Conversation with Alice Yu on effective environmentalism

December 2013

I recently had a conversation with Alice Yu about effective environmental advocacy. Alice studied environmental engineering at Cornell and has taken a much more comparative, effectiveness-oriented approach to environmental issues than most environmentalists I know. So I thought effective altruists might be interested in her thoughts on how to best save the planet.

Note: like most of the conversations I have, this was completely off-the-cuff and unprepared, and focused mainly on the big picture. Citations can be found fairly easily with searching, and I hope to update with a “further reading” section at some point.


Why aren’t more effective altruists interested in environmentalism?

Some people regard it as a “known issue”, i.e., there are already people working on it, so extra contributions will be less useful. Also, there is a perception that many environmentalists hold the environment as a sacred value, in conflict with the more anthropocentric and utilitarian views of effective altruists. Some environmentalists become misanthropic, i.e., holding the environment so sacred as to view human influence as a net negative.

Actually, though, this attitude of sacred value can lead to ineffectiveness–e.g. environmentalists notice that two alternatives have different negative environmental effects, and rather than thinking about how the trade-offs between them work, decide to fight both of them. An example is the divide over power generation: many environmentalists oppose e.g. nuclear power, but the effects of coal power are so much worse that almost anything that averts coal power usage is a net environmental positive.

However, the sacred value mode may be better for the environmentalist movement as a whole. It’s more motivating to most people; being rational about the current state of things is depressing because they’re so bleak; and appeals to sacred values get more public support. For instance, protests against the Keystone oil pipeline were largely conducted via this route. It’s an open question whether it’s possible to get the kind of support needed for big rallies without appealing to sacred values.

Are there subsets of the environmentalist community with more effectiveness-oriented discussion?

When you actually dig in with someone who’s cared about this for a long time, they can sometimes understand the issues quite well. However, it really varies–they can also be disappointingly ignorant. Four or five years ago, Alice was a fan of, whose name comes from the long-term target CO2 level of 350 parts per million. For stopping climate change this number is essentially the only thing that matters; this focus stops environmentalism from being hijacked/conflated with political and social justice issues and thus overly politicized.

Of course, the battle not to politicize global warming may already be lost. Many proposed solutions require stronger government, and things that sort of used to be bipartisan, e.g. cap-and-trade, no longer are. However, there are a lot of non-political options in industry: for example, there’s a lot of potential in cleantech. (Although the government has been subsidizing cleantech, this hasn’t been a great driver of improvement since they’re not very good at subsidizing the right stuff. Elon Musk has been extremely effective, and his government loans were arguably not a driving force since he paid them all back very early. At any rate, cleantech subsidies are no worse than e.g. oil or agriculture subsidies.)

What are the biggest bottlenecks in environmentalism?

In industry, bottlenecks are not in science, or in technology, but rather in market penetration and business. For instance, solar costs are at parity with coal, but they’re incomparable for different reasons. Most people’s model of the environmentalism of power has two components: cleaner generation and reducing consumption. The generation side is usually imagined as a pool where you can put power in and take it out later. But grids are actually way more complicated: people neglect the requirement that grid power in/out must balance not just on the long term, but at every instant. This makes power storage very difficult. Solar and wind are variable and intermittent, so storage becomes much harder (and also introduces more chaos into grid management). For instance, to ensure a steady power supply with wind and solar in the grid, you need ancillary services–paying always-available power to be on “standby” in case your wind or solar power falls suddenly. This is not a big deal when the clean energy is a small portion, but becomes difficult if you hit 30% wind/solar.

Despite these difficulties, places in Europe force wind/solar usage through regulation, by e.g. forcing companies to buy wind/solar power when it’s available despite higher costs. Obviously this is an unsatisfactory long-run solution. Ideally, we’d be able to store the generated power in order to smooth out spikes and troughs. However, storage is quite expensive. The best current option is “pumped hydro” (storing power in potential energy of water), but this is topography-specific and difficult to scale. For instance, Alice did a consulting project during school asking whether pumped hydro could be implemented in Ithaca, NY, which has almost the ideal topography for it (two lakes at different elevations already). But it turned out that this wasn’t enough: even with the most extreme parameters, pumping the entire lower lake into the upper to store power and vice versa, they couldn’t get enough storage.

Other, still unsatisfactory, storage options include compressed air (which is too inefficient) and batteries (which are too expensive). Light Sail Energy is a promising company that’s working to make compressed air more efficient. If they succeed, this would drastically increase the potential market for clean power.

Would technical solutions to those problems be a high-leverage point to attack environmental issues?

Yes. One potential attack, as discussed above, is better power storage systems. Another is smartening grids: making them more responsive and more stable. The US grid is very archaic, which damages adoption of new power technologies. (Places in Europe are probably better, while developing countries are worse.)

Interestingly, however, there are lots of smart grid opportunities in developing countries. This is because of the same technological leapfrogging that resulted in extremely high market penetration of mobile phones: it’s much more costly, relatively, to modernize the US grid than to simply build a modern grid when you’re already going to build a grid from scratch. This is good in a couple ways: developed countries have entrenched paradigm of a few centralized power plants and a lot of transmission infrastructure, because coal plants have economies of scale and zoning problems. But wind and solar can be much more distributed because they’re modular.

However, it’s unclear whether working in the developing world would have the highest impact, since most power consumption currently comes from the US. Better might be smart grids, electric cars, etc. One interesting idea is that as electric cars become more popular, car batteries could conceivably be used as temporary power storage. By a very naive estimate, existing car batteries could power the whole US for a while. Since cars spend most of the time idle, if the grid were smart enough to handle it, it would be technically feasible to use car batteries to store spikes in clean power and smooth production. These could be potentially very valuable as ancillary services, but the main hurdle is that the grid needs to be smarter.

For a case study in how backwards current grid management tech is, here’s how electricity billing currently works: you have a meter on the side of your house, a person physically goes and checks your meter to find out your monthly usage. However, this is expensive, so they only actually check your usage every other month. On the off months, they predict what they think you’re going to use, and bill based on that, and then the next month they charge or rebate the difference between predicted and actual usage. As a result, when Alice moved into an apartment whose previous residents were energy hogs, she got an unexpectedly giant bill the first month, and the next month’s bill was literally negative.

Smart grid monitoring technology

One big goal of grid management, as mentioned before, is smoothing out electricity demand to reduce the need for costly standby power generation. We’re now seeing the beginning of smarter grids that can help balance demand by doing different billing for peak/off-peak hours. This actually does provide substantial incentive for people to use more off-peak electricity: although constantly monitoring power usage takes a lot of attention, power is a surprisingly significant expense for many people. As a result, funding smart meters would be highly leveraged. - They’re concrete, actionable, and under-advocated. - The causality for benefits is hard to evaluate, main immediate benefit is flattening demand graph. - It’s pretty easy to make the financial case for smart meters. - As a result, they’re already gaining penetration; the benefit of advocacy would be in accelerating it.

Another interesting idea is allowing utilities to reduce your power consumption during spikes for you. For instance, you could allow them to temporarily shut off your air conditioner during times of peak demand.

Environmentally friendly diets

Figuring out which diets are environmentally friendly can be difficult. For instance, many environmentalists also prefer to eat local and organic. However, this doesn’t usually correlate with the most environmentally friendly options. This is because transportation is a relatively small part of the environmental impact of raising food, so the dominant effect is from how efficient the supplier is, not where they’re located.

The local/organic movement is actually substantially worse for this in some cases. For local food, more demand pressures farmers to have smaller farms that are closer to cities; often the terrain close to cities isn’t optimal for farming, which costs efficiency, and furthermore, pressure for smaller farms destroys economies of scale. Huge industrialized farming is just more efficient.

Many people have the impression that organic food is also more environmentally friendly, but this isn’t true either. For instance, a common misconception is that organic food must be grown without pesticides or herbicides. In fact, it’s just restricted to “natural” pesticides and herbicides, which are not necessarily any better for the environment because the criteria for “natural” are poorly chosen. Also, since organic herbicides are less effective, many organic farmers are forced to over-till their land to get rid of weeds, which isn’t good–it releases carbon and takes more energy. Amusingly, because of this there’s an environmentalist movement towards “no-till” farming, but there are very few organic no-till farms because if you don’t till you can’t get a good enough herbicide.

Rather than eating more local or organic food, the biggest effect you can have through your diet is by eating less meat. Animal products are just an order of magnitude more environmentally costly than plants (remember high school biology: if you move up a trophic level, you lose an order of magnitude of energy efficiency). So if you want to green your diet, change what you eat, not the label that’s on the thing. If you switch to local/organic everything, you won’t change the environmental impact much–you have to change consumption to get anywhere.

Interestingly, learning more about environmentalism and agriculture did more for Alice’s eating habits than animal suffering ever did. Besides energy consumption, animal husbandry has a lot of other side effects: toxic animal waste getting in the water, etc. Plus it takes way more space: if everyone in the world had a plant-based diet, the US would have enough arable land to feed the entire world.

In terms of environmental costliness, the ranking goes roughly:

Alice also knows some other environmental vegetarians, although the proportions (who’s vegetarian for environmental vs. animal-suffering reasons) are unclear. But the upshot is that effective altruists interested in veg*n advocacy shouldn’t neglect the environmental angle.


Enjoyed this post? Get notified of new ones via email or RSS. Or comment:

Public (as ) submit ⤇

Daniel Haran

Urban politics are one of the highest leverage points for effective environmentalism. Dense, walkable neighbourhoods are inherently more efficient. I recommend reading Carbon Zero if you’re interested:


What about the effects of non-organic pesticides on the soil and ecosystems around the farms? I guess I don’t know for sure that organics are better, but that’s a HUGE part of the problem with large industrial farms. Also: monoculture.

Alice Yu

@Kathleen: Thanks for the comment; I should clarify that I do not wholly adore large industrial farms. My point is that purely in terms of energy expenditure/climate change impact, large industrial farms are more efficient than small local farms, due to economies of scale and the relative insignificance of energy spent on transportation (something like 90% of energy spent on procuring food happens before it ever leaves the farm gate).

There are, of course, plenty of worthy factors other than carbon footprint that go into food choices (environmental and otherwise), and it was outside the scope of this convo to cover them. Also want to heavily re-emphasize that changing your diet is far more environmentally beneficial than switching to local/organic, even generously assuming local/organic is indeed better in every way. Some foods are just inherently more environmentally positive than other foods.


Thanks for this, it’s shifted me to be more interested in environmental causes. The things that have put me off the environmental movement before are:

Do you know of any other environmentalists who are EA-minded?


Richard: not so much–they’re hard to find! I’ll probably try to have conversations with any others I meet as well.


Just came across this discussion through a post by Chris Watkins on the EA facebook group. Here’s my thoughts:

Agriculture has a carbon impact, but also a biodiversity impact, which threatens the functioning of ecosystem services. Without a stable ecosystem with good soil quality and biodiversity, we are more prone to malfunctioning of the ecosystems eg floods, droughts, pest plagues, poor pollination etc.

Two notable examples are the Irish potato famine (caused by a monoculture of one strain of potato) and the American dust bowl (caused by poor soil health). So, when assessing an agricultural system we need to consider how well the biodiversity is being preserved. This includes the soil biodiversity – fungi, bacteria etc, which keep soil functioning and able to absorb and retain water).

One concept which can help in thinking about a farm’s impact is that of ‘land sharing’ or ‘land sparing’. A farm run on ecological principals will ‘share’ its land and have a positive effect on biodiversity eg have hedges which support birds, insects etc, preserve soil by addition of bulky composts etc. Another strategy is to have a highly productive, but high input and low biodiversity system which produces a large amount of food, thereby ‘sparing’ land elsewhere on which wildlife can survive eg in nature reserves.

The above needs to be considered when thinking about efficiency. So, a high input, monoculture may have a higher efficiency per unit area than a more diverse farm. However, a well run farm on ecological principals is likely to be doing much more ‘land sharing’. For example, I buy apples from a small orchard. The orchard is not as productive as a large industrial scale orchard – but it has high biodiversity - supporting much wildlife, and using almost no inputs. Therefore, this orchard has a positive environmental impact while also producing some food. By buying the apples, I am supporting the owner to preserve the environment. This must be a more positive impact than buying from a farm with a negative environmental impact, no matter how efficient it is.

The danger of going for the ‘land sparing’ choice, is that there is no mechanism to make sure that the ‘spared’ land is actually used for wildlife preservation, and so the ecosystem services can still be threatened as much as if the land hadn’t been ‘spared’. With land sharing, in contrast, the ecosystem services are preserved, within farms, and agricultural production can in many circumstances, be an environmental good rather than a environmental bad.

Also – when talking efficiency, we must take care on units – as well as measuring per unit area of land, we can measure by unit of labour, or per unit of input eg fertiliser/pesticides. Large monocultures can be very efficient per unit area, but lower per unit input eg of industrially produced fertiliser, pesticides, machinery, fuel etc, which all have an environmental impact.

Of course, the biggest effect of our dietary choices is still the amount of animal product consumed, and a switch to more plant protein would allow for more of both land sharing and sparing.

Hope that contributes to the discussion a bit.