JOINT IMPACT OF SERVICE EFFICIENCY AND SALVAGE VALUE ON THE MANUFACTURER’S SHARED VEHICLE-TYPE STRATEGIES

. With the rapid development of the sharing economy, many traditional automobile manufacturers have been choosing to provide the car sharing service. Some manufacturers share GVs, while others introduce EVs in the sharing market. We develop a model that a monopoly manufacturer who simultaneously sells GVs and EVs and discuss which type of vehicles should the manufacturer launch in the sharing market considering the service efficiency and the salvage value. Our findings are that no matter which type of vehicles the manufacturer shares, EV sales remain the same, but GV sales are reduced. This means that the manufacturer’s EV-sharing strategy always promotes EVs’ adoption. It is found that when both the service efficiency ratio of EV to GV and the salvage value gap between them are low or high, the manufacturer launches EVs; otherwise, the manufacturer launches GVs. We also find that the equilibrium vehicle-type strategy can maximize the manufacturer’s profit while being the most environmentally friendly only if the valuation of shared product is high. Through numerical analysis, we know that, although the manufacturer’s GV-sharing strategy worsens the environment, it always improves the social welfare. Notably, the manufacturer’s EV-sharing strategy is not always beneficial for the environment, especially if the service efficiency ratio is relatively high. Similarly, the manufacturer’s EV sharing does not always improve the social welfare, especially if the service efficiency ratio is in the middle range. The findings not only contribute to guiding the manufacturer’s vehicle-type strategies for car sharing, but also providing potential policy implications for the government’s effort in promoting EVs’ adoption.


Introduction
Carbon emissions from human activities have create a significant burden on the environment, which leads the global warming and the intensification of haze weather.Transportation is the second largest carbon sector in the world, which is responsible for 24% of direct carbon emissions from fuel combustion [13].Many countries have advocated and taken action to reduce carbon emissions.For instance, China vigorously develops and promotes new energy vehicles (EVs) to peak carbon dioxide emissions in 2030 and become carbon neutral by 2060.As a low-carbon and green travel tool, EV is considered as one of the effective means to reduce 40% carbon emissions compared to the gasoline vehicle (GV).Many countries have formulated corresponding incentive policies to promote EVs [19,28].Accordingly, automakers have begun to produce and sell EVs on a large scale in this trend to respond to the national advocacy and consumers' need for low-carbon travel.According to Pontes [26], the sales of EVs in 2021 have reached 6.5 million units, a substantial increase of 108% compared with the sales in 2020.
Although the sales of EVs have risen rapidly, there is still a big gap with GVs.As one of the most developed countries in the EV's industry, China's sales reached 3.521 million units in 2021, while the sales of GVs reached 26.275 million units [6].There are some problems that hinder the development of EV, such as a relatively high price, a limited driving range and a lack of charging stations [1].Although the above problems are alleviated with the innovation of the charging battery technology, a prominent problem arises from the rapid technological iteration of EVs: the low salvage value.According to AutoHome [3], statistics show that the sales price of Toyota C-HR is 242 266 and that of Toyota C-HR EV is 183 442.The salvage value of both Toyota EVs are nearly 90 000 after three years.It is easy to know that the value of EV is falling faster than that of GV.The low salvage value of EV has been an important factor restricting the development of EV [23].
In addition to selling EVs, with the development of the sharing economy, automakers have attached importance to car sharing and attempt to promote EVs through car sharing.However, every coin has two sides, and automakers' promoting EVs is no exception.On one side, compared with the slow adoption of EVs in the sales market, the sharing market seems to be more effective in promoting EVs.Firstly, consumers can use EVs without bearing a high economic burden.Since the car sharing gives consumers the right to use instead of ownership, they do not have to worry about the high sales price when purchase EVs and the low salvage value when resale EVs.Then, consumers can freely choose the type of shared vehicles according to their travel distance (a GV for longer trips and an EV for shorter ones).Therefore, consumers do not need to worry about the limited driving range of EVs and the lack of fast charging stations.It seems that the car sharing is the best way to promote EVs.As a result, many automakers have been launching EVs in the sharing market.For instance, Daimler's Car2Go offered two-seat EVs called "Smart" [33].BMW offered EVs in the car-sharing program in the San Francisco in August 2012 [1].Shanghai Automotive Industry Corporation built an electrical car sharing platform EVCARD in Shanghai in 2013 [8].On the other side, although the car sharing can solve the burden and anxiety that EVs bring to consumers in the sales market, there are still some issues hindering automakers from launching EVs in the sharing market.For instance, the high cost and the low salvage value put a heavier burden on automakers who launch EVs than those who launch GVs.Moreover, EVs take a long time to charge and serve fewer people than GVs at the same period.So, the sharing service efficiency of EVs is lower than that of GVs.In practice, some automakers have replaced EVs by GVs.Car2go replaced EVs with GVs in California in May 2016 [1].It is a symbolic stepback for the adoption of EVs.
In retrospect, above two sides show that there is ambiguous for automakers to launch EVs or GVs in the sharing market.To respond this, Bellos et al. [4] consider the joint impact of fuel efficiency and driving performance to study the manufacturer's vehicle-type strategy, while only focusing on GVs with different performance and ignoring EVs.Abouee-Mehrizi et al. [1] study the vehicle-type strategy of the car-sharing company rather than that of the manufacturer.Liu et al. [21] examine the effect of consumers' behavior and the subsidy on the manufacturer's vehicle-type strategy.However, there is few scholars to analyze the impact of the vehicle salvage value and the sharing service efficiency on the manufacturer's vehicle-type strategy in the sharing market.To fill this research gap, we propose the following questions: (1) Which type of vehicles should the manufacturer launch in the sharing market?Can car sharing promote the adoption of EVs? (2) How the salvage value and the service efficiency affect the manufacturer's vehicle-type strategy in the sharing market?(3) Whether EVs' adoption in the sharing market is environmentally friendly and increases social welfare?
To answer the aforementioned research questions, we develop a game-theoretic model that a monopoly manufacturer who simultaneously sells GVs and EVs.It is optional for the manufacturer to enter the sharing market.
If the manufacturer enters the sharing market, it needs to choose which type of vehicles to launch.We consider three scenarios: (1) Scenario N, in which the manufacturer does not provide the sharing service; (2) Scenario G, in which the manufacturer shares GVs; (3) Scenario E, in which the manufacturer shares EVs.Based on the above considerations, this paper contributes to investigating which type of vehicles should the manufacturer launch in the sharing market to respond to the service efficiency and the salvage value.Employing a game-theoretic approach, we find the following results.
Firstly, we analyze the conditions for the manufacturer providing the sharing service.The manufacturer can sell EVs and GVs simultaneously when both the cost gap and the salvage value gap between them are small.When consumers have a low valuation on the shared vehicles and the sharing service efficiency is not too low, the manufacturer provides the car sharing.There impose stricter constraints on consumers' valuation of shared vehicles and the service efficiency if the manufacturer shares EVs.Secondly, we find that sales prices of both EV and GV increase.No matter which type of vehicles the manufacturer shares, EV sales remain the same, but GV sales are reduced.As the sharing market cannibalizes the sales market demand, the profit of the sales market decreases, but the manufacturer' total profits increase.Thirdly, we discuss the manufacturer's equilibrium strategy considering the service efficiency and the salvage value.When both the service efficiency ratio of EV to GV and the salvage value gap are low or high, the manufacturer should launch EVs.When the service efficiency ratio is low and the salvage value gap is high, the manufacturer should launch GVs.The same happens when the service efficiency ratio is high and the salvage value gap is low.When the valuation discount of shared product is low or high, the manufacturer launches EVs; otherwise, the manufacturer launches GVs.When consumers' low-carbon awareness is low, the manufacturer launches GVs; otherwise, the manufacturer launches EVs to maximize profits.Finally, we find that the equilibrium vehicle-type strategy can maximize the manufacturer's profit while being the most environmentally friendly only if the valuation of shared product is high.Through numerical analysis, we know that the manufacturer's EV-sharing strategy is not always beneficial to the environment.When the service efficiency ratio is relatively high, the environment deterioration impact is the largest when the manufacturer launches EVs.The manufacturer's GV-sharing strategy always improves the social welfare.When the service efficiency ratio is in the middle range, the social welfare when the manufacturer shares EVs is lower than that without the sharing service.
The rest of this paper is organized as follows.In Section 2, we review the related literature.In Section 3, we present the models in three scenarios.In Section 4, we compare three scenarios.In Section 5, we discuss the manufacturer's vehicle-type strategies.In Section 6, we analyze the impact of manufacturer's vehicle-type strategies on the environment and the social welfare.In Section 7, we provide an extension of our model where the manufacturer simultaneously shares two types of vehicles.Finally, in Section 8, we present conclusions, managemental implications, and suggest future research directions.

Literature review
Our paper contributes to two different streams of literature focusing on vehicle-type comparison and car sharing, respectively.The relevant researches are reviewed below to clarify our contribution.

Vehicle-type comparison
EV is widely used in the transportation industry, which gradually replaces GV.Some scholars compared the two kinds of vehicles based on their performance.Shi et al. [29] and Yang et al. [31] found that EV could reduce carbon emissions but increased other harmful gases emissions compared to GV. Yu et al. [32] found that harmful gases emissions from EVs were significantly lower than those from GVs with respect to Chinese energy-saving policies and actual emission reduction techniques.Other scholars discussed the economics of EV and GV for consumers.Parker et al. [25] compared EV's and GV's total cost of ownership consisting of capital expenditures, operating costs, and salvage value.They found that over 17% of consumers could save money by purchasing an EV.Suttakul et al. [30] found that EV's total cost of ownership was lower than GV's and the salvage value played a significant role in EV costs.Chen et al. [7] studied the impacts of the subsidy on consumers' vehicle-type choice and found that when subsidy was sufficiently high, rich consumers chose high-end EVs, poor consumers purchased GVs, and medium income consumers used low-end EVs.Zhang and Huang [34] examined impacts of both the consumer subsidy and the R&D subsidy on the manufacturer's vehicle-type choice strategies.They found that when the technology capacity for improving EV energy-saving level was sufficiently large, the manufacturer produced GVs and EVs simultaneously.When the technology capacity was extremely high, it produced EVs only.However, all the above papers discuss the vehicle-type comparison and how the manufacturer or consumers to choose vehicle-type strategies in the sales market rather than the sharing market.

Car sharing
Car sharing, as an important part of the sharing mobility, has been studied by many scholars.Some scholars studied its impact on the automotive industry [10,24,27].They all found that car sharing reduced the sales of vehicles.Then, other scholars examined whether the manufacturer provided the sharing service.Ke et al. [17] indicated that the manufacturer who provided the sharing service always benefited from car sharing.Yu et al. [33] studied whether the manufacturer provided the sharing service considering the vertical difference of products and found that when the vertical difference fell into an interval, the manufacturer who introduced carsharing programs was more profitable.Based on Yu et al. [33], Jia et al. [15] studied how service level affected the manufacturer's optimal decision and modes selection, and found that the manufacturer should not provide the sharing service when service level was less than a critical value.Ke et al. [16] found that the manufacturer could obtain more benefits from car sharing if the market size was sufficiently large and generated a larger profit with producing less vehicles.Agrawal and Bellos [2] investigated the impacts of the sharing mode on environment and found that the manufacturer who provides the sharing service was more profitable, as well as the better environment, under strong pooling effect.Mao et al. [22] discussed how the manufacturer's operating model changed under the effect of the COVID-19 pandemic.Besides, scholars studied the manufacturer's strategy of sharing modes.Li et al. [18] found that the value perception factor and the marginal cost were two critical factors for the manufacturer to choose to cooperate with the business-to-consumer (B2C) platform or the C2C platform.Based on Li et al. [18], Zhang et al. [35] studied the manufacturer's strategy on choosing a self-built platform or a third-party platform when the manufacturer provided the sharing service.Moreover, Bellos et al. [4] studied the shared vehicle-type strategy by accounting for the trade-off between fuel efficiency and driving performance and concluded that the manufacturer should launch more fuel-efficient low-end cars in the sharing market.In contrast, in our study, we investigate the manufacturer's shared vehicle-type strategy considering different types of vehicles with the same performance.
Our paper has distinct difference from related literature.Although many scholars have studied vehicle-type comparison and car sharing in various views, there are few literatures focusing the manufacturer's vehicle-type strategy in the sharing market.Our study is most closely related to Abouee-Mehrizi et al. [1] and Liu et al. [21].Abouee-Mehrizi et al. [1] studied whether carsharing company should launch EVs in the car sharing market and showed that only if the charging speed was fast enough, the number of charging stations and the range of EVs were sufficient, it was optimal to launch EVs.Different from Abouee-Mehrizi et al. [1], we study the vehicle-type strategies of the manufacturer for car sharing rather than that of car-sharing company.Liu et al. [21] examined the impacts of consumers' behavior and the subsidy on the manufacturer's vehicle-type strategy, while we focus on the service efficiency and the salvage value.In conclusion, although there are related research focusing on the vehicle-type strategy for car sharing, the model in our study is the first attempt to study the manufacturer's shared vehicle-type strategy considering the service efficiency and the salvage value.Therefore, this study adds theoretical knowledge to the manufacturer's vehicle-type strategy for car sharing and attempts to fill a gap in the literature.The differences between our study and related works are summarized in Table 1.

Model
We consider there is a monopoly manufacturer who simultaneously produces and sells GVs and EVs.The manufacturer produces vehicle  with marginal cost   and sells vehicle  to consumers at price   ( ∈ {, }, where  represents the GV and  represents the EV.To meet consumers' increasing demand for shared travel, the manufacturer considers whether to provide the sharing service and which type of vehicles to share.If the manufacturer provides the sharing service, we assume the manufacturer rents vehicle  out at price   .

Consumer utility
When consumers have use needs, they choose to purchase or rent to meet their needs.We assume that consumers' use need, which represents the frequency of using car by consumers.We follow Bellos et al. [4] and Li et al. [18] and adopt the parameter (0 <  ≤ 1) to characterize the frequency.Consumers decide whether to buy a GV or EV according to the net utility   ( ∈ {1, 2}, where 1 represents the sales market, 2 represents the sharing market).Consumers are heterogeneous in their valuation  of the vehicles, where  is uniformly distributed  [0, 1].Since EV can reduce carbon emissions, consumers can get an extra value , which (0 <  ≤ 1) is low-carbon awareness.Thus, consumers who purchase a GV and consumers who purchase an EV get a usage value  and (1 + ), respectively.The net utility of consumers who purchase a GV and consumers who purchase are is the salvage value of vehicle , which is obtained by selling it in a secondary used goods market.According to McKinsey [23], in addition to inadequate charging facilities and endurance concerns, the salvage value affects consumers' willingness to pay.When consumers purchase the vehicle , they not only obtain the whole vehicle's usage value, but also own its salvage value   .Consumers who purchase different types of vehicles derive corresponding salvage value.The battery is a critical component of EV, but the battery's fast decay and rapid technological iteration make the salvage value of EV be lower than GV (  >   ) [5,14,20].We assume that the salvage value   is observable.
If consumers rent the manufacturer's vehicle, considering the psychological gap caused by non-ownership of products, consumers' usage valuation of renting vehicles is lower than that of purchasing vehicles when the manufacturer provides the sharing service.This consideration is in line with Li et al. [18].Thus, consumers who rent a GV can get a usage valuation , where  (0 <  ≤ 1) is the valuation discount coefficient of the shared product and captures consumers' acceptance of shared product.Due to the low-carbon awareness , consumers obtain the utility of (1 + ) when renting the manufacturer's shared EVs.The consumers' utility for renting the shared GV and the shared EV are

Manufacturer's decision
To maximize profits, the manufacturer who simultaneously produces and sells vehicles also considers whether to provide the sharing service and which type to launch.If the manufacturer does not provide the sharing service, the profits come from two parts: the EV sales market and the GV sales market.So, the manufacturer's profits are   = ∑︀  (  −   ) 1 .Due to the high cost of batteries, we assume that the marginal cost of EV is higher than that of GV [28].
If the manufacturer provides the sharing service, the manufacturer can obtain the profit from the sharing market.When consumers rent a GV, they only get the right of use.If one consumer stops using the shared product, the other consumer can use it.Thus, there exists a pooling effect that one product can be used to serve more than one consumer on the supply side [33].To capture the pooling effect, we assume   (0 <   < 1) represents the service efficiency of vehicle .To be noticed that there also exists a depooling effect (  ≥ 1) when the delivery of shared vehicles does not form economies of scale.When consumers have use needs, consumers cannot find shared vehicles and the use value of shared vehicles cannot be released.Then, the manufacturer should launch more vehicles to meet consumers' needs [11].A large   represents a low service efficiency.Oppositely, a small   means a high service efficiency.Compared to the GV, the slower charging speed and longer recharge time of EV limit consumers' available time.Therefore, shared EVs' service efficiency is far less than shared GVs' (  >   ).Since the manufacturer retains the product's salvage value   , the profits of the manufacturer are expressed as   = ∑︀  (  −   ) 1 +    2 − (  −   )   2 .The sequence of events as follows.At step 1, the manufacturer decides whether to provide the sharing service.If the manufacturer does not provide the sharing service, at step 2, it sets the sale price   .At step 3, consumers decide which type to buy.If the manufacturer provides the sharing service, at step 2, it decides which vehicle's type to share.At step 3, it sets the sales price   and the rental price   .At step 4, consumers make their choice to buy or rent and which type to buy. Figure 1 illustrates the sequence of events and Table 2 summarizes the notation we use herein.

Scenario analysis
In this section, we first analyze the benchmark where the manufacturer does not provide the sharing service (Scenario N), then we analyze the scenario that the manufacturer shares GVs (Scenario G) and the scenario that the manufacturer shares EVs (Scenario E).Since   >   , we assume   =   + .For simplicity, we let   = 0,   = .

Scenario N
In this scenario, the manufacturer simultaneously sells EVs and GVs, but does not provide the sharing service.When the consumer has a use need, she has two options: buying an EV or a GV.If the consumer purchases the EV, since EV can reduce carbon emissions, she can obtain more value compared to purchase the GV.Consumers decide whether to buy the EV or GV according to the net utility.The consumers' utility for making different decisions is shown below.
Then the profits of the manufacturer are expressed as We identify the conditions that ensure both demands and prices of two types of vehicles are positive and obtain Lemma 1.
According to Lemma 1, the conditions that the manufacturer simultaneously sells EVs and GVs are that the cost gap between them is not too large (0 <   −   ≤ ), and the salvage value gap is small (max{0,  1 } <  ≤  −   +   ).Although consumers have additional willingness to pay for the low-carbon performance of EVs, this additional willingness to pay is limited.If the cost gap is large, the additional willingness to pay cannot offset the impact of EV's higher sales price.Consumers have no incentive to purchase the GV since the low-cost performance.From Lemma 1, consumers with high use need can tolerate the larger cost gap compared to consumers with low use need.This is because consumers with high use need have higher valuation on EV and reduce the per cost of travel.Thus, the manufacturer should strive to reduce the marginal cost of EV to promote EV adoption.In addition, the salvage gap between GV and EV also has an impact on the manufacturer's product line strategy.If the salvage value gap is high ( >  −   +   ), consumers will buy the GV rather than the EV, which results in a negative EV sales market demand.The manufacturer will only produce GV.
From Lemma 1, we know the product's marginal cost and salvage have an important impact on ensuring that both demands and prices of EVs and GVs are positive.

Scenario G
In Scenario G, the manufacturer launches GVs in the sharing market to meet consumers' increasing demand for shared travel.Thus, the consumer has another option aside from buying: renting a GV.When the consumer rents the GV, she obtains the utility  since she acquires the right to use the vehicle but not the right to own it.The net utility of different consumer decisions is as follows.
(i) Purchase an EV:

According to the utility functions in this scenario, we can get 𝑣
The profit of the manufacturer is expressed as The analysis is the same as Lemma 1, we identify the conditions for the manufacturer to provide car sharing and get Lemma 2.
the manufacturer not only sells two types of vehicles, but also provides shared GVs.Among them,  2 = . Same as Lemma 1, both the cost gap and the salvage value gap cannot be too large, the manufacturer can provide shared GVs while selling two types of vehicles.Further, Lemma 2 states the conditions that the manufacturer shares GVs.When the valuation discount coefficient of shared products  is not too high ( ≤ 4 (1+) 2 ), the GV sales market still has positive demand.When the discount coefficient of shared products  is high ( > 4 (1+) 2 ), there is little difference in consumers' valuation of buying and renting vehicles, which makes the GVs sales market demand shift to the sharing market due to the low expenditure.In addition, since the salvage value means a reduction in actual expenditure, consumers also consider the impact of the salvage value when they make their purchase decision.When the salvage value is high ( ≥  −   +   ), consumers have more incentive to purchase GVs rather than rent GVs, which makes the manufacturer lose the demand of the sharing market.When the salvage value is in the middle range, consumers still choose to purchase GVs no matter whether the service efficiency of shared vehicles is high or low.When the salvage value is low, consumers are reluctant to buy GVs and more willing to use shared GVs.
We further analyze how the service efficiency affects the manufacturer's strategy.The service efficiency not only reflects the balance of the supply and demand in the sharing market, but also affects the volume of vehicles launched by the manufacturer.If the service efficient is too low (  >  2 ), the manufacturer needs to launch more vehicles to meet the random demand of consumers in geographical locations, which in turn increases the manufacturer's fixed cost and the rent.When the service efficient is not high, the sharing market demand decreases (   2  < 0), consumers cannot meet their usage needs by renting the shared GV then switch to purchase the GV.As a result, the GV sales market demand increases.As the service efficient further rise, the high service efficient makes the shared GV more attractive, which leads to the result that the sharing market completely cannibalizes the GV sales market.
Further, we obtain the following proposition by comparing the equilibrium demand, price, and profit between Scenario G and the benchmark model.Proposition 1.Compared to the scenario where the manufacturer does not provide the sharing service, when the manufacturer provides shared GVs: (i) The EV's sales remain the same, but the GV's sales decline, i.e.,   1 =   1 ,   1 <   1 .(ii) Both the EV's and GV's sales prices increase, i.e.,    >    ,    >    .(iii) The manufacturer's total profits increase, but the profit in the sales market decreases, i.e.,    >    ,    1 <    .
Proposition 1 shows that the sharing market cannibalizes the GV sales market when the manufacturer provides the sharing service.When the manufacturer launches shared GVs, some price-sensitive consumers choose to use shared GV to reduce expenditure.In order to promote the use of shared GVs, the manufacturer raises the sales prices of both GV and EV (   >    ,    >    ).The sharing service not only attracts consumers who were reluctant to purchase vehicles originally, but also shifts the demand from the GV sales market to the sharing market (  1 <   1 ).Although the sales price of EV increases, the EV sales market demand is constant.For one thing, consumers who purchased GVs are not sensitive to the price and more concerned about the ownership of product and the characteristics of environmental protection, so they do not attempt to use shared GVs.For another, the manufacturer raises the sales prices of GV and EV at the same time, the prices of the two vehicles increases by the same amount (∆ =    −    =    −    ).Therefore, the EV sales market demand remains unchanged.It is worth noting that the manufacturer's decision on sharing GVs reduces the share of GV sales market.Although the manufacturer tries to cover its loss by raising the EV's and GV's sales prices, the profit generated by raising prices does not offset the loss due to reduced demand.So, the profit in the sales market decreases (   1 <    ).However, shared GVs attract consumers who originally purchased GVs as well as those who did not purchase or rent.This enables the manufacturer to capture additional demand.Thus, the manufacturer gains more profits in the sharing market, which not only cover the loss in the GV sales market, but also increase the manufacturer's total profits (   >    ).

Scenario E
In Scenario E, the manufacturer launches EVs in the sharing market.Similar to Scenario G, the consumer also has another option in addition to purchasing: renting an EV.Consumers decide whether to buy or rent the vehicle according to the net utility.The consumers' utility is as follows.
(i) Purchase an EV: Thus, the profit of the manufacturer is expressed as Similarly, in order to ensure that the manufacturer can enter the sharing market, we determine the conditions and derive Lemma 3.
, the manufacturer provides the EV sharing service.Among them, .
Along with Lemmas 1 and 2, both the cost gap and the salvage value gap cannot be too large, the manufacturer can provide the sharing service while selling two types of vehicles.
Compared to Lemma 2, only when consumers have a lower value perception for shared products (0 <  < 4 (1+) 2 (1+) ), the GV sales market demand   1 is positive.Since EV can reduce carbon emissions, more consumers who originally bought GVs turn to rent them when the manufacturer shares EVs.As the sharing market encroaches on the GV sales market, consumers still have incentive to purchase GVs since consumers have a lower valuation on shared EVs.
When the manufacturer provides shared EVs and the salvage value gap is large, the service efficiency should not be too low (  ≤  2 ) for ensuring that the sharing market demand is positive.As we mentioned above, the sharing market cannibalizes the GV sales market demand.When the salvage value gap is large, consumers have more incentive to buy GVs.If the service efficiency is not high, some consumers cannot meet their mobility needs by using shared vehicles and turn to purchase vehicles.Thus, the sharing market demand decreases but the GV sales market demand increases.When the salvage value gap is small, consumers prefer renting shared EVs in the sharing market.In addition, there also exists a constraint that the service efficiency cannot be too high (max{0,  1 } <   ).If the service efficiency is high, the GV sales market is replaced by the sharing market.
Since EV can reduce carbon emissions, consumers have an additional willingness to pay for EVs.When the manufacturer provides EV sharing and the salvage value gap is small, it is more likely to happen that the sharing market replace the GV sales market.Therefore, Lemma 3 imposes stricter constraints on the service efficiency compared to Lemma 2 ( 1 >  1 ,  2 <  2 ), as shown in Figure 2.
)︂ • Lemma 3 shows the conditions that ensure all market demands are positive when the manufacturer shares EVs.Then, we compare the price, demand and profit between Scenario N and Scenario E and derive Proposition 2.
Similar as Proposition 1, the EV sales market demand is constant but the GV sales market demand decreases.In addition, we also find that the increase in the prices of GV and EV is the same (∆ =    −   =    −   ).Since consumers who purchased EVs are not sensitive to price and attach importance to ownership, they still choose to purchase EVs regardless of the existence of shared products.In order to maximize profits, the manufacturer raises the sales prices.Consumers who originally bought GVs choose to use shared EVs due to the lower rent, which makes the sharing market cannibalize some of the GV sales market.Compared to Scenario G, the EV sharing market cannibalizes more GV sales market (  1 −   1 >   1 −   1 ).To make up for the demand loss, the manufacturer raises the sales price of GVs (   <    ).Summarizing the above analysis, the manufacturer's participating EV sharing increases the usage of EVs and promotes the adoption of EVs but does not boost the sales of EVs.
Note that the manufacturer makes more profits by providing the sharing service than by not sharing, but the profit from the sales market decreases.Although the manufacturer raises the sales prices of GV and EV, the profit obtained by raising sales prices cannot offset the loss due to the decreased demand.Therefore, the profit from the sales market decreases.However, the sharing market can attract more consumers who are unable to travel by buying vehicles.Thus, the additional profit from the shared EVs makes up for the loss from the GV sales market and increases the total income.

Equilibrium analysis
In this section, we discuss how the salvage value and the service efficiency affect the manufacturer's vehicletype strategy for car sharing.For the convenience of analysis, we let  equal to   ( > 1), where  denotes the service efficiency ratio of an EV to a GV.
Firstly, we examine the impact of the service efficiency on the manufacturer's profit of sharing different types of vehicles and obtain Proposition 3. Proposition 3. When both the service efficiency ratio and the salvage value gap are low or high, the manufacturer should launch EVs.As the service efficiency ratio is in the middle range, the manufacturer should launch GVs.
Proposition 3 indicates that the manufacturer chooses which type of vehicles to launch in the sharing market by accounting for the trade-off between low cost of GV and high rent of EV.Compared to EV, GV has lower production cost and higher salvage value, the manufacturer can reduce its fixed cost.GV also has higher service efficiency than EV, the manufacturer can launch fewer vehicles in the sharing market to meet consumers demand, which also lowers fixed investment in the sharing market.However, due to the environmental performance, consumers have extra willingness to pay for EVs, so the manufacturer can get more rent by launching EVs.
As shown in Figure 3, when both the service efficiency ratio and the salvage value gap are low ( <  1 ), the manufacturer should launch EVs.A low service efficiency ratio means that the charge time of EV is short and there is no significant difference between the operation time of GV and EV.It also means the volume of vehicles launched by the manufacturer is close whether it shares GVs or EVs.If the salvage value gap is low, GV completely loses its advantages in the sharing market, while EV can attract more consumers due to its environmental performance.Although the cannibalization of the sharing market into the GV sales market is even more serious if the manufacturer shares EVs, it is more profitable for the manufacturer to launch EVs.As the service efficiency ratio increases ( 1 <  ≤  2 ), EV requires longer time to be charged and has less time to operate.To meet consumers demand, the manufacturer needs to launch more EVs in the sharing market, which increases fixed investment.If the salvage value gap increases, the manufacturer prefers to launch GVs.As the salvage value gap further increases ( >  2 ), consumers are more motivated to buy GVs and thus there is no demand for the EV sales market.Although the fixed cost of shared GV is lower for the manufacturer, shared GV is less attractive to consumers.Only by launching EVs with environmental performance can the manufacturer attract more consumers and gain more profits.Therefore, the dominant strategy for the manufacturer is launching EVs.
Then, we discuss how the valuation discount coefficient of shared product affects the manufacturer's profit when it launches different types of vehicles in the sharing market and obtain Proposition 4.
Proposition 4. When the valuation discount coefficient is in the middle range ( 1 <  ≤  2 ), the manufacturer should launch GVs; otherwise, the manufacturer should launch EVs.
Proposition 4 show when the valuation discount coefficient is low, consumers have a low valuation of shared vehicles and prefer to buy them.Compared to GV, consumers have additional low-carbon preferences for EV.This additional preference of shared EV mitigates the loss of demand due to consumers' discrimination against shared vehicles.Therefore, the manufacturer makes more profits by launching EV than GV (   >    ,    2 >    2 ).As the valuation discount coefficient  increases in Figure 4, the competition between the sharing market and GV sales market intensifies.Since the rent of GV is less than EV, the manufacturer can attract more consumers and gain more profits by launching GVs (   <    ,    2 <    2 ).As  further increases in Figure 4, there is little difference in consumers' valuation on buying and renting vehicles, consumers find it more suitable to rent vehicles and have less incentive to buy them.To maximize the profit, the manufacturer chooses to share EVs since the rent of EV is higher than GV (  >   ).Thus, the manufacturer's total profits and the profit in the sharing market is the highest (   >    ,    2 >    2 ).Next, we investigate the effect of consumers' low-carbon awareness on the manufacturer's profit and get Proposition 5.
Proposition 5. When consumers' low-carbon awareness is low, the manufacturer should launch GVs.When consumers' low-carbon awareness is high, the manufacturer should launch EVs.Proposition 5 and Figure 5 intuitively show the effect of consumers' low-carbon awareness on the manufacturer's profit.When both consumers' low-carbon awareness and the salvage value gap are low, consumers'  additional willingness to pay for EV's environmental performance is low, resulting that EV loses its advantage.In this case, the manufacturer should launch GVs with lower cost.As consumers' low-carbon awareness increases, EV can attract more consumers.So, the dominant strategy for the manufacturer is launching EVs.As the salvage value gap increases, although EV can attract more consumers, the manufacturer chooses to share GVs because of their higher salvage value.The manufacturer can reduce sharply the cost by launching GVs.
Finally, we compare the EV sales market demand, the GV sales market demand and the sharing market demand respectively in three scenarios and obtain Proposition 6. Proposition 6. (i) The EV sales market demand remains the same in three scenarios, i.e., (  1 =   1 =   1 ).(ii) When both the service efficiency ratio and the salvage value gap are low ( ≤  3 ), the GV sales market demand in Scenario N is the most, while Scenario E is the least.When the service efficiency ratio is in the middle range ( 3 <  ≤  4 ), the GV sales market demand in Scenario N is the most, while Scenario G is the least.When the service efficiency ratio is high ( >  4 ), the GV sales market demand in Scenario SE is the most, while Scenario G is the least.(iii) When the service efficiency ratio is low ( ≤  5 ), the sharing market demand in Scenario E is more than Scenario G. Otherwise ( >  5 ), the sharing market demand in Scenario G is more than Scenario E.
As we mentioned above, the sharing service does not affect the EV sales market demand.No matter whether the manufacturer provides sharing services or not, consumers do not change the original intention of buying EV.Both Proposition 6 and Figure 6 shows that, when the salvage value gap is low, consumers with low valuation are more willing to use shared vehicles instead of buying GV, which makes the sharing market cannibalize the GV sales market.So, the GV sales market demand in Scenario N is the most.In two scenarios where the manufacturer provides the sharing service, when the salvage value gap and the service efficiency ratio is low, more consumers who originally bought GVs turn to use shared EVs in Scenario E. Therefore, the GV sales market demand in Scenario E is the least and the sharing market demand in Scenario E is more than Scenario G.As the service efficiency ratio increases, EV needs more time to recharge and the manufacturer needs to launch more EVs to meet consumers demand, which increases the fixed cost and the rent.The high rent leads some consumers to purchase GVs instead of renting them, so the GV sales market demand in Scenario G is the least and the sharing market demand in Scenario G is more than Scenario E. As the salvage value gap increases, the GV sales market demand in Scenario E is the most.This is due to the high rent resulted from the EV's low service efficiency.In addition, the rent is so high that some consumers prefer to buy GVs with high salvage value.Intuitively, the EV sharing market is cannibalized by the GV sales market.The increase of GV's salvage value reduces the manufacturer's fixed cost and makes the manufacturer lower the rent.Thus, the GV sales market demand in Scenario G is less than Scenario N.

Environmental impact
In recent years, the EV's industry has developed rapidly since EV can reduce carbon emissions.So, in this section, we discuss the effect of manufacturer's different strategies on the environment.To measure vehicles' environmental impact, we assume   represents the vehicle's environmental impact factor.Thus, when the manufacturer does not provide the sharing service, the total environmental impact (TE) is     1 +     1 .If the manufacturer shares GVs, the total environmental impact is   (  1 +   2 ) +     1 .If the manufacturer shares EVs, the total environmental impact is     1 +   (  1 +   2 ).The lower of the environmental impact, the more environmentally friendly it is, and vice versa.
In order to further study the effect of manufacturer's different strategy on the society, we introduce the social welfare (SW), which consists of the consumer surplus (CS), the manufacturer's total profit and the total environmental impact in our model.The consumer surplus is defined as the aggregate utility of all consumers in the sales market and the sharing market.Thus, when the manufacturer does not provide the sharing service, the consumer surplus is ∫︀  The parameter  means that consumers' acceptance of shared product, which affects consumers' choice of travel mode and the manufacturer's vehicle-type strategy for car sharing.Proposition 7 indicates that when the valuation discount coefficient is low, consumers with low valuation choose other ways to satisfy their needs, such as walking and taking public transportation.This leads to a low environmental impact in the sharing market.Compared to Scenario N, since the GV sales market demand shifts to the sharing market, the environmental impact in the GV sales market is reduced.The reduced environmental impact in the GV sales market is larger than the environmental impact produced by the sharing market.So, it is better for the environment that the manufacturer provides the sharing service, as shown in Figure 7. Compared to GV sharing, EV sharing attracts more consumers not only who originally bought, but also who chose other ways.This makes the environmental impact in Scenario E be larger than Scenario G.But the manufacturer chooses to share EVs instead of GVs in order to maximize its profits.As  increases, the uncovered segments of consumers' use needs shrink.More low-value consumers attempt to use shared vehicles to meet their needs.Although the manufacturer provides the sharing service to meet consumers' needs, it also increases carbon emissions.Therefore, it is environmentally friendly that the manufacturer does not provide the sharing service, but it shares GVs in order to maximize profit.This means that the sharing service and the manufacturer's equilibrium decision are not necessarily good for the environment.As  increases further, the sharing market further cannibalizes the GV sales market, while also attracting more low-value consumers.Since the EV's lower carbon emissions, if the manufacturer shares EVs, the reduced environmental impact in the GV sales market offsets the increased environmental impact in the sharing market.Therefore, EV sharing is the most environmentally friendly.Only when  >  2 , the equilibrium strategy of EV sharing can maximize the manufacturer's profit while being the most environmentally friendly.Furthermore, we analyze the impact of different parameters on the total environmental impact through numerical analysis.According to Mao et al. [22], we know that people are gradually accepting car sharing and nearly 80% may consider it.So, we let  = 4  5 .Although consumers' low-carbon awareness has been improved Figure 8.Effect of service efficiency on total environmental impact.
as the government and environmental protection organizations strongly advocate protecting the environment in recent years, it is still at a low level.According to Shao et al. [28], we set consumers' low-carbon awareness as 1  10 .We let the cost of a GV   equal 1 5 , which represents the low production cost.As we all known, the cost of an EV is higher than that of a GV.According to Shao et al. [28] and Huang et al. [12], we can obtain the cost of an EV is 1.16 times that of a GV.It is easy to obtain that the cost of an EV   = 11 50 .According to Jia et al. [15], we can obtain that the service efficiency of shared GV   = 17  20 .Since the service efficiency of shared GV is higher than that of shared EV, we set   = 19  20 .In our model, the salvage value is exogenous variable.We let  equal 1  25 according to our assumption that the salvage value is a bounded value which can make the demands of all market nonnegative.According to Shao et al. [28], the environmental impact of GV   = 3 20 .We can know EV reduces carbon emissions by 40% compared to GV [9].It is easy to get the environmental impact of EV   = 9 100 .Figure 8 shows that when there is little difference in the service efficiency between EV and GV, the sharing market attracts both consumers with low valuation and consumers who originally bought.If the manufacturer shares GVs, the total environmental impact increases since more low-value consumers use shared GVs.So, the total environmental impact in Scenario G is the largest.If the manufacturer shares EVs, the environmental impact in the GV sales market decreases since some consumers who originally purchased GVs turn to use shared EVs.The reduced environmental impact can offset the increased environmental impact of the sharing market.Therefore, the total environmental impact in Scenario E is the lowest.When EV's service efficiency is low, the manufacturer needs to launch more shared vehicles than consumers' demand, which results in the depooling effect.Some consumers who originally rented EVs switch to purchase GVs, and thus the total environmental impact in Scenario E is larger than that in Scenario N and even larger than that in Scenario G.
Figure 9 shows that the total environmental impact in Scenario G is the largest and that in Scenario E is the lowest.The total environmental impact increases as the salvage value gap increases in all scenarios.Since the sharing market attracts low-value consumers, it increases the environmental impact.So, the total environmental impact in Scenario G is the largest.But the sharing market also attracts consumers who originally purchased GVs.If the manufacturer launches EVs, the reduced environmental impact in the GV sales market can offset the increased environmental impact in the sharing market.Therefore, the total environmental impact in Scenario E is the lowest.When the salvage value gap increases, consumers are more willing to purchase GVs than to use shared vehicles and purchase EVs, which results in an increase in the total environmental impact.Figure 10 shows that the total environmental impact in three scenarios decrease with the increase of consumers' lowcarbon awareness.This is because the improvement of consumers' low-carbon awareness increases the demand of both the EV sales market and the EV sharing market.
Finally, we analyze the influence of different parameters on the social welfare.Figure 11 shows that when the service efficiency gap between EV and GV is not large, the social welfare is the highest when the manufacturer  shares EVs.EV can increase consumers' willingness to pay and the consumer surplus.The manufacturer can also acquire more profits because of the higher rent, and total environmental impact is the lowest.So, the social welfare is highest when the manufacturer shares EVs.As the service efficiency gap increases, the manufacturer needs to launch more volume of EVs to meet consumers' needs, which increases the burden of the manufacturer and reduces profits.In addition, the total environmental impact in Scenario E is larger than that in Scenario N and even higher than that in Scenario G. Therefore, the social welfare is the highest when the manufacturer shares GVs. Figure 12 shows that the social welfare is the highest if the manufacturer shares EVs, but it decreases with the increase of salvage value gap.When the salvage value gap is not large, consumers are more willing to use an EV since the low carbon emissions.This decreases the environmental impact.The manufacturer can raise the price and rent of GV to increase profits in this instance.Therefore, the social welfare in this scenario is the highest.However, as the salvage value gap increases, some consumers turn to use GVs, which reduces the social welfare.

Extension
In our model, we assume that the manufacturer only launches one type of vehicles and investigate which type of vehicles should the manufacturer launch in the sharing market.In practice, some manufacturers launch EVs and GVs simultaneously in the sharing market.So, we further investigate how the change in the above assumption may affect the findings we obtained in previous sections.
Through calculation and analysis, the conditions that the manufacturer produces and shares two types of vehicles simultaneously are similar as those that the manufacturer only shares one type of vehicles.We provide details in Appendix B. Next, we further discuss how the change in the assumption may affect the manufacturer's profits and the environment.
Figure 13 shows that whether there is a difference in the service efficiency between EV and GV or not, the manufacturer always obtains the most profits by launching two types of vehicles simultaneously in the sharing market.This is because when there are two types of vehicles in the sharing market, it can attract more low-value consumers who do not use a vehicle, which increases the manufacturer's profits.Although the profit of the sales market decreases due to the sharing of vehicles, the profit reduced in the sales market is less than that gained in the sharing market.In addition, as the service efficiency ratio increases, so does the profit of the manufacturer  who launches two types of vehicles in the sharing market.This is because, firstly, the low service efficiency of shared EV shifts the demand in the EV sharing market to the GV sharing market and the GV sales market.Although the profit in the EV sharing market decreases, the profit in the GV sharing market and GV sales increase even more.Secondly, due to the high service efficiency of shared GV, one GV can serve more consumers, which reduces the volume of shared GVs launched by the manufacturer, thus reducing fixed investment and improving car-sharing profits.
As Figure 14 shows, the profit of the manufacturer who launches two types of vehicles in the sharing market is the largest regardless of the salvage value.This is because the sharing market attract more consumers, which brings the manufacturer more profits.Although the various salvage value changes consumers' options, these options can be always satisfied by the manufacturer.With the increase of the salvage value gap, GV has more attraction in the sales market, which decreases the demand in both the EV sales market and the EV sharing market.The increase in the salvage value gap also reduces the fixed cost of the manufacturer who shares GVs.The manufacturer reduces the rent of shared GV to attract more consumers and thus increase total profits.Then, the effect of the salvage value gap on the profit of the manufacturer who shares two types of vehicles is non-monotonic.As the salvage value gap increases, the profit decreases first and then increases.When the salvage value gap is small, the increased profit from the reduction of rent cannot compensate for the loss caused by the change in demand.On the contrary, when the salvage value gap is large, the increased profit can cover the loss.
The total demands in all markets increase when the manufacturer launches two types of vehicles in the sharing market.As Figure 15 shows, compared with the secinario where the manufacturer only shares EVs, the  environmental impact is larger when the manufacturer launches two types of vehicles in the sharing market.GV sharing attracts low-value consumers who did not use a vehicle, resulting in the environmental deterioration.Moreover, some consumers who originally used shared EVs choose shared GVs, which also worsens environmental impact.As the service efficiency ratio increases, the environmental impact also increases, as shown in Figure 15.Due to the low service efficiency of shared EV, some consumers who used shared EV are more willing to purchase GV or use shared GV, which leads to the increase of the environmental impact.Figure 16 shows that the environmental impact also worsens with the increase of the salvage value gap.For one thing, consumers with higher valuation are more willing to buy a GV in the sales market because of its high salvage value.For another, due to the low cost of GV, the manufacturer reduces the rent of shared GV and attract more consumers to use shared GVs in the sharing market.This is not beneficial for the environment.

Conclusion, managemental implications and limitations
This study is inspired by different types of vehicles adopted by automakers in the car-sharing market.We develop a model that a monopoly manufacturer who simultaneously sells GVs and EVs.We discuss which type of vehicles should the manufacturer launch for car sharing considering the sharing service efficiency and the vehicle salvage value.Firstly, we analyze three scenarios where the manufacturer does not provide the sharing service, shares GVs, and shares EVs, respectively, and derive the conditions for the manufacturer entering the sharing market.Then, we compare the price, demand and profit in different scenarios.Next, we obtain equilibrium strategies of the manufacturer considering the profit maximization.Finally, we discuss the impacts of the manufacturer's different vehicle-type strategies on the environment and the social welfare.

Conclusion
Our results can be summarized as follows.Firstly, we analyze the conditions for the manufacturer providing the sharing service.The manufacturer can sell EVs and GVs simultaneously when both the cost gap and the salve value gap are small.Otherwise, the high price and the low salvage value make consumers be no incentive to buy EVs even though they are more willing to pay for EVs with the lowcarbon performance.When consumers have a low valuation on the shared vehicles and the service efficiency of shared vehicles is not too low, the manufacturer provides the sharing service.Compared with the condition for the manufacturer sharing GVs, there imposes stricter constraints on both the consumers' valuation of shared vehicles and the service efficiency if the manufacturer shares EVs.
Secondly, after comparing the price, demand and profit in different scenarios, we find that sales prices of both EV and GV increase.No matter which type of vehicles the manufacturer shares, EV sales remain the same, but GV sales are reduced.This means that the manufacturer's EV-sharing strategy always increases the usage of EVs.As the sharing market cannibalizes the sales market, the profit of the sales market decreases.But the car sharing also attracts consumers with low valuation, so the manufacturer' total profits increase.
Thirdly, we discuss the manufacturer's equilibrium strategies considering the service efficiency and the salvage value.When both the service efficiency ratio and the salvage value gap are low or high, the manufacturer should launch EVs.When the service efficiency ratio is low and the salvage value gap is high, the manufacturer should launch GVs; the same happens when the service efficiency ratio is high and the salvage value gap is low.When the valuation of shared product is low or high, the manufacturer launches EVs; otherwise, the manufacturer launches GVs.When consumers' low-carbon awareness is low, the manufacturer launches GVs; otherwise, the manufacturer launches EVs to maximize profits.
Finally, we find that the equilibrium vehicle-type strategy can maximize the manufacturer's profit while being the most environmentally friendly only if the valuation of shared product is high.Through numerical analysis, we know that the manufacturer's EV-sharing strategy is not always beneficial for the environment.If the service efficiency ratio is relatively high, the environment impact is the largest when the manufacturer launches EVs.Due to the low service efficiency of EV, there exists the depooling effect that causes consumers who originally rented EVs turn to buy GVs.This increases carbon emissions and damages the environment.The manufacturer's GV-sharing strategy always improves the social welfare.When the service efficiency ratio is in the middle range, the social welfare of the manufacturer's EV sharing is lower than that without the sharing service.

Managemental implications
Based on these results, this study has the following managemental implications.For the manufacturer, the sharing service efficiency and the vehicle salvage value are two important factors to decide which type of vehicles should it launches in the car sharing market.A high service efficiency means the maturity of the sharing market, that is, the high matching efficiency of the supply and demand.So, the manufacturer does not need to launch too many shared vehicles and reduce the fixed investment accordingly.A high salvage value can reduce the fixed cost of the manufacturer and the actual expenditure of consumers.Therefore, when considering which type of vehicles to share, the manufacturer prefers GVs with high salvage value and without charging.As consumers' low-carbon awareness rises, they have high valuation on both the low-carbon EVs and the sharing mobility.Therefore, the manufacturer begins to share EVs with the shift in the consumption trend towards sustainable products and services.To obtain more profits from EV sharing, the manufacturer should shorten the charging time and improve the driving range of EVs, which can reduce not only consumers' purchasing anxieties, but also the manufacturer's fixed investment.
In addition, we find that the car sharing is not necessarily environmentally friendly.If the manufacturer provides the carsharing service, the profit-maximization vehicle-type strategy is not always the best for the environment.Therefore, there need some measures to encourage the manufacturer to choose the vehicle-type strategy based on the environment protection rather than the profit maximization.For instance, the government has been implemented policies such as tax cuts and subsidies, which reduce the cost of shared EVs.Moreover, car-sharing platforms should improve the matching efficiency of supply and demand.It can not only reduce the operating cost of car sharing, but also meet the travel needs of more consumers.

Limitations
This study also has certain limitations.Firstly, to analyze the impact of the service efficiency and the salvage value on the manufacturer's vehicle-type strategy, this study does not consider subsidies and other government incentive measures.In practice, to encourage consumers to use EVs, the government usually provides subsidies to consumers.Subsidies can reduce consumers' actual expenditure and increases their purchasing motivation accordingly.In addition, in the case of EV sharing, the manufacturer can also obtain subsidies, which reduces the fixed cost.Therefore, we will further consider the impact of subsidies on the manufacturer' vehicle-type strategy in future work.Secondly, our model consisting of the sales market and the sharing market only considers the demand cannibalization and competition from the internal product lines of the enterprise, without considering the competition between enterprises.Thus, future work will be to consider the sharing market and analyze the manufacturer's vehicle-type strategy facing external competition.

Appendix A.
In this appendix, we maximize the interests of the manufacturer, and obtain feedback equilibrium solutions.
Proof of Scenario N. To ensure EV promotion and the positive effect of consumer's low-carbon awareness, we assume that  is in the region ( ≥ (1−1)−(−) ).That is, consumers will obtain more utility when they purchase EVs.To ensure that the sales market demand of GVs is positive,   1 ≥ 0. If   1 < 0 and   1 < 0, consumers will not purchase the manufacturer's vehicles.So .Finally, we use the consumer utility formula to derive total environmental impact, consumer surplus and social welfare., we can obtain .
Next, we substitute .
Proof of Scenario E. Similar to the proof of Scenario G.
Proof of Proposition 6.First, we compare the EV sales market demand in three scenarios.From Appendix A, it is easy to obtain that (i) Purchase an EV:   1 = (1 + ) −   +   .(ii) Purchase a GV:   1 =  −   +   .

Figure 2 .
Figure 2. Feasible and unfeasible regions for Scenario G and Scenario E.

Figure 7 .Proposition 7 .
Figure 7. Effect of valuation discount coefficient on total environmental impact.

Figure 9 .
Figure 9.Effect of salvage value on total environmental impact.

Figure 10 .
Figure 10.Effect of low-carbon awareness on total environmental impact.

Figure 11 .
Figure 11.Effect of service efficiency on social welfare.

Figure 12 .
Figure 12.Effect of salvage value on social welfare.

Figure 13 .
Figure 13.Effect of service efficiency on total profits.

Figure 14 .
Figure 14.Effect of salvage value on total profits.

Figure 15 .
Figure 15.Effect of service efficiency on total environmental impact.

Figure 16 .
Figure 16.Effect of salvage value on total environmental impact.

Table 1 .
Differences of the relevant literature.

Table 2 .
Summary of notation.
Parameters Definitions   ∈ {, },  represents gasoline vehicle,  represents new energy vehicle   ∈ {1, 2}, 1 represents the sales market, 2 represents the sharing market  Consumer's valuation for purchasing/renting product,  ∼  [0, 1]  Consumer's low-carbon awareness  Valuation discount coefficient of shared product  Manufacturer's profit  Demand of product  in the market   Price of product   Marginal cost of product   Rental price of product   Salvage value gap between product   Salvage value of product   Consumer's use need  Service efficiency ratio of EV to GV  Service efficiency of product   Environmental impact factor of product To ensure EV promotion and the positive effect of consumer's low-carbon awareness, we let   .Then, we get the demand of products in the sales market +   ) and   * 2 .The profit of the manufacturer is expressed as   = (  −   )  1 + (  −   )  1 +     2 − (  −   )    2 .