How to make sense of big, scary climate costs

Climate costs can seem scary, but it’s all in how you look at them. Bills image from

The Daily Telegraph yesterday published a set of Big Scary Numbers around ambitious action on climate change, leading to lower economic growth, lower wage growth and the winding down of the coal industry.

In response the federal environment minister Greg Hunt stated that “What Labor refuses to learn is that you can tackle climate change without a carbon tax. You can tackle climate change without a [A]$633 billion hit to the economy.”

Where do such numbers — and associated apocalyptic scenarios — come from, what do they mean, and is ambitious action a recipe for a significant hit to the economy and our living standards?

The source of the numbers

The numbers come from work done during the term of the previous government, which can be found at the Treasury website as well as the website of the Climate Change Authority.

The Telegraph highlights the most ambitious of the possible emissions reductions targets (40-60%), and then presents these as serious possibilities arising from climate policies adopted at the recent ALP national conference.

The paper states that these confronting results have been “buried in more than 300 spreadsheets” on the authority website, as if it was hoped nobody would ever notice. While the tactic of “hiding something in plain sight” can sometimes be an effective way of concealing something, the placing of detailed analysis on publicly accessible government websites seems less like hiding something, and more like transparency as it should be practised.

Shadow environment minister Mark Butler has responded, calling the modelling “ridiculously outdated” and noting that the ALP did not adopt emissions reductions targets at its national conference (although significant targets for renewable energy were a signature part of the conference agenda).

Numbers: big, small and scary

Two extremely simple rules of thumb apply to discussions of impacts on the entire economy, especially over multiple years or decades.

The first is to note whether the numbers are being presented as amounts or as percentages (or proportions).

The second is to note whether the numbers are in real (inflation-adjusted) or nominal (face value) terms. In both cases, each approach is neither right nor wrong, but if you want to make the numbers look Big and Scary, presenting nominal amounts will do the job most effectively. Many of the numbers presented by the Tele— such as the A$633 billion quoted by the Minister — are exactly of this form.

The Australian economy is large, measured at approximately A$1,500 billion annually. The A$633 billion impact quoted by Mr Hunt certainly looks like a substantial chunk taken out of that: between a third and a half of economic activity sacrificed.

But as the Tele clearly states, the A$633 b impact is cumulative, not annual. We can do a very loose exercise to make a point.

If we assume the economy stays that size from now until 2030, and the A$633 b impact is deducted in equal increments each year, it adds up to a loss of around 3% of total economic activity each year.

A more serious exercise might go as follows. Take an economic growth trajectory out to 2030 with and without a particular emissions reduction target, and calculate the actual accumulated differences in economic activity.

I did this using the Treasury’s 1.2% growth rate forecast for GDP with a particular target, compared to 1.4% growth as the reference case. Of course, the discrepancy between reference GDP and target-restricted GDP gets larger in each successive year. Using this method, I calculated that the reduction in cumulative GDP, expressed in present value terms, was less than 2%. (I used a 3% discount rate so as to not to favour the present too much relative to the future.)

Looking at proportions like this takes large nominal dollar figures and puts them in, arguably, a more realistic, less alarmist context.

Models, assumptions and projections

Economic models are often used in policy evaluation when the policy in question is expected to have large and durable impacts. There is little option if we seek to quantify the impacts of a major policy, other than firing up a computer model and running it through its paces.

Even in the best such exercises, results are contingent upon (i) model design, (ii) the assumptions fed in, and (iii) the scenarios analysed.

The suite of models used in this analysis are perfectly credible. This does not mean they are perfect, but they are professionally legitimate.

The assumptions used, say, for electricity costs might have been valid a few years ago. As Mr Butler points out (as do others), renewable energy costs have been falling fast and the modelling results may well be outdated. And of course, results are contingent upon the scenario assumed, such as high emissions reductions, which is in effect the Tele’s key point.

Model projections are not forecasts. Over long time frames, Treasury notes “assumptions necessarily become more speculative.”

The upshot is, whether deliberately or accidentally, the Telegraph presented the numbers in as dramatic and confronting a way as possible. But turning them into a percentage makes them much less alarming.

This alone does not warrant a particular course of climate action, but it does suggest such action isn’t going to economically cripple Australia. As Treasury notes, other global factors can have bigger effects on the path of economic activity.

The ALP may have been over-emphatic in distancing itself from these numbers; but it would be a reasonable presumption that the modelling should be re-done to assess impacts of changed policy settings (revised scenarios) using new data (revised assumptions). And the results should be presented in a calm, proportionate manner.