學術論文

Does inward foreign direct investment (FDI) promote or inhibit the technological innovation of local firms? Prior research has generated mixed views and findings on this important question. To address this puzzle, we propose that FDI spillover and crowding-out effects may apply to different phases of innovation (i.e., idea generation and R&D implementation) respectively, which then leads to differential impacts of FDI on incremental and radical innovation. The results from a panel dataset of Chinese listed firms show that industry-level FDI is positively related to incremental innovation but has an inverted U-shaped impact on radical innovation. Furthermore, we find that technology gap reinforces the positive impact of FDI on incremental innovation while making its inverted U-shaped impact on radical innovation more pronounced, and when FDI local market focus is higher, the inverted U-shaped relationship between FDI and radical innovation is steeper. These findings reconcile the inconsistency regarding how FDI may affect local innovation and provide a novel analysis framework for the FDI literature.

While prior research has emphasized the economic threats posed by political risk, it is unclear how geopolitical risk (GPR), a supranational-level risk, affects global supply chain decisions. Drawing on the political economy perspective, we posit that GPR presents both opportunities and threats for multinational corporations (MNCs), depending on the political affinity between MNCs’ home and host countries. We also identify the risk-mitigation roles of MNCs’ political lobbying and market diversification. Using panel data from publicly listed MNCs in the US, we find that host-country GPR increases MNCs’ first-tier supply base when home–host country political affinity is high, but decreases when political affinity is low. Moreover, the positive effect of high-affinity host-country GPR on MNCs’ supply base is stronger, and the negative effect of low-affinity host-country GPR is weaker for MNCs with high levels of political lobbying or market diversification. These findings enrich the international business research and political economy perspective by elucidating both the opportunities and threats of GPR, and highlight the importance of risk-coping capabilities in managing GPR. These findings also provide insights for MNCs to adapt their strategies amid GPR by leveraging home–host political affinity, engaging in political lobbying, and pursuing market diversification to mitigate geopolitical challenges.

Our paper examines analyst reports and online stock opinion articles which recommend buying stocks that, based on the literature, trade at high prices and earn low future returns ("short-leg securities"). Using a textual analysis, we test whether the justifications primarily (1) emphasize safe-haven qualities, (2) indicate exuberance, or (3) highlight lottery-like features. Our results strongly point to (3). We subsequently validate our text-based inferences through a survey of institutional and retail investors with long positions in short-leg securities. Overall, perceived upside potential appears to play a material role in driving investor demand for stocks in the short legs of anomalies.

Hate speech is a major problem on social media platforms. Automatic hate speech detection methods relying on machine learning models, which learn from manually labeled datasets, have been proposed in both academia and industry. However, there is increasing evidence that hate speech detection datasets labeled by general annotators (e.g., amateurs or MTurk workers) contain systematic bias, as they cannot effectively consider language use differences among different speakers. When such biased datasets are used to train machine learning models, the resulting models will also be biased. Unlike general annotators, experts can produce much less biased annotations. However, expert annotations cannot be efficiently obtained in large quantity. This paper bridges the gap by adopting a weakly supervised learning method for hate speech detection using a small number of expert annotations. We propose a novel design that uses contrastive learning and prompt-based learning based on large language models, incorporating a group estimator, a pair generator, and knowledge injection. Using real-world Twitter posts written by African American English speakers and other racial groups as an example, extensive experiments were conducted to demonstrate the superior performance of the proposed method. The proposed approach was also evaluated on data in the LGBTQ+ community and achieved consistent results. The study has important academic and practical implications for hate speech detection and large language models.

Problem definition: Internal theft poses a significant challenge in retail firms’ operations. Owing to a lack of effective monitoring tools, a firm cannot observe every action in daily operations of its employees, providing opportunity for wrongdoing, such as capacity and cash stealing. As a result, a common practice is to increase the price of goods to offset the loss in revenue due to the increasing threat of theft. However, we show that such practices are not optimal. Methodology/results: We model the internal theft problem in retailing as a principal-agent model, where the principal (firm) contracts an agent (retail manager) for capacity planning and daily sales. The agent is subject to moral hazard and may steal the capacity (procurement budget or company asset) before demand realization (ex ante stealing) or steal the sales revenue after demand realization (ex post stealing). We solve for the optimal capacity, price, and agent’s commission decisions to maximize the principal’s utility. We find that capacity and price decisions are not monotone in terms of the severity of moral hazards. In particular, the principal should first decrease and then increase (increase and then decrease) the price (the capacity) when ex post stealing becomes more prevalent. We also provide an optimal commission scheme to the agent, which is simple and can be easily implemented. Finally, we investigate the sensitivities of price and capacity decisions to demand uncertainties in the presence of moral hazard. Managerial implications: Simply increasing retail prices and shifting the margin to consumers to combat loss in revenue caused by internal theft can amplify the agency problem in some scenarios because it leads to a significant loss in demand and insufficient commission to the agent. Retail firms should instead focus on jointly optimizing capacity and price and providing their employees with appropriate commissions.

Problem definition: How should retailers leverage aggregate (category) sales information for individual product demand prediction? Motivated by inventory risk pooling, we develop a new prediction framework that integrates category-product sales information to exploit the benefit of pooling. Methodology/results: We propose to combine data from different aggregation levels in a transfer learning framework. Our approach treats the top-level sales information as a regularization for fitting the bottom-level prediction model. We characterize the error performance of our model in linear cases and demonstrate the benefit of pooling. Moreover, our approach exploits a natural connection to regularized gradient boosting trees that enable a scalable implementation for large-scale applications. Based on an internal study with JD.com on more than 6,000 weekly observations between 2020 and 2021, we evaluate the out-of-sample forecasting performance of our approach against state-of-the-art benchmarks. The result shows that our approach delivers superior forecasting performance consistently with more than 9% improvement over the benchmark method of JD.com. We further validate its generalizability on a Walmart retail data set and through alternative pooling and prediction methods. Managerial implications: Using aggregate sales information directly may not help with product demand prediction. Our result highlights the value of transfer learning to demand prediction in retail with both theoretical and empirical support. Based on a conservative estimate of JD.com, the improved forecasts can reduce the operating cost by 0.01–0.29 renminbi (RMB) per sold unit on the retail platform, which implies significant cost savings for the low-margin e-retail business.

Problem definition: Emergency department (ED) delay announcement systems are implemented in many countries. We answer three important questions pertaining to the operations and effectiveness of such systems by studying the public hospital network and ED waiting time (WT) announcement system in Hong Kong’s “universal” public healthcare system: (1) How many patients are aware of (and sensitive to) the ED WT announcements? (2) How sensitive are these patients to the announced WT? (3) How can the Hong Kong government improve the WT announcement system? Methodology/results: We study over 1.3 million patient visits to the 17 tier 1 public EDs. We structurally estimate the fraction of patients sensitive to the announced WT and their sensitivity to the announcements as well as patient characteristics that lead to higher sensitivity. In the patient’s ED choice decision, we estimate the trade-off between the travel distance to an ED and the expected WT at the ED. We find that 3.1% of the patients are sensitive to the announced WT, and they are willing to travel an additional 4.8 km to save one hour of waiting. Urgent patients are less likely to be sensitive to the delay announcement than less urgent patients, but those that are sensitive are more WT averse than their less urgent counterparts. Counterfactual analysis shows that the average actual WT and number of patients who leave without being seen can be reduced by 4.6% and 8.5%, respectively, by increasing the fraction of sensitive patients to 15.0% and, simultaneously, reducing the announced WT assessment window to one hour from the current level of three hours. Further improvement can be achieved by providing predicted WT information based on the current level of ED crowding or less extreme past performance—median WT rather than the currently used 95th percentile. Managerial implications: The Hong Kong government should utilize the two levers of the announcement system: the sensitive fraction of patients and information recency. Increasing the sensitive fraction can benefit the system when it is below a certain threshold level. However, administrators should exercise caution when the sensitive fraction becomes large and consider implementing additional measures to mitigate the negative effects of information delay. The sensitive group of patients can unfairly be punished for their proactiveness. Shortening the announced WT assessment window and providing predicted WT are possible alternatives that not only improve overall performance but also exhibit strong robustness to increases in the sensitive population.

Using administrative data on the Chinese National College Entrance Examination, we study how left-digit bias affects college applications. We find strong discontinuities in students’ admission outcomes at ten-point thresholds. Students with scores just below multiples of 10 make more conservative college application choices that place them into less selective colleges and majors. In contrast, students who score at or just above multiples of 10 aim at and achieve higher but are at greater risk of overshooting. The discontinuity reveals that despite the educational and labor market consequences, students’ self-evaluation based on exam scores is subject to information-processing heuristics.

For a large sample of countries, this article shows that non-banks curtail their syndicated lending by significantly more than banks during financial crises in borrower countries. Differences in the value of lending relationships explain most of the gap. Relationships with non-banks are less valuable in general and thereby do not improve borrowers’ access to credit during crises. Non-banks are also less likely to form lasting relationships with borrowers. These findings imply that the rise of non-banks could increase the importance of transaction-based lenders and exacerbate the repercussions of financial shocks.