Workshop on "Recent Advances in Public Economics and Quantitative Methods"

This joint workshop organized with KU Leuven's faculty of Economics and Business features three keynote talks, to be held by Prof. Benny Geys (BI Norwegian Business School), Prof. Ana Camanho (University of Porto) and Prof. Kristof De Witte (KU Leuven, Maastricht University), and two student presentation sessions (one in the morning and one in the afternoon). The day will conclude with the PhD thesis defense of Giovanna D'Inverno, a student from the Cycle XXX "Economics" Curriculum (more information regarding the defense can be found here: https://www.imtlucca.it/news-events/events/thesis-defense) The abstracts of all talks can be found below, while the full program can be found on the poster linked below. Welcome: 9:00 am Prof. Massimo Riccaboni (IMT School) Morning keynote speakers: 9:05 am Prof. Benny Geys (BI Norwegian Business School) Title: Abandon Ship? Party Brands and Politicians' Responses to a Political Scandal Abstract: How do politicians react to a political earthquake? In this article, we study politicians' - rather than voters' - responses to the main political scandal in Italian recent history (Tangentopoli), and overcome endogeneity concerns by analysing the local implications of this national corruption scandal. We find that local politicians withdraw support for incumbents in parties hit by Tangentopoli - inducing early government terminations in such municipalities. Moreover, politicians in parties hit by the scandal exhibit higher rates of party switching and lower re-running rates. By decreasing the value of the party "brand", scandals thus become transmitted across politicians and levels of government via partisan cues. Prof. Ana Camanho (University of Porto) Title: Regulation and benchmarking: a formative evaluation of Portuguese wastewater services using composite indicators Abstract: This paper describes a framework to enhance the evaluation of performance of wastewater companies based on the indicators collected by the Portuguese water and waste services regulation authority (ERSAR). Our assessment is focused on wholesale and retail services, contributing to increased accountability in the water sector. It considers three main dimensions of performance: protection of user interests, operator sustainability, and environmental sustainability. We develop a composite indicator to provide a summary measure of utilities performance. The composite indicator is estimated using a directional distance function with weight restrictions in the form of assurance regions type I, to allow the specification of alternative perspectives concerning the importance of the different dimension of company performance. We illustrate our approach with data of 19 wholesale companies and 264 retail companies in the year 2014. In a second stage, a truncated regression model is used to analyze the impact of contextual factors on companies' activity. The information provided by this analysis constitutes the basis of a formative evaluation framework that the regulator can use to disseminate best practices and motivate continuous improvement. Afternoon Keynote Speaker 13:30 Prof. Kristof De Witte (KU Leuven, Maastricht University) Title: The persistent effect of funding for disadvantaged students. Evidence from a Stochastic Frontier Model Abstract: This paper examines the effect of additional funding for disadvantaged students on underachievement of primary school pupils. We apply an innovative approach that accounts for heteroscedasticity and separates persistent underachievement, time-varying underachievement, pupil heterogeneity and random noise. This contrasts to earlier literature that has traditionally measured underachievement as the difference between capacity test scores, such as IQ, on the one hand, and achievement test scores or school grades on the other. We exploit rich data from the Flemish region of Belgium, where schools receive additional funding for disadvantaged students if an exogenous threshold is met. Our results suggest that additional funding reduces underachievement and that teachers use the additional resources to minimize underachievement among those student who underachieve most. Session 1: 10:15 am Helena Kreuter (IMT Lucca) Title: Low-wage import competition and populist backlash: the case of Italy Abstract:In this paper we empirically study the role of trade globalization in shifting the electoral base toward populism. We proxy trade shock with swiftly rising import competition from China and compare the voting pattern at the parliamentary national elections from 1992 to 2013 in about 8,000 Italian municipalities differently exposed to the trade shock. To address the potential endogeneity of import exposure we instrument it with Chinese export flows to other high-income countries and estimate the model in first differences. Our results show that trade globalization increases support for populist parties and they are robust to a large number of sensitivity checks. Moreover, we show that voters' protest reaction also takes the form of an increase in invalid ballot papers and a drop in turnout. To rationalize these findings, we further offer evidence that low-wage import competition worsens labor market conditions - higher unemployment and lower income - and is associated with a rise in inequality, as predicted by trade theory. Vitezslav Titl (KU Leuven) Title: Political Connections and Competition on Public Procurement Markets Abstract: We study the effects of political connections on the public procurement market. Our main findings suggest that political connections -- measured both by political affiliation of firm representatives and implicit connections based on frequency of contracts between a procurer and a supplier -- increase the final price of procurement contracts allocated to connected firms. Our evidence also suggests that public procurement authorities restrict competition in order to help connected firms. Moreover, the projects awarded to connected firms tend to be of lower quality as measured by the probability of repairs. To quantify the total distortion, we estimate a structural model of the public procurement market where connected firms are being favored by the procurers. Subsequently, we calculate a lower bound of the total welfare loss caused the political connections. Our conservative estimate suggests that tenders allocated to favored firms are overpriced by 3.8\% which would sum up to 5 billion CZK in the total sample of contracts in value of 227 billion CZK (due to data restrictions, we used sub-sample in the estimation of the structural model, in which the total welfare loss sum up to half a billion CZK). Sara Landi (IMT Lucca) Title: Central Government and Contestable Municipalities. Do higher national transfers improve local public good provision? Abstract: Transfers from the central government are critical in the provision of public goods at local level as they constitute the main source of revenue in the municipality budget. Theoretical models give discordant interpretations to politically driven transfers. Those might be used to increase their reelection probability by targeting swing voters (see Lindbeck and Weibull 1987), as well as to reward core supporters (see Cox and McCubbins 1986). Moreover, the central government could also allocate transfers to benefit (political) friends or to penalize (political) enemies locally. Consequently, political alignment between higher and lower strata of governance is expected to increase the amount of transfers.​ In this paper we would explore the relationship between extra transfers (i.e. they exceed the amount of transfers the municipality needs to cover ordinary expenses) to political aligned municipalities and local electoral consensus. Is contestability associated with higher national transfers when major and government have the same political affliation? In the run-up to Election Day, does the central government distribute pork barrel evenly across municipalities or are swing municipalities preferred? Do voters reward national governments for increasing transfers to their jurisdictions? Falco Bargagli Stoffi (IMT Lucca) Title: Machine Learning Zombie Hunting Abstract: The debate on zombie firms, i.e., non-viable albeit surviving firms, is growing in importance in recent years. The debate revolves around the way to identify zombie firms from a financial perspective, as well as on how to estimate their impact on a country welfare. This holds true in particular for those countries, such as Italy, where a high number of financial zombies is detected, and where their aggregate effects on productivity and capital misallocation may be particularly disruptive. In this paper we adopt machine learning techniques to investigate the ability of main indicators to identify the persistence of a firm in a high-failure-risk situation. Namely, we assess which are the more salient indicators to identify what we call surviving zombies: firms that are predicted to have a high failure risk in three consecutive years, but manage to remain on the market. In order to classify surviving zombies, we pick from machine learning toolbox to assess the failure risk of firms in a non-parametric way. The technique that we use seems to outperform otherwise standard econometric models in predicting firms' failures in presence of missing-not-at-random observations. Eventually, we build and evaluate the performance of Bayesian Additive Regression Trees (BART) on a unique dataset of 123,815 Italian manufacturing firms in the period 2009-2016. Finally, we compare with a counterfactual to assess the aggregate impact of surviving zombies on the Italian manufacturing sector in terms of employment, revenues and value added. We argue that the study of surviving zombies is central to highlight where market selection mechanisms fail. The use of machine learning non-parametric techniques paves the way to the implementation of relatively more target-specific policies. Jan Niederreiter(IMT Lucca) Title: Learning from Experience in the Pharmaceutical Industry Abstract: Innovation has often been conceptualized as a pure "lottery model" where previous innovation does not influence current- and future innovation in the same market. In this paper we employ an experimental design that varies by the number of players and by the degree of uncertainty to study investment behavior and learning under different scenarios, namely share- and lottery contests. In line with previous findings, we show that excess of expenditures in share contests are rapidly converging to their predicted equilibria whereas pervasive overdissipation remains in winner-take-all pure lottery games. A significantly lower and stagnant fraction of equilibrium plays combined with a significantly higher rate of non-participation in the pure lottery contests advocates that the difference between contest structures are driven by differences in learning paths. A learning parameter estimation using the experience weighted attraction (EWA) model shows that lottery participants rely more on own realized payoffs while contestants in share contests use more adaptive learning techniques. These results are robust to changes in the number of players. We hypothesize that the strong reliance on experiential learning in high uncertainty R&D domains makes it difficult to learn, due to the random nature of the contest outcomes. Our setting mimics environments with high outcome uncertainty, such as R&D endeavors in the pharmaceutical industry, where it is challenging to infer optimal investment choices based on past experience. Session 2 14:00 Fritz Schilz (KU Leuven) Title: Prediction errors and school rankings Abstract: School rankings based on value-added (VA) estimates are subject to prediction errors, since VA is defined as the difference between predicted and actual performance. We introduce a more flexible random forest (RF), rooted in the machine learning literature, to minimize prediction errors and to improve school rankings. Monte Carlo simulations demonstrate the advantages of this approach. Applying the proposed method to data on Italian middle schools indicates that school rankings are sensitive to prediction errors, even when extensive controls are added. RF estimates provide a low-cost way to increase the accuracy of predictions, resulting in more informative rankings, and better policies. Kaat Iterbeke (KU Leuven) Title: Pupils' and teachers' perspectives on homosexuality - a comparative analysis in eight European countries Abstract: Homonegativity remains a burning issue in many European societies. In this paper, we provide the first large-scale comparative analysis of teachers' and pupils' perspectives on homosexuality in eight European countries. Results indicate significant differences across countries. Therefore, panel data is applied in order to identify an underlying mechanism explaining the differences and changes in the pupils' and teachers' perspectives. Policy recommendations may be drawn from the analysis as the topic of sexual diversity is still not well integrated into most curricula of education. Boukje Compen (University of Antwerp) Title: The role of teacher professionalization in financial literacy. A literature review. Abstract:During the last two decades, the global recognition of the importance of financial literacy has strongly increased. Since financial literacy levels have consistently been found to be insufficient, national authorities started integrating financial education in school curricula. While existing literature emphasizes the importance of well-trained teachers for effective financial education, the process of teacher professional development (TPD) seems to have received little attention. By combining a general TPD model with literature on TPD needs for financial education in specific, this review aimed to reveal the extent to which professional development in the financial education context has been examined. A systematic literature review was performed that led to 966 unique records, of which 51 eventually met all inclusion criteria. As assumed, results demonstrate that very few articles directly aimed to evaluate professional development initiatives in financial education. Furthermore, for various reasons, it proved difficult to derive effective TPD features from previously tested financial education programmes. All in all, the review reveals a strong demand for research investigating the professionalization process of teachers in financial education. Joana Elisa Maldonado (KU Leuven) Title: Tools to Involve Parents in Financial Education. A Literature Review. Abstract:In today's increasingly complex market economies, financial literacy has become the basis of generating, managing and utilising earnings. Incorporating financial skills and knowledge in the compulsory education could ensure a minimum level of financial literacy and amplify the effects of education interventions beyond a single generation by involving parents. While economic survey studies have shown that parents determine the financial capability of their children and parental involvement programmes in schools are becoming more popular, the research on the impact of these interventions remains limited. Existing studies on parental involvement are predominantly qualitative and correlational, yet, robust experimental methods are necessary to identify causal links between interventions and learning effects. This explorative literature review examines which tools for parental involvement in school-based education effectively increase children's or parents' knowledge, skills or desirable attitudes and behaviours. In a first part, the paper illustrates the role of parents in financial education in the four dimensions of behaviour, motivation, reciprocity and persistence. In a second part, the paper reviews the parental involvement tools used in the selected experimental literature within a framework that classifies the tools by interaction partner of the parent, i.e. teacher or child, and by location of the involvement, i.e. home-based and school-based tools. The review finds that information and homework are the most promising tools for parental involvement in financial education. Further, parent events have proven to be successful in combination with other involvement tools. Overall, only few studies decompose the effects of the different programme components. The paper therefore concludes that further experimental evidence on the learning effects of different parental involvement tools is needed, in particular on experiential learning methods, such as family tasks, educational items and family events. Kenneth De Beckker (KU Leuven) Title: Identifying vulnerable groups An application to financial literacy Abstract: Targeted policy interventions are more effective than one size fits all initiatives. This paper proposes the use of k-means cluster analysis to identify vulnerable groups with respect to financial literacy. Using a sample of 12 countries, we distinguish 8 groups and examine their socio-economic characteristics. The results suggest that individuals in the most vulnerable groups are on average, single, low-educated, unemployed and earn a low income. This contrast with those in the strongest group: individuals with the highest financial knowledge, financial behaviour and financial attitudes scores are on average highly-educated males who live together with a partner. Financially speaking they earn a high income and hold several financial products. Applying the insights from this paper on a national scale will not only lead to more effective but also to more efficient policy initiatives to improve financial literacy.