资源描述：
Yan Carrière-Swallow and Vikram Haksar The Economics and Implications of Data An Integrated Perspective Strategy, Policy, and Review Department No. 19/16INTERNATIONAL MONETARY FUND The Economics and Implications of Data An Integrated Perspective Yan Carrière-Swallow and Vikram Haksar September 2019 Strategy, Policy, and Review DepartmentCopyright ©2019 International Monetary Fund Cataloging-in-Publication Data IMF Library Names Carrière-Swallow, Yan, author. | Haksar, V. Vikram, author. | International Monetary Fund. Strategy, Policy, and Review Department, issuing body. | International Monetary Fund, publisher. Title The economics and implications of data an integrated perspective / Yan Carrière-Swallow and Vikram Haksar. Other titles International Monetary Fund. Strategy, Policy, and Review Department Series. Description Washington, DC International Monetary Fund, 2019. | At head of title Strategy, Policy, and Review Department. | Departmental paper series. | Includes bibliographical references. Identifiers ISBN 9781513511436 paper Subjects LCSH Big dataEconomic aspects. | Data structures Computer scienceEconomic aspects. | Data protectionLaw and legislation. | Freedom of ination. Classification LCC QA76.9.B45 C37 2019 Publication orders may be placed online, by fax, or through the mail International Monetary Fund, Publication Services P .O. Box 92780, Washington, DC 20090, U.S.A. Tel. 202 623-7430 Fax 202 623-7201 E-mail publicationsimf.org www.imfbookstore.org www.elibrary.imf.org The Departmental Paper Series presents research by IMF staff on issues of broad regional or cross-country interest. The views expressed in this paper are those of the authors and do not necessarily represent the views of the IMF, its cutive Board, or IMF management.Contents Acknowledgments . iv Abstract. iv 1. Introduction .1 2. Context and Framing .3 3. The Building Blocks of a Market for Data . 7 Supply The Decision to Produce Data 8 Demand The Economic Value of Data .9 The Price of Data .12 Economic Characteristics of Data 13 4. Macroeconomic Implications of Data Proliferation 19 Growth.19 Equity 23 Stability 25 5. Data Policy Frameworks 29 Market Opacity 29 Concentration and Market Power 32 Financial Instability 34 International Fragmentation.35 6. The Case for an Integrated Approach 41 References 43 Boxes Box 1. Data Sharing in Banking From Credit Bureaus to Open Banking .27 Box 2. The European Union’s General Data Protection Regulation .38 Box 3. Data Localization 39 iiiAcknowledgments We thank Alessandro Acquisti, Tamim Bayoumi, Giovanni Dell’Ariccia, Federico Diez, Romain Duval, Dong He, Chad Jones, Ross Leckow, Martin Mühleisen, Aditya Narain, David Rozumek, and Fabián Valencia for excel- lent discussions and comments on previous drafts. Remaining errors are the authors’ responsibility. The views expressed in this paper are those of the authors and do not represent the views of the IMF, its cutive Board, or its Management. Abstract Is data the new oil of the digital economy We provide a synthesis of data’s functions in the modern economy, as an in the production function and as a means of shifting ination across agents. We consider three characteristics of data that a potential basis for policy interventions 1 non-rivalry and the associated returns to scale and scope, 2 privacy exter- nalities, and 3 partial excludability. We describe how data policy frame- works affect objectives for efficiency, equity, and stability, and discuss their implications for cross-border activities and the financial sector. An integrated approach to data policy is required to balance the complex trade-offs that arise across objectives. We identify four growing concerns that modern data policies must address, including via domestic and international cooperation among agencies. ivTHE ECONOMICS AND IMPLICATIONS OF DATAIs data the new oil Data has taken on a critical role with the rise of the digital economy. In the past decade, companies with data at the core of their business models have come to dominate the rankings of the world’s most valuable corporations. The literature that studies the economics of data spans separately growth, privacy, competition, inclusion, and financial stability aspects. This paper provides an analytical review that integrates these perspec- tives and assesses the implications of data for macroeconomic growth, equity, and stability. Its contribution is to describe the main trade-offs facing poli- cymakers as they design data policy frameworks for the increasingly complex global data economy. The proliferation of data in the economy presents a tremendous opportunity to boost growth through efficiency and innovation. But to do so without compromising other objectives, we argue that current policy frameworks must be modernized to tackle four growing challenges. First, data markets are opaque and may be leading to too much data collection and too little privacy. Rights and obligations over data must be clarified for the market to function efficiently, and the way in which these are assigned will impact growth and equity. Second, incumbents have an incentive to hoard data, potentially sti- fling competition and reducing the social benefits that could flow from wider access. A range of policies can be deployed to encourage data sharing that can promote competition and innovation. Third, it is unclear that companies are doing enough to protect the data they hold, creating risks to stability that should be mitigated with measures to ensure that all market participants invest adequately in cybersecurity. Finally, without some coordination across countries, there is a risk that global data markets could become fragmented, impeding potentially large gains from cross-border data activity including trade and finance. Introduction CHAPTER 1 1Data has long been of value in economic activity. For a colonial bank to extend a loan to a farmer, it would consult the data in the local registry to learn how much land its potential borrower owned. Ancient states con- ducted censuses to gather ination on their subjects to facilitate the task of levying taxes. Many market activities have been recorded to build trust and provide assurance on the exchange of goods and services. The collec- tion of personal data has always involved a trade-off between respecting the individual’s desire for privacyincluding from governmentand reaping the commercial and social benefits that can be derived from its collection and dissemination. What is new about data that requires policymakers to rethink its economic implications Two recent technological trends are widely recognized as hav- ing led to an explosion in the economic relevance of data. First, technolog- ical progress has drastically reduced the costs of collecting and storing data. Widespread digitalization leads to more data being produced as a byproduct of economic and social activities, including aspects of human interactions and experiences that used to be conceived as being entirely qualitative. Second, advances in analytic techniques have allowed for more advanced process- ing to extract greater value from available data. General purpose technolo- gies including artificial intelligence and machine learning have pushed the use of data across sectors, with prediction algorithms deployed to develop self-driving cars, identify promising new drugs, deliver targeted advertis- ing, and to improve the efficiency of operations. For several of the world’s most valuable publicly traded firms, data is central to their highly profitable business models. 1 1 In their report to the US Securities and Exchange Commission from July 2019, Alphabet GOOGL reported that advertising revenuesgenerated by the company’s data-driven ad targeting servicesreached 32.6 billion in the latest quarter, making up 83.7 percent of total revenue. Context and Framing CHAPTER 2 3These trends are changing how consumers, companies, and policymakers measure and analyze the economy. Previous IMF work has studied the impli- cations of big data and digitalization for the compilation of economic statis- tics IMF 2018a and for real-time policy making IMF 2017b. 2 This paper focuses on the economic characteristics of data and their impli- cations for macroeconomic growth, efficiency, and stability. Although the proliferation of data is a recent development, the economic literature con- tains many relevant insights that we draw on. At the outset, we should emphasize that while the literature offers a rich set of qualitative mechanisms and trade-offs, their quantification remains at early stages, as are efforts to describe optimal policies by incorporating multiple mechanisms into unified models. There is a long tradition of study in economics on the importance of imper- fect inationespecially when it is held asymmetricallyas a key friction undermining allocative efficiency and compounded by search and transac- tions costs Stiglitz 2002. Likewise, much has been written about the role of knowledge and ideas in driving economic growth Arrow 1962; Romer 1990. An emerging literature has focused directly on the economics of data and digitalization more broadly. Jones and Tonetti 2018 and Farboodi and Veldkamp 2019 present growth models with data in the production func- tion and study its implications for long-term growth. Acquisti, Taylor, and Wagman 2016 provide a synthesis of the insights uncovered in the literature on the economics of privacy, which has a rich tradition going back to Posner 1981. And more broadly, Goldfarb and Tucker 2019 provide a review of how digital technologies are changing economic decisions by shifting the costs of search, duplication, transportation, tracking, and verification. With the literature on data large and growing rapidly, the main contribution of this paper is to integrate perspectives to in the design of data policy frameworks. We focus on the macro-financial implications of data prolifera- tion through its impact on efficiency, equity, and stability. We start by asking What does data do in the economy We consider two functions that have been emphasized in the growing literature on the topic. First, data is an in the modern production function that firms combine with factors such as labor, capital, land, and oil to produce a wide range of goods and services, and to innovate. Second, data shifts ination across economic agents, with implications for efficiency, equity, and competition. We then argue that data has three economic characteristics that create important challenges for public policy. 2 The IMF 2017a has also explored how the combination of big data with artificial intelligence, distributed computing, and cryptography could change the provision of financial services, with a special emphasis on cross-border payments. THE ECONOMICS AND IMPLICATIONS OF DATA 4First, data is nonrival While using oil means others can no longer use it, the same data can be used by many. Like a new idea, society will benefit most from data when it is widely shared, because more users will be able to use it to increase efficiency and innovate. But while technology makes data non- rivalry possible, policies and private decisions affect whether it will be so in practice. Under current policies, it is unlikely that private firms have incen- tives to grant competitors access to the data they have collected, such that data hoarding practices may be limiting market contestability and the social benefits that could be derived from data. Second, data involves externalities The collection, sharing, and processing of personal data by one agent imposes costs on others by affecting their privacy. An implication is that a market for data lacking sufficient user control rights––where data collectors do as they please with the data they collect––is likely to lead to excessive data collection and too little privacy. A related policy challenge is to clarify the rights and obligations of participants in data markets. Third, data is only partially excludable The storage of data on interconnected systems means that controlling access to data requires continuous invest- ment to prevent its loss through cyber-attacks. A key policy question is to what extent private data collectors and processors have adequate incentives to invest in protecting their data, particularly in the case of individual data about others. There is an emerging consensus that private reputational effects are insufficient, and policy measures are needed to ensure that sensitive data is protected adequately. Effective data policy requires an integrated perspective to balance competing objectives promoting growth and competition through data access, ensuring incentives exist for data to be collected and processed, promoting stability by adequate investment in cybersecurity, and ensuring that individual privacy preferences are respected. In most cases, this will require cooperation at the national level across agencies that may not have traditions of interaction consumer protection and privacy agencies, competition authorities, minis- tries of finance and economics, statistics offices, central banks, and finan- cial regulators. It is natural that national priorities will vary across objectives, such that there can be no one-size-fits-all data policy framework. However, there is also a risk that the global data economy becomes fragmented. Because data is inherently mobile, such a scenario could stifle the potential benefits of international trade in data. We argue that international coordination is needed to achieve the minimum data policy principles that are compatible with productive cross-border data economies.Context and Framing 5This section lays out the basic ingredients needed for thinking about the economic decisions involved in a stylized market for data. When we refer to data, we refer to the factual representation of a characteristic, action, or natu- ral occurrence. Data can be quantitative or qualitative in nature, and may be stored on analog that is, paper, stone tablets or digital media. Data is a of ination, and a rich literature on the economics of inationand particularly on incomplete or imperfect ination thus offers many useful insights for thinking about the economics of data Stiglitz 2002 provides an overview. We follow Jones and Tonetti 2018 in distinguishing data from ideas, which are another of ination. The intuition is that an idea is ination that provides a set of instructions for completing a task, such as a recipe for producing a drug or a blueprint for building a machine. In contrast, data is an ingredient that can be used in the completion of a task, more akin to a raw material. This remains a broad definition of data and necessarily makes much of the discussion in the following sections
展开阅读全文