Brendan Keenan: 'Broadband expense means it must be part of a wider plan'
It would be hard to find a more succinct definition of politics than that supplied by an anonymous government source on the great broadband conundrum. "It doesn't matter what it costs - the children need to do their homework."
Oh, but it does. I'm sure he, or she, knows that, but the dilemma is real enough. Ministers must grapple with squaring reality and politics, while opposition politicians, including those who have grappled with reality in government, need never care what anything costs.
Unless, that is something has already cost too much and government can be attacked for, ahem, putting politics first. The broadband question is doubly sensitive because the cost overruns at the children's hospital are running simultaneously as a hot political issue.
That started out as €700m and is now approaching €2bn. Unfortunately, one would wonder about that figure too. Experience almost universally suggests that once a project gets seriously out of hand during construction, there is no limit to what it may eventually cost.
If the history of such projects is any guide, the hospital could end up costing €3bn. It is a victim of two kinds of politics; the jockeying for constituency advantage where not even children's health is a deterrent for vote-grubbing, and the necessity to start with a low figure to get as much as possible into the capital programme.
The old solution was simply to add on the additional costs as they emerged on an annual basis. That, after all, is how the national accounts are presented. There would be a brief flurry when it was realised that part of the trumpeted increase in that year's capital budget was just extras costs, but with so much else going on at budget time, it would only be brief.
With any luck, revenues would exceed forecasts and the budget targets would not be affected. That, after all, is what has been happening for most of the past 20 years - and still is.
Last week the Central Bank tiptoed onto the Fiscal Advisory Council's lawn with its cautious analysis as to what should be done about the surplus tax revenues since 2015 - estimated at €7.5bn. In truth, they have already been spent because the budget was supposed to balance around now even if there had been no windfall, and balance is all that has been achieved.
One difference from the past is that, while the capital budget is always there to be pared if things go wrong, now there is no room for a significant overrun. This produced the unprecedented spectacle of other spending being cut to cover the overruns on the hospital.
No supplementary budgets, except for the health service itself. It was a timely reminder that, despite the recent rapid growth, the debt burden is still too large to risk any significant increase in borrowing.
All of which makes the broadband project even more troublesome. Minister for Finance Paschal Donohoe says they will not make the same mistakes as with the hospital.
No-one could call the €3bn price tag unreasonably low, but it might not be high enough.
Supplying broadband for just about everyone presents an entirely different set of problems than building a hospital; even one on a difficult site, incorporating existing hospitals and trying to be the most advanced of its kind. In this case there is not even an agreement on what kind of broadband it should be.
This column is not equipped to deal with that kind of question but it has acquired a certain familiarity with broadband, living in a district which has just been rewired for very fast stuff.
It was an extraordinary business. Every pole now has more wires than it carries for the electricity supply. Most of them also have some kind of elaborate, expensive-looking gizmos attached to them. If that is what is envisaged, one can see that €3bn might not go that far.
The economics are simpler, and not very encouraging. The bold programme of rural electrification is often cited in the broadband argument. But it is easy to forget that urban dwellers subsidise rural ones in their electricity bills, even though almost everyone is on the grid and the initial costs have long been paid.
Rural broadband will no more pay for itself than rural electricity and a demand, in the early years at least, may well be less than that for light and heat.
With broadband, it will not be so easy to hide a subsidy in the bills of other users and no one knows how many will sign up and much a public subsidy would cost. Conveniently, the reservations of the Department of Public Expenditure and Reform have surfaced. They are right to be concerned.
After the hospital farrago, there is a renewed commitment to detailed design, careful appraisal, risk assessment, efficient procurement and so on. But all of that, even if achieved, is not the end of the story.
The much vaunted cost-benefit analysis process is necessary - and appears to have been shamefully ignored too often in the past - but it may not be sufficient. These are particularly difficult matters for small countries, which tend to have their populations concentrated around the major city and sparse elsewhere.
Ireland has more awkward geography than most. Compared with bigger countries there may well be extra costs to be borne per head to create and operate an acceptable modern society. Insofar as possible, they need to be quantified and the hard choices which involve identified. They cannot just be airily waved aside. 'Airy' seems the right word for the Taoiseach's weekend comments that €3bn can surely be found, which may illustrate a more insidious risk.
The country borrowed €100bn to rescue the banks and the public sector, and we seem to have survived. The serpent can easily whisper in the ministerial ear. "What's another €3bn?"
We should know the answer, both from the Bible and experience. Yield to that temptation and one ends up fiscally naked. Despite all the good news, we are already at the fig leaf stage when it comes to debt. Yet Mr Varadkar and that anonymous politician, who must be close to him, have a point - one borne out by rural electrification. The children do need to do their homework and not be disadvantaged as compared with other countries' children in the way they do it.
These wider benefits require political judgement which even the most sophisticated bean-counting cannot supply. It is justifiable to start with a picture of the future and the infrastructure it will require rather than costs and revenues.
One might well criticise the details of the national development plans, but not the concept. What is missing is detail on what they are meant to achieve on a national level and how that justifies the costs. Too often, the local trumps the national.
It is hard to believe, for instance, that the concentration on one-off housing in the last 20 years is anything but a cost to the country, diverting resources from more productive uses and doing nothing to foster a modern sophisticated society.
Not only that, it is now adding to the cost of broadband coverage, with little extra benefit. The kind of politics which produced that has given us "one-off" investment programmes which are often of dubious benefit to the community as a whole.
The road system, comes to mind. They are often the wrong kinds of road in the wrong places. The cheapest are favoured over the most needed and local interests over national requirements.
It turns out that broadband does not really lend itself to that kind of approach. Whatever the system, it has to be rolled out nationally. The more that local interests are appeased, the more it will cost and the greater the inevitable cross-subsidies. No wonder it has taken so long to reach decisions.
Clearly, a modern country needs a modern broadband system. Better to start with what a modern country needs. That immediately brings in the national spatial strategy.
No broadband plan will make sense unless it fits into such a strategy, and we know what has happened to that. Choices had to be made as to who gets what, and where, but they were ducked. Three billion can be found, but it can also be squandered.
It is not the amount that matters so much as how it is spent.