2018-11-30

New technologies, not Paris climate agreement, will do the job?

13403371_1121168774614721_7270794426077298794_o Or so says Dieter Helm in the FT (arch).Someone called Simon Evans doesn't like it, and although he says it is packed full of wrongness, in an entire Twit thread he lands few if any hits. But I think DH gets some major points right; principally that the entire Kyoto-Paris-etc negotiating theatre is a waste of time and money; and that the response to GW so far, throwing money at subsidies, has been inefficient. What he mostly says is the very limited amount of money that current customers and voters are actually prepared to pay is spent wisely. The money needs to go on those things that might actually make a real difference. I don't think SE likes the "very limited" aspect, but otherwise, the call to spent money wisely should hardly be controversial2; though when phrased as a criticism of the way money is currently being spent it does become so, since it implies that "the" money is currently being spent unwisely. Which I agree with.

But sadly, as soon as DT gets to solutions, he falls into exactly the same trap as the people he is criticising: Instead of putting all the money in the conventional wind and solar panels boxes, some of it should go on research and development. Did you spot the error? The chances are that if you're a researcher you didn't :-). The error is once again trying a top-down directed approach to what should be done. The answer is a carbon tax, and let the market sort things out. Alas people like DT like that idea no more than people like SE like it, because it gives them nothing to pontificate about.

Fortunately there seems to be a genuine chance that solar will spread quickly enough so solve all our problems, despite the general incompetence. And people are currently doing a really bad job on working out the costs of GW1, so this may be for the best. Although effectively saying "don't worry, all will be well" seems a poor plan, given past experience.

Refs


Carbon budgets and carbon taxes.
* It took me ages to find this so I'll put it here, but its NSFW: Oglaf/Intermission.
What made solar panels so cheap? Thank government policy?
As Congress Tackles Climate, Markets Are The Engine But Policies Set The Direction - Forbes

Notes


1. As ATTP notes in "10% of GDP?"
2. Though now I think about it, I'm actually opposing attempts to spend money "wisely", if "wisely" is read as "after careful consideration by some central authority". But I am arguing for trusting to the "wisdom" or perhaps knowledge of individuals and smaller entities.

24 comments:

Phil Hays said...

Subsidies are making solar volumes increase and driving private research and development.

https://www.sciencedirect.com/science/article/pii/S0301421518305196/pdfft?md5=76728a988c2db77621a3e48aefbeec5e&pid=1-s2.0-S0301421518305196-main.pdf

Solar is spreading fast because of subsidies.

So subsidies are inefficient.
.
.
.
Right. Makes ... perfect ... sense. Well, not really. Not any sense at all.

PaulS said...

Not going to bother reading the article but, safe in the knowledge that you accurately represent his views, I will declare them ridiculous and wrongheaded.

Firstly, nothing in the Paris Accord precludes research into new technologies. In fact, the UNFCCC published this document briefing on the important role of technological innovation within the Paris commitments, stressing that such innovation was vital for meeting the specified goals. So, he is presenting a false dichotomy and a false representation of the Paris agreement.

Secondly, as Phil Hays says, subsidies incentivise private research and development.

William Connolley said...

> https://ac.els-cdn.com/S0301421518305196/1-s2.0-S0301421518305196-main.pdf?_tid=6d8b4f9f-1402-4775-9791-e85484cd8f7b&acdnat=1543686754_eaad0516c30a1071717af77d476bb315

Thank you for the direct link. I do hope you haven't forgotten this discussion.

As to the subsidies, I fail to follow your argument. We know that solar has been subsidised, and we know that it is growing rapidly. But we have no real idea at all whether subsidies were the most efficient method to achieve that growth.

> nothing in the Paris Accord precludes research into new technologies

Of course this is true. How could it be otherwise? But the article is more about priorities: where to focus our efforts. And indeed this is a theme of economics: given limited resources, how best to "allocate" them?

PaulS said...

But we have no real idea at all whether subsidies were the most efficient method to achieve that growth.

Which means for all we know they may have been the most efficient possible method. So how can you claim 'throwing money at subsidies has been inefficient'

But the article is more about priorities: where to focus our efforts.

The title says 'New technologies, not Paris agreement' - that's representing the two things in opposition. But, in any case, nothing in the Paris agreement precludes prioritising technological research.

Phil Hays said...


Oh yes, I remember that you failed to understand why Pigovian subsidies can be produce a smaller impact to the economy to Pigovian taxes in the case of very low volume for the alternative.

Start with the absurd case. One person is willing to try an alternative to a polluting technology. 5 billion people are not. Is it not easier and more efficient to subsidize the one person, than to tax then rebate on the 5 billion?

At nearly equal volumes of between the incumbent technology and the alternatives, the tax is more efficient.

William Connolley said...

> for all we know they may have been

Well, we could guess. Or we could look at how well subsidies usually work; or ask economists, who are the people that study this kind of stuff. People are very keen that climate scientists should be asked about climate science, but curiously reluctant to ask economists about the economy.

> Start with the absurd case

The problem is that your case is indeed absurd, so teaches us nothing. Why not start with a non-absurd case: people are sensitive to the prices of things.

Sam said...

Helm is consistently pretty bad on technology. Especially that bit where he goes on about "The world in which we have to tackle climate change is made up of robots, 3D printing, artificial intelligence and with this comes enormous flexibility in energy consumption.", which is basically a line he's been repeating in one form or another for the last 3 or 4 years with very little to really back it up.

For one thing, 3D printing has already completely died on it's arse, outside of a few niche use cases. AI (and I'm assuming he knows very little about neural networks, here) seems likely to massively increase energy demand, since it needs to be run on extremely large server farms (such as the one I sit about 10 meters away from) which eat lots of energy, and which like their power supplies to be consistent. See also robots. Battery tech really isn't doing anything special either. At least he's not a blockchain evangelist I suppose.

Which is to say, it's all well and good if technology does come to the rescue, but you do kind of need to think about what'll happen if these things don't actually work. I'm quite firmly a tech-pessimist

Phil Hays said...

People are indeed sensitive to the prices of things.

Government can change the prices of things several ways. Let us look at a real problem. Transportation is a significant source of carbon dioxide pollution. Like anything, we could just use less, or change technologies to use the same at a lower amount of pollution.

Electric cars are about 2% of the US market by sales. The sales have been rising rapidly partly due to the subsidies. As cars have a useful lifetime of over a decade, a much lower fraction of cars on the road are electric. So the subsidy impacts the consumer at the point of buying a new vehicle, and the tax impacts owning a vehicle over years. So how much carbon tax would be needed to have the same impact on the real cost of ownership of a new vehicle bought today as the US Federal $7500 subsidy does today?

How much money would that carbon tax on transportation fuels collect vs how much money would the subsidy cost?

If we want to have a minimum sized government, which choice would we select?

William Connolley said...

> how much carbon tax would be needed

I don't know. You certainly can't do it by pure maths; you need to factor in the problem that many EC drivers may not even care greatly; I see many Tesla-head posts that imply as much.

But the great thing about the carbon tax is that you don't have to keep making these fiddly little decisions for every little thing that you can think about. You do it once, for everything. OK, "once" is an exaggeration, because you'll want to keep nudging the tax up, but it's definitely a lot better than going tech-by-tech and getting snarled up in a pile of special interests. Which is why the answer to "minimum sized government, which choice would we select?" is "a carbon tax, of course, as previously discussed".

Phil Hays said...


He who refuses to do arithmetic is doomed to talk nonsense.

John McCarthy

William Connolley said...

As previously discussed, that quote is more subtle than you think.

Phil Hays said...

There are subtle points, but 100:1 ratio of numbers isn't subtle.

William Connolley said...

I don't know where you get the 100:1 from, but the point is, that arithmetic in the way you've done it is not the point. Taxing fuel incentives people with cars to make a particular journey by bus or walk or not make the journey. It makes them think about fuel costs when changing car. And so on. Whereas subsidising EC's only affects those who can pay the high price for a subsidised EC. So anyone who refuses to do arithmetic is indeed doomed; but so is anyone who insists on doing the wrong arithmetic.

Phil Hays said...


Why do you say EVs are high priced?

They were yesterday, but yesterday is gone.

https://assets.bwbx.io/images/users/iqjWHBFdfxIU/im5dj3BPnJWA/v1/-1x-1.png

What about the future?

https://www.bloomberg.com/news/articles/2018-03-22/electric-cars-may-be-cheaper-than-gas-guzzlers-in-seven-years

Other people say 10 years, 5 years and 12 years. And that is selling price, as electric power is cheaper than gasoline almost everywhere and as maintenance on an EV is less than on an ICE, for some people EVs are slightly cheaper than ICEs now, depending on details.

Don't think that I'm arguing against a carbon tax. Carbon tax is worthwhile now, and will be more worthwhile in the future. Subsidies become less efficient as the market share rises. But right now, subsidies are more than 100 times more efficient in creating a price difference for new technologies such as EVs and solar PV.

William Connolley said...

Errm your link is to a graph of battery prices. Not EC prices.

Phil Hays said...

Battery prices used to be a huge fraction of EV prices. See second graph in second reference.

https://assets.bwbx.io/images/users/iqjWHBFdfxIU/i4Tt7LfOeNGo/v1/-1x-1.jpg

And notice this is selling price, not TCO.

crandles said...

Phil Hays wrote

"Start with the absurd case. One person is willing to try an alternative to a polluting technology. 5 billion people are not. Is it not easier and more efficient to subsidize the one person, than to tax then rebate on the 5 billion?"

Setting up a new taxation system for 5 billion people would indeed be expensive. There certainly is something in the idea that the more people that have to be dealt with the more the administration costs. Typically taxes cost about 4% of amounts collected to administer however there is quite a wide range eg petroleum revenue tax is one of the cheapest to collect as large amounts from small number of oil companies. VAT is one of the most expensive.

That may seem at first to back up what Phil is saying. However, you should note that it is easy to find oil companies. With your absurd case you would have to set up your subsidy system then communicate its availability to a very large number of people in order to find the one person willing to take it up. If you know who the one person is in advance then perhaps that would be cheap to administer but that seems unrealistic. I am not sure that could be done for anything remotely as cheap as 4% of the amount involved. With a taxation system, people try to understand their taxes and how to minimise them so this does a lot of the communicate problem.

If there is not much in it either way but at some point you have to eliminate the subsidy system and change to a taxation system with all the problems of such a changeover, isn't there quite a bit to be said for starting with a system that won't need to be changed?

William, in this case, it seems preferable to use the absurd case against Phil Hays than just dismiss it as absurd.

Phil Hays said...

What an amusing straw man, crandles. Of the five billions, very few would even vaguely be interesting in a brand new technology, and the few interested are likely fairly easy to find. If the cost of getting the first candidate is much higher than the subsidy, then raise the amount of subsidy. In practice getting the first candidate(s) for a new technology usually isn't a problem. Look at electric cars before the EV1, for example. Or for that matter, drivers of EV1s.

http://www.ev1.org/

Personal disclosure. I had the privilege of driving pre-production serial numbers 6 and 8 of an electric car as part of my work a decade ago for a major supplier to the automotive industry. I'm still under a NDA, not much I should say.

You do, however, have a point that the switch between tiny volumes and a subsidy and high volumes with a carbon tax has the potential to create problems of all sorts. With EVs the need for subsidies is rapidly declining as the production ramps up and the costs go down, mostly due to increasing production volume. A carbon tax at this point would do little to incentivize sales. Yet a carbon tax is needed longer term, but not for new technology adoption. The biggest selling problem is the sticker price, not the total cost of ownership. A carbon tax improves the comparison between an ICE and an EV for TCO, but not for MSRP. Ending of subsidies will hurt the MSRP comparison.


Spend some time behind the wheel of an electric. Find out why Tesla is eating the lunch of BMW, Mercedes, Audi and Porsche sales in the USA. BMAP and friends are bringing out electrics... They would like to eat lunch again someday soon. Or at the other end, commute for a few weeks in a Nissan LEAF or a Volkswagen e-Golf. Sure, can't do 0-60MPH(100kph) in less than 3 seconds, but quiet, smooth and convenient and very well behaved. The realization that you will never have to stop for gas as you start each day with a full battery may well sell you. No oil changes. No greasy spot on the garage floor. And don't let the spouse drive it, you may never get a chance again.

https://www.caranddriver.com/reviews/2015-tesla-model-s-p90d-test-review
https://www.caranddriver.com/volkswagen/e-golf

EVs are currently mostly production limited, you will not see many TV advertisements as they can usually sell all they can make. Wait times and available vehicles vary by market, a few months isn't uncommon. In the USA subsidies are phasing out (if the fossil fueled Republicans don't end them sooner), but market is might be beyond the point of caring. And not to replace with a carbon tax, as global warming is a Chinese plot.

https://www.reuters.com/article/us-usa-trump-autos/white-house-seeks-to-end-subsidies-for-electric-cars-renewables-idUSKBN1O22D4

Subsidies worked, are working, and are probably the best policy choice for starting up needed new technologies. The future beyond the next decade is different, subsidies for EVs and renewable power need to be phased out, and should ideally be replaced with a carbon price. Ideally a market price, but a tax adjusted sanely wouldn't be that different. Of course, sane and politics don't often go well together.

rconnor said...

> “people are sensitive to the prices of things”

Depends on the context.

Personal car driving habits and fuel prices - low price elasticity.

Car purchase decision and upfront cost – higher price elasticity.

Home heating and cooling loads and fuel prices – low price elasticity.

Utility generation portfolio and capital cost of new generation technologies – higher price elasticity.

A carbon tax would impact gas prices directly and capital cost of new gen indirectly. Subsidies, whether direct on the capital purchase or in R&D that can be used to drive down capital costs, impacts car costs and new gen costs directly and gas prices and fuel (gas/electric) prices indirectly. In other words, a carbon tax (mainly) hits on areas where there is predominately low price elasticity and subsidies can hit on areas where there is higher price elasticity.

This is not to say that a carbon tax is not a smart option. It’s a necessary correction to a previous market failure of hiding externalities and works really well when combined with other mitigation strategies (like subsidies). We can (and should) do both, ideally where we fund the subsidies through the revenue from the carbon tax.

However, this is to say that the idea that we can slap in a carbon tax and let the market sort out the rest is misguided. Which kinda goes back to my recurring question to you WMC, “carbon tax…and then what?”.

Russell Seitz said...

If Phil want to reinvent himself as an easy credit Tesla salesman, le Chef Frog is hiring--

The Gilet Jaune tide could be stemmed by offering the irate drivers EV's that run on the clean, green French nuclear grid instead of devouring diesel.

M. Macron had better do something soon- his approval rating has fallen ten points lower than Trumps !

Silly Dreamer said...

If there is one thing I am keen on, its a carbon tax. And pretty well precisely for the reasons William Connolly says.

I get continually frustrated by the (seemingly) willful misunderstanding of how a carbon tax works.

Firstly, it can be done in a revenue neutral way - every cent collected can be given back as a reduction in general taxation without reducing the benefit by one iota.

Secondly the main benefits of a carbon tax flow not from the changes induced in individual behaviour, but rather through the changes in the behaviour of power companies, manufacturers etc. If a power generator can see the cost of power from a coal fired power plant getting more expensive every year, they won't invest in one, and will look to wind/solar/nuclear/hydro instead. If a car maker can see that 10 years from now petrol powered cars will be so expensive that consumers will shun them, they'll start making electric vehicles now.

But I'm not totally trusting of markets, and do favour a bit of regulation if needed.

Phil Hays said...

"Firstly, it can be done in a revenue neutral way - every cent collected can be given back as a reduction in general taxation without reducing the benefit by one iota."

Yes. Like France did. Cut taxes on the wealthy, then added carbon taxes which are a larger impact on lower income people to reduce the deficit. Understand why that might get the lower income people upset?

https://www.vox.com/energy-and-environment/2018/7/20/17584376/carbon-tax-congress-republicans-cost-economy

"A carbon tax is, in and of itself, somewhat regressive. It hits the poor harder than the rich because the poor spend a larger percentage of their income on energy services."

"Carbon taxes are a useful tool but a dangerous fetish

To me, all of the above suggests a simple conclusion: Carbon taxes are good policy, an important part of the portfolio, but unlikely ever to be sufficient on their own. It’s worth getting a price on carbon anywhere it can be gotten, but climate hawks should not believe, and definitely shouldn’t be saying in public, that a carbon price is enough, that it’s worth trading anything and everything for, that when we implement it, we are done."

William Connolley said...

> vox

Well, it's DR, I'm bound to disagree. In this case, i think that judging carbon taxes at $50 a failure because they don't get you down to -80% in the USA by 2050 is a mistake. The section on transport is weird, and appears to ignore EC's entirely. And I don't think fetish-like language is useful; all it means is "I can't be bothered to write coherently".

Phil Hays said...

The section on transport looks about as I understand the situation.

EVs are only 2% of car sales, and the average service life of a car is roughly 20 years. The average age of a car is use is 11.4 years in the US, and roughly 8.1 years in the UK.

https://www.telegraph.co.uk/politics/2018/04/12/average-age-car-uk-roads-highest-level-since-turn-millennium/


If EV sales follow an optimistic doubling of every 3 years, it would take nearly 18 years to replace most new cars, and another almost decade to get to half EVs on the road. If all goes well, the 2040's will be when EVs start to impact gasoline usage in a major way. Unlikely to be faster, could be decades later. Especially if we get rid of subsidies for EVs too soon.


Carbon tax would reduce usage of existing vehicles slightly, as there is a tiny impact observed from higher fuel prices, and would bias ICE sales to slightly smaller and more fuel efficient cars. But has little overall impact on transportation, at least until EVs are a large fraction of total cars being sold.

"Personal car driving habits and fuel prices - low price elasticity." rconnor.

https://en.wikipedia.org/wiki/Price_elasticity_of_demand


Need to make a change? Need to change technology, at least for transportation. Carbon tax is a very blunt instrument, a big hammer when a tiny screwdriver would be more effective. But when all you can see is a hammer, everything looks like nails.

As for goals, I don't think an endpoint warming of 1.5C is possible, and 2C is very unlikely. Maybe 5C. The higher the warming, the higher the social cost of carbon, the higher the carbon tax (or alternative subsidies) should be.