FEMM - Working Paper Series - 2025

25004

Jens Robert Schöndube, Barbara Schöndube‐Pirchegger

Availability of AI tools and their effect on the auditing process

In this paper we model the interaction between an auditor and a client firm. The client firm’s manager can either report truthfully or commit fraud. The auditor needs to plan a two stage audit that allows to detect fraud. In the first stage an AI tool is employed that provides a signal about the quality of the client’s internal control system (ICS). Classifying the ICS as weak or strong, the signal alters the auditor’s expectations regarding the client’s fraud probability. In the second stage, the auditor decides about her audit effort conditional on the information provided by the AI. Comparing the AI setting to a benchmark setting without AI use, we find that employing the AI tool reduces the manager’s incentives to commit fraud. At the same time it reduces the equilibrium effort provided by the auditor. As a consequence, the probability that actual fraud is detected remains unchanged. We extend our model and allow the AI tool to be customized such that it can either focus on detection of the weak ICS, the strong ICS, or on both equally. We find that the AI specification that minimizes ex ante probability for fraud not necessarily coincides with the specification that minimizes auditing costs. It follows that the auditor in charge of customizing the AI cannot necessarily be expected to do so in a fraud minimizing way.

25003

Max-Frederik Neubert, Barbara Schöndube-Pirchegger

Costs and benefits of discretion in performance evaluation and patterns of bias

This paper investigates incentive effects from subjective performance evaluation (SPE) in an agency setting. An employee (agent) is evaluated by his superior (principal) via a subjective, potentially biased, performance report. We assume that this subjectiveness in evaluation affects the utility of both players, causing costs from biasing the report to the principal and benefits (costs) from over- (under-) evaluation to the agent.
If the superior chooses the reporting bias sequentially optimal, we find that benefits from subjective, as opposed to objective performance measurement, do not outweigh its costs. If, in contrast, the supervisor is able to commit to an ex ante optimal bias choice, SPE can be beneficial if the agent’s preference for over-evaluation is sufficiently strong. While a centrality bias arises independent from the supervisor’s ability to commit, a leniency bias results only along with an ex ante optimal bias.

25002

Laura Daniela Bohorquez Camacho, Marcel Lichters, Pedro J. Amor

Automated emotion recognition in marketing research: A systematic literature review of current image and video-based methods

This study reviews the recent literature on Automated Emotion Recognition (AER), focusing on image and video-based methods applied in marketing research. The literature overview highlights the transformative potential of AER, including real-time, unobtrusive, and scalable applications. It identifies key tools, including Noldus' FaceReader and iMotions' Facial Expression Analysis, as significant contributors to insights in diverse contexts such as e-commerce, social media, and online platforms. The analysis also addresses theoretical challenges, such as the limitations of Ekman's basic emotion theory and the contextual dependence of facial expressions. Practical recommendations for AER use include incorporating multimodal approaches and ensuring cultural and contextual inclusivity in training datasets. Thus, the current work advances the discourse on leveraging AER for refined marketing strategies.

25001

Mohamad Alhussein Saoud

The 2015/2016 New Year’s Eve Cologne Event and Anti-refugee Crimes in Germany

During New Year’s Eve 2015/2016, the German city of Cologne witnessed mass sexual assaults and thefts by perpetrators described as having an Arab-African appearance. This paper studies whether the event in Cologne led to a backlash in crimes against refugees in Germany. Difference-in-differences regressions reveal a significant jump in anti-refugee crimes immediately after the event. This rise is driven by assaults and miscellaneous crimes and is more pronounced in North Rhine-Westphalia (where Cologne is located), in wealthier counties, in counties with a higher share of refugees, and in the counties that had a refugee reception center. The immediate rise in anti-refugee crimes is also higher in counties where a higher share of German suspects has been involved in crimes against foreign victims. Regarding longer-term repercussions, I find evidence for an anniversary effect a year later, i.e., a rise in anti-refugee crimes after the next New Year’s Eve.