Testing for Local Management Autonomy: A Question of Perception?

Abstract

Surveys based on standard questionnaires have long been a major research tool of the social scientist. The great majority have focused on one type of respondent — the workers of a given firm, the voters in a particular constituency, or whatever. For one sort of survey, namely, those that seek to know about the individuals in question, this is plainly sensible. But another sort of survey uses a respondent to provide information about the organisation for which he/she can be taken to be an authoritative informant. Questionnaires sent to the head offices of companies or unions and asking about the organisation's policy are a good example. How do we know whether the replies are in some sense representative of the organisation or are just the views of the respondent chosen?