Looking for a middle class bias: salary and co-operation in social surveys

Abstract

The aim of this paper is to test the existence of middle class bias in survey cooperation. We do this by carrying out a record check study. Our analysis uncovers no evidence of middle class bias. Instead we find a negative gross bias in estimates of the proportion of persons with highest salaries. We also find that high salary earners are more likely to be hard refusers. We argue that this 'elite resistance' is due to specific attitudes rather than more transient features of an interaction. We suggest that these attitudes could be overcome by tailoring of advance communication.