Iacopo Odoardi

Looking for Regional Convergence: Evidence from the Italian Case with Multivariate Adaptive Regression Splines

This paper examines the role of data mining analysis in explaining the Italian regional dualism with the aim of suggesting economic policies to fill the existing socio-economic gaps. We analyze the 2004–2014 period exploiting the capacity of MARS model in finding relationships among data. In Italy, the presence of a North-South divide is well-known for decades and present for several social and economic aspects. Recent studies prove that strong differences exist also in the regional human capital. Thus, we search for the causes of the local differences, also considering the entrepreneurial vitality and the international trade leverage. Among several variables, MARS is useful in showing the actual determinants on which to intervene. This is possible by comparing regions grouped homogeneously into clusters using recent data. MARS results are used for policy suggestions with the aim of filling the income gap.

Understanding the Support of Savings to Income: A Multivariate Adaptive Regression Splines Analysis

The understanding of the complex socio-economic phenomena requires a deep insight of the dynamics through several correlated variables. Our aim is to demonstrate how some relevant macroeconomic variables could affect the evolution of the economic development path. This research arises in a broader analysis on the role of the wealth of households held in specific forms unexploited and potentially useful if properly integrated in productive cycles. At the basis of these assumptions, however, there are complex dynamics on the formation and composition of wealth, which does not include business capital and savings in the bank channel. We want to demonstrate the utility of aggregate savings for the composition and the very existence of wealth, which can be stimulated through policy instruments. We need to ask how to prove empirically that, in modern economies, encouraging savings and the accumulation of private wealth represent, under certain conditions, a not fully considered support to economic development. In this phase we test if statistical techniques inspired by Artificial Intelligence can be better exploited respect to classic approaches, given the quality of the data available. We use multivariate adaptive regression splines model, in a comparison with a multivariate model, to examine the relationship of aggregate savings, and then other related variables, on GDP in the US for the period 1970-2012.

The North-South Divergence in Italy during the Great Recession

The economic recession that followed the 2007 crisis has widened the economic gaps between the wealthiest and the relatively poorer regions in Italy. The Great Recession has changed the importance of local economic strengths and hindered the possibilities of economic recovery, especially in the Mezzogiorno of Italy. We seek the local strengths present in Italian regions in the post‐crisis period by comparing two macro areas to observe strong and weak points for intervention. A first analysis using multivariate adaptive regression splines (MARS) is used to filter the relevant determinants in a large dataset, and a panel data analysis serves to obtain group‐specific results. Some effects of the prolonged recession are confirmed in all regions, while some weaknesses of the South, such as financial markets, play an increasing role in the regional development scenarios.

A Data Mining Analysis of the Chinese Inland-Coastal Inequality

As in many countries, even in China the socio-economic changes have affected income inequality in recent decades. The various economic opportunities have led to different paths of development causing severe disparities in GDP per capita level. In addition to the well-known Chinese rural/urban inequality, in this work we study the inland/coastal differences. There are many known causes of inequality, but we aim to discover the actual determinants of the local GDP and, therefore, of income in a period that includes the international economic crisis started in 2007. With this aim, we use different variables to obtain clusters of the Chinese provinces in the period 2004–2015 and, subsequently, we investigate the determinants of income with a multivariate adaptive regression splines (MARS). There is an extensive economic literature on the Chinese case: MARS allows us to integrate this literature enabling us to find which GDP determinants are the most relevant in the certain areas of China.

Improving the Relationship between Welfare Economics and Ethics

In this essay, the essential aspects of Welfare Economics are summarized in order to analyse the possible links with Ethics, and consequently to provide useful suggestions for Economic Policy. The Introduction contains the principles of neoclassical Welfare Economics, without considering the circumstances in which the so‑called failures of competitive markets occur. The second part of this paper contains a description of two possible solutions to the problem concerning the integration of Welfare Economics and Ethics. The first solution is the determination of the optimal combination of general competitive equilibrium principles and social ethics ones, without removing the theoretical structure of Welfare Economics. This first proposal, while retaining some economic liberalism principles, assigns a central role to the government, which is delegated to superimpose on the Pareto criterion a distributive justice rule. Another solution, which is suggested by Sen, is essentially directed to modify Welfare Economics by ethical criteria aimed at improving every individual deprivation, and it is based on the distinction between the two concepts of ‘utility’ and ‘agency’. We also proposed exploiting the original considerations written by Smith about social aspects as a useful integration and support to Sen’s approach.

Regional Income Differentials in Italy: A MARS Analysis

We analyse the influence of 45 regressors that explain the existence of income differentials and other socio-economic spreads in the Italian regions. We use a multivariate adaptive regression splines analysis as a data-driven approach to detect relationships among variables in big data sets. The focus on regional areas allows us to consider economic contexts historically extremely distant in terms of growth and development. The considered time series include the pre-crisis economic period, the negative effects of the crisis and the originated recession. The independent variables chosen are based on the endogenous growth theory according to the knowledge economy, thus focusing on human capital and other intangible assets. Macroeconomic data and entrepreneurial competitiveness are also considered as control variables. The results are consistent with the findings of the recent economic literature.

A Comparative Analysis of Literacy Rate in Contributing to Social Exclusion Insights

Our contribution aims to analyse across countries the relationship between the phenomena of social exclusion and literacy levels, and to consider the implications of this relationship on economic growth. A low level of literacy, affecting the employment status, precludes the possibility to enter and operate freely in society, while poverty and persistent social exclusion of a person or family make difficult to address appropriate educational and training paths. Therefore, our study has rejoined two issues which influence almost all decisions adopted by policy makers, especially in the Western world. The opening issue is the level of education, which should constitute the human capital of a country, through appropriate investment, and the second one is the relational conditions of social dynamics, which highlight the so-called social capital. Together these two types of intangible capitals constitute a strong support for the long-term development of a country. Our quantitative analysis is also addressed to detect differences and peculiarities among the different national realities, with the ultimate purpose to recognize which socio-economic variables affect more directly the processes of education.