NBNs, CBAs, and the folly of prediction

The NBN. You’ve probably heard of it. Which is to say, you’ve probably heard how much it costs and how much faster it is than what we have now. You might have even heard that the Coalition have an alternative NBN out now. Again, you’d know basic things like cost and how much faster it is.

But do you know enough to make an informed decision? Do our leaders even know enough?

There’s a great summary of the two policies on The Conversation: the ALP’s fibre to the home (FTTH) and fibre to the node (FTTN). In short, the former connects high speed optical fibre directly to each house, while the latter connects it to a shared “node,” which houses within a kilometre of the node will then access via Telstra’s existing copper network.

FTTH will deliver speeds up to 8x faster than current peaks, while FTTN is a more modest doubling. FTTN will, however, be rolled out five years faster and $17B cheaper. That the Coalition’s alternative will arrive sooner and cheaper is intuitive since the final outcome isn’t as advanced (otherwise it would basically be a slam dunk for them).

Also, importantly, while FTTN may present as a sensible stepping stone toward FTTH, the Australian Competition and Consumer Commission has advised that FTTN is “not so much a stepping stone towards a future FTTH network as a distinct network alternative.” Malcom Turnbull disagrees, and claims that they can build in the option of upgrading when needed. Basically, don’t confuse FTTN as an incremental step toward FTTH. It’s not clear.

As you might expect with something so complex and political, there are problems with the way it’s being discussed in the public arena.

The first problem is to do with the costs – $37B vs $20B – that are being waved around as justification and/or argument against each alternative. That we are only talking about costs is a big problem. That we are only talking about a subset of the costs – the upfront installation costs – makes it an even bigger issue. Each of the NBN alternatives will be in place for decades, which means maintenance, upgrades etc. And these costs are likely to be significantly different for each project alternative. So, if we’re going to consider these projects exclusively on costs, then we should at least be looking at all the costs over a reasonable timeframe.

We also aren’t considering the benefits of each project alternative. The argument that the Coalition’s NBN will take “less time and money” is only relevant if the outcome is the same in both. But it isn’t. The ALP’s will be four times faster. We are not comparing like with like. Worse still, we’re almost totally ignoring the main reason behind this project –i.e. the benefits it yields.

In short, neither has been subjected to a cost-benefit analysis. A CBA wouldn’t definitively end the debate; but it would at least, hopefully, put an end to out-of-context figures being floated in the debate.

A more abstract and interesting problem is around the prediction inherent in each alternative.

The idea of the NBN is to future-proof our broadband needs as the digital economy grows. To that end, the Coalition’s plan, to roughly double current peak broadband speeds, suggests that they take a much more moderate view of future needs.

Either way,  when it comes to prediction, who better to turn to than Nassim Taleb, author of Fooled by Randomness and The Black Swan? When it comes to predicting the future, especially around technology, he points out that there is a significant fundamental flaw in the process:

There is actually a law in statistics called the law of iterated expectations, which I outline here in its strong form: if I expect to expect something at some date in the future, then I already expect that something at present… Prediction requires knowing about technologies that will be discovered in the future. But that very knowledge would almost automatically allow us to start developing those technologies right away. Ergo, we do not know what we will know.

In short, we can only make predictions on what we know. If something is expected, but doesn’t exist yet, then it is known. We don’t know what we don’t know, which is to say that we don’t know what we might soon expect. We don’t even know what proportion of things we know, or how much impact what we don’t know will have.

When discussing the future and future technology, we are, at best, speculating.

Insofar as discussions about each NBN alternative is concerned, talk of “sufficiency,” then, is speculative. It’s an estimate, based on what we think we know about the future.

Some think that either NBN might fall short of the future needs we already know about. And that’s before we add in everything we don’t know.

Even when we discuss the known, prediction is fraught with difficulty. Ken Olsen, head of Digital Equipment Corporations, said: ”There is no reason for any individual to have a computer in his home.” Similar predictions have been made for every technology from the telephone to the television to the iPod to the internet itself.

Fortunately, in each of those instances, incorrect predictions by individuals didn’t overpower growth in demand and changes in behaviour. But given the costs and scale here, that may not be an option.

At last year’s FODI, Tim Harford implored us to make more mistakes. Or, to put it more accurately, to create environments that allow us to make mistakes, learn from them, and improve outcomes. Arguably, the Coalition is trying to do that, but without a proper understanding of the costs of expansion let alone maintenance, it seems difficult to get too excited about such a conservative estimate of growth (especially when it still costs over half of the ALP’s NBN but delivers only a quarter of the speeds).

Future-proofing is hard enough at the best of times. When you do it on such a large scale, and when you really only have one shot, it’s even harder. There’s a lot we need to know about the NBN when making decisions. A lot more than we currently know, anyway.