Publications

Visual design characteristics of products play an important role in consumer preferences for many categories. However, characterization of quantification of visual design is a challenging problem. We provide a method to automatically discover and quantify visual characteristics (attributes) from image data using a disentanglement-based approach. While the deep learning literature has shown that supervision is required to obtain unique disentangled representations, ground truth visual characteristics are typically unknown in real world applications. Our method does not require such supervision, and instead uses readily available structured product characteristics as supervisory signals to enable disentanglement. No prior knowledge on design characteristics is required, yet we are able to discover human interpretable and statistically independent characteristics. We apply this method to automatically discover visual product characteristics of watches, and discover 6 human interpretable visual characteristics providing a disentangled representation. We conduct visual conjoint analysis to obtain consumer preferences over visual characteristics. Our generative method is also able to create novel visual designs that correspond to ideal points of different consumer segments.

One of the most widely discussed issues in banking regulation and research is transparency. Yet, whether depositors – banks’ most important claimholders – are affected by transparency, is an empirical open question. Analyzing US commercial banks from 1994 to 2019, we show that uninsured deposit flows are more sensitive to information about bank performance when banks are more transparent. We also link transparency to deposit rates, banks’ investment funding patterns, and profitability. In addition, we find consistent evidence from a differences-in-difference analysis using the Sarbanes-Oxley Act of 2002 as a shock to transparency. Overall, our findings demonstrate that transparency is important in shaping depositors’ behavior and highlight its potential costs.

We study how innovative firms manage their innovation portfolios after filing for Chapter 11 reorganization using three decades of data. We find that they sell off core (i.e., technologically critical and valuable), rather than peripheral, patents in bankruptcy. The selling pattern is driven almost entirely by firms with greater use of secured debt, and the mechanism is secured creditors exercising their control rights on collateralized patents. Creditor-driven patent sales in bankruptcy have implications for technology diffusion—the sold patents diffuse more slowly under new ownership and are more likely to be purchased by patent trolls.

This paper studies competition between Örms when consumers observe a pri- vate signal of their preferences over products. Within the class of signal structures which induce pure-strategy pricing equilibria, we derive signal structures which are optimal for Örms and those which are optimal for consumers. The Örm-optimal policy ampliÖes underlying product di§erentiation, thereby relaxing competition, while ensuring consumers purchase their preferred product, thereby maximizing total welfare. The consumer-optimal policy dampens di§erentiation, which inten- siÖes competition, but induces some consumers to buy their less-preferred prod- uct. Our analysis sheds light on the limits to competition when the information possessed by consumers can be designed áexibly.

The Black Lives Matter (BLM) movement, which sought to combat systemic racism and brutality against Black individuals, saw significant corporate engagement following George Floyd's murder. This corporate response was unprecedented in both scale and visibility. Companies issued public statements, with some pledging financial support for racial equity initiatives and committing to internal reviews and policy changes. CEOs like Satya Nadella of Microsoft and Tim Cook of Apple publicly denounced racial injustices. However, these actions revealed several dilemmas. Critics from both sides labeled the corporations' efforts as "woke capitalism" or mere "virtue signaling," questioning their authenticity given historical practices contradictory to their current stances. Furthermore, internal disparities became evident as some employees felt companies' public support did not translate into equitable workplace policies. For instance, claims of superficial diversity efforts and slow progression in meaningful hiring practices persisted. Additionally, companies faced the challenge of balancing stakeholder expectations, particularly younger, socially conscious employees, with broader customer bases and political pressures. These complex dynamics underscored the intricate nature of corporate engagement in social justice issues and raised questions about the substantive impact of such support.

The New York Federal Reserve (New York Fed) played a critical role during the COVID-19 crisis by implementing most of the Federal Reserve's emergency lending facilities. These facilities aimed to stabilize financial markets by providing liquidity and preventing further economic collapse. Utilizing large financial institutions as counterparties, the New York Fed could quickly revive previous programs and establish new ones. However, as the COVID-19 pandemic unfolded, there was a recognized need to diversify the pool of counterparties beyond the traditionally narrow set of large financial institutions. This diversification effort aimed to include smaller financial institutions, including those owned by veterans, women, and minorities, that were typically excluded but were essential in providing a broader market participation. The challenge for the New York Fed was to define parameters and bring on new counterparties while efficiently operating multiple new facilities and programs.

For the future of its diversity program, the New York Fed needs to address several critical questions. It must consider what the goals of its diversity efforts should be and determine how to effectively reach new counterparties. Moreover, the New York Fed must establish criteria for involving smaller and less-established financial entities to ensure the effectiveness of the programs remains uncompromised. 

In 1942 more than 110,000 persons of Japanese origin living on the U.S. West Coast were forcibly sent away to ten internment camps for one to three years. This paper studies how internees’ careers were affected in the long run. Combining Census data, camp records, and survey data, I develop a predictor of a person’s internment status based on Census observables. Using a difference-in-differences framework, I find that internment had long-run positive effects on earnings. The evidence is consistent with mechanisms related to increased mobility due to re-optimization of occupation and location choices, possibly facilitated by camps’ high economic diversity.

We model the world economy as one system of endogenous input-output relationships subject to distortions and study how the world’s input-output structure and world’s GDP change due to changes in distortions. We derive a sufficient statistic to identify distortions from the observed world input-output matrix, which we fully match for the year 2011. Our main empirical result is to determine how changes in internal distortions (affecting transactions across sectors within countries) impact the whole structure of the world’s economy and show that they have a much larger effect on world’s GDP than external distortions (affecting transactions across countries).

There is an increasingly prevalent expectation in contemporary society that employees be passionate for their work. Here, we suggest that employers and employees can have different understandings of passion that potentially conflict. More specifically, we argue that although employers may often be well-intentioned, their emphasis on employee passion may at times amount to normative control and reflect a means to attain valued work outcomes. In contrast, employees may primarily view their pursuit of passion as an opportunity to self-actualize, and thereby, view passion as an end in itself. We propose that when employees notice that these two understandings of passion diverge, they experience uncertainty in adjudicating which understanding of passion—their own or their employer’s—to privilege. Critically, employees may feel responsible for and subsequently seek ways to reduce this uncertainty, and doing so places added demands that impedes employees’ ability to perform. We discuss why employers may not necessarily recognize how their understanding of passion can create challenges for employees, and examine the difficulties employers face in attempting to resolve the tensions employees experience. Subsequently, we develop an agenda for future research that highlights how individual, organizational, and cultural differences may lead to variation in divergent understandings of passion, and the critical role managers could play in helping address employees’ uncertainty.

Investment funds that claim to focus on socially responsible stocks have proliferated in recent times. In this paper, we verify whether ESG mutual funds actually invest in firms that have stakeholder-friendly track records. Using a comprehensive sample of self-labelled ESG mutual funds (as identified by Morningstar) in the United States from 2010 to 2018, we find that these funds hold portfolio firms with worse track records for compliance with labor and environmental laws, relative to portfolio firms held by non-ESG funds managed by the same financial institutions in the same years. Relative to other funds offered by the same asset managers in the same years, ESG funds hold stocks that are more likely to voluntarily disclose carbon emissions performance but also stocks with higher carbon emissions per unit of revenue. Despite these findings, ESG funds hold portfolio firms with higher average ESG scores. We show that ESG scores are correlated with the quantity of voluntary ESG-related disclosures but not with firms’ compliance records or actual levels of carbon emissions. Finally, ESG funds appear to underperform financially relative to other funds within the same asset manager and year, and to charge higher fees. Our findings suggest that socially responsible funds do not appear to follow through on proclamations of concerns for stakeholders.