Saturday, November 7, 2015
Treasury's effort was little more than a propaganda exercise about the need to restrain government spending, and showed clear signs of government interference. It was widely criticised for purporting to tell us what could happen to the economy over the next 40 years while making no allowance for the effects of climate change and other environmental problems.
By contrast, CSIRO's peer-reviewed modelling exercise attempts to look at what may happen to the economy out to 2050, after accounting for the economic effects of climate change – and our efforts to reduce it – plus other environmental problems such as energy use, water use and use of other natural resources.
This attempt to integrate changes in the natural environment with standard economic modelling is a heroic effort, the first time to my knowledge it's been attempted for our economy.
It's contained in CSIRO's first Australian National Outlook report, but also reported separately in this week's issue of the prestigious scientific journal, Nature.
The project was directed by Dr Steve Hatfield-Dodds, a former Treasury economist now with CSIRO, with participation by another three economists and 13 scientists, mainly from CSIRO.
The question they sought to answer was whether the mounting ecological pressures in Australia can be reversed while our population continues growing and our material living standards continue rising.
To put it another way, can economic growth be "decoupled" from natural resource use and environmental stress?
The modelling takes a fairly conventional "computable general equilibrium" model of our economy, but surrounds it with eight other models of different aspects of the environment – global climate change and economic growth, water use, energy use, transportation, land use, material flows and biodiversity – which have effects on the economy.
But can any person or model accurately predict what will happen in the future? Of course not. So the exercise identifies 18 different plausible "scenarios" of how things may unfold and runs each of them through the nine-model set-up.
Each scenario combines differing global drivers of change with differing domestic drivers. The global drivers cover differing rates of growth in the global population by 2050 – it may grow to 8 billion, 9 billion or 11 billion – and differing rates of greenhouse gas emission.
Limiting global warming to 2 degrees above pre-industrial levels by 2100 would require "very strong" efforts to "abate" (reduce) emissions. Limiting it to 3 degrees would require either a "strong" abatement effort if the global population was allowed to grow to 11 billion, or a "moderate" effort if the population grew only to 9 billion.
That leaves "no abatement action", with the global population growing to 11 billion and global warming reaching 6 degrees. Gasp.
The domestic drivers of change cover differing degrees of improvement in agricultural productivity, differing land-use changes from the development of reforestation markets for sequestration of carbon dioxide or for protection of biodiversity, individuals' take-up of opportunities to use energy and water more efficiently, how much of our improving productivity we take as reduced working hours rather than higher real incomes, and how much of our consumer spending we devote to buying "experiences" rather than goods.
(Turns out those last two drivers made little difference to environmental outcomes, according to the model.)
It's assumed that Australia's abatement effort is at the same rate as the global effort. Up to half our net reduction in emissions is achieved by "carbon sequestration" – withdrawing carbon dioxide from the atmosphere and storing it in plants – achieved by reforestation of cleared land.
So, we build this amazing nine-model model, then run each of the 18 different scenarios through it. What results do we get?
In all scenarios, the economy and living standards are projected to grow strongly. The value of economic activity (gross domestic product) is projected to rise 10-fold over the 80 years to 2050 (the exercise actually starts in 1970, with actual data up to 2012).
This increase in GDP is driven by a 2.9-fold increase in population, leaving a 3.2- to 3.6-fold increase in GDP per person.
On some scenarios, net greenhouse emissions fall to zero or lower by 2040. From four times the global average today, our emissions per person could fall below the global average by 2050.
Apart from reforestation, emission reduction comes from reduced emissions (within Australia, not elsewhere) and from the economy's reduced resource-intensity (that is, fewer natural resources being used to generate each dollar of GDP).
National water extractions are projected to maybe double in 2050, but up to half this increase could be met by desalination in coastal cities and water recycling for industrial use.
Water stress – seen in rain-fed water use in water-limited catchments – improves or is stable in seven of the 18 scenarios.
Pressures on biodiversity (preservation of species) could also be reduced despite economic growth and increased agriculture. But carbon and biodiversity tree-planting could increase the pressure on river-based water systems.
Overall, 13 of the 18 scenarios show improvement in a least one environmental indicator, but only three – each requiring "strong" or "very strong" abatement effort and development of reforestation markets – show improvement in all three environmental indicators.
So the modelling suggests economic growth can continue without worsening – and even while improving – pressures on the natural environment, but only if we and the rest of the world greatly increase our efforts to reduce emissions.
Now, I should warn you that modelling exercises – economic and scientific – are always subject to limitations and open to criticism. They rely on many assumptions and are widely misused by vested interests.
I'm sure in 20 years' time, this CSIRO modelling will look very crude. Right now, however, it's a wonder of the modern world.