The aim of this study is to provide a comprehensive description of the statistical methodology used to produce estimates for various labor market variables at both the City and FUA levels, along with an analysis of the results obtained. To achieve this goal, small area estimates were computed using a unit-level multivariate model. This model was specifically designed to enable coherent estimation of the variables of interest collected by the Labour Force Survey, exploiting information derived from administrative data and statistical Registers. The use of such administrative data at the unit-level represents a novel approach to estimation based on Italian Labour Force Survey data. The estimator used in this work is based on a multivariate model implemented through the Mind R package, which was developed by Istat. The method presented in this study represents an extended multivariate version of the conventional linear mixed model at the unit level. To ensure consistency across different domains, a single cross-classification model was employed, encompassing all relevant domains of interest. The outcomes of this analysis reveal significant improvements in efficiency compared to direct estimates. This is particularly noteworthy in the estimation of unemployed individuals (both total and by gender), where direct estimates are prone to relatively high sampling errors.

本研究的目的是全面描述用于对城市和 FUA 层面的各种劳动力市场变量进行估计的统计方法，并对所得结果进行分析。为了实现这一目标，使用单元级多元模型计算小区域估计值。该模型专门设计用于利用从行政数据和统计登记册中获得的信息，对劳动力调查收集的感兴趣变量进行连贯估计。在单位层面使用此类行政数据代表了一种基于意大利劳动力调查数据的新估算方法。这项工作中使用的估计器基于通过 Istat 开发的 Mind R 包实现的多元模型。本研究中提出的方法代表了单元级别传统线性混合模型的扩展多元版本。为了确保不同领域之间的一致性，采用了单一交叉分类模型，涵盖所有相关的感兴趣领域。该分析的结果表明，与直接估计相比，效率显着提高。这在失业人数（总数和性别）的估计中尤其值得注意，直接估计容易出现相对较高的抽样误差。