Deliverable 4.3 - Report on survey discontinuities

van den Brakel, Jan, Krieg, Sabine, Smith, Paul A., Tzavidis, Nikolaos, Ertz, Florian and Munnich, Ralf (2019) Deliverable 4.3 - Report on survey discontinuities Makswell 37pp.

Abstract

The MAKSWELL project was set up to help strengthen the use of evidence and information on well-being and sustainability for policy-making in the EU, as the political attention to well-being and sustainability indicators has also been increasing in recent years. Traditionally sample surveys provide the primary source of data that are used for measurement frameworks for well-being and sustainability. However, survey organisations and national statistical institutes frequently review and occasionally change their approaches to collecting survey data. Although such changes are motivated by the need for a more efficient approach to allocating resources, they can lead to breaks in the series of published estimates known as discontinuities. This report presents approaches to estimating and adjusting for discontinuities in survey data. Alternative methodological approaches are presented, some of which assume the availability of experimental or pilot data under the new design, whilst other approaches make use of time series methods for quantifying discontinuities and hence not requiring the use of pilot data. The methods we present are illustrated using real data from the UK and the Netherlands. The targets of estimation are defined both at national (aggregate) level and subnational (domain) levels. The report presents estimation and inference both under the design-based and model-based frameworks. The methodological tools are general and therefore transferable to other survey settings.

This record has no associated files available for download.

Contributors

Author: Paul A. Smith

Author: Nikolaos Tzavidis

Download statistics