The complex interactions between human behaviors and social economic sciences is critically analyzed in this paper in view of possible applications of mathematical modeling as an attainable interdisciplinary approach to understand and simulate the aforementioned dynamics. The quest is developed along three steps: Firstly an overall analysis of social and economic sciences indicates the main requirements that a contribution of mathematical modeling should bring to these sciences; subsequently the focus moves to an overview of mathematical tools and to the selection of those which appear, according to the authors bias, appropriate to the modeling; finally, a survey of applications is presented looking ahead to research perspectives.

This study examines the impact of education on criminal activity in Italy. A number of hypotheses are tested regarding the effects of education and past incidence of crime on criminal activity, using annual data for the 20 Italian regions over the period 1980 to 1995. Empirical results show that education is negatively correlated with delinquency and that crime rates display persistence over time. The results are robust to model specifications and endogeneity.

This paper compares the performance of binomial and multinomial logit models in the context\ud of building early warning systems (EWS) for systemic banking crises. We test the hypothesis\ud that the predictive performance of binomial logit models is hampered by what we define as\ud the crisis duration bias, arising from the decision to either treat crisis years after the onset of\ud a crisis as non-crisis years or remove them altogether from the sample. In line with our hypothesis,\ud results from a large sample of world economies suggest that i) the multinomial\ud logit outperforms the binomial logit model in predicting systemic banking crises, and ii) the\ud longer the average duration of the crisis in the sample, the larger the improvement

Using a conviction-based measure, we find that local (state-level) public corruption exerts a negative effect on the lending activity of US banks. Our baseline estimations show that the difference in public corruption between, for example, Alabama, where corruption is high, and Minnesota, where corruption is low, implies that banks headquartered in the former state grant 0.55% less credit (or $3.52 million for the average bank) ceteris paribus. Using proxies for relationship lending and monitoring, we also find that these bank characteristics weaken the negative effect of public corruption on lending. These results are robust to tests that address endogeneity, to the use of perception-based measures of corruption, and after controlling for credit demand conditions. In further analysis, we show that these effects are more evident for smaller banks and banks operating in a single state. These findings provide evidence that public corruption could facilitate information asymmetry in the lending market and, thus, could hinder local development by reducing bank credit.