Tom Schwantje

Presentations: 10th CSEF-IGIER Symposium on Economics and Institutions*, EEA Congress 2025, CEPR-IMO Workshop (Bocconi), CEPR Governance in a Polarized World (POLECONFIN), EBRD Office of the Chief Economist Seminar, CEPR Symposium, Working Group in African Political Economy; 8th EBRD-CEPR Research Symposium: The Frontiers of finance in emerging markets; Groningen internal seminar; SIOE 2026 Conference. (* if by coauthor)

Abstract: How do organizations adapt internally when ethnic divisions intensify? We develop a model in which a bank jointly chooses the ethnicity of its branch manager and how much lending authority to delegate. Managers who share local borrowers' ethnicity acquire better information about them but are also more prone to favor them at the bank's expense. Conflict changes the relative importance of these two frictions, shaping both optimal managerial appointments and delegation. We test the model using a new panel of 979 Ethiopian bank branches surveyed before and after the onset of the 2020 civil war, exploiting variation in banks' exposure to ethnic conflict across their branch networks as a source of plausibly exogenous variation. Where the bank's and city's ethnicities differ, a one-standard-deviation increase in exposure raises the probability of appointing a manager sharing the city's predominant ethnicity by 6.7 percentage points, or 36 percent, while reducing delegated lending authority. Conflict-exposed branches are also more likely to be staffed by experienced insiders reassigned within the bank. We use an LLM-based CEO vignette exercise to further examine the underlying mechanisms.

Presentations: CSAE Research Workshop; Oxford Applied Micro Group; Workshop on Frontiers in Measurement and Survey Methods; Empirical Management Conference; Frontiers in Management Research Workshop.

Abstract: It is well understood that most organisational practices are communicated and understood through rules, but these have traditionally been difficult to measure well. This paper develops a novel Bayesian survey method to empirically measure the use of managerial rules within organisations. The method implements an adaptive survey, incorporating Bayesian optimisation, to first understand decision-making in a potentially high-dimensional space. It uses the information from this method to then directly ask about the rules driving patterns in these decisions. I apply this tool to measure the rules used by senior HR managers at Ethiopian firms. Descriptively, I find that this measure robustly, positively correlates with firm performance and helps explain variation in performance across firms, in addition to more traditional measures of firm management. Using quasi-exogenous variation from firms' exposure to COVID-19, I further find that while managerial rules enhance firm performance, they also reduce resilience to economic shocks, suggesting a trade-off in the use of more rigid practices. The methods and findings in this paper motivate further research into the causal mechanisms underlying the effectiveness of managerial rules, and into understanding the factors influencing the heterogeneous adoption of such rules across organisations.

Presentations: City St George's (University of London)*, 10th CSEF-IGIER Symposium on Economics and Institutions, Princeton University*, the NBER Organisational Economics Working Group Spring 2025*, King's College London*, Queen's University (Firms and Inclusive Growth Workshop), Imperial College London*, CEPR Symposium*. (* if by coauthor)

Abstract: We introduce a new method for measuring managerial traits of young professionals: using management vignettes in a video studio. This method — analysed through the lens of a Bayesian hierarchical model — allows us to identify four distinct managerial archetypes (which we term ‘rule-based’, ‘affiliative’, ‘power-based’ and ‘discretion-based’). We find that past labour market exposure (including exposure induced through a previous field experiment) correlates strongly with the propensity to act as a rule-based manager. We then use the videos to run an incentivised experiment with firm managers, to elicit preferences over young professionals. Strikingly, we find that firms consistently prefer the rule-based managerial style for entry-level managerial positions. Empirically, our results highlight an underexplored mechanism for labour market exclusion among young professionals. Methodologically, we demonstrate the value of controlled vignette scenarios for assessing managerial traits. Our findings underscore the importance of managerial training in shaping labor market outcomes, and offer new avenues for studying the development of managerial talent.

Presentations: CSAE Research Workshop (2023), Food for Thought Seminar (Bocconi), CSAE Conference 2025, Milan Experimental Coffee, CEPR-EBRD 3rd Accelerating growth for women entrepreneurs joint research conference (EBRD). Upcoming: CEPR Paris Symposium.

Abstract: We conduct a field experiment within a business plan competition to examine how institutional features influence inclusive decision-making. In the control group, over the course of independent sequential assessments, evaluators become less likely to recommend female candidates. This pattern reduces the quality of their decisions, as measured by expert assessments and machine-learning methods. Informing judges of the organisation's commitment to equal opportunity entirely offsets this decline, whereas requiring judges to justify decisions to peers has a more muted effect. Our results show that decision fatigue can undermine both decision quality and inclusivity, but simple organisational messaging can resolve this.