Last year, Sir Nicholas Stern provided a review for the UK Treasury on the economics of climate change. In his commentary, Stern suggested that the costs of stabilising carbon dioxide emissions within the preferred range were about 1% of GDP. This is being taken by many commentators as being an appropriate number to apply to the Australian economy.

There are two reasons why this may not be so. The first is that the Australian economy is a high carbon use economy. We would therefore expect that the economic cost would be larger than the average 1%. The second is that when we look more closely into Stern’s own review, we see that the 1% economic cost is only the mid point of a very broad range.

On page xv of the Executive Summary of his Review, Stern shows as Figure 4 a scatter plot of potential costs. (This is also shown as Figure 10.1 of the full report on page 242). In his words “this shows that the full range of estimates drawn from a variety of stabilisation paths and years extends from a -4% of GDP to +15% of GDP” although he says that most estimates are “clustered in the range of -2% of GDP to 5% of GDP, with an average around 1% of GDP”.

This means that the 1% estimate can have the statistical accuracy of me throwing darts in an English pub on Friday night. I would need to go back the next morning to see where they landed on average. The actual number for the Australian economy of the costs of reducing carbon use could be notionally anywhere from -4% of GDP to +15% of GDP. Well, how can we find out what it might be?

Stern provides an estimate of what he believes is the long term social cost of carbon. He provides that estimate on page xvi of the Executive Summary as $US85 per tonne of CO2. He says this is “higher than typical numbers in the literature, largely because we treat risk explicitly and incorporate recent evidence on the risks”.

This means that we can assess the impact of reducing carbon use to the extent desired by Stern by estimating the impact of a $US85 per tonne of CO2 impost upon the use of carbon in the Australian economy. What would be the impact in terms of GDP and what would be the impact in terms of jobs? What we are talking about here is what would be the impact on the Australian economy if Australia entered in an international carbon trading system where permits for the use of all carbon in the Australian economy were priced at the price that Stern has nominated of $US85 a tonne of CO2.

Let us look at the use of carbon in the Australian economy(1). One of the first things we learn is that Australia is likely to be a very high carbon use economy.

In Table 1 below we compare the coal consumption in the Australian economy to that that of the United Kingdom. The UK was of course the original country of the industrial revolution. This revolution was based on the burning of coal. These days though UK coal consumption is dwarfed by that in Australia. Annual UK coal consumption is now some 58.5 million tonnes per annum. Australian coal consumption is almost 2.3 times as large at 132.2 million tonnes per annum.

On a per capita basis, the comparison is even more dramatic. Australia uses 6.5 tonnes of coal for every man, woman and child in Australia every year. The UK uses only 1.2 tonnes. These days the UK gets its energy from North Sea Gas and also from nuclear energy. Whereas Australia gets no electric power from nuclear energy, the UK generates 23.3% of its electric power from this source. Clearly, Australia will be a much higher carbon use economy than the UK and the Stern carbon price of $US85 per tonne will have a much higher impact on the Australian economy.

CountryCoal Consumption
Millions of Tonnes
Coal Consumption
Tonnes Per Person
United Kingdom58.4901.19

Table 1 – Coal Consumption, Australian and the United Kingdom. Source: ABN AMRO Morgans

In Table 2 we see Australian carbon consumption on an annual basis. Each year Australia uses 18.46 million tonnes of natural gas. Natural gas is around 78% carbon so this equates to 14.43 million tonnes of carbon burned every year in the form of gas.

In addition, Australia uses 33.42 million tonnes of oil. Oil is around 86% carbon so this equates to around 28.9 million tonnes of carbon burned every year in the form of oil and petroleum products. In addition, each year Australia uses 132.2 million tonnes of coal. Coal is around 73% carbon so this equates to 97.2 million tonnes of carbon that Australia burns every year in the form of coal. The total of all of these forms of carbon used in Australia on an annual basis is 140.47 million tonnes per year. That means that each person in Australia uses 7.0 tonnes of carbon per year.

Fuel type    Tonnes of Consumption (millions)Proportion of CarbonTonnes of Carbon (millions)
Natural Gas18.460.7814.43
Total Carbon140.47

Table 2 – Australian Annual Carbon Consumption. Source: ABN AMRO Morgans

How do we equate Stern’s price of carbon dioxide emissions of $US85 per tonne of CO2 to an Australian dollar cost per tonne of carbon? Well, $US85 per tonne of carbon dioxide equates to $US311.67 per tonne of carbon. We calculate this price by multiplying the price of carbon dioxide by the atomic weight of carbon dioxide divided by the atomic weight of carbon. $US311.67 per tonne at an exchange rate on 24 April of US82.53 cents per $A equates to an Australian price of $A377.64 per tonne of carbon.

If Australia uses 140.47 million tonnes of carbon and has to pay a cost on each tonne of $A377.64, then this equates to a total cost of $A53.05 billion. In 2006, Australian GDP was $A1,001.9 billion. This means that the total cost is not 1% of GDP but 5.3% of GDP. This is as we expect, higher than Stern’s estimate of 1% of GDP. We would expect this because Australia is a high carbon use economy. Our estimate is however, within the range specified by Stern of -4% of GDP to +15% of GDP. It is at the top end of the smaller cluster specified by Stern as from -2% of GDP to +5% of GDP.

An economy is made up of people. If we say that the Australian economy will fall by 5.3% of GDP because of the increased cost of carbon, what this means quite bluntly is that 5.3% of people will lose their jobs. When the growth rate of an economy slows, then employment growth also slows. With a large fall in GDP of around 5.3%, unemployment could be expected to rise by the same amount. We last experienced this in Australian the recession of the early 1990’s.

In Table 3 we show the total employment effects of a carbon price of $US85 per tonne of CO2. In February 2007, the Australian economy on a trend basis employed 10.35 million people. Of these people, 7.39 million have fulltime jobs and 2.95 million have part-time jobs. If 5.3% of these people lost their jobs, then the total job loss would be 548,000 people. Of these, 392,000 would be fulltime jobs and 156,000 would be part-time jobs. This seems to be a high price for trading carbon. The question is, would the sacrifice of jobs be worthwhile?


Total Employed (thousands)Full-time Employed (thousands)Part-time Employed (thousands)
February 200710,350.47,398.92,951.5
Job loss of 5.3% would be548.02391.75156.27

Table 3 – Total Employment Effects at $US85 per tonne of CO2. Source: ABN AMRO Morgans

The whole purpose of increasing carbon costs as outlined by Stern is to increase the price of high carbon use relative to low carbon forms for producing energy such as nuclear, hydro-electric and wind power. This makes a lot of sense in a country like the United Kingdom where the subsidy effect of high carbon prices will generate a flow of funds to nuclear power generation and wind farms.

As Stern says, after carbon pricing and technology “the third element is the removal of barriers to behavioural change”. At the moment, however, every State Premier in Australia has declared that he would not like to have a nuclear power plant built in his back yard. Hence, the job sacrifice caused by carbon trading in the Australian economy would reap little economic benefit. It is likely that most of the Australian subsidy from high carbon prices would be paid away within the Kyoto Protocol network to French, or British, or German nuclear power generators.

It is unlikely that an Australian government of any persuasion would enter into international carbon trading of Stern’s price of $US85 per tonne. Stern realises this and suggests that “the social cost of carbon would start in the region of $25-$30 per tonne of CO2”. He argues however that this low short-term cost would have to rise. He says “the social cost of carbon is likely to increase steadily over time because marginal damages increase with the stock of greenhouse gases in the atmosphere”.


Total Employed (thousands)Full-time Employed (thousands)Part-time Employed (thousands)
February 200710,350.47,398.92,951.5
Job loss of 1.87% would be193.42138.2655.15

Table 4 – Total Employment Effects at $US30 per tonne of CO2. Source: ABN AMRO Morgans

In Table 4 we see the employment effects of reducing the carbon price to $US30 per tonne of CO2. We estimate that this price would mean a short term cost of 1.87% of GDP. Again, this doesn’t make a lot of sense if we can’t build nuclear power plants in Australia but let’s just say we are prepared to pay this 1.87% of GDP for reason of political correctness.

Table 4 shows us that 1.87% of GDP loss would result in a loss of 193,000 in jobs. Of these 138,000 would be full-time jobs and 55,000 would be part-time jobs.

Other Australian Studies

The impact of carbon pricing upon the Australian economy has also been explored by the Australian Bureau of Agricultural and Resource Economics (ABARE) in its July 2006 report “Economic Impact of Climate Change Policy”. Their document does not refer directly to the Stern Report because Stern published his document 3 months later in October 2006.

The ABARE report differs from ours in that it assumes real wages will fall instead of employment. ABARE calculates falls in GDP and real wages associated with reductions in CO2 and the greenhouse gas emissions. Many of these falls are greatly in excess of our own calculations. A summary of their results based on Table C, page 3 of their report is shown in our Table 5 below.

Decline in CO2 %Decline in GHG %Fall in GDPFall in Real Wages

Table 5 – Australian economy alternative scenarios 2050. Source: ABARE

We can see from Table 5 that reductions in greenhouse gas in excess of 60% by 2050 as commonly talked about in the public domain imply a loss of GDP of 8.3% to 10.7% and a fall in real wages of 17.4% to 20.8%. Under no scenario is the reduction in GDP limited to only 1%. Reducing carbon emissions is at all times purchased at significant economic cost. The ABARE study suggests that these economic costs persist even when “clean coal”, nuclear energy and non-hydro renewables are included in the scenarios.

Job Numbers Fall or Real Wages Fall

I have assumed in my scenario that real wages are fixed. This is because no political party advocating that we should sign the Kyoto Protocol currently advocates a cut in real wages to maintain employment as GDP falls from carbon trading. Because real wages are fixed, a fall in GDP must result in a loss of jobs.

There is a trade-off here. Employment could be maintained if real wages fell by the percent of GDP fall divided by the wages share of the economy. The December quarter 2006 National Accounts show the wages share (compensation of employees) at 53.8% of total income. We can approximate the fall in real wages required to maintain employment by dividing the fall in GDP caused by carbon trading by 0.538. This approximation produces a decline in real wages that is slightly less than that implied in the ABARE study (the ABARE study implies division by ratios rising through the range from 0.415 to 0.514).

This means that instead of a 5.3% fall in jobs in Table 3, we could have a 9.8% fall in real wages. Instead of a 1.87% fall in jobs in Table 4, we could have a 3.5% fall in real wages. Using the trade-off implied in the ABARE study, this fall in real wages rises to 11.4% and 4% respectively. A 3.5% fall in real wages could be accomplished by having a 2-year wages freeze for all Australian workers simultaneously with the beginning of CO2 emissions trading at $US30 per tonne. A two-year wages freeze with average inflation at 2.5% would theoretically reduce real wages by 5.1% but we have assumed that only 69% of wages and salaries are effectively frozen. Again, no party advocating the signing of the Kyoto Protocol is advocating such a wages freeze. This suggests that job losses of the size we have calculated would follow.


Here we have reviewed what we believe to be the full cost of adopting the Stern report within the Australian economy. We estimate the full cost would be equivalent to a loss of 5.3% in Australian GDP. This estimate is based upon Stern’s own estimate of the long term social cost of carbon of $US85 per tonne of CO2.

This is as we expected, higher than Stern’s estimate of 1% of GDP. We would expect this because Australia is a high carbon use economy. Our estimate is however, within the range specified by Stern of -4% of GDP to plus 15% of GDP. It is at the top end of the smaller cluster specified by Stern as from -2% of GDP to +5% of GDP.

In the short term, should Australia sign the Kyoto Protocol and enter into carbon trading at a price of $US30 per tonne of CO2, this would have a short term cost of 1.87% of Australian GDP. This would result in a loss of 193,000 jobs of which 138,000 would be fulltime and 55,000 would be part-time. This is based on Stern’s own estimate of an initial cost of traded carbon in the range of $US30 per tonne.

These job losses could be reduced by introducing a 2-year wages freeze with the beginning of carbon trading in the region of US$30 per tonne of CO2. These sacrifices only make sense if we follow Stern’s recommendation and allow production of low carbon energy in all its forms including nuclear.

It may of course prove that objectives similar to that of Stern can be achieved in Australia by a carbon trading system other than that described within the Kyoto Protocol. Such a system may not be so initially costly to the Australian economy.