One of the more curious recent opinion poll findings has been that people are unwilling to pay to control immigration. A Yougov poll found that 62% would pay nothing to reduce net migration from the EU, and that only 15% would pay 5% of their income to reduce it to zero. Support for immigration controls, it seems, falls away when people are asked to pay for them.

But of course, such controls do have a cost. Although the near-term impact of immigration (pdf) on overall wages (pdf) is roughly zero, it's quite plausible that tough migration controls would reduce innovation and productivity growth in the longer-run, leaving us worse off. And even if you disregard this, there are shorter-term costs: the deadweight cost of paying for border guards; the loss of export earnings as universities take in fewer overseas students (which should mean a weaker pound and hence higher import prices); and the loss of net tax revenue because migrants make a positive contribution to the public finances.

All this poses the question: given that immigration controls have obvious costs which most people are unwilling to pay, why is support for such controls so strong when people are not confronted with their costs?

One possibility is that voters assume that others will bear the cost. This might be reasonable for the low-paid, non-tax-payers and for older people. But it's plain wishful thinking for others.

Another possibility is simply that people are lousy at making connections in economics, and so don't link controlling immigration with costs of doing so.

However, I want to suggest another possibility - that the supposedly impartial media is deeply misleading here.

Discussion programmes - not just about immigration but anything else - tend to follow a strict format based upon the adversarial prosecution-defence model; there's an advocate of a position and an opponent - often with both being overconfident blowhards.

What this misses is that policies have costs. The question is not simply: is this policy right or wrong? But rather: is the cost of this policy worth paying?

So for example, it would be reasonable to support immigration controls because their economic cost is worth paying to shore up social capital. It would also be reasonable to oppose them because you place more weight upon their costs - in money and freedom - than to the risks to social capital and of discontent as the popular will is disregarded; this is my position.

Reasonable people should take either position on immigration with a heavy heart.

But this sort of position is often excluded by the adversarial model of discourse which invites only pro- and anti- stances. Such a model encourages fanaticism at the expense of trying to assess costs and benefits. It is yet another way in which the "impartial" media actually serves to coarsen discourse.

But there is an alternative. Rather than have a prosecutor-defender model, discussion programmes could have an examining magistrate model in which examining the evidence is more important than having a ding-dong. This might not make for good TV or radio. But the BBC must ask: is it in the business of entertaining fools, or in the business of improving political discourse?