Distinguishing production and retail digitization alters investment incentives: simultaneous digital transformation and a revenue-/cost-sharing contract raise supply‑chain investment and profits, and under moderate production-digitization efficiency the contract also aligns digital upgrades with carbon reductions.

Main Finding

A two-tier supply-chain game-theoretic model shows that distinguishing production digitization (manufacturer) from business digitization (retailer) reveals strong synergy: simultaneous digital transformation (DX) raises investment levels and joint profits more than isolated moves, and centralized decision-making raises them further. A revenue-sharing and cost-sharing (RC) contract can fully coordinate the chain and produce Pareto improvements. Environmentally, DX reduces total emissions only if production-digitization investment efficiency is sufficiently high; at moderate efficiency levels the RC contract can simultaneously advance both digitization uptake and emission reductions.

Key Points

• Model setup and scenarios

  • Two-tier supply chain: manufacturer (production digitization) and retailer (business digitization).
  • Policy environment: carbon cap-and-trade (C&T) affects firms’ carbon costs and trading opportunities.
  • Decision regimes analyzed: unilateral DX (one party only), simultaneous decentralized DX (both decide independently), and centralized (joint optimization).
  • Coordination mechanism proposed: RC contract combining revenue-sharing and cost-sharing.

• Main theoretical results

  • Production and business digitization are complementary (synergistic): each raises the marginal return to the other.
  • Simultaneous decentralized DX leads to higher investment levels than isolated unilateral DX; centralized decision-making yields the highest investment and joint profit.
  • The RC contract can implement the centralized (first-best) outcome under decentralization and is Pareto-improving for both firms.
  • DX increases sales and supply-chain revenue; coordination reduces internal conflict and can accelerate DX adoption.
  • Environmental impact is ambiguous: total emissions fall only if production-digitization investment efficiency exceeds a threshold. If efficiency is low, demand-side gains (from business DX) can increase emissions (rebound effect).
  • At moderate production-digitization efficiency, the RC contract can achieve both increased DX investment and net emission reduction.

• Contributions claimed

  • Integrates digital transformation, supply-chain coordination, and carbon cap-and-trade into one framework.
  • Decomposes SCDX into upstream production vs downstream business digitization, revealing distinct mechanisms (cost-side vs demand-side) and interaction effects.
  • Extends contract-theoretic coordination literature with an RC contract tailored to dual-party DX problems.
  • Clarifies that DX does not automatically equal greening; investment efficiency is the critical boundary condition.

Data & Methods

• No empirical micro-data: the paper is theoretical/analytical.
• Methods
  • Game-theoretic modeling of a two-tier supply chain under carbon cap-and-trade.
  • Comparative equilibrium analysis across decision regimes: unilateral, simultaneous decentralized, and centralized optimization.
  • Design and analysis of a revenue-and-cost-sharing (RC) contract to achieve coordination.
  • Comparative statics to identify how parameters (notably production-digitization efficiency and carbon-trading price) affect investments, profits, and emissions.
  • Numerical simulations to illustrate results and study sensitivity (e.g., effects of carbon trading price).
  • Robustness checks / model extensions discussed (details in paper).
• Key modeling assumptions (high level)
  • Production digitization reduces unit production cost and (potentially) emissions per unit, with an efficiency parameter.
  • Business digitization increases demand/market size or willingness-to-pay for the retailer.
  • Carbon cap-and-trade imposes carbon costs or allows selling excess quota.
  • Investment in DX is costly and long-term; firms decide investment levels strategically.

Implications for AI Economics

• For firm strategy and organization

  • Distinguish upstream AI/automation (production-side) from downstream AI (customer analytics, personalization, intelligent logistics): their returns operate through different channels and are mutually reinforcing.
  • Joint investment (or coordinated contracts like RC) unlocks higher investment and higher joint surplus than isolated action; firms should consider revenue- and cost-sharing contracts to internalize cross-party externalities.
  • Managers should evaluate production-side AI efficiency carefully: if automation/AI adoption on the production side yields strong cost and emission efficiencies, joint DX is both profitable and green; if not, demand-side AI can increase emissions through higher volumes.

• For policy design

  • Carbon pricing and cap-and-trade design interact with digitalization incentives. A well-calibrated carbon price can strengthen the emission-reduction case for production-side AI and improve the effectiveness of coordination contracts.
  • Policies that raise the effective efficiency of production digitization (e.g., subsidies for low-emission automation, R&D tax credits, standards for energy-efficient AI/automation) make DX more likely to reduce emissions.

• For environmental outcomes and rebound risks

  • Digital/AI investments are not inherently carbon-reducing: the net effect depends on production-efficiency gains versus demand-stimulating effects (rebound). Empirical estimation of production-digitization efficiency is crucial.
  • Coordination mechanisms (contracts, possibly supported by policy) can help align incentives so that DX adoption also advances environmental goals at moderate efficiency levels.

• For research agendas in AI economics

  • Empirically estimate the “efficiency threshold” for production-side AI/automation above which DX reduces emissions—this threshold is central to policy and firm decisions.
  • Extend models to multi-period, multi-firm, platform-mediated supply chains and incorporate data externalities, fixed sunk costs, and labor-market impacts from AI adoption.
  • Study how information asymmetries (about DX returns or carbon efficiencies) affect contract design and adoption timing.
  • Evaluate real-world implementations of RC-like contracts and policy instruments that combine carbon pricing with digitalization subsidies.

One-line summary: Coordinated DX across upstream production automation and downstream business AI—enabled by revenue- and cost-sharing contracts and supported by suitable carbon pricing—can maximize supply-chain profits and, when production-digitization is efficient enough, also reduce emissions; without sufficient production efficiency, DX may not be green.