McKinsey 2008 Research in Review: Stabilizing at 450 ppm has a net cost near zero.

The McKinsey Global Institute has done some of the most comprehensive and credible recent analyses on energy efficiency potential and carbon mitigation cost curves (see “Must read McKinsey report shatters myths on cost of curbing climate change“). They have summarzed their work in “2008 Research in Review,” so this is a good opportunity to create one universal link for their work.

One core MGI factoid you can use: Nearly 40% of the U.S. emissions reduction potential by 2030 is from energy efficiency (see here).

MGI is best known for its comprehensive cost curve for global greenhouse gas reduction measures (click to enlarge), which concluded measures needed to stabilize emissions at 450 ppm have a net cost near zero — the same conclusion as the International Energy Agency and IPCC.


Another 2008 MGI report has its own stunning conclusion:

The macroeconomic costs of this carbon revolution are likely to be manageable, being in the order of 0.6–1.4 percent of global GDP by 2030. To put this figure in perspective, if one were to view this spending as a form of insurance against potential damage due to climate change, it might be relevant to compare it to global spending on insurance, which was 3.3 percent of GDP in 2005. Borrowing could potentially finance many of the costs, thereby effectively limiting the impact on near-term GDP growth. In fact, depending on how new low-carbon infrastructure is financed, the transition to a low-carbon economy may increase annual GDP growth in many countries.

The new analysis explains that “at a global, macroeconomic level, the costs of transitioning to a low-carbon economy are not, in an economic ‘welfare’ sense, all that daunting — even with currently known technologies.” Indeed, 70% of the total 2030 emissions reduction potential (below $60 a ton of CO2 equivalent) is “not dependent on new technology.”

The report notes that “we have been fairly conservative in our assumptions about technological progress in these projections.” For instance, the analysis appears to ignore the potential of concentrated solar thermal electricity entirely (see “Concentrated solar thermal power Solar Baseload — a core climate solution“).


On U.S. efficiency potential, “McKinsey worked with leading companies, industry experts, academics, and environmental NGOs to develop a detailed, consistent fact base estimating costs and potentials of … more than 250 options, encompassing efficiency gains, shifts to lower-carbon energy sources, and expanded carbon sinks” (see “McKinsey: Fighting climate change is affordable“).

One key conclusion is well worth bearing in mind as a new Administration (hopefully) contemplate banning all new coal plants that do not capture and store most of their carbon:

Improving energy efficiency in the buildings-and-appliances and industrial sectors, for example, could (assuming substantial barriers can be addressed) offset some 85% of the projected incremental demand for electricity in 2030, largely negating the need for the incremental coal-fired power plants assumed in the government reference case.”

Again, no breakthroughs are needed:

The United States could reduce GHG emissions in 2030 by 3.0 to 4.5 gigatons of CO2e using tested approaches and high-potential emerging technologies. These reductions would involve pursuing a wide array of abatement options with marginal costs less than $50 per ton, with the average net cost to the economy being far lower if the nation can capture sizable gains from energy efficiency. Achieving these reductions at the lowest cost to the economy, however, will require strong, coordinated, economy-wide action that begins in the near future.

Yes — existing or in-the-pipeline technology can get us very far for the next quarter century (duh and more duh).

Here is McKinsey’s own summary pointers to its recent work:

Using carbon and energy resources more productively

With the global financial system in crisis and the economy in a downturn, there is a risk of ebbing momentum for investing in clean energy and tackling climate change. Some argue that action is too expensive when the economy is weak; others say it will hurt economic growth and force consumers to make unwanted changes in their lifestyle. Our joint research with McKinsey’s Climate Change Special Initiative found that we can dramatically reduce greenhouse gases and grow the global economy for less than we think. Meanwhile, our work on global energy demand trends examined the highly attractive economics of investing in energy productivity–the level of output we achieve from the energy we consume. Other research examined how Europe can further capture its energy productivity opportunity and explored energy demand trends in developing countries. We also identified areas for action to overcome today’s barriers and highlighted how companies can not only make large energy-cost savings but also create lucrative new markets in energy-efficient technologies and services. Look for an updated perspective on energy supply and demand trends and oil prices early next year.

Kudo to McKinsey Global Institute for all of this important research.

Energy efficiency overview:

23 Responses to McKinsey 2008 Research in Review: Stabilizing at 450 ppm has a net cost near zero.

  1. Larry Coleman says:

    Joe, you say that “the analysis appears to ignore the potential of concentrated solar thermal electricity entirely,” but CSP is included on the graph of abatement options (at about $45/ton CO2).
    I was surprised to see automotive hybrids coming in at more than $50/ton. I suppose that’s because, although my Prius did not cost that much more than a comparable non-hybrid, nor does it reduce CO2 output by very many tons of CO2 per year.

  2. Ben says:

    Great news in this article. A late xmas present for all of us climate sinners.

    Larry, good analysis.

  3. john says:

    Part of the reason for the PHEV appearing to cost so much is the quetsion of how much more it costs over a traditional ICE, and that increment has been vastly overblown.

    For example, in assessing how much more a hybrid cost, the 2004 Prius was compared to a Corolla, and based on that, it was assigned a incremental cost of more than $5000 dollars.

    However, the 2004 Corolla is not comparable to a Prius — it did not have a touch screen; it did not have as much passenger room; it did not have the range of options a Prius had.

    In reality, the Prius was mid-way between a Corolla and a Camry, which suggests the Prius’ incremental cost was more like $2000.

    At this increment, the abatement costs assigned to a Hybrid would be much less.

    As McKinsey says, they were quite conservative in terms of technological advances.

    You can bet that battery costs will drop, as battery performance improves, and that will make PHEVS a much better deal for carbon abatement — I would predict carbon costs for PHEVs would be less $25 by 2018

  4. Modesty says:

    Why do these conservative estimates get to own the graphics space, though?

  5. Joe says:

    Sorry, Larry, I just don’t see the CSP.

  6. Peter Wood says:

    The 27 Gt CO2-e of abatement is equivalent to 7.4 Gt C, i.e. 7.4 wedges. If technologies such as CSP or geothermal can also be deployed, then that could mean more wedges. Conversely, if the baseline includes less growth (i.e. due to the financial crisis), less wedges are needed. This suggests that a stabilisation target of less than 450 could be achieved at low cost.

    Interestingly, if McKinsey was right, and if we set the carbon price at $50 now, we will get more abatement, provided we also address the market failures (e.g. in energy efficiency) that need to be addressed to realise the negative cost abatement on the left hand side of the cost curve.

  7. Joe says:


    Note that MGI curve is for 2030. We’ve got a lot more to do by 2050. And then 2070. And then 2090.

  8. Jim Bullis says:

    Sorry Joe, but when you delete a serious question it makes the search far more compelling.

  9. Larry Coleman says:

    Joe, to see the CSP I went to the McKinsey link you supplied and looked at the graph there. It has more detailed annotation than the one you posted. Specifically, it is Exhibit B of the Executive Summary, which is at:

    [JR: Thanks. That’s a trivial amount. So they almost completely ignore CSP.]

  10. Jim Bullis says:


    My apologies for thinking you had deleted my comment re McKinsey chart, which I actually left on the old post about this.

    My comment which is appropriate here I repeat:
    Once more I went on the wild goose chase to find the basis of the oft shown chart by McKinsey.

    The report I found offered nothing in support of this chart you call a “cost curve.”

    I did find that McKinsey combines “management and economics” to find global solutions etc. Huh? Seems like they left out technical expertise. No wonder there is nothing more to the chart.

    Only about a third of the specific bars on the chart carry a notation. And these notations are not discussed anywhere that I can find.

    Has anyone actually found the detail on this?

    End of paste.

    I will now chase again given the Larry Coleman link.

  11. Jim Bullis says:

    Ok, Larry Coleman gives me faith that McKinsey knows the labels for most of their bars.

    On to the next question: How can we find out what the labels actually mean? For starters, two big bars tell us that home electronics and commercial electronics can save us lots of CO2 and that there will be a huge cost benefit per Giga– of CO2. How does this happen? Do we stop watching football?

  12. Jim Bullis says:

    The chart that the Larry Coleman link hooks to is different than the one above. Huh? Notice the one above fails to identify the first big bar. (The one that bans football watching on TV.)

    The chart on this present blog, though missing a lot of bar identifications, can be made comprehensible (sort of) if we take their “conservative” qualifyer to mean that they apply their judgment about likelihood that the specified actions will be taken.

    For example, the gain from converting from coal to natural gas could be a bar about as wide as a third of the chart, but in the judgment of the chart makers this is not going to happen.

    Similarly the gain from us all driving Prius cars could be about a fifth of the chart, but again, the McKinsey wisdom says it is not going to happen. Maybe this shows that the more likely actions by the US carmakers in electrifying cars will not matter much, which is a conclusion I also came to.

    McKinsey also sheds doubt on themselves by giving great expectations of progress from new coal CCS and coal CCS retrofit, which are solutions having at best a doubtful future.

    McKinsey needs to do some real work here.

  13. Larry Coleman says:

    The gains in home electronics as well as the other categories are described in some detail in the full report, which is where one would expect it, at:

    The gains in electronics are from improved efficiency of the increasing numbers of computers, tvs, etc., expected to be sold, and other approaches

  14. Peter Wood says:


    Note that MGI curve is for 2030. We’ve got a lot more to do by 2050. And then 2070. And then 2090.

    Agreed. This is especially the case if we initially overshoot a particular stabilisation target, which is likely, given the political situation at the moment.

  15. john says:

    I had the opportunity to review and comment upon some of the McKinsey cost curves and I think one of the most important observations they make is this:

    “Borrowing could potentially finance many of the costs, thereby effectively limiting the impact on near-term GDP growth. In fact, depending on how new low-carbon infrastructure is financed, the transition to a low-carbon economy may increase annual GDP growth in many countries.”

    Like any I/O or econometric model, there are myriad assumptions embedded within it. If we use smarter policies, we will get more carbon than projected, AT A LOWER COST.

    The McKinsey findings are big — but the acknowledgment that smart, creative policies could get us even more carbon for less money — and that fiscal and financial strategies are at the core of getting more for less — is even bigger.

  16. David B. Benson says:

    I’ll go further. It may be there is monetary profit to be had while causing such a stabilization to occur. Irrespective of emission scenarios.

  17. Jörg Haas says:

    Don’t try to figure out too much about the presented version of the cost curve. A new, completely revised and updated version is in the making and will be released in Brussels on January 26. The publication will contain much more detail than the present version. Stay tuned.

  18. Jim Bullis says:

    Thank you Jorg Haas.

    I look forward to reading a more substantial analysis.

  19. Bob Wallace says:

    Early posters – please don’t confuse a Prius with a PHEV. The Prius packs only enough batteries for a half-dozen miles or so on batteries alone. A true PHEV would be expected to give a 30+ miles range per charge.

    Increasing the battery pack to get from a Prius to a PHEV would be quite expensive given today’s battery prices, but doing so would greatly reduce the amount of petroleum burned. And if, as expected, an ever larger amount of the replacement electricity comes from green sources a lot of carbon would remain under the ground where it belongs.

  20. Bob Wallace,

    I appreciate the way you state the difference between a PHEV and a Prius. Thus, you avoid confusing the two different goals of (1) cutting use of foreign oil and (2)reducing the CO2 emissions. While these goals could be substantially complementary, the move to propel cars with electricity has serious consequences if your “expected” green sources do not pan out. At least you clearly state the condition whereby things could be ok.

    For those of us who begin with the world as it is and form our expectations based on economic and political reality, it seems best to not get carried away with the plug-in until the “green” conditions come to pass.

    It is anticipated that the use of electricity in the USA will increase significantly by 2030 and that 81% of the incremental load will be filled by coal fired generation. This is a DOE projection as reported in the McKinsey analysis report discussed above. Conversion of heat from coal to electric energy was 33% efficient in the USA in 2005 (see

    If anyone is interested in a definite analysis, the reference case of a production Prius versus a specific conversion is an interesting basis for discussion. The production Prius has a measured engine efficiency of 38% whereas as a Hymotion conversion reduces the efficiency of this engine to 32%. (These are measures by Argonne for a “UDDS” driving cycle.) Further, coal produces more CO2 per BTU than gasoline. Therefore, the effect of converting to plug-in operation will be to substantially increase the emitted CO2 whether it is operating on the battery or the engine.

  21. Everette Carnes says:

    Not being an expert like everyone else I form an opinion based on what I think is so. I do not believe we can achieve energy independence at the same or a reduced cost. It will cost more, perhaps much more, to satisfy our hungers with non-polluting forms of energy than it has with fossil fuels. Or perhaps I am wrong and we can do it for less. We do not have time to debate that point. We can either pay the cost of making the fastest possible transition to new energy sources, whatever the cost turns out to be, or we can learn to live in deeper water and breath thicker air.

  22. Jörg Haas says:

    The new version of the McKinsey Cost Curve has now been published. Download at

  23. Mike says:

    I found this article while looking for an actual listing of the “250 abatement opportunities”..Seems like they could have been more generous with the details. Still looking.

    Anyway, I don’t believe their numbers for CSP. The slight positive cost versus polluting technology is OK, but the amount of potential is ridiculous. Solar Base Load can go to any scale that is needed, and everything I’ve found so far points to a penny or two per kwh to move it 4500 km.

    What else did they get wrong?