Double-calibration estimators accounting for under-coverage and nonresponse in socio-economic surveys

TL;DR

This paper introduces a design-based double-calibration estimation method that simultaneously addresses under-coverage and nonresponse in socio-economic surveys without relying on models, enhancing survey accuracy.

Contribution

It proposes a novel two-step calibration approach that unifies adjustments for under-coverage and nonresponse in a model-free framework, with theoretical and empirical validation.

Findings

Estimator is approximately unbiased under certain conditions.

Method performs well in simulations with artificial populations.

Case study confirms practical applicability to real survey data.

Abstract

Under-coverage and nonresponse problems are jointly present in most socio-economic surveys. The purpose of this paper is to propose a completely design-based estimation strategy that accounts for both problems without resorting to models but simply performing a two-step calibration. The first calibration exploits a set of auxiliary variables only available for the units in the sampled population to account for nonresponse. The second calibration exploits a different set of auxiliary variables available for the whole population, to account for under-coverage. The two calibrations are then unified in a double-calibration estimator. Mean and variance of the estimator are derived up to the first order of approximation. Conditions ensuring approximate unbiasedness are derived and discussed. The strategy is empirically checked by a simulation study performed on a set of artificial populations. A case study is lead on Danish data coming from the European Union Statistics on Income and Living Conditions survey. The strategy proposed is flexible and suitable in most situations in which both under-coverage and nonresponse are present.