Caterina Giusti

Small Area Model-Based Estimators Using Big Data Sources

Abstract The timely, accurate monitoring of social indicators, such as poverty or inequality, on a finegrained spatial and temporal scale is a crucial tool for understanding social phenomena and policymaking, but poses a great challenge to official statistics. This article argues that an interdisciplinary approach, combining the body of statistical research in small area estimation with the body of research in social data mining based on Big Data, can provide novel means to tackle this problem successfully. Big Data derived from the digital crumbs that humans leave behind in their daily activities are in fact providing ever more accurate proxies of social life. Social data mining from these data, coupled with advanced model-based techniques for fine-grained estimates, have the potential to provide a novel microscope through which to view and understand social complexity. This article suggests three ways to use Big Data together with small area estimation techniques, and shows how Big Data has the potential to mirror aspects of well-being and other socioeconomic phenomena.

Resistance to outliers of M-quantile and robust random effects small area models

The presence of outliers is a common feature in real data applications. It has been well established that outliers can severely affect the parameter estimates of statistical models, for example, random effects models, which can in turn affect the small area estimates produced using these models. Two outlier robust methodologies have been recently proposed in the small area literature. These are the M-quantile approach and the robust random effects approach. The M-quantile and robust random effects approaches are two distinct outlier robust small area methods and a comparison between these two methodologies is required. The present paper sets to fulfill this goal. Using model-based simulations and showing an application to real income data we examine how the alternative small area methodologies compare.

Small Area Estimation of Poverty Indicators

The estimation of poverty, inequality and life condition indicators all over the European Union has become one topic of primary interest. A very common target is the core set of indicators on poverty and social exclusion agreed by the Laeken European Council in December \(2001\) and called Laeken indicators. They include measures of the incidence of poverty, such as the Head Count Ratio (also known as at-risk-of-poverty-rate) and of the intensity of poverty, as the Poverty Gap. Unfortunately, these indicators cannot be directly estimated from EU-SILC survey data when the objective is to investigate poverty at sub-regional level. As local sample sizes are small, the estimation must be done using the small area estimation approach. Limits and potentialities of the estimators of Laeken indicators obtained under EBLUP and M-quantile small area estimation approaches are discussed here, as well as their application to EU-SILC Italian data. The case study is limited to the estimation of poverty indicators for the Tuscany region. However, additional results are available and downloadable from the web site of the SAMPLE project, funded under the 7FP (http://www.sample-project.eu).

Multilevel mixture factor models for the evaluation of educational programs’ effectiveness

Factor models aim at explaining the associations among observed random variables in terms of fewer unobserved random variables, called common factors. When data have a hierarchical structure, multilevel mixture factor models are a powerful and flexible tool useful to correctly take into account the correlation between first-level units due to the data structure, and to evaluate the presence of latent sub-populations of units with some typical profile at different levels of the analysis.

Small area estimation based on M-quantile models in presence of outliers in auxiliary variables

When using small area estimation models, the presence of outlying observations in the response and/or in the auxiliary variables can severely affect the estimates of the model parameters, which can in turn affect the small area estimates produced using these models. In this paper we propose an M-quantile estimator of the small area mean that is robust to the presence of outliers in the response variable and in the continuous auxiliary variables. To estimate the variability of this estimator we propose a non-parametric bootstrap estimator. The performance of the proposed estimator is evaluated by means of model- and design-based simulations and by an application to real data. In these comparisons we also include the extension of the Robust EBLUP able to down-weight the outliers in the auxiliary variables. The results show that in the presence of outliers in the auxiliary variables the proposed estimator outperforms its traditional version that takes into account the presence of outliers only in the response variable.

Estimating the at Risk of Poverty Rate Before and After Social Transfers at Provincial Level in Italy

Considering the local areas where citizens live is fundamental to investigate deprivation and social exclusion, particularly in a period of increasing financial difficulties and reduction of public funding. In this work we estimate the at risk of poverty rate of Italian households before and after social transfers at provincial level. To obtain these estimates we use data coming from the EU-SILC 2013 survey and data coming from the population census and administrative archives in a small area estimation framework, since the design of EU-SILC survey does not allow for reliable direct estimation at provincial level. Our results, besides indicating the essential role of social transfers in the reduction of the at risk of poverty rate, allow a sub-national analysis of the phenomenon of interest that would be lost by using traditional statistical techniques.

Estimation of Income Quantiles at the Small Area Level in Tuscany

Available data to measure poverty and living conditions in Italy come mainly from sample surveys, such as the Survey on Income and Living Conditions (EU-SILC). However, these data can be used to produce accurate estimates only at the national or regional level. To obtain estimates referring to smaller unplanned domains small area methodologies can be used. The aim of this paper is to provide a general framework in which the joint use of large sources of data, namely the EU-SILC and the Population Census data, can fulfill poverty and living conditions estimates for Italian Provinces and Municipalities such as the Head Count Ratio and the quantiles of the household equivalised income.