A MULTI–DIMENSIONAL FRAMEWORK FOR ASSESSING FTA RULES OF ORIGIN UTILIZATION: THE CASE OF THE VIETNAM–EU FREE TRADE AGREEMENT

ABSTRACT

Objective: to develop a multi-dimensional analytical model for diagnosing the under-utilization of preferential Rules of Origin (RoO) in Free Trade Agreements (FTAs) and apply it to the Vietnam-EU FTA (EVFTA). Methods: a comparative case study analysis of five Vietnamese export sectors (seafood, footwear, wood/furniture, garments/textiles, electronics) using a proposed four-dimensional framework encompassing Legal/Institutional (D1), Supply Chain/Structural (D2), Administrative/Procedural (D3), and Enterprise Capacity (D4) factors. Results: The interaction between D1 and D2 creates a "structural ceiling" for the Preference Utilization Rate (PUR). Sectors with high D1×D2 compatibility (seafood, footwear) achieve PUR of 84–99%, while the garments sector, characterized by fundamental incompatibility, remains capped at 28–32% despite four years of intensive D3/D4 improvements. Conclusions: The framework effectively explains cross-sectoral PUR variation. Overcoming the structural ceiling in constrained sectors requires long-term upstream industrial investment (e.g., weaving, dyeing), not just administrative fixes. The paper also argues for excluding sectors with zero preference margins (electronics) from PUR denominators to avoid measurement bias.

АННОТАЦИЯ

Цель исследования — разработать многомерную аналитическую модель для диагностики причин недостаточного использования преференциальных правил происхождения (RoO) в соглашениях о свободной торговле (ССТ) и применить её к ССТ между Вьетнамом и ЕС (EVFTA). Методология основана на сравнительном анализе пяти экспортных секторов Вьетнама (рыба, обувь, дерево/мебель, текстиль/одежда, электроника) с использованием предложенной четырехмерной модели, включающей правовые/институциональные (D1), структурные (D2), административные (D3) и фирменные (D4) факторы. Результаты демонстрируют, что взаимодействие D1 и D2 создает «структурный потолок» использования преференций (PUR). В секторах с высокой совместимостью (рыба, обувь) PUR достигает 84–99%, тогда как в текстильном секторе с несовместимостью D1×D2 PUR остается на уровне 28–32%, несмотря на все усилия по улучшению D3 и D4. Выводы. Предложенная модель эффективно объясняет межотраслевую вариацию PUR. Для преодоления «структурного потолка» в проблемных секторах необходимы не административные реформы, а долгосрочные инвестиции в развитие смежных производств (ткачество, красильное дело). Также предлагается исключать сектора с нулевой преференциальной маржой (электроника) из знаменателя PUR для корректировки измерений.

Keywords: rules of origin; FTA preference utilization; EVFTA; Vietnam; supply chain structure; structural ceiling; compliance costs

Ключевые слова: правила происхождения; использование преференций в рамках соглашения о свободной торговле; EVFTA; Вьетнам; структура цепочки поставок; структурный потолок; затраты на соблюдение требований.

1. Introduction

The proliferation of preferential trade agreements has made rules of origin (RoO) one of the most consequential non-tariff instruments in contemporary trade policy. Yet, a persistent paradox characterizes many developing economies: despite significant tariff preferences embedded in FTAs, firms systematically fail to claim them. The Vietnam-EU Free Trade Agreement (EVFTA, effective 1 August 2020) illustrates this paradox sharply. With nearly 99% of tariff lines eligible for elimination, EVFTA offers transformative market access. In practice, the overall preference utilization rate (PUR) rose from 14.8% in 2020 to 35.2% in 2023 before plateauing at approximately 35–36% in 2024 — meaning roughly two-thirds of eligible exports still forego EVFTA benefits [1, 2]. By comparison, EU FTAs globally achieve an average PUR of approximately 67% once mature [3], and Vietnam's own GSP utilization rate before EVFTA stood at approximately 60% [4].

The academic literature on FTA under-utilization is rich but fragmented, focusing on single-dimensional lenses: administrative compliance costs [5, 6, 7], institutional opacity [8, 9], supply chain structure [10, 11], and firm-level capacity [12, 13]. Each stream illuminates a real constraint; none provides an integrated framework for diagnosing which constraint is binding for a given sector, or how multiple constraints interact to generate observed PUR outcomes.

This paper addresses that gap. The central research question is: what combination of dimensions best explains cross-sector variation in EVFTA preference utilization, and what does this imply for policy sequencing? The paper makes three contributions: (i) a theoretically grounded multi-dimensional framework with explicit interaction logic; (ii) a cross-sector comparative diagnostic revealing the structural ceiling mechanism; and (iii) a methodological argument for correcting PUR measurement in sectors with zero preference incentive.

2. Materials and Methods

The paper employs a comparative multiple case study design [14] across five EVFTA export sectors. Qualitative comparative analysis is appropriate because the objective is diagnostic — identifying which dimensional combination is binding — rather than estimating effect sizes. The five sectors (garments/textiles, footwear, seafood, wood/furniture, electronics) were selected through theoretical sampling [15]: each represents a distinct predicted outcome in a four-group taxonomy derived from the framework, together covering all EVFTA RoO criterion types (Wholly Obtained-WO, Regional Value Content-RVC, Change in Tariff Heading-CTH, fabric-forward double transformation). They collectively account for approximately 68% of Vietnam's merchandise exports to the EU in 2020–2024.

This paper proposes that RoO utilization efficiency is jointly determined by four dimensions arranged in a two–tier causal structure. The foundational tier (D1 and D2) determines structural feasibility; the operational tier (D3 and D4) determines how close to the structural ceiling actual utilization reaches.

This paper proposes that RoO utilization efficiency is jointly determined by four dimensions arranged in a two–tier causal structure. The foundational tier (D1 and D2) determines structural feasibility; the operational tier (D3 and D4) determines how close to the structural ceiling actual utilization reaches.

Figure 1. Analytical framework for evaluating the effectiveness of EVFTA rules of origin utilization

D1 (Legal/Institutional framwork) captures the domestic framework implementing EVFTA origin obligations: the clarity and consistency of Circular 11/2020/TT–BCT (Vietnam's transposing regulation), C/O issuance system efficiency, and availability of self–certification (REX). D2 (Supply chain structure) captures macro–level compatibility between a sector's input–sourcing geography and applicable RoO – operationalising Cadot et al.'s (2006) structural barrier concept at the sector level. D3 (Administrative procedures) captures practical C/O certification efficiency, costs, and SME accessibility. D4 (Firm capability) captures firm–level RoO knowledge, compliance expertise, financial resources, and crucially – control over input sourcing decisions, which the CMT production model forecloses.

The framework's core theoretical proposition: when D1 and D2 are compatible, D3 and D4 are the binding constraints and can be addressed through administrative reform and capacity building. When D1 and D2 are incompatible, they impose a structural ceiling on PUR that D3/D4 improvements can approach but cannot exceed. This generates a testable four-group taxonomy: Group 1 (WO criterion + domestic production → PUR ceiling near 100%); Group 2 (RVC/CTH criterion compatible with existing supply chain → ceiling 65–99%, D3/D4 determine actual PUR); Group 3 (complex criterion incompatible with input-sourcing geography → structural ceiling at low PUR); Group 4 (MFN already zero → no economic incentive to utilize).

Table 1.

 The four–dimensional framework: definitions, and indicators

Source: Authors, synthesized from [10, 6, 13, 7, 8].

Primary data sources include: Ministry of Industry and Trade C/O issuance statistics (2020–2024); EuroCham Business Climate Index surveys (Q2 2024, Q2 2025) [16]; CIEM five-year EVFTA assessment (2025) [2]; and sector association reports [17, 18, 19, 20]. Secondary sources include Eurostat, GSO manufacturing surveys [21], and peer-reviewed literature. Sector-level PUR estimates are triangulated from these sources.

3. Results and Discussion

3.1 Cross-Sector Diagnostic

Table 2 presents the cross-sector diagnostic. The PUR variation is striking and structurally patterned: sectors with natural D1×D2 compatibility (seafood, footwear) achieve near-optimal PUR, while the sector with fundamental incompatibility (garments) is held below 32% despite four years of intensive D3/D4 interventions (340–350 training workshops annually; full eCoSys digitalization) [16]. The electronics case is analytically distinct: low PUR is not a compliance failure but a rational response to a zero preference margin.

Table 2.

Cross–Sector Four–Dimensional Diagnostics (EVFTA, 2020–2024)

Sources: EVFTA Protocol 1; [1, 2, 16, 17, 18, 19, 20, 21, 22].

3.2 Analysis of Key Sectors

Seafood (Group 1). The Wholly Obtained criterion (D1) combined with 100% domestic sourcing (D2) generates a near-100% structural ceiling. The observed 84–90% PUR is attributed to D3 friction from the EU's "yellow card" for IUU fishing, which adds documentation costs for open-sea catch [23]. This is an addressable administrative constraint, not a structural one.

Footwear (Group 2a). The RVC ≥40% criterion (D1) is easily met with domestic value-added of 45–50% (D2). The FOB production model gives firms full input control (D4). With EU MFN tariffs of 8–17%, the preference margin is large enough to justify compliance costs [6], resulting in near-optimal PUR of 97–99%.

Wood/Furniture (Group 2b). This sector shows a D4 split. Large firms with integrated processing chains achieve 65–75% PUR. However, ~40% of output comes from village workshops lacking chain-of-custody (CoC) systems [20], effectively excluding them from preferences.

Garments/Textiles (Group 3). This case critically tests the structural ceiling proposition. The fabric-forward rule (D1) requires yarn and fabric to be sourced from Vietnam, the EU, or Korea. However, D2 is fundamentally incompatible: in 2024, 67% of fabric was imported from China (ineligible), with domestic fabric only meeting ~18–20% of needs [17, 22]. Even with perfect D3 and D4, the structural ceiling is arithmetically capped at 28–32%. The CMT model (65–70% of output) exacerbates this, as buyers supply the fabric [17], leaving Vietnamese firms with no control over compliance (D4 symptom with D2 root).

Electronics (Group 4). While D2 is structurally weak, the determinative factor is the zero preference margin under the WTO ITA [24]. Low PUR is a rational economic decision, not a failure. This case highlights a critical measurement issue.

3.3. Discussion

The Structural Ceiling Mechanism. The cross-sector evidence strongly supports the proposition that D1×D2 incompatibility imposes an upper bound on PUR. Between 2022 and 2024, extensive D3/D4 improvements in the garments sector—digitalization, training, faster C/O processing—failed to raise the PUR ceiling by more than 2–3 percentage points. This falsifies single-dimensional predictions and confirms that structural barriers are impervious to administrative fixes alone.

Comparison with Existing Frameworks and Measurement Correction. The proposed framework subsumes single-dimensional approaches. The cost-benefit paradigm (D3) is binding in Group 2, but not in Group 3. The institutional opacity stream (D1) explains cross-country, not cross-sector, variation. The supply chain stream (D2) explains the ceiling, while the firm-capacity stream (D4) explains within-group variation.

The electronics case reveals a systematic measurement bias. Standard PUR includes all eligible trade, but eligibility is distinct from incentive. Including Group 4 sectors with zero preference margins deflates the PUR signal. CIEM [2] estimates electronics accounts for 20–25% of eligible trade value; excluding it would materially raise Vietnam's measured performance and provide a more policy-relevant indicator.

Policy Implications. The diagnostic taxonomy generates differentiated and sequenced policy priorities. For Group 3 (garments), administrative and capacity–building reforms are necessary to approach the structural ceiling but cannot raise it. The pre–condition for PUR above 35% in garments is upstream industrial development: specialised industrial zones for weaving and dyeing, FDI attraction targeting Korean and EU fabric producers (the successful Toray Top Textiles USD 203 million investment in Nam Dinh in 2024 provides a replicable model), and acceleration of the national textile material centre announced in September 2024 (B–Company, 2026). These are long–horizon investments requiring sustained inter–ministerial coordination between MOIT and MPI. For Group 2 (footwear, wood), the priority is removing remaining D3 friction: a binding 48–hour SLA for C/O processing, full eCoSys integration with VNACCS/VCIS, and REX self–certification extension. For Group 1 (seafood), IUU card resolution is the priority – a diplomatic and fisheries management challenge, not an RoO one. For Group 4 (electronics), the appropriate response is reclassification in PUR measurement and monitoring of upstream investment to build future RVC, not RoO–related interventions. A cross–cutting implication: Vietnam's existing Korea–cumulation provision is substantially under–utilised in garments. Maximising Korean fabric sourcing (currently ~10% of imports; eligible under Protocol 1, Art. 3(7)) is the fastest near–term route to raising garment PUR without upstream industrial change.

Limitations. This study has three main limitations. First, sector-level PUR data in Vietnam are not systematically published, requiring triangulation and introducing some estimation uncertainty. Second, the framework is tested in a single FTA context; its portability to FTAs with different rules requires further validation. Third, the dimensions are treated as analytically distinct, though in practice they interact dynamically.

4. Conclusion

This paper has proposed and operationalized a four-dimensional framework for diagnosing FTA RoO utilization efficiency, applying it to five sectors of the EVFTA. The framework's central proposition—that D1×D2 incompatibility creates a structural ceiling on PUR that D3/D4 improvements cannot exceed—is strongly supported by the evidence. The garments sector's persistent ceiling at ~30%, juxtaposed with footwear's near-optimal 97–99%, provides the cleanest empirical illustration.

The paper advances two further contributions. Methodologically, it argues for excluding sectors with zero preference margins (electronics under the WTO ITA) from PUR denominators to correct a systematic downward bias. Policy-analytically, it translates the diagnostic taxonomy into sequenced interventions: upstream industrial investment for Group 3, administrative streamlining for Groups 1 and 2, and measurement reclassification for Group 4. Vietnam's target of ≥60% aggregate PUR by 2030 requires coordinated action calibrated to the binding constraint in each sector.

The proposed framework is replicable across other developing-country FTA contexts where restrictive RoO interact with weak upstream industries. Future research should validate it against panel data from multiple FTAs and explore the dynamic feedback between dimensions.

References:

1. Ministry of Industry and Trade (Vietnam). EVFTA implementation statistics 2020–2024. Hanoi: MOIT, 2024.
2. Central Institute for Economic Management (CIEM). *Five-year assessment of EVFTA implementation in Vietnam (2020–2025). Hanoi: CIEM, 2025.
3. BDI / European Commission. The utilisation of EU free trade agreements. Brussels: BDI/European Commission, 2022.
4. VCCI WTO Center. FTA utilization survey: Vietnamese enterprise awareness and EVFTA practice. Hanoi: VCCI, 2023.
5. Anson J., Cadot O., Estevadeordal A., de Melo J., Suwa-Eisenmann A., Tumurchudur B. Rules of origin in North–South preferential trading arrangements // Review of International Economics. 2005. Vol. 13, no. 3. P. 501–517.
6. Keck A., Lendle A. New evidence on preference utilization (WTO Staff Working Paper ERSD-2012-12). Geneva: WTO, 2012.
7. Albert J., Nilsson L. The cost of using EU trade preferences // World Economy. 2016. Vol. 39, no. 3. P. 318–339.
8. Brenton P., Manchin M. Making EU trade agreements work: the role of rules of origin // World Economy. 2003. Vol. 26, no. 5. P. 755–769.
9. Cadot O., Gourdon J. Non-tariff measures, preferential trade agreements, and prices // Review of World Economics. 2016. Vol. 152, no. 2. P. 227–249.
10. Cadot O., Carrère C., de Melo J., Tumurchudur B. Product-specific rules of origin in EU and US preferential trading arrangements // World Trade Review. 2006. Vol. 5, no. 2. P. 199–224.
11. Hayakawa K., Laksanapanyakul N. Firm-level analysis on FTA usage in Thailand // Economic Modelling. 2015. Vol. 48. P. 142–148.
12. Takahashi K., Urata S. On the use of FTAs by Japanese firms // Business and Politics. 2010. Vol. 12, no. 1. Article 2.
13. Hayakawa K. Does firm size matter in using FTA? // Developing Economies. 2015. Vol. 53, no. 3. P. 173–197.
14. Yin R. K. Case study research and applications: design and methods. 6th ed. Thousand Oaks: Sage, 2018.
15. Eisenhardt K. M. Building theories from case study research // Academy of Management Review. 1989. Vol. 14, no. 4. P. 532–550.
16. EuroCham. Business Climate Index Q2 2025. Hanoi: European Chamber of Commerce in Vietnam, 2025.
17. Vietnam Textile and Apparel Association (VITAS). Annual textile and garment industry report 2024. Hanoi: VITAS, 2024.
18. Vietnam Leather, Footwear and Handbag Association (LEFASO). Annual report 2024. Hanoi: LEFASO, 2024.
19. Vietnam Association of Seafood Exporters and Producers (VASEP). Seafood export report 2024. Ho Chi Minh City: VASEP, 2024.
20. VIFORES / HAWA. Vietnam wood and furniture industry report 2024. Hanoi: VIFORES/HAWA, 2024.
21. General Statistics Office of Vietnam (GSO). Industrial production and trade statistics 2024. Hanoi: GSO, 2024.
22. B-Company. Vietnam textile and garment industry report 2025–2026. Hanoi: B-Company, 2026.
23. European Commission. Progress report on IUU fishing: Vietnam. Brussels: DG MARE, 2023.
24. World Trade Organization (WTO). Ministerial Decision on the expansion of trade in information technology products. WT/MIN(15)/25. Geneva: WTO, 2015.