Parliament is voting on whether to renew Trident. And they are, of course, doing it the wrong way.
It should be done by a demand-revealing referendum.
Here's how. Trident will cost about £450 per voter - £20bn. So, we should ask each voter: what is the benefit to you of Trident? For concreteness, imagine there are just three voters: Alice, Simon and Julian. Column two in our table shows the (imaginary) answers they give.
Now, for two of them, Trident's cost exceeds its benefit. So a simple majority decision would not renew Trident.
However, the aggregate net benefit is positive, because the net gains to Julian exceed the losses to Simon and Alice. This alone shows that majority rule is inconsistent with potential Pareto efficiency. The solution to this is to levy an extra tax on Julian, equal to the net losses to Alice and Simon. If Julian were to pay £100, he'd still be better off with Trident.
Now, the beauty of this system is that it forces people to debate honestly and rationally.
Take Alice, for example. Let's say she were to exaggerate her net costs of Trident, claiming the gross benefits were zero and net cost therefore £450. We then wouldn't renew Trident - because the £1350 cost exceeds the £1100 benefit. Alice would get her way. But she'd have to pay a tax of £200 - the net benefit the others would get, had she not voted.
So, she has an incentive to reveal the true strength of her feelings. If she overstates her opinion, she risks paying a tax, and if she understates it, she risks not getting her way. The same, of course, is true for everyone.
Now, you can no doubt pick holes, or suggest improvements, in this method. For me, though, it has the great benefits of being democratic (in the sense of involving everyone), of reconciling efficiency with democracy, and of compelling people to think properly about costs and benefits, rather than just being windbags.