Executive Summary
The taxation of multinational digital enterprises represents one of the most consequential mechanism design challenges in contemporary international economics. The OECD/G20 Inclusive Framework's Two-Pillar Solution — Pillar One reallocating taxing rights to market jurisdictions and Pillar Two establishing a 15% global minimum effective tax rate — was hailed as a historic agreement when endorsed by 137 jurisdictions in October 2021. Yet as of early 2026, implementation remains incomplete, with Pillar One's Multilateral Convention unsigned by several major economies and unilateral digital services taxes proliferating despite commitments to their withdrawal.
This paper analyses the Two-Pillar Solution through the lens of mechanism design theory, examining whether the framework satisfies the conditions for incentive compatibility, individual rationality, and budget balance that would make it self-enforcing. We find that while Pillar Two's minimum tax achieves approximate incentive compatibility through its Income Inclusion Rule (IIR) backstop mechanism, Pillar One's profit reallocation formula faces fundamental incentive compatibility challenges arising from heterogeneous country preferences and information asymmetries about multinational profit allocation. The EU Tax Observatory estimates that incomplete implementation of the Two-Pillar framework leaves approximately $120–180 billion in annual tax revenue uncollected from profit shifting by the world's largest multinationals.
The Tax Competition Problem: A Race to the Bottom
International corporate taxation has long exhibited the characteristics of a multi-player game with suboptimal equilibria. When jurisdictions independently set corporate tax rates to attract mobile capital, the resulting tax competition drives rates below the socially optimal level — a dynamic formalised in the seminal models of Zodrow and Mieszkowski (1986) and Wilson (1986). Empirical evidence confirms this prediction: the global average statutory corporate tax rate fell from 40.1% in 1980 to 23.4% in 2025, with effective rates for the largest multinationals averaging just 17.2% according to the IMF Fiscal Monitor.
The digitalisation of the economy has intensified this dynamic. Digital business models — characterised by asset-light structures, user-generated value, and intellectual property mobility — enable profit shifting at scales impossible in the pre-digital era. The OECD estimates that base erosion and profit shifting (BEPS) by multinational enterprises costs governments $100–240 billion annually in lost corporate tax revenue, equivalent to 4–10% of global corporate income tax collections. For developing countries, the losses are proportionally larger: UNCTAD estimates that developing nations lose $47 billion annually to profit shifting by the top 100 multinational enterprises alone.
The strategic structure is a multi-player Prisoner's Dilemma. Each jurisdiction prefers a world in which all maintain reasonable tax rates (generating adequate public revenue) but has an individual incentive to undercut competitors (attracting mobile tax base). The Nash equilibrium involves suboptimally low rates and extensive profit shifting — a stable but collectively damaging outcome.
Pillar Two: The Global Minimum Tax as a Mechanism
Pillar Two's Global Anti-Base Erosion (GloBE) rules, establishing a 15% minimum effective tax rate for multinationals with consolidated revenues exceeding €750 million, can be understood as a mechanism designed to eliminate the most destructive forms of tax competition. The mechanism operates through two interlocking rules: the Income Inclusion Rule (IIR), under which a parent jurisdiction imposes a top-up tax on low-taxed foreign income, and the Undertaxed Profits Rule (UTPR), providing a backstop where the parent jurisdiction does not apply the IIR.
From a mechanism design perspective, Pillar Two exhibits several desirable properties. Incentive compatibility: the IIR/UTPR backstop structure means that no jurisdiction can attract real investment solely through tax rates below 15%, because the tax benefit is recaptured by the parent or UTPR jurisdiction. This converts what was previously a dominant strategy (offering sub-minimum rates) into a dominated strategy. Individual rationality: jurisdictions maintaining rates at or above 15% face no compliance burden and lose no revenue, making participation costless for the majority. Approximate budget balance: the top-up tax mechanism redistributes revenue from low-tax jurisdictions to IIR/UTPR-applying jurisdictions without requiring explicit transfer payments.
However, Pillar Two's mechanism is imperfect. The Qualified Domestic Minimum Top-up Tax (QDMTT) — allowing source jurisdictions to capture the top-up tax before it flows to parent jurisdictions — was introduced to address participation incentives but complicates the mechanism's revenue allocation. Furthermore, the 15% threshold, while representing a political equilibrium among Inclusive Framework members, sits below the rates that would maximise collective welfare: the EU Tax Observatory estimates that a 21% minimum would generate approximately $220 billion in additional annual revenue and substantially eliminate remaining profit-shifting incentives. The 15% rate reflects bargaining power dynamics rather than efficiency analysis.
Pillar One: The Profit Reallocation Challenge
Pillar One's Amount A — reallocating a portion of residual profits of the largest and most profitable multinationals (those with global revenues exceeding €20 billion and profitability above 10%) to market jurisdictions — faces more fundamental mechanism design challenges. The core difficulty is that Pillar One requires agreement on a formula for allocating profits across jurisdictions, and different jurisdictions have systematically different preferences over formula parameters depending on their economic structure.
Market jurisdictions (large consumer markets like India, Brazil, and Indonesia) prefer formulae weighting revenue-based factors, which allocate more profit to countries where products and services are consumed. Headquarter jurisdictions (the US, UK, and Netherlands) prefer formulae weighting headcount, assets, or existing nexus rules. Resource-rich jurisdictions prefer carve-outs for extractive industries. This preference heterogeneity means that no single formula satisfies the Gibbard-Satterthwaite theorem's conditions for strategy-proofness: jurisdictions have incentives to misrepresent their economic characteristics to influence formula outcomes.
The Multilateral Convention (MLC) implementing Pillar One, opened for signature in October 2023, reflects a laboriously negotiated compromise. Yet as of early 2026, key jurisdictions — including the United States, where Congressional ratification faces significant obstacles — have not signed. Meanwhile, the proliferation of unilateral Digital Services Taxes (DSTs) continues: the OECD counts 47 jurisdictions with enacted or proposed DSTs, generating an estimated $15–22 billion annually but creating compliance complexity and trade friction.
Unilateral DSTs: Defection from the Cooperative Equilibrium
The persistence of unilateral DSTs despite the Inclusive Framework agreement illustrates a classic cooperation enforcement problem. DSTs — typically levied at 2–5% on gross revenues of large digital companies — offer immediate, visible revenue to implementing jurisdictions. The commitment to withdraw DSTs upon Pillar One implementation was conditional on a timeline that has repeatedly slipped, eroding the credibility of the cooperative commitment.
Game-theoretically, the DST proliferation represents a trigger strategy breakdown. The implicit agreement was: "cooperate on Pillar One, and we'll remove DSTs." But as Pillar One implementation delays accumulated, the discount factor for future cooperation benefits declined below the threshold for sustaining cooperation in the folk theorem framework. Jurisdictions rationally concluded that the expected present value of future Pillar One revenue did not justify foregoing current DST revenue.
The resulting fragmentation imposes real costs. UNCTAD estimates that compliance with multiple overlapping DST regimes costs the largest digital multinationals $2.8–4.1 billion annually in direct compliance costs, with economic distortions — including cascading taxation of intermediate digital services — reducing global digital trade efficiency by an estimated 3–5%.
Redesigning the Mechanism: Lessons from Theory
Mechanism design theory suggests several principles for improving the international tax cooperation framework:
1. Dominant Strategy Implementation. Where possible, design mechanisms where cooperation is a dominant strategy rather than a conditional best response. Pillar Two's IIR/UTPR structure approximates this: cooperation (maintaining rates ≥15%) is optimal regardless of others' actions. Extending this principle to Pillar One would require a backstop mechanism — analogous to the UTPR — that captures reallocated profits if the primary mechanism fails.
2. Simplification and Transparency. The complexity of Pillar One's Amount A calculation (requiring determination of global consolidated profits, residual profits above a 10% benchmark, allocation keys across market jurisdictions, and marketing and distribution profits safe harbour provisions) creates information asymmetries exploitable by sophisticated multinationals. Simpler formulae, while less precisely targeted, reduce manipulation opportunities and lower compliance costs. The Vickrey-Clarke-Groves mechanism teaches that efficiency and simplicity in information requirements are often complementary.
3. Credible Commitment through Institutional Architecture. The absence of binding enforcement for Pillar One commitments — unlike WTO dispute resolution for trade agreements — undermines cooperative stability. Establishing a dedicated international tax dispute resolution body with binding authority would provide the institutional infrastructure necessary for sustained cooperation.
4. Compensation Mechanisms for Transition Losers. Any reallocation of taxing rights creates winners and losers. Jurisdictions that lose revenue under new allocation formulae (typically small headquarter jurisdictions and tax havens) have strong incentives to defect. Explicit compensation mechanisms — potentially funded by a portion of the efficiency gains from reduced profit shifting — could expand the set of individually rational outcomes.
Implications for GDEF's Finance & Economy Working Group
The international tax architecture is at a critical juncture. The partial success of Pillar Two demonstrates that well-designed mechanisms can achieve international tax cooperation even among jurisdictions with divergent interests. The continuing challenges of Pillar One illustrate the limits of negotiated formulae in environments with heterogeneous preferences and weak enforcement. GDEF's Finance & Economy Working Group will convene a dedicated symposium on digital tax mechanism design, drawing on the analytical frameworks presented here, to develop proposals for strengthening the multilateral tax cooperation architecture.
References & Sources
- OECD, Two-Pillar Solution to Address the Tax Challenges Arising from the Digitalisation of the Economy. OECD/G20 Inclusive Framework on BEPS. oecd.org/tax/beps
- IMF, Fiscal Monitor: Taxing Times Revisited. October 2025. imf.org/en/Publications/FM
- EU Tax Observatory, Global Tax Evasion Report 2025. Paris School of Economics. taxobservatory.eu
- UNCTAD, World Investment Report 2025: International Tax Reforms and Investment. United Nations Conference on Trade and Development. unctad.org
- Zodrow, G.R. and Mieszkowski, P. (1986). "Pigou, Tiebout, Property Taxation, and the Underprovision of Local Public Goods." Journal of Urban Economics, 19(3), 356–370. doi.org/10.1016/0094-1190(86)90048-3
- Gibbard, A. (1973). "Manipulation of Voting Schemes: A General Result." Econometrica, 41(4), 587–601. doi.org/10.2307/1914083
- Devereux, M.P. et al. (2021). Taxing Profit in a Global Economy. Oxford University Press. global.oup.com
- Clausing, K.A. (2020). "Profit Shifting Before and After the Tax Cuts and Jobs Act." National Tax Journal, 73(4), 1233–1266. doi.org/10.17310/ntj.2020.4.14
- Vickrey, W. (1961). "Counterspeculation, Auctions, and Competitive Sealed Tenders." Journal of Finance, 16(1), 8–37. doi.org/10.1111/j.1540-6261.1961.tb02789.x