Today's voters have learned to be the look-out for cheap talk from politicians on the campaign trail. But there's a race on and what counts as cheap shifts from announcement to announcement. The media should know better than to reproduce the cheap talk and play into the spinners' game.

What is 'cheap talk'? Pronouncements and postures that clearly fit a purpose but are only pronouncements and positioning. My favourite example comes from biology: male sparrows fluff up their feathers when they meet, apparently to make themselves look bigger and tougher than they really are. It seems to be an instinctual, inherited behaviour. Unfortunately for the instinctual fluffers, sparrows have learnt to tell the difference between fluffers and real brawn. Fluffing in sparrows, it seems, no longer serves a purpose but survuves because it once did in the sparrows' ancestral past. Have humans caught up with the sparrows yet in their ability to see through cheap talk?

This morning's example of Sparrow Fluffing came from David Willetts, the Tory spokesman on families, on the Today program. (He is also, which makes it all the funnier - I have written about this here -  prone to adopt arguments about biological signalling quite literally).

Willetts's cheap talk was the Tories new 'Pro-marriage' policy. Cheap to the politicians, that is. Saying "We're pro marriage!" needs a reply to the question "What does that mean in terms of policy?" So they feel obliged to dream up some micro-tax fluff that seems to cost no one anything and will almost certainly not in fact significantly change marriage behaviour.

The Today programme did roll out Tim Harford to claim that small bribes of this sort really do change behaviour, but the Australian evidence he cited almost certainly applies to the timing of marriages and births than to their happening (ie, people who have already decided to marry or have a child do so sooner, but it makes little difference to the actual decision). He used as back-up evidence the claim that people will hang-on and die later to avoid inheritance tax. But this too suggests it is about when not whether...  Will we have a politician saying: "We are anti-death, and in order to discourage it, we will offer tax breaks to those who die later in the year..."

Willetts recognised the weakness of any empirical claim in the opening of his defense of the Tory plan, saying "We want to send a strong signal that we approve of marriage". This was remarkably honest of him: he does not want to encourage marriage so much as send a signal to that effect. And the economics of signal-sending is clear: make it too cheap to send, and no one believes you; make it too expensive and it might not be worth sending.

£500m levied from banks to pay for £150 per year for the virtuous married should be treated with the same contempt as feather fluffing. Financial services paid about £25bn to the Treasury in taxes of all forms in 2009 on a total income of about £140bn. That's an average, aggregate 18% tax rate. Even if the £500m levy is not offset by a giveaway, it is a tiny tax on the finance sector - it changes the average tax rate from 17.65% to 18.02%.

Cheap policy equals cheap talk; the marriage policy is  a clear case of feather-fluffing and we as voters should not honor it with attention. If the Tories really are deeply pro-marriage, let them come up with a real policy, one that shows actual commitment. How about a hefty divorce tax that also reduces the budget deficit? That would send a signal all right. And if all they wanted to do was remind people that they are nostalgics about the pre-60s golden-age of the family, that those are the sorts of people they are, then they didn't need a tinker with policy at all. We know that already. And, if as seems even more likely, all they wanted to do was send a signal that they are that sort of nostalgic and that they are appealing to an important part of their party base, then we should doubly ignore the national air-time devoted to stroking the party machine.

Another cheap policy announcement came from Labour this morning: a national interest criterion in take-over battles could trigger a requirement for a 66% majority shareholder approval. Nice one, this, because it seems not to cost anyone anything while also preserving the union – 'til death or bankruptcy do them part – of iconic brands with the nation. Cadbury's and England may have divorced recently, but we won't let the next one get away.

True, there's a real problem with takeovers. The overall evidence on big takeovers suggest that most of them destroy value. They are massive transactions designed to take money away from shareholders and transfer it to private equity magnates (see Peter Johnson's analysis of the takeover of EMI by Guy Hands -- "part of the erotica of naked capitalism"), bankers, lawyers, strategy consultants and senior management. Of course the system that separates reward from responsibility badly needs to be changed

But just as the Tories are sending a nostalgist signal about a world of good marriages before the 1960s, so Labour is sending us back to a coddled 1960s corporatist Britain when Whitehall could be counted on to prop up national treasures like British Steel and British Leyland. But we know where that leads: sleepy management that rests on its laurels and squanders the treasure.

The good thing about globalised trade is that it really has energised all those senior managers to be frenetically active and competitive – it's just that their aims have been all wrong. The aim of corporate policy ought not to be to put them back to sleep, but to get them to direct their energy away from the financial "innovation" at the heart of today's mergers and takeovers to something that actually benefits us all. But that is a program that involves real thought about the reform of the City, which Mandelson shows no appetite for tackling. Here, then, is another exercise in feather-fluffing.

The marriage policy announcement is cheap-talk, and the takeover policy announcement looks cheap but hides its costly, backward looking underbelly. The common lesson is that policy designed to send a signal rather than change the way we do things should be very purposefully ignored by voters. Who knows, the politicians, unlike the sparrows, might even be taught to stop puffing themselves up.