Consumer protection in the age of big data.

Author:Helveston, Max N.
Position::Abstract through II. The Data Revolution Will Change Insurance Markets, for Better and for Worse B. Big Data's Big Benefits: Actuarial Fairness & Loss Prevention, p. 859-886

The Big Data revolution is upon us. Technological advances in the degree to which third parties can record information about individuals, along with increases in the use of predictive analytics, are transforming the way that business is conducted in practically all sectors of the economy. This is particularly true in the insurance industry, where a firm's ability to forecast the future is the central determinant of its profitability.

Scholars and the media have touted the potential benefits of Big Data analytics--it will enable businesses to tailor their practices to suit consumers' preferences and increase the efficiency of their operations. The Big Data movement's potential negative impacts, however, have garnered significantly less attention. Commentators have focused on privacy and data security concerns as the primary problems associated with Big Data analytics. There have been essentially no attempts to assess how these developments affect consumers' other interests or, more broadly, the extent to which they justify additional regulation of markets.

This Article fills this gap. It identifies eight societal interests that will be affected by insurers' uses of data--actuarial fairness, loss prevention, autonomy, non-discrimination, justice, utility maximization, privacy, and good faith--and describes how regulators could act to ensure that markets generate an optimal balance of these values. While laissez-faire regulatory approaches are superior for some types of insurance, more extensive state interventions are needed for products that are sold to individual consumers. Where additional regulation is needed, community rating rules, authorization requirements for policy modifications, and claims handling standards are the mechanisms best suited to guaranteeing that insurance markets continue to advance public interests in the Big Data era.

TABLE OF CONTENTS INTRODUCTION I. BIG DATA AND PREDICTIVE ANALYTICS IN THE COMMERCIAL SPHERE A. Defining Big Data and Predictive Analytics B. The Data Revolution and the Datafication of Commerce C. How Commercial Uses of Big Data Harm Individuals II. THE DATA REVOLUTION WILL CHANGE INSURANCE MARKETS, FOR BETTER AND FOR WORSE A. Big Data's Impact on Contemporary Insurance Practices B. Big Data's Big Benefits: Actuarial Fairness & Loss Prevention C. Advanced Insurance Analytics' Threat to Traditional Market Goals 1. Personal Liberty and Autonomy Norms 2. Anti-discrimination Norms 3. Equality Norms 4. Utility Maximization, Privacy, and Good Faith Norms D. Summary III. MEETING THE REGULATORY CHALLENGE: MODERATING INSURERS' USES OF DATA A. The Current State of Insurance Regulation and Big Data B. The Normative Goals of Regulation C. The Future of Regulation 1. The Possibility of Federal Involvement in Insurance Markets 2. The Key Components of Reform: Community Rating, Policy Content Review, and Prohibitions on Consumer Profiling CONCLUSION INTRODUCTION

The Big Data revolution is upon us. Technological advances in data collection and storage, along with increases in the use of predictive analytics, are transforming the way that business is conducted in all sectors of the economy. (1) Much attention has been given to the benefits that Big Data analytics will generate; it will provide businesses with insights about their customers, enabling them to tailor their practices to better satisfy consumers and identify ways to increase the efficiency of their operations. (2) The negative impact that this movement could have on consumers, however, is still being explored. Governmental bodies and scholars have primarily focused on the privacy and data security problems presented by businesses' use of Big Data analytics. (3) This Article is one of the first comprehensive assessments of how these developments threaten the public's interests within a specific market. It also describes how regulation could address these problems.

While the use of Big Data in any industry has the potential to bring about these harms, this Article focuses on analyzing how the data revolution will affect insurance markets. (4) Insurers, always interested in refining their predictive capabilities, have been aggressively integrating Big Data methodologies into their business operations. (5) Auto insurers have begun to directly monitor policyholders' driving practices and use this information to calibrate personalized premium rates. (6) Many casualty insurers are using data culled from social networking sites to inform their sales, advertising, and product development practices. (7) Some companies have gone as far as scrutinizing individuals' actions on social networking websites and their other online activities to evaluate the likelihood that policyholders' claims are fraudulent. (8)

The combination of recent developments in data science and insurers' existing predictive analytics practices has the potential to catalyze incredible advances in efficiency and innovation, creating tangible benefits for consumers and providers alike. But it also poses substantial threats to consumer welfare. Many of these dangers are common to all commercial uses of Big Data: aggregating large amounts of personal data increases the magnitude of security breach losses; (9) compiling information from a large number of sources makes it less likely that individuals' consent-based constraints on the use of their information will be respected; (10) and using data mining to inform sales, pricing, or employment decisions increases the likelihood that companies will violate anti-discrimination laws. (11) These issues, however, assume greater significance in the context of insurance markets. One standard response to these concerns--that market forces will punish companies that abuse Big Data analytics--has less force in insurance markets due to the high degree of uniformity across insurers. (12) Additionally, the necessary and non-substitutable nature of many insurance products prevents individuals from being able to completely withdraw from consumer insurance markets. (13)

Insurance markets possess certain characteristics that will cause insurers' embrace of Big Data analytics to threaten public interests. Competitive pressures and the increased availability of data will inevitably lead the industry to begin collecting and analyzing massive amounts of information about applicants' social and commercial behaviors. Having one's ability to obtain insurance depend on the degree to which their behaviors fall within certain parameters imposes market mechanisms on individuals' personal lives in potentially objectionable ways. At its most extreme, it would grant insurers the power to effectively compel individuals to take actions or force them to waive their rights. (14) In addition to endangering individual autonomy, allowing insurance companies to analyze this type of information would injure societal commitments to justice and equality. (15) It would convert insurance from a mechanism that mitigates the advantages and disadvantages that people have due to luck into a mechanism that exacerbates them. (16) Finally, permitting insurers to use data in this way would destroy individuals' privacy interests, generate patterns of behavior that do not maximize societal utility, and injure good faith contractual norms. (17)

State actors appear to be unaware of many of these problems. (18) Neither state governments nor the federal government have implemented rules that restrict insurers from integrating Big Data methodologies into their core operations. For the vast majority of lines of insurance, there is essentially nothing limiting the amount of data that insurers can collect about individuals and very little controlling their use of consumers' personal information. (19)

Designing regulations to address these issues is complicated by the fact that allowing insurers to conduct these types of analyses would benefit consumers in certain ways and harm them in others. Ideally, regulatory measures would permit insurers to use analytics to the extent that the associated welfare gains outweigh losses. Identifying this threshold requires a regulator to identify the different types of interests that will be affected by insurers' practices and make a normative judgment about their relative importance.

In this context, a regulator would have to weigh the gains associated with increases in actuarial fairness and risk reduction against the injury to autonomy, anti-discrimination, equality, utility maximization, privacy, and good faith norms. Analyzing the regulatory problem through this rubric shows that a laissez-faire approach is merited when it comes to insurers' uses of data in commercial lines of insurance. The state, however, must take an active role in regulating policies marketed to individual consumers. In order to be effective, regulation of consumer insurance markets will have to address how insurers may use data when performing underwriting, rate setting, policy construction, and claims management functions. The ideal regulatory mechanisms for constraining insurers' behaviors in these areas are community rating rules, authorization requirements for policy modifications, and claims handling standards. These approaches provide frameworks that can be tailored to effectuate different conceptions of the ideal balance of public values.

This Article proceeds as follows: After providing background information on the influence that Big Data and predictive analytics are having on commercial activities in Part I, Part II of this Article discusses the impact that these changes will have in insurance markets. More specifically, it will describe how allowing insurers to have unrestrained access to and use of consumer data would improve actuarial fairness and resolve several problems in insurance markets, but would injure a number of other societal interests. Part III concludes by providing an overview of the current state of insurance regulation, discussing the normative...

To continue reading