In previous posts (here and here) I've explained how the Commerce Commission sets regulated telecommunications prices. It's not the sexiest of subjects, but I've done it because you need to get a feel for the status quo if you want to be able to take a view on the government's recent proposals on how existing (copper based) broadband products are priced and what the new Ultra Fast (fibre) Broadband network will cost you.
Here's the benchmarked international data that formed the basis of the decision (it's on p30).
The median of these overseas costs, as a starting point for setting the price, was 0.66 cents a minute, but in the event the decision was made to take the 75th percentile of this range, and to set the price at 1.13 cents a minute. The rationale (as discussed on pp36-39) was that any adverse effect of setting the price too low (by deterring people from building infrastructure of their own, since they could free-ride on the too-cheap equipment of others) would be higher than the adverse effect of setting the price too high, so the Commission opted for the relative safety of a somewhat higher price.
It also meant that the price come out around the same as the prices in the UK and Australia, which for a variety of reasons might have been expected to have been pretty good sighting shots for what the equivalent costs were likely to be in New Zealand.
The result, in short, looked rather reasonable, even if the outcome was a lot lower than the 2.65 cents a minute Telecom was arguing for back then.
You might feel that this benchmarking approach is too rough and ready a route to follow, and others certainly did at the time. There were submissions that the raw overseas costs ought to have been adjusted for the fact that things like labour costs and the cost of capital varied a lot from one country to the next. But when the Commission looked at making those sorts of adjustments, they didn't make much net difference. In aggregate, they tended to cancel out (as noted in p33 of the decision).
The Commission investigated one adjustment in particular that looked like it might be material, namely the impact of network density. In fairly densely populated countries, the argument went, you can run out compact, shorter networks more cheaply than you can in lighter density countries (like New Zealand), where you might have to run out your network, at high expense, for miles and miles into the wop-wops. But when you went through the data, you didn't in fact find that there was any systematic relationship between cost and density (pp33-34). Again, you end up back at square one, the original unadjusted benchmark costs.
So there's the process - you get the overseas prices (when they're cost based), you pick a point within the range of prices that you think is fair for New Zealand conditions, and you're done. It may have its rough edges, but it produces numbers that look broadly reasonable. And (a) because they're reasonable and (b) any other way is highly expensive (and arguably no more accurate), these prices are the ones that have been used to set regulated prices.
Next post - the government's proposed changes.