All KLU Publications

Extreme weather is often perceived as a major threat to the reliability of rail transportation. This study investigates regional rail operations in central Germany and examines whether severe weather conditions causally affect train arrival delays. We combine several years of operational data (2017–2022) with weather observations from the German Weather Service and define four component-specific treatments for adverse conditions: extreme temperature, strong wind, heavy rainfall, and snow presence. To estimate causal effects, we employ a Double Machine Learning framework based on partial linear regression (PLR–DML) and Causal Forest–DML. By reframing the weather–delay relationship as a causal question rather than a purely predictive one, the study provides evidence on whether extreme weather constitutes a materially relevant source of arrival delays. Across all four weather components, estimated average treatment effects are small in magnitude; analyses of individualized and group-level effects reveal no consistent or robust patterns of heterogeneity across lines, seasons, weekdays, or hours of the day. Robustness checks indicate that these operationally negligible estimates are not sensitive to outliers, rare-event imbalance, or to reasonable perturbations consistent with plausible unobserved confounding. Exploratory mediation analyses are consistent with the interpretation that passenger-flow variables do not materially amplify the already negligible estimated effects of severe weather on delays. Overall, the results suggest that, in this regional network, severe weather does not materially increase arrival delays. The findings underscore the importance of rigorous causal diagnostics for distinguishing perceived from materially relevant sources of delay in transportation reliability studies.

Since the first publications on humanitarian operations, researchers have been concerned about the impact of our research on humanitarian practice. This paper seeks to better understand the gap between research and practice and, thus, how to narrow it. We incorporate not only the research perspective, through a review of 640 abstracts in humanitarian operations and a survey of the author community, but also the practitioner perspective, through 18 interviews with experienced humanitarian practitioners and a review of 131 practitioner conference proceedings. We assess the extent of practice impact, and find limited but important instances. We identify new barriers to achieving practice impact from the practitioner perspective. Finally, we develop a framework for describing the depth and reach of practice impact achieved by a research project, with a description of many ways in which such impact may be (and has been) achieved. We hope these efforts enable greater researcher accountability to practitioners and clearer paths to practice impact for the field.

Effective resource allocation is crucial for optimizing business processes. Yet, most existing methods focus solely on single-process optimization, overlooking the interdependencies present in multi-process environments. This limitation results in inefficient resource allocation, and scalability challenges. To address this gap, we propose MuProMAC (Multi-Process Multi-Agent Coordination), a novel reinforcement learning-based method designed to optimize resource allocation across multiple interdependent business processes. Unlike prior methods, MuProMAC is the first online resource allocation method that explicitly models the interdependencies between processes and dynamically balances competing resource demands to minimize global average cycle time. We evaluate our method in five multi-process scenarios with different levels of resource contention, comparing it against state-of-the-art online resource allocation methods and three simple baselines. Our results show that MuProMAC is consistently among the top-performing methods in shared-resource environments. It achieves low cycle times and stable performance across different workload conditions, outperforming existing methods through its strong adaptability to evolving business processes and increasing complexity.

Experiments are a powerful method for establishing causal inference and have increasingly been used in international business (IB). Yet the method remains in its infancy, standards remain unclear, and IB is particularly challenging for experimental designs despite its promise. IB experiments are also scattered across a wide range of (non-IB) journals. We therefore conduct a comprehensive review of experiments in IB to assess how, where, and with what rigor experimental methods are currently being used, and where methodological weaknesses may limit their contributions to IB and adjacent fields. We identify and analyze 212 experiments since 2000 across IB and management journals. Our review highlights both the opportunities and methodological challenges of experimentation in IB, with particular attention to issues of external validity, internal validity, and transparency. We provide an IB-specific design guide to address these challenges, illustrate how experiments can strengthen theory building in cross-border contexts, and outline opportunities for impactful future research at the intersection of IB and experimental methods.

Biomethane production has expanded rapidly in France, making it the leading producer in Europe in 2024, largely supported by feed-in tariffs. Recent institutional reports nevertheless argue that these policies generate excessive profits, particularly for agricultural producers. This paper revisits these claims by examining how inter-organizational arrangements shape profitability measurement in the biomethane sector. Drawing on the literature on transfer pricing and business models, we argue that profitability measured at the legal-entity level may diverge from underlying economic performance when production is organized across interconnected entities. Combining a quantitative analysis of ownership structures for 273 biomethane units with 23 in-depth case studies, we identify two main sources of distortion. First, arrangements affecting fixed assets and biomass inputs may disconnect recorded costs from market conditions. Second, productive activities such as biomass handling, transport, and digestate spreading may be performed outside the biomethane entity itself, shifting labor and operating costs across organizations. We further show that these arrangements are strongly shaped by ownership structures, giving rise to three distinct business models of biomethane production that bias profitability indicators in different ways. These findings question the robustness of firm-level profitability measures used to recalibrate feed-in tariffs and, more broadly, highlight the need to integrate organizational structures into renewable energy policy evaluation.

Process mining has grown into a mature research field with a wide range of techniques and applications. Much of this development builds on the pioneering work of Wil van der Aalst, whose contributions have shaped both the foundations and the growth of the discipline. Today, the field’s breadth raises the need for systematic methodological reflection to ensure that findings are robust and meaningful. In this paper, we provide a comprehensive discussion of threats to validity in process mining research. Building on the methodological framework of algorithm engineering, we analyze nine distinct validity concerns and examine how they apply across different streams of process mining. Our analysis highlights both established strengths and recurring challenges, drawing on examples from seminal contributions in the field, many inspired by Wil’s work.

As consumer concern about labor conditions in global supply chains grows, firms face increasing pressure to disclose fair wage and labor-related information. Yet, transparency can be risky because it may expose unfavorable supplier practices and cause reputational harm. Drawing on signaling theory, we examine how fair wage and labor-related supply chain transparency (SCT) influences consumer word-of-mouth (WOM), particularly when disclosures contain both positive and negative information. Across two vignette-based experiments in an online shopping context, we test whether distributive justice and trustworthiness explain consumer responses to SCT and leverage these mechanisms to further explain responses to mixed-valence SCT disclosures. Study 1 shows that a uniformly positive SCT disclosure, relative to nondisclosure, increases WOM through both distributive justice and trustworthiness. Study 2 examines a mixed-valence disclosure combining positive and negative information. Although mixed-valence disclosures lower distributive justice perceptions, they increase trustworthiness relative to nondisclosure, yielding a positive total effect on WOM. Together, the findings show that SCT functions as a costly and credible signal: even when transparency reveals negative wage and labor information, mixed-valence disclosure can strengthen trustworthiness enough to enhance consumer advocacy.

The Colombian road freight transport sector is a major contributor to greenhouse gas (GHG) emissions and plays a critical role in achieving the country’s net-zero target by 2050. While technological solutions such as vehicle electrification and alternative fuels are essential, substantial mitigation potential also lies in improving operational efficiency. A persistent source of inefficiency is the prevalence of empty trips, cargo-free freight movements that increase logistics costs, reduce fleet utilisation, and generate avoidable emissions.
This study addresses empty trips from a national, system-level planning perspective by introducing a spatio-temporal synchronisation model that identifies opportunities to coordinate independent freight movements travelling in opposite directions based on temporal and spatial compatibility. By relying on aggregated information rather than firm-level routing optimisation, the model enables scalable coordination across the freight network without requiring centralised control.
Two algorithms were developed to estimate current empty-trip rates from observed trip sequences and the minimum achievable rate under ideal synchronisation. Applied to more than 7.99 million trips from Colombia’s National Freight Registry, the results indicate that synchronisation could reduce empty kilometres by up to 39.34% for light and medium-duty vehicles and 37.37% for heavy-duty vehicles, corresponding to an estimated annual reduction of 130,470 tons of CO2e.
The study also introduces the concept of structural residual empty trip revenue (SRETR) to assess the economic potential of empty trips that remain unavoidable under optimal coordination. These insights support policymakers in prioritising interventions, addressing regional freight imbalances, and integrating operational efficiency measures into national decarbonisation strategies.

Task offloading strategies in Mobile Edge Computing (MEC) aim to reduce computation delay and energy consumption of mobile devices by offloading tasks to edge servers, which is key to improving MEC system performance and user experience. Recent efforts have focused on utilizing deep reinforcement learning (DRL) but DRL-based offloading strategies struggle to achieve optimal decisions in limited iterations due to complexity. Therefore, to address the complexity challenge, this paper proposes a Prune-based Deep reinforcement learning Offloading Algorithm (PDOA) to enhance MEC system performance. First, we construct a dynamic MEC system model and formulate the task offloading problem as a Markov decision process to minimize the total cost of the MEC system. Next, we propose a prune-based DRL offloading algorithm, which prunes DRL models to reduce the complexity and improve learning efficiency, thereby lowering system costs. The experimental results show that PDOA reduces the computational cost of MEC systems significantly compared with other methods and lowers system costs by over 10%. This optimization approach provides a novel research perspective for applying DRL models in MEC.

This paper addresses the planning of loading and unloading processes, also termed stevedoring, for Roll-on/Roll-off (RoRo) ships. We consider a setting where we sequence operations, assign tugs to handle trailers, and determine the cargo positions on the ship and in the yard. In real-world settings, the outcomes of these decisions are strongly affected by stochastic uncertainty, such as considering the type of cargo to unload or traffic congestion on the ship. To account for this, we formulate the problem as a sequential decision process and describe two feature-based policy function approximations for the decision-making. We propose an effective heuristic for designing policies in an offline learning process. We evaluate this approach in a case study at the Port of Kiel, Germany, simulating the service operations of a three-deck RoRo ship. Compared to industry benchmark policies, the proposed policies reduce the ship’s turnaround time by up to 26 min (8.8%). Additionally, we apply our methodology to illustrate the value of various levels of information availability. The results show that information on process completion times, especially for vehicle entry and exit at both the ship and the yard, has the potential to shave off turnaround time by an additional 8%. Yet, the availability of information on the incoming ship’s stowage plan is an important precondition for tailored policies capable of achieving such reductions, especially when there is a surplus of tug-handled cargo.

Despite extensive research streams on leadership and team processes, there is a surprising paucity of studies at their intersection. Both research streams share an increasing attention to the social interactions at the core of these phenomena. Leveraging this behavioral lens, this study draws on respectful inquiry theory to explore how specific leader communication behaviors affect team interaction dynamics during decision-making, as one important team process. We conducted a laboratory study with 22 four-person teams and a confederate leader who engaged in a hidden profile task in a personnel selection scenario. We manipulated the leader’s question asking behavior (open questions vs. statements only) and listening behavior (listening attentively vs. not listening) and randomly assigned teams to one of the four conditions. Team interactions were video-recorded and analyzed at the micro-level of communication. Specifically, we explored how leader communicative behaviors affected (1) the quality of team decision-making, (2) the conversational structure (via speaker turns), and (3) constructive communication patterns. We found that team’s yielded the lowest performance in the “disrespectful inquiry”-condition (i.e., asking questions but not listening). This condition was also characterized by increased levels of interaction amongst team members that could be interpreted as an attempt to compensate for the lack of functional leadership. By adopting a consistent, micro-level behavioral perspective, our findings bridge the literature of leadership and team interactions and suggest an update to extant theorizing on leadership substitutions.

In many decision processes, a decision maker or planner must review and optionally adjust the recommendations that are generated by a decision support system (DSS). When the DSS is well-tuned to its task, adjustments by a planner can be rare and may even degrade the DSS’s performance. Targeted automation could address these inefficiencies by predicting whether a planner will adjust a recommendation and improve the performance of the system. The remaining recommendations can be automated. However, as more recommendations are automated, fewer will receive planner input. This may starve the prediction model of the observations it needs for retraining. To maintain predictive performance, we must therefore address the loss that automation imposes on the model’s ability to learn from a planner’s decisions over time. Using 4 years of procurement ordering data from our research partner, a large materials handling equipment manufacturer, we develop and train a series of machine learning classifiers that predict individual instances in which a planner will improve a DSS-generated procurement order decision. We mitigate the performance erosion that automation engenders by structuring the selection of the model’s classification threshold similar to a newsvendor problem, accounting for the value of learning and balancing the costs and benefits of under or over automating. In our setting, this approach automates around 84% of all DSS recommendations while retaining three times more planner improvements than random automation. The models maintain their predictive performance over time, despite losing automated outcomes for retraining and substantial dataset shift. Our research contributes to a broader debate on the allocation of decision authority between humans and algorithms, and creates a framework for targeted automation in an operational setting that balances the net benefits of automation versus the long-term benefits of algorithmic learning.

Humanity urgently needs innovative solutions to tackle challenges like climate change and social inequality. The corporate sector is instrumental, requiring adaptability akin to the digital revolution, with established companies and startups mutually benefiting. In the logistics sector, innovation and embracing new technologies are critical for addressing environmental impacts. Kühne Logistics University (KLU) leads this initiative, proposing an Entrepreneurship and Innovation Center for Green and Responsible Logistics. This center aims to develop sustainable logistics solutions through collaboration among academia, industry, and entrepreneurs, symbolizing a pioneering venture in sustainability-focused social innovation. This case study delves into logistics’ role in global emissions, KLU's leadership in green and responsible logistics, and its commitment to sustainability. The center strives to merge academic research with practical application, support startups, students, and companies, and foster social innovation both within education and beyond.

Will we one day inhabit virtual worlds? From the buzz around Second Life in the early 2000s to Mark Zuckerberg’s much-hyped quest for the Metaverse, there seems to be a clear pattern. After being touted as revolutionary, virtual worlds fail to meet expectations and fade into obscurity. Yet the dream of an immersive, independent, and interconnected virtual social universe still has the potential to transform the economy, reshape society, and profoundly affect the future of humanity.

This book contends that despite shifting trends, the virtual universe remains crucial to the technology of tomorrow—and might arrive sooner than we think. Through vivid examples and detailed case studies, Andreas Kaplan, a pioneering researcher of virtual worlds since the early 2000s, provides a comprehensive examination of the Metaverse and its far-reaching implications. He delves into key technologies such as artificial intelligence, blockchain, and extended reality and discusses ethical issues around privacy, identity, and governance. Kaplan examines business models, digital ownership, and corporate strategies alongside applications spanning marketing, research and development, human resources, and supply chain management. The book concludes by contrasting potential futures, from utopian promise (the Petaverse) to dystopian peril (the Metaworse) and everything in between. Written in an engaging, reader-friendly style, The Virtual Universe goes beyond the hype—and the doubting—to uncover the true potential of the Metaverse.

We investigate how entrepreneurs’ pride expressions influence funding outcomes in new venture pitches. Drawing on Social Functional Theory of emotions, we argue that authentic and hubristic pride have opposing effects on investors’ decisions through perceptions of competence and warmth. Across three experimental studies with crowdfunders (N1 = 196, N2 = 593, N3 = 159), we examined verbal and bodily pride expressions. Verbal expressions of authentic pride indirectly increased funding intentions via perceived competence, whereas hubristic pride indirectly decreased funding through reduced perceptions of both competence and warmth. These effects did not vary across variations in campaign size and technological innovation. Bodily expressions replicated the negative indirect effect of hubristic pride compared to authentic pride. Our findings suggest entrepreneurs should convey authentic pride during pitches, as hubristic pride may undermine success by signalling arrogance and diminishing perceived competence and warmth.

Despite growing regulatory concerns about potential overcharging of sustainable investors, empirical evidence is lacking. In two controlled laboratory-in-the-field experiments with 415 professional financial advisors from Europe and the United States and an incentivized survey, we identify two distinct but interacting effects. First, advisors charge sustainable investors a premium. This premium persists even when accounting for differences in skill, effort, and costs. Second, advisors impose higher fees on clients with low financial literacy. These factors interact. Sustainable investors with low financial literacy are charged the highest fee, whereas those with high financial literacy do not pay a sustainability premium. Our findings suggest that advisors extract additional fees for sustainable investment mandates but avoid overcharging sustainable investors with high financial literacy.

Recent reviews portray humble leadership as a near-universal asset, yet a close inspection of 217 journal articles (274 studies) suggests the construct rests on shaky ground. Prevailing definitions conflate self-insight, appreciation of others, and teachability, variables rooted in other literatures, creating tautologies and valence-based halo. Measurement issues compound the problem: 84% of studies rely on surveys, so ratings of “humble leadership” are conflated by various mechanisms (e.g., evaluative judgments, performance-cue effects, omitted variables) and should not be used as independent variables. We argue that progress on this topic depends on shifting attention from traits and evaluations toward the behaviors that humility denotes: voluntary, public, status-minimizing acts through which leaders redirect credit away from themselves. With this definition, we integrate signaling theory with an idiosyncrasy-credit perspective, arguing that self-effacement behaviors function as costly signals: they translate into humility perceptions only when paired with accrued credit that is placed at risk. It follows, for instance, that signaling humility in the absence of established credit is likely to backfire. We develop testable propositions, a conceptual model linking accrued credit to signal credibility, and an incentive-compatible laboratory paradigm that enables identification of the causal effects of instrumented humility signaling. We conclude with recommendations for behavioral coding, archival text analysis, and experimentally grounded field designs that can replace halo-laden survey inferences with more credible evidence on when humility signaling helps, when it harms, and when it goes unnoticed.

Process mining aims to obtain insights from event logs through the automated analyses of recorded process data in information systems, with the ultimate aim to improve business processes running in organisations. However, real-life event logs are often incomplete, noisy, or ambiguous, such as missing timestamps or having ambiguous event labels, which traditional deterministic models cannot capture. Recent process mining developments have considered uncertainty in process mining artifacts more explicitly: in logs of recorded process behaviour, uncertainty may implicitly or explicitly influence process mining outcomes, while in process models, explicit uncertainty allows analysts to interpret and value outcomes. In this paper, we provide a conceptual foundation for uncertainty in process mining by introducing a four-level specification that separately addresses uncertainty in log attributes (e.g., activity labels of events, frequencies) and model elements (e.g., service times, read guards). For each type of uncertainty, we illustrate the levels with concrete examples to help understanding and application. We then provide a structured overview of the state of the art in stochastic process mining, classified using our specification, and present key open research challenges.

We investigate the nexus between the early-life disaster experiences of chief executive officers (CEOs) and their firms’ environmental performance metrics. We hypothesize that first-hand experience of the adversities of natural disasters in the formative years of a CEO can catalyze a transformation in their environmental cognizance and perspective. This transformation is postulated to have a beneficial influence on their corporations’ strategic frameworks for environmental risk mitigation. Our results show that entities steered by CEOs exposed to disasters in their early life have fewer incidences of environmental issues. These findings remain consistent even when controlling for other factors or using alternative methods. We suggest that CEOs with early disaster experience have an enhanced perception of risk ramifications, which inculcates a prudential approach to decision making, potentially heightening the environmental risk profile of their enterprises.

When binary classification models are wrong, managers face misclassification costs. Although false positive outcomes imply unnecessary mitigation efforts, false negative outcomes imply overlooking the class of interest. Humans calibrate these ai models supporting operational systems by adjusting the decision threshold that translates prediction probability into either class. Results of our controlled laboratory experiment show that, despite all relevant information being available, decision makers systematically deviate from the optimal cost-efficient threshold. We observe a significant interaction effect of class and cost imbalance on this deviation, which increases in high-stakes settings where more extreme thresholds are optimal. When unit costs are different, we find that participants anchor on the threshold where expected misclassification costs for false alarms and missed hits are equal, whereas mean anchoring cannot explain the pull-to-center behavior sufficiently. Surprisingly, we confirm that this impulse balance equilibrium also serves as attractive anchor in our setting, where decisions are made ex ante without loss aversion. To debias decision makers, simulated responses with behavior-aware costs show that subjects are nudged to make choices closer to the optimum. Managers should be aware of this boundedly rational behavior and complementary debiasing techniques, as sub-optimal threshold setting results in 53% higher misclassification costs, on average.

An estimated 305 million people required humanitarianassistance in 2025, even as official donor fundingcontinued to contract. Under these constraints,logistics — historically treated as an operational lineitem — is a determinant of whether responses happenat all, and transportation choices about modalities,providers, and financial arrangements shape deliverytimelines, geographic reach, and operational risk. Yetevidence on when donated transport is used, howdecisions are made on both sides of the partnership,and what effects these arrangements have onresponse capacity remains limited. This report providesempirical evidence on when these partnerships areactivated, how they create value, and where they canbe strengthened.

Studies have shown that anomie, that is, the perception that a society’s leadership and social fabric are breaking down, is a central predictor of individuals’ support for authoritarianism. However, causal evidence for this relationship is missing. Moreover, previous studies are ambiguous regarding the mediating mechanism and lack empirical tests for the same. Against this background, we derive a set of integrative hypotheses: First, we argue that perceptions of anomie lead to a perceived lack of political control. The repeated failure to exert control in the political sphere leads to feelings of uncertainty about the functioning and meaning of the political world. This uncertainty heightens people’s susceptibility to authoritarianism because, we argue, the latter promises a sense of order, meaning, and the guidance of a “strong leader.” We support our hypothesis in a large-scale field study with a representative sample of the German population (N = 1,504) while statistically ruling out alternative explanations. Adding internal validity, we provide causal evidence for each path in our sequential mediation hypothesis in three preregistered, controlled experiments (conducted in the United States, total N = 846). Our insights may support policymakers in addressing the negative political consequences of anomie.

High-performing employees (HPEs) contribute disproportionate value to organizations not only through exceptional task performance but also through discretionary helping behaviors toward coworkers. Yet prior research offers mixed evidence regarding HPEs’ engagement in organizational citizenship behavior directed at individuals (OCB-I), portraying them alternately as highly productive “all-stars” or isolated “lone wolves.” We argue that this inconsistency reflects differences in how HPEs interpret coworkers’ reactions to their success, particularly envy, and how these interpretations are shaped by relational context. Drawing on the envy-attribution model and Emotions as Social Information (EASI) theory, we develop a moderated mediation framework in which HPEs’ perceptions of coworkers’ envy, whether benign or malicious, explain when high-performance status translates into OCB-I, and how relational identification shapes these interpretations. Across two vignette experiments (N = 837), a two-wave field study (N = 233), and a laboratory experiment (N = 197), we find convergent evidence that perceiving benign envy increases OCB-I, whereas perceiving malicious envy decreases it. Relational identification selectively reduces perceptions of malicious envy and buffers its negative indirect effect on OCB-I but does not significantly shape perceptions of benign envy or its positive behavioral consequences.