Use of Poset Theory with Big Datasets: A New Proposal Applied to the Analysis of Life Satisfaction in Italy

The aim of this work is to propose a tool for measuring a complex concept, and to apply it to big sets of data measured on ordinal and/or dichotomous scales. An important field of application are subjective data, that are often based on opinions or personal evaluations. Many national and international surveys employ this kind of data, measured among thousand of individuals. Thanks to the use of the “average rank” as a synthetic measure of a complex concept, we believe that poset theory could be a very useful approach for dealing with ordinal data avoiding the use of scaling procedures. Because classic poset approaches are at their best when applied to few data at a time, our idea is based on a procedure for sampling units from a big population using a simple criterion to summarize the resulting values appropriately. Applying the central limit theorem enables a comparison of the results obtained from different groups using statistical tests on the means. We used our Height of Groups by Sampling (HOGS) method to compare the average rank among groups that are defined by one or more socio-demographic variables influencing the level of the complex concept we wish to measure. The application of the HOGS procedure to life satisfaction in Italy generated convincing results, revealing significant differences between regions, genders and levels of formal education. We compared the results given by HOGS with the non linear principal component analysis and obtain an easy readable output with convincing precision and accuracy; we are confident that the HOGS procedure can be applied to many other concepts investigated in the social sciences.     

Is a loosely specified regulatory intervention effective in disciplining management commentary? The case of performance indicator disclosure

This paper offers a contribution to the call for research on the effectiveness of regulatory interventions governing management commentary disclosure. Specifically, we focus on the mandatory requirement concerning performance indicator disclosure introduced by the Modernisation Directive (2003/51/EC). In keeping with other regulators, the European legislator opted to implement a regulatory approach based on a rule with loose specifications. To understand the effects of this Directive, we have investigated the Italian context, in which neither the national legislator nor the standard setter have supported companies with detailed specifications or guidelines aimed at integrating the European provision. We have compared companies’ disclosure practices before and after the adoption of the Directive, investigating the number of disclosed indicators and also their modalities of presentation, as they are considered essential to guaranteeing the quality of this disclosure. Our results document that a mandatory intervention, even if based on loose specifications, is associated with an increase in disclosure practices. Nevertheless, such regulation does not seem able to guarantee high quality disclosure practices. In fact, before and after the regulation, companies primarily disclose common financial measures. Moreover, the usefulness of such disclosure is undermined by a limited compliance with the international guidelines concerning the modalities of presentation. These results reveal some weaknesses in the European approach to performance indicator regulation. In general, the EU legislator fails to explain the purpose and the objective of performance indicator disclosure and does not take into account the differences between financial and non-financial indicators. Furthermore, it does not provide firms with clear guidelines concerning the presentation modalities.     

The role of the board in shaping foundations’ strategy: an empirical study

This paper investigates the relationships of the choice of philanthropic strategy with board capital (diversity and networks), board activities (board processes, internal board committees, and board effectiveness), and CEO leadership. Using a sample of 110 Italian foundations, the research shows that board processes have the strongest positive association with an evolved strategic approach to philanthropic institutional grant-giving, while board diversity and strong CEO leadership are associated with the strategic approach only under certain conditions. In particular, good governance processes (e.g., training the board, self-evaluation of trustees, setting the stage for effective board and committee meetings, implementing control software, and steering meetings to improve the board’s analysis) are positively associated with evolved strategic approaches to philanthropy (e.g., signaling other funders for the best grantees, improving the performance of grant recipients, and advancing selected social fields’ state of knowledge and practice).     

Minority and majority private equity investments: firm performance and governance

This paper adds to the literature on the determinants of the effects of private equity (PE) investments. Using an original dataset of 191 target firms in Italy, we study the effects on performance and governance of the stakes acquired by the PE investor. We employ a difference-in-differences approach and compare target and control firms sharing similar characteristics and performance in the years preceding the deal. We find that PE investment has a positive effect on profitability, sales, and employment; these effects are larger for minority investments. We argue that this signals effective governance that follows from complementing rather than substituting incumbent managers in minority investments.     

Measuring freedom of choice: An alternative view of a recent literature

A recent literature has emerged in social choice theory which attributes intrinsic importance to freedom in the evaluation of states of affairs. The literatures philosophical basis lies in Berlins notion of positive liberty. Accordingly, axiomatic measures of availability of choices are developed and the information they convey used for ranking states on the basis of the extent of liberty they offer to individuals. I am grateful to Nancy Cartwright, Dan Hausman and Marco Del Seta for their suggestions to previous drafts of this paper. The usual caveat applies. The support of the Centre for Philosophy of Natural and Social Science at the London School of Economics is also acknowledged.     

Small area estimation for longitudinal surveys

Over the last few years many studies have been carried out in Italy to identify reliable small area labour force indicators. Considering the rotated sample design of the Italian Labour Force Survey, the aim of this work is to derive a small area estimator which borrows strength from individual temporal correlation, as well as from related areas. Two small area estimators are derived as extensions of an estimation strategies proposed by Fuller (1990) for partial overlap samples. A simulation study is carried out to evaluate the gain in efficiency provided by our solutions. Results obtained for different levels of autocorrelation between repeated measurements on the same outcome and different population settings show that these estimators are always more reliable than the traditional composite one, and in some circumstances they are extremely advantageous.The present paper is financially supported by Murst-Cofin (2001) Lutilizzo di informazioni di tipo amministrativo nella stima per piccole aree e per sottoinsiemi della popolazione (National Coordinator Prof. Carlo Filippucci).     

Technology spillover and regional convergence process: a statistical analysis of the Italian case

Recent literature broadly highlight the importance of modelling technological innovation effects on economic growth. This paper develops a methodology that allows to measure technology contribution to economic convergence; the choice of a regional framework also allows to underline interregional knowledge transmission as a the major channel of technological progress. Moreover, the specification of a dynamic growth model enables to evaluate both the regional convergence and the effect of innovation on long-run labour productivity without resorting to any technology index measurement. We contribute to the methodological literature also by comparing different dynamic panel data estimation procedures and by detecting both the presence of small sample bias and the existence of a nearly unit root autoregressive process in labour productivity series. The results of an empirical analysis on Italian regions show how most of innovation resources derives from relevant spillover mechanisms. Furthermore, technology spillover intensity seems to be strongly affected by geography and productive structure of regions.Financial support by Fondazione Cassa di Risparmio di Rimini and Uniturim S.p.A. is gratefully acknowledged.     

The univariate distribution function for a particular bilinear model

The aim of the paper is to find the univariate stationary distribution of a particular bilinear process. In this context, we propose a novel approach to derive the distribution function. It is based on a recursive formula and allows to relax the conditions on the moments of the process. We also show that the derived approximation converges to the true distribution function. The accuracy of the recursive formula is evaluated for finite sample dimensions by a small simulation study.Received: February 2003, Revised: May 2004,     

Optimal use of two auxiliary variables in double sampling

Double sampling scheme is used when cheap auxiliary variables may be measured to improve the estimation of a finite population parameter. Several estimators for population mean, ratio of means and variance are available, when two dependent samples are drawn. However, there are few proposals for the case of independent samples. In this paper both cases of dependent and independent samples are dealt with. A general approach for estimating a finite population parameter is given, showing that all the proposed estimators are particular cases of the same general class. The minimum variance bound for any estimator in this class is provided (at the first order of approximation). Furthermore, an optimal estimator which reaches this minimum is found.     

A general algorithm to fit constrained DEDICOM models

The DEDICOM model is a model to analyze square tables describing asymmetric relationships among n entities. Its importance in the asymmetric multidimensional scaling literature is due to the fact that several authors showed a large class of models to be simply a constrained version of DEDICOM. A typical example is the Generalized GIPSCAL proposed by Kiers & Takane. In this paper we present a new algorithm capable to fit, in the least squares sense, any DEDICOM constrained model.