EV Charging PPPs Get a New Subsidy Blueprint—Performance Pays
The race to electrify transportation is no longer about who builds the best battery or fastest sedan—it’s increasingly about who builds the infrastructure that keeps those vehicles moving. And in that race, one quiet but powerful factor is beginning to shape outcomes more decisively than steel, cables, or even kilowatts: how governments pay for it. A new study published in the Journal of Industrial Engineering and Engineering Management offers a detailed, rigorously modeled roadmap for how public-private partnerships (PPPs) in EV charging can be structured—not just to build more stations, but to build better ones, faster, and with higher long-term value for society.
At first glance, the problem seems straightforward. There are still too few public chargers, too many bottlenecks in deployment, and too much uncertainty around returns for private investors. But beneath the surface, the real challenge is behavioral: when private operators shoulder the cost and risk of building and maintaining charging networks, their incentives don’t always align with the public good. They may prioritize high-traffic corridors and premium pricing over equitable access or grid resilience. Left unchecked, this misalignment can slow adoption, deepen regional disparities, and ultimately undermine the climate goals that electrification is meant to serve.
Enter performance-based subsidies—payments not simply for constructing a station, but for delivering measurable outcomes: uptime, user satisfaction, integration with smart grids, even support for demand-response programs. This is where the work of Zhang Yiting, Wang Xueqing, Ma Rui, Liu Yunfeng (Tianjin University), and Wang Dan (Chongqing University) breaks new ground. Rather than offering broad policy prescriptions, they construct a Stackelberg game—a hierarchical model where the government moves first (setting subsidy rules), and private operators respond (choosing effort levels)—to simulate how performance-linked payments can reshape behavior.
What emerges is not a one-size-fits-all prescription, but a nuanced framework revealing when and how performance incentives actually move the needle—and when they don’t.
One of the study’s most striking findings is the critical role of operator cost structure. Not all charging companies are the same. Some—perhaps backed by tech-savvy startups or integrated energy giants—can deploy software updates, dynamic pricing, or predictive maintenance at relatively low marginal cost. Others, particularly smaller regional firms, face steep internal costs for every extra hour of technician time, every additional layer of data analytics, or every upgrade to interoperability standards.
The model shows that when operators’ effort-cost coefficient is low, performance-based subsidies act like turbochargers. A small increase in the incentive rate (say, per-percentage-point gain in station uptime or customer satisfaction) triggers a disproportionately large boost in effort. Operators respond by optimizing routing for service crews, investing in real-time monitoring dashboards, or even co-locating chargers with retail to improve user dwell time and safety.
But flip the script—and the effect evaporates. When effort is expensive (think: remote locations, legacy hardware, labor shortages), operators become far less responsive to performance tweaks. No matter how generous the bonus, the internal cost of chasing that last 5% of reliability or responsiveness simply isn’t worth it. In those cases, the study warns, flat subsidies or upfront capital grants may actually deliver more social value—freeing the operator from the impossible math of high-effort marginal returns.
This insight alone could save governments millions in misallocated incentives. Instead of rolling out a uniform “pay-for-performance” program nationwide, a smarter approach would segment bidders during procurement: categorize them by projected operational complexity, then tailor the subsidy structure accordingly—performance-heavy for agile players in dense urban zones, fixed-support-heavy for those tackling underserved rural corridors.
Then there’s price. Not electricity price—but the toll users pay to charge. One might assume that higher charging fees hurt demand, and thus hurt social welfare. But the model reveals a surprising twist: it depends on how users react, and how much the operator benefits from extra revenue.
Let’s say the government allows a higher per-kWh rate—but ties part of the operator’s subsidy to demand retention. If the price increase is modest and the operator responds by improving service (faster repairs, better app integration, cleaner facilities), riders may shrug and keep plugging in. In fact, the extra revenue lets the operator double down on quality—creating a virtuous cycle.
But if prices jump too high, or service doesn’t improve in lockstep, demand collapses—and so does welfare. Crucially, the researchers identify a tipping point: a precise price band where welfare climbs with price, then abruptly declines. Beyond that threshold, consumer surplus evaporates faster than operator gains accumulate.
This has direct relevance for today’s debates over “value-based pricing” at fast-charging hubs. Tesla’s Supercharger network, for example, already varies prices by location and time of day. Other networks are experimenting with membership tiers and idle fees. The study doesn’t say “don’t price dynamically”—it says calibrate carefully, and never decouple pricing from performance obligations. A higher price should earn the operator a higher subsidy—but only if riders don’t flee.
Perhaps the most counterintuitive result involves uncertainty—specifically, how unexpected shifts in demand (say, due to a sudden EV tax credit, a viral range-anxiety story, or a major automaker’s recall) affect optimal policy.
Common sense says: when demand is volatile, play it safe. Stick to fixed payments. Avoid performance metrics that could swing wildly month to month.
The model says the opposite—but only under certain conditions. When operator effort is cheap, moderate demand declines actually increase optimal effort. Why? Because the operator, facing falling revenue, leans harder on controllable levers—like service quality or uptime—to retain customers. In that scenario, a well-designed performance subsidy becomes a stabilizing force: it rewards the operator precisely when they’re working hardest to weather the storm.
Even more intriguing: there exists a sweet spot for demand volatility itself. Too stable, and operators get complacent. Too chaotic, and planning becomes impossible. But in a Goldilocks zone of moderate, predictable fluctuation, welfare actually peaks. The implication? Governments shouldn’t aim for ultra-stable demand forecasts. Instead, they should steer variability—using pilot programs, targeted incentives, or even public messaging—to keep the ecosystem in that productive tension zone.
Of course, real-world policy can’t chase theoretical optima without considering hard constraints—namely, budget ceilings and minimum profit guarantees for private partners. Here, the researchers extend their model to six realistic scenarios, from “plenty of public funds + low private expectations” (the ideal) to “tight budget + high profit demands” (the nightmare).
Their guidance is refreshingly pragmatic: when the ideal performance plan is unaffordable, don’t just slash the incentive rate. Instead, layer in a fixed base subsidy, then modulate the performance multiplier to hit both fiscal and participation targets. Think of it like a salary-plus-bonus structure: the fixed portion ensures survival; the variable portion preserves motivation.
This hybrid approach appears in nascent form in programs like California’s CALeVIP, which mixes upfront vouchers with bonus payments for stations meeting utilization or uptime benchmarks. The Tianjin-Chongqing team’s work suggests such hybrids aren’t just politically expedient—they’re economically optimal under real-world constraints.
Zooming out, the study arrives at a broader philosophical point: performance-based subsidies aren’t just financial tools—they’re coordination mechanisms. They translate abstract public goals (equity, resilience, speed of rollout) into concrete, auditable actions private firms can execute.
But—and it’s a crucial but—those metrics must be chosen wisely. The model assumes performance (μ) is a direct function of effort (e): more effort → higher uptime, faster repairs, better user ratings. In practice, however, not all “performance” is equally valuable. A station that’s 99% available but located 20 miles off the highway serves fewer societal needs than one at 95% uptime on a key freight corridor—even if the latter scores lower on a generic reliability index.
The researchers acknowledge this limitation: their model simplifies performance into a single linear function. Future work, they suggest, should incorporate multi-dimensional metrics—geographic equity, grid support capability, compatibility with low-income EV models—ideally weighted by local priorities. One jurisdiction might prioritize “stations per capita in disadvantaged ZIP codes”; another, “MW of controllable load for ancillary services.” The subsidy contract would then reflect that diversity.
Already, signs of this evolution are appearing. The U.S. National Electric Vehicle Infrastructure (NEVI) program mandates that 75% of corridor chargers be located within one mile of a highway interchange—explicitly privileging accessibility over pure volume. Meanwhile, the UK’s Office for Zero Emission Vehicles now requires grant recipients to report not just usage, but average queue times and complaint resolution rates.
What the Tianjin-Chongqing analysis provides is a rigorous framework to go further: to test, in silico, how shifting metric weights affects operator behavior, public benefit, and fiscal sustainability—before committing hundreds of millions to a program design.
So where does this leave policymakers today?
First: diagnose before prescribing. Before drafting subsidy terms, governments should gather data—or require bidders to disclose—estimates of their marginal effort costs. A simple questionnaire probing internal processes (How many field techs per 100 stations? What’s your mean-time-to-repair target? Do you integrate with utility demand-response platforms?) could generate enough signal to segment operators meaningfully.
Second: embrace adaptive pricing. Rather than fear price variability, design contracts that allow—and even encourage—dynamic rates, as long as they’re paired with strong consumer protections (e.g., caps on off-peak premiums) and performance backstops (e.g., bonuses tied to price-adjusted satisfaction scores).
Third: rethink risk. Instead of viewing demand uncertainty as a threat to be minimized, consider how moderate, managed volatility can enhance responsiveness. Pilot programs could even introduce controlled variation—say, by staggering incentive rollouts across regions—to keep operators agile.
Fourth: hybridize intentionally. When budget or participation constraints bind, combine fixed and variable payments deliberately—not as a compromise, but as a strategic lever. Use the fixed component to guarantee baseline network build-out; reserve performance payouts for differential achievements: ultra-fast deployment, integration with renewables, or service in equity-priority zones.
Finally—and perhaps most importantly—measure what matters. Move beyond kWh delivered and station count. Track outcomes like median charge time during peak hours, percentage of stations interoperable with all major EVs, or reduction in “charge anxiety” reported in user surveys. Then, feed those metrics back into the subsidy formula.
The electrification of transport is a marathon, not a sprint. But the pit stops—where drivers plug in, cool down, and decide whether this whole EV thing is worth the hassle—will determine who finishes strong. Getting those stops right requires more than concrete and conduit. It demands smart contracts that align profit with public purpose. Thanks to this new research, we now have a far clearer map for how to draw them.
Zhang Yiting, Wang Xueqing, Ma Rui, Liu Yunfeng
College of Management and Economics, Tianjin University, Tianjin 300072, China
Wang Dan
School of Public Policy and Administration, Chongqing University, Chongqing 400044, China
Journal of Industrial Engineering and Engineering Management, Vol. 38, No. 5 (2024)
DOI: 10.13587/j.cnki.jieem.2024.05.015