The impact of innovations
Finally, there is a pervasive question that runs throughout all democratic innovation research: what effects do different innovations have on politics and government that add or detract from their democratic performance? Evidence is hard to come by (Rowe and Frewer 2004; Abelson and Gauvin 2006: 33), partly because there is little impact research (which requires before-and- after comparisons), partly because research trying to sort out cause and effect complexities in the real world is a minefield of problems, and partly because the initial impact of some innovations may soon wear off, while others may take a generation or more to take effect. Some innovations may be of value in themselves (such as improving citizen knowledge and understanding of politics) even if they have no measurable impact on democratic practices and behaviour (such as increasing election turnout).
Innovations may have no measurable impact for one of two main reasons. In the first place, it may be that the innovation is simply ineffective because it is incapable of producing the desired effects. Its effects may be too weak, or the logic or methods underlying it may be faulty. In the second place, it could be that some innovations might produce their desired effects, but for the actions of interested parties who succeed in negating its effects - governments, parties, bureaucrats, the mass media, or opposing interest groups. It is important to distinguish between the two possible reasons, since failure on the first count means returning to the drawing board and re-thinking the reasoning behind the innovation, while the second is a political problem that requires a different kind of response.
Equally, some innovations may have a placebo effect, caused not by the innovation itself but simply because the authorities have taken trouble to try to reach out to citizens and express an interest in their opinions. The warm and fuzzy feelings and the belief that things are getting better, even if they are not, may be what research measures, rather than any specific benefit flowing from the specific innovation. When dealing with innovation effects we must remember the Hawthorne experiments (Landsberger 1958).
In any case, the evaluation of innovations is inherently problematic insofar as there is unlikely to be universal agreement about what makes democracy better and worse (Diamond and Morlino 2004; Buhlmann et al. 2007), and hence the merits of any democratic innovation may be disputed. If ‘democracy’ is an inherently contested concept, then what makes it better or worse is also essentially contested. The best we can do, perhaps, is evaluate innovations against their stated aims, and leave readers to judge for themselves whether the aims are for the better or worse.