學術論文

Differentially private multiple testing procedures can protect the information of individuals used in hypothesis tests while guaranteeing a small fraction of false discoveries. In this paper, we propose a differentially private adaptive FDR control method that can control the classic FDR metric exactly at a user-specified level α with a privacy guarantee, which is a non-trivial improvement compared to the differentially private Benjamini-Hochberg method proposed in Dwork et al. (2021). Our analysis is based on two key insights: 1) a novel p-value transformation that preserves both privacy and the mirror conservative property, and 2) a mirror peeling algorithm that allows the construction of the filtration and application of the optimal stopping technique. Numerical studies demonstrate that the proposed DP-AdaPT performs better compared to the existing differentially private FDR control methods. Compared to the non-private AdaPT, it incurs a small accuracy loss but significantly reduces the computation cost.

The scalability of a marketplace depends on the operations of the marketplace platform and its sellers’ capacities. In this study, we explore one strategy that a marketplace platform can use to enhance its scalability: providing an ancillary service to sellers. In our model, a platform can choose whether and when to provide this service to sellers and, if so, what prices to charge and which types of sellers to serve. Although such a service helps small sellers, we highlight that the provision of such a service can diminish the incentives of large sellers to make their own investment, thereby reducing their potential output. When the output reduction by large sellers is substantial, the platform may not want to provide the ancillary service, and, even if it does, it may choose to set a price higher than its marginal cost to motivate large sellers to scale. The platform may also choose to strategically delay the provision of the service.

We study whether and how banks’ social performance affects depositors, who hold demandable debt with pervasive government protection. Exploiting the regulatory releases of bank performance ratings for community development and a difference-in-differences design, we find a decline in deposit growth following the release of negative bank social performance. In addition, deposits that are impacted by the negative events flow to nearby banks with high social performance. Further analyses find that the results hold similarly among insured and uninsured deposits and are primarily driven by banks with a large proportion of deposits from high-trust and pro-social counties, and in poor information environments. Overall, we contribute to the literature by documenting the importance of social performance to non-shareholder stakeholders and providing implications for bank stability.

We undertake the first empirical analysis of profit shifting by U.S. firms during foreign tax holidays. We show that foreign tax holidays have become a prevalent and powerful tax planning strategy among U.S. firms. We find that U.S. firms significantly increase their outbound profit shifting while participating in foreign tax holidays. However, we also find that profit shifting associated with tax holidays comes at the cost of increased tax uncertainty. Our results have important implications for policy making and for understanding firm behavior.

Modern consumers are concerned about not only their material payoff, but also the fairness of the transaction when making purchasing decisions. In this paper, we investigate how consumers’ inequity aversion affects a manufacturer who sources inputs from upstream suppliers. We find that, when the manufacturer sources from a single supplier or when consumers observe the manufacturer’s cost, inequity aversion hurts both the supplier’s and manufacturer’s profits. However, when the manufacturer sources from multiple suppliers and consumers do not observe the manufacturer’s cost, inequity aversion reduces both the suppliers’ and manufacturer’s margins, which significantly alleviates the double marginalization problem, increases consumer demand, and improves channel efficiency. As a result, inequity aversion benefits the suppliers, manufacturer, and consumers alike, leading to a “win–win–win” outcome. By comparing cases in which consumers observe and do not observe the manufacturer’s cost, we also find that, when faced with inequity-averse consumers, a manufacturer may find it optimal to withhold its cost information to help secure lower procurement costs from upstream suppliers.

This paper studies career spillovers across workers, which arise in firms with limited promotion opportunities. We exploit a 2011 Italian pension reform that unexpectedly tightened eligibility criteria for the public pension, leading to sudden, substantial, and heterogeneous retirement delays. Using administrative data on Italian private-sector workers, the analysis leverages cross-firm variation to isolate the effect of retirement delays among soon-to-retire workers on the wage growth and promotions of their colleagues. We find evidence of spillover patterns consistent with older workers blocking the careers of their younger colleagues, but only in firms with limited promotion opportunities.

Customer referencing is a strategy that firms can use to disclose their connections with reputable customers as a means of enhancing their own reputations. We study the capital market benefits of naming reputable nonmajor customers in firms' financial reports to provide empirical evidence on whether this form of customer referencing has important practical implications. We predict and find that firms enjoy a lower cost of equity when they engage in customer referencing in their financial reports, consistent with the argument that this form of voluntary disclosure increases investor attention and customer certification. In cross-sectional analyses, we predict and find that the benefits of customer referencing are more pronounced for firms that (1) lack major customers or reputable major customers, (2) name customers whose reputations exceed their own, and (3) face higher competition. Overall, our study provides evidence that communicating certain interorganizational connections can generate capital market benefits for disclosing firms.

While venture capital firms are increasingly relying on recommendation models in investment decisions, existing startup recommendation models fail to consider the uniqueness of venture capital context, including two-sided matching between investing and investee firms and a lack of information disclosure requirements on startups. Following the design science research paradigm and guided by the proximity principle from social psychology, we develop a novel framework called SocioLink by depicting and analyzing various relations in a knowledge graph via machine learning. Our experimental results show that SocioLink significantly outperforms state-of-the-art startup recommendation methods in both accuracy and quality. This improvement is driven by not only the inclusion of social relations but also the superiority of modelling relations via knowledge graph. We also develop a web-based prototype to demonstrate explainable artificial intelligence. This work contributes to the FinTech literature by adding an innovative design artifact—SocioLink—for decision support in the investment context.

Informative online ratings enable digital platforms to reduce the search cost for buyers to find good sellers. However, rating inflation, a phenomenon in which average rating increases and rating variance across listings decreases, threatens the informativeness of ratings. We empirically identify the consequences of rating inflation by conducting a quasi-experiment with a digital platform that exogenously changed its rating display rule in a treated neighborhood, which resulted in rating inflation. Using a differences-in-differences approach, we find that platforms benefit from one aspect of rating inflation: user purchases and seller sales increase because of the increased average rating. However, they also face negative consequences: rating inflation causes a decrease in user trial and a greater concentration of sales among popular restaurants. Overall, our results illustrate the potential consequences of rating inflation that platforms need to consider when designing and managing their rating system.