The Impact of High Environmental Standards in Trade Clauses on Bilateral Aquatic Product Value Chain Linkages

Abstract

Aquatic product value-added trade constitutes a vital component of agricultural food security. Particularly in developing coastal nations, aquatic products serve as the backbone of the agricultural sector. However, illegal, unreported, and unregulated (IUU) fishing activities not only disrupt the global marine aquatic products value chain but also accelerate the degradation of marine ecosystems and the depletion of marine resources, posing severe challenges to sustainable fisheries and environmental governance. In 2022, the World Trade Organization reached a consensus on fisheries subsidy negotiations, while regional agreements such as the United States–Mexico–Canada Agreement (USMCA) and the Comprehensive and Progressive Agreement for Trans-Pacific Partnership (CPTPP) incorporated fisheries into relevant provisions under environmental rules. This indicates that high-standard environmental trade agreements are emerging as crucial tools for cross-border fisheries governance. This study employs open economy theory and a deep text protocol database to conduct an empirical analysis of the impact of high-standard bilateral environmental provisions on the interconnections within the aquatic products value chain. Findings reveal that environmental provisions significantly strengthen these linkages by lowering market access barriers, promoting technology spillovers, and reinforcing horizontal and vertical labor division. Heterogeneity analysis further shows that the extent of these effects varies with trade provisions, political distance, and network position. These insights offer new perspectives for seafood exports and upstream–downstream coordination in aquatic products, providing policy implications for regions seeking to enhance their value chain advantages.

1. Introduction

Aquaculture relies heavily on freshwater resources and marine ecosystems [1], with both the processing and treatment of aquatic products consuming substantial organic water resources. To address increasingly scarce water resources, countries have implemented comprehensive multi-sectoral actions to reduce water vulnerability and enhance resilience, thereby safeguarding the sustainable development of the aquaculture industry. For instance, the WTO’s recently concluded Agreement on Fisheries Subsidies curbs excessive water exploitation by preventing illegal fishing. High-standard agreements like the CPTPP and USMCA also establish clear provisions regarding wastewater discharge standards in aquaculture, aquatic environmental governance, and clean water resource utilization. The environmental protection provisions within these high-standard agreements not only influence national water resource management models but also profoundly reshape the aquatic product value chain. The core issue of this paper is to examine whether stricter environmental management provisions will affect the degree of linkage within bilateral seafood value chains, thereby promoting the sustainable development of aquatic product value chains.

The aquatic products value chain is intrinsically linked to global agricultural trade. As globalization advances toward regional integration, aquatic products—a critical component of agricultural trade—are increasingly integrated into global value chains through fishing, processing, and cold-chain distribution. This process fundamentally involves the utilization of water resources. Trade liberalization has intensified the spatial exchange of water resources. For instance, fisheries may operate in freshwater-rich regions like South America or Northern Europe, processing occurs in Asia–Pacific or Africa, while final distribution takes place in North America—forming an interdependent global supply network. Classical trade theory posits that international trade enables countries to leverage comparative advantages, fostering both horizontal and vertical divisions of labor. Existing research largely agrees on the positive impact of regional trade agreements on value-added trade [2,3], as they reduce tariff barriers and enhance bilateral market access, thereby promoting seafood exports and growth. However, developing nations’ reliance on cheap labor and abundant marine resources for extensive fishing severely undermines the sustainable development of aquatic products and marine ecosystems. Based on this, high-standard trade rules have led to stricter and more complex market access requirements [4]. Early studies on regional openness focused on “shallow integration” under the WTO framework, often employing difference-in-differences approaches to treat WTO accession as a quasi-natural experiment analyzing trade-related welfare effects. For instance, Handley and Limão (2017) showed that China’s WTO accession boosted exports and lowered U.S. consumer goods prices [5]. Similarly, Amiti and Konings (2013) found that regional trade liberalization increases firms’ use of intermediate goods and export intensity [6].

With the evolving landscape of international openness, the focus of research has shifted from multilateral liberalization under the WTO framework to regional “deep integration” characterized by high-standard provisions. Many scholars have employed trade gravity models to examine the impact of high-standard environmental provisions on trade creation, trade diversion, and trade substitution effects. The first strand of relevant literature focuses on the economic impacts of such environmental integration. For example, Rocha et al. (2024) found that bilateral environmental provisions generate substitution effects on trade in polluting goods, thereby stimulating trade in cleaner goods [7]. Brandi et al. (2020) indicate that environmental trade provisions may generate a trade substitution effect for developing countries [8]. Environmental provisions also cover a significant number of digital products, Elsig and Klotz (2021) showed that related rules have spillover effects and that bilateral deep integration encourages more states to join FTA networks [9]. Fabrizi et al. (2024) further find that stricter green policies are associated with increased innovation, which in turn drives better export performance [10]. Since deep environmental openness involves not only reducing border tariffs but also opening domestic markets, recent research has examined its internal effects, such as promoting intra-regional technology spillovers, enhancing host-country market transparency, and fostering government procurement cooperation [11,12,13].

The second body of literature pertinent to this study examines the impact of bilateral environmental openness on the restructuring of aquatic products value chains. Chandran and Sudarsan (2012) argue that the India-ASEAN Free Trade Agreement could lead to substantial imports of aquatic products from ASEAN into India, thereby redistributing value chain segments [14]. Due to the severe problem of illegal fishing worldwide, the EU’s red and yellow card system for seafood has indirectly curbed the aquatic products value chains of affected countries [15]. Luo et al. (2017) suggest that China’s Belt and Road Initiative will drive the restructuring of aquatic products value chains between China and ASEAN member states [16]. Erokhin et al. (2021) argue that the Asia-Pacific, where RCEP members have abundant marine resources and Japan and South Korea possess advanced aquaculture technologies, will play a key role in restructuring the global aquatic products value chain [17].

In summary, the contributions of this study are threefold. First, it broadens the research scope. Unlike previous studies that focus solely on the economic effects of regional openness, this paper examines the degree of bilateral aquatic products value chain linkage from an environmental openness perspective and elucidates the underlying mechanisms, thereby contributing to the development of a secure and efficient industrial and supply chain system. Second, in terms of practical significance, existing literature largely addresses aquatic products trade under “shallow integration”. By contrast, this study leverages a deep trade agreement database to analyze value chain restructuring under regional “deep integration”. Third, the study systematically explores the theoretical mechanisms through which deep regional openness affects the aquatic products’ value chain linkage intensity and investigates how heterogeneous provisions enhance bilateral linkage, thus refining the analytical dimensions of existing research.

2. Theoretical Analysis and Research Hypothesis

Bilateral environmental trade provisions function similarly to a compliance pass. For instance, the EU’s red-and-yellow card system targeting illegal fishing of aquatic products significantly curtails trade in seafood exports to Europe. Conversely, once a country completes institutional reforms, market access is restored and enhanced, thereby reducing de facto non-tariff barriers that had risen. Environmental labels formed by high-standard environmental rules also serve to alleviate buyers’ “trust deficit” regarding water resource sustainability and labor compliance. This boosts prices and demand, expands effective markets, thereby lowering entry costs and amplifying economies of scale. These deep environmental trade provisions require regional integration of fisheries markets while strengthening bilateral trust mechanisms through transparent texts. The following sections will explain how environmental trade provisions deepen bilateral fisheries value chain linkages by reducing market barriers, starting with non-tariff barriers and progressing to tariff barriers.

Aquatic products are primary agricultural commodities and are significantly affected by market access restrictions. High-standard environmental regulations are pivotal to advancing regional integration, providing institutional support for economic and regulatory coordination. Provisions such as carbon tariff exemptions, rules of origin, and reduced technical barriers enhance cross-border processing, cold-chain logistics, and distribution, thereby reinforcing the aquatic products value chain linkages. Carbon tariff reductions will generate green production effects, lower import costs, and enhance exporter competitiveness, driving significant export growth and deepening integration within global value chains. Under the CPTPP, Canada will remove most tariffs on aquatic exports, Japan will eliminate 66% of fishery tariffs immediately, Vietnam will liberalize 83% of tariff lines, and Malaysia will achieve full duty-free status. Beyond tariffs, institutional barriers such as sanitary and phytosanitary measures (SPS) under environmental provisions and environmental regulatory assessments also shape trade patterns [18]. For perishable products like aquatic products, stringent SPS standards heighten quality risks and impose hidden costs [19]. Japan’s strict inspection requirements exemplify such barriers, weakening the aquatic products value chain linkages in Northeast Asia. Recent evidence shows that RCEP reduces both tariff and non-tariff barriers, fostering member consultations and mitigating the adverse impacts of SPS measures [20].

Second, rules of origin constitute a critical institutional variable influencing cross-border production and distribution networks [21]. The environmental provisions also establish institutional arrangements for the origin of materials. In the absence of unified international trade regulations, fragmented origin rules hinder the effective division of labor among nations. Regional bilateral trade agreements establish unified arrangements for rules of origin, lowering market access barriers. This allows enterprises to allocate activities such as fishing, primary processing, and deep processing across different countries. For aquatic products enterprises in particular, this means they can locate fishing operations in regions abundant with marine resources while positioning midstream processing in labor-rich countries, thereby forming tightly integrated cross-border industrial synergies.

Finally, Environmental Trade Provisions often imply improved predictability [22]. Formal bilateral or regional agreements establish standardized and specific texts. These texts incorporate transparency rules and dispute resolution mechanisms, reducing policy uncertainty in cross-border environmental governance and thereby enhancing predictability regarding illegal fishing quotas, resource utilization, and environmental compliance. In other words, bilateral trade agreements lower the policy costs of entering foreign markets, enabling aquatic products enterprises to plan more accurately in advance for fishing seasons, catch volumes, processing capacity, and optimal export timing. This stability in expectations helps strengthen cross-border investment and the cohesion of upstream-downstream cooperation, thereby fostering greater coupling and mutual trust within the value chain.

Based on this, the following hypothesis is proposed:

Hypothesis 1:

Environmental Trade Provisions strengthen bilateral aquatic products value chain linkages by lowering bilateral market access barriers.

The integration of aquatic products into global value chains relies not only on product penetration facilitated by reduced market access barriers but also requires global technology spillovers and cross-border knowledge flows. These elements drive effective integration in upstream production processes, thereby enhancing total factor productivity in aquatic products and improving the quality of export products. The flow of technological capital and cross-border knowledge cooperation between host countries and home countries requires certain prerequisites to be realized. Bilateral environmental agreements provide the platform for such technological cooperation. The provisions of these agreements span multiple domains, including technical assistance, intellectual property rights, and standard-setting. They primarily strengthen bilateral technological spillovers through trade channels, investment channels, and institutional channels, thereby reinforcing the aquatic products value chain linkages.

In terms of trade channels, driven by trade liberalization, export expansion positions enterprises more centrally within trade networks. This enhances their understanding of new markets, technologies, and management theories, preventing the “innovation laziness” that can occur in a single domestic market. It generates an “export-induced learning” effect that domestic producers lack [23]. Under trade stimulus, developing-country aquatic products enterprises gain easier access to modern fishing vessels, hatchery technology, and other high-quality intermediate inputs. They can refine domestic fishery management systems by benchmarking against developed-country practices, continuously internalizing, absorbing, improving, and even surpassing these technologies. Such technology spillovers acquired through trade channels enhance the yield and quality of domestic products, strengthen backward linkages in the aquatic products sector by leveraging foreign intermediate inputs, and enable domestic fishery exports to incorporate higher-value-added elements.

Provisions under the traditional WTO framework emphasize trade liberalization, whereas international economic cooperation viewed through an environmental openness lens prioritizes transnational capital flows—particularly green technology-intensive foreign direct investment. An attractive investment environment, strong intellectual property protection, and transparent competition policies are key drivers of technology inflows. Regarding investment channels, Javorcik (2004) shows that many countries have strong incentives to attract foreign capital to expand domestic production capacity in line with development needs [24]. Using Lithuanian data, the study further demonstrates that technology spillovers can foster linkages between upstream industries and local suppliers.

Based on this, the following hypothesis is proposed:

Hypothesis 2:

Environmental Trade Provisions strengthen the aquatic products value chain linkages by enhancing bilateral technology spillover effects.

Li et al. (2010) introduced the concept of deepening cross-cutting division of labor within value chains, positing it as the foundation for bilateral value chain linkages [25]. The original factor endowment theory provided the theoretical basis for the vertical division of labor model in global value chains, where labor-intensive countries concentrated on low-end segments of the value chain while capital-intensive countries focused on high-end segments. Cross-cutting division of labor in value chains refers to the dynamic interweaving of horizontal and vertical divisions among enterprises across nations. This occurs when firms simultaneously engage in vertical specialization (upstream-downstream relationships) and horizontal differentiation (product specialization) within the global value chain system, thereby forming intricate networks of cross-industry and intra-industry linkages.

First, environmental provisions standardize market access thresholds, laying the groundwork for deepening vertical division of labor systems. Moreover, under high-standard environmental provisions like RCEP and CPTPP, intermediate goods such as fish fry, fishing vessels, and fishing gear can circulate freely across multiple countries. The value-added segments, grounded in value-added trade theory, achieve vertical integration through “rough processing”, “refined processing”, and “deep processing”, thereby forming a “stepping-stone” regional vertical division of labor production system. Simultaneously, differing end-stage demands across nations generate horizontal intra-industry trade at the terminal level, such as canned fish and fish-based health supplements. Thus, regional trade agreements significantly advance the development of vertical division of labor systems.

Second, high-standard environmental provisions generate intra-industry technological spillovers that lay the groundwork for deepening horizontal division of labor. Keller (2004) demonstrates that technology diffusion is a key driver of global industrial transformation [26]. These agreements streamline foreign investment procedures and approvals, boosting greenfield investments and acquisitions across borders. Developed countries introduce advanced technologies—such as deep-sea fishing, intensive processing, and cold-chain preservation—to developing nations. As mentioned earlier, such technological spillovers not only enhance the productivity and product quality of local enterprises, enabling them to meet the entry thresholds for developed markets. More critically, the overall improvement in technological capabilities across the entire industrial chain reduces productivity gaps between partners. This shift allows the division of labor to evolve from a single vertical model toward a horizontal one, ultimately leading to an intertwined value chain structure.

Finally, high-standard environmental provisions encompass a series of provisions on investment protection, intellectual property rights, and other matters, which inherently reduce bilateral and even multilateral trade policy uncertainties. Multinational corporations are particularly sensitive to business environments, further incentivizing them to pursue global industrial layouts. They play a central role in strengthening industrial chain clustering worldwide. In host countries, multinational firms can leverage technological standards to foster upstream production clusters and create vertically integrated divisions of labor. At the same time, they capitalize on branding and end-market management to expand intra-industry partnerships, thereby promoting both horizontal and vertical interconnections.

Based on this, the following hypothesis is proposed:

Hypothesis 3:

Environmental Trade Provisions will strengthen linkages within the aquatic products value chain by deepening cross-sectoral division of labor.

3. Methods and Data

3.1. Empirical Model

Following the approach outlined by Kabir et al. (2017) [27], we specify the following high-dimensional fixed-effects econometric model to further test the theoretical hypotheses of this paper:

$$Gvc\_lin{k}_{ijt}={\beta }_0+{\beta }_{1}Dept{h}_{ijt}+{\beta }_2{X}_{ijt}+\left[{\eta }_{it},{\eta }_{jt,}{\eta }_{ij}\right]+{\xi }_{ijt}$$

(1)

As shown in Equation (1), subscripts i and j denote two distinct countries, while t represents time. $GVC\_lin{k}_{ijt}$ is the dependent variable in this study, indicating the degree of bilateral value chain linkage. ${\beta }_0$ represents the constant term. $Dept{h}_{ijt}$ is the core explanatory variable, reflecting bilateral openness. $X_{ijt}$ represents the joint control variable at the bilateral level. $\left[{\eta }_{it},{\eta }_{jt,}{\eta }_{ij}\right]$ denotes the high-dimensional fixed effects introduced to mitigate potential endogeneity issues in Equation (1). The fixed effects model, a common statistical regression method in econometrics, can be understood as a specialized form of control variables. It is important to note that when employing high-dimensional fixed effects in this study, certain control variables may be absorbed due to multicollinearity issues. The benchmark regression results will present outcomes both with and without the inclusion of high-dimensional fixed effects.

3.2. Data

3.2.1. Dependent Variable

Bilateral Value Chain Linkages in Fisherie ($Gvc\_link$): Building upon the methodology of Duval et al. (2016) with enhancements [28], we constructed the following indicators to reflect bilateral value chain linkages:

$$GVC\_lin{k}_{ijt}={\displaystyle \frac{DV{A}_{ijt}+DV{A}_{jit}}{DV{A}_{it}+DV{A}_{jt}}}$$

(2)

The denominator in Equation (2) represents the total value created by country i and country j in year t. In the numerator, $DV{A}_{it}$ denotes the domestic value added absorbed by country j from country i’s exports, with $DV{A}_{jt}$ having a similar economic meaning. The decomposition method for domestic value added follows approaches outlined in (Koopman et al., 2010; Koopman et al., 2014; Wang et al., 2017) [29,30,31], utilizing raw data from the OECD Global Input-Output Tables for the period 1995–2020.

3.2.2. Independent Variables

Environmental Trade Provisions (EP): Regarding measurement methods for environmental trade provisions, Horn et al. (2010) pioneered the use of the HMS approach to construct a cross-regional “high-standard trade provisions” index, primarily composed of 52 core trade provisions [32]. Hofmann et al. (2017) expanded this framework to develop a deep trade provisions index [33]. This study utilizes trade agreement text data to measure regional opening processes, employing a measurement methodology that incorporates both static and dynamic dimensions. The static dimension measures bilateral openness using dummy variables: if an agreement signed between country i and country j at time t includes environmental provisions among the 52 specific clauses, this variable is coded as 1 for year t and all subsequent years; otherwise, it is coded as 0. To more precisely control for other variables, this study defines time point t as the agreement’s effective date rather than the signing date. While static indicators are widely applied, their limitation lies in the inability of dummy variables to capture nuances within specific provisions. Therefore, this study further constructs dynamic indicators to measure regional openness depth. The core explanatory variable in the benchmark regression model is (the logarithm of the sum of specific rules covered under environmental provisions plus 1). To conduct heterogeneity analysis, this study further decomposes the single-provision dimension for regression statistical analysis.

3.2.3. Control Variables

This paper incorporates the following control variables to mitigate endogeneity issues. Bilateral demand is controlled by taking the logarithm of the sum of the two countries’ per capita GDP (LnGDP), sourced from the World Bank. The Index of Economic Freedom (Free_index) serves as another crucial control variable. This study employs the absolute difference between the two countries’ economic freedom indices, log-transformed, to account for disparities in their economic freedom levels, with data provided by the Heritage Foundation. Furthermore, within international trade theory, the concept of iceberg transport costs is frequently referenced, as bilateral trade costs can weaken industrial linkages between countries. Therefore, this paper controls for bilateral geographic distance (LnDis), shared language (Comlang_off), and shared ethnicity (Smctry) to account for trade friction between countries, with data sourced from the CEPII database.

3.2.4. Mechanisms and Grouping Variables

To validate the theoretical mechanisms proposed in this paper, it also incorporates mechanisms and grouping variables such as bilateral fisheries trade costs, technological spillovers, value-added trade network centrality, bilateral political distance, and trade agreement network centrality. The analysis in this paper covers a total of 76 countries. Descriptive statistics for the relevant variables are presented in

Table 1. Descriptive Statistics of Variables.

Definition	Variable	Obs	Mean	Std. Dev.	Min	Max
Independent Variables	Dummy_Environmrnt	74,100	0.341	0.435	0	1
	Total_Environment	74,100	1.008	1.271	0	3.318
Dependent Variable	GVC_link	74,015	0.006	0.027	0	0.833
Control Variables	LnGDP	72,150	10.146	1.035	5.964	12.317
	lnDis	72,150	8.515	0.943	4.088	9.894
	Free index	71,092	2.255	0.817	0	4.05
	Comlang off	72,150	0.079	0.269	0	1
	Smctry	72,150	0.012	0.108	0	1
Mechanisms and Grouping Variables	Cost	12,140	5.709	2.130	1.226	9.995
	Spill	74,100	4.021	1.303	1.115	10.599
	DW	74,100	0.120	0.081	0.013	0.363
	IdealDis	69,185	1.213	0.348	0.322	5.399
	Center	74,100	14.930	8.829	0	83

Note: The variation in observed values across different datasets stems from the presence of missing data.

.

4. Results

The benchmark regression results are shown in

Table 2. Benchmark Regression Results.

	(1)	(2)	(3)	(4)	(5)	(6)
	GVC_Link
Dummy_Environment	0.0074 ***		0.0013 ***		0.0017 ***
	(18.25)		(5.46)		(5.80)
Total_Environment		0.0032 ***		0.00028 **		0.00045 ***
		(16.94)		(2.37)		(5.45)
LnGDP			0.003 ***	0.0029 ***
			(23.62)	(22.56)
Free_index			−0.0006 ***	−0.0003 ***
			(−4.95)	(−4.96)
LnDis			−0.0059 ***	−0.0054 ***
			(−45.64)	(−44.26)
Comlang_off			0.0046 ***	0.0046 ***
			(12.16)	(12.39)
Smctry			0.0078 ***	0.0079 ***
			(7.91)	(7.98)
N	74,100	74,100	67,304	67,304	74,100	74,100
R-squared	0.0122	0.0177	0.0609	0.0607	0.708	0.708
			Fixed effects
Exporter#Year	NO	NO	NO	NO	YES	YES
Importer#Year	NO	NO	NO	NO	YES	YES
Exporter#Importer	NO	NO	NO	NO	YES	YES

Notes: * p < 0.1, ** p < 0.05, *** p < 0.01. Values in parentheses are t-statistics. The same convention applies to the tables below.

. Columns 1 and 2 include no control variables or fixed effects. Whether using dummy variables or the number of deep provisions in environmental agreements as the core explanatory variable, the findings indicate that deep environmental integration significantly promotes bilateral aquatic products value chain linkages. Columns 3 and 4 present results after incorporating relevant control variables. The core explanatory variables remain significantly positive, indicating that the regression results are not confounded by control variables. The significantly positive coefficient for LnGDP suggests that bilateral economic scale stimulates fishery consumption, thereby deepening forward linkages in the fishery value chain—consistent with economic expectations. The significantly negative coefficient for Free_index indicates that greater disparities in economic policies and political systems between countries make it less likely to establish upstream-downstream linkages in aquatic products. The significantly negative coefficient for LnDis aligns with spatial economics and international trade theory, suggesting that bilateral geographic distance increases trade costs, hindering horizontal or vertical input-output linkages between countries. The positive coefficients for Comlang_off and Smctry indicate that linguistic and cultural similarities between countries also promote extensive linkages within the aquatic products value chain. Columns 5 and 6 present results after incorporating high-dimensional fixed effects. Following the inclusion of fixed effects, all control variables were absorbed, as the fixed effects contained richer information to mitigate endogeneity issues. However, the core explanatory variables remained highly robust. Notably, the R-squared value increases from approximately 0.06 without fixed effects to around 0.7. This indirectly demonstrates that the high-dimensional fixed effects model employed in this study provides a robust explanation for bilateral linkages within the fishery value chain, encompassing explanatory variables at both the individual level and the bilateral common level.

5. Robust Check

5.1. Parallel Trend Test

If signing relevant deep trade provisions is viewed as a quasi-natural experiment, then the prerequisite for using this method is that there is no pre-existing trend in the degree of linkage within the aquatic products value chain prior to the bilateral signing of trade agreements. Event study methodology provides the theoretical foundation for conducting such tests. First proposed by Jacobson et al. (1993) [34], this approach has been widely applied to validate the effectiveness of policy evaluations (Chen et al., 2025; Fox and Swearingen, 2021) [35,36]. Moreover, it can reveal dynamic effects in causal relationships between variables. Therefore, this paper establishes the following model for testing:

$$Gvc\_lin{k}_{ijt}={\beta }_0+{\beta }_k{\displaystyle \sum _{-5}^{5}Dept{h}_{ijt}}+{\beta }_2{X}_{ijt}+\left[{\eta }_{it},{\eta }_{jt,}{\eta }_{ij}\right]+{\xi }_{ijt}$$

(3)

The difference between Equations (1) and (3) lies in the introduction of relative time dummy variables, which construct dummy variables for the lead time preceding the agreement’s implementation and the subsequent period following its implementation. Using the year of agreement entry into force as the dividing point, assign a value of 0 to the trade clause intensity index prior to the agreement’s implementation. If the coefficients are close to 0 with minimal fluctuation prior to the agreement’s entry into force, this indicates that no pre-existing bilateral trends existed before the regional trade agreement was signed. Consequently, endogeneity issues arising from sample self-selection can be ruled out.

The relevant results are presented visually in Figure 1. The solid black line represents the regression coefficient, while the dashed black line indicates the 90% confidence interval. Prior to policy implementation, the coefficient fluctuated around zero and remained within the confidence interval, indicating insignificance. Following policy implementation, the coefficient became significantly positive and continued to increase starting from the third period. This may also suggest that the economic effects of bilateral environmental trade agreements exhibit a certain degree of lag.

5.2. Placebo Test

In our empirical study, regional integration is treated as an exogenous trade policy in a high-dimensional fixed-effects regression. The core assumption of exogenous policy rests on the parallel trends hypothesis. However, parallel trends tests only demonstrate the absence of self-selection effects in aquatic products value chains prior to deep regional opening. During the same period, other aquatic products policies may have interfered with the study’s conclusions—such as fishing quota systems, fuel policies, and domestic aquatic products subsidy regimes—all of which can influence fisheries value chain spillovers. Therefore, to eliminate interference from related factors, this study conducted a placebo test to further enhance the credibility of the findings. The basic principle involves randomly shuffling the original bilateral groups, randomly selecting hypothetical experimental and control groups, and then performing statistical analysis using the same high-dimensional fixed effects as the baseline regression. This process was repeated 500 times, yielding 500 sets of regression coefficients and p-values. Their distribution is shown in Figure 2, where the mean of regression coefficients forms a normal distribution around zero, indicating that the coefficients from the randomized groups lack economic significance. Furthermore, the true regression coefficients lie far outside the extreme distribution range of the “spurious coefficients,” confirming that the genuine regression results are not coincidental. Finally, observing the p-values reveals that the vast majority exceed 0.1, lacking statistical significance. In summary, the placebo test results indicate that the benchmark regression findings in this paper can, to a certain extent, exclude interference from other factors.

5.3. Lagged Core Explanatory Variables

This paper conducts regression analyses after lagging the core explanatory variable by three and five periods. The rationale is threefold. First, economic and trade policies naturally involve implementation lags. Regional trade agreements (RTAs) covering fisheries quarantine inspections, industrial specialization, and aquaculture investment planning require preparation and adaptation periods for member states. In addition, fisheries-related agreements often grant developing economies grace periods of more than two years. Lagging the explanatory variable therefore enables a more comprehensive assessment of the policy effects. Second, lagging mitigates endogeneity concerns arising from reverse causality—namely, current levels of fisheries value chain integration cannot influence past bilateral economic and trade policies. Third, lagging facilitates examination of whether bilateral policies generate long-term effects. If significant, this would indicate that the impact of bilateral RTAs on fisheries value chains is enduring rather than transient. The estimation results are presented in

Table 3. Lagged Core Explanatory Variables.

	(1)	(2)	(3)	(4)
	GVC_Link
L3.Total_Environment	0.0007 ***
	(4.58)
L5.Total_Environment		0.0008 ***
		(5.13)
L3.Dummy_Environment			0.0025 ***
			(4.96)
L5.Dummy_Environment				0.0026 ***
				(5.12)
N	32,800	29,827	32,800	29,827
R-squared	0.80	0.84	0.80	0.84
		Fixed effects
Exporter#Year	YES	YES	YES	YES
Importer#Year	YES	YES	YES	YES
Exporter#Importer	YES	YES	YES	YES

. Columns (1) and (2) use dynamic indicators of trade agreement depth and breadth as the core explanatory variables, whereas Columns (3) and (4) employ a dummy variable indicating whether a trade agreement was signed. Across all specifications, the coefficients are significantly positive, with no meaningful difference between the three-period and five-period lag estimates. This provides indirect evidence that bilateral deep integration exerts a sustained and positive effect on the development of the fisheries value chain.

6. Mechanism Analysis

Theoretical analysis suggests that bilateral environmental trade agreements deepen bilateral fisheries value chain linkages by lowering market access barriers, increasing technological spillovers, and promoting deeper cross-sectoral division of labor within value chains. In this subsection, this paper empirically tests the above three theoretical hypotheses.

6.1. Lowering Market Access Barriers

In international trade theory, the bilateral iceberg trade costs for fisheries represent the most direct measure of market access barriers. Typically, bilateral trade costs encompass explicit expenses such as tariffs on aquatic products and fisheries transportation insurance. Given this characteristic, some literature employs the product catalogs exempted from tariffs under various agreements to gauge the difficulty of specific products entering a country (Caliendo and Parro, 2015) [37]. However, deeper regional openness extends beyond tariff reductions to encompass institutional trade barriers arising from implicit procedures such as customs clearance approvals and quarantine inspections for aquatic products. Therefore, following the methodology of Tombe and Zhu (2019) [38], this paper employs the trade flow approach to measure bilateral fisheries trade costs, as detailed in Equation (4):

$${\overline{\tau }}_{ni}^j\equiv \sqrt{{\tau }_{ni}^j{\tau }_{in}^j}={\left({\displaystyle \frac{{\pi }_{nn}^j{\pi }_{ii}^j}{{\pi }_{ni}^j{\pi }_{in}^j}}\right)}^{1/2\theta }$$

(4)

Equation (4) is derived from the gravity model. The superscript j denotes fisheries, while subscripts n and i represent countries. ${\overline{\tau }}_{ni}^j$ represents the symmetric bilateral trade cost for aquatic products, measuring the entry cost for fisheries in bilateral markets. ${\pi }_{nn}^j$ denotes the proportion of aquatic products imported by country n from itself relative to its total aquatic product imports, while ${\pi }_{ni}^j$ denotes the proportion of aquatic products imported by country n from country i relative to its total aquatic product imports. Parameter $\theta$ represents the trade elasticity, set to 4. After calculating fishery trade costs using OECD input-output tables, the relevant mechanism tests were conducted following the methodology of Zeng et al. (2025) [39]. The specific results are shown in

Table 4. Lowering Market Access Barriers.

	(1)	(2)	(3)	(4)
	Cost		GVC_Link
Dummy_Environment	−0.3099 ***		0.0254 **
	(−2.84)		(2.31)
Total_Environment		−0.1016 ***		0.0042 **
		(−3.08)		(2.32)
Dummy_Depth×Cost			−0.0060 ***
			(−4.53)
Total_Depth×Cost				−0.0021 **
				(−2.36)
Cost			−0.0096 ***	−0.0083 ***
			(−8.36)	(−8.31)
N	10,583	10,583	10,583	10,583
R-squared	0.94	0.94	0.87	0.87
		Fixed effects
Exporter#Year	YES	YES	YES	YES
Importer#Year	YES	YES	YES	YES
Exporter#Importer	YES	YES	YES	YES

. Column 1 indicates that bilateral environmental openness significantly reduces bilateral fishery trade costs, decreasing losses and transaction frictions. The interaction coefficients between the core explanatory variable and the mechanism variable in Columns 3 and 4 are significantly negative, indicating that bilateral environmental openness enhances bilateral fisheries value chain linkages by reducing bilateral fisheries trade costs. On one hand, simplified and transparent procedures for sanitary inspections and subsidies enable cross-border cooperation in upstream capture and aquaculture products, allowing optimal allocation of high-quality marine fishery resources within the region. On the other hand, the division of labor in downstream sectors—including cold chain transportation, deep processing, and cross-border e-commerce sales—tends to stabilize, strengthening the integrity and complementarity of the fisheries value chain. Therefore, based on the above, Hypothesis 1 holds.

6.2. Increasing Technological Spillovers

According to De Loecker’s (2007) “export-induced learning” theory [23], technology is primarily embedded within products to achieve broad spillover effects. Therefore, this paper constructs the following bilateral fishery technology spillover-related indicators, following the methodology of Hausmann et al. (2007) [40]:

$$Spil{l}_{mt}=\sum _i^N\left({\displaystyle \frac{EX{P}_{imt}/{\displaystyle \sum _m^M}EX{P}_{imt}}{{\displaystyle \sum _i^N}EX{P}_{imt}/{\displaystyle \sum _i^N}{\displaystyle \sum _m^M}EXP}}\times pcGD{P}_{it}\right)$$

(5)

$$Spil{l}_{ijmt}={\displaystyle \frac{EX{P}_{mt}^{ij}+EX{P}_{mt}^{ij}}{{\displaystyle \sum _m^M}\left(EX{P}_{mt}^{ij}+EX{P}_{mt}^{ji}\right)}}\times {Spill}_{mt}$$

(6)

In Equations (5) and (6), M denotes the product, N represents the country, EXP signifies export value, pcGDP indicates per capita GDP for each country, and Spill represents the bilateral fishery technology spillover indicator. Similarly, the specific results are shown in

Table 5. Technological Spillovers.

	(1)	(2)	(3)	(4)
	Spill		GVC_Link
Dummy_Environment	0.1805 ***		−0.0019 **
	(2.61)		(−2.14)
Total_Environment		0.0231 **		−0.0008 **
		(2.44)		(−2.33)
Dummy_Environment×pill			0.0010 ***
			(6.52)
Total_Environment×Spill				0.0003 ***
				(4.75)
Spill			0.0007 ***	0.0006 ***
			(8.71)	(8.68)
N	74,100	74,100	74,100	74,100
R-squared	0.77	0.77	0.75	0.75
		Fixed effects
Exporter#Year	YES	YES	YES	YES
Importer#Year	YES	YES	YES	YES
Exporter#Importer	YES	YES	YES	YES

. Column 1 indicates that bilateral deep regional opening significantly enhances bilateral fishery technology spillovers, potentially facilitating the sharing of energy-saving and emission-reducing fishing technologies, disease prevention and control traceability measures, and improvements in processing techniques through imitation of mid-to-downstream products. The interaction coefficients between the core explanatory variable and the mechanism variable in Columns 3 and 4 are significantly negative, indicating that bilateral environmental provisions enhance bilateral fisheries value chain linkages through increased bilateral technology spillovers. This technology diffusion generates synergies in production and sales: it improves upstream production efficiency and export-quality aquatic products while positively influencing the responsiveness of downstream market sales, collectively deepening bilateral fisheries value chain linkages. Therefore, Hypothesis 2 holds.

6.3. Strengthen Cross-Sectoral Collaboration Within the Fisheries Value Chain

In our theoretical analysis, this paper posits that bilateral environmental trade agreements promote fisheries value chain linkages by reinforcing both horizontal and vertical divisions of labor. This argument is subsequently tested from the perspective of value-added trade networks. The closer the centrality of one country’s value-added trade network is to that of its partner, the greater the similarity in their overall industrial structures within the fisheries supply chain, thereby increasing the likelihood of horizontal division of labor. By contrast, larger disparities in centrality reflect more divergent industrial structures, favoring vertical division of labor—for instance, one country specializing in capture fisheries while the other focuses on sales. Following the methodology of Amador and Cabral (2017) [41], this study first constructs a binary bilateral value-added trade matrix using a 0.5% threshold. Subsequently, the value-added trade network centrality of each country is calculated based on this matrix. Finally, two groups are formed: one with lower absolute centrality differences between countries (Low_DW) and another with higher absolute differences (High-DW). These groups are then used as core explanatory variables in regression analyses with the dependent variable. The specific results are shown in

Table 6. Strengthen cross-sectoral collaboration within the fisheries value chain.

	(1)	(2)	(3)	(4)
	Low-DW		High_DW
Dummy_Environment	−0.0027 *		0.0091 ***
	(−1.71)		(3.59)
Total_Environment		−0.0029 **		0.0021 **
		(−2.08)		(2.13)
N	36,368	36,368	37,726	37,726
R-squared	0.73	0.73	0.73	0.73
		Fixed effects
Exporter#Year	YES	YES	YES	YES
Importer#Year	YES	YES	YES	YES
Exporter#Importer	YES	YES	YES	YES

. The results in Columns 1 and 2 are significantly negative, indicating that for the two groups with low similarity in fishery industrial structure, bilateral deepening of openness further promotes the vertical division of labor model. For instance, reducing barriers like quarantine inspections enables resource-rich countries to focus on upstream fishing and aquaculture, while labor-abundant countries concentrate on mid-to-downstream processing and market development. With trade barriers lowered, this vertical division becomes tighter, enhancing bilateral fisheries value chain linkages. The results in Columns 3 and 4 are significantly positive, indicating that for groups with high similarity in fishery industrial structure, deeper bilateral openness promotes the formation of horizontal division of labor. This occurs because fishery technology spillovers generate overlapping demands between partners, facilitating complementary fishery value chain systems across countries in end-consumption and processing segments. Therefore, the above conclusions demonstrate that the intersecting systems of vertical and horizontal fisheries division of labor promote bilateral fisheries value chain linkages, validating Hypothesis 3 of this paper.

7. Heterogeneity Analysis

7.1. Analysis of Heterogeneity in Agreement Clauses

The results of benchmark regression and robustness tests indicate that bilateral depth environmental provisions significantly enhance the linkage of bilateral fisheries value chains. However, depth environmental provisions encompass multiple specific details with varying characteristics. Therefore, following the methodology of Monteiro (2016) [42], this study decomposes the core explanatory variable into provisions related to domestic law (Domestic_Environment), exception and safeguard provisions (ES_Environment), multilateral environmental provisions (Mul_Environment), environmental capacity-building provisions (Capacity_Environment), environmental governance provisions (Law_Environment), and emerging theme provisions (Emerging_Environment) with the dependent variable. The specific results are shown in

Table 7. Heterogeneity in Agreement Clauses.

	(1)	(2)	(3)	(4)	(5)	(6)
	GVC_Link
Domestic_Environment	0.0010 ***
	(3.28)
Law_Environment		0.0093 **
		(2.44)
Mul_Environment			0.0003 *
			(1.83)
Emerging_Environment				0.0007 *
				(1.90)
ES_Environment					0.0006 ***
					(3.98)
Capacity_Environment						0.0006 *
						(1.74)
N	74,100	74,100	74,100	74,100	74,100	74,100
R-squared	0.73	0.73	0.73	0.73	0.73	0.73
Fixed effects
Exporter#Year	YES	YES	YES	YES	YES	YES
Importer#Year	YES	YES	YES	YES	YES	YES
Exporter#Importer	YES	YES	YES	YES	YES	YES

. The results in columns 1–6 indicate that environmental governance provisions and domestic environment-related provisions have a greater promotional effect on fisheries value chain linkages. Regarding the content of bilateral openness, provisions concerning domestic laws and regulations undoubtedly exhibit a deeper level of openness compared to other areas, often focusing on implicit costs within upstream-downstream industrial linkages. From an economic perspective, negotiations on superficial openness—such as tariffs and goods trade liberalization—have largely reached consensus among most member states, leading to diminishing marginal effects across member countries. Given the evolving landscape of global economic development, the characteristics of transnational corporations have grown increasingly complex, deeply intertwining host country and home country industries. From an implementation perspective, governance measures targeting specific segments within deep opening are more effective at integrating various links across the value chain. Collectively, this evidence demonstrates that deep opening is more conducive to fostering bilateral fisheries value chain linkages than shallow opening.

7.2. Heterogeneity in Political Distance Among Member States

The provisions on deep environmental commitments cover numerous sensitive areas, including subsidies, labor standards, government procurement, and intellectual property. This necessitates shared political philosophies between the two countries to ensure more effective implementation at the operational level. On one hand, closer political alignment facilitates consensus on fisheries technical regulations, catch quotas, and cold chain transportation methods, thereby amplifying the integration effects across the fisheries value chain. On the other hand, closer political alignment facilitates effective information sharing and communication regarding marine resource conservation—such as endangered marine species—which fosters seamless integration between upstream and downstream value chains. If end consumers demand higher environmental sustainability standards, but this information is not shared with upstream producers, the productivity gains from the fishing industry cannot translate into meaningful economic value. In summary, the degree of political distance implies implicit implementation costs for bilateral trade agreements. A relatively effective metric is needed to reflect the institutional and regulatory differences arising from political divergence between countries. Therefore, this paper adopts Bailey et al.’s (2017) methodology [43], using the UN General Assembly voting database to measure political distance between countries. The underlying economic logic is that shared positions on resolutions indicate greater similarity in political systems and economic trade policies, and vice versa. As shown in

Table 8. Heterogeneity in Political Distance Among Member States.

	(1)	(2)	(3)	(4)
	Low-IdealDis		High_IdealDis
Dummy_Environment	0.0043 ***		0.0010 *
	(6.18)		(1.69)
Total_Environment		0.0010 ***		0.0003 **
		(9.55)		(1.99)
N	34,806	34,806	34,379	34,379
R-squared	0.82	0.82	0.81	0.81
		Fixed effects
Exporter#Year	YES	YES	YES	YES
Importer#Year	YES	YES	YES	YES
Exporter#Importer	YES	YES	YES	YES

, regression results for the low bilateral political distance group exceed those for the high group (0.0043 > 0.0010, 0.001 > 0.0003). In other words, the positive effect of bilateral deep environmental provisions on fisheries value chain linkages is negatively moderated by bilateral political distance. The signing of bilateral trade agreements significantly reduces institutional barriers in the fisheries value chain linkages, as demonstrated in the baseline regression. However, the results in Table 8 further reinforce a key finding: the full promotional effect of bilateral deep opening on deepening fisheries industrial chain cooperation is only realized when the political distance between the two countries is low. Therefore, the process of deep bilateral opening requires continuous communication and refinement of details to reduce subsequent implementation costs and maximize economic benefits.

7.3. Center–Periphery Heterogeneity

Regional openness between bilateral and multilateral agreements exhibits certain network effects [44]. A country that signs more trade agreements occupies a central position within regional openness, tending to establish closer industrial cooperative relationships with other nations. Conversely, a country with fewer trade agreements occupies a peripheral position in the regional openness landscape, incurring higher costs for industrial linkages with other nations. Countries positioned at the center of regional openness networks possess extensive experience in implementing agreements and more open economic and trade policies, providing long-term institutional safeguards for transnational fisheries industry cooperation between them. However, trade agreements between regionally open central countries and periphery countries may weaken the actual effectiveness of fisheries value chain integration due to differences in specific implementation levels. Therefore, this paper categorizes countries into central-central and central-periphery groups based on the median absolute value of the difference in environmental trade agreements signed by the two countries in a given year. This aims to examine whether differences in agreement networks affect the promotion of bilateral deep opening on fisheries value chains. As shown in

Table 9. Center–Periphery Heterogeneity.

	(1)	(2)	(3)	(4)
	Center–Center		Center–Outside
Dummy_Environment	0.0023 ***		0.0013 *
	(3.84)		(1.65)
Total_Environment		0.0002 ***		0.0001
		(3.82)		(0.73)
Constant	0.0059 ***	0.0060 ***	0.0034 ***	0.0034 ***
	(17.49)	(19.70)	(34.36)	(46.66)
N	43,967	43,967	29,417	29,417
R-squared	0.77	0.77	0.75	0.75
		Fixed effects
Exporter#Year	YES	YES	YES	YES
Importer#Year	YES	YES	YES	YES
Exporter#Importer	YES	YES	YES	YES

, within the center–center group, deep interregional openness significantly promotes fisheries value chains. However, in the center–periphery group, although the economic effect is positive, it fails to pass statistical tests. This implies that the closer the regional openness network positions of cooperating parties, the more effectively the spillover effects of “interconnectivity” can be leveraged, facilitating seamless coordination among member states in fisheries management policies, fishing technologies, and fishery resource development. Conversely, when the gap in regional openness network positions between parties is significant, the positive promotion of fisheries value chain linkages may be weakened due to the presence of communication, coordination, and implementation costs.

8. Conclusions and Discussion

This study finds that “deep environmental provisions” between regions significantly enhance the linkage of bilateral fisheries value chains, consistent with existing research findings on the positive spillover effects of environmental trade provisions [45,46]. First, environmental provisions strengthen bilateral value chain integration by lowering market access barriers for aquatic products. This occurs through requirements for member states to improve fisheries resource management efficiency, combat illegal fishing, and strengthen traceability systems. Second, unlike “shallow integration,” the deep openness represented by environmental provisions encompasses multiple domestic liberalization measures such as government procurement, market access, and foreign investment protection. This further promotes mutual technological spillovers among fisheries products, achieving forward linkages within the value chain. Finally, deep opening effectively promotes cross-division of labor across national fisheries value chains, thereby strengthening bilateral fisheries’ value chain linkages. However, the heterogeneity observed in this study indicates that deep regional provisions exert significantly heterogeneous effects on the evolution of bilateral fisheries value chains. For instance, the stronger the correlation between such environmental provisions and domestic reforms, the more pronounced their promotional effect on bilateral fisheries value chain linkages. Simultaneously, environmental trade provisions are not universally beneficial. When bilateral political distance is high, favorable value chain linkages are less likely to form—for instance, trade agreements between developing and developed countries may exhibit weaker implementation outcomes. The centrality of bilateral trade agreements also significantly influences their effectiveness. Countries occupying central–central network positions, being closer than central–peripheral nations, tend to stimulate greater mutual spillovers within aquatic product value chains.

The findings of this study offer significant insights for nations seeking to deepen environmental openness and achieve proactive water resource utilization. For example, developing countries should further clarify the depth of environmental provisions in the text and strengthen the agreement’s central role within their networks. This includes, but is not limited to, promoting clean energy and the circular economy, advancing international green technology sharing, and improving water resource regulation. Despite the extensive empirical research conducted in this paper, certain limitations remain. Primarily due to data constraints, the study did not differentiate between processing and fishing segments within the aquatic products value chain. Consequently, further refinement is needed in the construction of national water resource policy management systems.