Events

Privacy-Seeking Behavior in the Personal Data Market
Speaker: Joy Wu (ISTO)

Firms are looking to commercialize, trade, and monetize the personal data they collect and receive from consumers. Internet users regularly choose to disclose and share their personal data in return for goods and services. This study examines whether a data recipient's ability to exploit data in a secondary market can motivate users' privacy behavior. An online experiment elicited individuals' willingness and reservation prices for sharing their personally-identifiable psychometric data when faced with real privacy consequences. I found that individuals' information disclosure behaviors were misaligned with their willingness to allow data recipients to monetize their data and trade with a third party. Individuals behaved more privately — by refusing to share data or by demanding greater benefits in exchange for privacy losses — when they became more aware of a data recipient's ability to sell their data for money. Moreover, when individuals considered allowing access to and exposing their data to many recipients, the privacy responses were weaker than the responses to  just one recipient's exploitation abilities.

Performance-related CEO Dismissal and Innovation Performance
Speaker: Ali Samei (TUM)

Among different types of CEO turnovers, performance-related CEO dismissals are usually a response to the request by unhappy shareholders to turn around a troubled firm. This may have negative long-term consequences for firms if the pressure to deliver short-term returns disincentivizes newly appointed CEOs from pursuing relatively risky and uncertain, but important, long-term growth strategies. In this study, using a sample of CEO turnover events among S&P 1500 firms over 18 years, we find that (only) performance-related CEO dismissals have a long-lasting negative effect on the amount of innovation a firm produces starting from the year immediately following a dismissal. Our results also show that the higher percentage of informed institutional investors, higher voting power of board directors, and the existence of a family relative of the CEO among the board of directors, will reverse or weaken the negative effect of performance-related CEO dismissal on innovation. We provide several robustness tests to rule out alternative explanations. The paper thus provides important insights into the potential negative long-term consequences of the CEO dismissals for firms and how these consequences can be mitigated.

Gender Bias in Remote Service Delivery – Evidence from Healthcare

The COVID-19 pandemic resulted in a massive shift to the remote delivery of services. In this paper, we examine if the gender of the service provider moderates this transition. Specifically, we utilize data on in-person and virtual primary healthcare to study if the gender of the physician affects the shift to virtual healthcare. We find that female physicians experienced a 5.7% larger reduction in the delivery of services via conventional methods during the pandemic. Although female providers delivered a greater extent of their services digitally (2.2%), they suffered a net decrease in the services they provided (3.4%). For female physicians, the likelihood of having a child in the household was correlated with the amount of virtual services provided. However, correlations between being a parent and digital delivery of service were absent for male physicians.  Relative to their male colleagues, female healthcare providers with lower autonomy (such as nurse practitioners and physician assistants) did not use a significantly higher amount of telemedicine. Their digital service provision was also uncorrelated with being a parent. Finally, women’s presence in firms increased remote service delivery – physicians of both genders in majority-female clinics provided significantly more telemedicine. Overall, female physicians experienced a 2.36 percentage point lower reimbursement relative to male physicians during the initial period of the COVID-19 pandemic. These results underscore the gendered difference of the shift to remote services, implications for mothers, low autonomy workers, and institutions.

Does Telemedicine Transcend Disparities or Create a Digital Divide? Evidence from the COVID-19 Pandemic

We examine telemedicine utilization during the COVID-19 pandemic. Advocates have argued that telemedicine can overcome barriers in accessing healthcare and protect patients from contracting COVID-19. Rural and poor patients, for example, would not need to make expensive and time-consuming trips to healthcare facilitates when using telemedicine. Conversely, telemedicine adoption may depend on broadband access and technology skills, which could create a digital divide and exacerbate disparities. We study these questions using data on virtual and conventional care from a large commercial insurer. Telemedicine utilization soared during the pandemic. We further find that telemedicine utilization was concentrated in urban and affluent markets. We attribute this to two factors. First, telemedicine use was correlated with broadband penetration. Second, telemedicine adoption was much higher for patients with an established healthcare provider relationship (i.e., received care in the same health system in the previous year). We also find that telemedicine utilization was lower among older patients and comorbidities; cohorts with the greatest risk of severe illness and death from COVID-19. Without further intervention, telemedicine could exacerbate existing health care disparities.

Contact Person: Lucy Xiaolu Wang

When publishers sell impressions, advertisers cannot target viewers without information from the publisher.  By withholding information from advertisers (or limiting their ability to make contingent bids), the publisher can pool impressions for a given advertiser.  Pooling impressions increases market thickness but reduces efficiency.  We show that it is optimal for the publisher to pool high value impressions, in order to maintain market competition, but separate low value impressions, to maintain efficiency. (Joint work with Stephen Morris, MIT, and Tibor Heumann, PUC Chile)

Contact Person: Marina Chugunova

Innovation under regulatory uncertainty and the role of expectations: Evidence from the U.S. drone market
Presenter: Virginia Herbst (TUM)
Discussant: Lucy Xialou Wang (Max Planck Institute for Innovation and Competition)

There and back again: Disruptive transitions in dyadic role relationships
Presenter: Maren Mickeler (LMU ISTO)
Discussant: Daniel Obermeier (TUM)

There is a well-documented gap in the observed number of scientific works produced by male and female scientists. We first document that the gender gap in observed scientific productivity may be in part explained by a gender gap in the attribution of scientific credit. We find that women are approximately half as likely to be named on any given patent or publication produced by their team as their male counterparts, and women are twice as likely to be left off the coauthor list of the most cited scientific works produced by their team, relative to the papers that receive no citations. The gender gap in attribution is found across scientific fields and career stage. We then investigate some possible explanations for the gender gap in scientific credit.

Contact person: Cristina Rujan

Abstract follows soon.

Contact Person: Lucy Xiaolu Wang

We conduct a welfare analysis of R&D subsidies and tax credits incorporating externalities, limited R&D participation and financial market imperfections, estimating the model using Finnish R&D project level data. Firms with no immediate R&D history face more severe financial market imperfections than firms with a history of R&D. The intensive, not the extensive margin of R&D is important for policy. Financial market imperfections play no role in determining R&D. Tax credits and subsidies increase R&D investments compared to laissez-faire but less than first best. Neither R&D support policy improves welfare. (Joint work with Tuomas Takalo, Bank of Finland, and Tanja Tanayama, EIB) 

Contact: Rainer Widmann

We wrote the attached draft article, Valuing the Vaccine, in the context of debates over whether contracts to COVID-19 vaccine manufacturers for less than $40 per course were overcompensating those producers. In short, we argue that the right lodestar for valuing medical innovations is social value—not compensating R&D costs—and that even with low-end estimates of social value, current prices reward developers with a small fraction of those estimates.

During the talk, I will also explain our plans to extend this argument in a broader paper, tentatively titled Valuing Medical Innovation. Although cost-based pricing has attracted significant attention from scholars concerned with U.S. pharmaceutical pricing, our arguments for value-based pricing are not limited to the COVID-19 vaccine context. In some cases, such as drugs without rigorous evidence of comparative clinical efficacy, value-based pricing suggests that current rewards are too high. In other cases, such as for preventative medicines and treatments that require lengthy clinical trials, evidence suggests that current rewards are too low. 

Contact Person: Lucy Xiaolu Wang

We investigate whether excessively optimistic beliefs play a role in the persistent demand for doctoral and postdoctoral training in science. We elicit the beliefs and career preferences of doctoral students through a novel survey and randomize the provision of structured information on the true state of the academic market and information through role models on nonacademic careers. One year later, both treatments lead students to update their beliefs about the academic market and impact career preferences. However, we do not find an effect on actual career outcomes two years postintervention.

Link to paper: https://research.upjohn.org/up_workingpapers/334/

Contact: Fabian Gaessler

How do children influence productivity, promotions, and participation in science? We investigate this question by analyzing biographies, patents, and publications for 82,094 American scientists in 1956, at the height of the baby boom. Output data indicate that mothers reach peak productivity in their mid 40s, nearly a decade after other scientists. Event studies of marriage show that mothers become more productive 15 years into marriage, when children are less work. Differences in the timing of productivity have important implications for tenure. Just 27% of academic mothers achieve tenure, compared with 48% of fathers and 46% of other women. Examining selection, we find that female scientists are more educated, half as likely to marry, one third as likely to have children, and half as likely to survive in science compared with men. While mothers who survive are positively selected, employment data indicate that a generation of baby boom mothers was lost to American science. (Joint work with Scott Kim)

Contact Person: Felix Pöge