This is an election where it’s really tough to work out what’s going to happen.  We are all agreed on that, at least.

But there is still no true consensus as to why this is so.

One reason often put forward is that an abnormally large number of voters still seem curiously undecided.  They are numbered in their millions.  Does that mean that – within the short time remaining – there could be a sudden surge of minds being made up in a way that could dramatically tilt the outcome one way or the other?

If so, this is going to be decided by the undecided. So, who are they, and what sort of indecision are they afflicted by?

I’ve doorstep canvassed for three decades.  I’ve met a lot of ‘don’t knows’ and have learnt that they are far from a homogeneous group.  They include:

We haven’t decided yet” – that’s a polite English euphemism for “well not for you”.

Can’t really say” – which means “none of your business – and most likely not you anyway”.

Not sure this time” – which means “I’m inclined to vote your way, but I still need convincing”.

“Haven’t a clue, mate” – which means “I’ll be abstaining”.

“Do you know, I really don’t know now” – which, finally, means “I actually don’t know”.

So I’m wary of polling on the ‘don’t knows.’  Granted, a pollster isn’t the same as a candidate and the answering may be more truthful when asked by the former.  But even so, we should be cautious before running off with conclusions.

It probably is true that the proportion who genuinely have yet to make up their minds is, at this stage, running higher than at similar stages before previous elections. That’s the polling evidence and it is probably not wrong.  But we should dig a bit deeper into the data.

The first thing to note is that nine polling organisations have now assessed the level of ‘don’t know’ and they are all agreed on something:  amongst those who voted in 2010, it is former Liberal Democrats who are now the most undecided – at an average of 18% – followed by former Labour voters – 16% – and former Conservatives – 10%.

But the variation between the pollsters is huge – with the Liberal Democrat figure, for example, varying between 7% and 35%!

The crucial distinction that has to be made is between those who really still do not know (the minority) and those who have more or less decided but could still change (the majority).  The issue is more about committing to an inclination, rather than wondering what to do at all.

There is separate polling on this.  Pollsters have been asking those who now expressing a party preference if they could still change their mind.  And here’s the interesting finding – almost one-third say this could happen:  32% of UKIP identifiers, 30% of Labour identifiers and 28% of Conservative identifiers could still change their mind.

This matches a recent Survation poll, which found 67% of voters saying they have now decided for sure, as against 31% who have a preference but could still change (and just 2% who really didn’t know at all.)

So these voters are not as much ‘don’t knows’ as they are ‘loosely affiliated’.

What might make the loosely affiliateds change their vote over the remaining 80 days?  It is sometimes suggested that, in this election, where local constituency perceptions will count for so much, votes may be changed in the end by tactical considerations: people voting against parties they don’t want to win rather than for the ones they do. However, polling on this suggests the inclination to ‘go tactical’ is not strong.  Only 5% of UKIP supporters showed any inclination to do this, which isn’t surprising as they are the insurgent force. Only 7% of Conservative voters are so inclined, as opposed to 12% of Labour voters.

So there will be a bit of tactical voting but it won’t be widespread and it won’t be decisive. The chances are, therefore, that these loosely affiliated voters are waiting for sufficiently reinforcing messages from the party to which they are already inclined.

The parties all know which social groups dominate in this yet-to-be-secured category: they are predominantly women – in a recent poll 35% said ‘undecided’ (mostly meaning they could still change) and young voters – where the level was 25%.

So what conclusions can be drawn from this?

There isn’t a vast army of totally ‘don’t knows’ out there, many are loosely affiliated supporters of a party, waiting to be wooed back by reinforcing messaging. However, the most resolute ‘completely don’t knows’ are highly likely not to vote at all.

The parties know this and as a result, their campaign messages are going to be aimed at their ‘soft supporters’ – those inclined their way but waiting for that reinforcement. This group predominantly comprises women and younger voters. The majority of those who have an inclination, as opposed to having made a decision are not currently wavering between the two major parties, so much as between one of them and a smaller party (or parties).

Eventually, you get a feel from the doorsteps, as well as an ability to translate the real meaning of what voters say to you. The feel at the moment is that all politics has become more decoupled from people’s daily reality.  So it’s no surprise that there is less decision and more inclination. Millions of voters are awaiting confirmatory messages, not just about their own preference, but about whether the process really relates to them.

That’s what they don’t know.