Optimal government subsidy strategies for improving charging stations in ride-hailing services

Technical Summary: Optimal Government Subsidy Strategies for Improving Charging Stations in Ride-Hailing Services

Problem Statement
The integration of New Energy Vehicles (NEVs) into ride-hailing services has created a critical dependency between ride-hailing efficiency and charging infrastructure. While governments worldwide have shifted subsidy focus from vehicle purchases to charging infrastructure, the optimal strategy for supporting ride-hailing specifically remains unclear. Existing literature often overlooks the interplay between ride-hailing platforms, heterogeneous charging providers (fast vs. conventional), and specific market conditions such as urban economic development levels. This paper addresses the gap by investigating how different government subsidy strategies—charging subsidies (for drivers), usage subsidies (for station operators), and building subsidies (for infrastructure construction)—impact the equilibrium outcomes of a competitive ride-hailing and charging ecosystem.

Methodology
The authors develop a game-theoretic model involving four key entities: a ride-hailing platform, a conventional charging service provider (Provider-c), a fast-charging service provider (Provider-f), and the government. The model incorporates:

• Two-Stage Decision Process: Drivers first decide whether to become active based on income and charging convenience, then select between fast or conventional charging based on price, waiting time, and travel distance.
• Subsidy Scenarios: Three distinct strategies are analyzed:
  1. NC Strategy: Charging subsidy ($\tau$) for drivers using fast-charging stations.
  2. NU Strategy: Usage subsidy ($\mu$) for Provider-f based on station utilization.
  3. NB Strategy: Building subsidy ($\eta$) for Provider-f based on the number of stations constructed.
• Extensions: The basic model is extended to include scenarios where the platform invests in fast-charging stations via cooperative shareholding and scenarios accounting for heterogeneity in drivers' sensitivity to the number of charging stations.
• Analysis: The study employs backward induction to derive equilibrium decisions (prices, station counts, fares) and conducts sensitivity analyses on parameters such as network externalities, charging time cost differentiation ($\Delta$), and building cost coefficients ($k$).

Key Contributions

1. Substitutability and Dominance: The paper establishes that charging subsidies and usage subsidies are mutually substitutive, exerting similar impacts on market equilibrium. In contrast, building subsidies play a distinct, positive role. The study identifies a critical threshold: when building subsidies exceed a certain level, the building subsidy strategy dominates both charging and usage subsidies, yielding superior outcomes for all entities.
2. Non-Linear Profitability: Contrary to intuitive expectations, the research finds that fast-charging providers do not always increase profitability as the charging time cost differentiation ($\Delta$) increases. High building costs ($k$) combined with high differentiation can lead to reduced market capacity and lower profits.
3. Marginal Thresholds: The analysis reveals that the marginal building subsidy thresholds required to ensure a "win-win" situation for all entities do not intersect. This implies that achieving a universally beneficial outcome requires specific, non-overlapping subsidy levels depending on the strategy chosen.
4. Regional Policy Implications: The study provides a framework for tailoring subsidy strategies to local economic conditions. Developed regions, characterized by higher service rates and sensitivity, are better suited for usage or charging subsidies. Conversely, underdeveloped regions may achieve better results with building subsidies to overcome initial infrastructure barriers.

Key Results

• Impact of Subsidies: Increasing charging or usage subsidies prompts the fast-charging provider to raise additional fares, whereas building subsidies directly stimulate station expansion without altering the fare structure.
• Win-Win Conditions: A "win-win" scenario for the platform, both providers, and passengers is achievable under the building subsidy strategy if the subsidy level ($\eta$) surpasses the maximum of specific thresholds ($\max\{\eta_2, \eta_5, \eta_6\}$).
• Platform Investment: When the platform cooperates with the fast-charging provider via shareholding, the region for a "five-way win" (including passengers) shrinks under the building subsidy strategy but expands under the usage subsidy strategy as the shareholding ratio increases.
• Driver Sensitivity: Heterogeneity in drivers' sensitivity to charging station availability confirms the robustness of the basic model. Drivers prioritizing fast-charging efficiency ($\rho_1$) expands the multi-win region, whereas sensitivity to conventional stations ($\rho_2$) can constrain the effectiveness of usage subsidies.

Significance and Claims
The paper claims that its findings offer critical guidance for government authorities and ride-hailing managers. It asserts that a "one-size-fits-all" subsidy approach is ineffective; instead, policy must be contingent on local economic development levels and the specific market structure (e.g., whether the platform invests in infrastructure). The study emphasizes that while building subsidies are generally more robust and positive, their dominance is conditional on exceeding specific thresholds. Furthermore, the research highlights that ride-hailing platforms are motivated to invest in fast-charging infrastructure, particularly in developed urban areas, to enhance service efficiency. Finally, the paper notes that the pursuit of charging efficiency by drivers acts as a catalyst for the sustainable development of urban transportation, provided that subsidy strategies are aligned with these market dynamics.

Limitations
The authors acknowledge that the model currently excludes privately owned NEVs that do not participate in ride-hailing, which represent a significant customer base for charging providers. Additionally, the model simplifies the market competition to a specific set of providers, whereas real-world markets may involve more complex three-way or multi-party competition dynamics.