By Liana Downey (CEO of Blueprint Institute)
Into every modellers life will come that sinking feeling — the moment when a politician, board member or client has a clever idea, has made an announcement to the press, and is hoping you can run the numbers to back it in. Sometimes you get lucky, and the numbers work. Other times that’s the cue for a difficult conversation to let them know things don’t stack up. At which point they may ask, isn’t there a way to make it work? A bit of massaging?
In the best case, you're lucky to have, or to be, a boss who is willing to hold the line, who considers rigour fundamental, and won’t do analytic backflips to try to make something fit. Integrity matters, especially in policy, where massaging numbers can lead to bad decisions with decades of impact for millions.
I sympathise with the analysts at Frontier Economics— this was always going to be tough, with announcements made well before any supporting analysis. Credit where it's due, they’ve been clear about much of what they have and haven't included (including omitting some critical transmission costs associated with adding nuclear to the grid). They've used Plexus, which is the right software to replicate AEMO modelling. They've been honest about comparing different scenarios and explicit about their assumptions, although not their sources.
But beyond those headline good practices, there are analytic flips which render the analysis essentially worthless. We should be having honest and clear discussions based on realistic assumptions about a range of technologies including nuclear (something that Blueprint Institute has looked at before). But this isn’t it.
It doesn't actually model the Coalition’s Nuclear plan. The analysis includes 13 GW of nuclear capacity, way more than the 6.6 GW the coalition had committed to.
It doesn't compare like-for-like – the equivalent of comparing costs for building a 2-bedroom apartment with a 5 bedroom house.
It uses indefensible assumptions on the costs of nuclear, including cost declines for nuclear with no historical precedent. Their capital costs are about 20% lower than the current average US capital costs for nuclear, with no clear explanation for why, especially problematic given labour is a key cost. Indeed, they seem to be cherry picking from jurisdictions like the United Arab Emirates, where an average construction worker in earns about 9-13 times less than in Australia.
It uses magical thinking on the timing of nuclear. When it comes to timing, they seem to have picked the one best example from South Korea, who not only manufacture their own facilities, but are well into their own learning curve. They also seem to be excluding transmission connection times, which have consistently blown out domestic renewables production start dates.
Those assumptions are particularly problematic, because they don't test sensitivities — so we don’t know how the analysis changes if they use numbers that more realistically reflect the global experience of developed economies on timing or cost.
Low voltage transmission costs are excluded. Many commentators, including Blueprint Institute, have criticised AEMO’s modelling for lack of adequate transparency on transmission costs for renewables. This repeats the problem, but this time for nuclear. There are real costs in adding nuclear to the grid as well — especially low voltage transmission costs.
Crucially, their model doesn't achieve target emissions reductions. Year-on-year emissions matter, not just what happens in 2050. But if you care about economics, you have to care about emissions reduction, given the costs of adverse weather events are $38Bn p.a. in Australia.
I get it — it's good to be optimistic, but it’s also good to have your wits about you. It's a bit like when the guy tells you, oh for sure, he can definitely have the bathroom done by Christmas, you'd be wise just to do the rounds and see how well that's worked out for the neighbours.
Perhaps most importantly, there is no evidence at all that this policy will reduce energy prices. Rather the report emphases that any attempts to tie increased prices with renewables (or other changes to the grid) represents a misunderstanding about how prices work in the most of Australia. Prices are rising because of oligopolistic behaviour of a few players, exacerbated by a NEM pricing mechanism that is no longer fit-for-purpose. As the report says, "it is inevitable that the current NEM pricing mechanism will be reformed in the near future … it makes little sense attempting to forecast prices for the next 25 years at this stage.”
Exactly. This proposal makes no claim that it will lower prices. Just as well, because the flexibility required to come up with the headline numbers would make any circus performer envious.
Liana Downey is the CEO of Blueprint Institute, an independent think tank, and leader of Australia’s first Greenhouse Gas Abatement Cost Curve with McKinsey & Company. Contact Liana at: [email protected]
Icons created by Eucalyp - Flaticon.