Episode Summary
Executive Summary: Catherine Tucker argues that privacy, data, and antitrust policy must distinguish among different kinds of data and markets rather than treating “tech” as one category. She says privacy regulation can have real costs, including slowing adoption of valuable technologies like electronic medical records, while also noting privacy benefits remain hard to measure. She is skeptical of claims that ad-supported platforms or data accumulation automatically imply monopoly power, emphasizing consumer anchoring to “free,” weak or local network effects, and the need for sector-specific privacy rules.
Main Topics: How Tucker entered privacy and data research (Priority: 5/5): Tucker explains she began studying network effects during her Stanford economics PhD, then turned to data sharing and privacy because privacy regulation provided an exogenous shock for measuring network effects. Privacy regulation and healthcare technology (Priority: 5/5): She discusses research showing privacy rules can slow adoption of electronic medical records and reduce the effectiveness of data-sharing in high-risk pregnancies, potentially affecting neonatal outcomes. Measuring privacy trade-offs and data longevity (Priority: 5/5): Tucker says economists can often measure privacy costs more easily than benefits, and argues that policymakers should focus more on long-lived data versus short-lived advertising data. A taxonomy for data privacy (Priority: 4/5): She and Amalia Miller propose thinking about data through three lenses: persistence, potential harm if exposed, and spillovers to others, rather than treating all data alike. Advertising-supported platforms and consumer pricing (Priority: 4/5): Tucker defends ad-supported internet platforms, citing large consumer value, and argues that moving users from free to paid models is difficult because consumers are anchored to zero-price services. Antitrust, data, and barriers to entry (Priority: 5/5): She argues data is not automatically a barrier to entry; its competitive value depends on uniqueness and whether one firm uniquely observes intent signals in a digital footprint. Platform entrenchment, network effects, and crypto (Priority: 4/5): Tucker says social media and ride-hailing incumbents may be more fragile than critics claim because of local network effects, multi-homing, and platform substitution; she is similarly skeptical that Facebook can easily leverage its base into new markets like Libra.
Key Arguments: Privacy regulation can impose real economic and human costs by slowing adoption of valuable technologies, especially where timely data sharing matters most, such as high-risk medical care. Economists are good at measuring privacy costs but have not yet done enough to quantify privacy’s benefits, making policy trade-offs hard to assess. Not all data is equal: advertising data is often short-lived and less sensitive than biometric, genetic, or health data that can persist and create future harm. A useful framework for privacy policy is to assess data by persistence, exposure risk, and spillovers to relatives or others. The advertising-supported internet creates substantial consumer value, and consumers are strongly anchored to zero-price services, making simple transitions to pay models unrealistic. Claims that data alone creates insurmountable competitive barriers are overstated because most data is non-unique and many ad signals are observable by multiple firms. Big tech should not be treated as a single market: Facebook, Google, Amazon, Uber, and others face different economic dynamics and competitive pressures. Social media network effects can be fragile because they are local and shaped by social identity; ride-hailing platforms also face multi-homing and price-based switching. Sector-by-sector privacy regulation is preferable to a universal rule because different data types warrant different levels of protection. Large platforms cannot automatically leverage a user base into a successful new market; failures like Google+ and the challenges facing Libra illustrate limits to platform extension.
Data Points: Podcast date: Monday, October 1st, 2019 - Opening introduction to the interview Value of Facebook to average user: around $50 a month - Cited from MIT research calibrating consumer value of the platform Value of digital maps: $3,600 per year - Cited as another example of digital platform value Traffic change after paywall: 98.7% drop in traffic - Tucker’s recollection of research on a newspaper adding a blunt paywall Pizza experiment: slice of pizza - In the privacy paradox study, pizza changed MIT undergraduates’ willingness to share data Time horizon for risky data: 20 years - Used to contrast short-lived advertising data with long-lasting sensitive data Duration of career experience studying the topic: nearly 20 years - Tucker notes how long she has studied these issues
Pivotal Quotes: "I think it was important we were measuring the benefits of technology there." — Catherine Tucker: On electronic medical records and the need to recognize technology’s value, not just privacy costs "I think there is a paradox, which is almost the opposite of a market failure in the provision of privacy." — Catherine Tucker: On the difficulty of defining privacy as a classic market failure because behavior and stated preferences diverge "I just think that's wrong to lump everything together in that way." — Catherine Tucker: On treating all data as a homogeneous source of competitive advantage in antitrust debates
Implications: Policymakers should avoid one-size-fits-all rules for privacy and competition. The interview suggests better policy depends on data type, data lifespan, and market structure, with more caution about unintended harms from regulation and overbroad antitrust assumptions.
About Two Think Minimum
Podcast of the Technology Policy Institute of Was…