Who’s violating your privacy? Maybe it’s your smartphone provider. The developer of your laptop’s search engine. Or the manufacturer of your smart TV. Whether you know it or not, these and other companies profit by tracking your digital behavior.
Call it the economics of privacy. The subject is a longstanding focus of Alessandro Acquisti, the T. Wilson (1953) Professor in Management and a Professor of Information Technology at MIT Sloan. He’s also the leader of a new research group at the MIT Initiative on the Digital Economy (IDE): Data, Markets and Privacy.
Alessandro is a newcomer to MIT, having joined the institute in July of this year. Previously, he was a professor of IT and public policy at Carnegie Mellon University. Alessandro spoke recently with IDE contributing writer Peter Krass.
Q: Why launch a new IDE research group on privacy now?
I started working on the economics of privacy dating back to my Ph.D. days. My dissertation was on this topic. As for the IDE, the economics of privacy and consumer data is certainly part of the broader focus on the economics of digitization. Yet it has some peculiar features, and these justify a separate research group.
If the IDE is interdisciplinary, then research on privacy is even more so. That’s because understanding privacy in our digital age requires understanding many different factors: the technology behind privacy invasions and privacy protections; the tradeoffs which arise for individuals, companies and society from sharing personal information; consumers’ decisions and behavior that show the extent to which they care about privacy; and finally, the policy implications of letting markets set the rules for data collection and protection, or letting policymakers set rules when the market may have failed to achieve a desirable balance between data collection and data protection.
Q: What changes have you seen during your long involvement with privacy research?
In the more or less 20 years I’ve been working in this space, privacy issues have emerged every few years in different contexts. From cookies, back when I started 20 years ago…to social media, around 2004, 2005…to behavioral advertising, around 2007, 2008…on and on, until the main issue today: large language models [LLMs] and privacy.
There is no shortage of interesting topics in this field. The issues keep evolving. But they are still ultimately about the boundaries of public and private, and how these evolving boundaries — which are affected by technology — end up impacting the welfare of both individuals and societies.
Q: What future privacy research are you planning?
I intend to continue studying what I like to call the allocation of benefits from the data economy. To the extent that the collection and analysis of people’s data create wealth — and I believe most of us would agree that they do — how is that wealth allocated back to different stakeholders, including the very consumers who are the subjects of the data?
This is an important policy question that has been under-studied, because we tend to assume that consumers benefit from their data being collected in the form of free content. Two examples are free online searches from Google and free interactions with ChatGPT. And there is some truth to that!
However, we don’t fully understand whether consumers in these exchanges receive fair compensation for the data that was collected from them. That data fuels both these services and the organizations that provide them.
Q: You’re also interested in privacy regulation?
Yes, another area of investigation is the study of the economic impacts of privacy protections. By protections, I’m referring both to privacy regulations, such as the GDPR, and industry privacy interventions. The latter includes technologies or self-regulatory initiatives such as Apple’s ATT [App Tracking Transparency] framework.
The conventional wisdom among economics scholars is that privacy protections, by reducing the amount of consumer data available in the marketplace, create friction and reduce economic efficiency. I’ve been interested in scrutinizing the extent to which this conventional wisdom is actually correct. I wonder, for instance, whether the economic frictions which originate from the deployment or adoption of privacy protections and technologies are only short-term frictions. In the long term, however, they may lead to more privacy protection without significant economic harm to many market stakeholders.
A related area is what I refer to as privacy-preserving analytics. This term includes tools such as differential privacy, homomorphic encryption and federated learnings. All these technologies, tools and processes can help protect individuals’ data in databases while still allowing analytics to be applied to these databases. I’m interested in the extent to which the application of these tools degrades or reduces the quality of the analysis one can ultimately produce with more privacy-preserving datasets.
Q: Any field experiments you can tell us about?
Yes, there’s one I’ve been working on for several years and that just launched before I joined MIT. It’s an ambitious, large and complex online field experiment that can aid not just my research team, but also many other collaborators.
[The research] investigates the impact that either exposure to, or protection from, online tracking, targeting and advertising has on consumers’ online behavior and welfare.
This experiment is based on a client-server infrastructure we’ve created, and it includes browser extensions, email extensions and mobile apps. With the participants’ consent, this infrastructure will not only monitor their online behaviors under strict privacy and confidentiality guarantees, but also expose them to different experimental treatments and conditions.
The experiment and experimental platform were launched in May of this year. We’re currently recruiting more participants daily, and we aim to recruit several hundred participants for each of several experimental conditions. I’m excited about analyzing this data together with other IDE scholars.
· Meet Alessandro Acquisti.
· Learn about the new IDE research group: Data, Markets and Privacy.