In Chapter Seven we discussed the role played by government in promoting research use in practice settings. We now consider its own use of research in developing and implementing policy. In doing so, we aim to tease out the lessons from previous chapters about improving research use in government settings. We focus largely on national policy making, although we comment on the relevance of our analysis for regional and local policy settings towards the end of the chapter.

As we noted in Chapter Five, good empirical evidence about what works to improve research use in policy contexts is relatively thin on the ground, so this chapter not only draws on the documented and evaluated studies of research use summarised in Chapter Five, but also seeks to capture experiential knowledge about what has been tried and what seems to be effective. Since the mid 1990s we have been involved in a range of seminars, meetings and advisory groups centred on the core issue of improving research use; these have revealed a wealth of experiential knowledge on which we can draw. We know from these discussions that there has been a good deal of activity aimed at encouraging and increasing research use in policy making and in policy implementation, often driven by the twin aims of improving the availability of research and developing more evidencebased policy processes (see earlier discussion in Chapter One).

Similar to the situation in practice contexts (Chapter Seven), many of the initiatives to increase research use in policy settings draw on combinations of the mechanisms highlighted in Chapter Five (dissemination, interaction, social influence, facilitation and incentives/reinforcement). In order to understand the rationale for these combinations we need to surface the assumptions and frameworks that underpin them. We do this by highlighting first the roles envisaged for research and, within these broad roles, the frameworks that shape research enhancement activities. In discussing these frameworks we aim to capture the extent to which the various descriptive research– policy models outlined in Chapter Four have informed or supported prescriptions for developing research-informed policy.

In terms of the three roles for research outlined in Chapter One – consensus supporting, contention rousing and paradigm challenging (see Box 1.1) – discussion of the research–policy relationship is often underpinned by an implicit assumption that it is the consensus-based use of research that needs to be improved.