Publications

In finitely repeated public goods games, contributions are initially high, and gradually decrease over time. Two main explanations are consistent with this pattern: (i) the population is composed of free-riders, who never contribute, and conditional cooperators, who contribute if others do so as well; (ii) strategic players contribute to sustain mutually beneficial future cooperation, but reduce their contributions as the end of the game approaches. This paper analyzes experimentally these explanations, by manipulating group composition to form homogeneous groups on both the preference and the strategic ability dimensions. Our results highlight the role of strategic ability in sustaining contributions, and suggest that the interaction between the two dimensions also matters: we find that groups that sustain high levels of cooperation are composed of members who share a common inclination toward cooperation and also have the strategic abilities to recognize and reap the benefits of enduring cooperation.

In this paper, we investigate what are the drivers of cross-border equity acquisitions made by Sovereign Wealth Funds (SWFs) of the Gulf Cooperation Council (GCC) countries. GCC SWFs are considered as relatively opaque investors and strongly politicized, raising some concerns for perceived political and security risks. Using both Logit and ordered Logit models, we test if the usual determinants of SWFs investments still stand when we look at large or majority acquisitions made by GCC SWFs. Unlike results found in the literature investigating the determinants of SWFs cross-border investments, we find that GCC SWFs do not take into account the financial characteristics of the target firm by taking majority stakes, apart from its financial wealth. The economic, institutional and financial factors of the target country as well as the existence of trade agreements between both countries do not matter in their acquisition/control decision. We also find that firms operating in strategic sectors are targeted by GCC SWFs for diversification purposes but not for the purpose of acquisition or control. Overall, our results lend support to the hypothesis that GCC SWFs differ from other institutional investors in terms of acquisition decision strategy and that financial and commercial motives are not the exclusive target of their acquisition strategy.

In the context of credit scoring, ensemble methods based on decision trees, such as the random forest method, provide better classification performance than standard logistic regression models. However, logistic regression remains the benchmark in the credit risk industry mainly because the lack of interpretability of ensemble methods is incompatible with the requirements of financial regulators. In this paper, we propose a high-performance and interpretable credit scoring method called penalised logistic tree regression (PLTR), which uses information from decision trees to improve the performance of logistic regression. Formally, rules extracted from various short-depth decision trees built with original predictive variables are used as predictors in a penalised logistic regression model. PLTR allows us to capture non-linear effects that can arise in credit scoring data while preserving the intrinsic interpretability of the logistic regression model. Monte Carlo simulations and empirical applications using four real credit default datasets show that PLTR predicts credit risk significantly more accurately than logistic regression and compares competitively to the random forest method.

Most studies examining the impact of migrants on crime rates in hosting populations are in the context of economic migrants in developed countries. However, we know much less about the crime impact of refugees in low- and middle-income countries—whose numbers are increasing worldwide. This study examines this issue in the context of the largest refugee group in any country—Syrian refugees in Turkey. Although these refugees are much poorer than the local population, have limited access to formal employment, and face partial mobility restrictions, we find that total crime per person (including natives and refugees) falls due to the arrival of the refugees. This finding also applies to several types of crime; the only exception is smuggling, which increases due to the population influx. We also show that the fall in crime does not result from tighter security; we find no evidence of a change in the number of armed forces (military and civil personnel) in the migrant-hosting regions.