Externalities from Digital Consumer Profiling

The emergence of externalities resulting from digital consumer profiling demonstrates how invisible data markets are established in the current state of the digital economy. There are a multitude of ways for companies to monetise a person’s behaviour through the collection and analysis of individual behavioural data generated from an individual’s interaction with a webpage

The development and use of profiling were one of the most important economic activities to develop and grow within the global arena throughout the 21st century. However, the economic value of both individual consumers paying to use these services and business organisations charged for those services does not reflect the total economic impact from this activity, which results in many different forms of externality that are not being priced by economists.

To illustrate the economic impact of this activity, in 2023, the data broker industry globally was estimated to be valued at over $300 billion, where companies (i.e. Acxiom, Oracle Data Cloud) have collected and analysed behavioural data for hundreds of millions of individuals worldwide. As such, Meta's advertising revenues, which are primarily used for behavioural profiling, were reported to have reached $131.9 billion in 2023; this was approximately 98% of the total sales revenue for Meta.

The core economic logic that can be established from these facets of data profiling is that profiling is not merely a by-product of the provision of digital services. Profiling is the primary product of the provision of digital services. The economic forces driving profiling can be viewed as a result of the properties of data that are established once data has been collected. Once data has been collected, it is non-rival (an individual can share his or her browsing history with an advertiser without losing anything), however, it is also crucially non-excludable by most users, as most users lack the technological knowledge, understanding, and capacity to utilise the capabilities to stop companies from collecting, analysing, and using this data against them. Beyond targeted advertising and price discrimination, extensive behavioural profiling also increases the potential consequences of data breaches and identity fraud, making identity theft coverage an increasingly relevant form of financial protection for consumers participating in the digital economy.Firms are able to gain an enormous amount of value from the use of data generated by the consumer, while consumers generally receive a service that is either subsidised partially or in some cases not at all. Negative externalities from this asymmetry are well-documented, as over 75% of consumers, according to a study published in the Journal of Marketing Research, suffered measurable psychological distress due to targeted advertising developed on the basis of inferred health conditions. Those costs are borne entirely by the consumer and health systems, not by the advertisers profiting from their inference. Similarly, algorithmic price discrimination resulting from profiling redistributes income regressively, effectively forcing lower-income consumers to pay more for identical products than their higher-income counterparts because of an automatic algorithm that views them as less sensitive to price as a result of an urgent need for the product.

Positive externalities from this asymmetry exist, although they are routinely overstated, as many of the efficiencies realised by companies that can more accurately understand the behaviour of their customers will reduce wasted advertising spending and redirect funding towards products that are truly relevant to those customers. By analysing consumer data, firms can better understand customer behaviour and tailor products, services, and marketing strategies to meet changing consumer preferences more effectively. Theoretically, this creates a more efficient market for customers with the potential for lower prices and/or better quality of service. Spotify, for example, is able to offer a recommendation engine that was trained on hundreds of billions of listening events to its customers to assist them in discovering artists they would not have otherwise encountered.

However, it should also be noted that the majority of the benefits from profiling will not be shared equally across all markets, as they will be concentrated within markets with significant amounts of data generated, and not make much of an impact in markets with limited data generated.

Governments have attempted to internalise the negative externalities created from the use of data through the implementation of data protection legislation, such as the General Data Protection Regulation (GDPR) which has been in effect since 2018, and placing an estimated compliance cost on companies of approximately €200 billion during its first year of implementation according to the International Association of Privacy Professionals. Although there are mixed results in the literature as to whether or not the GDPR has resulted in a reduction of the external negative costs of the practice of profiling, a 2022 research study published in the RAND Journal of Economics indicates that the GDPR reduced the amount of tracking that websites do by 12.5 percent during the period of the implemented regulation; however, the regulation also resulted in a 47 percent decline in the rate of new applications being created within the markets impacted by GDPR thus illustrating the possibility that the regulation will create its own allocative distortions.

In the United States, there is still a fragmented regulatory response, with the California Consumer Privacy Act being a stand-alone state regulation being used currently along with a patchwork of other state-specific regulations with no overall federal regulatory scheme. Therefore, it follows that the continued regulatory inconsistency acts as a negative externality for companies doing business across state lines, and creates an opportunity for arbitrage whereby companies that are operating across jurisdictions will have a disincentive to comply with any one specific regulatory regime. Therefore, the burden of regulatory compliance will disproportionately impact the smaller competitors that do not have the infrastructure in place to comply with the various compliance regulations.

Further, the most likely stable economic equilibrium may be based on first-party data that would be collected transparently with true consent. As the current practice of using third-party cookies comes to an end, the means of digital advertising are likely to evolve to utilise technologies that are focused on the preservation of privacy and confidentiality of their consumer base. As these businesses begin to utilise first-party data to create value for their consumers and themselves, an interesting dynamic will no longer reside solely in academic journal publications but will move into antitrust courtrooms.