The financial market plays a crucial role in supporting firms in advancing basic technologies. This study investigates how the investors in the A-share market in China value the firms with different layers of artificial intelligence (AI) technology. Specifically,we classify firms with different layers of AI technology into three groups-basic,mixed,and application-and examine the initial public offering (IPO) prices and the subsequent returns of these AI firms. 

Our sample consists of all AI firms that went public on China’s A-share market between 2007 and 2024. We identify the scope of AI firms by ①reviewing the “main business” and “business scope”sections of listed companies in the“cninfo.com”database to identify AI firms initially; ② reading“the main business of the issuer and changes since its establishment” in the IPO prospectuses to further identify AI business of these firms around the IPO time. The sample comprises 300 AI firms that conducted IPOs on the Shanghai and Shenzhen A-share markets,including the Growth Enterprise Market and STAR Market,from 2007 to 2024. Based on the Artificial Intelligence Standardization White Paper (2021) and ISO/IEC FDIS 22989 standards,firms are then classified into basic,mixed,and application layers. IPO initial day return is used to measure issuance pricing efficiency,and post-IPO long-term return after the initial day is used to measure price efficiency on the initial day.

Our empirical analyses reveal the following findings:①AI firms in the basic layer exhibit significantly higher returns on the IPO day,outperforming firms in the application layer by 106.4%. The evidence indicates relatively lower offering prices for firms in the basic layer; ② firms in the basic layer exhibit significantly higher post-IPO long-term returns after the initial day,with 20-day Buy and Hold Abnormal Returns (BHAR) higher by 11.5%,60-day by 19.7%,120-day by 18.4%,and 240-day by 20.8%. The evidence suggests the underpricing of firms in the basic layer on the initial day; ③ Firms in the basic layer possess more invention patents but fewer total,utility,and design patents,consistent with the explanation of higher information asymmetry in pricing basic technologies. Firms in the basic layer also show weaker pre-IPO profitability but higher future earnings growth,aligning with the explanation tied to historical issuing PE limits in China’s A-share market and investor myopia; ④ Registration-based reform significantly reduces differences in initial returns and post-IPO long-term returns after the initial day across the AI firms with different technology layers. It suggests that the reform improves pricing efficiency for firms in the basic layer. This improvement is associated with the removal of pricing restrictions under the reform and the shift of focus from profitability metrics to the quality of information disclosure.

Our study contributes to the literature by ① addressing the lack of research distinguishing basic,mixed,and application technology and examining the difference in valuation between the firms with different layers of technology and ② exploring the capital market’s support for the AI industry,a vital area for China’s economic growth. Additionally,this paper provides several practical insights for different stakeholders:① From the perspective of firms,basic technology can drive future financial performance and firms’ value growth. Therefore,companies should increase their investment in basic technology. At the same time,firms with basic technology should strengthen information disclosure to alleviate undervaluation. ② From the perspective of investors,since the IPO pricing of firms with basic technology remains relatively low,investors should enhance their understanding of the economic significance of basic technology and the value of related firms. It could facilitate the creation of higher investment returns. ③ From the perspective of policymakers,efficient pricing is important for the development of basic technology. The registration system reform has strengthened the capital market’s support for basic technology. Future policies should continue to optimize asset allocation and increase the efficiency of market pricing. The findings in this paper are based on the Shenzhen and Shanghai A-share markets in China,where the information environment and disclosure practices differ from those in the developed Western markets,which may have different IPO pricing strategies. Therefore,further research is needed to answer the question of  whether our findings can be generalized to other capital markets.

This study explores how micro,small,and medium-sized enterprise (MSME) entrepreneurs respond to external crises and the psychological implications of their strategies. While prior research predominantly focuses on firm-level responses to crises,this study highlights the unique challenges faced by MSME entrepreneurs,whose personal resources,decision-making processes,and well-being are closely intertwined with their firms’ crisis responses. Understanding these dynamics is crucial given MSMEs’significant role in the global economy and their heightened vulnerability during crises. The research provides insights into how entrepreneurs make strategic decisions under pressure and how these decisions affect their mental health,filling an important gap in the literature on adversity and entrepreneurship.

Drawing on the Conservation of Resources (COR) theory,the study distinguishes two primary adaptation strategies:investment-oriented and conservation-oriented approaches. Investment-oriented adaptation involves actively investing resources,such as launching new initiatives or adopting innovative technologies,to mitigate the effects of the crisis and create value. Conservation-oriented adaptation,on the other hand,focuses on reducing costs and conserving existing resources,such as scaling down operations or cutting employee hours,to survive the crisis. The researchers hypothesized that resource availability,such as higher educational levels,innovation experience,and government relief benefits,would drive investment-oriented adaptation,while resource loss,particularly the adverse impacts of the crisis,would lead to conservation-oriented adaptation. Additionally,the study predicted that conservation-oriented strategies would be more strongly associated with anxiety than investment-oriented strategies,reflecting the greater psychological toll of managing resource depletion.

To test these hypotheses,the research employed two complementary studies. Study 1 used a time-lagged survey of Chinese MSME entrepreneurs during the COVID-19 pandemic,utilizing archival data collected before and after the outbreak. This study provided a detailed understanding of entrepreneurs’ responses within a single crisis context. Study 2 expanded on these findings by surveying entrepreneurs from multiple countries who faced various  crises,such as economic downturns and natural disasters. This broader approach allowed the researchers to examine the generalizability of their model across different cultural and economic contexts,enhancing the robustness of their findings.

The results revealed distinct predictors for the two adaptation strategies. Investment-oriented adaptation was positively influenced by resources such as innovation experience,higher educational levels,and government relief benefits. In contrast,conservation-oriented adaptation was driven by resource loss,particularly the adverse impacts of the crisis. These findings emphasize that the two strategies are governed by different mechanisms,underscoring the relevance of COR theory in understanding entrepreneurs’ responses to crises. Additionally,the findings suggest that multiple factors must be considered to fully explain why entrepreneurs adopt specific adaptation strategies during crises.

The study also investigated the impact of these strategies on entrepreneurial anxiety. Both investment-oriented and conservation-oriented adaptations were associated with heightened anxiety,but the psychological strain was more pronounced for conservation-oriented strategies. This aligns with the idea that focusing on resource depletion and cost-cutting may exacerbate stress,whereas investment-oriented approaches may provide a sense of purpose and control. These findings offer valuable insights into the mental health challenges confronted by entrepreneurs during crises and highlight the importance of considering psychological outcomes when evaluating adaptation strategies.

Interestingly,some predictors did not perform as expected. Entrepreneurial experience was not significantly related to investment-oriented adaptation in either study,potentially because external crises like the COVID-19 pandemic involve unprecedented challenges that render prior experience less relevant. Similarly,financial resources did not significantly predict investment-oriented adaptation,possibly because entrepreneurs with substantial financial resources may feel less urgency to act proactively. These non-significant findings highlight the need for future research to explore moderators and alternative explanations,as well as to consider the context-specific nature of crises.

The two studies demonstrated consistent findings for some variables,reinforcing the validity of the theoretical framework. Innovation experience and the adverse impact of crises emerged as robust predictors of adaptation strategies across different types and contexts of crises. However,there were also contextual variations. For instance,the effects of educational levels and government relief benefits on investment-oriented adaptation were significant in Study 1 but not in Study 2,suggesting that these relationships may depend on the type and context of the crisis. 

The study has important implications for theory and practice. Theoretically,it advances the literature on adversity and entrepreneurship by integrating COR theory to explain entrepreneurs’ crisis responses,providing a nuanced understanding of how resource availability and loss drive distinct adaptation strategies.Practically,the findings highlight the importance of resource accumulation during stable periods to enable investment-oriented responses in crises. Entrepreneurs should focus on continuous innovation and enhancing their educational qualifications or those of their decision-making teams to prepare for unforeseen challenges. Policymakers can play a crucial role by providing targeted relief benefits and fostering innovation-friendly environments to support entrepreneurs during crises. Moreover,addressing the mental health challenges associated with conservation-oriented strategies through community and governmental support programs can complement resource-focused interventions,ensuring a holistic approach to helping entrepreneurs and their firms navigate crises.

The persistent disconnection between management theories and practices has been a longstanding concern in organizational research.This disconnection becomes particularly problematic in the era of Artificial Intelligence (AI),where technological advancements are fundamentally reshaping organizational elements and logic.Conducting phenomenon-based research on human-AI collaboration has emerged as a crucial pathway for advancing theoretical development and real-world impact.

This paper begins by reviewing the current state of human-AI research,identifying three emerging themes central to understanding human-AI dynamics:human reaction to AI,human-AI interaction,and human-AI collaboration.The “human reaction to AI” theme predominantly examines individual responses to AI (and AI-generated information) during episodic interactions,characterized by vignette-based scenarios where AI does not directly impact work-related outcomes.This body of literature can be categorized into two main streams:research on AI/algorithmic decision accuracy,focusing on human assessment of computational reliability and precision; and research on AI decision legitimacy,focusing on human acceptance of moral judgment.While these studies provide valuable insights into human trust formation and AI perception,they often lack ecological validity.The “human-AI interaction” theme investigates how AI directly influences work-related outcomes,also during episodic interactions.This body of research often adopts controlled laboratory settings,exemplified by studies on ChatGPT’s impact on professional writing output and human reasoning capabilities.These studies highlight the growing relevance of AI’s immediate impact on both individual and organizational performance,emphasizing the necessity of understanding how AI interacts with human work in real-time.The “human-AI collaboration” theme addresses the evolving nature of human-AI coexistence in workplaces.AI has increasingly become embedded in work processes,creating complex,real-world challenges about how humans and AI collaborate over time.This stream of research examines dynamic,longitudinal interactions where humans and AI systems reciprocally influence outcomes across multiple dimensions,including work design,learning strategies,and long-term performance.These studies,grounded in specific organizational contexts,identify theoretical mechanisms and intervention strategies through fieldwork.

To demonstrate how phenomenon-driven research advances theoretical development,this paper analyzes an exemplary ethnographic study from the Academy of Management Journal.This study reveals how researchers can construct robust theory by systematically investigating the human-machine interface as it unfolds within authentic organizational settings.It also illuminates how richly contextualized insights from immersive fieldwork can effectively bridge theoretical development with practical implications.

After reviewing existing studies on human and AI,we further revisit seminal theories in human-computer interaction literature.These classical frameworks advocate for a system-level perspective that considers technological integration within broader institutional structures and organizational dynamics.By tracing the evolution and enduring relevance of these foundational theories,we call for a theoretical approach that recognizes the deeply embedded and interconnected nature of human-AI collaboration within complex organizational ecosystems.The goal is to inspire contemporary researchers to embrace more holistic,context-sensitive approaches when investigating emerging human-AI phenomena.

Informed by the latest studies as well as grounded in cutting-edge integration of technology and organizational settings,we here propose several promising research avenues that shed light on both theoretical advancements and managerial applications.First,we encourage exploring how employee experience and collaboration with AI jointly interact,focusing on how different experience levels impact collaboration outcomes.We should further unravel how to enable employees to maximize AI’s potential as well as alleviate the negative side.Second,we suggest that future researches may investigate how AI feedback,trust,and job security concerns affect performance in complex tasks,and how adaptation to and dependency on AI may impact employees’ skill development in the longer term.Third,we encourage researchers to examine how distinct collaboration modes between AI and humans in different contexts could optimize performance.Last,researchers can delve into users’ responses to AI suggestions and explore mechanisms to balance trust in AI and independent judgments in terms of decision-making.

The era of AI requires management scholars to bridge the gap between theory and practices by updating,transforming and transcending existing research paradigms.By uncovering deeper insights into the intricate dynamics of human-AI interaction,we shall contribute to richer and more comprehensive theoretical layers and potentially great theoretical breakthroughs.

In this study,we investigate the evolution of Sino-U.S. competition in the chip equipment industry,focusing on the dynamic interplay of technological nationalism policies. We aim to understand how the United States’ restrictive measures and China’s support strategies interact over time and impact technological development in this strategically critical sector. We observe that the world has entered an era characterized by technological geopolitical uncertainty,where a “tech cold war” is reshaping global competition. The United States employs technological nationalism to restrict China’s access to key technologies,leveraging tools such as the “entity list.” Meanwhile,China counters with policies like the National Integrated Circuit Industry Investment Fund (the “Big Fund”) to strengthen its domestic capabilities. Our study critiques existing research that treats U.S. restrictions as static,one-time events. Instead,we adopt a dynamic perspective,analyzing the evolution of these policies and their long-term implications for both nations. Using the semiconductor industry as a case study,we examine this interplay at the level of specific technological fields.

To achieve our goals,we use a knowledge graph approach. We decompose global semiconductor manufacturing patents into 88 technological fields,mapped based on their co-occurrence relationships. This allows us to explore how U.S. restrictions and Chinese support policies target specific technologies. By comparing the patent knowledge networks of U.S. entity list firms and Chinese firms supported by the Big Fund,we identify patterns of co-evolution. This analysis provides insights into the overlap and divergence between the two countries’ strategies.

We find that U.S. and Chinese policies are not static but evolve over time. The United States has adopted  increasingly specific restrictions,particularly during 2018—2020. However,these measures showed reduced continuity post-2020,influenced by shifts in political leadership. In contrast,China’s support policies through the Big Fund are remarkably consistent,focusing on building self-reliance in semiconductor design,manufacturing,and packaging. We also observe that the “needle-for-needle” adversarial strategy between the two countries peaked in 2020 and subsequently declined,suggesting adjustments in both nations’ approaches.

Our findings reveal that both countries prioritize policies targeting high-centrality technologies:Primarily targeting fields with high technological importance,particularly those critical to national security. Emphasize commercially promising technologies rather than exclusively addressing “choke-point” areas. Despite rhetoric about tackling weaknesses,we find that China invests more heavily in areas where it already has competitive strength.

Our analysis highlights that technological,market,and political factors shape policy decisions differently:Policies from both nations strongly focus on high-centrality fields in the knowledge network. Competitive fields attract more intervention from both countries. However,U.S. policies exhibit a greater focus on restricting scientific innovation,while China emphasizes supporting areas with immediate market potential. While U.S. measures show discontinuities influenced by political cycles,China maintains a steady trajectory of support,reflecting its centralized approach to industrial policy.

We find that U.S. and Chinese strategies exhibit a moderate degree of co-evolution. While both countries target similar technological areas,their approaches reflect different priorities. The United States emphasizes maintaining its dominance in critical areas,while China seeks to reduce its dependence on foreign technologies. Our results suggest that while the term “decoupling” implies complete separation,the reality is more nuanced. Both nations exhibit overlapping interests,leading to a complex dynamic rather than outright divergence.

We propose several avenues for future investigation:①Policy Drivers: Understanding the factors that influence countries’ choices of target technologies. ②Firm-Level Responses: Examining how firms navigate conflicting pressures from domestic support and foreign restrictions. ③ Micro-foundations of Decoupling: Analyzing how firm strategies and technological developments contribute to broader patterns of decoupling.

In conclusion,we have shown that Sino-U.S. technological competition is a dynamic and evolving process. We provide evidence of co-evolution in their policies,with both nations targeting high-centrality and competitive fields. While the United States demonstrates fluctuations in its restrictions,China maintains a consistent trajectory of support. This interplay shapes the global semiconductor industry and has far-reaching implications for technological development. Our findings highlight the importance of understanding these dynamics at the level of technological fields rather than industries. By doing so,we reveal the subtleties in how policies are formulated and their impacts on innovation. We believe this approach can provide valuable insights for policymakers and researchers alike. We invite future studies to build on our methodology and findings to deepen the understanding of technological nationalism and its role in shaping global technological landscapes.

The past decade has witnessed a rapid growth of money supply in China and consequently a large amount of money has flown into the real estate market leading to a real estate bubble,and shifting the economy from real to virtual.This has raised special attention from policy makersto professional economists and also has ignited heated debate on the direction of monetary flows and the efficacy of monetary policy.

Following Brunnermeier and Sannikov (2016),we construct a stochastic growth model with endogenous asset portfolio choice.In the model,each household owns productive capital and is subject to idiosyncratic shocks.Constrained by financial frictions,households are unable to insure against such idiosyncratic risks in full,so they need to do diversification through careful asset allocation.Money is risk-free and functions as a store of value,as illustrated in the conventional literature.Our innovation relative to the previous work is to include a (risky) bubbly capital for the purpose of examining the dynamic structural change of the real sector and virtual economy.The bubbly capital does not create real wealth but gives its holder“money illusions”.The return rate for bubbly capital is monotonically increasing in its social average investment rate.That is,the more household purchase the bubbly capital,the higher its return.

The effect of money growth on the economy has two competing forces.The first is the aggregate effect,i.e.,an increase in the money growth rate leads households to hold more non-money assets,thus pushes up the aggregatesavings,accumulates both productive and bubbly capital and finally promotes economic growth.The second is called the structural effect,i.e.,an increase in the money growth rate reduces the proportion of productive capital relative to bubbly capital via household portfolio choice and therefore worsens the phenomenon of“making the economy shift from real to virtual”and finally slows down economic growth.Risk is an important factor to determine the relative magnitude between these two effects.When the bubbly capital is very risky (more volatile),the aggregate effect looms large and increasing the money growth rate may boost economic growth.When the bubbly capital is not very risky,the structural effect becomes the dominant force.In this case,investing in the bubbly capital is more profitable and inflation can only transform more capital to become bubbly rather than productive and eventually hinders economic growth.Further policy simulation results show that the combination of macro-prudential policy tools and monetary policy can prevent the economy from shifting to virtual,as well as maximize social welfare.More precisely,①when the volatility of bubbly capital is sufficiently high and exceeds some bound,there exists an optimal money growth rate maximizing the social welfare; ②as long as the volatility of the bubbly capital is below this bound,increasing money supply results in welfare loss.We use the Chinese housing market as a laboratory to conduct experiments.We find that taxing on the real estate could level up the effectiveness of monetary policy through cutting down the return on holding real estate and thus suppressing the structural effect of increasing money growth.Further welfare analysis illustrates that there may exist a combination of positive real estate tax rate and positive money growth rate which maximize the total social welfare.

Our paper makes contributions in several aspects.Firstly,the paper analyzes the reason and the underlying mechanism behind the phenomenon of“making the economy to shift from real to virtual”from the perspective of risk.We find from the comparative statistics that the risk of bubbly capital is key to understanding this phenomenon.The existing literature only considers the return rate rather than the risk of investment.Secondly,the paper embeds portfolio choice into an economic growth model and highlights the dual effects of money growth rate on economic growth,i.e.,the aggregate effect and the structural effect.The existing literature mainly considers a single effect.For example,Brunnermeier and Sannikov（2016）only emphasize the aggregate effect by showing that an increase in the money growth rate could decrease the riskless return rate,thus leading households to be exposed more to risky investment and promoting economic growth.However,after introducing another bubbly capital into the economy,we find that increasing money growth rate would change the Sharpe ratio of productive and bubbly capital and generate a structural effect,which pushing the economy to shift from real to virtual from bad to worse.Thirdly,the paper constructs aneconomic growth model with portfolio choice to offer a unified framework for exploring the effects of the combination of macro-prudential policy tools and monetary policy from a long-run growth perspective.The existing papers analyzing the macro-prudential policy are mainly based on the DSGE framework,so they are unable to discuss either the growth issue or the asset allocation problem.The framework in this paper can examine the interactions between portfolio choice,optimal monetary policy and macro-prudential policy.

In the modern international monetary system,only very few international currencies with the function of storing value and as the medium of exchange are the central bank currencies (notes and coins),and most are government bonds and commercial bank deposits of the countries that issue international currency.From 1976 to 2022,foreign holdings of U.S.financial assets exhibited the following characteristics.The largest proportion of foreign official holdings of U.S.securities are U.S.government bonds.Most foreign official holdings of U.S.government bonds are treasury bonds.The largest proportion of foreign private holdings of U.S.securities are U.S.corporate bonds and stock other than Treasury securities.The second is U.S.liabilities reported by U.S.banks and securities brokers.The third is U.S.liabilities to unaffiliated foreigners reported by U.S.non-banking concerns.The fourth is U.S.Treasury bonds.

In a closed economy,the government can obtain government bonds seigniorage,and commercial bank can obtain commercial banks seigniorage.In an open economy,the governments of countries that issue international currencies can gain government bonds seigniorage from foreign-owned government bonds,referred to as government bonds international seigniorage,and commercial banks can gain seigniorage from foreign-owned deposits,referred to as commercial bank international seigniorage (Canzoneri et al.,2013; Cova et al.,2013).Therefore,international seigniorage gained by countries that issue international currencies includes central bank international seigniorage,government bonds international seigniorage,and commercial banks international seigniorage.

The Federal Reserve Banks do not pay interest on Federal Reserve bank notes in circulation outside the U.S.The international seigniorage gained by the Federal Reserve is calculated with the opportunity cost of seigniorage every year,i.e.,the market yield of U.S.Treasury securities at 1-year maturity.The outstanding stock of Federal Reserve bank notes outside U.S.comes from the International Investment Position (IIP) of the United States.

Government bonds international seigniorage.Foreign holders of large amounts of U.S.government bonds used for storing value both increased the price of U.S.government bonds and reduced the cost of financing (Greenspan,2005; Craine and Martin,2009; Francis and Warnock,2009; Krishnamurthy and Vissing-Jorgensen,2012).Government bonds international seigniorage is the total of the interest expense saved from financing in international financial markets at a low cost by the government.The categories of foreign-owned U.S.government bonds are very diverse,and the detailed statistics of the balance every year are difficult to gather.Here the data on foreign-owned U.S.government bonds every year (including treasury bonds and other government bonds) comes from the International Investment Position (IIP).Some reports have presented a numerical value for the cost of financing saved by U.S.government bonds (Greenspan,2005; Craine and Martin,2009; Francis and Warnovk,2009).However,the level of interest rates of U.S.government bonds is either high or low in different periods.The cost of financing saved may be underestimated in a period with a high interest rate,and overestimated in a  period with a low interest rate.In order to overcome this deficiency,a relatively precise way has been adopted in this paper,where the cost of financing saved by U.S.government bonds account for 10 percent of market yield on U.S.Treasury securities at a 1-year constant maturity.In normal circumstances,market yield on U.S.Treasury securities at a 1-year constant maturity is around 5 percent,and the cost of financing saved is about 50 p.m.,which is relatively close to the empirical analysis results (Greenspan,2005; Craine and Martin,2009; Francis and Warnock,2009).

Commercial bank international seigniorage.In the dollar-dominated international monetary system,foreign holding of commercial banks deposits reduces the financing costs of U.S.commercial banks.International seigniorage gained by U.S.commercial banks is the interest expense saved from the international market due to lower financing cost.It equals the amount of foreign-owned liabilities of commercial banks,multiplied by the savings in financing cost.In this paper,the data of foreign-owned liabilities of commercial banks come from the IIP statistics of U.S.liabilities reported by U.S.banks and securities brokers.Following the same reasoning above,the financing cost saved by U.S.commercial banks equals about 12% of the market yield on U.S.Treasury securities with a 1-year maturity.

The international seigniorage gained by the U.S.includes international seigniorage gained by the Federal Reserve,government bonds international seigniorage,and commercial bank international seigniorage.The proportions of the three components had small variations but remained roughly equal during 1976 to 2013.First,the international seigniorage gained by commercial banks is the largest component,with the mean,the maximum,and the minimum at 37.45%,47.81%,and 27.05%,respectively.The second largest is the international seigniorage gained by the Federal Reserve with the mean,the maximum,and the minimum at 34.47%,44.44%,and 24.28%,respectively.The least is the government bonds international seigniorage with the mean,the maximum,and the minimum at 28.67%,40.38%,and 22.80%,respectively.

Analysis of the international seigniorage gained by the U.S.shows that the central bank’s international seigniorage is only a part of the international seigniorage.Commercial bank international seigniorage and government bonds international seigniorage are equally important.

Innovation is a central part of economic growth.Existing theories elaborate on the causes and consequences of firm innovation,but less discuss the determination of the direction of firm innovation.How innovation decisions are made and whether the direction of innovation is steered are questions well worth answering.There is often a lack of incentives in the market for the production of high-cost,low-polluting technologies.Demand for cleaner technologies is always low.This is an important manifestation of market failure.Firms prefer the low-cost but highly polluting production technologies to the costly green production technologies.Relying on market forces alone cannot force enterprises to abandon highly polluting production technologies and switch to cleaner technologies.Strict environmental protection policies will help to reverse this market failure.China has gradually introduced four batches of inspections since July 2016,with each batch lasting one month,which is named Central Environmental Protection Inspectorate (CEPI).The policy not only raises the penalties for polluting behavior of enterprises in the system development,but also realistically supervises enterprise pollution,which effectively reduced pollution.The CEPI will have a twofold impact on enterprises’ innovation.One is to increase the relative cost of using polluting technologies for enterprises,forcing them to innovate in a green way.The second is to increase R&D investment in cleaner technologies,crowding out the R&D resources that firms invest in other technologies.This paper explored how the environmental regulation policy affects the direction of firms’ technological innovation,and found that the central environmental inspection policy changes firms’ innovation decisions.Firms always react to the change of costs and benefits.Under the strict environmental regulation policy,firms re-examined the benefits of adopting different technologies.That increased the demand for cleaner technologies,which spurred the effect of promoting innovation shifts.This paper firstly demonstrated the promotional effect of the environmental policy on the “Green Patents” with General DID.This paper secondly set other “non-green patent” that are not related to environmental protection as a control group.Environmental policy reversed the low demand for cleaner technologies in the market.The applications for “non-green patent” increased by 2.74% after the CEPI,while “Green Patents” increased by 3.39% more.Strict policies can reverse the problem of insufficient demand for clean technologies and inhibit the short-sighted behavior of enterprises blindly pursuing low-cost technologies.

“Green Patents” had an overall lower output size than other “non-green patents” before the policy intervention,and this trend remained stable over the 21-month period before the policy intervention.After the policy intervention of the CEPI,this difference has changed significantly.The portion of “green patents” that was originally lower was gradually filled in,and the difference between the two gradually disappeared.In the long term,the effect of the central environmental protection inspections will remain in place until 2021.This suggests that short-term policy shocks can have a long-term impact after a period of effective implementation.

At the same time,we also found that the shift in technological innovation brought about by environmental policies comes at a cost.In this paper,based on the semantic analysis of patent applications,the patents were categorized into six groups according to the remoteness of semantic similarity:95%~100%,75%~95%,50%~75%,25%~50%,5%~25%,and 0%~5%.The results showed that enterprises have adjusted their R&D allocation to meet the requirements of environmental protection inspections,squeezing out some of the R&D of other types of patents.More resources are allocated to environmental protection-related technologies,showing the “crowding out effect”.The cost of environmental protection inspections to reduce pollution is to force some patents to be replaced by cleaner technologies.

This paper also did a comparative analysis using the IPC classification.The main conclusion that policy promotes innovation steering still holds when the IPC classification is used.However,in comparison,the innovation steering effect produced by policy is significantly higher when measured based on the text analysis method.The conclusions drawn from the method based on textual analysis were more reliable.Because some of the technologies of synergistic pollution reduction are not counted when counting green patents with the IPC classification method.In addition,applicants have a tendency to categorize technologies when applying for patents.Artificial categorization will bring about greater errors and deviate from the problem that the technology is actually intended to solve.

This paper supplemented the robustness test on the definition method and observation window.Changing the way of defining green patents does not change the main conclusions.Narrowing the observation window to 2015 to 2018,the central environmental protection inspection policy can still lead to faster growth of green invention patents.In both the short and long term,the CEPI had led to a change in the cost-benefit relationship of technology,triggering a reorientation of corporate innovation.

This paper adds new evidence to the benefits of environmental regulation and expects to inform policymakers’ decisions.This paper employs textual analysis to provide a new perspective for research related to innovation,hoping to avoid the deception and arbitrariness in the application of scientific and technological achievements.That will improve the accuracy and efficiency of the research on China’s innovation output.

This paper investigates the influence of algorithmic decision-making on employees’ perceptions of fairness within the context of organizational management,particularly in light of the rapid advancements in artificial intelligence (AI).Leveraging the Stereotype Content Model (SCM),the study explores differential fairness perceptions between algorithmic and human decision-makers,particularly in adverse outcome scenarios.Findings suggest that algorithms are generally viewed as fairer than human managers,with perceptions influenced significantly by the type of task being evaluated.

As AI technologies continue to permeate various business operations,organizations increasingly deploy algorithms for diverse managerial functions,including human resources management,task allocation,and performance evaluations.This shift raises critical questions about how employees perceive the fairness of decisions made by algorithms as opposed to human managers.Given that fairness perceptions are pivotal to employee satisfaction,organizational commitment,and performance,understanding these dynamics is essential for effective organizational management.

Existing research on algorithmic decision-making offers mixed insights into its impact on perceived fairness.Some studies argue that algorithms,by relying on data and objective models,minimize human biases and enhance fairness.Others,however,highlight potential shortcomings such as neglect of qualitative information and lack of transparency,which can lead to perceived unfairness.Additionally,the literature suggests a gap in understanding the effects of decision outcomes on fairness perceptions,particularly when outcomes are unfavorable,thus forming the basis of this study’s inquiry.

This study employs scenario-based experiments to examine how employees perceive fairness when confronted with unfavorable decisions executed by either algorithms or human managers.These experiments are designed to cover a range of task subjectivities,from highly objective,data-driven tasks to those requiring significant human judgment and intuition.

The experimental results reveal that decision-maker type and task nature significantly affect fairness perceptions.In scenarios involving objective tasks,algorithms are perceived as fairer due to their presumed impartiality and lack of human error.For subjective tasks,algorithms are still viewed more favorably,but this is attributed to human managers being perceived as potentially indifferent or lacking empathy.This dichotomy underscores the complexity of fairness perceptions and suggests that while algorithms may excel in objectivity,they may fall short in areas requiring emotional intelligence.

This research adds depth to the discussion on AI’s role in management by delineating how task type and outcome influence fairness perceptions differently under algorithmic versus human decision-making.It offers insights that could help organizations better integrate AI into their management practices,ensuring that fairness perceptions are carefully managed to maintain employee satisfaction and performance.

Future researches could broaden the investigation into other organizational contexts and include longitudinal studies to assess how fairness perceptions evolve with long-term exposure to algorithmic decision-making.Moreover,further studies could explore how increased transparency and employee involvement in algorithm development might enhance trust and fairness perceptions,fostering a more equitable organizational environment.

Relational structures consisting of different types of interactions among several groups of entities are very common nowadays.As a useful tool for analyzing this type of data,multi-layer networks have gained increasing attention in recent years due to their ability to capture the complexity of real-world systems.Multi-layer academic networks are a specific type of multi-layer network that consists of multiple layers of relationships among academic entities,such as researchers,institutions,papers,or journals.Typical examples of multi-layer academic networks include the collaboration network that represents co-authorship relationships among researchers,the citation network that represents citation relationships among papers,and the journal citation networks that represent citation relationships among journals.They have been used for various purposes,such as identifying research areas,evaluating research impact,predicting scientific trends,studying the diffusion of scientific knowledge,and supporting science policy and decision-making.Overall,multi-layer academic networks provide a powerful tool for understanding and analyzing the complex relationships that underlie academic communities and their impact on scientific knowledge production and dissemination.

In this work,we collect data from 42 statistical journals published between 1981 and 2021 from the Web of Science(www.webofscience.com).Our LMANStat dataset includes basic information on 97,436 papers,including their title,abstract,keywords,publisher,published date,volume and pages,document type,citation counts,author information (name,ORCID,address,region,and institution),as well as their reference lists.Based on this information,we construct multi-layer academic networks,including collaboration network,co-institution network,citation network,co-citation network,journal citation network,author citation network,author-paper network,and keyword co-occurrence network.These networks change dynamically over time,providing a dynamic analytical perspective during analysis.Moreover,we also include rich nodal attributes of authors,such as the authors’ research interests,to enhance the usefulness of our dataset.The LMANStat dataset is publicly available on GitHub,and can be accessed directly at https://github.com/Gaotianchen97/LMANStat.

We present a comprehensive overview of our methodology,which covers the complete workflow from data collection to data cleaning,as well as the construction of multi-layer academic networks.Subsequently,we provide detailed explanations regarding author and paper identification,the extraction of author attributes,and the construction of multi-layer academic networks.Next,we validate the dataset through various potential scenarios for exploring and analyzing our multi-layer academic networks.To emphasize the usability of our dataset,key insights into the characteristics of the data are also provided,aligning with historical research findings and the consensus among statisticians.More importantly,the LMANStat dataset is extensively utilized by our research team to validate its usability.In our multi-layer academic networks,the collaboration network and citation network are the most commonly used networks.Therefore,we utilize them for verification.Additionally,we also consider the journal citation network with journals as nodes and the keyword co-occurrence network with keywords as nodes to validate the LMANStat dataset.

For the collaboration network,the scale-free phenomenon can be detected through a log-log degree distribution plot,which is also referenced in research on collaborative networks.The average number of authors per paper shows an increasing trend by year,indicating collaboration trends in statistical research.By visualizing a sub-network of the collaboration network,we identified the top 4 authors with the highest degrees,whose innovative methods and insights have had a profound impact on the field of statistics.As for the citation network,the in-degree of a paper is crucial as it represents the number of times the paper has been cited within the network.A higher in-degree implies a greater number of citations for a paper.Within our dataset,the average in-degree per paper is 5.31,which correlates closely with the Impact Factor (IF) of the selected journals.

In addition,journal citation networks are often employed for ranking journals,which is considered an important indicator for evaluating the quality and impact of publications in specific research fields.Therefore,we validate the accessibility of the journal citation network through journal ranking.By calculating the PageRank centrality of each node (journal),we can effectively rank journals based on their importance.Interestingly,we observe the phenomenon that the ranking of journals based on PageRank centrality closely aligns with the expectations and intuitions of statisticians.This suggests that the PageRank-based approach provides a ranking that resonates well with the perceptions of experts in the field.

It is important to note that the multi-layer academic networks presented in this paper are all dynamic in nature.Taking the citation network as an illustration,we showcase the dynamic nature of the network.The visualization includes snapshots of the network from different time periods:1980—2006,1980—2010,and 1980—2020.It is evident from the visualization that the citation network exhibits a community structure that undergoes constant changes over time.This community has shown continuous growth over the years,as indicated by the increasing number of papers associated with variable selection.In conclusion,it is worth emphasizing that the networks within this dataset are all dynamic,thereby enabling the exploration of dynamic nature.

In conclusion,the paper utilizes statistical publication data collected from the Web of Science to provide a large-scale,high-quality multi-layer academic network dataset (LMANStat dataset).The study further validates the quality and usability of these constructed multi-layer academic networks from multiple perspectives.It discusses feasible research directions and application scenarios,including but not limited to exploring community structures within academic networks,tracing the development and evolution of research topics,investigating mechanisms behind citation counts of papers,discussing the impact of international and inter-institutional collaborations,exploring career planning and development of researchers,and establishing more diversified journal ranking systems.