Decoding the 2+7 Framework: Cross-Border Service Trade in Asia

By Senior Research Fellow, Cross-border Service Economics

The 2+7 cooperation framework, tabled by Premier Li Keqiang at the 16th China-ASEAN Summit in Brunei in October 2013, is most often read as a diplomatic architecture: two political consensuses paired with seven priority fields spanning the political sphere, business, connectivity, finance, maritime cooperation, security, and people-to-people, scientific and environmental cooperation. Read that way, it belongs to the world of summit communiqués. Read through the lens of cross-border service trade, the same framework reveals a structural asymmetry that has shaped a decade of regional integration. Its seven pillars were calibrated primarily around goods movement, physical infrastructure, and security posture, while the regulatory plumbing that actually governs whether a digital service can cross a border — data localization rules, licensing discretion, Mode 1 and Mode 3 commitments — sits only obliquely inside the architecture. That misalignment is the central constraint this article disassembles, and the measure by which the framework’s second decade will be judged.
 

The Architecture of the 2+7 Framework Through a Service-Trade Lens

 

The Two Consensuses and Their Asymmetric Weight

 

The “2” in the framework denotes two political consensuses: first, a commitment to deepen strategic trust and good-neighborliness between China and ASEAN; second, a parallel pursuit of political-security cooperation and economic development. For goods trade, this dual track translated cleanly into tariff schedules, rules-of-origin harmonization, and the China-ASEAN Free Trade Area (CAFTA) upgrade targeting 1 trillion US dollars in bilateral trade by 2020. The translation into service-trade terms is far less direct. Political trust does not, by itself, remove a foreign-equity cap in Indonesian logistics, nor does a security consensus neutralize a data-residency requirement in Viet Nam. The consensus layer establishes the diplomatic precondition for negotiation, but it carries no binding force on the regulatory instruments that govern service supply.
 

This is where the analytical separation between goods trade and service trade becomes essential. Goods move through customs checkpoints; services move through regulatory regimes. A tariff line is a single, transparent number. A service-trade barrier is a composite of licensing procedures, movement-of-natural-persons rules, technology-transfer conditions, and digital regulatory measures that resist single-parameter description. The two consensuses of the framework operate on the diplomatic stratum above these regimes, and their effect on services is therefore mediated — and diluted — by each member state’s domestic regulatory apparatus. A consensus to cooperate, in service-trade terms, is only as strong as the domestic statutes it never touches.
 

The Seven Pillars and Where Services Actually Sit

 

The seven cooperation fields can be sorted by their direct service-trade content. Connectivity, finance, and the people-to-people cluster carry the heaviest service logic; the political, business, and security pillars carry service implications but are framed in goods-and-infrastructure language. The connectivity pillar produced the Asian Infrastructure Investment Bank (AIIB) and the Silk Road Fund — instruments that finance ports, railways, and power grids, which are goods-anchored even when they enable service delivery downstream. The finance pillar speaks to currency cooperation and the AIIB’s capital structure, not to the cross-border supply of financial services themselves. The people-to-people pillar funds education and training centers, which sit adjacent to Mode 4 (presence of natural persons), but it does not bind the movement-of-natural-persons commitments that would make professional exchange frictionless.
 

The honest reading is that the seven pillars address the preconditions for service trade — physical connectivity, financing, trust — without directly dismantling the regulatory barriers that constrain it. The framework is a scaffold, not a key. Its service-trade effect is real but indirect: better roads lower the cost of delivering a logistics service; cheaper financing expands commercial-presence investment; trust creates the political space for regulatory negotiation. None of these, however, removes a single line from a negative list or eliminates a data-localization mandate. That work was left to downstream instruments, and the gap between the framework’s ambition and its regulatory reach is where the service-trade deficit lives.
 

Measuring the Service-Trade Barrier Gap in the Framework’s Footprint

 

STRI Readings Across the China-ASEAN Corridor

 

To quantify what the framework leaves unaddressed, the WTO-World Bank Services Trade Restrictiveness Index (STRI) offers the cleanest diagnostic. The STRI scores economies from 0 (fully open) to 100 (fully closed) across five sectors and 34 subsectors, capturing the policy restrictiveness of each regime through surveys administered across 133 economies. The ASEAN average tells an uncomfortable story. On Mode 1 — cross-border supply, the mode most relevant to digitally delivered services — the ASEAN average sits at 70.1, against a world average of 58.5. On Mode 3, commercial presence, ASEAN averages 54.4 against a global 44.5. On Mode 4, presence of natural persons, the gap is narrower but still negative: 51.2 versus 44.8. Every mode of supply is, on average, more restrictive in the framework’s footprint than in the rest of the world.
 

The intra-regional variance is equally revealing. Malaysia registers a Mode 1 score of 86.7, Indonesia 83.1, Viet Nam 76.6, Thailand 75.1, and Singapore 74.8. Singapore’s high Mode 1 figure is something of an artifact — its restrictiveness concentrates in Mode 4, where it scores 70.7, reflecting tight labor-market protection rather than digital barriers per se. Strip out that artifact and the picture sharpens: the larger ASEAN economies that anchor the framework’s economic logic are precisely those with the most restrictive cross-border service regimes. Indonesia and Thailand top the restrictiveness chart across multiple sectors; Indonesia peaks in professional services, Malaysia in distribution, the Philippines in finance, and Thailand in transport. Tourism services in ASEAN run 31.8% more restrictive than the world average, finance 28.1%, and professional services 24.4% — figures that sit uneasily alongside three decades of regional liberalization rhetoric.
 

The Digital Service Restrictiveness Deficit

 

The OECD Digital Services Trade Restrictiveness Index (DSTRI), which scores 129 economies on a 0-to-1 scale where 0 signals an open regulatory environment and 1 a closed one, exposes the digital dimension more sharply. The European Centre for International Political Economy (ECIPE) Digital Trade Restrictiveness Index, covering 65 economies, places four ASEAN members among the most restrictive regimes globally: Indonesia ranks 4th, Viet Nam 5th, Thailand 10th, and Malaysia 11th. These are not marginal positions. They place the framework’s core economic partners in the top tier of digital-trade restrictiveness worldwide.
 

The barrier inventory behind these rankings is specific and operational. Data localization mandates require certain categories of data to be stored on domestic servers before processing. Cross-border data transfer rules condition export on adequacy determinations, consent mechanisms, or security review. Licensing discretion grants regulators opaque authority over market entry for digital service providers, often without published decision criteria. Foreign-equity ceilings cap commercial-presence investment in logistics, finance, and telecommunications at thresholds that block majority control. Source-code disclosure requirements and technology-transfer conditions attach to certain digital infrastructure investments, raising the cost of market entry for platform and cloud providers. Each of these is a discrete regulatory instrument, and each operates independently of the diplomatic consensus that the 2+7 framework establishes. Removing one does not neutralize the others; the binding constraint is whichever instrument is most restrictive for a given service category.
 

Where Goods Logic Breaks Down for Cross-Border Services

 

The temptation, when assessing a framework like 2+7, is to apply the analytical apparatus that works for goods trade: identify tariff peaks, compute trade-weighted averages, model welfare gains from liberalization. That apparatus fails for services in three structural ways, and tracing each failure clarifies why the framework’s goods-calibrated design underdelivers on its service potential.
 

First, services lack a single price instrument. A tariff is a number; a service barrier is a vector. The STRI captures this by weighting dozens of regulatory measures across modes and sectors, but the composite score obscures the specific binding constraint. A Mode 1 score of 86.7 in Malaysia could be driven by licensing opacity, by local-presence requirements, or by data rules — and the corrective policy differs for each. Parameter disassembly, not aggregate scoring, is what reveals the constraint a service exporter actually faces.
 

Second, services are consumed at the point of production. A goods shipment can be inspected at a border crossing; a cloud service is delivered into the buyer’s infrastructure in real time, with no customs post in between. This means the regulatory frontier for services is not the border but the domestic rulebook — data protection law, telecommunications licensing, professional qualification recognition, payment-systems access. The 2+7 framework’s connectivity pillar builds roads and ports, but the binding constraint on a cross-border data service is a domestic data-protection statute, not a missing railway. Goods logic points the analyst toward infrastructure gaps; service logic points toward regulatory gaps, and the two point in different directions.
 

Third, services trade is dominated by Mode 3 — commercial presence — which moves through foreign-direct-investment channels, not through trade channels. The framework’s business pillar addresses trade in goods; its finance pillar addresses infrastructure financing. Neither directly engages the foreign-equity caps, board-nationality requirements, and local-content rules that govern whether a service firm can establish commercial presence in the first place. The ASEAN Framework Agreement on Services (AFAS) and its successor, the ASEAN Trade in Services Agreement (ATISA), address these instruments, but the framework itself does not. The 10th AFAS package, completed in 2023, still leaves more than a fifth of subsectors unbound on Mode 1 for Indonesia (26%), the Philippines (42%), Singapore (27%), Thailand (31%), and Viet Nam (27%) — and the share of fully unrestricted Mode 3 subsectors sits below 10% for Indonesia, Malaysia, the Philippines, and Thailand. Thailand’s entire committed subsector set carries foreign-equity limits ranging from 25% to 70%.
 

These three failures map onto the framework’s design logic. The framework was built to coordinate goods-and-infrastructure flows between sovereign states; the binding constraints on service trade sit inside each state’s domestic regulatory regime. The gap is structural, not incidental, and it will not close by adding more diplomatic summits.
 

Case Verification — RCEP, ATISA, and the Digital Economy Framework

 

RCEP’s E-Commerce Chapter as a Partial Fix

 

The Regional Comprehensive Economic Partnership (RCEP), signed on November 15, 2020, is the most consequential downstream instrument of the 2+7 framework’s economic logic. Covering 15 economies, roughly 2.27 billion people, and about 30% of global GDP, RCEP is the largest free-trade bloc by population and economic weight. Its e-commerce chapter — Chapter 12 — represents the first region-wide binding digital trade rule set among major East Asian economies, and it operationalizes several pillars the 2+7 framework left in declarative form.
 

RCEP’s service-trade relevance is concrete and article-level. Article 12.5 mandates paperless trade; Article 12.6 addresses electronic authentication and signatures; Article 12.10 requires parties to establish a domestic framework for electronic transactions; Article 3.14 provides trade facilitation for certified operators; Article 10.17 addresses investment facilitation. The intellectual property chapter runs to 83 articles, the most extensive among RCEP members to date, and includes transitional arrangements and technical-assistance annexes for the least-developed members. These provisions create a baseline of transparency and interoperability that the 2+7 framework’s pillars gestured toward but never codified into binding text.
 

The fix is partial for identifiable reasons. RCEP’s cross-border data flow provisions allow parties to impose restrictions for legitimate public-policy objectives, a carve-out that preserves domestic regulatory discretion and leaves the hardest questions unanswered. Several ASEAN members — Cambodia, the Philippines, Thailand — use positive-list scheduling for services, meaning liberalization applies only to listed sectors and remains subject to inscribed limitations. The agreement defers hard decisions on mandatory data localization and source-code access to future negotiation rounds. RCEP raises the floor; it does not dismantle the walls, and the walls are precisely where the service-trade deficit lives.
 

The Unfinished ATISA and DEFA Agenda

 

ATISA, in effect since 2015, represents ASEAN’s own attempt to deepen service integration beyond the AFAS architecture. Its shift from the positive-list AFAS modality to a negative-list approach — requiring members to disclose all non-conforming measures across national treatment, market access, local presence, and board-nationality requirements — is a genuine institutional advance. The negative-list structure locks in the regulatory status quo and, through a ratchet mechanism, binds future liberalization automatically, preventing backsliding.
 

The 10th package of commitments, finalized in 2023, improved on the 7th package but remains restrictive for the larger economies. The ASEAN Framework Agreement on E-commerce, by contrast, consists largely of soft-law provisions and is widely judged insufficient to deliver the proposed ASEAN-wide digital economic community. The Digital Economy Framework Agreement (DEFA), under negotiation, is the instrument most likely to close the gap — but its outcome is uncertain, and the four ASEAN members ranked among the world’s most restrictive digital regimes will shape its ceiling. A DEFA text that reproduces RCEP’s public-policy carve-outs without narrowing them would leave the binding constraints intact.
 

The case evidence converges on a single finding: the 2+7 framework established the diplomatic and infrastructure preconditions, but the regulatory dismantling required for genuine cross-border service integration has been left to downstream instruments whose ambition is bounded by the same domestic regulatory discretion the framework never constrained. Each successive instrument — AFAS, ATISA, RCEP, DEFA — has narrowed the gap incrementally, but none has closed it, and the cumulative effect is a region whose service-trade architecture remains far more restrictive than its goods-trade architecture.
 

Constraint Mapping and Corrective Pathways

 

Root-cause tracing of the service-trade deficit inside the 2+7 footprint points to four binding constraints, each paired with a corrective pathway that operates within the framework’s existing institutional footprint.
 

The first constraint is the goods bias in framework architecture. The seven pillars privilege physical connectivity and goods exchange; service regulatory convergence has no dedicated pillar and no standing review mechanism. The corrective pathway is institutional: embed a standing service-trade regulatory-coordination mechanism — modeled on the ATISA reservation-list review process — directly into the framework’s implementation architecture, so that service barriers receive the same ongoing ministerial attention as infrastructure financing. What is reviewed annually gets reformed; what is reviewed only at leadership summits does not.
 

The second constraint is the measurement gap. Framework progress is tracked in goods-trade volumes and infrastructure disbursements, not in STRI or DSTRI movement, and the absence of measurement allows restrictiveness to persist without political cost. The corrective pathway is analytical: adopt the OECD Digital STRI Simulator and the WTO-World Bank STRI as standing diagnostic instruments, with annual scorecard reporting tied to the framework’s review cycles. Transparency on barrier levels is itself a liberalizing pressure, because it converts an opaque regulatory regime into a comparable, rankable policy variable.
 

The third constraint is domestic regulatory discretion on data. Cross-border data flow rules, localization mandates, and source-code conditions sit outside any framework discipline, and the public-policy carve-outs in RCEP preserve them in amber. The corrective pathway is rule-based: negotiate binding cross-border data transfer disciplines — with narrow, transparent, and challengeable public-policy exceptions — as the substantive core of DEFA, and align them upward from the RCEP e-commerce chapter rather than reproducing its deferrals. A carve-out that cannot be tested is a carve-out that becomes permanent.
 

The fourth constraint is the Mode 4 gap. People-to-people cooperation is funded through education and training centers, but the movement-of-natural-persons commitments that would convert training into professional service delivery remain unbound across large shares of subsectors. The corrective pathway is mutual-recognition deepening: convert the eight existing ASEAN mutual recognition arrangements — covering engineering, architectural, accountancy, surveying, nursing, medical, dental, and architectural services — into binding licensing-access commitments, and extend the model to China-ASEAN professional service corridors under the framework’s people-to-people pillar. Training without licensing access produces graduates who cannot deliver services across borders; the constraint is not capacity but authorization.
 

None of these pathways requires renegotiating the 2+7 framework itself. They require redirecting the institutional energy already flowing through its pillars toward the regulatory instruments that actually govern cross-border service supply. The framework’s diplomatic capital is sufficient; what has been missing is the analytical insistence on spending it where the barriers actually sit.
 

The Structural Reading

 

The 2+7 cooperation framework is best understood not as a service-trade instrument — it is not one — but as the diplomatic and infrastructural envelope within which service-trade liberalization must occur. Its two consensuses establish trust; its seven pillars build the roads, the financing, and the goodwill. The barriers to cross-border service trade, and to digital service trade in particular, sit one layer below: in the licensing discretion, data rules, equity caps, and Mode 4 restrictions that the framework leaves to domestic regulators. A decade of implementation has confirmed that the framework’s goods-calibrated design can move container traffic and capital flows, but it cannot, by itself, move a cloud service across a regulatory border. Closing that gap is the unfinished work of East Asian service globalization — and the measure by which the framework’s second decade will be judged.
 

Reference
 

Source Reference Link: https://wiki.mbalib.com/wiki/2+7 合作框架

Link Brief: The mbalib wiki entry outlines the 2+7 cooperation framework as a special economic and trade coordination mechanism anchored by two principal consensuses and seven cooperation fields, commonly applied to regional industrial and infrastructure coordination. This article adopts the framework’s structural parameters as the analytical scaffold for dissecting cross-border service trade barriers.
 

Disclaimer
 

This article is for general reference only and does not constitute professional R&D guidance, production process advice or quality certification. All material performance data has specific test premises; readers should verify parameters against actual equipment and working conditions.