ABSTRACT

In Italy, the provision of educational ancillary services (like meals and school transportation) is in charge of the municipalities. We investigate whether municipalities differ in their efficiency when providing these services and whether such heterogeneity explains some portion of the variability observed in pupils’ test scores. This paper is the first application of a non-parametric order-$\mathrm{m}$ model and a two-stage multilevel regression model to a unique administrative dataset, made up of the entire population of Italian pupils tested in reading and mathematics at grade 5 (academic years 2012/2013 and 2014/2015). Results demonstrate that local governments have different efficiency levels in providing services to schools. The test scores’ variability among pupils is not explained by different efficiency levels of local government in producing ancillary services.

摘要

在意大利，提供教育辅助服务（如膳食和学校交通）由市政当局负责。我们调查了市政当局在提供这些服务时的效率是否存在差异，以及这种异质性是否可以解释在学生考试成绩中观察到的部分变异性。这篇论文是非参数阶次的第一个应用——$\mathrm{米}$模型和一个独特的行政数据集的两阶段多级回归模型，由在 5 年级（2012/2013 和 2014/2015 学年）进行阅读和数学测试的意大利学生的全部人口组成。结果表明，地方政府在为学校提供服务方面有不同的效率水平。地方政府在提供辅助服务方面的效率水平不同无法解释学生之间考试成绩的差异。