As the UK General Election approaches, we can expect a rising tide of promises (and warnings about what the other side will do), all calculated to persuade us to vote a particular way. The depth and rigour of the calculations may vary, from complex statistical analysis to gut feel but, in the end, there has to be some idea of what voters want, what will "buy" their votes.

It's said that "In a pure market economy, everyone is motivated by pure self-interest." But markets aren't pure and I think it's generally accepted these days that we aren't ideal economic units who always make choices solely for our own benefit. It may be the strongest influence, but self-interest is usually modified by personal values.

If you've got some money to invest you'll want to look for the best return, but you might also be interested in the ethics of the companies involved. So you don't necessarily take the best deal. That's a straightforward trade-off between monetary gain and intangible value.

More complicated is what you attend to in everyday situations; which values come to the fore.

Take someone who's out of work and living out of a food bank. Their immediate focus is on just getting by. They are preoccupied with getting enough. In contrast, someone who's just a little better off, say a low-paid, full-time worker, might feel secure enough to be concerned about comparisons with others. Maybe they resent the level of benefits paid to the unemployed - which might be more than their own earnings.

Put them into a top corporate job though and they'd soon be looking at how much their colleagues earn - or at similar jobs in other companies. How you feel about what you've been given depends on the comparison with what "they" have been given!

So the politician's calculation of what to give away is quite complicated. The extent to which anyone will appreciate the gift depends on their personal values and on how well off they perceive themselves to be compared with their neighbours.

The things we value change as our circumstances change. But there are some underlying principles that seem to persist for most of us. For instance, one of Robert Cialdini's six "weapons" of influence is based on our desire for consistency. No matter how uncertain you felt in making a decision (career change, a major purchase etc), once the choice is made you immediately feel much more confident about it, and firmly wedded to it as well. Your further thinking is now likely to support the commitment you've made. You won't easily change your mind and in fact you'll probably make other decisions that are consistent with it.

(So you can be manipulated into buying something expensive by being sold a cheap but similar thing first.)

This has a bearing on voting decisions. It's a serious business and you don't want to be seen to be taking it lightly or to be swayed by trivial inducements. You probably made your choice of party to support long ago and have put much effort into justifying that choice ever since. Now you feel a strong pull to act in line with your past behaviour.

So most of us vote the same way every time and elections are decided by the floating voters: the minority who can be persuaded to change their allegiance in response to manifestos or, quite likely, local issues. I'm sure that the campaign managers have detailed profiles of the floating voters in each constituency and will design their advertising and communications accordingly.

But wouldn't it be really useful to be able to profile ALL voters based on their past behaviours? We already have our supermarket purchases tracked through loyalty cards and pretty much all online activity is recorded by someone somewhere. Of course voting is secret, but I'm sure there are ways of guessing who you support based on other factors.

Who's good at that I wonder? Look out for Google as part of the next coalition!