‘All Those in Favour’: Computerised Trade Union Membership Lists as Sampling Frames for Postal Surveys

The computerisation of membership lists offers considerable scope for postal surveys of trade unionists. However, faith in the new technology can give rise to the illusion of increased accuracy. Yet, the computerised data bases, while undoubtedly quicker and easier to use, may be no more accurate than their manual predecessors and samples drawn from them can result in misleading response rates. Therefore, with reference to a specific case study of a major trade union, we argue that researchers should seek to establish empirically the probable error in their lists with a view to constructing realistic ‘benchmarks’. Then, these substantial new resources can provide a solid basis for worthwhile research.