modern empirical trade research. The fact that the biggest category is the N/A category with 13 articles indicates that author metadata is not fully filled or is irregularly filled in a part of the records. Although this is a limitation of the data, it does not change the general composition of the authorship distribution. The literature is structured in a highly structured way with a consistent core of contributors, which is supported by an increasing pool of other authors involved in the process of the field expansion. The ranking of the authors, in general, offers evidence that the research field is backed by a consistent epistemic community, though at the same time it possesses a gradual diversification in terms of participation as new themes and extensions of analysis are still integrated.

The thematic map in (Figure 6) depicts a well-organized and dynamic research framework that has a powerful trade-policy center and a growing resilience-based frontier. The themes related to trade agreements, free trade and trade liberalization fall under the motor themes quadrant and are highly centralized and highly dense, meaning that they are well developed and core in structuring the field. These themes align with the mainstream gravity-based application of FTAs and ensure that the analysis of trade agreements is the tool that pushes the literature.

The basic themes quadrant is defined by the very central and less internally developed themes, such as: gravity model, FTAs, China and international trade. These themes constitute the conceptual and methodological core of the field, which serves as a constructive component that bridges various streams of research. Their stance indicates that these subjects are necessary and universal, but still they are far and to be varied to suit new situations instead of being exhausted.

Supply chains, fault tree analysis, risk assessment, commerce, globalization and COVID-19 are some of the themes that fall under the niche themes quadrant. These themes are characterized by high levels of internal development and less centrality, which reflect specialized and relatively closed lines of research. Their location is evidence of the fact that resilience and risk-focused studies are interdisciplinary and rely extensively on engineering, operations management and sustainability sciences methodologies. These themes are yet to be central in the overall structure, but they bring advanced tools and perspectives that can help to enhance the field.

Lastly, the emerging or falling themes quadrant is where supply chain management and FTAs are under a more general conceptualization. The low density and moderate centrality of these themes indicate that they are either at a low stage of development or are in redefining stages as they become part of the larger analysis structure. With the current surge of resilience-oriented research, the placement of supply-chain-related words can be more predictive of an emergent trend than a downward trend toward further incorporation into the main trade-policy discourses.

All in all, the thematic map supports earlier results from the co-word network and clustering analyses of the resilience and risk-seeking themes, demonstrating a centralized trade-agreement core with mature, specialized and growing resilience. The trend of increasing supply chain complexity and resilience issues with higher centrality implies that future research will increasingly focus on assessing trade agreements in relation not only to trade creation but also to resilience, weakness and systemic vulnerability within global value chains.

The results show that the traditional gravity model literature and the newer literature on supply chain risk, sustainability and resilience are not mutually exclusive but rather co-exist. The gravity model, international trade, trade agreements and trade liberalization are all more or less strongly represented in the term, reflecting the continued strong influence of international trade economics. This finding is in line with the longstanding application of gravity models to trade creation, trade diversion, market-access effects and the different impacts of economic integration agreements on countries, sectors and trade margins (Baier and Bergstrand, 2007, 2009; Baier et al., 2014, 2018, 2019; Egger et al., 2022). Concurrently, the role of supply-chain management as a bridging concept and the emergence of concepts related to risk, sustainability and resilience in the existing trade-policy literature suggest an increasing linkage of logistics and production-network aspects with the common framework of trade policy literature.

This study then recognizes the principal shift as not from the gravity framework to the framework of vulnerability, continuity, adaptation and recovery to shocks, but rather the steady process of this extension. The methodological underpinning is still gravity modelling, which has been complemented by supply-chain research, and which incorporates new issues related to the structure of production networks, reliance on key suppliers, vulnerability to logistical constraints and cascading country and industry effects. This interpretation is confirmed by the studies that interpret the provisions of free-trade agreements in the design of the global supply chain and its implications on production costs, carbon emissions and the changes in the configuration of the supply chain during disruption (Kinoshita et al., 2023; Nagao et al., 2022).

The performance measures also indicate significant growth in the field over the last few years. The estimated inflection point is around 2023, and the fitted life-cycle model shows that growth was fast during the 2020s, around the time of the COVID-19 pandemic and the resulting global disruption. This estimated inflection point should be considered a modeled turning point and not the actual number of publications. This increased number of publications since 2020 is in line with growing interest in scholarly research on disruptions in transport systems, limited availability of intermediate goods and services, geopolitical tensions and global value chain vulnerabilities. However, bibliometric evidence cannot prove that the pandemic was responsible for the increase. Other areas of development such as nonagricultural conflicts, environmental events, the digital revolution, and evolving industrial policies could also have aided in the growth of the literature.

The reduction of average citations for recent publications should not be taken as a sign of less scientific relevance either. The longer periods of citation given to earlier gravity- and FTA-related publications, and the growth of this body of intellectual work, have helped them to find their place in the field. In contrast, resilience-focused studies have not yet had as much time to cite and are spread out among a wider variety of journals and subject areas. As a result, the pattern of citations observed must be interpreted as the sum of citation-window effects and thematic diversification.

From the thematic map and co-word network, it can be seen that “trade agreements” and “trade liberalization” are very central themes, while “gravity models” are methodological themes at the core of many different research streams. There is less focus on risk assessment, sustainability, supply chains and disruption-related studies. Their location indicates that resilience research has a fairly well-established internal structure and has not yet been integrated into the general framework of gravity-based trade analysis. The results therefore reinforce the cautious interpretation that the contents of resilience-related topics are increasingly moving closer toward the core of the trade-policy field, but not necessarily into the core. It would have to be a longitudinal thematic-evolution study over separate periods to demonstrate such a movement.

The first theoretical implication involves the outcome that is normally studied using gravity models. The explanations of bilateral trade flows by economic size, bilateral trade costs, market access and multilateral resistance are found in structural gravity. It is a theoretically consistent method to calculate partial- and general-equilibrium effects of FTAs, tariffs, non-tariff barriers and other trade policies (Baier, 2010; Anderson et al., 2018; Yotov et al., 2016). Many of the existing applications have greatly enhanced the identification of FTA effects by addressing the issue of endogeneity, unobserved bilateral heterogeneity, zero trade flows, general-equilibrium adjustment, and product-specific and country-specific effects (Baier and Bergstrand, 2007, 2009; Baier et al., 2018; Egger et al., 2022; Larch et al., 2023). In the literature, structural gravity is known to be one of the key tools in the partial- and general-equilibrium toolboxes of trade policy research (Anderson et al., 2021; Anderson and Yotov, 2025).

However, the typical gravity applications only measure the long-term, static, or average impacts on trade quantities and trade structure. They do not necessarily reflect the ability to maintain trade operations following a major disruption or how quickly trade operations resume after a disruption. The analytical angle of supply-chain resilience is a complementary dimension related to the ability of trade and production chains to anticipate disruption, absorb the impact, maintain critical activities, restore normal functioning, and adapt to the changes.

From a production-network point of view, the evidence is theoretical because production-networks are important. Such an event on one set of companies can be spread up and down the chain to suppliers and customers, causing ripple effects that result in losses not just at the point where the shock itself hit, but further spread. For instance, Carvalho et al. (2021) demonstrate that the impact of the Great East Japan Earthquake cascaded through buyer–supplier linkages and had a wider macroeconomic impact. Resilience needs to be understood as a property of the system of interconnected trade and production networks, and not just as a property of a bilateral relationship (Todo et al., 2015).

With the help of a gravity model that is equipped with resilience, three interrelated dimensions may be discerned. Exposure is defined as reliance on specific partners, suppliers, critical products, ports or transport corridors. Sensitivity is defined as the size of the losses to trade or disruption of trade logistics after shock. Adaptive capacity refers to the capacity to replace suppliers, find alternative trade routes, enter new markets, re-organize production, and re-open trade flows. In addition to providing FTAs with credit for their average trade-creation effects, this extension would provide credit for mitigating the initial impacts of shocks and speeding up the subsequent recovery.

The relation between FTAs and resilience is theoretically unclear. On the other hand, an agreement can enhance resilience through a reduction of bilateral trade costs, expansion of the pool of economically viable suppliers, easier market substitution, lower regulatory uncertainty and better customs and institutional cooperation. Deep trade agreements also can encourage involvement in GVCs through provisions that affect behind-the-border policies that affect services, investment, standards, and production ties. In a theory-consistent gravity framework, Kim et al. (2024) conclude that deep trade agreements have an impact on integration into agri-food global value chains.

In contrast, preferential integration could lead to risks for firms and nations since it can foster a focus on a specific group of partners in terms of imports, exports, or production. An agreement can bring efficiency gains to the agreement party and make the party more dependent on a given supplier, intermediate input, production center or logistical corridor. New findings suggest that tariff and digital-trade-cost cuts could raise supply- and demand-side risk exposure of industrial chains, and there is significant cross-country and cross-sector variation (Xiao et al., 2026).

Such fuzziness makes it difficult to conclude that integration of trade is either resilience-enhancing or resilience-reducing. While global supply chains can lead to shocks, they can also enable the access to other suppliers and markets in case of shocks or disruptions at the local level (Baldwin and Freeman, 2022). Likewise, the WTO (2021) posits that international connectivity can simultaneously provide vulnerability to shocks and provide resilience to respond to such shocks by engaging in diverse goods, technologies, suppliers and markets. The question, therefore, that is theoretically relevant is not only whether or not an FTA leads to more trade but whether or not the overall trade structure that is created is the appropriate one in terms of efficiency, diversification, substitutability, and systemic dependence.

The bibliometric results then lend support to the extension of the gravity framework from an analysis of trade performance under normal conditions to an analysis of trade continuity and recovery under disrupted conditions. No sacrifice of the existing gravity model is made in this extension. Rather, indicators of resilience, interactions with shocks, network structures, and recovery actions can be integrated into gravity-based identification approaches.

The results have direct implications for international logistics, as the ability of bilateral trade to withstand shocks goes beyond tariff and geographic considerations, and extends to the capacity of transport, border, and information systems to operate. The most common gravity models typically rely on distances as a proxy for trade costs. While distance is a factor, it is not a measure of customs-clearance time, port congestion, frequency of maritime services, variability of transport time, border reliability, digital documentation capacity, or alternative transport routes. World Bank's Logistics Performance Index (Arvis et al., 2023) encompasses six dimensions that are of special importance for this extension: customs and border-management efficiency, trade and transport infrastructure, ease of arranging international shipments, logistics-service quality, tracking and tracing, and shipment timeliness. Additionally, the report illustrates the significance of reliability, lead time spread and dwell time at ports and gateways. These dimensions offer measurable variables that can be added to future models of gravity and resilience.

Future gravity-based analyses could therefore complement geographical distance with measures such as:

1. customs-clearance and border-crossing time;

2. port and airport dwell time;

3. liner-shipping connectivity and service frequency;

4. transport-time variability;

5. number of accessible transport corridors;

6. digital-documentation and tracking capacity;

7. logistics-service quality;

8. availability of alternative ports, routes, and transport modes.

These variables would allow for the separation of the average cost of bilateral trade from the ability of the logistics system to sustain trade with the disruption of one route, port, supplier or border crossing. There are also managerial implications of the bibliometric link between the research on supply-chain management and trade policy. The firms that want to take advantage of an FTA would have to seriously examine the tariff preference granted through the agreement and the robustness of the sourcing and distribution arrangement. It can include qualified suppliers, the substitutability of critical inputs, rules of origin, availability of alternative transport routes, inventory policies, shipment visibility and the time it takes to change suppliers or markets. Kinoshita et al. (2023) depict the way FTA utilization is incorporated into global supply-chain design decisions based on production costs and carbon emissions, and Nagao et al. (2022) study supply-chain network redesign in light of the disruptions caused by the COVID-19 pandemic. The findings of these studies, however, show that the impact of an FTA is not limited to the tariff cut, but is influenced by the structure of the production and logistics networks (Nakamura et al., 2017, 2019; Nagao et al., 2021, 2023; Kotegawa et al., 2024).

The findings also suggest that resilience-conscious trade policy should extend beyond tariff liberalization. The potential contribution of an agreement to resilience depends on whether its provisions reduce administrative delays, informational asymmetries, regulatory incompatibilities, and logistical uncertainty.

Relevant provisions may include:

1. customs cooperation and coordinated border management.

2. pre-arrival processing and electronic documentation.

3. paperless trade and interoperable digital systems.

4. mutual recognition of standards and conformity assessments.

5. expedited treatment of essential and perishable goods.

6. regulatory cooperation and information-sharing mechanisms.

7. emergency consultation and crisis-response committees.

8. transit guarantees and cooperation on alternative corridors.

9. flexible administrative procedures during severe disruptions.

10. cooperation on transport infrastructure and port connectivity.

These measures can impact the time, certainty and speed of cross-border flows and could thus impact the resilience of trade relationships even if the tariff preferences are unchanged. Both evidence and OECD policy principles highlight the need for flexible, responsive and coordinated policy frameworks in supply chains and for intergovernmental, firm and international cooperation to strengthen supply-chain resilience. It also emphasizes the need for digitalization, trade facilitation and regulatory coordination.

These provisions may have varying application across agreements and industry. The pharmaceuticals, food, energy, semiconductors or other critical inputs may need more robust emergency-coordination and information-sharing mechanisms. The criteria that may matter more for landlocked economies could be related to transit guarantees and corridor diversification, while for maritime-dependent economies, port connectivity, shipping frequency and access to alternative gateways could be more important. It is therefore necessary to have sectoral heterogeneity. Previous empirical gravity research shows that FTAs can affect intermediate goods, sectors, margins of trade, and type of agreements (Baier et al., 2014, 2018; Cardozo et al., 2022). A resilience perspective adds to this heterogeneity concerning criticality of inputs, supplier substitutability, storage capacity, and transport dependency.

This means that there are no implications that call into question undiscriminating reshoring or the systematic reduction of supply chains. Reshoring will also not mitigate some of the foreign shocks that have occurred, will raise the domestic shock sensitivity, and will lose any diversification that international markets have provided. The OECD (2025) also summarizes that the implementation of effective supply-chain risk management should not involve withdrawing from international trade but that well-functioning and diversified global markets can help to build resilience. It is therefore appropriate to increase adaptability, transparency, substitutability, and recovery capacity, not to eradicate international interdependence.

To better understand the implications of the bibliometric results for the development of the field, the main findings of the present study are linked to existing theoretical arguments and implications from the international trade, logistics and supply-chain-resilience literature in Table 6. The table distinguishes and clarifies the evidence collected through bibliometric analysis from interpretations from previous research. This distinction is important since although the location of keywords and the level of thematic proximity point to the intellectual structure of the literature, this is not enough to confirm the economic mechanisms by which FTAs impact logistics performance or supply-chain resilience.

To illustrate the relevance of the bibliometric results in relation to the development of the field, Table 6 links the key aspects of the present study with already existing theory and practice from the world of international trade, logistics and supply-chain resilience. The table differentiates between evidence produced in the process of conducting the bibliometric analysis and interpretations that have been produced in previous research. The distinction is relevant as keyword centrality and thematic proximity reveal the intellectual structure of the literature without a clear indication of the economic transactions on which FTAs rely to impact logistics performance or supply-chain resilience.

Table 6 illustrates that continuing to pursue research grounded in gravity doesn't contradict the growing trend of supply-chain resilience as a research priority. While structural gravity offers theoretical and econometric tools to reveal the impact of FTAs on bilateral trade costs and trade allocation, resilience and network approaches add exposure, shock propagation, supplier substitution and recovery as additional outcomes. The use of these perspectives would then take FTA assessment beyond its traditional theoretical framework, rather than substitute it. It would also allow for the testing of whether an agreement could increase trade and increase adaptive capacity or whether increased trade is correlated with increased partner, product, input or logistical concentration.

The outcomes suggest a research direction that integrates structural gravity estimation, shock analysis, input–output linkages and network approaches. A first empirical approach would test the hypothesis that bilateral trade between FTA partners decreases less during a disruption identified by an external force than bilateral trade between nonpartners. This could be examined by cross-regressing FTA membership with the shocks, while holding fixed other factors such as exporter-time, importer-time and country-pair fixed effects. But resilience should not be equated with the amount of drop. As for future research, a minimum of four outcomes should be distinguished:

1. Resistance: the reduction in trade during the initial shock;

2. Persistence: the duration of the negative effect;

3. Recovery: the time required to return to the pre-shock trade trajectory;

4. Adaptation: the capacity to switch suppliers, products, markets or transport routes.

Potential dependent variables of interest would include the percentage reduction in bilateral trade, volatility of trade following the shock, recovery time, likelihood of supplier substitution, change in number of trading partners, change in concentration of imports or exports, and trade diversion to other markets. The more granular the level of trade analysis (by product and industry), the better, since there are significant variations in the degree of resilience that is needed across industries. Where there are few substitutes for the critical intermediate input, it cannot be treated like a final consumer good with a number of different suppliers. The differentiation between intermediate and final goods proposed by Cardozo et al. (2022) is a good starting point and the concepts of agreements, sectors and countries for which integration is less or more vulnerable could be further developed by Baier et al. (2018) and Baier and Regmi (2023). Depending on the data available, structural gravity could calculate FTA effects on bilateral trade costs and geographies of trade. Direct and indirect dependence on foreign intermediate inputs could be identified using input–output analysis. Network analysis can then be used to identify central suppliers, bottlenecks, communities and shock propagation channels. Ren et al. (2024) demonstrate that the robustness of RTA networks is related to their structure and the role of highly connected countries. Armed with these strategies, it would be possible to assess bilateral adjustment and systemic interdependence.

To complement gravity, event-study and difference-in-differences designs could be used to analyze trade dynamics leading up to, during and after significant disruptions. Examples of suitable shocks include pandemics, natural disasters, geopolitical conflicts, disruption of transport, sanctions or a sudden policy restriction. The credibility of these designs would rely on the exogeneity and geographic/sectoral dimension of the shock, and on the validity of the comparison group and the way in which anticipatory adjustments are treated.

The results have to be interpreted with caution due to a number of limitations. First, the intellectual and thematic association that is shown by the bibliometric co-occurrence does not constitute a causal link between FTAs, logistics performance and supply-chain resilience. The FTAs may have a reinforcing or weakening effect on resilience is thus a research proposition based on a theoretical and empirical understanding and not a causal consequence from the present study. Secondly, only publications indexed in Scopus might not be found if they are included in other databases, such as Web of Science and Dimensions. Third, recent studies have shorter citation windows, and this does not allow for direct comparison with the older publications. Fourth, the results of bibliometric analyses are dependent on the research design, keywords, author terminology and quality of the available metadata. A major restriction is the acronym “FTA,” which can stand for FTA and fault tree analysis. The isolated acronym might bring back records unrelated to trade agreements because FTA is often applied in the engineering and supply chain risk assessment context (Chen et al., 2009; Kumar Mangla et al., 2016; Chodur et al., 2018; Gupta et al., 2022). Therefore, a robustness analysis removing the records of the fault-tree analysis should be conducted before interpreting the corresponding risk cluster as converging research from the free-trade-agreement with research on supply-chain resilience.

A bibliometric analysis in Scopus was conducted to analyze the development of the literature in the field of FTAs, gravity models and supply-chain resilience. The study aimed to understand if resilience-related research is moving away from traditional trade-policy analysis or if it is increasingly aligned with the existing gravity-based literature by performing a performance analysis, a keyword co-occurrence graph, centrality analysis, thematizing and analyzing the study's life-cycle. The results suggest that the “Gravity” model is still the major methodological and conceptual basis of the field. Terms associated with international trade, trade agreements, trade liberalization and gravity modeling have core roles in the bibliometric structure, supporting the ongoing dominance of the international trade economics tradition. Concurrently, a more focused research field on supply chain management, sustainability, risk management and disruption concepts has emerged, which are yet more closely intertwined. The bibliometric results indicate, then, not a replacement, but a sort of coexistence and gradual extension of the gravity framework in the direction of the study of vulnerability, shock transmission, adaptation and recovery.

The primary theoretical issue is whether the gravity model approach to estimating FTAs can be extended beyond the traditional welfare gains, trade diversion and trade creation. While the structural gravity model can be used for estimating impacts of trade agreements on trade costs, access to markets and trade allocation, the supply-chain and production-network models can be used to complement the analysis of exposure to disruption, supplier dependence, shock propagation, substitutability and recovery capacity. The integration of these views would enable the assessment of whether an FTA would simultaneously boost trade and enhance resilience, and whether it would lead to more trade with a narrower group of partners, products, inputs, ports or transport corridors. The research also sheds light on relevant implications of logistics and trade policy. In addition to tariffs and geographical distances, bilateral trade relationships are resilient if customs is efficient, borders reliable, ports are connected, transport time is not too variable, documentation is digital, shipments are visible, and there are alternative routes and suppliers. Therefore, provisions on customs cooperation, regulatory compatibility, digital trade, transit facilitation, mutual recognition, information sharing, emergency coordination and other aspects should be taken into account when future trade agreements are assessed. Such measures can impact trade and logistics systems capacity to keep functioning and recover during disruption.

The findings should not be concluded to be evidence that trade agreements always enhance the resilience of supply chains, however. Greater integration can lead to access to more alternative markets and suppliers but can also lead to reliance on fewer partners or critical inputs. Theoretically, FTAs have a mixed impact on resilience, so it needs to be empirically tested. Future studies might take a closer look at resistance during shocks, the sustainability of trade losses, the recovery rate from a disaster, partner substitution, supplier concentration, and export-market concentration by integrating measures of structure, event-study analysis, and input–output analysis. This study has several limitations. This analysis is restricted to Scopus and is more or less sensitive to citation windows, the quality of metadata, the choice of search terms, and acronyms. Specifically, “FTA” can mean FTA or fault tree analysis, so these may be misleading clusters of risks when reading some of these. Furthermore, bibliometric proximity is an indicator of intellectual connections without being a causal economic impact. Overall, the study shows that the gravity-FTA literature provides a mature analytical base in developing more explicit resilience-oriented research programs. Its biggest contribution is to highlight the conceptual links between the trade-policy evaluation and supply-chain analysis, and to demonstrate how future studies can go beyond trade-creation to incorporate exposure, continuity, adaptation and recovery. Because this extension is especially important for the design of trade and logistics policies that can support economic integration and resilience in a rapidly changing world, this report draws on the latest research on trade and transport. This report builds on the latest scholarship on trade and transport, with a particular focus on trade and logistics policies that can help economic integration and resilience in an uncertain global environment.

This study uses secondary, publicly available macroeconomic data and does not involve human participants, personal data, or animal subjects.