Publications

We construct a longitudinal data-set of news and media websites to study how online content providers adapted their responses to the GDPR over time, and whether restrictions on online tracking enforced by the regulation affected downstream outcomes such as the quantity of content those websites offer to their visitors and visitors’ engagement with such content. We find robust evidence of websites’ reactions to the GDPR in both the US and the EU. However, reactions differ between US and EU websites. We find a small reduction in average page views per visitor on EU websites relative to US websites near the end of the period of observation, but no statistically significant impact of the regulation on EU websites’ provision of new content, social media engagement with new content, and ranking in both the short-term and the long-term.
See below for working paper, and video of presentation at the FTC PrivacyCon’22.

Designing public spaces requires balancing the interests of diverse stakeholders within a constrained physical and institutional space. Designers usually approach these problems through participatory methods but struggle to incorporate diverse perspectives into design outputs. The growing capabilities of image-generative artificial intelligence (IGAI) could support participatory design. Prior work in leveraging IGAI’s capabilities in design has focused on augmenting the experience and performance of individual creators. We study how IGAI could facilitate participatory processes when designing public spaces, a complex collaborative task. We conducted workshops and IGAI-mediated interviews in a real-world participatory process to upgrade a park in Los Angeles. We found (1) a shift from focusing on accuracy to fostering richer conversations as the desirable outcome of adopting IGAI in participatory design, (2) that IGAI promoted more space-aware conversations, and (3) that IGAI-mediated conversations are subject to the abilities of the facilitators in managing the interaction between themselves, the AI, and stakeholders. We contribute by discussing practical implications for using IGAI in participatory design, including success metrics, relevant skills, and asymmetries between designers and stakeholders. We finish by proposing a series of open research questions.

Natural language processing (NLP) tools have the potential to boost civic participation and enhance democratic processes because they can significantly increase governments’ capacity to gather and analyze citizen opinions. However, their adoption in government remains limited, and harnessing their benefits while preventing unintended consequences remains a challenge. While prior work has focused on improving NLP performance, this work examines how different internal government stakeholders influence NLP tools’ thoughtful adoption. We interviewed seven politicians (politically appointed officials as heads of government institutions) and thirteen public servants (career government employees who design and administrate policy interventions), inquiring how they choose whether and how to use NLP tools to support civic participation processes. The interviews suggest that policymakers across both groups focused on their needs for career advancement and the need to showcase the legitimacy and fairness of their work when considering NLP tool adoption and use. Because these needs vary between politicians and public servants, their preferred NLP features and tool designs also differ. Interestingly, despite their differing needs and opinions, neither group clearly identifies who should advocate for NLP adoption to enhance civic participation or address the unintended consequences of a poorly considered adoption. This lack of clarity in responsibility might have caused the governments’ low adoption of NLP tools. We discuss how these findings reveal new insights for future HCI research. They inform the design of NLP tools for increasing civic participation efficiency and capacity, the design of other tools and methods that ensure thoughtful adoption of AI tools in government, and the design of NLP tools for collaborative use among users with different incentives and needs.

We present the design of a field experiment on the impact of tracking, targeting, ad-blocking, and anti-tracking technologies on consumers’ behavior and economic outcomes. The online data industry has often heralded the benefits of online tracking and targeting, particularly in the context of online advertising. Its claims are juxtaposed by the privacy concerns associated with the vast number of ad-tech companies tracking and analyzing consumers’ online behavior – often without consumers’ awareness. We use a field experiment to analyze the impact of online tracking and targeting as well as ad-blocking and anti-tracking technologies, focusing on consumers’ online behaviors (such as browsing and shopping), and their ultimate purchasing outcomes (as measured by amounts of money spent online, product prices paid, time spent on product searching, and purchase satisfaction). In this draft, we describe the rationale and motivations behind the study; the experimental design and the instrumentation infrastructure developed for the experiments; and the plans for data collection.
See below for working paper, and video of presentation at the Economics of Digital Services EODS Economics of Digital Services Inaugural Research Symposium.

Self-service machines are a form of pseudo-automation; rather than actually automate tasks, they offset them to unpaid customers. Typically implemented for customer convenience and to reduce labor costs, self-service is often criticized for worsening customer service and increasing loss and theft for retailers. Though millions of frontline service workers continue to interact with these technologies on a day-to-day basis, little is known about how these machines change the nature of frontline labor. Through interviews with current and former cashiers who work with self-checkout technologies, we investigate how technology that offsets labor from an employee to a customer can reconfigure frontline work. We find three changes to cashiering tasks as a result of self-checkout: (1) Working at self-checkout involved parallel demands from multiple customers, (2) self-checkout work was more problem-oriented (including monitoring and policing customers), and (3) traditional checkout began to become more demanding as easier transactions were filtered to self-checkout. As their interactions with customers became more focused on problem solving and rule enforcement, cashiers were often positioned as adversaries to customers at self-checkout. To cope with perceived adversarialism, cashiers engaged in a form of relational patchwork, using techniques like scapegoating the self-checkout machine and providing excessive customer service in order to maintain positive customer interactions in the face of potential conflict. Our findings highlight how even under pseudo-automation, workers must engage in relational work to manage and mend negative human-to-human interactions so that machines can be properly implemented in context.

Image generative AI (IGAI) could change how policymakers engage with the public to design public spaces, facilitating how designers translate the public’s desires into features. However, using IGAI has challenges, such as encoded biases, which might reinforce stereotypes and harm underrepresented communities. We conducted a case study to explore how using IGAI alters the co-design process of public parks through public engagement. We use data collected from interviews with immigrants discussing the Puente Hills Landfill Park design in Los Angeles, which will re-purpose a former landfill into a new public park. We use Dream Studio as a Design Probe, generating images from the interviewees’ insights and critically reflecting on the design process through internal interviews and a reflective workshop. We analyze our case in three domains: Opportunities, Risks and Challenges, and Features and Requirements. In the opportunities domain, we discuss how the enhanced translation of words to images changes the relationship between stakeholder engagement, multiplicity, and efficiency. In the risks and challenges domain, we discuss how IGAI might enhance power imbalances and biases. Finally, we reflect on what features would ease the safe and responsible use of IGAI to engage citizens in co-designing public parks.

Advocates for regulating behaviorally targeted advertisements tend to focus on ethical and legal justifications for regulation. Meanwhile, the advertising technology industry has staunchly opposed regulation by drawing on economic arguments, contending that such regulation would be harmful to advertisers, consumers, publishers, and data intermediaries alike – ultimately undermining innovation and accessibility of free products across the Internet. In this Article, we provide a critical economic perspective to the regulatory debate. We analyze the theoretical and empirical economic literature on the costs and benefits of privacy regulation in the context of behavioral advertising to evaluate the strength of economic arguments for and against regulation. We find that economic analyses traditionally used to make anti-regulation arguments are incomplete and myopic, as they focus only on the short-term effects of regulation and a subset of stakeholders affected by the online data economy, rather than how any value created by data-intensive behavioral targeting is distributed among the entire online advertising ecosystem. We argue that recent enforcement actions against ad-technology firms and movements across the world for online privacy regulations may be justifiable not only on ethical or moral grounds, but also on economic grounds. Our analysis complements and contributes to legal scholarship on privacy harms and antitrust regulation by incorporating economic literature on harms and benefits of privacy regulation across the data economy, and challenging a key assumption used by the ad-tech industry in the privacy debate: that economic arguments support anti-regulatory positions. Rather than resulting in a loss of welfare for consumers, regulation can result in a reduction of harms and a more balanced allocation of the costs and benefits of data accumulation.

We study the impact of increasing online intermediation in legacy industries. We develop a duopoly model where an intermediary platform enables firms to attract consumers from competitors in exchange for a fee. We show that the platform can induce firms to join, even when they cannot expect benefits from joining. When both firms join, they pass down the platform fee to consumers. As the popularity of the platform rises among consumers, it can raise its fee to extract a growing proportion of the benefits it creates. We test the predictions of our model by analyzing the adoption of OpenTable by restaurants in New York City. We show that after the platform becomes prevalent, restaurants that use it raise their prices by an amount equivalent to the fees charged by the platform. In contrast, we observe almost no effect of adoption on propensity to exit – a proxy for restaurants’ profitability.
See below for working paper and video of presentation at the StartML Workshop @ NeurIPS’21

Concerns regarding online tracking and excessive advertising have led to a marked increase in the adoption of Ad-Blocking tools. We conduct a field experiment to study users’ valuation of Ad-Blockers, and to study how exposing or shielding users from online advertising influences their online experiences, their attitudes towards online advertising, their valuation of ad-blocking tools, and their future usage of such tools. We find that for users currently using an ad-blocker, uninstalling them leads to a deterioration in their online experiences and lower satisfaction with recent purchases. For users that were not using Ad-Blockers, installing one led to fewer reported regrets with purchases, an improvement in subjective well-being, and a less positive view of online advertising. In terms of users’ valuation of Ad-Blockers, we observe a great degree of heterogeneity. Some users are not willing to uninstall their Ad-Blocker even if offered large payments (>$100). Conversely, a similar number of users are not willing to install an Ad-Blocker even if offered large payments. However, most users are willing to install/uninstall an Ad-Blocker in exchange for moderate payments (<$20). Our experimental treatment has a large effect on future usage of Ad-Blockers. Participants that we ask to install an Ad-Blocker are much more likely to use Ad-Blocker after the experiment ends than comparable participants in the control group. However, not all this effect can be attributed to the benefits of Ad-Blocking, as we also observed that the participants that we asked to uninstall their Ad-Blocker are more likely to continue not-using an Ad-Blocker after the experiment ends, although the magnitude of this effect was smaller.

We study the impact of the implementation of Apple’s App Tracking Transparency (ATT) framework on the Apple App Store ecosystem. We use comprehensive data on every app available in both the Apple App Store and Google Play Store ecosystems in the eighteen-month period around the implementation of ATT, and a difference-in-differences analysis to investigate whether the introduction of the privacy transparency framework affected the incentives for developers in the Apple ecosystem to create new apps, update their existing apps, or withdraw from the market. We also leverage data on the presence of Software Development Kits (SDK) in a select number of apps in each ecosystem to study how developers adapted specific functionalities in their products, such as the use of advertising platforms or payment systems. We find that the number of available apps in the Apple App Store ecosystem quickly recovers after an initial drop following the introduction of ATT. When analyzing the use of SDKs, we find a reduction in the use of Monetization and Ad Mediation SDKs, and an increase in the use of Authentication and Payments SDKs. Our results suggest developers did not withdraw from the market after ATT and instead adapted to operate under the conditions of a more protective privacy framework.

This study analyzes how scientists on Twitter react to high-visibility events. Using a longitudinal sample of 17,157 scientists and a matching-based framework, why study how unusual visibility events affect users’ subsequent behaviors and long-term visibility on the platform.

We study firm performance in the semiconductor industry after the introduction of the integrated circuit, comparing the outcomes experienced by diversifying firms and new entrants across different clusters. Over the long term, succesful firms were disproportionately Spinoffs of leading firms, or diversifiers with a transistor background. New firms in Silicon Valley were more likely to enter at the technological frontier. However, over the long term, location had no significance on becoming a top producer.

We study firm entry and inventor mobility in the Semiconductor Industry. Our results show most of the increased inventor mobility in Silicon Valley is due to inventors moving from parents to spinoffs, or from incumbents to recent entrants. Incumbents in Silicon Valley don’t seem to benefit from the greater mobility of inventors in the cluster, as they don’t hire inventors at a higer rate than incumbents in other regions, while they lose many inventors that leave to join new firms.