Like many others, I really don't understand how a 13% lead for the 'yes' camp in any referendum on enhancing the powers of the National Assembly has been covered as though it were a major problem. It's a very healthy lead – the sort that any government would be very happy with if it were about to call an election.
I do understand, though, why some are still so cautious. For those who got their fingers burnt in 1979 - and very nearly so again in 1997 - there is a real fear of failure. The danger, though, is that seeing only the echoes of the past, seeing only the similarities with what happened before, may blind us to the very real differences. Historical experience is a useful input to the present, but fear of a repeat of the past should not be allowed to determine our responses to such an extent that we wait for unanimity.
There are two key differences between the next referendum, whenever it happens, and the two previous referenda, and those differences need to be considered as carefully as the similarities. And then we need to make a measured judgement about the risks.
The first is that the Assembly now exists, and people have had ten years' experience of the new situation. The sky didn't fall in, and in a growing number of areas, the Assembly has started to make a real difference to people's lives. In some areas, there have been disappointments, of course, and it would be dishonest to deny that. But its influence on our daily lives is growing, and all the polls indicate that people understand that - and broadly welcome it.
The second key difference is that this is not an 'all or nothing' referendum, as were the two in 1979 and 1997. Even if the referendum were to be lost (and I don't believe that it would be), the Assembly would still be there, it would still be gaining powers and influence, just at a much slower pace. In short, there is less to lose.
That is not to say that a lost referendum wouldn't be a setback for Wales; it would. But the scale of the setback would be a great deal less that than which occurred in 1979 – and which so nearly happened again in 1997.
Some have argued that the opinion poll lead must be at least x% - put whatever number you like instead of the x – but I simply don't accept the validity of such a mechanistic approach. Setting an arbitrary threshold is no substitute for making a proper judgement of the situation at the time, and coming to a decision based on that judgement.
There is a danger that we see only the views of those with whom we concur, and rush into a judgement based on that. I attended the meeting of the Convention in Carmarthen a couple of weeks ago, and if the feeling of those who spoke at that meeting is typical of the views of the population as a whole, then the Carmarthen jury declared pretty clearly that the biggest problem with the proposed referendum is that it does not go far enough. I know, though, that other conclusions would probably be drawn from meetings elsewhere in Wales.
There are those who are starting to doubt the commitment of the One Wales government to holding the promised referendum. I do not share those doubts. For me, there were always two windows in which the One Wales government could have called a referendum. The first was immediately after taking office, in 2007; but the agreement of the partners to set up the Convention effectively ruled that one out. The second is after the Westminster election but before the 2011 Assembly elections.
That second window has been narrowed by the fact that Gordon Brown is now likely to hang on until the bitter end before calling the election, but it has not been closed by a long shot. Any referendum campaign is likely to be as painful for the Labour Party as it will be for the Conservatives, since both parties are hopelessly divided on the issue. Expecting Labour to expose and emphasise those internal disagreements in the period immediately before an election is just not realistic. And, whether we like it or not, the commitment of Labour to any campaign is a key element in ensuring success.