Abstract
This chapter aims to investigate how social science can shed light on the diffusion process of autonomous driving. Autonomous or automated driving is expected to improve the efficiency of road traffic and, consequently, of society. Therefore, it is one of the most crucial social applications of big data and artificial intelligence (AI) that is expected to be realized in our society. Two theories are employed to explain the diffusion: the theory of network effects and the tipping point theory. The former stems from economics, while the latter stems from sociology. Both theories investigate how interdependencies among people affect the adoption of a new innovative change. The existence of “data network effects” in data platform services is a key concept for applying the theory of network effects. Its influence is conspicuous in services that utilize data as a platform and affect the diffusion of services. The level of interdependencies among users is affected by various factors that represent social conditions. The tipping point theory assumes that four social factors shape public opinion. Depending on the level of the factors, public opinion does not change easily, but when social pressure increases, it suddenly changes dramatically. Taking autonomous driving as an example, this chapter discusses how interdependencies among users affect the diffusion process.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
DRIVE (2017) Autonomous vehicles without human intervention “useless”, Nissan claims. https://www.drive.com.au/news/autonomous-vehicles-without-human-intervention-useless-nissan-claims/. Accessed 29 Aug 2021
Katz ML, Shapiro C (1985) Network externalities, competition, and compatibility. Am Econ Rev 75(3):424–440
Littlechild S (1975) Two-part tariffs and consumption externalities. Bell J Econ 6(2):661–670
Mitomo H (1992) Heterogeneous subscribers and the optimal two-part tariff of telecommunications service. J Operat Res Soc Jpn 35(2):194–214
Mitomo H, Jitsuzumi T (1999) Impact of telecommuting on mass transit congestion: the Tokyo Case. Telecommun Policy 23, Elsevier Science
Oren SS, Smith SA (1981) Critical mass and tariff structure in electronic communications markets. Bell J Econ 12(2):467–487
Parker GG, Van Alstyne MW (2005) Two-sided network effects: a theory of information product design. Manage Sci 51(10):1494–1504
Parker GG, Van Alstyne MW, Choudary SP (2016) Platform revolution: how networked markets are transforming the economy—and how to make them work for you. W. W. Norton & Company
Rohlfs JA (1974) A theory of interdependent demand for a communications service. Bell J Econ Manage Sci 5(1):16–37
SAE International (2019, January 7) SAE Standards news: J3016 automated-driving graphic update. https://www.sae.org/news/2019/01/sae-updates-j3016-automated-driving-graphic. Accessed 29 Aug 2021
Scheffer M, Westley F, Brock W (2003) Slow response of societies to new problems: causes and costs. Ecosystems 6:493–502
Squire L (1973) Some aspects of optimal pricing for telecommunications. Bell J Econ Manage Sci 4(2):515–525
Turck M (2016) The power of data network effects. http://mattturck.com/2016/01/04/the-power-of-data-network-effects/. Accessed 29 Aug 2021
Value. Today (2021) World top 1000 companies list and world ranks as on Jan 1st, 2021 from Value. Today. https://www.value.today/. Accessed 29 Aug 2021
WIRED (2019) Ford taps the brakes on the arrival of self-driving cars. https://www.wired.com/story/ford-taps-brakes-arrival-self-driving-cars/. Accessed 29 Aug 2021
World Economic Forum (2015) Are we ready for self-driving cars? https://www.weforum.org/agenda/2015/11/are-we-ready-for-self-driving-cars/. Accessed 29 Aug 2021
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendices
Appendix: Formal Approach to Data Network Effects: A Case of User-Generated Data
Formulation of Network Effects
Among the most typical data businesses is when a company collects data from its users and creates a data platform (see Fig. 4). Users then get useful information from the data platform. They can access product or service information and additional information such as user evaluations. The abundance and reliability of such information rely on how many users contribute to the data platform. This type of service is expected to have demand externalities in that the benefit to a user depends on the total number of users. As the number of users increases, the benefit each user can obtain from the service also increases.
Suppose there is a data platform service composed of a set of users. The number of the total users is denoted by y, where y is a subset of the total potential users N. Each user acquires information from the platform. In return, they provide various data, including personal information, to the service. We assume a benefit to a user does not depend directly on the information provided but on the number of users because the service quality and quantity depend on how many people use the service.
Following Oren and Smith (1981) and Mitomo (1992), let us assume each user has a unique index i, and without loss of generality, distributed uniformly between 0 and N such that \(i \in \left[ {0, N} \right]\).
Assumption 1
Users are distributed in the order of the size of their potential demand.
Let the index of the user with the minimum potential demand be N and that of the user with the maximum potential demand be 0.
The potential demand for the service can be defined as \(v_{p} = D\left( {p, i, y} \right)\), where p is the unit price for the service. This demand function explicitly defines the dependencies of each user’s demand on the number of users. Most platform services today are provided for free, or users do not pay for the service directly. Alternatively, some services are provided at a flat rate. A two-part tariff can deal with both usage-sensitive and non-sensitive price settings. The total charge that the user i should pay for the service \(C\left( {v_{p} } \right)\) is represented by a combination of the usage and flat fees:
Assumption 2
The demand for the service is finite, even when the service is provided free of charge. That is,
where \(V\left( {i, y} \right)\) denotes the potential demand of user i for the service for the user set y.
The gross benefit for user i from consuming the service depends on the unit price and the potential demand, defined as.
Therefore, the net benefit from this service is
Since the gross benefit can be illustrated by the area under the relevant demand curve, the net benefit is formulated as the consumer surplus \(S\left( {p, i, y} \right)\) net of the fixed charge, \(F\).
For a user set to be feasible, the net benefit for the smallest user \(i = y{ }\) should be non-negative:
Per Mitomo (1992), stable and unstable equilibria can be defined in terms of the user set. For an equilibrium point \(y = y^{*}\), if \(dNB\left( {p, y, y} \right)/dy\) is negative, the user set \([0, y^{*}]\) is a stable equilibrium, and if positive, it is an unstable equilibrium. An unstable equilibrium defines a “critical mass,” a well-known concept in the diffusion theory.
Regarding user-generated data, each user benefits from the service depending on the number of users. Thus, interdependencies among users create a mass effect, resulting in the advantage of attracting many users. If the services provided by competitive suppliers are homogeneous, as in an online information retrieval system, the antecessor can take advantage.
Figure 5 illustrates a case of a single modal net benefit function. \(NB\left( {p, y, y} \right)\) has a single modal parabolic curve at each unit price level. The fixed price is F, which is a cutting plane parallel to the bottom plane. At the price \(p^{*}\), the curve has two points of intersection with \(F\). The lower intersection,\({ }y_{0}^{*}\), is defined as a “critical mass” and the upper one, \(y_{1}^{*}\), an ultimate expansion level of the user set. The supplier can attain a user set exceeding \({ }y_{0}^{*}\) to expand autonomously to \(y_{1}^{*}\), suggesting the existence of the first-mover advantage given the data network effects. If an antecessor can overcome difficulties associated with the start-up stage of business and reach a critical mass level, the business can acquire a dominant position.
Although the figure does not reflect the revenue from the business, the combination of a unit price and a fixed charge can cover a variety of tariff settings, and the supplier can select an appropriate setting as a strategic tool for attracting users. Early adopters are usually those with a greater demand for the service. It is less attractive in the early stage of service delivery because they do not know its usefulness. The supplier can apply a low introductory price or zero price to facilitate the subscription. Regarding a flat rate (F > 0, p = 0), the nearest parabola depicts the net benefit. A critical mass is lower than in the case of a positive unit usage charge. As an extreme case, the figure illustrates the advantage of freemium or an advertising model.
Competition in the Presence of Data Network Effects
As shown in the previous section, the antecessor has an advantage in providing the service over potential entrants and can occupy a dominant position. Suppose there is an entrant that seeks to provide a service identical to the antecessor’s service. From a marketing perspective, the entrant will employ a strategy of product differentiation to avoid fierce competition with the antecessor. If the service is homogeneous, a successful entry will be a cream-skimming entry. That is, the entrant would focus on large-scale users.
From the assumption, the consumer surplus or net benefit is monotone decreasing regarding user index i (Fig. 6). At \(i = y\), it should be equal to zero since the net benefit for the smallest user must be equal to zero at equilibrium.
The success of a new entry depends on the shape of the net benefit function. Figure 6 illustrates the consumer surplus function, defined as the gross benefit net of the total unit usage charge. Assume \(p^{0}\) and \(p^{1}\) are the prices for services provided by the antecessor and the entrant, respectively. If the services from the two suppliers are substantially homogeneous, the entrant cannot set the price higher than the antecessor. Thus, \(p^{0 } \ge p^{1}\). Since
and the demand is monotone decreasing regarding i, for \(p^{0 } \ge p^{1}\), we obtain that
It means that the consumer surplus curve for the antecessor is less steep than that for the entrant. Depending on the setting of the fixed charges, \(F^{0}\) and \(F^{1}\), an intersection can be found (Fig. 7a). It implies that the entrant can obtain the users \(0 \le i \le e\) and the antecessor’s share is \(e \le i \le y\). However, there is a case where the benefit from the antecessor’s service exceeds that from the entrant’s for all users (Fig. 7b). The success or failure of the entrant depends on the shape of the benefit function and the tariff setting. If the antecessor’s service is provided for free or at a low price by utilizing other revenue sources, such as advertisement, it would be difficult for the entrant to gain market share.
Efficiency Versus Dominance Over a Data Service Platform
The previous section overviewed the possibility of competition in the market. In reality, it is challenging to enhance competition in the dominant platform business even when a potential entrant seeks to start a competitive service as far as there is no substantial product differentiation. However, a policymaker intends to realize a liberalized market. Despite the challenge of intervening in the private business directly, policymakers can promote further service diffusion.
Given positive externalities, the equilibrium diffusion level tends to be lower than the socially optimal level (See Mitomo and Jitsuzumi 1999). Data network effects can apply to this case. Suppose there exists a potential user willing to use the service. He will perceive the benefit from using the system with the total number of users \(y + 1\). His perceived benefit is given by NB(y + 1). All users also benefit from his participation. Thus, the increase in the social benefit is
In addition to his benefit, \(NB(y + 1)\), the new user creates additional benefits to all other users, \(y[NB(y + 1)-NB(y)]\). He will not perceive this additional benefit created by his participation. Hence, the private benefit is lower than the social benefit created by him by a specific amount. The equilibrium point where the marginal private benefit is equal to the marginal (private) cost is lower than the socially optimal point where the social marginal benefit is equal to the (social) marginal cost. If left to the market mechanism, a lower diffusion level will be attained. The gap justifies policy support to bridge it and attain a socially optimal diffusion level.
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this chapter
Cite this chapter
Mitomo, H. (2022). Interdependency on the Data Platform and Its Effect on the Diffusion of Autonomous Driving. In: Jitsuzumi, T., Mitomo, H. (eds) Policies and Challenges of the Broadband Ecosystem in Japan. Advances in Information and Communication Research, vol 4. Springer, Singapore. https://doi.org/10.1007/978-981-16-8004-5_3
Download citation
DOI: https://doi.org/10.1007/978-981-16-8004-5_3
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-16-8003-8
Online ISBN: 978-981-16-8004-5
eBook Packages: Economics and FinanceEconomics and Finance (R0)