Tighter cross-border data rules chill Chinese digital takeovers: countries with stricter data-flow limits see fewer initiated and completed digital M&A deals, longer transactions and a shift toward joint ventures—effects strongest for data-heavy and knowledge-rich targets.

Main Finding

Stricter cross‑border data flow restrictions materially reduce and reshape Chinese firms' cross‑border digital M&A activity (2014–2022). Higher restriction levels both raise transaction/compliance costs—leading to fewer completed deals, lower completion rates, and longer deal durations—and lower expected benefits—leading to fewer initiated deals and a shift toward joint ventures or partial ownership rather than wholly‑owned acquisitions. Effects are stronger for targets that are data‑sensitive or knowledge‑intensive.

Key Points

• Index construction: The paper builds a Cross‑Border Data Flow Restriction Index (CDFRI) for major economies (2014–2022) using the OECD Digital Services Trade Restrictiveness Index (DSTRI) and the entropy‑weight method. Higher CDFRI = stricter restrictions.
• Empirical focus: Chinese firms' cross‑border digital M&A across targets in the sample economies, 2014–2022.
• Main mechanisms:
  • Cost channel: Restrictions increase transaction and compliance costs (localization, audits, liability exposure), reducing deal completion, lowering completion rates, and lengthening deal timelines.
  • Benefit channel: Restrictions reduce expected synergies (impeding algorithm transfer and post‑merger knowledge integration), resulting in fewer initiated deals and more governance forms that limit ownership exposure (e.g., joint ventures).
• Heterogeneity:
  • Data‑sensitive targets: stronger negative effect because of higher localization/audit/liability risk.
  • Knowledge‑intensive targets: stronger negative effect because restrictions obstruct algorithm transfer and knowledge integration post‑deal.
• Robustness: Results obtained using negative binomial regressions (deal counts), Probit models (deal/completion likelihood), and moderated regression analyses (heterogeneity).

Data & Methods

• Data inputs:
  • OECD DSTRI indicators (used to build CDFRI by entropy weighting).
  • Transaction‑level data on Chinese firms' cross‑border digital M&A (2014–2022), with target characteristics (data sensitivity, knowledge intensity), deal outcomes (initiation, completion, duration), and governance type.
• Index construction:
  • Entropy‑weight method aggregates DSTRI components into a single CDFRI per country-year.
• Empirical strategy:
  • Negative binomial regression for counts of deals (to model initiation/incidence).
  • Probit models for binary outcomes (e.g., deal completion).
  • Moderated regressions and interaction terms to test heterogeneous effects by target data sensitivity and knowledge intensity.
• Identification:
  • Controls for country, time, and deal/firm covariates; robustness checks reported (details in paper).

Implications for AI Economics

• Global AI diffusion and scale effects: Data‑flow restrictions fragment the data environment, impeding cross‑border transfer of training data and models, which reduces economies of scale in AI development and slows international diffusion of algorithms.
• Innovation and knowledge spillovers: Restrictions hinder post‑merger knowledge integration and algorithm transfer, lowering potential technology spillovers from cross‑border M&A and thus possibly reducing global productivity and innovation spillovers.
• Firm strategy and governance:
  • Firms may shift away from full acquisitions toward joint ventures, minority stakes, licensing, or purely local partnerships to mitigate data‑localization and compliance risks.
  • Expect increased investment in onshore data infrastructure, localized R&D, and legal/compliance capacity in target jurisdictions.
• Market structure and competition: Reduced cross‑border M&A can preserve or entrench local incumbents, slow international consolidation in digital sectors, and raise barriers to entry for foreign digital competitors.
• Policy trade‑offs and recommendations:
  • Policymakers face a trade‑off between data protection/security and international digital integration. To balance this, coordinated approaches could include data adequacy frameworks, interoperable compliance standards, safe‑harbors for trusted transfers, and bilateral/multilateral data‑sharing agreements to reduce transaction costs without compromising legitimate regulatory objectives.
• Directions for further research: Quantify welfare trade‑offs from restrictions, trace long‑run impacts on productivity and R&D, sectoral heterogeneity (platforms vs. enterprise software), and effects on cross‑border labor and knowledge flows.