Dr. Marina Chugunova

Artikel in referierten Fachzeitschriften

Chugunova, Marina; Luhan, Wolfgang J. (2024). Ruled by Robots: Preference for Algorithmic Decision Makers and Perceptions of Their Choices, Public Choice 2024. DOI

• As technology-assisted decision-making is becoming more widespread, it is important to understand how the algorithmic nature of the decision maker affects how decisions are perceived by those affected. We use an online experiment to study the preference for human or algorithmic decision makers in redistributive decisions. In particular, we consider whether an algorithmic decision maker will be preferred because of its impartiality. Contrary to previous findings, the majority of participants (over 60%) prefer the algorithm as a decision maker over a human—but this is not driven by concerns over biased decisions. However, despite this preference, the decisions made by humans are regarded more favorably. Subjective ratings of the decisions are mainly driven by participants’ own material interests and fairness ideals. Participants tolerate any explainable deviation between the actual decision and their ideals but react very strongly and negatively to redistribution decisions that are not consistent with any fairness principles.
• Also published as: Max Planck Institute for Innovation & Competition Research Paper No. 22-04

Sele, Daniela; Chugunova, Marina (2024). Putting a Human in the Loop: Increasing Uptake, but Decreasing Accuracy of Automated Decision-Making, PLOS ONE, 19 (2). DOI

• Automated decision-making gains traction, prompting discussions on regulation with calls for human oversight. Understanding how human involvement affects the acceptance of algorithmic recommendations and the accuracy of resulting decisions is vital. In an online experiment (N = 292), for a prediction task, participants choose a recommendation stemming either from an algorithm or another participant. In a between-subject design, we varied if the prediction was delegated completely or if the recommendation could be adjusted. 66% of times, participants preferred to delegate the decision to an algorithm over an equally accurate human. The preference for an algorithm increased by 7 percentage points if participants could monitor and adjust the recommendations. Participants followed algorithmic recommendations more closely. Importantly, they were less likely to intervene with the least accurate recommendations. Hence, in our experiment the human-in-the-loop design increases the uptake but decreases the accuracy of the decisions.
• Also published as: Max Planck Institute for Innovation & Competition Research Paper No. 22-20

Chatterjee, Chirantan; Chugunova, Marina; Ghosh, Mainak; Singhal, Abhay; Wang, Lucy Xiaolu (2023). Human Mediation Leads to Higher Compliance in Digital Mental Health: Field Evidence from India, Frontiers in Behavioral Economics, 2. DOI

Chugunova, Marina; Danilov, Anastasia (2023). Use of Digital Technologies for HR Management in Germany: Survey Evidence, CESifo Economic Studies, 69 (2), 69-90. DOI

• Using a survey with 57 German firms, we evaluate the level of digitalization of the human resource management (HRM) function and document perceived benefits and barriers of technology adoption from organizational and individual users’ perspectives. The results give reason for optimism. Most of the companies report that the core HR processes are digitized. We do not observe adverse effects of the digital HRM tools on users’ job satisfaction and work stress. Still, more than half of companies do not yet use digital tools for strategic HRM decisions. Respondents appreciate the increased speed and cost-efficiency of digital HR processes and associate them with a competitive advantage in talent acquisition. The most prominent barriers to adoption are lack of qualified professionals, high costs, and uncertainty regarding the legal framework. Additionally, we test whether small- and medium-sized enterprises differ systematically from larger organizations in how they use digital HRM tools.
• Also published as: Max Planck Institute for Innovation & Competition Research Paper No. 22-02

Chugunova, Marina; Nicklisch, Andreas; Schnapp, Kai-Uwe (2022). Redistribution and Production with a Subsistence Income Constraint: A Real-Effort Experiment, FinanzArchiv - Public Finance Analysis, 78 (1/2), 208-238. DOI

• A large body of experimental studies demonstrates that redistribution leads to inefficiencies due to distorted work incentives. Yet, this finding is typically obtained in environments where people are unconstrained in their labor-leisure allocation decisions. In this paper we study labor supply decisions in a framework with a subsistence income constraint and a redistribution system that supports disadvantaged members of a society in meeting the constraint. We document that while high-talent taxpayers perform equally well in all conditions, the less talented ones significantly decrease their performance in response to the introduction of the tax. The negative effect of taxation is mitigated if an income threshold is present and the tax is spent meaningfully.
• Also published as: Working paper // DFG Research Group 2104 ; No. 2017,18

Chugunova, Marina; Sele, Daniela (2022). We and It: An Interdisciplinary Review of the Experimental Evidence on How Humans Interact with Machines, Journal of Behavioral and Experimental Economics 99. DOI

• Today, humans interact with automation frequently and in a variety of settings ranging from private to professional. Their behavior in these interactions has attracted considerable research interest across several fields, with sometimes little exchange among them and seemingly inconsistent findings. In this article, we review 138 experimental studies on how people interact with automated agents, that can assume different roles. We synthesize the evidence, suggest ways to reconcile inconsistencies between studies and disciplines, and discuss organizational and societal implications. The reviewed studies show that people react to automated agents differently than they do to humans: In general, they behave more rationally, and seem less prone to emotional and social responses, though this may be mediated by the agents’ design. Task context, performance expectations and the distribution of decision authority between humans and automated agents are all factors that systematically impact the willingness to accept automated agents in decision-making - that is, humans seem willing to (over-)rely on algorithmic support, yet averse to fully ceding their decision authority. The impact of these behavioral regularities for the deliberation of the benefits and risks of automation in organizations and society is discussed.
• Also published as: Max Planck Institute for Innovation & Competition Research Paper No. 20-15
• Also published as: ETH Zurich Center for Law and Economics Working Paper Series No. 12/2020

Chugunova, Marina; Luhan, Wolfgang; Nicklisch, Andreas (2020). When to Leave Carrots for Sticks: On the Evolution of Sanctioning Institutions in Open Communities, Economics Letters 191. DOI

• When asked, people dislike punishment institutions, although punishment is more effective than rewards to maintain cooperation in social dilemmas. Which institution do they choose in the long run? We study migration patterns in a laboratory experiment that allows participants to migrate continuously between punishment and reward communities. The majority of participants initially chooses the reward institution, but a substantial number of subjects joins the less profitable punishment community subsequently. In this case, the mere threat of punishment establishes high contributions. Income differences and missing compensations for cooperators in the reward community are the key factors for the decision to migrate.

Herausgeberwerke

Lacetera, Nicola; Chugunova, Marina (Hg.) (2023). The Ethics and Behavioral Economics of Human-AI Interactions (Frontiers in Behavioral Economics - Special Issue, 2) 2023.

• https://www.frontiersin.org/research-topics/43382/the-ethics-and-behavioral-economics-of-human-ai-interactions/magazine

Beiträge in Sammelwerken

Bechthold, Laura; Chugunova, Marina; Friess, Svenja; Hoisl, Karin; Rose, Michael (2021). Women in Creative Labor: Inventors, Entrepreneurs and Academics, in: Ulla Weber (Hg.), Fundamental Questions. Gender Dimensions in Max Planck Research Projects (Schriften zur Gleichstellung, 51), 135-154. Baden-Baden: Nomos. DOI

Editorials

Chugunova, Marina; Lacetera, Nicola (2023). The Ethics and Behavioral Economics of Human-AI Interactions, Frontiers in Behavioral Economics 2. DOI

Andere Veröffentlichungen, Presseartikel, Interviews

Sele, Daniela; Chugunova, Marina (2024). Putting a Human in the Loop: Increasing Uptake, but Decreasing Accuracy of Automated Decision-Making, Oxford Business Law Blog 2024.

• https://blogs.law.ox.ac.uk/oblb/blog-post/2024/02/putting-human-loop-increasing-uptake-decreasing-accuracy-automated-decision

Chugunova, Marina (2023). When Humans Interact with Non-Human Agents / Wir und unsere automatisierten Partner, Highlights 2022 from the Yearbook of the Max Planck Society 2023, 34-35.

• Englische Fassung
• Deutsche Fassung

Diskussionspapiere

Azar, José; Chugunova, Marina; Keller, Klaus; Samila, Sampsa (2023). Monopsony and Automation, Max Planck Institute for Innovation & Competition Research Paper, No. 23-21.

• We examine the impact of labor market power on firms’ adoption of automation technologies. We develop a model that incorporates labor market power into the task-based theory of automation. We show that, due to higher marginal cost of labor, monopsonistic firms have stronger incentives to automate than wage-taking firms, which could amplify or mitigate the negative employment effects of automation. Using data from US commuting zones, our results show that commuting zones that are more exposed to industrial robots exhibit considerably larger reductions in both employment and wages when their labor markets demonstrate higher levels of concentration.
• Available at SSRN

Heursen, Lea; Friess, Svenja; Chugunova, Marina (2023). Reputational Concerns and Advice-Seeking at Work, Max Planck Institute for Innovation & Competition Research Paper, No. 23-17.

• We examine the impact of reputational concerns on seeking advice. While seeking can improve
  performance, it may affect how others perceive the seeker's competence. In an online
  experiment with white-collar professionals (N=2,521), we test how individuals navigate this
  tradeoff and if others' beliefs about competence change it. We manipulate visibility of the
  decision to seek and stereotypes about competence. Results show a sizable and inefficient
  decline in advice-seeking when visible to a manager. Higher-order beliefs about competence
  cannot mediate this inefficiency. We find no evidence that managers interpret advice-seeking
  negatively, documenting a misconception that may hinder knowledge flows in organizations.
• Available at SSRN

Sele, Daniela; Chugunova, Marina (2022). Putting a Human in the Loop: Increasing Uptake, but Decreasing Accuracy of Automated Decision-Making, Max Planck Institute for Innovation & Competition Research Paper, No. 22-20.

• Are people algorithm averse, as some previous literature indicates? If so, can the retention of human oversight increase the uptake of algorithmic recommendations, and does keeping a human in the loop improve accuracy? Answers to these questions are of utmost importance given the fast-growing availability of algorithmic recommendations and current intense discussions about regulation of automated decision-making. In an online experiment, we find that 66% of participants prefer algorithmic to equally accurate human recommendations if the decision is delegated fully. This preference for algorithms increases by further 7 percentage points if participants are able to monitor and adjust the recommendations before the decision is made. In line with automation bias, participants adjust the recommendations that stem from an algorithm by less than those from another human. Importantly, participants are less likely to intervene with the least accurate recommendations and adjust them by less, raising concerns about the monitoring ability of a human in a Human-in-the-Loop system. Our results document a trade-off: while allowing people to adjust algorithmic recommendations increases their uptake, the adjustments made by the human monitors reduce the quality of final decisions.
• Available at SSRN
• Forthcoming in: PLOS ONE, 19(2): e0298037

Widmann, Rainer; Rose, Michael; Chugunova, Marina (2022). Allegations of Sexual Misconduct, Accused Scientists, and Their Research, Max Planck Institute for Innovation & Competition Research Paper, No. 22-18. DOI

• We study academic consequences of non-academic misconduct for accused researchers at US universities. Focusing on allegations of sexual misconduct, we find detrimental effects on scientific impact, productivity and career. Other researchers are less likely to cite the perpetrator’s prior work after allegations surface. The effect is absent in male-dominated fields and weakens with distance in the co-authorship network, indicating that researchers learn about allegations via their peers. Although we find that alleged perpetrators tend to remain active researchers, they are less likely to be affiliated with a university and publish fewer articles following the incident.

Chugunova, Marina; Luhan, Wolfgang J. (2022). Ruled by Robots: Preference for Algorithmic Decision Makers and Perceptions of Their Choices, Max Planck Institute for Innovation & Competition Research Paper, No. 22-04.

• As technology-assisted decision-making is becoming more widespread, it is important to understand how the algorithmic nature of the decision-maker affects how decisions are perceived by the affected people. We use a laboratory experiment to study the preference for human or algorithmic decision makers in re-distributive decisions. In particular, we consider whether algorithmic decision maker will be preferred because of its unbiasedness. Contrary to previous findings, the majority of participants (over 60%) prefer the algorithm as a decision maker over a human—but this is not driven by concerns over biased decisions. Yet, despite this preference, the decisions made by humans are regarded more favorably. Participants judge the decisions to be equally fair, but are nonetheless less satisfied with the AI decisions. Subjective ratings of the decisions are mainly driven by own material interests and fairness ideals. For the latter, players display remarkable flexibility: they tolerate any explainable deviation between the actual decision and their ideals, but react very strongly and negatively to redistribution decisions that do not fit any fairness ideals. Our results suggest that even in the realm of moral decisions algorithmic decision-makers might be preferred, but actual performance of the algorithm plays an important role in how the decisions are rated.
• Available at SSRN
• Forthcoming in: Public Choice

Chugunova, Marina; Danilov, Anastasia (2022). Use of Digital Technologies for HR Management in Germany: Survey Evidence, Max Planck Institute for Innovation & Competition Research Paper, No. 22-02.

• Using a survey with 57 German firms, we evaluate the level of digitalization of the HR management function and document perceived benefits and barriers of technology adoption from organizational and individual users’ perspectives. The results give a reason for optimism. Most of the companies report that the core HR processes are digitized. We do not observe adverse effects of the digital HRM tools on users’ job satisfaction and work stress. Still, more than half of companies do not yet use digital tools for strategic HRM decisions. Respondents appreciate the increased speed and cost-efficiency of digital HRM processes and associate it with a competitive advantage in talent acquisition. The most prominent adoption barriers are lack of qualified professionals, high costs, and uncertainty regarding the legal framework. Moreover, we test if small and medium-sized enterprises differ systematically from larger organizations in how they use digital HRM tools.
• Available at SSRN
• Also published in: CESifo Economic Studies, Volume 69, Issue 2, June 2023, Pages 69–90

Chugunova, Marina; Keller, Klaus; Samila, Sampsa (2021). Structural Shocks and Political Participation in the US, Max Planck Institute for Innovation & Competition Research Paper, No. 21-22. DOI

• This paper examines the impact of the large structural shocks – automation
  and import competition – on voter turnout during US federal elections from 2000
  to 2016. Although the negative income effect of both shocks is comparable, we
  find that political participation decreases significantly in counties more exposed to
  industrial robots. In contrast, the exposure to rising import competition does not
  reduce voter turnout. A survey experiment reveals that divergent beliefs about the
  effectiveness of government intervention drive this contrast. Our study highlights
  the role of beliefs in the political economy of technological change.

Chugunova, Marina; Sele, Daniela (2020). We and It: An Interdisciplinary Review of the Experimental Evidence on Human-Machine Interaction, Max Planck Institute for Innovation & Competition Research Paper, No. 20-15.

• Today, humans interact with technology frequently and in a variety of settings. Their behavior in these interactions has attracted considerable research interest across several fields, with sometimes little exchange among them and seemingly inconsistent findings. Here, we review over 110 experimental studies on human-machine interaction. We synthesize the evidence from different disciplines, suggest ways to reconcile inconsistencies, and elaborate on political and societal implications. The reviewed studies show that people react to automated agents differently than to humans: They behave more rationally, and are less prone to emotional and social responses. We show that there are several factors which systematically impact the willingness to accept automated decisions: task context, performance expectations and the distribution of decision authority. That is, humans seem willing to (over-)rely on algorithmic support, yet averse to fully ceding their decision authority. These behavioral regularities need to be considered when deliberating the benefits and risks of automation.
• Available at SSRN
• Also published in: Journal of Behavioral and Experimental Economics, Volume 99, August 2022, 101897

Chugunova, Marina; Nicklisch, Andreas; Schnapp, Kai-Uwe (2017). On the Effects of Transparency and Reciprocity on Labor Supply in the Redistribution Systems, Working Paper / DFG-Research Group 2104, Nr. 2017-19.

• Although taxation distorts work incentives both for taxpayers and transfer recipients, its net effect on labor provision is shown to be less severe than predicted by the theory. It is likely that the reciprocity between taxpayers and transfer recipients plays an important role in
  mitigating the negative consequences of redistribution and maintaining a high level of effort. To check it we run a series of real-effort experiments exploring the production effects of taxation in the environment with unilateral monitoring: Taxpayers can continuously monitor the effort of the transfer recipient, which is designed to trigger reciprocity.
  Surprisingly, we find that monitoring decreases the total labor provision: recipients produce significantly less under monitoring, while the
  production of the taxpayers remains unchanged.
• https://epub.sub.uni-hamburg.de/epub/volltexte/2020/100489/pdf/2017_19.pdf

Chugunova, Marina; Nicklisch, Andreas; Schnapp, Kai-Uwe (2017). Redistribution and Production with the Subsistence Income Constraint: a Real-Effort Experiment, Working Paper / DFG-Research Group 2104, Nr. 2017-18.

• A large body of literature demonstrates that redistribution leads to inefficiencies due to distorted work incentives. Yet, this result is obtained under the assumption that people are absolutely free in their labor-leisure allocation decisions and that taxation is merely a wage cut. We challenge this assumption and study labor supply decisions
  in a framework with the subsistence income constraint and a redistribution system which supports disadvantaged players. The results of the real-effort experiment show that the introduction of the moderate
  subsistence income requirement causes a substantial increase in productivity among taxpayers, with slight additional boost if tax returns are transferred to recipients and not wasted unproductively. As for
  recipients, the prospect of receiving a transfer significantly enhances
  their productivity and spurs the overall efficiency leading to a self-sorting of recipients according to their skills.
• http://epub.sub.uni-hamburg.de/epub/volltexte/2020/100471/
• Published in: FinanzArchiv Jahrgang 78 (2022) / Heft 1-2, S. 208-238 (31)