Registration

If you are interested in receiving information about future workshops, please fill out the following survey form: here. Please feel free to share this information with anyone else who may be interested.

Workshop Schedule

Next workshop

Title	Going Public and Employee Innovation Activity
Speaker	Tony Tong
Affiliation	University of Colorado Boulder
Date&Time	Apr. 17 2024, 9:00am-10:00am, GMT+8 /Apr. 16 2024, 6:00pm-7:00pm, PST/Apr. 16 2024, 9:00pm-10:00pm, EDT
Abstract	While a prominent stream of research has studied how going public affects firm innovation, research on how it affects firm employees’ innovation activity remains scant. We address this gap by examining how going public affects where employees’ innovation activity is being directed, toward their firm’s projects (internal innovation) versus external entities’ projects (external innovation), and the mechanisms driving such changes. Applying an instrumented difference-in-differences technique to granular, activity-level data on employees’ contribution to their firms’ open source software projects and to the projects of other entities pre- and post-IPO, we show that going public decreases employees’ internal innovation activity but increases their external innovation activity. Further, these effects are moderated by employees’ internal and external collaboration ties, and are mainly driven by those who eventually depart the firm post-IPO. However, open source contribution from third-party nonemployee developers to the firm increases post-IPO, which may compensate for the reduced internal innovation by firm employees.

Past workshops

Speaker Affiliation Date	Title	Abstract
Chih-Sheng Hsieh National Taiwan University Mar. 13 2024	Collaboration in Bipartite Networks	This paper proposes a general framework for studying the impact of collaboration on team production. First, we build a micro-founded model for team production, where collaboration between agents is represented by a bipartite network. The Nash equilibrium of the game incorporates both the complementarity effect between collaborating agents and the substitutability effect between concurrent projects of the same agent. Next, we propose a Bayesian MCMC procedure to estimate the structural parameters, taking into account the endogenous participation of agents in projects. We then illustrate the empirical relevance of the model by analyzing the collaboration network of inventors in the semiconductor and pharmaceutical industries. We find that the estimated complementarity and substitutability effects are both positive and significant when the endogenous matching between inventors and patents is controlled for, and are downward biased otherwise. Moreover, we find that these effects are higher in the pharmaceutical than in the semiconductor industry. To show the importance of correctly estimating the structural model for policy analysis, we conduct a counterfactual study for an innovation incentive program. We find that the effectiveness of the innovation incentive tends to be understated when the complementarity effect is ignored and overstated when the substitutability effect is ignored. Moreover, we find that the higher the complementarity effect, the more effective the program. We also find the innovation incentive program is more effective in the pharmaceutical industry where the complementarity effect dominates the substitutability effect. We also derive the optimal incentive scheme and show that there is substantial improvement in patent innovations under the optimal incentive scheme.
Serene Qian Huang Peking University Jan. 10 2024	Navigating Emission Reduction: The Interaction of Disclosure Regulation and Policy Support in China	We examine the impact of a disclosure regulation on corporate carbon emissions and the critical factors that influence its efficacy. In 2021, the Chinese Securities Regulatory Commission (CSRC) announced a disclosure regulation urging firms to disclose, in their annual reports, the measures undertaken to reduce their carbon emissions and the outcomes. We argue that this ostensibly voluntary regulation operates as de facto mandatory due to the perceived political cost associated with non-disclosure. Using a difference-in-differences design, with the treatment firms defined as those complying with the regulation and the control firms as those indicating its inapplicability, we find a significant decrease in the treatment firms’ carbon intensity (emissions) compared to control firms. Motivated by field evidence, we examine and find that the emission reduction effect of the disclosure regulation is observed solely among firms benefiting from supportive policies that facilitate their carbon reduction efforts. Our findings underscore the importance of complementing carbon disclosure regulations with the necessary policy support.
Michael Sockin UT Austin Jan. 3 2024	Information Production in a Hybrid Economy	We analyze how state interventions impact information discovery in China’s hybrid economy. In our model, a local government makes public investment decisions based on a combination of its policy agenda and a market signal about economic fundamentals. Private firms, limited in their capacity to process information, must choose between focusing on the fundamental and the government’s policy agenda. We find that a moderate governmental response to its agenda leads firms to prioritize information about the fundamental, which enhances market-based information discovery that benefits both government and firm decision-making. In contrast, an intense government response may divert firms’ attention exclusively towards deciphering the government’s agenda. The crux of this dynamic lies in the dual accountability of the local governor, who must balance the central authority’s performance evaluation against the need to uphold household welfare. These dual roles shape the governor’s policy choices, ultimately influencing the efficacy of the market’s information discovery.
Bin Ke National University of Singapore Nov. 15 2023	Can Machine Learning Enhance Management Control Systems? The Case of Employee Selection	Theories from economics and management accounting recognize the importance of employee selection as a control system to align the objectives between firms and employees. However, despite large effort and cost spent on employee hiring processes, firms still find it difficult identifying right employees ex ante, resulting in low employee performance or high turnover ex post. This study employs machine learning models to enhance employee selection. We replicate and assess existing human-based hiring outcomes and find that the existing human-based approach suffers from poor ex post employee-firm match quality. We show that by incorporating employee turnover and performance scores, simple machine learning models are able to significantly enhance hiring quality. Our best model reduces employee turnover by over 30% and increases the percentage of high achievers by 16%.
Tong Liu National University of Singapore Oct. 25 2023	Protecting the Center: Environmental Regulation and Regional Inequality in China	This study documents substantial regional inequalities due to environmental regulation in China. In order to protect core areas in cities, the local governments of China are incentivized to enforce tighter regulations on peripheral areas that are located to the upwind direction of the core areas. As the wind patterns are relatively stable in the long run, different peripheral areas will face differential treatments, leading to substantial environmental and economic inequalities. For cities with prevailing winds, firms located to the upwind of the city center will downsize production, adopt clean technologies, or sort into the non-upwind peripheral areas with lenient regulation. As a result, the industrial value in upwind counties is 18% lower than that in non-upwind counties. In sharp contrast, for cities without prevailing winds, industrial activities are balanced in different directions around the core areas. The regional inequality grows larger over time as environmental quality becomes more important in recent years, and is more prominent in key regions with stricter regulation and enforcement. In addition, population and political power structure can also play an important role in shaping the inequality. The findings provide the first evidence on the effects of environmental regulation on inequality in a developing country, contribute to the understanding of the costs and political economy of environmental regulation, and provide useful policy insights into regulation and economic development.
Nan Jia University of Southern California Sept. 27 2023	Introducing Machine-Learning-Based Data Fusion Methods for Analyzing Multimodal Data: An Application of Measuring Trustworthiness of Microenterprises	Multimodal data, comprising interdependent unstructured text, image, and audio data that collectively characterize the same source (with video being a prominent example), offer a wealth of information for strategy researchers. We emphasize the theoretical importance of capturing the interdependencies between different modalities when evaluating multimodal data. To automate the analysis of video data, we introduce advanced deep machine learning and data fusion methods that comprehensively account for all intra- and inter-modality interdependencies. Through an empirical demonstration focused on measuring the trustworthiness of grassroots sellers in live streaming commerce, we highlight the crucial role of interpersonal interactions in the business success of microenterprises. We provide access to our data and algorithms to facilitate data fusion in strategy research that relies on multimodal data.
Lin William Cong Cornell University May 31 2023	The Rise of E-Wallets and Buy-Now-Pay-Later: Payment Competition, Credit Expansion, and Consumer Behavior	The past decade has witnessed a phenomenal rise of digital wallets, and the COVID-19 pandemic further accelerated their adoption globally. Such e-wallets provide not only a conduit to external bank accounts but also internal payment options, including the ever-popular Buy-Now-Pay-Later (BNPL). We examine, for the first time, e-wallet transactions matched with merchant and consumer information from a world-leading provider based in China, with around one billion users globally and a business model that other e-wallet providers quickly converge to. We document that internal payment options, especially BNPL, dominate both online and on-site transactions. BNPL has greatly expanded credit access at the extensive margin through its adoption in two-sided payment markets. While BNPL crowds out other e-wallet payment options, it expands FinTech credit to underserved consumers. Exploiting a randomized experiment, we also find that e-wallet credit through BNPL substantially boosts consumer spending. Nevertheless, users, especially those relying on e-wallets as their sole credit source, carefully moderate borrowing when incurring interest charges. The insights likely prove informative for economies transitioning from cash-heavy to cashless societies where digital payments and FinTech credit see the largest growth and market potential.
Wenlan Qian National University of Singapore May 10 2023	Scared Away: Credit Demand Response to Expected Motherhood Penalty in the Labor Market	We exploit a policy reform that exogenously deteriorates mothers’ job prospects. China switched from a one-child policy to two-child in 2016, which increased female workers’ childbearing and caring responsibilities. Using a leading peer-to-peer lending platform targeting college students in China, we find that loan applications from female college students decrease by 15.6% relative to male students after the reform. The drop suggests that female students can anticipate the poorer future job prospects; they reduce their expenditure and invest less in human capital accordingly. Applications for long-term and large-amount loans and loans for human capital investment purpose experience the largest decline. We also find that loan applications decrease after provincial governments’ staggered extension of maternity leaves and that the decrease is more prominent when the expected motherhood penalty is greater. The results are unlikely driven by credit supply channels.
Xueming Luo Temple University March 22 2023	Music-Motion Synchronicity: A Crossmodal Transformer Model of Customer Engagement with Social Media Videos	The use of short-form videos has been growing rapidly on social media platforms as a main content format. This paper proposes a novel framework, Music-Motion Synchronicity (MM Sync), to predict the video engagement of dance videos from a model-based perspective. Specifically, we customize a Multimodal Transformer model with crossmodal attentions to predict video engagement (i.e. share count and play count). Our proposed framework MM Sync captures the fit between two modalities (i.e. music and motion) of dance videos and can be used as an index to evaluate the video engagement. Our model demonstrates the best performance compared with unimodal models and classic deep learning models. Beyond prediction, we add interpretabilities to our model by conducting heterogeneity analyses across influencer types and music genres and by counterfactual analyses, which offer a decision-making tool of how to select music and influencers to enhance video performance. Finally, our findings and methods provide managerial implications for influencers, brands, and short-video platforms on how to improve the virality and engagement of dance videos in this booming entertainment market on social media.
Minyuan Zhao Washington University in St. Louis Feb 15th 2023	Investments as Bargaining Chips	Institutional environments are often considered exogenous in firms’ investment decisions. While the non-market strategy literature has discussed various approaches that firms may adopt to influence their interactions with institutions, such discussion is mostly absent in the analysis of firms’ location decisions. In this paper, we argue that in countries with significant government discretion, firms may obtain better treatment from the local institutions by locating R&D or manufacturing—the types of investments usually welcomed by the local governments—in the host countries. Using a sample of global patenting and litigation records of the Fortune Global 500 companies from 2007 to 2014, we find that firms tend to have a better chance at obtaining patents, or reversing unfavorable patenting decisions at patent authorities, after their increased local R&D or manufacturing presence. The results are only significant in countries with weak legal institutions, and in countries with relatively weak domestic industries. The effect of investment on firm-specific experience with local institutions thus offers an alternative explanation for location decisions that may seem suboptimal when environments are treated as exogenous.
Zhiguo He University of Chicago Jan 11th 2023	Homemade Foreign Trading	Using cross-border holding data from all custodians in China’s Stock Connect, we provide evidence that Chinese mainland insiders tend to evade the see-through surveillance by round-tripping via the Stock Connect program. After the regulatory reform of Northbound Investor Identification in 2018, the correlation between insider trading and northbound flows decays, so does the return predictability of northbound flows. The reduction of return predictability is especially pronounced among less prestigious foreign custodians and cross-operating mainland custodians, behind which mainland insiders are more likely to hide. Our analysis sheds light on the role of regulatory cooperation over capital market integration.
Shelley Xin Li University of Southern California Dec 14th 2022	Structured Sharing of Best Practices on Unstructured Information Sharing Systems: Evidence from an Enterprise Social Network	Unstructured information-sharing systems, such as enterprise social networks (ESNs), can supplement top-down knowledge transfer with peer-to-peer knowledge sharing. However, the large volume and uneven quality of peer-to-peer shared content can make it difficult to find relevant and trustworthy information. We examine data from a retailer that introduced structured sharing of best practices (SSBP), a mechanism to systematically highlight best practices from high-performing stores in the ESN it already used. This significantly increased the sales trends of the stores where it was implemented, consistent with gradual learning from SSBP. Improvement was greatest in stores that (a) perceived higher ESN information overload before the intervention, (b) had lower ex-ante offline exposure to best practices, or (c) were exposed to higher-quality posts during the intervention and was lowest in stores serving markets differing from those served by the best-practices stores. Our results also suggest positive externalities on knowledge sharing by non–best-practices stores.
Chicheng Ma University of Hong Kong Nov 23rd, 2022	The Legal Origins of Financial Development: Evidence from the Shanghai Concessions	The critical challenge to assessing the legal origins view of comparative financial development is identifying exogenous changes in legal systems. We assemble new data on Shanghai’s British and French concessions between 1845 and 1936. Two regime changes altered British and French legal jurisdiction over their respective concessions. By examining the changing application of different legal traditions to adjacent neighborhoods within the same city and controlling for an array of military, economic, and political influences, we offer new evidence consistent with the legal origins view: the financial development advantage in the British concession widened after Western legal jurisdiction intensified and narrowed after it abated.
Yanbo Wang University of Hong Kong Oct. 12, 2022	Has China’s Young Thousand Talents Program been Successful in Recruiting and Nurturing Top Chinese Scientists?	We study China’s Young Thousand Talents (YTT) Program and evaluate its effectiveness, or lack thereof, in recruiting elite expatriate scientists and in nurturing the returnee scientists’ productivity. We find that YTT scientists are generally of high caliber in research but, as a group, fall below the top category in pre-return productivity. We further find that YTT scientists are associated with a post-return publication gain across journal-quality tiers. However, this gain mainly takes place in last-authored publications and for high (albeit not top) caliber recruits and can be explained by YTT scientists’ access to larger funding and research teams. Our paper has policy implications regarding scientific talents’ global mobility, especially in the context that early-stage scientists have increasing challenges to access research funding in the US and EU.
David Yang Harvard University Sept. 21, 2022	Exporting the Surveillance State via Trade in AI	What are the international ramifications of China’s emergent leadership in facial recognition AI? We collect global data on facial recognition AI trade deals and document two facts. First, we show that China has a comparative advantage in this technology. It is substantially more likely to export facial recognition AI than other countries, and particularly so as compared to other frontier technologies. This comparative advantage may stem in part from the Chinese government’s demand for the technology to support its surveillance state — a form of “home-market” effect — as well as Chinese firms’ access to large government datasets. Second, we find that autocracies and weak democracies are more likely to import facial recognition AI from China, in particular those lacking domestic AI innovation or experiencing political unrest. No such political bias is observed in AI imports from the US or in imports of other frontier technologies from China. To the extent that China may be exporting its surveillance state via trade in AI, this can enhance and beget more autocracies abroad. Regulations of AI trade should thus be framed around regulations on products with global externalities.
Yuan Zhang University of Texas at Dallas May 25, 2022	The Bright and Dark Side of a News Paywall	We examine the implications of a news paywall for the capital market by exploiting the adoption of a digital subscription model by Caixin.com, a major financial media outlet in China as a quasi-experiment. Using a difference-in-difference research design, we find that news articles at Caixin.com have more original and detailed content and have higher associations with future earnings news after the paywall implementation than before, compared to control media outlets. However, news articles behind the paywall are associated with higher bid-ask spreads relative to other news articles. Further analyses show the paywall leads to a decrease in retail investors’ activities on social media, but an increase in forecast accuracy by analysts who are more likely to subscribe to Caixin.com. Overall, we provide novel evidence that a news paywall improves news quality but also contributes to an unlevel playing field for investors.
Eric Zheng University of Texas at Dallas Apr. 6, 2022	The Mercy of AI: Combating Natural Disasters through Lending	Natural disasters wreak economic damage and cause financial distress to victims. However, when those in need of money apply for loans, lending firms often tighten up their lending practice as a reaction to victims’ impaired repayment ability. In this study, we show how lenders’ use of AI can help those who suffer from natural disasters. Collaborating with a leading credit scoring company in China that offers AI-as-a-Service (AIaaS) for lenders, we are able to observe lenders’ use of AI services in assessing the risk of loan applications. Through a Stochastic Frontier Analysis, we find that lenders who use AI services fare better in reducing the default rate of applicants suffering from natural disasters. Notably, such a default reduction effect is more pronounced for underprivileged borrowers with lower credit scores. We explore possible mechanisms under play and discuss the societal implications of our findings.
Yanhui Wu University of Hong Kong Mar. 30, 2022	Social Media and Government Responsiveness: Evidence from Vaccine Procurements in China	This paper studies whether and how public opinion on social media affects local governments’ procurement of vaccines in China from 2014 to 2019. To identify causal effects, we exploit the regional variation in the eruption of social-media opinion on vaccine safety following sudden vaccination scandals. We find that governments in cities exposed to more intensive social media discussion increased the frequency and share of the more-transparent (open-bidding versus private arrangement) format of procurement for vaccine-related products. The effects are robust when we use early-stage social media penetration in China as an instrumental variable. Moreover, the effects are larger in cities where governments adopted more surveillance technology, the information environment was more strictly controlled, and local officials had stronger career concerns. Our overall findings shed light on the mechanisms and limitations regarding the effect of social media on policy compliance and government accountability.
Hong Ru Nanyang Technological University Feb. 23, 2022	Government Credit and Industrial Pollution	Using firm-level pollution data from Environmental Survey and Reporting Database in China and loan data from the China Development Bank (CDB), we analyze how government-subsidized credit from the CDB affects firm pollution activities from a supply chain perspective. Consistent with prior studies, we find that while SOEs in China produce substantially more pollution (e.g., emissions of chemical oxygen demand (COD), sulfur dioxide (SO2), ammonia-nitrogen (NH3-N), wastewater discharge, and waste gas emission) than private firms, CDB credit helps reduce the pollution from SOEs much more than the reductions from private firms. Moreover, we find that CDB credit to strategic industries at the top of the supply chain is associated with lower pollution levels for firms in downstream industries. We further shed light on the mechanisms underlying such CDB credit positive spillovers across the supply chain through emission abatement actions and outcomes. First, we find that CDB loans induce SOEs to use substantially less water and coals in their production than private firms. In particular, one standard deviation increase in CDB direct loan (upstream loan) is associated with an additional 2.35% (4.79%) reduction in freshwater usage and 8.4% (3.48%) decrease in coal usage compared to private firms. Second, the results suggest that CDB loans enable SOEs to reduce COD pollution through better treatment of COD {COD treatment/(COD treatment+COD emission)} compared to private firms. Our findings suggest the important role of government credit in reducing pollution across the supply chain.
Bohui Zhang The Chinese University of Hong Kong, Shenzhen Jan. 12, 2022	Knowledge is Power: A Field Experiment in Chinese and US Stock Markets	This study examines the causal impact of financial knowledge on stock pricing efficiency. We created an investor education website and conducted a field experiment by providing knowledge about the pricing implications of accounting accruals to investors in randomized stock groups via social media in both China and the US. We find that treatment stocks experience a reduction in accrual mispricing relative to control stocks, and that the effect is most pronounced when both conceptual and methodological knowledge about accruals is provided. The education effect is stronger among stocks with investors who are either less sophisticated or can study the knowledge more intensively. Finally, we document a real effect of financial knowledge on firms. Treatment firms experience a decline in discretionary accruals in the post-experiment period, especially when they are heavily owned by individual investors or less monitored by institutions.
Yue Zhang Westlake University Dec. 22, 2021	Recent progress and challenges in machine learning for Natural Language Processing (NLP)	In recent years, artificial intelligence has received increasing attention. Deep learning technology has enabled intelligent systems to perform tasks that are far more complex compared to what machines could do a decade ago. Investigating automatic understanding and generation of natural language texts, natural language processing has been a central topic of artificial intelligence since its dawn, and human language dialogue capabilities has been recognised as a major metric for evaluating artificial intelligence. Advances in the field allow intelligent systems to perform automatic speech to speech translation, question answering, essay scoring, automatic auditing and algorithm trading by news reading. The field of natural language processing has evolved since the early days of computer science, going through three main stages, where rule-based methods, statistical methods and deep learning methods dominate the literature, respectively. Researchers and engineers have seen a shift from linguistic feature engineering to parameter tuning in their daily work as the state-of-the-art approaches shift from statistical learning to deep learning. Today, deep learning not only allows natural language processing systems to perform much better on existing tasks such as syntactic parsing and automatic machine translation, but also enables new tasks to be investigated. This talk aims to introduce the field of natural language processing from a machine learning perspective, laying out the mathematical and algorithmic foundations for the major technologies of this field. The organisation follows the order of increasing complexity, which is also largely consistent with the development history of NLP technologies. In particular, an overview of major NLP tasks is introduced first, and the remaining talk evolves around machine learning methods.
Kenneth Huang National University of Singapore Nov. 17, 2021	Trade Protection and Firm Innovation: Impact from U.S. Anti-Dumping Sanctions on Innovation Output in China	This study examines how trade protections, such as anti-dumping sanctions on foreign exports, impact innovations developed by affected foreign firms. Using data on U.S. anti-dumping sanctions levied on Chinese exports from 1985 to 2015 and domestic patents developed by Chinese firms in China that are associated with the sanctioned products, we found that anti-dumping sanctions led to an increase in the overall number of domestic patents in China in technology classes that were most relevant to the sanctioned products. This result is consistent with our theory that anti-dumping sanctions enlarge the gap between the pre-innovation and post-innovation rents, thereby providing greater incentives for sanctioned firms to innovate. Furthermore, this effect was boosted by a major national-level pro-innovation campaign instituted by the Chinese central government to promote domestic innovations that further increased the post-innovation rent. However, we also found that anti-dumping sanctions decreased the production of novel patents in China that were most relevant to sanctioned products, which is consistent with the theory that anti-dumping duties reduce the resources available to firms to develop innovations. To sum up, these findings suggest that affected Chinese firms produce more innovations to escape competition and future sanctions, but that they can go only so far to produce more incremental innovations than novel innovations due to resource constraints. This study addresses the literature gap concerning how trade protections affect sanctioned firms’ innovation and generate important strategy and policy implications from such protections’ unintended consequences.
Yongxiang Wang The Chinese University of Hong Kong, Shenzhen Oct. 13, 2021	The Chinese Collectibles Bubble	Using novel, hand-collected data from the largest Chinese collectibles exchange, we examine an asset price bubble in the collectibles market in the 2010s. Because the securitized collectibles were traded outside of the exchange, the fundamental price of the securities was publicly observable. This feature, combined with plausibly exogenous shocks to the cost of information acquisition, barriers to arbitrage, Tobin tax, and maturity allows us to further examine bubble theories in ways typically possible only in laboratory experiments. Our results are broadly consistent with resale-option theory, the importance of limited arbitrage, and provide support for the external validity of several key findings of the experimental bubble literature.
Clive Lennox University of Southern California Sept. 15, 2021	A review of China-related accounting research in the past 25 years	The past 25 years have seen an exponential growth in the number of China studies in the leading accounting journals. The rise in China-related research mirrors the country’s increased importance on the global stage. We organize our review of the China literature around three central themes: 1) political and regulatory forces, 2) China’s relations with the outside world, and 3) the availability of novel data and regulatory shocks. The former two themes address research questions that are more China-centric, while the third exploits the China setting to examine questions that are less China specific. We highlight the contributions that China studies have made to the broader accounting literature, the limitations of the current literature, and we offer suggestions for future research directions.