Conclusions: developing comparative analysis
IR offers some concrete examples of ways in which an approach akin to CR has evolved. It may be that research is now open to the multi-level approach favoured by CR. Thompson (2003: 372) notes the development of new approaches using ‘multi- level case studies’ to address contemporary developments. Watson (2004: 454) similarly underlines the need to link a micro political study of events within firms to the ‘political economy of employment relationships’. The following remarks suggest how we can develop such a perspective.
There is evidence of a progressive research programme, in the terms used by Lakatos (1978) and as discussed elsewhere in relation to aspects of IR not addressed here (Edwards, 1993). ‘Progressive’ means a developing theme of research that builds on what has gone before, to improve conceptualization of the phenomenon in question and to advance explanation of its causes and consequences. The study of team work would be an example. We can also argue that empirical IR has had, in institutionalist analysis, an approach and set of concepts which can frame empirical inquiry and act as a link to more abstract issues addressed in CR’s contribution to the philosophy of social science.
Yet research programmes have often developed haphazardly. Though it is possible to identify clear progress, there are also limitations. Studies focus on particular issues and may thus not be directly comparable with each others. At one level, it is simply an issue of whether or not results are reported, but there are also deeper issues of the selection of specific issues to study and of research sites. We often lack the information to make systematic statements about certain features (such as types of teams) and their conditions and effects.
It is also the case that IR has rarely made explicit its institutionalist analytical base. As noted above, old institutionalism left little legacy. Kaufman’s (2004b) analysis of the origins and development of the field makes clear that several factors militated against the creation of a distinct paradigm. These included an interest in addressing practical real-world issues, distrust of frameworks from economics or sociology, which were felt to lack empirical realism, and a preference for inductive methods.
IR might then be said to have some unrealized potential (or even, causal power). One way of building on it is through debate between empirical researchers, institutional theorists, and experts in CR. But such engagement also needs to be taken further. It has been stressed above that explanation has been limited by the rarity of comparative studies that allow the researcher to specify the causal powers of a phenomenon in one setting through careful comparison with other settings. This is far from easy. Not only may suitable comparisons not exist, or be hard to access. There are two more fundamental issues. First, it may not be possible to identify likely candidates for study. Causal mechanisms can be deeply embedded in social arrangements, and finding a site for comparative study without already having detailed knowledge may be very hard. Second, causal influences also come in ‘bundles’ that have to be treated as wholes. All that said, however, past experience does show how cumulative information can be built up. It may be possible in the future for researchers to share existing knowledge and then devise a research programme with comparative questions clearly in view.
It has been suggested elsewhere that Qualitative Comparative Analysis (QCA) may have some value here, with some uses in an IR context being cited (Edwards, 2005b).