Portail HAL Rennes SB
Not a member yet
692 research outputs found
Sort by
The formal rationality of artificial intelligence-based algorithms and the problem of bias
No full textInternational audienceThis paper presents a new perspective on the problem of bias in artificial intelligence (AI)-driven decision-making by examining the fundamental difference between AI and human rationality in making sense of data. Current research has focused primarily on software engineers’ bounded rationality and bias in the data fed to algorithms but has neglected the crucial role of algorithmic rationality in producing bias. Using a Weberian distinction between formal and substantive rationality, we inquire why AI-based algorithms lack the ability to display common sense in data interpretation, leading to flawed decisions. We first conduct a rigorous text analysis to uncover and exemplify contextual nuances within the sampled data. We then combine unsupervised and supervised learning, revealing that algorithmic decision-making characterizes and judges data categories mechanically as it operates through the formal rationality of mathematical optimization procedures. Next, using an AI tool, we demonstrate how formal rationality embedded in AI-based algorithms limits its capacity to perform adequately in complex contexts, thus leading to bias and poor decisions. Finally, we delineate the boundary conditions and limitations of leveraging formal rationality to automatize algorithmic decision-making. Our study provides a deeper understanding of the rationality-based causes of AI’s role in bias and poor decisions, even when data is generated in a largely bias-free context
Business Goal Difficulty and Socially Irresponsible Executive Behavior: The Mediating Role of Focalism
No full textInternational audienceExecutive social irresponsibility has received increasing research attention in recent years, following the consensus for a broader stakeholder approach to managerial decision making. Despite the importance of the subject, there remains insufficient research on contextual factors that mold executives’ orientation toward social responsibility. Through three studies, we demonstrate that difficult business goals can reduce executives’ tendency to consider social responsibility in their decision making. Further, we find that focalism—a cognitive bias based on affective forecasting theory—can mediate positive relationships between business goal difficulty and socially irresponsible executive behavior. Our findings also suggest that, expanding executives’ thought processes beyond the narrow focus of a business goal achievement can be a good strategy in reducing socially irresponsible executive behavior, even in the presence of difficult goals
On the drivers of sustainable development: empirical evidence from developed and emerging markets
No full textInternational audienc
How Does Selling Capability Impact Firm Value? The Moderating Roles of Relative Strategic Emphasis, Market Volatility, and Technological Volatility
No full textInternational audienceFirms develop and deploy selling capability to create and sustain a competitive advantage. Previous studies have focused predominantly on static, input-based selling capability, paying little attention to dynamic, efficiency-focused selling capability. This treatise reconceptualizes selling capability from a dynamic and efficiency (input–output) perspective and investigates the effect of selling capability on firm value with the contingent role of internal [i.e. relative strategic emphasis (SE)] and external (i.e. market volatility and technological volatility) factors. Using data from 341 US-based manufacturing and service firms over the period 2014–2020 and an endogeneity-robust dynamic estimation technique, the authors find that selling capability positively affects firm value, and firms with a relative SE on value appropriation (i.e. advertising) as opposed to value creation (i.e. R&D) reap more rewards from selling capability. That is, relative SE positively moderates the nexus between selling capability and firm value. Furthermore, the results demonstrate that the interactive effect of selling capability and relative SE is weaker when an industry experiences higher market volatility but stronger when technological volatility is higher. Overall, this study demonstrates that a firm's selling capability should be managed dynamically in light of its (internal) relative SE and (external) environmental conditions. The results are robust to several additional sensitivity analyses.Source : Publisher
To Pool or Not to Pool in Carbon Quotas: Analyses of Emission Regulation and Operations in Supply Chain Supernetwork under Cap-and-Trade Policy
No full textInternational audienceCarbon emission control and maintaining economic growth drive firms and governments to strike a balance between these two aspects via environmental policies. Inspired by diverse carbon permit trading practices, we study the impact of two variant cap-and-trade schemes on supply chain performance and emission reduction. Focusing on a realistic supply chain supernetwork, we examine the operational decisions and emission mitigations of the system under the carbon caps-pooling and caps-nonpooling schemes which reflect two typical modes of carbon quota in practice. We exploit game theoretical models to characterize the strategic interactions of supernetwork node firms, and then transform them into equivalent variational inequality problems to derive the associated equilibria in different scenarios. We find that the carbon caps-pooling scheme yields a win–win outcome in terms of the environmental and economic performance and the carbon permit sale price affects the decreasing rate of carbon intensity more effectively than the purchasing price. Moreover, the caps-pooling scheme further restrains firms’ opportunistic behaviors of lowering production and selling surplus carbon permits at a high price. Notably, comparing the two carbon schemes implies that under the cap-and-trade policy governments should stimulate firms adopt the caps-pooling scheme with emission permits trading, and accordingly firms have more flexible control over carbon permits. We further conduct computational analysis to validate and illustrate the results, which indicates the robust effect of caps-pooling scheme on the supply chain supernetwork under distinct market structures
Hitting Net-Zero Without Stopping Flying: Increasing Air Travelers’ Likelihood to Opt-in to Voluntary Carbon Offsetting
No full textInternational audienceVoluntary carbon offset (VCO) programs give air travelers opportunities to neutralize their carbon footprint. Despite its potential, few existing studies have explained how to present VCOs that can effectively appeal to the sensibilities of travelers in different conditions. We designed three online experiments with strategies to motivate travelers to opt-in. We found travelers who receive concrete messages that emphasize specific actions are more likely to opt-in to VCOs when flying in the near future. In contrast, travelers receiving abstract messages that emphasize general initiatives are more likely to opt-in to VCOs when flying in the distant future. When travelers are allowed to choose their preferred carbon offset method, they are more likely to opt in, especially when they receive concrete messages that indicate specific actions but not general initiatives. These findings contribute to the aviation carbon offset literature and offer useful new insights for airline companies
Robust maritime disruption management with a combination of speedup, skip, and port swap strategies
No full textInternational audienceA large volume of global transportation is carried out annually by liner shipping companies and it includes a large portion of global trade. Accordingly, due to the countless number of these voyages, precise planning in this field is vital to prevent severe loss. One of the noticeable issues that occur during the voyage can be natural or human disruption which is interpreted as a “delay” in the liner shipping services. Therefore, there is a need to reschedule the pre-established plan to compensate for the delays and reduce the costs including the penalties and exceeding fuel consumption. In this paper, a novel recovery model is proposed for container ship problems. This mixed-integer programming model with the use of speedup, skip, and swap ports strategies not only attempts to mitigate the dire financial consequence of the disruption but also reduces the carbon footprint. Furthermore, to ensure customer satisfaction, the alternative transshipment decision for the skipped ports cargo is considered in the model. The nonlinear model is linearized through the exact techniques, then solved in CPLEX software. As a result, the primary deterministic model could act as a “wait-and-see” solution to reduce disruption losses by up to 66% using simultaneous recovery strategies of speedup, skip and swap. However, a robust optimization approach is proposed owing to the uncertainty in delay time, type, severity, and point of disruption. This approach enables the model to face a wide variety of disruptions (that are predicted under different scenarios each of which is associated with an occurrence probability) and recommends an augmented schedule that guarantees to be feasible, optimal, and resistant. The robust model is applied in a real case from the maritime industry, and the value of robustness is reported. The results demonstrate the superiority of this model compared with others
Food choice and the epistemic value of the consumption of recommender systems: the case of Yuka’s perceived value in France
No full textInternational audienceFood Recommender Systems (RecSys) are innovative knowledge systems that inform consumers of food choices according to criteria, including nutritional content, health concerns, production method, carbon footprint or other social and ethical considerations. They raise important questions at the intersection of technology accuracy and today evolving consumers’ knowledge seeking behaviours, which implies to unpack the epistemic value of food RecSys. This study investigates the drivers of the perceived value of food RSs consumption by proposing a model that establishes via PLS-SEM (n = 253) a positive relationship between the Yuka company’s food RecSys’ epistemic value and its perceived value. The model demonstrates that Yuka RecSys’ epistemic value relies on the disciplinary drivers of compatibility, self-confidence, and consumer innovativeness, and the problematising drivers of memory and learning, which come from using the application. The perceived value of food RecSys is found to relate to RecSys epistemic value beyond the functional accuracy aspects of recommendation algorithms. Results highlight the importance of developing a refined understanding of epistemic value considering the consumption of RecSys. RecSys’ developers, retailers, food manufacturers and policy makers must work on better mapping and adjusting information through consumers socialised RecSys’ usage to shape the design of the next generation RecSys
Common institutional ownership and mergers and acquisitions outcomes
No full textInternational audienceInstitutional owners frequently invest in a diversified portfolio of firms to avoid firm-specific risks. I investigate the particular scenario in which the institutional owners have shares in both the acquiring and the acquired target firms of an M&A deal. Using a quasi-experimental approach, I find that the acquirer pays less premium and performs better after the M&A effectiveness when the ratio between the value owned by common institutional shareholders in the acquirer and the value held by the same shareholders in the target firm is higher. The value paid is higher, and the performance worsens when this ratio is lower. The results suggest the common institutional owners can obtain benefits from promoting and implementing such M&A deals as a secondary compensation for their lack of control, usually at the expense of the controlling shareholders
The economic benefits of business mediation in the Brazilian scenario
No full textInternational audienceThis study aims to measure the economic benefits of business mediation in Brazil compared to lawsuits, investigating whether choosing mediation to resolve a business conflict is more economically advantageous than filing a lawsuit in the judiciary and whether opting for extrajudicial mediation to settle a business dispute is more economically advantageous than reaching an agreement in judicial mediation. The research, conducted in theSoutheast region of Brazil after the implementation of the Brazilian Code of Civil Procedure in 2016, employed quantitative methods, analyzing 397 judicial processes. The results indicated that extrajudicial mediation is the most cost-effective option, while the judicial lawsuit is the mostcostly. This article provides essential information for businesses to make informed decisions in the resolution of business disputes