Home > Economics - General, Environment > Reply to Davidson and Robson

Reply to Davidson and Robson

May 17th, 2007

Phillip Adams and Peter Dixon have prepared a reply (over the fold) to the opinion piece by Robson and Davidson in the Australian which offered a range of incoherent criticisms of proposals to reduce emissions of greenhouse gases. Disgracefully, but not at all surprisingly, the Oz has declined to print it, marking yet another step in its decline.

Admittedly, the debate is so one-sided that printing the reply would have made it obvious how ill-advised it was to publish the Davidson-Robson piece in the first place. Dixon is Australia’s pre-eminent economic modeller, and Adams is his successor as Director of the Centre of Policy Studies at Monash. They have published extensively in leading economic journals on modelling and climate change, and their expertise shows. Robson and Davidson have essentially zero professional expertise on these issues, and that shows too. Of course, they have exactly zero professional expertise in climate science, and that hasn’t stopped them claiming the entire profession is wrong, so we shouldn’t be surprised.

Tim Lambert cleans up what’s left

Insurance against catastrophic climate change: backing the Kyoto protocol

In an opinion piece (The Australian, 11/5), Alex Robson and Sinclair Davidson attempt to ridicule a petition currently being signed by university economists calling on the Australian Government to ratify the Kyoto Protocol.

We have signed the petition for three connected reasons: (1) compelling advice from the scientific community suggests that a sharp cut in world greenhouse gas (GHG) emissions would substantially reduce the risk of catastrophic climate change over the next century; (2) the Kyoto forum offers the best available possibility for Australia to play a constructive role in setting up world-wide arrangements for cutting GHG emissions; and (3) as part of a world-wide effort, Australia could achieve deep cuts in its own GHG emissions at only a moderate cost in terms of reduced economic welfare.

It is on point (3) that economists have particular expertise, justifying the presentation of an “economists� petition.

Cutting GHG emissions is like buying an insurance policy: we incur a cost (a loss in GDP) to reduce a risk (catastrophic climate change). In any insurance decision, the cost matters. If a worthwhile reduction in risk costs 50% of income, then living with the risk may be preferable. But if it costs 1% of income, then taking the insurance policy may be the best option. So what will it cost?

For the last 20 years, we have been undertaking economic modelling exercises for Australian and overseas organizations on the costs of GHG reductions. Our modelling, and that of other quantitative economists around the world, supports the claim in the petition that:
“Credible estimates suggest that a 50% emissions reduction is achievable for less than one year’s economic growth�.

Robson and Davidson have difficulty in figuring out what this means. Just to be clear, we will explain it in terms of the report by the Allen Consulting Group to the Business Roundtable on Climate Change (March, 2006).

Modelling we contributed to that report shows Australia’s real GDP growing between now and 2050 at an annual rate of 2.2% under the assumption of no new GHG policies. In this scenario, Australia’s GHG emissions by 2050 are 80% above their level in 2000.

In an alternative scenario, Australia undertakes policies to reduce its GHG emissions by 2050 to 60% below their level in 2000. Even with this very deep cut in emissions, Australia’s GDP grows between now and 2050 at an annual rate of 2.1 per cent. The implication is that a massive 60% cut in GHG emissions (relative to the 2000 level) costs about 20 months growth – the level of GDP that we would have reached on January 1, 2050 is not reached until September 1, 2051. A lesser cut would incur a lower cost. Taking account of non-linearities (the first 1% cut is much easier than the last 1% cut), a reasonable estimate for the cost of the 50% cut mentioned in the petition is 12 month’s growth.

Why do modelling results suggest that GHG emissions could be sharply reduced at seemingly moderate cost? Are these results plausible?

The main GHG-emitting activities are fossil-fuel-based provision of electricity and motor fuels. In Australia, these account for about 5.4% of GDP. Advice from scientists and engineers indicates that the adoption of current alternatives to fossil-fuel-based technologies would no more than double the costs of electricity and motor fuels. As a back-of-the envelope calculation, this suggests that Australia could make a 50% switch to alternative technologies at a cost of 2.7 per cent of GDP, a little over an average year’s growth.

But this is a pessimistic view of the costs of climate insurance. If the world embraced the need for deep cuts in GHG emissions, we would expect rapid technical progress in GHG-benign technologies which would reduce the costs of their adoption.

Professor Philip Adams and Professor Peter Dixon
Director and former Director of the Centre of Policy Studies
Monash University

Categories: Economics - General, Environment Tags:
  1. melanie
    May 17th, 2007 at 22:10 | #1

    I guess the only problem I have with the Adams Dixon piece is that they don’t appear to include the cost of appeasing the vested interests in the coal mining industry. If 5.4% of GDP is currently accounted for by domestic power generation, what about coal exports? How much do we need to add for those? Howard has certainly shown an unwillingness to raise the price of coal sufficiently to make alternative power attractive and/or slow exports to China, and it is not clear that Labor will be any different – Minister-in-waiting Combet has just been given a Newcastle seat. Qld Labor is still opening mines. Confrontation seems to be a politically unlikely strategy, so what about the cost of buying their co-operation?

  2. Grumpy Old Economist
    May 17th, 2007 at 23:37 | #2

    Adams and Dixon should be grateful that The Australian stopped themselves from publicly embarrassing themselves as, at best, being unable to do simple mathematics.

    Do on a spreadsheet a comparison of GDP when it grows at 2.1% and 2.2%. You will see that the difference in the first year is 0.1%. As they point out, the difference grows to 4.3% in 43 years in 2050, about two years growth. But GDP is also 4.2% lower in 2049 and 4.1% lower in 2048. GDP is lower in every year. So the difference is not ‘one or two year’s growth’. Moreover, 4% of GDP in 2050 amounts to about 11% of this year’s GDP. If we take the present value of the differences in GDP from now until 2050, it amounts to 23 percent of current GDP at a 6 percent discount rate (and much higher if you use a lower discount rate as Quiggin favours).

    Take their second calculation. They assume to implement the target would reduce GDP by 2.7%. But then GDP would be permanently 2.7% lower than otherwise. The present value of that is 0.027/(r-g). Say r = 6% (Quiggin would have a lower rate) and g = 2%, the present value would be 67.5% of GDP. The cost is higher than when the policy is phased in over 43 years.

    That is why the claim that this can be done for ‘one year’s growth’ is dishonest.

  3. mugwump
    May 17th, 2007 at 23:38 | #3

    “The main GHG-emitting activities are fossil-fuel-based provision of electricity and motor fuels. In Australia, these account for about 5.4% of GDP. Advice from scientists and engineers indicates that the adoption of current alternatives to fossil-fuel-based technologies would no more than double the costs of electricity and motor fuels. As a back-of-the envelope calculation, this suggests that Australia could make a 50% switch to alternative technologies at a cost of 2.7 per cent of GDP, a little over an average year’s growth.”

    Something smells fishy. Why doesn’t the same “back of the envelope calculation” also suggest that overnight removal of fossil-fueled-based electricity and motor fuels altogether will only cost 5.4% of GDP?

    Obviously, nearly 100% of the economy is dependent on these two things, even if they only constitute 5.4% of GDP when considered in isolation. Or do these economists think the whole economy is linear?

  4. Grumpy Old Economist
    May 17th, 2007 at 23:51 | #4

    Sorry, the present value in the first example should be 35 percent of current GDP.

  5. Ernestine Gross
    May 18th, 2007 at 00:00 | #5

    A few years ago, an architect tried to convince me that economics is not concerned with the material welfare of humans but with ‘GDP’, ‘interest rates’, ‘unemployment’. In other words ‘the economy’ is conceived as a basic macroeconomic model. This is how ‘economics’ has been taught by the media for decades. This is the language of politicians at budget time.

    Of course I accept that unemployment and interest rates and even GDP measures may be related to the welfare of humans but not necessarily.

    Suppose we would stop measuring (or publishig) GDP figures. Would anything change that is of importance for the welfare of humans? Not necessarily. Suppose the worst possible scenario of global warming would happen. Would anything change? Yes. The welfare of huge numbers of people would be catastrophically affected.

    The idea of separating the enviroment and hence environmental damage from ‘economics’ is, IMHO, incomprehensible.

    The problem is that GDP measures ‘market values’ and the market is incomplete. Global warming is, as is gradually filtering through to the media, a significant example of market failure. One cannot insure against the risk, as described by climat scientists, by giving up some GDP (‘cost’) because the markets aren’t there. And, if the risk of global warming is ‘significant’, then all prices that are currently in GDP measures are ‘wrong’. They provide ‘false’ price signals.

    Perhaps there is a communication problem in the sense that macroeconomic models, which for some policy questions (eg estimating tax revenue) may be considered essential, may be too esoteric for the population at large.

  6. May 18th, 2007 at 00:34 | #6

    What a sensible, practical response by Phillip and Peter.

  7. Valuethinker
    May 18th, 2007 at 00:42 | #7

    50% switch to alternative technologies would *not* reduce Australia’s final CO2 output by 60%?

    If output is 100 now


    it goes to 180 under Business As Usual (BAU)

    then if we halve it (costing 2.7% of GDP)

    then we are at 90?

    But we have another 30 to go?

    The actual necessary cut (per unit of GDP) is on the order of 80% (GDP doubles, carbon emissions fall by 40%). It’s not quite as bad as that: the ‘natural’ BAU fall in CO2 emissions per unit of GDP in the US is about 1.5% pa, due to greater efficiency, switches to cleaner fuels, and movements in the economy towards less GHG producing activities (healthcare, a low GHG activity, is 1/6th of the US economy).

    If my calculations are right, this underscores the scale of the task (switching the entire carbon-based energy component of the economy onto technologies that produce up to 80% less CO2).

    I still think it is perfectly doable. My rule of thumb is that half of that savings will come from applications of current technology– from painting rooftops white, to thermally efficient windows, building windmills etc.

    The other half will come from new technologies we don’t have yet.

  8. jquiggin
    May 18th, 2007 at 06:27 | #8

    GOE – the present value point (which Robson and Davidson tried too) is a silly presentational trick. You can make any economic flow look large by expressing it as a present value. So what? You then have to do the same for the costs of doing nothing, and in the end nothing changes, except that all the numbers are bigger by a factor of 100/6 and harder to communicate.

  9. May 18th, 2007 at 09:44 | #9

    A large proportion of the CO2 emission reductions can come from efficiency gains. There is a perception that industry is efficient already and there is not much to be gained.

    From what I have read the economy has reached a point now where it is efficient enough. ie: with cheap coal fired electricity available and energy being a small cost in most business there is no incentive to install the more efficient machinery that is available.

    However when companies DO implement the more efficient technologies the gains can be dramatic:

    “Georgia Pacific:

    Georgia-Pacific, in partnership with the U.S. Department of Energy’s Industrial Technologies Program, worked to identify energy conservation measures and operating practice improvements that would increase the overall energy efficiency of a mill producing printing paper, board, and tissue products in Crossett, Arkansas. An assessment team conducted a mill-wide energy survey at the facility and identified three heat recovery projects that could reduce annual natural gas use by 1.8 million MMBtu, saving about US$4.8 million each year. The overall payback period for the heat recovery projects will be less than 1 year. Operational improvements were also identified that could produce an additional 1.5 million MMBtu in natural gas savings at an added savings of $4.8 million annually.”

    The main point is that a large proportion of the cuts in emissions may not cost anything at all as companies can save large amounts of money and energy at the same time. While we are coddled with cheap coal power our industries will stagnate ,still shrilly demanding cheap power, while other countries move ahead and change to more efficient industry.

    Meanwhile I am trying to think of better ways 300 million could be spent on renewables.

    1. How about kick starting the solar tower in Mildura that is danger of stopping.

    2. Set up a fund to lure David Mills back with a 300 million dollar solar thermal plant to his revolutionary design.

    3. Fill in your own renewable project that is floundering due to lack of funds

    No the Queensland Government is investing it in the status quo and clean coal.

  10. Grumpy Old Economist
    May 18th, 2007 at 09:46 | #10

    No, it is not a presentational trick to express future flows in net present value. In fact, discounting is the only way to compare dollars received in different time periods. At least you recognise that GDP is a flow, not a stock – something which seems to escape Dixon and Adams.

    But what can be said about the claim that emissions can be cut substantially for the cost of ‘one year’s growth’? It is more than a presentational trick, but is plain wrong. In Dixon and Adam’s example, the cost in 2050 is 2 years of 2050 growth. The cost in 2049 is two years of 2049 growth. The cost in 2048 is two years of 2048 growth, And two years of 2050 growth is about 5 years of current growth – but that would be a presentational trick because it occurs far into the future and we have to discount it back.

    Its embarrassing to see how so many economists are prepared to publicly sign something without giving it much thought. It is too horrible to contemplate the alternative: that they gave thought and either didn’t realise or care how misleading it is. As for those who drew up the petition ….

  11. melanie
    May 18th, 2007 at 10:13 | #11

    #7 valuethinker, it is not the GDP that is 180 in 2050, it is the carbon emissions.

    #10 GOE, what will be the GDP losses be under the BAU scenario with GW? How much, for just one example, does the MDB currently contribute to GDP? Isn’t your present value technique a linear one? and how do you account for the non-linearities that the authors mention?

    Everyone, are my questions really dumb? I’m not an expert in this area (not familiar with the Adams-Dixon modelling at all), just reading out of curiosity – which is why I have not signed the petition. Can anyone give me a tutorial on my first question at #1?

  12. jquiggin
    May 18th, 2007 at 10:21 | #12

    GOE, I see no sign that anything has escaped Adams and Dixon. As your own presentation suggests, the best way of putting the issue is as a flow of annual costs, expressed as a per cent of GDP.

    Also, the calculation you present in the second para of #2 is incorrect. Doing a full PV analysis from now to 2050 or 2100, then converting to a constant percentage flow would produce a smaller annualised cost than the one given by Adams and Dixon, since costs grow over time (relative to GDP) with the required reduction in emissions. If we, or Adams and Dixon had done this, we might reasonably have been accused of misleading presentation, and of course, the choice of discount rate would have been a topic of contention.

    Finally, please read the discussion policy before commenting further. Comments in the snarky tone you are adopting are discouraged, particularly from pseudonymous commenters. If you want to participate further, along the lines you have taken so far, I request that you identify yourself.

  13. jquiggin
    May 18th, 2007 at 10:46 | #13

    Melanie, the main impacts on coal exports will arise from actions taken overseas. These are captured in the Allen Consulting model, since the scenario requires a global reduction in emissions.

    There’s a distributional question also implicit in your posts. If the coal companies, and other large emitters are fully compensated, the loss can be thought of as a roughly uniform impact for all Australians. If they bear some of the costs, the impact on their (domestic and overseas) shareholders is greater, and the impact on other Australians is less.

  14. DD
    May 18th, 2007 at 11:05 | #14

    I think it’s interesting that Davidson doesn’t seem to have gone out of his way to defend his position on this, which is interesting since he usually pops up on Australian economics blogs quite regularly.

    Just to avoid any confusion, a note that DD is not Derrida Derider, whom I often refer to as dd – JQ

  15. Sinclair Davidson
    May 18th, 2007 at 12:11 | #15

    Davidson’s day job has been unusually busy these last three weeks, what with auditing DEST publication returns, ARC Assessor Reports, Journal refereeing, a major proposal and his own ongoing work. Which part of the position needs defending?

  16. May 18th, 2007 at 12:13 | #16

    Its embarrassing to see how so many economists are prepared to publicly sign something without giving it much thought.

    For a quite funny example of people signing something that they have not really thought about check out this short video extract:-

  17. May 18th, 2007 at 14:37 | #17

    GOE stole what I was going to say. The appropriate way to compare costs of things over different times is with present value, and the present value of cutting co2 is high. And even higher if you use the Quiggin preferred discount rate (which I wouldn’t suggest).

    Of course, we also need to see the present value of the benefits from cutting co2 before we can make a judgement.

  18. May 18th, 2007 at 16:33 | #18

    I do PV, FV and NPV calcs regularly and I cannot see where GOE is wrong. I re-ran his / her numbers through a spreadsheet and they seem right. To say that, under their model, we would be only giving up 2 years of growth is simply wrong.
    Of course, as Ernestine correctly points out GDP is not everything and melanie (and John, obliquely) is also right that the model assumes that 2.2% is actually achievable despite any prospective climate change.
    Be that as it may, Adams and Dixon chose the numbers to model on and then, as far as I can see, got them wrong.

  19. Ernestine Gross
    May 18th, 2007 at 16:43 | #19

    Oh, in case there is any misunderstanding, IMHO, present value calculations, as understood in Finance, does not work either.

  20. May 18th, 2007 at 17:01 | #20

    That may or may not be the case – but I was trying to critique the model they used, not my own.
    Out of interest, what technique to you use to compare future costs or benefits to current costs or benefits?
    As a side note: if we assume they are equivalent then the gap between the two scenarios blows out even further. Assuming that future = present (on their scenario) then the difference in total production is nearly twice current GDP.

  21. jquiggin
    May 18th, 2007 at 17:13 | #21

    AR, I’ll try to spell it out more carefully.

    Let’s start with the basics. If you have a known constant flow of costs, and a discount rate, you can express it either as an annual amount, or a PV. Neither one is logically more correct than the other. The fact that you can derive one from the other using a spreadsheet reflects this. The only difference is that people are used to the annual flow measure (look, for example at the huge literature on effects of microeconomic reform which uses this measure exclusively), so the PV looks large.

    If you have a known series of costs, and a discount rate, you can express them either as a PV or as an annualised equivalent constant flow, or as an annualised equivalent constant percentage of income. Again, the only difference is one of ease of communication.

    In the present case, what we mostly have is comparative static estimates of the equilibrium cost of a policy that reduces emissions over time, relative to BAU, reaching a 50 or 60 per cent reduction in 2050. The output is the reduction in the flow of income over 2050 (around 2 per cent), but we don’t have the time path from the (near-zero) costs of the initial reductions to the 2 per cent level.

    GOE’s calculations (and Robson-Davidson’s) assume that the 2 per costs are incurred immediately, and forever, then use the approach in the first para to convert them to a PV. This is, quite simply, wrong. (You can get a back of the envelope estimate of how wrong they are from your spreadsheet, using linear interpolation).

  22. May 18th, 2007 at 17:52 | #22

    I am comfortable with the methodology being used: I am just trying to get across the mathematics.
    If you drop any compound growth rate from an annualised compound 2.2% to an annualised compound 2.1% (and they have to be compound to be correct) then the figures work the way GOE has calculated. Assuming GDP2007 = 100, then GDP2050 = 260.519 under the 2.2% scenario and 249.536 under the 2.1% scenario. Fair enough, the difference is two (approximately) year’s growth 4.4%.
    Problem is that, as GOE pointed out, this is a cumulative flow, not a total. The shortfall in 2050 must be added to the shortfall in all the preceding years if all of the lost production is going to be properly counted. The total lost production is not the 10.98 units of GDP in 2050, but also the 10.51 foregone in 2049, the 10.05 foregone in 2048 and so on.
    To say that “…a reasonable estimate for the cost [i.e. total production lost] of the 50% cut mentioned in the petition is 12 month’s growth…” is, as far as I can see, materially wrong. The cost in total production foregone is not “…12 month’s growth…”, but, with a zero discount factor, it is nearly 2 year’s current GDP or 1/2 of the average year’s production over the period.
    At higher discount factors the cost gets lower, but you need to go to (an unrealistic) 18.47% to get a total cost equal to two year’s (discounted) growth.
    Note that I am not arguing for or against a given strategy, just looking at the mathematics and the way they have been represented in the piece.

  23. May 18th, 2007 at 17:54 | #23

    I feel quite comfortable doing NPV calculations but can somebody give me a quick insight into how policy makers decide on a discount rate and why it is different to that used in finance. Try and dumb it down without losing the essence if that is possible.

  24. May 18th, 2007 at 17:57 | #24

    #1 Melanie: Most of our coal exports go to Japan, Korea and Taiwan. Very little goes to China. This is a common misconception.

    #13 ProfQ: What is the Allen consulting model? It seems to bear no resemblence to reality, past, present or future. If a price is put on carbon emissions in Australia and Japan, the surplus coal will be redirected to developing countries where it will be burnt in the same atmosphere (probably in a dirty and inefficient power station)

    The only thing that can prevent this happening is a global rollout of “clean coal technology” (Buckleys!) and/or an enforced reduction in carbon emissions in developing countries (Buckleys again!)

    We have to keep the coal in the ground

  25. May 18th, 2007 at 18:11 | #25

    The correct way to do it has been the subject of many a doctoral thesis and so a blog comment would be a little limiting.
    If we were to be completely honest, though, the way policy makers do it tends to depend more on the answer required than on the merits of the methodology.
    I normally find the better way to do it is to present a range. The more sensitive the present value is to discount rate change the more sceptical you should be about the model. Not throw it out, but be sceptical.

  26. jquiggin
    May 18th, 2007 at 18:22 | #26

    AR, you’re completely missing the point. I’m surprised at this.

    Perhaps it would help to note that nearly everyone in the debate, on all sides (Stern, ABARE*, Allen Consulting and many others) has used annualised flows, and no-one (except Robson-Davidson, GOE and you) has questioned this, for the reasons I spell out in #21.

    Rather than restating your own calculations, I suggest you reread the first para of #21 and then say what, if anything, you object to there.

    *ABARE played the present value trick quite a while ago, and I pulled them up at the time. They haven’t used it lately.

  27. jquiggin
    May 18th, 2007 at 18:26 | #27

    Carbonsink, it’s obviously necessary to bring developing countries into the post-Kyoto round, and this has always been recognised. Assuming the US comes onside after 2008, developing countries have little choice but to drive the best bargain they can, in terms of compensation, CDM and so on, then join in. The alternative is to face border taxes or complete loss of export markets.

  28. May 18th, 2007 at 19:02 | #28

    I do not object to anything there – it is correct. The problem comes when you look to measure the differential costs between the two strategies.
    Mathematically (and this is stretching my remembering of calculus) the comparison of the angle of slope between two differing strategies is fine – if the growth rate is the issue. The way they have expressed their conclusion, though, is not, as it appears to be, on the slope, but on the area under the curve.
    To be (hopefully) a little clearer (and to burn the mathematics) – the true opportunity cost here is not the foregone growth, but the foregone production, which is increased by the growth.
    I am sure I must have made a mistake but I cannot see it.

  29. jquiggin
    May 18th, 2007 at 19:22 | #29

    Adams and Dixon haven’t measured the area under the curve, they’ve measured the value of the difference at a point (2050) since this is what the model estimates. They’ve then worked back to derive the difference in slope.

    You can now work out the area under the curve if you want to, but this does not produce any new information, just a misleading way of presenting what we already know.

  30. May 18th, 2007 at 19:29 | #30

    We will have to disagree, then. To me, the cost of the strategy is not the foregone growth but the foregone production.
    To put it yet another way, we do not eat the increase in food output – we eat food.

  31. May 18th, 2007 at 20:07 | #31

    #27 ProfQ
    Assuming the US comes onside after 2008
    God lets hope so. I don’t have much hope when I see one of the more enlightened candidates co-sponsoring Coal-to-Liquid Fuel Promotion Act of 2007

    The alternative is to face border taxes or complete loss of export markets.
    Do you seriously believe the world will impose taxes on exports from (say) China if they refuse to enforce emissions reductions?

  32. jquiggin
    May 18th, 2007 at 20:18 | #32

    You’re still missing the point, AR.

    To use the food analogy, if you were reporting on hunger in a way you wanted people to understand would you rather say that the average person in a poor country falls short of their daily nutritional requirements by 500 calories (or maybe better by 20 per cent), or would you rather convert that shortfall to a lifetime estimate using a discount rate, and come up with a number like 36.5 million equivalent PV calories. If the latter, go with GOE. If the former, with Adams and Dixon

  33. melanie
    May 18th, 2007 at 20:18 | #33

    JQ #13, thank you for beginning to answer my questions. My main worry with respect to overseas emissions is that, since we currently account for about 30% of world coal exports (thanks carbonsink #24), we would not be contributing to global reductions by continuing to export. Carbonsink is possibly not an economist – s/he says keep it in the ground. Such a solution would cost a lot would it not? So, just as we export uranium without considering the global danger, we rely on our trade partners to be willing to bear these costs.

    I think distributional issues are important for soc dems. When you say full compensation would achieve a ‘roughly uniform impact for all Australians’ do you mean that poor Australians would face a higher burden relative to their incomes? (It seems to be implied by your next sentence.)

    Carbonsink #24, see my above to JQ. Why not keep exporting the coal and make the others bear the cost of reducing emissions? In addition Chinese and Indian demand is growing faster than Japanese. Under a BAU scenario they could be our main customers by 2050 (if anyone is using coal by then ;).)

    Andrew Reynolds, why is the under-curve area important? Doesn’t it still represent an increase in production over current levels? If the economy is still growing in per capita terms it seems possible that we’ll all be better off than in the BAU scenario. As Ernestine pointed out GDP isn’t everything and I’d rather live in a world not full of climate refugees and wars over water, or whatever. (not that I’ll be here though! I’m thinking about other people’s grandchildren.)

  34. jquiggin
    May 18th, 2007 at 20:20 | #34

    “When you say full compensation would achieve a ‘roughly uniform impact for all Australians’ do you mean that poor Australians would face a higher burden relative to their incomes? (It seems to be implied by your next sentence.)”

    Sorry to imply this. I meant an impact that is a roughly proportional reduction in all incomes (more precisely, all consumption).

  35. melanie
    May 18th, 2007 at 20:23 | #35

    JQ, that was quick! Thank you.

  36. May 18th, 2007 at 20:54 | #36

    #33 Melanie:
    Carbonsink is possibly not an economist
    No, carbonsink (he) is not an economist, but does know the economy is a wholly owned subsidiary of the environment.

    Yes it will cost a lot, about AUD 25 billion annually, or about 10% of our total exports, and around 2% of our GDP. Coal mining employs around 30,000 people, or 0.15% of our population. Phasing out coal exports may seem like economic madness (not to mention political suicide) today, but public opinion is moving very fast on this issue. In 3-5 years time it might be politically possible.

    QUESTION: If Australia made $25 billion annually from exporting asbestos, would we keep doing it? As Tim Flannery says, coal is the new asbestos.

    Getting China and India to “bear the cost of reducing their emissions” is a lovely idea (they might even consider doing it once they’ve raised their living standards to Western levels) but it ain’t gonna happen anytime soon. The reality is we have no control over how the coal is burnt once it leaves our shores, but we do have control over the coal in the ground. For God’s sake lets keep it there.

  37. melanie
    May 18th, 2007 at 21:43 | #37

    Carbonsink, that’s why I liked Ernestine’s original comment (#5). The problem is essentially political though. If we were exporting $25 bn annually in asbestos, I’m sure “they” would find reasons to continue justifying it. Despite all the known dangers of uranium, we’re committing ourselves to increased exports of that product. I agree with you that the ideal solution is to keep the coal in the ground, but it would, I think, cost this country a lot more than seems to be included in the Allen consulting model (ACM). Therefore, global agreements (implied by the ACM) offer better prospects. Less cost to Australians, more sharing of the costs around the world, more politically acceptable in coal producing countries.

    Besides, a well functioning global carbon market (BIG ‘if’ here – due to the necessity for political decisions on where to cap emissions) would enable us to export coal while still reducing global emissions. The EU market clearly undervalues carbon emissions due to the currently inappropriate cap – but that could be changed.

    If I’m thinking straight at this late point on Friday night, in the absence of a global agreement on emission reduction the Adams-Dixon model would need to add another 2% to bring the energy component up to 7.4% of GDP, then add in the cost not only of switching to renewables for domestic power generation, but the loss of export income. That would give a measure of the cost to Australians of reducing global emissions. Mind you, it’s not clear why Australians should bear all of that cost.

  38. melanie
    May 18th, 2007 at 22:20 | #38

    Sinclair Davidson #15, that was a neat bit of sophistry. Very Boyzone! We’re all busy these days with the same sort of crap. It really depends on your priorties – which, in your case, were clearly sufficient to go into print on the subject. Suggest you read the thread to find out what questions you need to answer.

  39. Grumpy Old Economist
    May 18th, 2007 at 23:11 | #39

    Deleted – if you want to participate further, please do so under your real name, not a pseudonym, as requested

  40. melanie
    May 18th, 2007 at 23:46 | #40

    OK GOE, I did that and I got a GDP at 2.1% growth rate that was 4 percentage points below what I got for 2.2%. Big deal. It is still very much larger than today’s GDP. The question I have for you is this: should we be prepared to sacrifice those 4 percentage points for an improved quality of life or not?

  41. Peter Wood
    May 18th, 2007 at 23:58 | #41

    It seems that the upstream vs downstream debate has been heating up! I think that in order to analyse this properly we need to look at the global problem of achieving cooperation on climate change (in this case).

    I did a post on this this morning on the CEDA thread, suffice to say, if we reduce our coal exports, I think it would be a positive step for global cooperation on climate change mitigation.

    We should do this in a sensible way though.

  42. May 19th, 2007 at 00:39 | #42

    The question I have for you is this: should we be prepared to sacrifice those 4 percentage points for an improved quality of life or not?

    Surely it depends on how much improvement.

  43. May 19th, 2007 at 00:41 | #43

    The correct way to do it has been the subject of many a doctoral thesis and so a blog comment would be a little limiting.
    If we were to be completely honest, though, the way policy makers do it tends to depend more on the answer required than on the merits of the methodology.

    Thanks Andrew. Thats not entirely satisfactory but I suppose it will have to do for now.

  44. Mark U
    May 19th, 2007 at 00:41 | #44


    I do not understand GOE or Andrew Reynolds point. I can’t see how GOE gets such big losses of GDP in his original comment. GOE’s original 67.5% of GDP at #2 seems to fail the reality check. I agree with the steps that Andrew Reynolds outlines at #22, but he then says this amounts to nearly 2 years of current GDP, which is precisely what Adams and Dixon have said.

    If, as suggested by GOE at #39, I calculate the discounted present value of GDP under the status quo (2.2%) and the emissions reduction (2.1%) scenario I get and add up the numbers, I get a loss of GDP of 175.6 out of a total GDP of 7296.3 over the whole period. This is 2.5% of total output over the 43 year period and is less than two year’s GDP in 2007 and one year’s GDP in 2050. This calculation corresponds to a zero discount rate. If I use a discount rate of 6%, the discounted loss of output is only 1.6% over the 43 year period.

    In other words, what Adams and Dixon have said seems to be perfectly correct.

  45. mugwump
    May 19th, 2007 at 03:57 | #45

    AR and GOE are correct. Quiggin’s analogy illustrates it well: if I am undernourished by 500 calories daily, then not only am I hungry everyday, I end up shorter and weaker for the rest of my life.

    Saying a 50% reduction in emissions only loses us 12 months growth in 2050 implies (in joe punter’s mind) that we’re just 12 months behind where we would have been under BAU. But that’s false: we’ll also be stunted and weak from 43 years of undernourishment.

    That’s AR’s and GOE’s point and at least to me it seems a much more accurate description of the situation.

  46. May 19th, 2007 at 08:16 | #46

    this kind of discussion would be more useful if corporation profits were more important than racial survival.

    but i take the view that constantly expanding human population is going to press up against the capacity of the planet to sustain us. the current water crisis is not the first signal we may already be too many, but it should inspire reflection to a much greater degree than has happened. at a time when energy production may be polluting the air to a dangerous degree, we are being told that our masters are ‘coping’ with the unlucky shortfall of rain by using more energy to produce drinking water.

    why does no one ask, why must we keep increasing in numbers?

    even more fundamental, why do we allow ourselves to be ‘farmed’ by a small group of people with no visible distinction of ethics, wisdom, or even education?

  47. May 19th, 2007 at 08:40 | #47

    #37 melanie wrote:
    enable us to export coal while still reducing global emissions
    How does that work? If you’re exporting the coal, its still getting burnt, so emissions are not reduced. Unless of course the “clean coal” fairytale materialises in the next 20 years.

    Lets say by some miracle a globally enforced emissions reduction treaty is signed next year. The net result will be a reduction in global demand for coal. The coal price will collapse, thousands of miners will lose their jobs, and more coal will stay in the ground. i.e. it will have exactly the same effect as phasing out coal mining through government legislation.

    Mind you, it’s not clear why Australians should bear all of that cost.
    Its not clear to me why Australians (or a very small subset of Australians) should be profiting by exporting a planet destroying substance.

  48. Grumpy Old Economist
    May 19th, 2007 at 08:41 | #48

    It’s a pity yopu are now deleting my comments as I think I have a way of explaining this that could resolve our differences.

    We are arguing how to best summarise the modeling of Dixon and Adams who model the costs of implementing the greenhouse gas reduction as lowering the growth rate from 2.2% to 2.1% from 2007 to 2050. They claim this cost is equivalent to two years economic growth.

    Say I had $100 in a savings account in 2007. if it was to grow at 2.2% it would reach $260.52 in 2050. if it were to grow at 2.1% it would grow to $249.54 and would reach $260 in 2052. That seems to be what Dixon and Adams have in mind. But that is the stock of savings. GDP is a flow concept.

    The relevant experiment to think about is you have a share which pays a $100 dividend this year and the dividend is expected to grow at 2.2% per year. At a 6% discount rate your share is worth $2,122.55.

    If instead, the dividend is expected to grow at 2.1%, your share would only be worth $2,089.50, 1.5% less. But would be misleading to describe this as ‘two years dividend growth’. It is a 1.5% fall in wealth, not income.

    So it would be accurate to say we could cut greenhouse gas emissions for 2% of our wealth. But we are used to expressing things in terms of GDP. The fall in wealth is equivalent to reducing the flow of GDP by 2% every year. Or in present value term, equivalent to 35% of this year’s dividend.

    It is an embarrassment that our ‘pre-eminent modellers’ could be so hopelessly confused between stocks and flows, between wealth and income.

  49. jquiggin
    May 19th, 2007 at 09:23 | #49

    GOE, I’ll respond to this comment after which I will firmly enforce a requirement to identify yourself if you wish to debate further (you can email me privately if you have a good reason for keeping your name confidential).

    You say:

    So it would be accurate to say we could cut greenhouse gas emissions for 2% of our wealth. But we are used to expressing things in terms of GDP. The fall in wealth is equivalent to reducing the flow of GDP by 2% every year.

    So far so good. 2 per cent is about the right number, and of course, it’s equivalent to one year’s economic growth, as claimed.

    But then you go on:

    Or in present value term, equivalent to 35% of this year’s dividend.

    But of course, as you just pointed, out this is misleading, since you are taking the ratio of a change in stock (wealth) to a flow (GDP), something that we are all unused to, and which can only mislead, without adding any new information.

    “It is an embarrassment that our ‘pre-eminent modellers’ could be so hopelessly confused between stocks and flows, between wealth and income.”

    As I’ve shown, the confusion is yours (and Robson-Davidson’s).

  50. jquiggin
    May 19th, 2007 at 09:37 | #50

    Since we have the example of a share, I’ll clarify the illustration. Suppose that corporate profits are initially tax free and the government introduces a 2 per cent tax on profits*. We could say that this will reduce post-tax earnings by 2 per cent every year, or that the value of equity will be reduced by 2 per cent. Both of these would be sensible ways of describing the situation, and both are commonly used.

    Alternatively, following GOE, you could say that the reduction in the value of equity is equal to 35 per cent of one year’s profits. I’ve never seen this description used in a situation of this kind, for the obvious reason that it confuses stocks and flows.

    * If you’re worried about general equilibrium effects, assume the tax is only applied to a small sector of the economy. If you’re not worried about general equilibrium effects, ignore this note.

  51. Grumpy Old Economist
    May 19th, 2007 at 11:39 | #51

    Thank you for leaving my comment up. I really do appreciate it. Telling you my name is a pretty silly requirement as I could just make up a name. My desire is for anonymity is because my experience is that economics professors can be extremely vindictive, especially to those pointing out the Emperor has no clothes.

    So we now have resolved our dispute. I think the term ‘one year’s economic growth’ is 2 percent of GDP for one year. you think it means two percent of GDP every year for 43 years.

  52. melanie
    May 19th, 2007 at 12:14 | #52

    #47 carbonsink, By imposing a global emissions cap you put a price on carbon for those countries that are currently emitters. It therefore encourages switching out of fossil fuel technologies, but enables those less able to switch quickly (e.g. China) to keep burning by buying emission entitlements in the market from those who don’t need them. As seen in Europe, where the Germans have reduced emissions quite a lot, the price of carbon has fallen, meaning that the emissions cap was set at too high a level. In the case of Australian coal, the industry would not collapse overnight, but coal would start to be phased out, provided the global cap was set at the right level. As I said, big if. Australian water markets have demonstrated the pitfallls already.

    Politically, it seems unlikely that anyone in this country will just ban coal exports. Without a carbon market, the alternatives are to tax them or to compensate the miners. A tax has the same effect of raising the price of carbon emission for end users, but it’s more likely that in the absence of a global agreement they’ll switch to alternative suppliers who don’t tax, thereby causing the collapse of the industry. Compensation would be paid out of government revenues, distributing the cost among the domestic population. With a global market, all coal users pay the costs of phasing out the industry.

    that’s my (non-expert) reading of the economic arguments anyway.

  53. jquiggin
    May 19th, 2007 at 12:17 | #53

    Melanie, I agree with what you say, although it is not necessarily true that China will be a net buyer. This depends on the detail of the initial allocation, which will doubtless be the subject of tough bargaining.

  54. Sinclair Davidson
    May 19th, 2007 at 12:53 | #54

    Melanie, I have nothing to add to GOEs comments on the confusion surrounding the estimates of costs.

  55. Ernestine Gross
    May 19th, 2007 at 14:53 | #55

    To all who consider growing parseley and other herbs in their backyards – or even on balconies or in kitchens – in an attempt to make a marginal contribution to the reduction of CO2 emmissions due to reduced transportation – may be creating confusions about the estimates of ‘costs’ in the eyes of Sinclair Davidson because GDP will decline, assuming all else constant. I don’t even dare talking about those rebels who are harvesting water from their roofs, using ‘primitive technology’ (ie no motorised pump) known as gravity feeding.

    GOE: You confused rows and columns in your earlier explanation of your arithmetic. Incidentally, Prof Quiggin’s first degree is in mathematics.


  56. Mark U
    May 19th, 2007 at 15:39 | #56

    The bottom line is that Adams and Dixon were correct. The loss of GDP is small relative to total GDP whether you look at GDP on a year-by-year basis or, alternatively, add up every year’s GDP and discount it by a “suitable” discount rate.

  57. May 19th, 2007 at 18:19 | #57

    #52 melanie, you misunderstand what I’m saying. My preferred option (of course) would be a globally enforced price on carbon emissions, that would be fantastic, but I just can’t see it happening … ever.

    Given that it won’t happen, the only way Australia as a coal exporter can reduce its emissions is by keeping the coal in the ground. If we reduce our domestic emissions, we’ll just export the surplus and someone else will burn it.

    Obviously its not possible to stop coal exports overnight, but a five year phase-out might be politically and economically feasible sooner than you think.

    BTW, most of Germany’s reduction in emissions since 1990 has been due to the closure of inefficient power stations in the former East Germany. CO2 per capita in France is close to half that of Germany because most French electricity is nuclear.

  58. May 19th, 2007 at 19:23 | #58

    Sorry, but I still disagree. The fundamental point (IMHO) is that growth in an underlying is being confused for the underlying itself.
    To go back to the food analogy, as I believe this works best – the difference is between an increase in the rate of food output and the actual quantity of food produced. The amount of food produced each and every year is what is available to be eaten.
    If I sacrifice .1% of growth each and every year then I am sacrificing .1% of food in the last year, but 0.1% of food every year – and even this understates it as it is a compounding amount.
    To say that the “cost” of the strategy is only a percentage of the growth is wrong – the cost is the production foregone.
    It is not the growth that is foregone – it is the production.
    My argument is that this line:

    As a back-of-the envelope calculation, this suggests that Australia could make a 50% switch to alternative technologies at a cost of 2.7 per cent of GDP, a little over an average year’s growth.

    is simply wrong. What is not foregone is growth – it is production. If it was put this way:
    As a back-of-the envelope calculation, this suggests that Australia could make a 50% switch to alternative technologies at a cost of 0.1% of GDP in 2008, increasing to 4.3 per cent of GDP in 2050. (I hope i have the numbers right – the spreadsheet is on another computer.)
    it would be right.
    It may well be worthwhile making the sacrifice – I am not expressing an opinion on this – but we should at least properly depict the true cost.

  59. jquiggin
    May 19th, 2007 at 19:36 | #59

    Andrew, your numbers are incorrect. The correct statement is that we could make a 50 per cent reduction in emissions, relative to BAU, at a cost of 0.1 per cent of GDP in 2008, rising to 2.7 per cent of GDP in 2050. (Reread the article to confirm this).

    Obviously, any PV number, expressed as an annual flow, will lie between 0.1 and 2.7. Bending over backwards to avoid understating the cost, Adams and Dixon report the 2.7 number. And still they get complaints and (ludicrous) accusations of incompetence from GOE-Davidson. Chalk it up as yet another example of tribalism/ideology trumping reason on the delusionist side of this debate.

  60. Sinclair Davidson
    May 19th, 2007 at 19:55 | #60

    I’m sorry to quibble, John. I don’t think that GOE or I have referred to either profs Dixon or Adams as incompetent. I don’t recall doing so, nor can I find GOE doing so. A search of this thread finds no use of that word. While I disagree with their assessment of the costs, calling them incompetent would be wrong (if not rude and defamatory).

  61. jquiggin
    May 19th, 2007 at 20:01 | #61

    To take the very first sentence posted by the mysterious GOE

    Adams and Dixon should be grateful that The Australian stopped themselves from publicly embarrassing themselves as, at best, being unable to do simple mathematics

    Given that Adams and Dixon do mathematical modelling for a living, I think it’s fair to read an imputation of incompetence here. As you say, wrong, rude and possibly defamatory (not to mention ungrammatical). Scroll down and you’ll see plenty more. You might want to make it explicit that you disagree with GOE on this assessment, given your apparent endorsement at #54.

  62. melanie
    May 19th, 2007 at 20:18 | #62

    #57 carbonsink, interesting info. I see that if we cut our carbon emissions by 60% we sould still be emitting more per head than France, Switzerland and Sweden. Should we go nuclear?

    I don’t share your faith in the feasibility of shutting down coal exports in 5 years. Last year they were 11.1% of all our exports. In the December quarter the trade deficit increased to $3.5 bn (11.6 bn for the whole year) and the current account deficit was $15 bn, net foreign debt rose to $571 bn and the debt service ratio to 37.7% of foreign earnings. (the last figure is an estimate from the Economist Intelligence Unit, the actual ratio for 2005 was 33.8%). Put another way our foreign debt service last year was over $79 bn, of which coal paid almost 1/3. Coal is our single largest export commodity – over one and half times more than the next one, iron ore.

  63. Sinclair Davidson
    May 19th, 2007 at 20:46 | #63

    That’s a very long bow you’re drawing there. I don’t think that follows at all. I have scrolled down. I can see a fairly robust, even passionate, debate debate about what it means to reduce growth by 0.1 over 43 years. You say this is small, GOE (and Andrew Reynolds) say this is big. I agree with them. (For the record, I don’t know either of them). The more important issue is that there is still debate over the appropriate policy and the costs of that policy; even after Dixon and Adams have emailled their clarification to every academic economist in the country (except Alex and I – although many forwarded it on to us) and you posted it here.

  64. jquiggin
    May 19th, 2007 at 21:19 | #64

    Sinclair, how do you know you don’t know GOE?

    And, to be clear, do you agree or disagree with the claim that Adams and Dixon are unable to do simple mathematics?

  65. Sinclair Davidson
    May 19th, 2007 at 21:35 | #65

    That is a good question. You’re quite right. I don’t know which of our economist collegues (academic or private sector – my perception is the latter) GOE is in real life. I don’t know his (her?) “secret” identity.

    I think my comment at #60 answers your other question.

  66. May 19th, 2007 at 23:40 | #66

    #62 melanie:
    I think the Swedes and Swiss have lots of hydro as well as nukes. We used to have some hydro, but then it stopped snowing in the Snowies. I think economists call than an “externality”.

    RE: phasing out coal exports. Yes I know its mad, but we do lots of mad things that no-one seems the slightest bit concerned about. For instance, of the $11.6 bn trade deficit more than $8 bn is oil. Why? Australia’s oil production has been declining for 7 years, and our oil consumption has been growing rapidly over the same period. Yet we give a tax break to imported SUVs, the fuel excise has been fixed at 38c/L for six years (hello? market distortion anyone?), the FBT system encourages people to maximise annual kilometres driven, and we subsidise the manufacture of big gas-guzzling family cars.

    But I digress. An economist would probably tell you this is all the more reason to keep exporting coal, so we can pay for the oil to feed our big-ass cars.

    I’m sorry but its mad, and I’m convinced this quibbling over the costs of action on climate change will be seen as mad within the decade. Frankly, we really need a kick up the bum like a capital city running out of water. You’ll be amazed how quick the pollies can shut down the coal mines once the voters figure out that coal and climate change are connected.

  67. Mark U
    May 20th, 2007 at 00:12 | #67

    I have attached the numbers below. Looking at these numbers, can Sinclair, GOE or Andrew answer me this:
    (1)Is the level of GDP achieved in 2049 under the 2.2% scenario the same as that achieved in 2051 under the 2.1% scenario? (This is what Adams and Dixon said)
    (2)How can GOE get a figue of 67.5% of GDP out of this (see #2)?
    (3)Isn’t the sum of the loss of output over this period 198 (equal to 2 years of current GDP) and less than one years GDP in 2051)? Or 2.5% on average?
    (4)If future losses are discounted at, say 6%, won’t this make the proprotionate losses even smaller in discounted present value terms?
    (5)How can you say (as at #58 and #63)that any of these impacts is big in a relative sense?
    (6)If you still claim these losses are big, what is it that I am misinterpreting?

    Growth Status Quo Emissions Reduction Difference % Difference
    2.20% 2.10%
    2006 100 100 0.0 0.0%
    2007 102.2 102.1 -0.1 -0.1%
    2008 104.4 104.2 -0.2 -0.2%
    2009 106.7 106.4 -0.3 -0.3%
    2010 109.1 108.7 -0.4 -0.4%
    2011 111.5 111.0 -0.5 -0.5%
    2012 113.9 113.3 -0.7 -0.6%
    2013 116.5 115.7 -0.8 -0.7%
    2014 119.0 118.1 -0.9 -0.8%
    2015 121.6 120.6 -1.1 -0.9%
    2016 124.3 123.1 -1.2 -1.0%
    2017 127.0 125.7 -1.4 -1.1%
    2018 129.8 128.3 -1.5 -1.2%
    2019 132.7 131.0 -1.7 -1.3%
    2020 135.6 133.8 -1.8 -1.4%
    2021 138.6 136.6 -2.0 -1.5%
    2022 141.6 139.4 -2.2 -1.6%
    2023 144.8 142.4 -2.4 -1.7%
    2024 148.0 145.4 -2.6 -1.7%
    2025 151.2 148.4 -2.8 -1.8%
    2026 154.5 151.5 -3.0 -1.9%
    2027 157.9 154.7 -3.2 -2.0%
    2028 161.4 158.0 -3.4 -2.1%
    2029 165.0 161.3 -3.7 -2.2%
    2030 168.6 164.7 -3.9 -2.3%
    2031 172.3 168.1 -4.2 -2.4%
    2032 176.1 171.7 -4.4 -2.5%
    2033 180.0 175.3 -4.7 -2.6%
    2034 183.9 178.9 -5.0 -2.7%
    2035 188.0 182.7 -5.3 -2.8%
    2036 192.1 186.5 -5.6 -2.9%
    2037 196.3 190.5 -5.9 -3.0%
    2038 200.6 194.5 -6.2 -3.1%
    2039 205.1 198.5 -6.5 -3.2%
    2040 209.6 202.7 -6.9 -3.3%
    2041 214.2 207.0 -7.2 -3.4%
    2042 218.9 211.3 -7.6 -3.5%
    2043 223.7 215.8 -8.0 -3.6%
    2044 228.6 220.3 -8.3 -3.7%
    2045 233.7 224.9 -8.8 -3.7%
    2046 238.8 229.6 -9.2 -3.8%
    2047 244.1 234.5 -9.6 -3.9%
    2048 249.4 239.4 -10.0 -4.0%
    2049 254.9 244.4 -10.5 -4.1%
    2050 260.5 249.5 -11.0 -4.2%
    2051 266.3 254.8 -11.5 -4.3%
    Sum* 7823.1 7625.1 -198.0 -2.5%

    * (zero discount rate)

  68. Sinclair Davidson
    May 20th, 2007 at 08:41 | #68

    The answers to question (1) through (4) are approximately as you suggest. Lets get to question (5). One way of stating that cost to say climate change abatement is going to cost 34% of current GDP over 34 years. [On my spreadsheet I get 34.19%]. So on budget night we might hear the Treasurer say “This program is going to cost $10 billion over five years”. In this instance, assuming a one trillion dollar GDP, we could say “This policy is going to cost $340 Billion over 34 years”. I think that is a big number (even before debating the assumptions that underlie GE modelling etc). I can only speculate why you may think this is a small number. For a start you’re probably more concerned about AGW than I am, and think its money well spent. That is a value judgement that in a pure market situation wouldn’t matter. In a political environment it does.

  69. Ernestine Gross
    May 20th, 2007 at 11:24 | #69

    I understand Sinclair Davidson now agrees with Mark U, namely: “The bottom line is that Adams and Dixon were correct. The loss of GDP is small relative to total GDP whether you look at GDP on a year-by-year basis or, alternatively, add up every year’s GDP and discount it by a “suitable� discount rate�

    However, Sinclair Davidson would call a fraction of about 2/100 ‘large’ while Mark U would call it ‘small’. Nothing is lost if one gets rid off the adjectives, ‘large’ and ‘small’. This leaves the conclusion that Sinclair Davidson now agrees with Mark U, namely: “The bottom line is that Adams and Dixon’s work is correct.�

  70. jquiggin
    May 20th, 2007 at 12:38 | #70

    As regards the $x billion over 4 years, it’s a remarkably unhelpful way of presenting things, being neither an annual flow nor (except for a specifically time-limited program) a PV cost (even disregarding the absence of discounting).

    The supposed justification for doing it this way is that 4 years is the span of the forward estimates. But the real reason for doing it, just as in the case we are debating here, is that it is a way of making small numbers look bigger.

  71. Sinclair Davidson
    May 20th, 2007 at 13:07 | #71

    No. I am not calling the fraction 2/100 big. I’m calling the fraction 34/100 big. In any event if you look at the figures, the average 2.5% pa starts to look very big as we approach 2050, even without discounting.

    Turning John’s last point around, the 2 years growth (or one year) cost is a way of making a big number look small.

  72. melanie
    May 20th, 2007 at 14:58 | #72

    “One way of stating that cost to say climate change abatement is going to cost 34% of current GDP over 34 years.”

    SD, how is that a big number? There are a number of countries around the world that invest that percentage of their GDP in the future of their economy every year, year after year.

  73. Ernestine Gross
    May 20th, 2007 at 15:11 | #73

    Now Sinclair Davidson agrees with Mark U that an appropriate verbal description of the fraction 2/100 is ‘small’.

    I conclude that total agreement has been achieved. That is, Sinclar Davidson agrees with Mark U, namely: “The bottom line is that Adams and Dixon were correct. The loss of GDP is small relative to total GDP whether you look at GDP on a year-by-year basis or, alternatively, add up every year’s GDP and discount it by a “suitableâ€? discount rateâ€?

  74. Sinclair Davidson
    May 20th, 2007 at 15:51 | #74

    Where are you taking the “The bottom line is that …” quote from? I can’t see it before #69.

    Melanie, I don’t even understand what “There are a number of countries around the world that invest that percentage of their GDP in the future of their economy every year, year after year” means. Countries don’t invest. They are not conscious monoliths. Individuals, firms (and sometimes governments) invest. Countries cannot. Having said that, I’d be surprised if any economy had an aggregate savings (or investment) rate of 34% “every year, year after year”. From memory some of the east-Asian tigers got into the 20% range (Paul Krugman wrote of this as the perspiation theory).

    Ernestine – I think you’re trying to place words in my mouth (words in my comments box actually). Nice try. But no. Until I see the contet of 2/100 I’m not saying anything about it. As I said above, I’m calling the 34/100 number big, not the 2/100 number. I’m not even sure how you got the 2/100 number.

  75. Mark U
    May 20th, 2007 at 16:41 | #75


    I am still missing something. Where does the 34/100 come from?

  76. melanie
    May 20th, 2007 at 16:51 | #76

    SD #74, would you be Ok if I said ‘the people of a number of countries’? But it can be aggregated and shown as a percentage of national income. I can also assure you that between 30 and 40% is not uncommon in East Asia. You could fairly easily do the research.

  77. Ernestine Gross
    May 20th, 2007 at 16:52 | #77

    Sinclair – I am definitely not trying to put words in your mouth. I wrote what I concluded from the discussion. I can refine the “2/100” by saying ‘approximately 2/100’. See Mark U’s calculations and statements.

    I also don’t know where 34/100 comes from.

  78. Sinclair Davidson
    May 20th, 2007 at 17:27 | #78

    In #67 Mark shows some calculations. I did the same in excel. If we start with GDP = 100 in 2007. The PV of the difference between 2.2% growth pa and 2.1% growth pa discounted at 6% pa is 34. So the pv cost is 34/100.

    Melanie, in a previous life I used to teach a subject called Asian Capital Markets. I don’t recall national aggregate investment rates at that level year in and year out.

  79. Sinclair Davidson
    May 20th, 2007 at 17:36 | #79

    I found this list. A lot of low-income economies do have high gross fixed investment (this is for business only amd 2006 only).

  80. melanie
    May 20th, 2007 at 17:59 | #80

    The following data are from a dataset downloadable from the World Bank. I just picked the 5 most successful economies in East Asia. Hope the time series are long enough for you. The high numbers are often to be found in the 1980-90s rather than in the earlier stages of growth.

    Investment as a percentage of GDP (unweighted averages)
    Japan 1960-97: 32.3 (low 27.7, high 39.0)
    Korea 1970-99: 31.0 (low 21.3, high 39.9)
    China 1978-99: 37.1 (low 32.5, high 43.3)
    Singapore 1970-99: 39.5 (low 32.8, high 48.5)
    Hong Kong 1970-99: 27.7 (low 20.5, high 35.3)

    Australia 1960-97: 24.6 (I gather it has dropped rather dramatically since 1997)

  81. melanie
    May 20th, 2007 at 18:04 | #81

    SD, possibly you didn’t see such large numbers when you were teaching because most of the investment was not via capital markets, but via state-owned entities such as banks or, in Singapore, the CPF and SGIC (except in Hong Kong). In that sense it was the ‘countries’ that did the investment.

  82. Mark U
    May 20th, 2007 at 20:02 | #82


    If I do what you have said I get the calculations below. The 35.9 (which I assume matches up with your 34) should be looked at as a proportion of 2269.2. That is, we calculate the PV of the loss of output over the 43 year period as a percentage of the PV of total output over the 43 year period. This loss is 1.6%, which is even less than the 2.5% we get with a zero discount rate.

    Discount Rate 6%
    Growth Status Quo Emissions Reduction Difference % Difference
    2.2% 2.1%
    2006 100.0 100.0 0.0 0.0%
    2007 96.4 96.3 -0.1 -0.1%
    2008 93.0 92.8 -0.2 -0.2%
    2009 89.6 89.4 -0.3 -0.3%
    2010 86.4 86.1 -0.3 -0.4%
    2011 83.3 82.9 -0.4 -0.5%
    2012 80.3 79.9 -0.5 -0.6%
    2013 77.4 76.9 -0.5 -0.7%
    2014 74.7 74.1 -0.6 -0.8%
    2015 72.0 71.4 -0.6 -0.9%
    2016 69.4 68.7 -0.7 -1.0%
    2017 66.9 66.2 -0.7 -1.1%
    2018 64.5 63.8 -0.8 -1.2%
    2019 62.2 61.4 -0.8 -1.3%
    2020 60.0 59.2 -0.8 -1.4%
    2021 57.8 57.0 -0.8 -1.5%
    2022 55.8 54.9 -0.9 -1.6%
    2023 53.8 52.9 -0.9 -1.7%
    2024 51.8 50.9 -0.9 -1.7%
    2025 50.0 49.1 -0.9 -1.8%
    2026 48.2 47.2 -0.9 -1.9%
    2027 46.5 45.5 -0.9 -2.0%
    2028 44.8 43.8 -1.0 -2.1%
    2029 43.2 42.2 -1.0 -2.2%
    2030 41.6 40.7 -1.0 -2.3%
    2031 40.1 39.2 -1.0 -2.4%
    2032 38.7 37.7 -1.0 -2.5%
    2033 37.3 36.3 -1.0 -2.6%
    2034 36.0 35.0 -1.0 -2.7%
    2035 34.7 33.7 -1.0 -2.8%
    2036 33.4 32.5 -1.0 -2.9%
    2037 32.2 31.3 -1.0 -3.0%
    2038 31.1 30.1 -1.0 -3.1%
    2039 30.0 29.0 -1.0 -3.2%
    2040 28.9 28.0 -0.9 -3.3%
    2041 27.9 26.9 -0.9 -3.4%
    2042 26.9 25.9 -0.9 -3.5%
    2043 25.9 25.0 -0.9 -3.6%
    2044 25.0 24.1 -0.9 -3.7%
    2045 24.1 23.2 -0.9 -3.7%
    2046 23.2 22.3 -0.9 -3.8%
    2047 22.4 21.5 -0.9 -3.9%
    2048 21.6 20.7 -0.9 -4.0%
    2049 20.8 20.0 -0.9 -4.1%
    2050 20.1 19.2 -0.8 -4.2%
    2051 19.3 18.5 -0.8 -4.3%
    Sum (zero discount rate) 2269.2 2233.4 -35.9 -1.6%

  83. Ernestine Gross
    May 20th, 2007 at 20:03 | #83

    Sinclair, your numbers don’t make sense even if one were to treat GDP as ‘future cash flows’ as understood in Finance.

    You refer to Mark U’s numbers. Using these numbers, the difference in the GDP index over 44 years (2006 = 100 to 2050)is 11 in 2050. The PV of 11 at 2006, using your 6% compound, is 0.897926. As a percentage of the GDP index at 2006, the number is 0.8% and NOT 34%.

    I don’t know why anybody would want to do such calculations though. Obviously, if the maximum projected growth rate of GDP is 2.2% then, applying a discount rate of 6% results in something less than 2.2% and not something bigger.

  84. Sinclair Davidson
    May 21st, 2007 at 06:55 | #84

    Sorry, Mark. I’m not following what you’ve done.

    Ernestine – there is a GDP diferential every year, the sum of the PVs of that differential is 34 (not just the last year). Why would the discount rate be less than a growth rate?

    Melanie, I’ll check my notes from the time. I used to make the students read the World Bank ‘East Asain Miricle’ publication. So I am aware the growth rates were high and capital markets played a very small role. Those numbers are a bit higher than I recall, but you’re probably right.

  85. Ernestine Gross
    May 21st, 2007 at 09:11 | #85

    Sinclair Davidson,

    Yes, there is a projected GDP differential every year and I believe you that the sum of the PV at 6% compound over 44 years is 34, on the GDP index.

    But where does the 34/100 (the ‘big number’) come from?

    I gather you relate 34 to the ‘current GDP'(index = 100). But, surely you would have to relate “34” to the discounted sum of GDPs for 44 years.

    How big is the number when you compare apples with apples? Would you still call it ‘big’?

  86. Sinclair Davidson
    May 21st, 2007 at 09:57 | #86

    That’s exactly were I got the 34/100 from. That is apples with apples. How big is the cost compared to our current GDP is a standard yardstick that everyone understands.

  87. jquiggin
    May 21st, 2007 at 09:58 | #87

    It’s been a long thread, but two points are clear. * If you compare stocks to stocks, or flows to flows, you’ll get a ratio close to 2 per cent.
    * If you compare stocks to flows, you can get any number you like

  88. Mark U
    May 21st, 2007 at 19:06 | #88


    You say you didn’t follow what I have done at #82. All I have done in the second set of numbers at #82 is discount each year’s GDP in #67 by 6% back to 2007 for each growth scenario. The final row is incorrectly labelled – it should say 6% discount rate. This last row shows the sum of discounted GDP in each year over the entire period. This brings out the key point that the sum of the losses in PV terms is 35.9 and this must be compared with the 2269.2, giving 1.6%.

    I can understand using current annual GDP as a yarstick but only to look at annual losses of GDP, not the sum of all annual losses over a 43 year period.

  89. Valuethinker
    May 22nd, 2007 at 17:34 | #89


    You need to differentiate between *metallurgical coal* and *steam coal*.


    according to these numbers, 2/3rds of Australian coal exports are metallurgical coal. $16.8bn v. $7.8bn.

    The latter Australia could give up exporting. But if this led to the burning of less efficient brown coal in the export countries, then it would be a net loss to CO2 reduction. AFAIK Australian steam coal is primarily bituminous ie relatively high grade (which also happens to be the best coal for the IGCC carbon sequestration technology, if it is ever implemented).

    Metallurgical coal is high value anthracite (hard coal with a high carbon content). It is the basis of the primary steel industry (secondary steel, produced by remelting scrap, is big in the US and other developed countries).

    You can’t actually abolish primary steel making nor is there any substitute for metallurgical coal in it. The secondary steel industry (naturally) can’t produce any more steel than is available in scrap from the primary industry, and also secondary steel doesn’t, I don’t think, have quite the structural properties of primary steel (the same problem occurs recycled v. primary aluminium – you don’t use the former in aircraft wings for example).

    Most of those exports go to Japan and Korea, but the steel is then shipped to China. Note also India is a relatively important customer.

    Metallurgical coal is actually in short supply. The original metallurgical coal fields in the UK and many in the US Appalachians are nearly exhausted.

    Cut that supply and the price of steel goes up (potentially a lot). This has all kinds of negative impacts:

    – countries use more concrete/cement (cement is the world’s leading greenhouse gas source, after cars)

    – price of wind power and other alternative energies goes up (they use a lot of steel)

    – more marginal deposits in other countries are extracted at a higher energy, CO2 cost, and at the cost of more workers’ lives

    What the developed world can do is mandate the best coal burning technologies in the recipient countries.

    But I can’t see Oz giving up exporting metallurgical coal. At the very least it would also significantly hurt the *iron* mining industry (since Australia’s other great export to the steel industry, is iron ore).

Comments are closed.