Based on the panel data of listed manufacturing companies in China from 2010 to 2019, the artificial intelligence (AI) index is constructed using the industrial robot data provided by the International Federation of Robotics, and the two-way fixed effect model is used to test the impact of AI on the total factor productivity (TFP) of enterprises. The results show that AI significantly improves the TFP of manufacturing enterprises, and this conclusion remains valid after robustness tests and endogeneity processing. AI promotes TFP by improving the level of human capital and technological innovation, and management and operational levels positively regulate the promotional effect of AI on the TFP of enterprises. Compared with manufacturing enterprises in the central and western regions, AI boosts the TFP of those in the eastern region; compared with non-state-owned enterprises, AI boosts the TFP of state-owned enterprises; and AI significantly boosts the TFP of high-tech and non-high-tech enterprises.

We (re)evaluate the general-equilibrium effects of (environmental) policies from the perspectives of input-output networks and firm dynamics and heterogeneity. Using China’s carbon emission trading system (ETS) as an example, we find that ETS leads to more patent applications, especially the ones associated with low-carbon technologies in the targeted sectors. The effects are muted at the firm level due to selection effects, whereby only larger firms are significantly and positively affected. Meanwhile, larger firms occupy a small share in number but a large share of aggregate outcomes, contributing to the discrepancy between the effects of ETS at the individual firm and aggregate sector levels. The effects also diffuse in input-output networks, leading to more patents in upstream/downstream sectors. We build and estimate the first firm dynamics model with input-output linkages and regulatory policies in the literature and conduct policy experiments. ETS’s effects are amplified given input-output networks.

The enforcement of peer-to-peer (P2P) registration requirements in mid-2018 triggered a P2P market meltdown, highlighting the inherent challenge faced by Chinese market participants in distinguishing between genuine and fraudulent fintech firms. The difference-in-difference results suggest that the too-big-to-fail (TBTF) perception can effectively halve investor outflows and borrower outflows during periods of uncertainty. Dynamic analysis further validates the parallel-trend assumption and underscores the persistent influence of TBTF perception. Moreover, the empirical findings suggest that, in the face of a market downturn, fintech market participants become unresponsive to all other certification mechanisms, including venture capital participation, custodian banks, and third-party guarantees.

This study explores the intricate relationship between emotional cues present in food delivery app reviews, normative ratings, and reader engagement. Utilizing lexicon-based unsupervised machine learning, our aim is to identify eight distinct emotional states within user reviews sourced from the Google Play Store. Our primary goal is to understand how reviewer star ratings impact reader engagement, particularly through thumbs-up reactions. By analyzing the influence of emotional expressions in user-generated content on review scores and subsequent reader engagement, we seek to provide insights into their complex interplay. Our methodology employs advanced machine learning techniques to uncover subtle emotional nuances within user-generated content, offering novel insights into their relationship. The findings reveal an inverse correlation between review length and positive sentiment, emphasizing the importance of concise feedback. Additionally, the study highlights the differential impact of emotional tones on review scores and reader engagement metrics. Surprisingly, user-assigned ratings negatively affect reader engagement, suggesting potential disparities between perceived quality and reader preferences. In summary, this study pioneers the use of advanced machine learning techniques to unravel the complex relationship between emotional cues in customer evaluations, normative ratings, and subsequent reader engagement within the food delivery app context.

This paper represents the first attempt to show how heterogeneous shock experiences help explain the enduring scars on household future behaviors. Using a large-scale household survey with 15,652 observations combined with geospatial transportation big data, we identify a novel belief-updating mechanism through which crises may exert prolonged impacts on household asset allocation and consumption patterns. An increase in the duration of previous lockdown experience is associated with a 10.52% escalation in enhanced anxiety for future precautionary saving motivations. This experience-based learning perspective supports the resolution of long-lasting overreactions to negative shocks via belief revisions and extends to households’ consumption behaviors. The lingering effects continue to skew households' beliefs even when conditions improve. Additionally, households with different individual-based shock experiences may exhibit varying perceptions of external shocks, resulting in disparate belief revision processes.

This study examines the effects of the robotic subsidy policy in China’s manufacturing sector. The demand-side subsidy policy aims at encouraging manufacturing firms to invest in robotics by lowering the cost of purchase. Our difference-in-difference analysis reveals distributional impacts of municipality-level robot subsidies on manufacturing firms of different scales. Although the subsidy brings a 14.2% increase in the application of robot patents, the facilitated access to robotics has not transformed into new firm entries. Strikingly, new firm entry decreases by 23.5% after the policy implementation. On the other hand, robot subsidies have increased the revenue, total asset, and employment of larger manufacturing firms by 9.8%, 6.9%, and 6.7%, respectively. To interpret the mechanism, we develop a simplified framework incorporating financial frictions into a task-based model. The model reveals that idiosyncratic borrowing costs lead to an inefficient equilibrium by generally depressing automation adoption and creating automation dispersion across firms. Such ex-ante distortion results in a uniform subsidy disproportionately benefiting firms with better capital access, thus creating a trade-off in terms of efficiency: while the subsidy can enhance overall automation, it simultaneously exacerbates automation dispersion. To quantify the efficiency implications, we embed this simplified model into a dynamic heterogeneous-agent framework, calibrated to the 2010 productivity distribution, financial frictions, and robot density in the industrial sector in China. Our dynamic model reveals that a 20% robot subsidy narrows the gap between mean and optimal automation level by 22% percentage points, while raises automation dispersion by 49%. This results in a 1.23% increase in aggregate output at the cost of a 2.40% decline in TFP. This dynamic model proposes a novel mechanism that automation exacerbates capital misallocation by enlarging asset accumulation dispersion between workers and entrepreneurs. Controlling for this dynamic feedback could enhance the subsidy-induced output gain by an additional 26%

In recent decades, over 40% of dollar-denominated corporate bonds have been issued by non-US firms. Strikingly, these foreign issuers face an extra discount of 20 bps than their US counterparts. While standard risks fail to account for the discount, the Economic Policy Uncertainty index from Baker, Bloom, and Davis (2016) can explain a substantial portion of this discrepancy, consistent with uncertainty-based model calibrations. Moreover, such foreign discount (USA effect) dominates the dollar safety premium (USD effect). My findings highlight the foreign discount effect in interna- tional corporate bonds, particularly amidst escalating global economic instability and uncertainty.

How does information sharing affect consumers' usage of FinTech credit? Using a unique dataset of ``Buy Now, Pay Later (BNPL)" users from a large digital platform and exploiting a credit reporting policy change, we document that consumers significantly reduce their usage of BNPL credit when the BNPL lender becomes subject to credit reporting regulation. This reduction is more pronounced among borrowers with previous default records, who also become more disciplined in repayment behaviors, than those without such records. The decrease in BNPL usage also leads to a reduction in online consumption, supporting the financial constraint hypothesis. Our results suggest that information sharing can help alleviate overborrowing and overspending, with stronger effects observed among younger borrowers, those who previously consumed more, or those with credit cards. We also highlight the synergies between BNPL lending and Big Tech platforms' ecosystems, which imperfectly substitute for formal enforcement institutions.

This study empirically investigates the impact of dialect diversity on corporate investment efficiency under different levels of economic policy uncertainty. Our findings reveal that local dialect diversity enhances investment efficiency during stable periods, but this advantage significantly diminishes under high economic policy uncertainty. This reduction primarily arises from underinvestment and overly cautious decision-making by fragmented management during periods of turmoil. Further analysis reveals that this reduction is exacerbated by stronger internal governance, which emphasizes checks and balances, and mitigated by stronger external governance, which focuses on supervisory power. Our results remain robust when using alternative measures of main variables and employing topography as an instrumental variable.

We investigate the causal impact of a company's environmental and social (ES) risk on the borrowing costs of its peer firms (that share lending banks). Using a regression discontinuity design based on the voting outcomes of ES-related shareholder proposals in US public companies' annual meetings from 2005 to 2021, we find that the passage of ES-related proposals leads to an average increase of 38 basis points in the loan costs for peer firms in the subsequent year. The negative spillover is more pronounced for peers with lower bargaining power in their banking relations or having lower ex-ante ES scores, on credit lines rather than term loans, and during the earlier years, validating that banks indeed channel the spillover. Notably, the spillover is particularly significant if the peer firms locate in the same states as the focal firm, or when the proposals reflect a higher degree of disagreement between the proposing shareholders and the managers, or for loans issued by banks lacking prior incentives or expertise in pricing ES risks (``non-ES banks''). We interpret these findings as evidence that the passage of ES-related shareholder proposals releases new information related to peers' ES risks and especially raises the awareness of ES risks among non-ES banks, prompting them to adjust loan rates for their portfolio companies accordingly.