As the automotive industry expands, its energy consumption and greenhouse gas emissions have increasingly threatened environmental sustainability. The global transition from conventional gasoline vehicles to electric vehicles is no longer an option but a necessity for decarbonizing transportation. In this thesis, I focus on two central challenges in the electric vehicle supply chain: how to select an appropriate battery research and development mode, and how to price electric vehicles when consumers exhibit safety anxiety related to battery incidents. Throughout this work, the term “electric vehicle” is used to refer to battery-electric passenger cars that rely on rechargeable power batteries, and the analysis treats the electric vehicle as a competing product against traditional fuel-powered vehicles.
Recent market data from 2024 reveal that while conventional fuel vehicle sales grew only marginally, electric vehicle sales increased by more than 35 percent in the same year. Despite this encouraging trend, battery fire incidents, thermal runaway events, and charging failures have created widespread public concern. A large proportion of consumers now cite battery safety as the primary obstacle to adopting an electric vehicle. This anxiety affects not only consumer purchase behavior but also the strategic choices of manufacturers and suppliers in the electric vehicle ecosystem. To restore consumer confidence, vehicle original equipment manufacturers and battery cell suppliers have invested heavily in technologies that reduce thermal runaway probability and improve battery monitoring. From my point of view, three archetypal R&D models dominate current industrial practice: manufacturer-led R&D, supplier-led R&D, and collaborative R&D between the manufacturer and the supplier. Each mode involves a distinct cost structure, risk profile, and innovation incentive. The first research question in this thesis asks how these R&D modes should be selected based on the R&D capabilities of the two supply-chain members. The second research question turns to the pricing arena, where I model the price competition between an electric vehicle manufacturer and a conventional fuel-vehicle manufacturer in the presence of consumer safety anxiety. I also explore whether the introduction of a safety early-warning system can improve the electric vehicle manufacturer’s position and how the level of such investment affects equilibrium pricing.
1. Literature and Industrial Context
The literature related to electric vehicle adoption has identified several critical drivers of battery-electric vehicle acceptance. Psychological factors, including perceived usefulness and environmental concern, strongly influence consumers’ attitudes. Objective product attributes, such as driving range, charging infrastructure availability, and price, also play central roles. In particular, consumers’ anxiety about battery safety has emerged as a distinct barrier, different from the more widely studied “range anxiety” and “resale anxiety.” While range anxiety arises from the fear of depleted battery charge, safety anxiety arises from the perceived risk of fire, explosion, or thermal runaway. The distinction is important because the remedies are different: range anxiety may be mitigated by more charging stations, while safety anxiety requires the development of safer battery chemistries, robust battery management systems, and proactive warning algorithms.
Prior research on supply-chain R&D decisions in the electric vehicle context mostly centers on government subsidy policies and the dual-credit policy. Some studies show that subsidies foster self-developed technology when government support exceeds a threshold; otherwise, firms prefer technology acquisition or collaboration. Other studies analyze whether manufacturers and suppliers should cooperate in new product development. However, the existing literature rarely connects R&D capability differences to mode selection in the specific context of power batteries for electric vehicles. This thesis fills that gap by building a three-mode game-theoretic model in which the manufacturer and the supplier decide their optimal R&D actions on the basis of their own profit-maximizing objectives.
On the pricing side, many papers examine the effect of green consumer preferences and subsidies on electric vehicle pricing. A smaller subset focuses on anxiety-driven behavior. In this thesis, I characterize price leadership and followership equilibria between fuel and electric vehicle manufacturers by using a two-stage extension game. I further compare equilibrium outcomes before and after the electric vehicle manufacturer invests in a safety early-warning system. In doing so, I offer strategic guidance for electric vehicle companies facing heterogeneous consumer safety perceptions.
2. Methodological Overview and Notations
I use Stackelberg and Bertrand game models throughout the thesis. For the R&D mode selection problem, I model a dyadic supply chain comprising one electric vehicle manufacturer and one battery supplier. The manufacturer assembles the electric vehicle, while the supplier produces the battery pack. To isolate the safety-improving effect of R&D, I let the market demand for the electric vehicle increase with R&D effort because safer and more reliable batteries increase consumer valuation. Table 1 lists the main symbols used in the R&D model. I intentionally normalize the non-battery production cost of the electric vehicle to zero because power batteries account for more than half of total vehicle manufacturing cost, and this simplification does not qualitatively change the findings.
| Notation | Description |
|---|---|
| \(\theta_M, \theta_S\) | R&D effort chosen by the manufacturer and the supplier, respectively |
| \(\lambda_M, \lambda_S\) | R&D capability of the manufacturer and the supplier |
| \(\mu\) | Fixed cost coefficient of R&D investment |
| \(k\) | Variable cost coefficient related to R&D resource input |
| \(c_B\) | Production cost per battery pack |
| \(p_I,p_O,p_C\) | Retail price of the electric vehicle under manufacturer-led, supplier-led, and cooperative R&D |
| \(w_I,w_O,w_C\) | Wholesale price of the battery pack in the three modes |
| \(\Pi_{M}^j,\Pi_{S}^j\) | Profit of the electric vehicle manufacturer and the supplier under mode \(j\in\{I,O,C\}\) |
R&D development time is stochastic. I define the expected development time as a decreasing function of R&D capability. Higher capability reduces both expected completion time and variable cost. The expected R&D cost of player \(i\in\{M,S\}\) is written as
\[
C_i(\theta_i)=\mu\theta_i^2 + \frac{k\theta_i^2}{\lambda_i},\qquad i\in\{M,S\},
\]
where \(\lambda_i\) is the R&D capability of the corresponding party. The term \(\mu\theta_i^2\) captures the fixed investment cost, and the second term captures resources consumed during the uncertain development period. The three demand specifications are as follows. When the manufacturer leads battery R&D, the electric vehicle demand is
\[
D^I(p^I)=1-p^I+\theta_M^I.
\]
When the supplier leads battery R&D, the demand is
\[
D^O(p^O)=1-p^O+\theta_S^O.
\]
When the two parties cooperate, the demand is
\[
D^C(p^C)=1-p^C+\theta_M^C+\theta_S^C,
\]
because cooperative development typically combines complementary knowledge and yields a larger improvement in battery safety. The two parties play a three-stage game. In stage one they simultaneously choose R&D efforts; in stage two the supplier sets the wholesale price of the battery pack; in stage three the electric vehicle manufacturer chooses the retail price of the electric vehicle. I solve all models by backward induction.
3. Equilibrium Results of Battery R&D Modes
In the manufacturer-led R&D mode, the electric vehicle manufacturer is the only investor in battery technology. The manufacturer buys battery packs from the supplier at wholesale price \(w_I\) and sells the electric vehicle at price \(p_I\). The supplier simply chooses the wholesale price to maximize its profit from battery sales. The manufacturer’s optimal R&D effort increases in its own R&D capability \(\lambda_M\), which is intuitive: a more capable manufacturer can translate R&D investment into effective safety improvements at a lower expected cost. I also find that the retail price of the electric vehicle rises with \(\lambda_M\) because the improved battery technology permits a premium price. Interestingly, the supplier’s profit in the manufacturer-led mode may decrease with \(\lambda_M\) when the manufacturer’s internal capability becomes too high, because the supplier’s bargaining power is weakened and its margin is squeezed. This insight warns battery suppliers not to rely solely on a strong electric vehicle manufacturer for technological innovation.
In the supplier-led R&D mode, the battery supplier independently invests in R&D to improve the safety and reliability of the power battery. The supplier chooses both its R&D effort \(\theta_S\) and the battery wholesale price \(w_O\), while the electric vehicle manufacturer chooses the retail price. The supplier’s optimal effort increases in its R&D capability, and both supply-chain members can benefit from a stronger supplier because a safer battery enlarges total demand for the electric vehicle. The profit of the supplier and the profit of the manufacturer are both increasing in \(\lambda_S\) in this mode. Hence, when the supplier has a marked technological advantage, supplier-led R&D creates a win-win situation.
In the cooperative R&D mode, the manufacturer and the supplier simultaneously decide their respective R&D efforts, and both share the total R&D cost. Cooperative R&D has two opposing effects. On one hand, it combines technological resources and diffuses innovation across the supply chain, which is beneficial when neither party is individually strong. On the other hand, it may create coordination costs and reduce each party’s incentive to exert effort. In my equilibrium analysis, I compare the three modes under different capability combinations and derive the conditions under which each supply-chain member strictly prefers one mode. To keep the exposition concise, I summarize the key results in the following propositions.
Proposition 1. From the electric vehicle manufacturer’s perspective, when the supplier’s R&D capability is weak, the manufacturer prefers cooperative R&D only when its own capability is also weak. Once the manufacturer’s own capability exceeds a threshold, it prefers manufacturer-led R&D. When the supplier’s capability is strong, the manufacturer’s optimal choice depends on its own capability: it prefers cooperation if its own capability is weak, prefers supplier-led R&D if its capability is moderate, and prefers manufacturer-led R&D only if its own capability is sufficiently high.
Proposition 2. From the battery supplier’s perspective, cooperation is attractive when both parties have low R&D capability. If the manufacturer’s R&D capability is very low, the supplier prefers cooperative R&D at low \(\lambda_S\), but shifts to supplier-led R&D when \(\lambda_S\) is high. If the manufacturer has moderate R&D capability, the supplier may prefer to let the manufacturer lead R&D rather than to invest itself. Such a situation arises when the supplier’s own capability is moderate; the supplier avoids heavy R&D costs and concentrates on battery production while benefiting from the manufacturer’s innovation.
Proposition 3. By comparing the two members’ preferred R&D modes simultaneously, I obtain conditions under which the manufacturer and the supplier attain consensus. First, when the supplier’s R&D capability is low and the manufacturer’s R&D capability is high, both members prefer manufacturer-led R&D. Second, when the supplier is very capable and the manufacturer is also sufficiently capable, both prefer manufacturer-led R&D because the manufacturer can use its superior capability to internalize innovation and the supplier can profit from wholesale orders without bearing R&D risk. Third, when the supplier’s capability is sufficiently high and the manufacturer’s capability is either weak or moderate, both may prefer supplier-led R&D. Finally, only when both R&D capabilities are low do the manufacturer and supplier agree on cooperative R&D. In that scenario, collaboration is a strategic necessity rather than a choice made from strength.
The R&D capability thresholds have explicit but lengthy expressions in the thesis. The central managerial lesson is that R&D mode selection is not universal; it must be tailored to the relative technological strengths of the supply-chain partners. A firm that blindly follows an “independent innovation” policy may earn lower profit than a firm that selects a supplier-led or cooperative strategy when its R&D capability is limited.
4. Numerical Analysis of the R&D Mode Model
I calibrate the R&D model with parameters from public industry data. In the baseline setting I use \(c_B=0.7\), \(\mu=0.035\), \(k=0.03\), \(\lambda_M=0.7\), and \(\lambda_S=0.2\). These values are consistent with the ratio of battery cost to the average retail price of an electric vehicle and with typical R&D expenditure intensities reported in the Chinese automotive industry. Table 2 reports how the manufacturer’s R&D capability affects the profits of both parties in the manufacturer-led and cooperative modes.
| \(\lambda_M\) | \(\Pi_{MI}\) | \(\Pi_{SI}\) | \(\Pi_{MC}\) | \(\Pi_{SC}\) |
|---|---|---|---|---|
| 0.50 | 0.0164 | 0.0961 | 0.0173 | 0.0328 |
| 0.60 | 0.0213 | 0.1606 | 0.0088 | 0.0216 |
| 0.70 | 0.0285 | 0.2892 | 0.0048 | 0.0159 |
| 0.80 | 0.0408 | 0.5913 | 0.0027 | 0.0126 |
| 0.90 | 0.0659 | 1.5438 | 0.0014 | 0.0105 |
Table 2 confirms that profits in the manufacturer-led mode increase with \(\lambda_M\), whereas profits in the cooperative mode decrease as \(\lambda_M\) increases. The reason is that a high-capability manufacturer contributes a great deal of effort in the cooperative mode and receives only a share of the joint reward, which lowers its net incentive. A low-capability manufacturer, by contrast, benefits from the supplier’s complementary knowledge and therefore prefers cooperation. These numerical results support all the analytical findings.
| \(\lambda_S\) | \(\Pi_{MO}\) | \(\Pi_{SO}\) | \(\Pi_{MC}\) | \(\Pi_{SC}\) |
|---|---|---|---|---|
| 0.16 | 0.0293 | 0.0257 | 0.0083 | 0.0371 |
| 0.20 | 0.0535 | 0.0347 | 0.0048 | 0.0159 |
| 0.24 | 0.1176 | 0.0514 | 0.0033 | 0.0072 |
| 0.28 | 0.3867 | 0.0933 | 0.0024 | 0.0029 |
| 0.30 | 1.0252 | 0.1519 | 0.0021 | 0.0016 |
The second numerical table demonstrates that supplier-led R&D becomes increasingly profitable for both the electric vehicle manufacturer and the supplier as \(\lambda_S\) rises. In contrast, cooperative R&D loses its appeal when the supplier is already highly capable. This result has a direct practical implication for battery suppliers: if the supplier dominates battery technology, establishing an independent R&D roadmap is preferable to a collaborative arrangement that may weaken the supplier’s own proprietary position.
I also extend the model by allowing the consumer preference for the independent R&D output to be less than one. In this extension, the demand coefficients of manufacturer-led and supplier-led R&D are \(\alpha\) and \(\beta\), respectively, both between zero and one, while cooperative R&D still has the strongest demand-enhancing effect. I calibrate the baseline \(\alpha=\beta=0.5\) and rerun the numerical analyses. The comparison reveals that although \(\alpha\) and \(\beta\) shift the numerical values of equilibrium decisions, they do not reverse the qualitative ranking of the three R&D modes. Therefore, my main managerial conclusions remain robust to the assumption that cooperative R&D generates the greatest improvement in consumer-perceived electric vehicle safety.
5. Price Competition with Consumer Safety Anxiety
The second part of this thesis moves from the supply chain’s upstream R&D decisions to the downstream market competition between an electric vehicle manufacturer and a conventional fuel-vehicle manufacturer. Consumers may hesitate to purchase an electric vehicle because of battery safety incidents. I model this hesitation through a consumer safety trust parameter \(\theta\in(0,1)\). A lower \(\theta\) corresponds to stronger safety anxiety; when \(\theta=1\), the electric vehicle is perceived to be as safe as a conventional gasoline vehicle in terms of driving utility. Let \(u_e\) represent the baseline driving utility that a consumer derives from using a vehicle. Let \(u_g\) be the consumer’s green preference, which is uniformly distributed on \([0,1]\). Then the utilities from buying an electric vehicle and a fuel vehicle are formulated as
\[
U_{EV}=\theta u_e – p_c + u_g,\qquad
U_{ICE}=u_e – p_o.
\]
Consumers choose the electric vehicle if \(U_{EV}\ge U_{ICE}\). This gives the demand functions
\[
q_c=1-(1-\theta)u_e – p_c + p_o,
\]
\[
q_o=(1-\theta)u_e – p_o + p_c,
\]
where both demands are positive in the interior equilibrium. The profit functions are
\[
\Pi_c=(p_c-c_c)q_c,\qquad \Pi_o=p_o q_o,
\]
where \(c_c\) is the unit production cost of the electric vehicle and the fuel vehicle’s production cost is normalized to zero. I solve three price-competition structures: (i) simultaneous moves, (ii) a sequential game in which the electric vehicle manufacturer is the price leader, and (iii) a sequential game in which the fuel-vehicle manufacturer is the price leader. The equilibrium prices for each structure are summarized in Table 3, with \(a=(1-\theta)u_e\) for notational convenience.
| Game structure | Electric vehicle price | Fuel-vehicle price |
|---|---|---|
| Simultaneous | \(\displaystyle p_c^S=\frac{2-a+2c_c}{3}\) | \(\displaystyle p_o^S=\frac{1+a+c_c}{3}\) |
| Electric vehicle leader | \(\displaystyle p_c^L=\frac{2-a+c_c}{2}\) | \(\displaystyle p_o^F=\frac{2+a+c_c}{4}\) |
| Fuel-vehicle leader | \(\displaystyle p_c^F=\frac{3-a+3c_c}{4}\) | \(\displaystyle p_o^L=\frac{1+a+c_c}{2}\) |
The corresponding equilibrium profits are obtained by substituting these prices into the profit functions. A key finding from the simultaneous-move game is that the fuel-vehicle market demand exceeds the electric vehicle demand whenever consumers have non-negligible safety anxiety. Because the electric vehicle is perceived as less safe, only green-minded consumers with a sufficiently high \(\theta\) will accept it. In contrast, in the sequential games, the electric vehicle can outsell the fuel vehicle when \(\theta\) is high enough. For instance, when the electric vehicle manufacturer leads the market, the demand of the electric vehicle can be greater than that of the fuel vehicle if consumer safety trust is above a certain threshold. This outcome occurs because the electric vehicle leader sets a sufficiently low margin to attract safety-conscious consumers, while the follower fuel-vehicle manufacturer responds by lowering its price, which partially dampens its profit margin.
I then compare the three game structures using a two-stage extended game in which each manufacturer first declares its intention to be a leader or a follower. The equilibrium leadership choice is summarized in Proposition 4.
Proposition 4. Consider the market with consumer safety anxiety. If the consumer safety trust parameter \(\theta\) lies in a high interval, both the electric vehicle manufacturer and the fuel-vehicle manufacturer are indifferent between being the price leader and being the price follower. In this case, either the \((L,F)\) or the \((F,L)\) equilibrium can emerge. When \(\theta\) decreases to an intermediate interval, the fuel-vehicle manufacturer prefers to lead, while the electric vehicle manufacturer prefers to follow. The consequent \((F,L)\) equilibrium means that the electric vehicle adopts a price-following strategy. This pattern is consistent with observed market dynamics: traditional fuel vehicle giants often price first and electric vehicle entrants undercut them to attract consumers. Only when the production cost of the electric vehicle is sufficiently high and consumer safety trust is low does the electric vehicle manufacturer choose to lead. In that case, the electric vehicle manufacturer relies on a high-price strategy to maintain brand positioning rather than on price discounting.
The managerial intuition is straightforward. In the early years of electric vehicle diffusion, high battery costs pushed electric vehicle prices above fuel-vehicle prices. Later, battery price reductions and increased competition triggered aggressive price cuts by electric vehicle manufacturers. Safety anxiety, however, may have made price cuts less effective because consumers are unwilling to purchase an unsafe product even at a lower price. Therefore, a price-following strategy is not always optimal. The electric vehicle manufacturer should change its role only when its own safety reputation and cost structure justify such a move.
6. Extension with Safety Early-Warning Systems
To actively address safety anxiety, many electric vehicle manufacturers have begun to install safety early-warning systems that monitor battery voltage, temperature, and other real-time signals. These systems use artificial intelligence algorithms to identify abnormal battery states before fires or thermal runaway events occur. In the model, I denote by \(T\) the investment level of such a safety early-warning system. The consumer utility of buying an electric vehicle becomes
\[
U_{EV}=\theta u_e – p_c + u_g + K T,
\]
where \(K\) measures the effectiveness of the safety warning system in restoring consumer confidence. Investing in the system also incurs a convex cost \(\frac{1}{2}\beta T^2\). The presence of \(K T\) enlarges the electric vehicle demand by an amount \(K T\) and reduces the fuel-vehicle demand by the same amount in the full-market model. I solve the simultaneous and two sequential price games under this safety-warning extension. The equilibrium profit expressions include an additional term \(K T\) in the demand intercept, and each manufacturer’s profit is adjusted by the investment cost.
Proposition 5. When the safety warning system investment level is high, both the electric vehicle manufacturer and the fuel-vehicle manufacturer may be willing to act as either price leaders or price followers. A high investment level effectively offsets consumer safety anxiety; therefore, the electric vehicle manufacturer can compete on an equal footing with fuel vehicles. When consumer safety trust is moderate and the safety warning investment is at an intermediate level, the electric vehicle manufacturer chooses the price-following equilibrium. By contrast, when the warning system investment is too low, it cannot eliminate safety anxiety, and the electric vehicle manufacturer may choose to be a price leader despite its disadvantage, because defensive price cutting would erode the limited margin needed to sustain the R&D investment.
The comparison of the safety warning model with the baseline model reveals a critical threshold behavior. \(T\) influences the market demand for the electric vehicle as follows. If \(T\) exceeds a threshold \(2-(1-\theta)u_e-c_c\), the positive safety signal dominates and consumers favor the electric vehicle. If \(T\) is below that threshold, the fuel-vehicle demand remains larger than the electric vehicle demand. Therefore, introducing a safety warning system is not automatically beneficial. Only when the investment is sufficiently high can it credibly signal product safety to consumers. A low-level or token system may be perceived as greenwashing and can even backfire.
| Investment level \(T\) | Electric vehicle profit | Fuel-vehicle profit | Market demand comparison |
|---|---|---|---|
| High \(T\) | High margin and high volume | Lower volume | Electric vehicle may dominate |
| Moderate \(T\) | Moderate margin but price pressure | High volume | Fuel vehicle dominates |
| Low \(T\) | Low margin and low volume | Stable volume | Fuel vehicle clearly dominates |
Table 4 provides a conceptual summary of how the safety-warning investment changes profitability and market demand. The equilibrium threshold expressions depend on the production cost of the electric vehicle, the baseline driving utility, and the safety trust parameter \(\theta\). In the numerical example, I calibrate \(c_c=0.2\), \(u_e=3.8\), \(\theta=0.7\), \(K=0.1\), and \(\beta=0.9\). The numerical sensitivity analysis shows that as \(\theta\) increases from 0.45 to 1, the equilibrium electric vehicle price rises by about 0.4 to 0.5 units and the electric vehicle manufacturer’s profit increases by several hundred percent. The fuel-vehicle price and profit decline monotonically with \(\theta\). This result explains why electric vehicle makers spend enormous resources on battery safety campaigns and early-warning algorithm improvement: a one-unit improvement in consumer safety trust substantially shifts market power from traditional automakers to electric vehicle makers.
| \(\theta\) | \(p_c^S\) | \(p_o^S\) | \(\Pi_c^S\) | \(\Pi_o^S\) |
|---|---|---|---|---|
| 0.45 | 0.1033 | 1.0966 | 0.0093 | 1.2026 |
| 0.60 | 0.2933 | 0.9066 | 0.0087 | 0.8220 |
| 0.75 | 0.4833 | 0.7166 | 0.0802 | 0.5136 |
| 0.90 | 0.6733 | 0.5266 | 0.2240 | 0.2773 |
| 1.00 | 0.8000 | 0.4000 | 0.3600 | 0.1600 |
Table 5 displays the numerical profit results for the simultaneous-move game. When \(\theta\) rises from 0.45 to 1, the electric vehicle manufacturer’s profit rises almost fortyfold, while the fuel-vehicle manufacturer’s profit falls by more than 86 percent. This powerful sensitivity implies that public perceptions of electric vehicle safety are not merely a PR issue; they have first-order effects on the profitability and sustainability of the entire electric vehicle supply chain.
Another important finding is obtained by comparing the leader and follower profits. I compute the equilibrium profit matrices and derive conditions under which \((L,F)\), \((F,L)\), or \((L,L)\) emerges. The extended game has no pure-strategy \((F,F)\) equilibrium equilibrium because both manufacturers have an incentive to deviate if their opponent also chooses to follow. The \((L,L)\) equilibrium corresponds to the simultaneous price game, and it is more likely when the electric vehicle production cost is high and the safety trust parameter is low. This aligns with the current Chinese market: in the high-end segment, electric vehicle makers maintain high price points and fuel-vehicle makers do not always undercut them. In the mass market, however, electric vehicle price cuts force fuel vehicle makers to react, but these fuel vehicle makers sometimes increase prices to protect brand equity. My model explains this seemingly counterintuitive price hike: when consumers of fuel vehicles are highly loyal to the traditional driving experience and are less swayed by electric vehicle discounts, the fuel-vehicle manufacturer can exploit its price-leader position and increase its margin.
7. Overall Managerial Implications
This thesis yields several strategic recommendations for managers in the electric vehicle industry. First, battery R&D decisions should be based on a rigorous assessment of both the manufacturer’s and the supplier’s R&D capabilities. If the electric vehicle manufacturer has a dominant capability advantage, it should not outsource battery innovation because the supplier may fail to meet the required safety specifications and the manufacturer may lose its core technological differentiation. If the supplier has a dominant capability advantage, the manufacturer can save R&D resources by delegating battery innovation to the supplier, provided that the wholesale price contract aligns their incentives. Cooperative R&D should be reserved for situations where both parties have comparably low capabilities, so that joint investment can overcome their individual limitations.
Second, safety anxiety cannot be overcome by reducing price alone. A lower price can enlarge demand only if the target consumer segment has already accepted the battery safety risk. For electric vehicle manufacturers, investing in observable safety attributes such as state-of-the-art battery management systems, thermal spread prevention, and real-time early-warning platforms is a more effective demand booster than unilateral price discounting. The safety early-warning system should be sufficiently robust; otherwise consumers may ignore it or remain skeptical.
Third, the price-leadership role of the electric vehicle manufacturer should be dynamically adjusted according to market conditions. In the early adoption stage, electric vehicle manufacturers can act as price leaders to signal premium technology, particularly in the high-end segment. In the growth stage, when consumer safety trust is moderate and price elasticity is high, the electric vehicle manufacturer should follow the fuel-vehicle manufacturer’s price and exploit the follower’s information advantage. In the mass-adoption stage, if battery safety is fully proven and consumer anxiety is low, the electric vehicle manufacturer can again become a leader if its cost structure permits.
Fourth, conventional fuel-vehicle manufacturers should not ignore the safety anxiety channel. When an electric vehicle rival introduces a credible safety warning system, fuel-vehicle manufacturers will experience a downward shift in demand. They may respond by emphasizing their own safety records and their mature fuel infrastructure, or they may enter the electric vehicle market themselves through hybrid models. The model suggests that taking a price-leadership position can be profitable when consumer safety trust is moderate, because fuel vehicles retain a perceived safety advantage that justifies a premium.
8. Robustness Extensions and Discussion
I examine the robustness of the R&D-mode results by considering that independent manufacturer R&D and supplier R&D may be less effective than cooperative R&D in improving consumer-perceived electric vehicle quality. I introduce parameters \(\alpha\) and \(\beta\) representing the relative consumer preferences for the independent R&D outcomes. As long as \(\alpha\) and \(\beta\) are between zero and one, cooperative R&D gives the largest positive demand shift. My numerical analysis using \(\alpha=\beta=0.5\) shows that the qualitative ordering of the three modes does not change. The precise value of \(\alpha\) and \(\beta\) only alters the cutoff points at which the manufacturer or supplier switches strategies. Thus, the central conclusion that cooperation is optimal only when both parties have low internal capabilities is robust.
I also discuss the possibility that the electric vehicle supply chain may involve more than one battery supplier. However, the bilateral monopoly assumption captures the fundamental trade-off between vertical integration and innovation outsourcing without the extra complexity of head-to-head supplier competition. Extending the model to multiple suppliers would introduce horizontal spillovers but would not overturn the basic insight that R&D capability asymmetry shapes governance choices.
The pricing model assumes that consumers are heterogeneous only in green preference, while safety anxiety is represented by the same trust parameter for all consumers. In reality, consumers differ in their risk perceptions, family income, and charging accessibility. My focus on aggregate demand functions is consistent with the quality-differentiated Bertrand literature and allows explicit equilibrium comparisons. Future research can incorporate individual-level safety anxiety distributions, test how publicized safety incidents shift \(\theta\), and estimate the value of \(\theta\) from real market data.
9. Conclusions and Future Directions
This thesis investigates the battery R&D mode selection and the pricing strategy of electric vehicles from the perspectives of safety improvement and consumer safety anxiety. The findings can be summarized as follows.
First, when the R&D capabilities of the electric vehicle manufacturer and its battery supplier are both high, independent R&D is preferred by both sides. When both capabilities are low, cooperative R&D dominates. There are also asymmetric consensus zones. If the supplier is weak and the manufacturer is strong, both parties prefer manufacturer-led R&D. Conversely, if the supplier is strong and the manufacturer is not strong enough to lead independently, both parties prefer supplier-led R&D. These results have important implications for the architecture of the electric vehicle supply chain. A manufacturer that lacks advanced battery knowledge but has strong brand and assembly skills should rely on a high-capable battery supplier, rather than attempting to “do everything in-house.” On the other hand, core electric vehicle firms such as those that treat battery technology as a key strategic asset should continue to internalize battery R&D even if supplier-led solutions appear cheaper in the short run.
Second, from the pricing model, consumer safety anxiety determines whether the electric vehicle manufacturer should be a price leader or a price follower. With high consumer safety trust, multiple pricing equilibria are possible and the manufacturer can be flexible. With moderate or low safety trust, the electric vehicle manufacturer often prefers to follow the fuel-vehicle manufacturer’s price. An effective safety early-warning system can alter these equilibria by restoring consumer confidence. The positive effect of the warning system is non-monotone: low levels are useless, high levels are powerful, and moderate levels trigger price-following behavior because they increase demand only partially. This explains why electric vehicle manufacturers are investing heavily in artificial-intelligence-based battery monitoring systems rather than merely installing simple alarms.
Third, when safety anxiety is present, the fuel-vehicle market demand tends to be larger than the electric vehicle market demand if both manufacturers set prices simultaneously. In sequential leadership structures, however, the electric vehicle demand can exceed the fuel-vehicle demand if consumer safety trust is sufficiently high. This finding highlights the importance of order of moves in an emerging market. Electric vehicle manufacturers should carefully observe whether their products have crossed the consumer safety threshold before deciding to take price leadership. Launching a price war too early may compress margins and slow future R&D investment in battery safety.
From a policy standpoint, this thesis suggests that governments should support not only battery production capacity but also credible safety-labeling and third-party testing schemes. Public-transparency initiatives can improve consumer safety trust \(\theta\), which the model shows has a large positive effect on electric vehicle market profitability. Subsidies that merely reduce prices are less potent than policies that assist manufacturers in building reliable early-warning systems and certified safety standards.
Future research can extend this work in several directions. The R&D model can be generalized to include a manufacturer choosing between multiple battery suppliers with different capabilities. The pricing model can incorporate supply chain coordination contracts between electric vehicle manufacturers and their battery suppliers after the retail pricing stage. Empirical studies can measure the consumer safety trust parameter and identify how safety incidents change its magnitude. Another promising avenue is to study dynamic R&D decisions over time, because battery technology improves with learning, and consumer safety anxiety may decline cumulatively as real-world accident rates fall. I believe that this thesis provides a useful analytical foundation for these future studies and offers practical guidance for firms operating in the fast-growing global electric vehicle market.