May 2, 2016 by


CREDIT: AP Photo/John Peterson

Microsoft co-founder Bill Gates plays table tennis against prodigy Ariel Hsing, unseen, along with Berkshire Hathaway Chairman and CEO Warren Buffett, unseen, outside the Borsheims jewelry store, a Berkshire Hathaway subsidiary, in Omaha, Neb., Sunday, May 1, 2016.

Bill Gates keeps saying confused and confusing things about climate policy and clean energy.

Gates has positioned himself as a major player and spokesman in this arena with his “Breakthrough Energy Coalition,” a $2 billion effort he is spearheading to research and develop breakthrough “energy miracles.” That means his confused — and often simply wrong — pronouncements could have serious consequences in the public debate and are worth exploring in detail.

Gates’ latest bizarre position, as one media headline explained, is to dismiss a carbon tax — “Bill Gates: Carbon tax not right for the US.”

This is based on an extended interview that Gates did with the magazine Technology Review (TR), which may be his most revealing set of comments to date. When TR naturally asked about Gates whether a price on carbon is needed “to help create a business rationale for investing more in clean technologies?” and “Do you favor a clean, transparent tax?” he replied:

Some countries will do a pure carbon tax, and there’s a certain beauty to doing it that way, but the consensus that I think people will reach here in the U.S. will be to focus more on supply side.

Significantly, Gates explains, “The supply side is about funding basic government research, and then creating startup companies with that research.” For Gates, carbon prices and clean energy tax credits are “demand side” policies because they increase the demand for clean energy, whereas supply side policies in theory increase the supply of clean energy technologies available.

What Gates is saying is that he believes the consensus that Americans will reach is to focus policy more on funding basic research and creating startup companies than on policies to boost demand — what are typically called deployment programs. This is just a restatement of Gates’ view that existing technology is not up to the job and that we need a bunch of energy miracles. That view has no basis in fact (as I’ve explained at length here and here), which is why there is zero chance American policy will focus more on energy miracles than energy deployment.

As I discussed in debunking the recent similarly confused Foreign Affairs piece, we have dawdled for so long on serious climate action that we must rapidly slash CO2 emissions, which requires over 100 times more money be spent on deployment (demand) than R&D into breakthrough technologies (supply). A carbon tax would be a big help in achieving rapid CO2 cuts.

I quoted leading climate expert Ken Caldeira, who explained four years ago, “Globally, deployment costs will be in the trillions of dollars, while R&D costs might be in the tens of billions.” Significantly, Caldeira works with Gates, helping to run the Fund for Innovative Climate and Energy Research, which Gates provided half the money for.

Gates, however, appears to be someone who doesn’t really listen to the advice of experts. This affliction — common to billionaires (I’m looking at you Donald Trump) — shines through from the very beginning of this interview:

Gates Reinvents The Wheel, Or, Rather, The Kaya Identity

Gates conceptualizes the CO2 problem through a formula that he claims credit for, but that in fact — as leading climatologist Michael Mann and others have pointed out — was developed by someone else two decades ago who has his own Wikipedia entry.

Technology Review understands what Gates has done but since the billionaire’s climate policy can’t be explained without this formula, they try to split the baby:

Gates is animated by an equation he claims to have come up with (although it resembles another equation, called the Kaya Identity, well known to climate scientists) that considers total carbon dioxide emissions as the product of the factors P (population), S (services consumed per person), E (the energy used to supply those services), and C (the amount of carbon emitted per unit of energy).


One of the dangers of reinventing the wheel, of course, is that you may not realize that one of the features of the original design — like its roundness — is actually central to its value. So you might instead come up with a square wheel. That’s kind of what Gates does here.

Gates explains to Technology Review how to use this formula to figure out the implications of a global effort to take CO2 levels down to zero:

If you take my equation and look at the first three factors, P is going to go up by about 1.2, S is going to go up by about a factor of 2, and E, assuming lots of breakthroughs — well, we’ll be generous and say that’s 0.6. So basically, when you summarize the equation, it suggests that C must be about zero.

The bold-face sentence is incorrect. Gates is saying that to get “Energy Per Service” to 0.6 — a 40 percent drop — would require “lots of breakthroughs.” In fact, it is going to happen without any breakthroughs at all!

That might be clearer to Gates if he used the original Kaya Identity, where the term is a Energy Per GDP, which is easy to quantify and indeed for which there is a vast historical dataset for the world and most countries. Gates never formally defines “Energy Per Service,” which kind of makes it a hand-waving term not so useful in a mathematical equation. Indeed, Gates himself is somewhat confused about the term, as is clear from his Annual Letter, where he writes:

Let’s take a look at (E). That’s the energy needed per service. There’s some good news here. Fuel-efficient cars, LED light bulbs, and other inventions are making it possible to use energy more efficiently…. All of these efforts help cut down on energy use.

Unfortunately, they don’t get us to zero. In fact, most scientists agree that by 2050 we’ll be using 50 percent more energy than we do today.

Uhh, no. Although these paragraphs are supposed to be about E — “energy needed per service” — the final bold-faced line is just about total Energy! Yes, scientists agree that we will be using 50 percent more energy globally by 2050 — but NOT because “energy needed per service” is going up! It’s because Population (P) and Services Per Person (S) are going up.

The whole point of this equation is to disaggregate the effects of overall growth in the population and economy (the P and S terms) from what we can do with technology — namely improve energy efficiency (E) or reduce the carbon intensity of energy sources (C). But Gates then either forgets or ignores or confuses the whole point by leaping to total energy in 2050 to somehow argue that it all doesn’t matter.

The original Kaya Identity simply uses Energy Per GDP (aka energy intensity), and once you do that you have access to reams of data. “Since 1973, energy intensity has declined at a rate closer to 2% per year,” as the US Energy Information Administration explained in a detailed analysis three years ago. “The 2013 Annual Energy Outlook Reference case projects that this average annual decline of 2% will continue through 2040.”

In short, we have every reason to believe that from 2015 to 2050, the energy intensity of the U.S. economy will drop 50 percent, without any miracles or indeed without any major push on energy efficiency — since the EIA doesn’t model any climate or energy policies that don’t exist today. Can you imagine how fast energy intensity or energy per GDP could drop if we actually had a national push on efficiency?

After Gates (incorrectly) asserts that a 40 percent drop in E by 2050 would require “lots of breakthroughs” — basically dismissing/ignoring energy efficiency as a strategy — he puts the entire burden of the climate technology effort on finding zero-carbon energy (generation) miracles: “C must be about zero.”

Immediately after that, TR naturally asks him when must C be zero. The correct answer is — for the entire world in a typical 2C pathway — the end of the century. You can see that in this typical pathway from Climate Interactive, one of the world’s leading organizations with energy and CO2 models that governments and others rely on for making decisions about what can plausibly be achieved through various energy technology options and policies.


But — and here is where it gets truly confusing — Gates apparently wants to make the problem seem harder than it is. So he says “rich countries need to be net zero by 2050,” and TR uses that as a pull quote, as if even that claim were true, which it isn’t:


To be clear, if you just focus on what rich countries need to do, then, as Climate Interactive explains in analysis of a 1.8°C “Ratchet” Success Scenario — the correct answer is that the U.S. and EU and other developed countries need to be “80% below 2005 [levels] by 2050].”

And an 80 percent drop in C is not the same thing as a 100 percent drop to zero. But Gates’ confusion about what the industrialized countries have to do is even bigger. That’s because the Kaya Identity and its poor cousin, the Gates equation, apply either globally — or to whatever set of countries you are talking about. Let’s go back to Gates’ original choice of values for the variables in his equation:

If you take my equation and look at the first three factors, P is going to go up by about 1.2, S is going to go up by about a factor of 2, and E, assuming lots of breakthroughs—well, we’ll be generous and say that’s 0.6. So basically, when you summarize the equation, it suggests that C must be about zero.

Now Gates’ statement that “S is going to go up by about a factor of 2” is true for the whole world, but S (services consumed per person) — best thought of as GDP per capita — is projected to go up 80 percent by 2050. So we should use 1.8 instead.

Let’s use the no-miracle estimate for E of 0.5 (a 50 percent drop in energy intensity) in the rich countries by 2050. That means to get an 80 percent drop in CO2 levels, the industrialized world needs slightly more than an 80 percent drop in carbon intensity of its energy sources by 2050. (If we had a major effort by the rich countries to push energy efficiency, we might only need a 70 percent drop in the carbon intensity of energy sources.)

In any case, no miracles are required. Ironically, if Gates were actually correct that we needed the entire industrialized world to have entirely carbon-free energy sources in 2050, that would only make a stronger case against Gates’ claim that we should focus on energy miracles rather than deployment.

To paraphrase what I wrote last week, I hope readers see how absurd it is to assert that the rich countries could plausibly expect to zero out CO2 by 2050 using technologies that do not exist today. The reality of the climate challenge is that only technologies that can be deployed at trillion-dollar scale in the next three decades can contribute to such a goal. But the technology development and deployment cycle is simply far too long for a technology that doesn’t exist today to plausibly make a vital contribution to cutting CO2 by mid-century.

Indeed, Climate Interactive has put together a (very optimistic) scenario for new technology that makes this point. If we had breakthrough technology miraculously pop out of the R&D pipeline really fast (by 2020) and then miraculously enter the market at half the price of new coal plants and then miraculously achieve mass penetration decades faster than every other major energy source in history — it still has essentially no impact on emissions reduction by 2050:


A wildly optimistic scenario of the impact of a breakthrough energy technology (gray line on left) on one business-as-usual warming scenario. The breakthrough occurs in 2020; leads to a practical, scalable new energy source half the price of new coal; and it is adopted by the marketplace decades faster than any other major energy source in history. Even so, it has virtually no impact on CO2 emissions by mid-century and only a very modest impact on temperatures (figure on right) by 2100.

BOTTOM LINE: Gates is just wrong about everything here. He is wrong that energy miracles are needed by the industrialized countries to achieve CO2 levels in 2050 consistent with beating the 2°C target. He is wrong that achieving that target requires focusing on R&D rather than deployment. He is wrong that there is some sort of consensus to that effect. He is wrong that a carbon price isn’t important in achieving the rapid reduction the rich countries need. He is wrong to make it seem like boosting energy efficiency is not as vital a strategy as reducing carbon intensity.

Leave a Reply

Your email address will not be published. Required fields are marked *