Any takers for $50 billion? That’s the next notch in the flood recovery cost lottery some economists and journalists are playing.

Experts who tentatively advanced costs in the single-digit billions a fortnight ago have been left red-faced. Early suggestions of $10 billion have long since been left behind. At the taxpayer-subsidised Australian, the January 14 editorial offered between $10 billion-$20 billion; its anonymous author must have felt this was a suitably scary number to fit the “NBN = evil” narrative. But even that has been left looking decidedly wimpy. Shane Oliver of AMP trumped everyone by wheeling out “above $30 billion” — albeit that was the total cost including lost output, a distinction quickly lost in media coverage.

Only Saul Eslake has provided a solid dose of common sense in all this.

By the time Friday arrives and Wayne Swan rises at the CEO Institute annual conference in Brisbane to offer some “preliminary estimates” of the cost, who knows what number might be being bandied about.

One assumes that economists will insist on a cost-benefit analysis for the billions — at least — expected to be required to repair rail and road infrastructure and amenities damaged by flooding. Whether all infrastructure damaged in the flooding should be repaired or replaced, and repaired or replaced to the same, lower or better standard, and what criteria should inform the priority accorded reconstruction (for example, should railway lines servicing Queensland lucrative, and carbon-intensive, coal mining industry take precedence over, say, bridges important to the transport needs of small communities) are all issues that, in the absence of an unlimited budget and an unlimited workforce, will need to be resolved between  the federal, state and local governments. In fact, it’s odd that we’re yet to hear the now fashionable demand for CBAs for taxpayer-funded infrastructure spending.

Then again, this wasn’t the approach taken after the Victorian bushfires when, in the face of appalling devastation, our political leaders simply committed to full rebuilding of the affected communities, regardless of the cost. Perhaps the response to the Queensland floods, too, will be a commitment to restore communities, and the amenities and infrastructure they need, to how they were before the floods, or better, regardless of the cost.

That may not be the economically rational thing to do but it may well be the socially (and politically) rational thing to do, to signal support for the communities affected by these disasters, which are usually regional and rural, from the majority of taxpayers, who have (normally) escaped devastation because they live in cities.

I mention the issue not to be snarky — well, only a little — nor to suggest reconstruction be delayed while economists conduct thousands of rigorous CBAs for every stretch or road, bridge, rail line, school, hospital, electricity sub-station damaged by floods. It’s to note how arbitrary and selective governments can be about infrastructure spending because, in the long run, the priorities around infrastructure provision are the same whether it is in response to a flood or it’s a greenfields project.

The media, politicians and in fact all of us are poor at adopting a rational and consistent approach to infrastructure spending — witness the persistent hostility to road user charging, for example. Handing responsibility for spending, as well as evaluation, over to an Infrastructure Australia-style body of experts would deliver more rational spending choices. But no one would be pleased — not politicians, and certainly not voters.