Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/92156 
Year of Publication: 
2004
Series/Report no.: 
ISER Working Paper Series No. 2004-06
Publisher: 
University of Essex, Institute for Social and Economic Research (ISER), Colchester
Abstract: 
This paper investigates the factors affecting the contact and the co-operation of the interviewees in the British Household Panel Survey, in the German Socio Economic Panel Survey and in the European Community Household Panel for the UK and for Germany. The coexistence of two independent panel surveys in the UK and in Germany gives the opportunity to investigate if differentials in the contact and co-operation rates are due to differences in the data collection, personal and household characteristics and/or differences in their impact between countries or between surveys in a same country. If the differentials are explained mainly by differences in the characteristics then it is possible to reduce differentials just by harmonising the data collection. If instead differentials are attributable to heterogeneity in the response behaviour across countries or surveys in a same country, then the harmonisation of the data collection process has a more ambiguous effect. We model the response at individual level as the occurrence of two sequential events: the contact and the co-operation. We explain the contact and the co-operation probabilities in wave t using a set of individual and household characteristics observed in wave t-1, a set of variables characterising the collection process in wave t and t-1, and an interviewer random effect. Moreover, we investigate differences between surveys in the refusal and in the failed contact probabilities by trying to disentangle the part due to differences in the distribution of the explanatory variables and the part due to differences in the model coefficients.
Document Type: 
Working Paper

Files in This Item:
File
Size
144.8 kB





Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.