The impact of tokenization on ride-hailing blockchain platforms

The effects of tokenization on ride-hailing blockchain platforms

Let’s once again analyze an article about blockchain. This article is relatively new and was published on POMS in 2023.

Since the focus of this article is different from the previous articles focusing on fakes, you need to read his INTRODUCTION carefully.

Introduction:
Blockchain has received increasing attention from academia and industry in recent years.
In the traditional peer-to-peer market/sharing economy, relying on a centralized platform to collect and authenticate information reduces supply and demand It reduces the cost of end-to-end communication and search, thereby increasing mutual trust between parties and supervising the execution of transactions.
However, such traditional centralized platforms charge high transaction fees (usually paid by the supplier).

Blockchain provides another option for this type of transactions. Based on distributed ledgers and smart contracts, blockchain can "reduce rent seekers, lower fees, and increase transparency" (DACSEE (Decentralized Online Car Hailing Platform) ) founder Lim Chiew Shan).

However, the application of blockchain faces some challenges: (1) Although blockchain can ensure the security of stored information, it cannot guarantee the authenticity and availability of inputs, leading to the "garbage-in-garbage-out" problem . (2) Incentive mechanisms are needed to ensure users’ active participation in the blockchain system.

In recent years, tokenization has become the mainstream of blockchain development, providing convenience for the transaction of illiquid assets (Babich & Hilary, 2019).

Although the introduction of tokens seems promising, in reality, their application has no advantages. Cryptocurrencies, for example, have been facing severe political regulation around the world due to concerns that they could destabilize financial markets.

This concept, which lacks support from the real economy and is instead supported by speculative activities, is more likely to lose the ability to develop sustainably in the frenzied pursuit of profits by capital, leading to the bursting of bubbles. The risk of this bubble bursting attracted government attention, leading to tighter regulation.
For example, the Chinese government has implemented a series of legislation to ban cryptocurrency activities, which has promoted the development of nontoken-based (non-token) blockchains, such as Dianrong, IBM and other P2P network lending Platforms. This type of platform does not issue tokens, or only uses tokens as trading tools.

Unlike non-token blockchain platforms, platform maintenance costs and profits are covered by charging commissions. Token blockchain platforms retain a portion of issued tokens (cryptocurrencies), and as these tokens become widely used and accepted, the value of these tokens increases, providing profitability to the platform. However, since token systems generally do not charge commissions, they cannot make money through commissions.

In the ride-hailing market, centralized ride-hailing platforms, such as Uber and Didi, entered the market 10 years ago and have occupied most of the market share so far. These platforms provide pricing and matching services for drivers and passengers, charging drivers a commission for their services.
In contrast, blockchain platforms, such as DACSEE, DAV, TADA, Swarm City, RideNote, DRIFE, etc., allow users to match their needs according to their own preferences instead of passively accepting the platform's assignments. .

This article is to explore the potential benefits of blockchain technology, explore its impact on the peer-to-peer market, and analyze the advantages and disadvantages of this type of token platform compared to non-token platforms. This article focuses on the ride-hailing market (which has shifted from a traditional market to a peer-to-peer market).

After introducing a token platform model, we also studied the optimal token bonus (token bonus) considering the impact of mining on platform profits.
Then, in order to avoid the policy’s strict prohibition on the token market, the article also models a non-token platform and studies the optimal token bonus (token bonus) and commission.
The author believes that although all previously mentioned blockchain-based ride-hailing platforms are token-based, non-token platforms are also an option.
Traditional online ride-hailing platforms, such as Didi, Lyft, etc., are seeking blockchain solutions to deal with the threats of emerging decentralized online ride-hailing (Fromgeek.com, 2022).

For these existing online ride-hailing platforms, if they want to introduce blockchain technology, non-token platforms are a more realistic choice, because such platforms do not need to change their profit models, and there is already a huge demand for this type of platform. And for mature platforms, the transition to a token blockchain will be operationally and legally more complex.

Furthermore, the article compares the social welfare of the two types of systems and conducts numerical analysis based on the Beijing market.

Conclusion of the article:
The article found that, first, the token blockchain platform will set a higher mining bonus (mining bonus), so it can attract more drivers as miners (Miner), reducing the lack of mining power. However, this also reduces transaction volume and the probability of drivers successfully finding customers.
Therefore, if transaction costs or mining shortage costs are high, the platform should increase incentives for miners before reducing mining bonuses over time.
Second, non-token platforms should be careful not to set a commission that is too high, because if a commission is set that is too high, a very high mining bonus will be needed to attract enough drivers, And this leads to lower profitability.
Third, the token system will bring more social benefits unless the non-token platform charges a significantly higher price. However, the article proves that this is the case (the non-token platform charges significantly higher prices) higher prices) are unrealistic for the Beijing market case. For the Beijing market, tokens can bring higher profits and matching efficiency. Therefore, in order to alleviate concerns about government activities in cryptocurrency, systematic legislation is necessary for the cryptocurrency market.
Fourth, previous research believes that blockchain efficiency and decentralization are substitutes (cannot have both). Unlike previous research, the article believes that it is possible to simultaneously achieve blockchain efficiency and decentralization.

Article contribution:
First, the article designs an analytical model to study the optimal strategy of token platforms in the peer-to-peer market. In addition, the article also examines how these strategies are affected by market conditions and business environment characteristics (market competition, platform maturity, opportunity costs, fixed costs, pricing rules, etc.).
Second, because cryptocurrency is strictly restricted by policies, the article reshapes the token model in a non-token structure and derives optimal strategies under different circumstances.
Third, the applicability of token and non-token platforms under different circumstances (mining shortage, matching efficiency, system efficiency, social welfare) is studied. We examine which of various situations, tokens or non-tokens, leads to higher social welfare. Different ways (adjustment of blockchain decentralization level, price, mining system design) have also been proposed to improve social welfare. And numerical analysis proves that token platforms can generally perform better in social welfare and matching efficiency.
Fourth, the research results provide insights into understanding how to achieve efficiency and fairness in blockchain platforms. The application of blockchain in the sharing economy will be very advantageous, providing potential opportunities for future blockchain development and research. The article further analyzes the impact of different blockchains and mining systems on user asset purchase and sharing decisions, and gives suggestions on the operating mechanism of the peer-to-peer market.
Fifth, in order to increase the accessibility of token and non-token platforms, the article provides feasible guidance for users, platform operators, and government participants of the blockchain platform. .

Literature Review
The literature review includes two parts, one is "The impact of blockchain on platform operations and the sharing economy", the other is "The intermediary role of sharing platforms".

1. The impact of blockchain on platform operations and the sharing economy
   Few studies have paid attention to the impact of blockchain on platform operations in the sharing economy, so let’s discuss the blockchain first Chain's impact on business processes.
   Frizzo-Barker et al. (2020) believe that current blockchain-related research is still in its early stages, so the management issues of blockchain application in various industries urgently require more quantitative research. Babich and Hilary (2019) believe that blockchain can help sharing platforms reduce the marginal cost of adding new participants and overcome the uncertainty of supply chain finance in terms of information asymmetry and moral costs.
   Research on blockchain in reducing information uncertainty: Pereira et al. (2019) believe that if the benefits of reducing uncertainty and increasing transparency exceed the benefits of writing smart contracts and When the cost of implementing a distributed storage and verification ledger is involved, then a decentralized blockchain system will perform better.
   Research on the impact of blockchain on the sharing economy: J. Sun et al. (2016) believe that blockchain realizes a trustless, distributed, automated and transparent business model, laugh Eliminating the middleman (traditional third-party platform) reduces operating costs and improves efficiency. Through case studies, Pazaitis et al. (2017) found that blockchain-driven platforms can increase the value of productive communities, but the ethical issue of trust still needs to be resolved. Hawlitschek et al. (2018) pessimistically predict that the application of blockchain in the sharing economy is less promising, arguing that the expected benefits of establishing a trustless and decentralized system are not enough to bring about returns.
   Unlike this article, none of the above studies are analytical.

2. The intermediary role of sharing platforms
   Hu (2019) pointed out that there are three themes of research on the intermediary role of sharing platforms: (1) matching and pricing; (2) information and payment; (3) There are self-scheduling suppliers.
   Since this article is about the first topic, only the literature on this topic is reviewed.
   Literature on matching: In order to quantitatively analyze the efficiency of the system, many studies model the matching process as a queuing system. The supply side is the service provider, and customers arrive with a Poisson distribution.
   Many studies have considered the heterogeneity of both supply and demand sides, such as customer impatience with congestion, time-related demand uncertainty, uncertainty in the number of suppliers during congestion, etc. Some studies explore matching priorities between groups with different preferences. Chen and Hu (2020) studied the fixed pricing strategy of online ride-hailing platforms with forward-looking drivers and customers. They found that fixed pricing and price adjustments based on waiting costs will lead to short-sighted behavior of users.

Literature on price: Research trends shift from fixed pricing to stage pricing, and from uniform pricing to pricing based on services and products. For example, Cachon et al. (2017) found that when labor prices become more expensive, dynamic pricing is better than fixed pricing for consumers on ride-hailing platforms. Bai et al. (2018) believe that platforms should charge commission rates based on time, charging higher commission rates during peak periods and lower commission rates during off-peak periods. L. Sun et al. (2019) study the matching problem through per-service pricing (specific trip details and driver location).

The above research is based on the intermediary role of sharing platforms and their decision-making. All questions are thought about in a focused way. However, blockchain operates in a decentralized manner. Without a platform directly involved in pricing and matching, the market composition may be difficult to control, hindering the adoption of blockchain in the sharing economy.
This article mainly studies the adoption of blockchain in the peer-to-peer market and compares the performance of tokens and non-tokens.

Model
Basic model, mainly the optimal decision-making of the token platform. The nature of a token is difficult to define due to its highly diverse functions and applications. In real life, tokens may be adopted in non-token platforms as an incentive for community users, where there is a fixed exchange rate between the token and fiat currency. In the study of this article, it is assumed that the token platform issues an ICO in the initial stage and does not fix its exchange rate.

Recently, two types of blockchain platforms have been implemented, tokens and non-tokens. Both types of platforms are based on blockchain technology including mining, so for blockchain platforms, platform operations and miners play an important role.

Four types of participants in the blockchain platform:
(1) Passengers: The purpose is to reduce costs under the corresponding service level. The core decision variable is price.
(2) Driver: The goal is to maximize profits. For non-token platforms, the driver’s profit is obtained from the platform’s redistribution of customer payment amounts, that is, the platform will charge the payment amount Partly as commission, the driver is paid an amount - the amount of the commission. With token platforms, drivers receive all payment amounts because there are no commissions to pay anymore.
Each user (passenger, driver) decides whether to buy a car. If he buys it, he first meets his own needs. He can also use the excess supply capacity to participate in this as a driver. platform. If he decides not to buy a car, he is a passenger on the platform.
User heterogeneity at the level of self-use can be explained by how frequently the purchased car is used, which will also affect their decision-making. Since heavy users are more likely to become drivers, the goal of all users is to maximize their utility.
(3) Miners, they use digital equipment to obtain mining bonuses. For blockchain platforms, mining is very important. Because there is no centralized platform to manage it, every transaction is effectively recorded on a distributed ledger. Therefore, mining is a process that can expand the system's data storage memory, which requires individual users to participate as miners and use terminal devices with computing capabilities (such as laptops, mobile phones, etc.) to provide storage capabilities. Because blockchain systems store transactions more slowly than other systems, mining bonuses are crucial to ensuring that the blockchain has enough miners to process all transactions.
In this study, the mining process is required for both token and non-token platforms, as both rely on blockchain technology.

It is assumed that both drivers and passengers can become miners as long as they have terminal equipment.

In the base model, it is assumed that only registered drivers are allowed to mine. They upload driving data, calculate, and receive rewards corresponding to their mining.
In the real world, customers may be able to participate in mining during their ride time, but their rewards are generally lower than drivers. For example, drivers’ mining bonuses on AlphaCar (a Chinese decentralized ride-hailing platform) are 100 times higher than those of passengers. The advantage of this setup is that it is easier for drivers to record continuous driving data. This also helps protect privacy, as external miners are eliminated and only the platform’s drivers have access to transaction data.

However, in real life, external miners or consumers can sometimes participate in mining. Therefore, in the extension, the assumption that only platform drivers can become miners is removed.
(4) Platform operators, both types of platforms pursue maximizing profits.
The token platform does not charge commission fees. It makes profits by holding a certain number of tokens for appreciation. This is similar to investing in a stock that has huge room for appreciation, which is why people generally regard this token as a new asset. The value appreciation of this type of currency depends on the total transaction volume of the platform, and the key profit of this type of platform is to achieve the appreciation of the token by encouraging community transactions. At the same time, the platform will reward miners with tokens to maintain the efficiency of the blockchain, because lack of computing power will directly lead to system congestion.
For non-token platforms, the operator’s income comes from commission income, which means that expanding the total transaction volume will help the platform’s profitability. As with token platforms, miners are rewarded to avoid computing power shortages.
Therefore, both types of platforms will use "mining bonuses" to motivate users.

The article assumes that the pricing power belongs to the user (this assumption will be relaxed in the appendix). In addition, the non-token platform will set a commission, which is first set to be exogenous (this assumption will be relaxed in the appendix).

Model T: Token Blockchain Platform
The platform serves as an information center and hands over pricing (price in tokens) and matching to users.

Online ride-hailing and token circulation process:

1. The driver sends the positioning to the platform.
2. Customers need to purchase tokens from the platform in advance before making a request.
3. The platform receives advance payment from customers.
4. The driver then picks up the passenger at the passenger pick-up point. After arriving at the destination, the passenger pays the token to the driver. The platform does not need to charge a commission.

Although different platforms have slightly different pricing processes, the final taxi price depends entirely on the user, not the platform.
For example, on the Swarm City platform, a passenger sends a taxi request containing ride details and proposes a taxi price (in tokens).
If the driver agrees to the deal, the match is complete.
On the RideNode platform, passengers first upload a screenshot of their price on a centralized service platform (such as Uber), then the driver will propose a price, and then the passenger decides whether to accept the price.

All transactions are processed directly between users. This is also how token platforms operate in real life.
In addition to the fare paid by the passenger, the driver can also receive a "mining bonus" token. Users can buy or sell tokens on the platform.

Token Insurance:
In the early stages of a token platform, there was an ICO that was similar to a company's initial public offering. The total number of tokens is n. Because the process of issuing tokens is simple and the investment is small, there is no need to consider the cost of issuing tokens.
Typically, issued tokens are distributed three-way. First, the n1 quantity will be retained and used to sell to users to pay for taxi rides. Second, the n2 amount will be used as the driver’s “mining bonus”. Third, the number of n-n1-n2 will be held by the platform. Assume n and n1 are given.

In fact, because the focus is not on the size of n (because more tokens are issued, it will only dilute the value of the tokens), but on the proportion of the three-way distribution, so n can also be normalized to 1. However, using n is more intuitive. So the article still uses n.
The reason n1 is fixed is because, in order to ensure smooth transactions, there needs to be enough tokens in circulation.

In order to maximize profits, the platform decides the number of n-n1-n2, as well as the number of "mining bonuses" for n2.

A higher bonus would encourage users to participate in the platform as drivers and miners. Especially in the first stage after the introduction of blockchain, it is very important to provide a high bonus to ensure that there is enough mining power to record transactions.
The time length of the first stage is l. The number of tokens available to pay mining bonuses per unit time is n2/l.

Given the expected number of drivers and the monetary value of each token, the monetary value of each driver's unit time can be recorded as mT (n2 can correspond to mT one-to-one).

As you can see from the analysis, it is easy to treat mT as a decision variable, with n2 being the dependent variable of mT, rather than the other way around.

In real life, blockchains always reduce bonus payouts over time, which may require multi-stage modeling. However, in order to facilitate processing and be more intuitive, the model in this article only focuses on one stage.

The number of drivers qTd.
The number of passengers qTc.
qTd+qTc=1.

The taxi fare per unit time: pT, that is, the taxi fare is positively related to the ride time.

In the basic model, it is assumed that the ride price pT is exogenous. On the one hand, this is for ease of handling, and on the other hand, because with competition from external taxi services, the ride price per unit time is often at a reasonable level. In the extension, endogenous pT is also considered.

The total demand per unit time is$\lambda T$.
This is the supply for one hour$\mu T$。

Therefore the expected available time occupied by each driver is
$\alpha T=\lambda T/\mu T$。

Because usually, after a passenger makes a request, the order can be completed quickly to meet the demand, so it is assumed that the total transaction value per unit time is$\lambda T\ast pT$. The value of the platform increases with the total transaction value, so assuming that this relationship is linear, the value of the platform is: z*total transaction value=$With\ast \lambda T\ast pT$, z>0. In the real world, z represents the price-to-sales ratio in financial markets.

Therefore, the expected value of unit token is: the value of the platform/n, that is,
$\varphi =With\ast \lambda T\ast pT/n$。

The driver’s decision-making relative to the passenger.
Assuming that users are heterogeneous, their usage rate per unit time $\theta$ is evenly distributed between 0-1.
The total market capacity is 1.

The cost of purchasing a car is fixed, and this cost is amortized into the cost F per unit time. Each driver will cater to their own needs first and provide services only during their own free time.
Drivers earn profits through two channels: (1) using the car themselves; (2) providing services to obtain taxi fees.

b is the average usage benefit per unit time.

The total usage benefit achieved by the driver by satisfying his own needs is b*$\theta$.
When the driver provides services, there is an opportunity cost c. Therefore, the driver's income per unit time of providing services is p-c.
When the driver provides services, the driver will also receive a mining bonus mT per unit time.
Therefore the driver’s total utility function is:
$Ud=b\ast i+(1-\theta )\ast (\alpha T\ast (pT-c)+mT)-F$

Passenger’s utility function:
$Uc=(b-pT)\ast i$

The indifference point of the driver’s utility function and the passenger’s utility function is$\theta T$, therefore, some users become passengers and some become drivers. Therefore, the number of passengers and drivers qTc and qTd are determined. Through the number of passengers and drivers, the supply quantity can be determined$\mu T$sum demand and demand$\lambda T$, and the probability that the driver successfully gets a passenger α T \alpha T λ$T$ a>$\alpha T$。

Then, the article gives:$\alpha T=\theta T^2/(1-\theta T{)}^2$, I don’t know why it is in the form of a square.

Total user welfare = driver utility + passenger utility
Driver utility is$Ud(\theta )$在0- $\theta$T's contribution.
乘客效用于$Uc(\theta )$在$\theta$Integral on T-1.

Goals of the platform:
Treat digital tokens as a new type of asset and therefore assume that the expected net return is proportional to its value. Let r be the expected net return per unit token per unit time. The expected value of unit token is defined earlier as $\varphi$.
Therefore, the unit net income of the n-n1-n2 number of tokens held by the platform per unit time is $(n-n1-n2)\ast \varphi \ast r$。

Platform profit = net income from tokens held by the platform - mining bonus
In order to ensure the efficiency of the platform, the platform needs to use mining bonuses to motivate drivers to mine.
Assume that the mining time required for each transaction is at least t. Therefore, the necessary mining time for the total demand is $\lambda T$*t，如果总供应$\mu T>\lambda T$*t$ . In order to meet the efficiency requirements of the system.
When the system is inefficient, the system will get a penalty $(\mu T-\lambda T\ast t{)}^{+}\ast \sigma$ 。

Therefore, the objective function of the platform is:
$at{x}_{mT}(\pi T)=(n-n1-n2)\ast \varphi \ast r-(\mu T-\lambda T\ast t{)}^{+}\ast p$

Here, the article does not take into account the mining bonus, only the possible penalties that the system may suffer.
The article believes that $The\; size\; of\; \sigma$ is related to the privatization level of the platform. When the privatization level is lower, it means that the platform is less open and has a higher possibility of being invasion, so leading to a higher $\sigma$. In reality, an accurate estimate $\sigma$ is very difficult, but should be set large enough to ensure that the system has enough mining capacity. Here, in fact$\sigma$ can also be regarded as a guarantee$\mu T>\lambda T\ast The\; Lagrange\; multiplier\; of\; t$.

The minimum mining time t required for a unit transaction can reflect the transaction processing speed of the system and also reflects the decentralization level of the system. Previous research has shown that higher levels of decentralization result in a system that processes transactions more slowly. Because reaching an agreement in a decentralized network requires the consent of more participants, this means that each transaction requires more mining time and computing power.

In order for the decentralized decision-making of the blockchain system to work, a consensus mechanism needs to be introduced.

As a process of verifying transaction information, the consensus mechanism determines the conditions under which new blocks (transaction records) can be created.

The expected value of the unit token determined previously is$\varphi =With\ast \lambda T\ast pT/n$.
Cause average exists$\lambda T$ drivers will receive mining bonus mT per unit time. Therefore, in l unit time, there are$n2=\mu T\ast mT\ast l/\varphi$。

So, substituting can get
$at{x}_{mT}(\pi T)=(n-n1-n2)\ast \varphi \ast r-(\mu T-\lambda T\ast t{)}^{+}\ast p$

Result analysis:

Then, the author conducted numerical analysis and found that mT has an optimal advantage when it affects the platform's profits. A mT that is too low will make the system inefficient, and a mT that is too high will be too expensive for the platform. However, for total social welfare and user benefits, they both increase monotonically with mT, so the author believes that in order to improve overall social welfare, the government should subsidize the platform.

The optimal token bonus:
The author analyzed the σ \sigma for different $The\; range\; of\; \sigma$, the increasing and decreasing relationship between mT and n1/n.
This is about the trade-off between the cost of the shortage of mining power and the loss of holding less currency.
As the platform chooses to sell more tokens to passengers (n1/n), on the one hand, the platform has an incentive to hold fewer tokens and increase mining bonuses to maintain system operating efficiency On the other hand, the platform will also hope to reduce mining bonuses and hold more tokens to increase profits.
Therefore, when the shortage cost is small, mT decreases with n1/n, and when the shortage cost is large, mT increases with n1/n.

The author also analyzed that for a relatively low $The\; range\; of\; \sigma$, the increasing and decreasing relationship between mT and z in the range of different t (the mining time required for each transaction).
It’s also about the trade-off between the cost of a shortage of mining power and the loss of holding less currency.
Therefore, when shortage cost $When\; \sigma$ is low, (1) when t is small, mT decreases with z (P/S ratio), when t is large, mT decreases with The z (price ratio, P/S ratio) increases incrementally.

Then, the relationship of mT with the opportunity cost c (positive direction), the mining time required for each transaction t (positive direction) and the driver's fixed cost amortization F (negative direction) is also analyzed. And the optimal platform profit changes with c (negative), t (negative) and F (negative).

Model N: Non-token Blockchain Platform
Due to policy restrictions on cryptocurrencies, non-token platforms are also worthy of analysis.

Platforms will also set up a "mining bonus", but unlike token platforms, they will charge a commission. The article assumes that commissions are exogenous.

Suppose the commission rate is $\gamma N$, so for each payment from the passenger, the driver can only receive $1-The\; proportion\; of\; \gamma N$.
Therefore, the driver’s utility function becomes;
$Ud=b\ast i+(1-\theta )\ast (\alpha T\ast (pN\ast (1-\gamma N)-c)+mN)-F$.
The passenger’s utility function remains unchanged and is still:
$Uc=(b-pT)\ast i$

The source of income of the platform is only commission, which is$\lambda N\ast \gamma N\ast pN$, flat stand ineffectiveness is$-\sigma \ast (t\ast \lambda N-\mu N{)}^{+}$
Determine the equation of the quotient:
$\pi N=\lambda N\ast \gamma N\ast pN-p\ast (t\ast \lambda N-\mu N{)}^{+}-mN\ast \mu N$。

The authors then use numerical analysis to show how mN affects the profit of the token platform. Studies have shown that the commission rate generally charged by peer-to-peer sharing platforms in the world is 30%-40%. Didi charges a commission rate of about 30% in Beijing, so it is assumed that pN = 1.3pT.

Text analysis completed mN meeting adjudication$\gamma N$ incrementally, if the platform charges a higher commission rate, then the driver will receive less income, so a Higher bonuses will motivate enough drivers to participate in the platform.

Then the author conducts a comparative analysis of the performance of the two platforms.
Consider two situations:
(1) Assuming that the two systems have the same matching efficiency, the system with a lower taxi price will have a higher mining bonus. Compensate drivers.
(2) Assuming the same mining bonus is provided, the token platform will have lower passenger surplus and lower profitability.

Then optimal strategies for the two types of systems were compared for four main characteristics (measures of performance).
1. Social welfare
On non-token platforms, commissions need to be paid to the platform, so the taxi price can only be increased. If the taxi price does not increase enough, the social welfare of the non-token platform will be lower than that of the token platform. Only when pN is large enough, the social welfare of non-token platforms will be greater than that of token platforms. In other words, the total increased profits of the platform and drivers exceed the losses of passengers.
Then, the numerical analysis given by the author shows that a 30% increase in taxi prices is not enough to make the social welfare of non-token platforms comparable to that of token platforms. Moreover, when the driver's opportunity cost c is higher, the mining time t required for a unit transaction is higher, and the fixed cost amortization F is smaller, then the social welfare of the non-token platform will be worse than that of the token platform. Therefore, in these cases, the government should encourage the application and development of token platforms. Moreover, the government can also promote the development of blockchain technology by reducing asset prices (F).

2. Matching efficiency
The greater the probability that the driver successfully finds a passenger, the higher the matching efficiency.
Unless the mining time t required for each transaction is very high, the probability is significantly higher for token platforms than for non-token platforms. This also supports the statement that "sharing leads to greater efficiency."
3. Balancing efficiency and fairness
The article believes that in the two systems, there is a shortage of mining capabilities (the greater the shortage of mining capabilities, the smaller the number of nodes. The lower the level of centralization) decreases as the mining time t required for each transaction increases (blockchain efficiency, the smaller t, the higher the efficiency). That is, the level of decentralization will decrease as the efficiency of the blockchain decreases. This is contrary to previous research conclusions, which concluded that blockchain efficiency and decentralization level are proxies. Using a relatively small number of nodes can ensure higher transaction speeds, but the level of decentralization will be questioned.
Explanation: Both systems adapt to the increase in mining time per transaction by increasing mining bonuses, resulting in more users choosing to become drivers, thereby reducing the lack of mining capabilities.
4. The profitability level of the platform
Since the profitability of non-token platforms is highly related to z and r, it is difficult to compare. Industry, country, and market conditions will all It has a comparative impact on the profitability of the two types of platforms. Through numerical analysis, the article concludes that for the Beijing market, token platforms are better than non-token platforms in terms of profits for drivers and platforms.