Do we need to keep increasing energy consumption?

post by JamesOz · 2021-06-16T15:39:19.900Z · EA · GW · 5 comments

Epistemic Status: Unsure, I thought this article was interesting and wanted to start a discussion around it. I’ve spent less than 2 hours thinking or writing about this so I’m not firm in my views

A lot of EA work on climate, namely by Founder’s Pledge, seems to be centred on the assumption that energy consumption will double up from 2020 to 2100, with a necessary increase of low-carbon energy production by a factor of 17 compared to 2020. The reasons being our increasing population and desire to lift billions out of energy poverty so in the future, 11 billion people can enjoy a reasonable standard of living. For these reasons, it makes sense for EA climate solutions to be focused on providing low-carbon forms of energy that will both provide abundant cheap energy to increase wellbeing but also not emit copious amounts of greenhouse gases into the atmosphere. 

However, recent research (see blog post here and paper here) might suggest we don’t need such a dramatic increase in energy production. An excerpt from the blog above:

 “We found that using 60% less energy than today, decent living standards could be provided to a global population of 10 billion by 2050. That’s 75% less energy than the world is currently forecast to consume by 2050 on our present trajectory – or as much energy as the world used in the 1960s.”

And their definition of decent living standards they’ve used can be seen here:

 “There’d be adequately sized housing that maintains a comfortable temperature year-round, with clean, running hot water. A washing machine, fridge-freezer, laptops and smartphones in every home. Enough hospitals and schools to guarantee universal access, and three times as much public transport per person as is currently provided in the world’s wealthier countries.”

Which in my opinion, seems fairly reasonable. Where it gets more divisive is along the following lines:

"There’s no longer room for second homes, second cars, 20 minute power showers in the second bathroom, biannual upgrades of electronic gadgets, new shoes for every season, or plates piled high with red meat seven nights a week."


Given this fairly recent (Nov 2020) research, it makes me think if we need to question our assumption of doubling energy consumption by 2100. Up until now, EA climate work has predominantly been focused on the supply-side of the issue e.g. increasing the supply of low-carbon energy sources via tax credits, R&D budgets and so on. There has been little on the demand-side of things as suggested by the article above e.g. advocating for a simpler lifestyle for wealthier countries which could actually dramatically lower our future energy consumption and hence reduce the scale of decarbonisation we need to undergo. Other benefits of a low-energy lifestyle would be apparent in less resource depletion and environmental degradation. In the scenario where we double energy consumption by 2100 using low-carbon energy, this would still involve massive amounts of lithium, cobalt and other rare mineral extraction and other forms of environmental degradation which in turn could lead to other environmental or socio-political issues. 

In short, I see a few clear benefits of reducing societal energy consumption:

  1. It will reduce carbon emissions and mitigate the worst impacts of climate change. Obviously great if we want to reduce both near-term and long-term consequences to human and non-human life.
  2. It will reduce the scale of decarbonisation we need to deploy and clean-tech that we have to develop via R&D. If we used 60% less energy than today in 2100, we would only need to scale up our low-carbon energy production by approx 3.5x (from 17,500TWh to 64,000TWh) instead of 17x from 17.5TWh up to 308,000TWh. This would significantly reduce the amount of innovation required in yet unproven (at scale) technologies e.g. carbon removal.
  3. It will reduce the significance of other potentially dangerous environmental/political consequences, such as resource depletion.

At first thought, there seems to be a few issues that might arise to make this kind of low-energy society advocacy intractable, namely:

  1. It’s politically unappealing. From my own experience, I’m fairly sure no one in the UK will want to be told to not have a second home, not to  buy new clothes regularly, etc. This makes me think it could be politically very challenging to get any country or group of citizens to agree to this on a large scale.
  2. How would you enforce such a lifestyle without an authoritarian state that hands out rations of phones, clothes and cars?


To clarify, I'm not talking about recommending individuals prioritise lifestyle changes over other forms of climate advocacy or donating, as argued by this great report by Founder's Pledge. Instead I'm referring to advocacy, lobbying or systemic changes that would lead to societally lower demands for energy, rather than individuals switching off extra light-bulbs.

Generally, I’m curious to hear what others think about two issues, which are slightly distinct:

  1. Should we be focusing on low-energy lifestyle advocacy within EA Climate work to reduce the colossal decarbonisation challenge?
  2. More broadly, do we think increasing energy consumption indefinitely will be detrimental to humanity (on any timescale) due to other environmental concerns such as resource depletion or will the benefits outweigh the risks?


Comments sorted by top scores.

comment by jackva · 2021-06-16T20:17:42.865Z · EA(p) · GW(p)

Thanks, James!

More on this later, but for now just two points:

I. Doubling is not dramatic: Doubling of energy supply is not a dramatic increase in at least two ways:

  1. It looks quite conservative when considering the demographic and economic dynamics you mention (60% population increase, hopefully at least a tripling in GDP per capita, i.e. something like a 5x larger economy). Saying one expects energy demand to only double by end of century assumes a lot of reductions in energy intensity, i.e. increased efficiency, structural change, and, possibly, demand reductions.

  2. Relatedly, it is by far not the at the upper end of plausible futures the IPCC and many other bodies consider. Indeed, it would not be terribly surprising if energy demand by end of century increased by much more than just a doubling and this is something our responses should be robust to.

II. Carbon intensity of energy to ~0 is the sine qua non of climate success.

Per the Kaya Identity, the only way to get to zero emissions is when the carbon intensity of all economic activity is zero, it's the only necessary condition and it's also sufficient. Because there is also carbon removal and the goal is net-zero not zero it's not quite as logically necessary (though it's still sufficient).

Replies from: JamesOz
comment by JamesOz · 2021-06-16T20:39:40.354Z · EA(p) · GW(p)

Quick reply from me too - You're right, doubling isn't so dramatic so I'll amend that sentence. What I really meant to say was that we have to scale up our low-carbon energy production from roughly 17,500 TWh in 2020 to 308,000TWh in 2100, an increase of almost 17x, which seems more dramatic to me! Will reply to the following later.

Replies from: jackva
comment by jackva · 2021-06-16T21:13:51.488Z · EA(p) · GW(p)

This also strikes me as pretty relevant in this context, essentially the IPCC's scenarios do not include futures where energy demand does not increase and a doubling (compared to 2010) is roughly in the middle of considered scenarios (of course, this is very simplistic, not all of those scenarios are equally plausible, nor does the IPCC necessarily capture the entire range of possilble futures, but it gives a good sense of how unlikely a scenario such as the one the paper you cite uses is in the overall range of views).

comment by jackva · 2021-06-17T12:30:04.623Z · EA(p) · GW(p)

On your questions more directly:

Q1: In a world where a focus on lifestyle advocacy makes a large difference to emissions (i.e. is cost-effective), I am fairly unconcerned about climate -- this is not a world with a lot of climate risk.
Conversely, in the worlds where most of the risk is -- high growth pressures and low willingness to pay for climate -- such a strategy will not be cost-effective whereas a strategy focused on making low-carbon energy the option of choice irrespective of concern about climate will (what I called the "shit hits the fan principle" in the GWWC talk).

Q2: We know that there is at least one energy source that could reliably and sustainably power civilization for centuries (nuclear fission) and likely there are several more (solar, nuclear fusion, advanced geothermal). This mostly seems a problem if one wanted to power the entire civilization only with intermittent renewables in their current state (e.g. without them becoming more resource-efficient).

comment by Denkenberger · 2021-06-19T00:01:22.445Z · EA(p) · GW(p)

I've worked in energy efficiency, so I've thought about this a lot. Pure energy efficiency is getting the same utility with less energy. However, energy conservation is generally regarded as situations where you have to give something up, such as thermal comfort, convenience, travel, second homes etc. One successful example of energy efficiency is appliances in the United States such as clothes washers, refrigerators, and dishwashers now use about 1/4 as much energy as they did a few decades ago at negative costs for CO2 saved. I think there are still many opportunities to reduce energy use cost effectively and get the same utility. But once you go to non-cost-effective energy efficiency or directly limiting activities, the economic costs (taking into account non-monetary factors) get extremely high. I've run a few numbers and have gotten around $1,000 to $10,000 per ton CO2, versus ~$100 per ton CO2 for things including renewable energy and air capture. So I don't think we should be directly limiting activities.