数据投资、数据共享与数据产权分配

Data Investment,Data Sharingg and Data Ownership Allocation

数据产权是数据要素有效配置的前提。数据具有非竞争性,为了提高效率需要促进数据流转与共享,但这往往使数据收集者因为竞争因素受损,从而抑制数据投资,引发新的效率损失。本文从促进数据共享和激励数据投资两者的权衡取舍切入,研究了数据产权的不同配置方式对资源配置效率的影响。研究表明,数据产权分配给用户更有利于实现数据共享,但是会导致数据收集者不愿意进行必要的投资;产权分配给数据收集者可以为其提供投资激励,但会抑制数据共享。在数据收集成本和数据通用性较高时,将数据产权分配给数据收集者以激励投资更有利于提高社会福利。本文进一步讨论了在数据产权配置基础上引入价格上限管制,并要求收集者不得拒绝数据交易的规制政策的效果,发现在一定条件下,将确权与规制相结合可以在保障数据投资的同时促进数据共享,从而优于单纯的数据产权配置。

The economic value of data has become increasingly prominent.Data ownership is a prerequisite for the effective allocation of data resources.Data are nonrival,and in order to enhance efficiency,it is necessary to promote data circulation and sharing.However,this often leads to disadvantages for data collectors due to competitive factors,thereby suppressing data investment and causing new efficiency losses.This paper investigates the impact of different configurations of data ownership on resource allocation efficiency,focusing on the trade-off between promoting data sharing and incentivizing data investment.We develop a duopoly model with horizontal differentiation,incorporating data collection and transaction.The model consists of two competing service providers located at the endpoints of a Hotelling bench.One service provider also acts as a data collector,while the other service provider lacks data processing capabilities and can only utilize processed data.Based on this framework,the paper compares the efficiency of different data ownership allocations.When data ownership belongs to the data collector,the data collector has the right to decide whether to collect and resell data.When data ownership belongs to the users,the data collector can choose whether to collect data,but the users have the right to determine which service providers can access the data.We further discuss the effects of mandatory data sharing when data ownership belongs to the data collector,specifically,through the implementation of price caps and the requirement that the data collector cannot refuse data transactions.The study finds that when data ownership belongs to the users,data are always fully shared.However,if the investment required for data collection is high or the data have high generality(i.e.,data are valuable for the collector's competitor),the data collector tends to invest less in data collection.On the other hand,when the data collector has data ownership,it is more willing to invest in data collection.However,if the data have low generality,the data collector will not sell the data to another service provider,leading to insufficient data sharing.If mandatory data sharing is required when data ownership belongs to the data collector,within certain parameter ranges,it will incentivize the data collector to invest in data collection and ensure data sharing,which can achieve an intermediate effect between the two modes of ownership allocation and cover the advantages of both configurations.Yet when the data have very strong generality to the extent that professional data collectors emerge,mandatory data sharing can reduce data collection and result in welfare losses.We discuss the welfare consequences of different ownership configurations.From the perspective of user surplus,when the data collection cost and generality are low,data ownership should belong to the users to promote data sharing.When the data collection cost and generality are high,data ownership should belong to the data collector,with mandatory data sharing required to promote data sharing while incentivizing data investment.When the data have very strong generality,data ownership should belong to the data collector,and mandatory data sharing should not be required to avoid hindering the normal operations of the professional data collector.When the data collection cost is very high,it becomes a challenge to facilitate data sharing,regardless of whether mandatory data sharing is required or not,and any ownership allocation will fail to incentivize data investment.From the perspective of total social welfare,the optimal configuration of ownership is generally similar.However,in certain parameter ranges,although the users with data ownership can improve user surplus,it can lead to a decline in social welfare.Therefore,if total social welfare is considered,ownership should be allocated to the data collector in a larger parameter range.Finally,this paper examines the impact of data-sharing aversion among users on the robustness of the conclusions.We demonstrate that when the utility loss caused by privacy concerns is relatively small,the introduction of data-sharing aversion does not affect the overall conclusions.