SHANE HEITZMAN, Associate Professor of Accounting

Banks must investigate and disclose suspicious financial activity to the government. These disclosures become the backbone of financial intelligence on money laundering, corruption, drug trafficking, terrorism, tax evasion, and other crimes. However, banks vary in the quantity and quality of information they are willing to offer. In this talk, I use data from the Financial Crimes Enforcement Network (FinCEN) to describe the geography, typology, and trends in banks’ suspicious financial activity disclosures, and discuss how these disclosure practices depend on attributes such as size, profitability, risk, and other factors. I discuss related trends in financial transparency laws, the externalities of financial crime and its enforcement, and practical implications for business decisions.

Shane Heitzman’s research interests span securities law, tax policy, corporate finance, governance, mergers and acquisitions, and financial crime. His research is published in top accounting and finance journals and he serves an associate editor for the Journal of Accounting and Economics and The Accounting Review. He received his Ph.D. from the University of Arizona. Prior to joining USC, Professor Heitzman was on the faculty at the University of Rochester.

ANGELA ZHOU, Assistant Professor of Data Sciences and Operations

How can we improve our decision-making over time to improve revenue or welfare? Can we achieve improvements by learning from imperfect past experiences, or collecting new data under a careful budget? Causal inference demonstrates the impacts of decisions in various areas like business, healthcare, and policy. However, it often relies on assumptions that don't hold up in real-world scenarios. In these cases, combining domain knowledge with robust optimization and machine learning techniques can make causal inference more practical and realistic. Improving sequential decision-making based on data is crucial for addressing key challenges like customer retention, managing chronic health conditions, or introducing new policies and products. In my talk, I'll discuss new ways to learn safe and personalized decision-making strategies from existing data. Though prior methods assumed (perhaps wrongly) data was collected at random, the methods I develop will allow for some human bias in prior treatment decisions. Our new approach helps identify weaknesses in decision-making strategies, develop safer and more reliable strategies, and enhance standard AI and reinforcement learning tools. I'll demonstrate our methods using a healthcare dataset, which, despite its flaws, is valuable for developing medical AI for ongoing care. Our safe and robust techniques ensure that we make the most of this valuable data responsibly.

Angela’s research interests are broadly in the design and theoretical analysis of methodology for data science in boundary-spanning impactful applications in business, policy, and healthcare. She develops tools combining optimization, operations, machine learning, and causal inference to better personalize and tailor decisions to individuals. Her work is published in both top-tier operations journals and top-tier machine learning and artificial intelligence venues. In particular, she initiated the development of robust data science methodology to salvage causal inference under weakly wrong assumptions.

Angela is an Assistant Professor at USC Marshall Data Sciences and Operations. Previously she was a research fellow at the Simons program on causal inference and a Foundations of Data Sciences Institute (FODSI) postdoc. She earned her PhD from Cornell University in Operations Research and Information Engineering. Her work has been supported by the prestigious NDSEG fellowship.

HOVIG TCHALIAN, Associate Professor of Clinical Entrepreneurship

In my research, I study how the social conversation around technologies impacts their adoption. In this talk, I look at the impact of digital technologies on craft authenticity. Webster’s word of the year for 2023 was “authentic,” the characteristic of being genuine or real. But ironically, authenticity claims hide as much as they reveal — they’re at least somewhat opaque, especially in craft. Craft has long relied on human skill and talent, which is tacit or unable to be easily codified or fully explained.

And while craft has long been used technology, its mediation by digital technologies has increased its opacity while also making it less reliant on human knowledge and skill. I address this negotiation with digital technology, an important yet overlooked research question in the study of craft authenticity. I propose a framework for how craft producers can navigate the changing tensions of craft authenticity and offer examples to help clarify its application. I conclude by discussing how the question applies across other “authenticity markets,” including AI.

EMILY NIX, Assistant Professor of Finance and Business Economics

Managers and CEOs must provide thoughtful leadership in challenging situations, from reacting to worker misbehavior to deciding how to position their firms in an increasingly polarized socio-political environment. Yet leaders often have little hard evidence to rely on when making these decisions.

This presentation will explore two challenging choices managers face. First, in an increasingly polarized sociopolitical environment, firms are more frequently publicly engaging in politically and socially controversial issues, including guns, climate change, global conflicts, and abortion rights. Such engagement may be a positive signal for value-aligned current or prospective workers but could alienate those with differing viewpoints. The proliferation of firms and universities engaging in sociopolitical dialogue raises an important question: What are the consequences for these firms from engaging in socially or politically controversial topics, particularly when it comes to their workforce?

Second, #MeToo demonstrated that workplace harassment and assault are all too common. How do these incidents impact victims, perpetrators, and the broader workforce? What role can management play in exacerbating or mitigating these costs?

My research provides answers to these questions, leveraging unique big data from tax records, the police, Indeed, and Glassdoor to provide new insights to guide future choices.

Emily Nix holds a Ph.D. in economics from Yale University. She is a labor economist who studies the economic impacts of violence against women, the gender income gap, and inequality. Her research and expertise have been featured in Bloomberg, The Economist, the Financial Times, the Guardian, NPR, the Washington Post, and more. Her papers have been published in top economics journals, such as the Quarterly Journal of Economics, Review of Economic Studies,and Journal of Labor Economics.She has served as a consultant to both the Federal Reserve and the World Bank. Nix is also an award-winning teacher who made news in 2020 when she created a DIY light board to enhance the remote learning experience for her students, and has won multiple Golden Apple awards, the Dr. Douglas Basil Award, and an Excellence in Teaching Award.