Another bucketload of rain emptied over another part of Great Britain has reignited the debate about how and whether the state should spread the risk of flooding.

On the Today programme this morning, Mary Dhonau, a flood campaigner, went to the heart of the matter:

The statement of principles [the deal between the insurance industry and the government, where the industry provides cover and the government provides money and pays for flood defences] is going to expire. It was only ever a temporary sticking plaster…

Now I hear that the talks have broken down. This has the potential to be huge for many flood victims. 200,000 [households] deemed at significant risk of flooding could find that their flood insurance is removed altogether, and then – if we enter a free market – we could enter a crazy market where the normal man on the street will be unable to afford flood insurance. They could also be penalised with huge excesses.

The average insurance claim is about £30,000, so unless we can knock heads together and get government and the industry talking again, and find a suitable solution, then the solution for all the flood victims that I care so much about looks really bleak.

The problem is that there isn't widespread agreement that it would be a bad thing for people who live on flood-plains to be charged the market cost of insuring their homes.

The justification for the state of affairs at the moment is two-fold. Firstly: people currently live in flood plains, in huge numbers. Rendering their housing situation precarious, as it would be if they were unable to purchase flood insurance, would obviously be a really bad idea. And secondly, the country as a whole is experiencing a squeeze on affordable housing which would only get worse if the large number of houses in flood plains became uninhabitable.

However, there are arguments for the alternative side: in the "crazy market" which Dhonau fears, the cost of those houses would fall to a level which rendered the insurance affordable to people moving in to the area. Your choice would be normal priced house with normal priced insurance, or very cheap house with very expensive insurance. That solves the second of the fears, because we wouldn't be rendering those houses unusable – merely ensuring that the owners take the risk they have chosen to assume on themselves.

The big problem isn't the market in equilibrium, but the market at the point of the changeover. That is: the hitch in the policy is that there are 200,000 homes which were bought with the assumption that they would get subsidised insurance, and now may not. When that sort of thing happens to one person, the public policy response is usually "unlucky" – but when it happens to hundreds of thousands, there needs to be a more nuanced response.

As we wrote in July – when Caroline Spelman was apparently on the cusp of announcing a solution, before she was sacked from her role as environment secretary – the problem becomes even more acute [2] if we factor in the fact that flood-prone areas are likely to grow in number:

The problem is that large swathes of the UK are prone to serious flooding. And as climate change bites, that's only going to get worse. It doesn't necessarily mean your house is definitely going to go underwater – if that were the case, you really should move – but it may be enough to render many places uninsurable.

And what then? It's all very well telling, say, the entire population of London, Kent and Essex east of the Thames Barrier that they are prone to flooding, but that isn't going to lead to them moving. Or, even worse, it might; Britain would be subject to development pressures like never before if that were the case.

With the comparatively small numbers involved, it may be possible to come up with some sort of legacy insurance – where your rate is subsidised provided you moved into the area before a certain date – which would impact the resale value of your home, but not render you without insurance while you still live there. But that solution isn't really scaleable.

Instead, the situation we are in is that we want people to carry on living in flood plains, because moving hundreds of thousands of people is basically impossible, but we also want them to have insurance, because otherwise we are one flood away from an even bigger crisis. And with those two priorities, it seems like spreading the cost out over the entire nation is the only real solution we have.

In the long run, there are things we can do: we can exempt new builds on flood plains from that subsidy; we can require gradually stricter "flood-resilience", as Mary Dhonau's house has, to qualify for the subsidy; we can even phase out the subsidy entirely, ensuring that we don't render any one generation suddenly homeless or uninsured. But in the short run – and the "principles" expire in just six months – there isn't much we can do other than carry on as we are. While that does mean we continue to (slightly counter-productively) subsidise people to live in flood plains, it is the least worst option.