"Please tell me how much confidence you have in each one..." There are problems being confident in any survey that is premised on such a question - in particular, over what domain (or in what respects) is the confidence to be projected? The government may not retain my confidence as much as the military in that I believe they will 'spin' whatever they do, withhold evidence of their errors if they can, may make major errors of judgment on 'policy' - and, relatively speaking, I believe the military may spin less, admit mistakes more readily and, given their decision-making is at a more operational than policy level, make fewer errors of (operational) judgment. But we are not comparing like with like and the differences in confidence here reflect this:- they don't necessarily reflect a view that the military would do better than the government at the government's job so much as less confidence that the government job will be well done in comparison to a military task. My relative lack of confidence is quite consistent with supporting the present government as against the viable alternatives. Nor do the results easily translate into degrees of 'distrust'. I don't positively _trust_ that, say, local bureaucracy will get a key decision right - but that does not mean I positively distrust local bureaucracy: it may be I adopt a 'wait-and-see' attitude and judge things case by case. [The negation of 'I believe x' is not 'I disbelieve x' but 'It is not the case that I believe x']. This kind of 'lack of confidence' might be better described as 'having a healthily sceptical or critical outlook'. For these reasons alone, it is unclear how useful an indicator the poll is. Donal "Scepticism: Always 'extreme', unless 'healthy'." London