1. Because all the data cited in the paper pre-date the Woolf reforms, the contemporary legal terminology has been retained.

2. H Genn Hard Bargaining: Out of Court Settlement in Personal Injury Actions (Oxford: Oxford University Press, 1987).

3. J Baldwin ‘Review of Genn, Hazel, Hard Bargaining’ (1988) 7 Civil Justice Quarterly 276–77. See also the reviews by Z Bankowski (1989) 47 Juridical Review 122-3, N H Andrews (1988) 47 Cambridge Law Journal 506-7 and B Thompson (1988) 4 Professional Negligence 123.

4. See M Galanter ‘Why the “haves” come out ahead’ (1975) 9 Law and Society Review 95–160 on the structural inequalities of civil litigation.

5. Genn Hard Bargaining at p 168.

6. For example: ‘In the context of the present study it is difficult to assess in any reliable way whether the “reasonable” approach to negotiation displayed by many solicitors is necessarily better or worse, more or less effective in achieving fair and speedy settlements of claims than the aggressive, uncompromising approach. Indeed it is unlikely that any one approach will be “best” for every type of claim.’ Genn Hard Bargaining at p 46.

7. See, eg, the references to Genn's work in standard law texts like P Cane Atiyah's Accidents, Compensation and the Law (London: Butterworths, 1993, 5th edn) p 226 and K M Stanton The Modern Law of Tort (London: Sweet and Maxwell, 1994) p 165. Also the extract reproduced in B A Hepple and M H Matthews Tort: Cases and Materials (London: Butterworths, 1991, 5th edn) pp 913–916 which is edited to emphasise Genn's critique of co-operative behaviour by solicitors.

8. The combination of methods here should not be seen as triangulation, whose logical difficulties have been well-documented by Bloor but rather as an exemplification of the complementary roles that different methods may play in answering real-world questions. Here the econometrics looks for predicted outcomes and the interview data are used to consider processes capable of explaining the failure to locate these. Interview data must, of course, be analysed with particular care because of the way in which interviews are presentations of self through a re-working of events outside or prior to the research encounter. M J Bloor ‘Techniques of validation in qualitative research: a critical commentary’ (1997) pp 37–50 in G Miller and R. Dingwall (eds) Context and Method in Qualitative Research (London: Sage, 1997).

9. Genn Hard Bargaining p 53.

10. P Pleasence, S Maclean and A Morley Case Profiling: The Case For Research (London: Legal Aid Board Research Unit, 1996).

11. The Legal Aid Board has been moving towards a system where public funding to support litigants will be directed mainly through law firms who have met certain quality (franchise) standards. This can, then, be taken to be a sample of better-organised firms. by implication offering an above-average service.

12. Further details of the data and econometric analysis can be found in P Pleasence Report of the Case Profiling Study: Personal Injury Litigation in Practice (London: Legal Aid Board Research Unit, 1998). P Pleasence ‘The Hazards of Litigation: A Study of Legally-Aided Personal Injury Claims’ (1999), Journal of Personal Injury Law 1–8. P Pleasence and R Bowles, Strategies Within Personal Injury Litigation (1997) paper presented at the Legal Aid Board Research Unit conference ‘Civil Justice at the Crossroads’ (London School of Economics, 23 October 1997).

13. Specialism is measured here by reference to the volume of legally-aided personal injury work undertaken.

14. Z=3.49, p0.001

15. Felstiner, Dingwall and Durkin conducted a total of 220 interviews with a range of stakeholders in asbestos disease litigation - solicitors and barristers specialising in both plaintiff and defence work, insurance claims managers, union officials, expert witnesses from medicine and engineering, victim organisations and victims - in all three major jurisdictions of the United Kingdom. These were mainly conducted in the informants’ workplaces, taped and transcribed in full. Because there were only a small number of women, all informants will be characterised as male in order to preserve anonymity.

16. Indeed, the iconicity of Genn's text is underlined by the informant who observed: ‘the longer we are doing [personal injury work], the more we are hard bargaining’.

17. In these extracts, the following conventions are used:… indicates that there was a pause or hesitation in the original speech; […] indicates that a passage has been omitted; Abbreviations such as ‘EL’ (Employer's Liability) are those of the informants.

18. One of the most commonly stated reasons for issuing proceedings in the LAB study's case file data was the slowness of insurance companies to enter into meaningful communication about claims

19. R Dingwall ‘Accounts, interviews and observations’ in G Miller and R Dingwall (eds) Context and Method in Qualitative Research (London: Sage, 1997) pp 51–65.

20. Genn, above n 2, p 80.

21. R Dingwall ‘Atrocity stones and professional relationships’ (1977) 4 Sociology of Work and Occupations 371–396.

22. The analysis of ‘repeat players’ and ‘one-shotters’ is derived from Galanter, above n 4.

23. At one time, Inland Revenue tax records were kept more or less indefinitely and provided an independent record of employment history in industries where few company personnel records had survived a 40–50 year period of bankruptcies, mergers and restructuring. Rather than specifically and expensively proving each period of employment, repeat players accepted these records as an economical solution to the question of allocating time-exposure, and hence liability, between members of the ‘club’ . Shortly before Felstiner and his colleagues began fieldwork, the law was changed so that tax records could be destroyed after seven years. The Revenue set about this task with great enthusiasm, saving itself the storage costs and disrupting the established settlement process, a nice example of unanticipated consequences. Ironically, one of the major record repositories survived for quite a long time because it was so heavily contaminated with asbestos that no one was allowed to enter it!

24. This may reflect the different accounting rules. As civil servants, the MoD claim managers would be concerned to ensure that they could not be accused by auditors of giving taxpayers’ money away too freely. By ensuring that payments were ordered by a court, the claim managers were automatically exonerated from any allegation that they had accepted liability too readily or settled on terms that were too generous. Private sector self-insurers and nationalised industries had more flexibility to respond to wider considerations of the balance of long-run costs to the enterprise and the benefits of enhanced employee morale through perceived employer generosity in times of trouble.

25. Galanter ‘Why the “haves” come out ahead’ op cit.