Climate research in effective altruism

Editorial note

The Global Health and Development (GHD) team at Rethink Priorities (RP) periodically assesses its approach to impactful research and exploring promising lines of work going forward. This report is an output of one such exercise for the cause area of climate change and, therefore, differs from the conventional research outputs of the Global Health and Development team, including in its more subjective tone and its attempted overview of the approaches of the effective altruism (EA) aligned climate community. 

This output was originally produced over a two-week period in January 2023, though some information may be more recent, e.g., if an organization we describe subsequently provided information to us directly during their review of this document prior to its publication.

Greer Gosnell — RP’s resident climate specialist — is the point person for this work and the lead author on this output. Its contents should be considered to be her perspective. There may be some diversity of views within the organization.

Key takeaways

• Founders Pledge (FP) and Giving Green (GG) are the two most prominent EA organizations with a track record of conducting research and making recommendations on climate change.
  • GG focuses on mitigation of greenhouse gas emissions to minimize human suffering, while FP focuses on minimizing expected climate damage, or “probability-weighted damage across worlds.”
  • My best guess is that these two organizations have likely collectively influenced $50 million to $100 million in donations to date (updated in the second half of 2023), with further funding of about $10 million to $20 million committed.
  • Some other currently or historically relevant EA organizations include SoGive, Open Philanthropy, Effektiv Spenden, and Let’s Fund. Effektiv Spenden, a regrantor, relies significantly on FP’s research, and Let’s Fund has made at least one climate recommendation in the past but is likely no longer involved in climate research.
• In my view, a major missing component of most research conducted in EA climate and global health and development (GHD) to date is the limited consideration of co-benefits across these and other related topics (e.g., particulate matter reduction or animal welfare improvements arising from efforts to mitigate climate change, and vice versa).
• The EA climate sphere may also benefit from “offering” research and recommendations tailored to different worldviews or preferences, such as neartermist versus longtermist or level of risk aversion.
• Assuming our ongoing research supports our current belief that there are neglected climate change interventions that are both high-impact and cost-effective relative to other GHD interventions, RP’s GHD team may develop and pursue its own climate research strategy that capitalizes on RP’s comparative advantages. Based on the research that informed this report, initial ideas include:
  • Research to better gauge comparative impacts and giving dynamics within and across climate change and GHD interventions;
  • Research to resolve decision-relevant uncertainties in the EA climate sphere;
  • Strategizing and outreach to grow the pipeline of high-skilled (non-)EA climate researchers;
  • Collaboration with reputable organizations and academics on primary research to improve our understanding of interventions’ impact and cost-effectiveness; and,
  • Assessing, conducting, or supporting EA-style investigations into plausibly high-impact interventions. ​​

Introduction

This document contains information on EA-aligned organizations that have been substantively involved with climate research, as well as some reflections on my climate-related observations in EA thus far while at RP. I joined RP in October 2021 after leaving my academic career as an environmental, behavioral, and (field) experimental economist. During my time in academia, I developed a passion for rigor, skepticism, and contextual understanding in answering research questions, which I aim to bring to my current role.

I had not spent much time reading the climate-related research or philosophical perspectives of the EA community prior to joining RP. My research at RP has been client-driven, so my exposure to EA thinking primarily comes from conversations with clients and EA experts, as well as limited reading, watching, or listening, a vast majority of which has generally taken place in the context of attempting to answer client-motivated research questions. As such, by no means do I consider myself an expert on the full EA climate landscape, and I am humble about my contributions thus far and open to learning. While not comprehensive, my exposure to the climate landscape in EA seems sufficient to contribute meaningfully to conversations around its direction.

My initial exposure to climate expertise in EA — having spent the better part of 10 years working alongside some of the most thoughtful and well-respected academics in environmental economics and policy — left me wondering whether EA climate is under-resourced in terms of personnel, time, and expertise. This document, therefore, attempts to gather a baseline understanding of the theories of change and recommendations in the EA climate space thus far,¹ with some initial musings on potential ways in which to bolster the reputation, coherence, and impact of the EA climate community. While I cannot promise a thorough response given capacity constraints, I invite others to challenge my positions humbly, thoughtfully, and with a sincere drive toward the collective goal of improving the flow of philanthropic resources to the most impactful cause areas and interventions — climate and/or otherwise — that we are capable of identifying. The differences in our perspectives and experiences can be an asset, in this sense, as we iteratively brainstorm, research, and refute the best objections to the various approaches, perhaps ultimately spreading risk by diversifying the collective approach to climate among organizations affiliated with or sympathetic to EA.

Finally, to date I have only dedicated about three weeks of committed time to researching and writing this report, and it is, therefore, subject to change. We remain in conversation with several of the major players in the EA climate space to brainstorm ways in which to improve our collective research through fruitful collaboration and knowledge sharing, including through a series of roundtable discussions among these players to identify opportunities for collaboration that can improve our organizations’ respective and the community’s collective thinking with regard to philanthropic climate change impact.²

To date, EA-aligned organizations’ climate strategies and theories of change have varied, and Founders Pledge and Giving Green have produced the most research and directed the most funding

Founders Pledge (FP) and Giving Green (GG) are the only two EA organizations of which I am aware that focus heavily on climate change recommendations. GG focuses on mitigation of greenhouse gas emissions to minimize human suffering, while FP focuses on minimizing expected climate damages. My best guess is that these two organizations have collectively influenced $50 million to $100 million in donations as of late 2023 (medium confidence). Some other relevant EA organizations either currently or historically involved in climate change include SoGive, Open Philanthropy, Effektiv Spenden, and Let’s Fund.

Both FP and GG host climate funds where donors can give flexibly to the recommended organizations as FP and GG see fit (e.g., depending on room for more funding or political contexts).³ However, the approach behind the two funds differs in that Giving Green restricts their funds to its recommended charities while FP allocates funds to recipients that it views to be the most impactful at the time of allocation.⁴ Beyond their respective funds, these organizations estimate they have influenced much more via individuals making direct donations to their recommended charities, as well as through broader citations of their work.

FP has granted $14.9 million via its Climate Change Fund (Founders Pledge, n.d.-x) — which it founded in the fall of 2020 — of the $17.6 million it had raised by the end of 2023 (Founders Pledge, 2023). FP estimated it had influenced $47.7 million to $78.4 million when accounting for indirect channels (private communication, December 2023). GG’s most recent estimates suggest they have influenced $17.3 million, with $620,000 regranted out of $670,000 granted to the Giving Green Fund (private communication, late 2023). Taking these numbers at face value would suggest that the research and recommendations of these two organizations have together (re)directed $50 million to $100 million as of late 2023.

I focus my attention on these two organizations, though I include a final section that touches on four additional organizations that either have focused significantly on the cause area albeit less than FP or GG (SoGive, Open Philanthropy), have more recently entered the EA climate landscape (Effektiv Spenden), or have ceased to research climate change but made recommendations historically (Let’s Fund). I may add organizations as I learn more or designate separate sections to organizations that become increasingly prominent in the EA climate landscape.

Founders Pledge

These takeaways are primarily based on:

• Founders Pledge Climate Change Fund (Founders Pledge, n.d.-x)
• Volts podcast: Johannes Ackva on effective climate altruism (Volts, 2022)⁵
• The Climate Change Fund at Two (Ackva, 2022)
• Guide to the changing landscape of climate philanthropy (Ackva et al., 2021)
• Conversation with Johannes Ackva and Megan Phelan (February 2023)⁶
• How to evaluate relative impact in high-uncertainty contexts? An update on research methodology & grantmaking of FP Climate (Ackva et al., 2023)

Research approach: Comparative, comprehensive, and credence-driven, with primary focus on evidence-based grantmaking

Two prominent past and present FP climate researchers — John Halstead and Johannes Ackva, respectively — have been at the helm of FP’s climate research approach since ~2017-2018.⁷ Ackva still leads FP’s climate research and, in December 2022, hired a second climate-specific researcher at FP, Megan Phelan, who was recently awarded a PhD in materials science. At the time of our interview with Johannes Ackva (February 2023), FP also had one part-time climate researcher, and was hoping to hire a third full-time climate researcher in the near future.

In a relatively recent (May 2023) EA Forum post, Founders Pledge provides an overview of its motivation and methodology (Ackva et al., 2023), and we supplement the information in that post with notes from our interview with Johannes Ackva. In the post, FP makes clear that its “goal is not to serve arbitrary donor preferences, but rather to guide donors to the most impactful opportunities available” based on “credible comparative methodology to estimate relative expected impact.”

In our interview in February 2023, Ackva told us that given past constraints on researcher time,⁸ FP’s research approach had primarily involved “landscaping” of philanthropic climate giving (e.g., Ackva et al., 2021) and “looking at correlates of impact”⁹ and “additionality.”¹⁰ To identify potential grantees, the organization narrowed in on a space where they expected funding to produce significant benefits in excess of what would occur counterfactually, identified promising organizations in that space, and then invited them to submit proposals (though he said it may broadcast requests for proposals in the future).

As indicated in the EA Forum post (Ackva et al., 2023), more “recent increases in resourcing” have allowed Founders Pledge to make concrete advances in developing a credible comparative methodology that will support its grantmaking going forward. In the post, the organization describes its methodology as:

• Comparative: Its methodology asserts that bottom-up cost-effectiveness analysis (CEA)¹¹ and plausibility checks cannot alone lead to claims of high impact in a climate context characterized by vast and often irresolvable uncertainty and where theories of change are “indirect.” Therefore, credible “comparative judgments of relative expected impact” that minimize the degrees of freedom associated with bottom-up approaches can lead to better decision-making.
• Comprehensive: The “crowdedness in climate” requires that there be an “analysis of funding additionality (‘how likely would this have been funded anyway?’), activity additionality (‘how many other organizations would do roughly the same thing absent funding org X?’), and policy additionality (‘how many of the avoided emissions would have been avoided through other policies otherwise?’)” to approximate the cost-effectiveness of giving or grantmaking.
• Credence-driven (aka Bayesian): FP aims to assign “credences to central questions of impact” to form bigger picture assessments of high expected marginal impact.

Ackva et al., 2023¹²

When asked how much time the organization spends evaluating a given intervention or organization, Ackva responded that its “general rule is to spend time in proportion to how much is at stake” so that time spent would scale more or less linearly with the size of the opportunity. He also mentioned that spending more time is not always associated with a reduction in uncertainty and that FP is “building a systematic framework for research prioritization to answer the question of how long you should spend on reducing a given uncertainty depending on funding opportunity and constraints” (conversation in February 2023).

When asked about how Founders Pledge’s climate research differs from Giving Green’s, Ackva responded that (i) FP focuses on impact maximization without accounting for donor-specific preferences, (ii) it uses a “comparative methodology to identify the highest impact opportunities” as opposed to assessing whether particular interventions or approaches meet a given bar, (iii) it argues that giving to its Climate Fund is the highest impact vehicle given its active grantmaking from the Fund. He also mentioned that the organization’s primary role is grantmaking,¹³ so “they often have time-sensitive things to consider, such as a grantee that would crumble without funding.”

Ackva emphasized that none of the work FP publishes is commissioned, which ensures the independence of their outputs and recommendations. He said that FP sometimes does research specific to the interests of its philanthropic partners but that this work would never be published. The team at FP intends to publish most of their remaining research, but the process is slower than they would like due to resource constraints; as of December 2023, they are working to hire a communications specialist to address existing lags.

Target outcome: Minimize climate damages

In his thoughtful podcast with David Roberts (formerly of Vox; Volts, 2022), Ackva discusses his reasoning behind focusing on reducing risk rather than emissions, which is to “hedge against the failure of mainstream approaches” since the “success of mainstream solutions is often correlated with a world with lower risk of high temperature increase.” In other words, if mainstream approaches succeed in reducing emissions, then we are likely in a world that satisfies the conditions necessary to avoid the worst climate outcomes (e.g., a world with successful global climate cooperation). Thus, minimizing climate damages would require that we prepare for worlds in which mainstream solutions do not succeed, which are characterized by high climate damage.

To do so, he says, we should focus not on increasing the likelihood that mainstream solutions succeed (e.g., investing in more solar energy deployment), but rather on neglected technologies that hedge against the worlds in which mainstream solutions fail and climate damages are, therefore, large (e.g., investing in the development of advanced nuclear, geothermal, or carbon capture technologies).

Theory of change: Redirect funding toward avoiding carbon lock-in, accelerating technological innovation, shaping policy, and growing organizations

In the Volts (2022) podcast, Ackva (quite rightly, I think) says that from an EA perspective, devoting resources to addressing climate change could potentially be justified given (i) the correlation between its solutions and those of other pressing global problems, primarily energy poverty and air pollution; (ii) its role as a “threat multiplier,” so that reducing its effects will reduce global catastrophic risk and improve political resilience; and (iii) its high tractability — via the opportunity to leverage vast societal resources — relative to other cause areas, which could make it attractive from a cost-effectiveness perspective.

He expounds a theory of change (ToC) that focuses on redirecting climate funds¹⁴ and attention toward minimizing climate damages — which are highly nonlinear as temperatures rise — rather than maximizing emissions reductions. For instance, funding is currently disproportionately allocated to emissions reductions in the US and the EU, though they are expected to be responsible for just 15% of future emissions (Ackva et al., 2021).

For instance, given the nonlinear nature of climate damages as the global surface temperature increases, he argues that we should devote resources to solutions and hedges, such as advanced nuclear, that can reduce the likelihood of high-damage scenarios if mainstream policy solutions (e.g., renewable energy RD&D investment and COP-style emissions reduction commitments) ultimately fail. We must, he says, “fund solutions that are robust against the worst worlds.”

The four broad solutions Founders Pledge currently believes are most crucial for avoiding highly damaging climate futures are (Ackva et al., 2023):

• “Policy leadership and paradigm shaping”
• “Accelerating the innovation of neglected-yet-critical technologies”
• “Advocacy to avoid carbon lock-in in emerging economies”
• “Supporting the growth of promising organizations”

Funded organizations: FP’s grants have shifted (and will continue to shift) with perceived needs over time and the evolution of its ToC, and its Climate Change Fund grants cover a wide range of geographical areas and interventions

FP published its first climate recommendations in 2018 in support of the Clean Air Task Force (or CATF, which it still recommends) and the Coalition for Rainforest Nations, the latter of which it no longer recommends due to decreases in its cost-effectiveness estimates by two orders of magnitude.

It established its Climate Change Fund in 2020 (which Ackva manages) and distributed funds to seven organizations over its first two years¹⁵ (Ackva et al., 2023):

• Clean Air Task Force: $4.85 million
  • Expansion of CATF to emerging economies ($4 million, multi-year grant)
  • US federal advocacy for neglected decarbonization approaches ($850K)
• Qvist Consulting Ltd:¹⁶ $1.6 million
  • Repowering coal in emerging economies (technoeconomic research, “mainstreaming”)
• TerraPraxis: $1.19 million
  • Organizational launch and nuclear hydrogen report ($250K)
  • COP26 event on nuclear repowering ($134K)
  • Fundraising and organizational capabilities to catalyze growth ($404K)
  • Capacity for better energy modeling of advanced nuclear power ($400K)
• Future Cleantech Architects: $670K
  • Scaling their team and work on breakthrough technology innovation
• Carbon180: $400K
  • US federal advocacy for carbon removal
• UN Climate Change High-Level Champions: $301K
  • Communications to encourage stronger international collaboration on emissions reduction
• Economics of Energy Innovation System Transition: $50K
  • Dissemination of a report on targeted technology support policy

Thus, based on information available in early 2023,¹⁷ a majority of FP’s funding has been allocated toward the promotion of neglected decarbonization technologies (often via growing organizations in the US and Europe), though this allocation is highly adaptable depending on continual research. For instance, in 2022, FP shifted much (from ~⅕ to ~⅔) of their funding toward “avoiding carbon lock-in in emerging economies,” particularly China, India, and countries in Southeast Asia (Founders Pledge 2022). The organization recommends giving to its Climate Change Fund, as opposed to particular organizations, given that it allows FP to make large, coordinated grants to grantees (including early-stage organizations that lack fundraising capabilities), respond to time-sensitive giving opportunities, and pivot to account for landscape shifts such as (geo)political updates.

FP has likely influenced additional giving to the above organizations through its research (e.g., see their CEAs for CATF and Carbon180) and spotlighting of particular organizations. For instance, Sebastian Schienle mentioned in conversation that his donor-advising organizations (e.g., Effektiv Spenden) have primarily relied on FP’s research for their climate recommendations, and Giving Green recommended CATF based on FP’s recommendation in its initial launch in 2019 (though its later recommendations of Carbon180 and CATF derived from the organization’s own research).

Giving Green

These takeaways are primarily based on:

• High-impact climate nonprofits for climate donors (2022, summary blog)
• Climate change mitigation strategies research (2022)
• Conversation with Justin Labeille, Dan Stein, and Kim Huynh (January 2023)

Additional items of potential relevance not yet reviewed:

• How to Think Beyond Net Zero (business strategy white paper, 2022)¹⁸

Research approach: Assess broad climate landscape to identify solutions and perform ITN assessment, then conduct shallow and deep dives to identify recommended organizations working on these solutions

Giving Green is transparent about its research approach on its website,¹⁹ and their team has been forthcoming with additional information in conversation with our team.

Giving Green’s approach includes the following steps:

• Assess the climate philanthropy landscape and generate a list of possible solutions²⁰, filtering by solutions that might be (some combination of) “overlooked, underfunded, and promising”
• Research the “scale, feasibility, and funding need” (which includes some qualitative ratings) of these solutions to evaluate and create a shortlist of the most promising solutions for philanthropy
• For each item on this shortlist of solutions, develop a longlist of organizations working on these solutions
• Conduct shallow dives for the most promising organizations to gauge effectiveness, “alignment with strategy-specific theory of change (ToC),” and RFMF, possibly among other considerations
• Conduct deep dives to develop organizational ToC and assess its credibility, undertake a cost-effectiveness analysis (CEA) where applicable,²¹ refine the RFMF estimate, consider positive and negative spillover effects, and decide whether to recommend
• Publish the recommendations and conduct outreach to media, donors, and partners in the community²²

Figure 1: Giving Green’s six-step process

Note. Screenshot from “2022 updates to Giving Green's approach and recommendations.”

Giving Green has expressed interest in collaboration with RP on the following dimensions (conversation from EAG DC 2022 with Justin Labeille, reiterated in conversation in January 2023):

• Sharing quick takes on the promisingness of different climate change opportunities
  • Since both of our organizations are capacity-constrained, this practice can help us continually orient our focus to the most promising opportunities.
• Sharing draft research that has not yet been published
  • Given that publishing timelines might be protracted or even put on hold, we can internally share draft research that is not yet public.
• Sharing draft research for feedback
  • If there is a topic we’ve both spent some time investigating, it might be useful to have informal “peer reviews” on some of our research before it’s published.

Target outcome: Mitigation of greenhouse gas emissions to reduce climate-induced human suffering

Giving Green’s mission is to “reduce human suffering due to climate change” (Giving Green, 2022). However, it has some uncertainty regarding whether maximizing emissions reductions is indeed the best approach to minimize human suffering from climate change, and it says the following about the use of this proxy as its target outcome:

...[W]e don’t have a formalized definition of what human suffering means in practice, since it’s hard to define and measure, and is generally open to interpretation. As a proxy, we use greenhouse gas (GHG) atmospheric levels as our outcome of interest. [...] There may, however, be ways to improve the accuracy of this proxy. For example, we might consider using climate damage functions or general estimates of the social cost of carbon. Other climate change organizations compare funding opportunities based on broader outcomes (e.g., Founders Pledge focuses on “climate damage”), which gives us some confidence we may be able to better define and use our actual outcome of interest when making recommendation decisions. (Giving Green, 2022 )

If the goal is to minimize human suffering, it seems to me that some focus on climate change adaptation should be present, or at least considered, relative to the benefits of greenhouse gas abatement.²³ I would expect that conversations with targeted experts could help to narrow in on the optimal means — according to the ITN framework and cost-effectiveness estimates — by which philanthropists could meaningfully reduce human suffering from climate change. However, I also would be curious to hear some justification for why human suffering is the only consideration and wonder if a more inclusive target — including the suffering of other beings or even the capability of humans to flourish — might be more appropriate.

As of early 2023, Giving Green also wants to do more thinking around the tradeoffs between targeting climate impacts in the near term vs. the more distant future. This consideration might affect whether it targets short-lived pollutants like methane or long-lived pollutants like carbon dioxide, or whether it recommends giving toward mitigation today or in the future, for instance, via existing carbon offsets or investments in direct air capture in the future, respectively. The organization currently does not include a discount rate for future emissions reductions, though it acknowledges that doing so may be important if, for example, it “think[s] it’s more important to reduce GHG atmospheric levels as soon as possible to avoid tipping elements” that threaten the stability of the global climate system (see this Carbon Brief, 2020, explainer for more information; quote from Giving Green, 2022). However, it acknowledges that there is sufficient uncertainty around this question that spending significant time considering and “formalizing” a stance is not a high priority, particularly as most of its (potential) recommendations involve mitigation of emissions on a similar timeline.

Theory of change: Recommend organizations, increase donations to organizations, allow organizations to increase their impact

Giving Green is also very clear and concise about its theory of change, which “follows a five-step process”: (i) “produce high-quality recommendations,” (ii) “supporters make donations,” (iii) “recommended organizations increase their activities,” which (iv) “remove or avoid atmospheric [greenhouse gases],” which (v) “reduces human suffering due to climate change.” It provides the following figure to visualize its ToC (Giving Green, 2023):

Figure 2: Giving Green’s theory of change

Note. Screenshot from “Giving Green’s Research Process.”

Recommended organizations: GG’s top recommendations focus on policy advocacy and research for a low-carbon energy transition and alternative proteins, while their business recommendations emphasize carbon removal

Giving Green’s recommendations depends on the relevant audience. The organization targets one set of recommendations to individual donors and another to businesses with net-zero goals, and it has also provided some (minimal, to date) guidance to investors, as well as donors in Australia who solicited their guidance.

Giving Green’s current recommendations²⁴ include the following:

Top recommendations:

• Clean Air Task Force (GG’s deep dive research): US advocacy organization (currently scaling globally)
• Evergreen Collaborative (GG’s deep dive research): US advocacy organization
• Good Food Institute (GG’s deep dive research): Nonprofit aiming to make alternative proteins competitive with conventional meat products
• Good Energy Collective (GG’s deep dive research): Nonprofit supporting advanced nuclear reactors as part of the clean energy transition
• Industrious Labs (GG’s deep dive research): Corporate and government advocacy organization focusing on low-carbon commitments to facilitate industrial decarbonization

Business-specific recommendations (see white paper):

• Direct policy engagement
• Support for technological innovation
  • Carbon removal funds (Frontier, Milkywire)
• Create or contribute to a climate action fund
  • The Giving Green Fund, for instance, donates to the top recommendations listed above
• Improve conventional offsetting
  • High-quality offset (Tradewater, BURN) or carbon removal (Mash Makes, Charm Industrial, Climeworks) projects
• Consider cost-effective opportunities to reduce own emissions²⁵
• Buy quality offset

Giving Green recently opened its Giving Green Fund for strategic high-impact giving so that individuals need not choose an organization but can instead give to the Fund to ensure that their donations are being allocated most effectively over time.

Some other organizations in the EA climate sphere of which we are currently aware: SoGive, Open Philanthropy, Effektiv Spenden, and Let’s Fund

I have spent less time seeking and investigating the publicly available documentation of the following organizations’ strategies and approaches thus far, though I am including my current impressions based on either direct communication or scanning through some documentation of which I am already aware. In three conversations in early 2023 — one with Dan Stein, Justin Labeille, and Kim Huynh of Giving Green, one with Luca Righetti of Open Philanthropy, and one with Johannes Ackva and Megan Phelan of Founders Pledge — we confirmed that the organizations listed here are likely the only relevant additional organizations involved in EA climate research, to their knowledge. The organizations are ordered according to my loose perception of their relevance in terms of EA climate impact to date and in the future.

SoGive

SoGive, which now operates mainly through contracted research (previously through volunteer-based research), has to date been a charity evaluator. However, it is currently transitioning to a model closer to RP’s (as communicated by Sanjay Joshi, founder of SoGive, to both Mel Basnak — who was a senior research manager on RP’s Global Health and Development team at the time — and me during EAG DC 2022, and as stated in its 2023 plans).²⁶

SoGive quotes some of its past climate work in its 2023 plans, including a critique of the analytical approach of Giving Green²⁷ (by Alex Lawsen) and a review of Cool Earth (by Sanjay Joshi). When asked about major players in the historic development of the EA climate space, at least one individual mentioned the importance of SoGive, highlighting its work (with an emphasis on these two Forum posts and their early critique of the Founders Pledge recommendation of the Coalition for Rainforest Nations) as having been quite influential. While I have not yet read the Cool Earth review, I did read the Giving Green critique and found reading the piece and some of the comments²⁸ very helpful both for understanding the dynamics across organizations and some of the key players in these organizations, at least historically.

SoGive mentions some climate work it plans to do in the future, with a special focus on red-teaming John Halstead’s book-length report on climate change. The organization also links to its public agenda which contains other climate priorities, including to “calculate a[n] SCC which includes tail risks” and conduct a “review of REDD+.” However, as of recent communication in January 2024, this backlog of potential research topics is second-priority compared to research that is donor-driven and will directly impact major gifts.

Open Philanthropy

My knowledge of Open Philanthropy’s (OP) climate strategy originally derived largely from direct experience working with its team to answer research questions that informed their strategy and subsequently from a conversation with Luca Righetti — who was the main researcher leading OP’s climate projects in 2022 — during the research for this report.

Historically, climate change has fallen under Open Philanthropy’s Global Health & Wellbeing (GHW) Framework. Open Philanthropy is therefore not just looking to compare climate interventions against each other but also to compare climate interventions against other GHW cause areas, such as global aid policy, science innovation, and farmed animal welfare. Open Philanthropy evaluates GHW grants largely by estimating the income and health benefits of an intervention and then comparing these benefits to the opportunity cost of not pursuing other interventions (Oehlsen, 2024).

In 2022, OP began to more deeply investigate how to value climate mitigation under its GHW cause prioritization framework by estimating the expected income and health damages of a marginal ton of CO2. Following further internal and external scrutiny, OP may publish these findings and its “social cost of carbon” in the future. For now, Luca Righetti’s slides from EAGx Latin America 2023 (Righetti, 2023) are useful in getting a preliminary sense of OP’s findings (although the slides should be interpreted as Righetti’s views and not OP’s). A key takeaway is that OP finds that despite recent progress on climate mitigation, the marginal costs could still be high under its GHW framework because many damages will disproportionately hurt the global poor (who are typically undercounted when only looking at dollar damages).

In addition to the importance of climate change mitigation as a whole, OP also wants to consider whether an intervention appears neglected and tractable, following the ITN framework. As such, OP is continuing to review specific approaches to climate that might be more neglected and still tractable. For these reviews, OP follows a “shallow/medium structure,” meaning that if it decides something is worth investigating, it starts with an approximately three-week investigation and, if the ITN picture merits further consideration, conducts another months-long investigation, perhaps averaging around three months. The GHW cause prioritization team will then deliberate findings and invite external scrutiny before deciding whether to launch a program area or make a larger set of ad hoc grants.

Grants made

To date, OP does not have a Program Officer who is focused on climate change full-time. Given limited grants made in the climate domain, OP’s grants page also does not list climate change as a focus area.

However, some program areas do intersect with climate change. Most prominently, Open Philanthropy has made a few small and contextual grants related to geoengineering as part of its Scientific Research program area. These grants have mostly focused on building up scientific capacity in LMICs to better understand the costs and benefits of solar geoengineering in the developing world (e.g., the Degrees Initiative). It has also made a few grants to strengthen global governance and deliberation (e.g., the Global Commission on Governing Risks from Climate Overshoot). Righetti notes that OP’s grants do not favor or oppose solar geoengineering but rather aim to help the world make more informed decisions, especially in relation to how solar geoengineering affects LMICs.

In addition, Open Philanthropy’s regranting challenge has supported the Tara Climate Foundation, with a focus on limiting coal lock-in in Asia and improving air quality. Righetti has also noted that some future climate-related grants may fall under OP’s science innovation (e.g., clean energy RD&D), South Asian air quality (e.g., coal phaseout), and farmed animal welfare (e.g., alternative proteins) grantmaking.

Effektiv Spenden

Effektiv Spenden primarily “leverages existing analysis, where available” (email communication with Schienle, August 2022), but at times complements available analysis with its own work where the organization views gaps in the coverage of EA climate (and biodiversity) research. It is not a research organization or evaluator, though Schienle mentioned that, at the request of a donor, his team had done some “high-level work in coming up with plausible interventions and theories of change, and areas with high potential for impact at the intersection of climate change and biodiversity,” which were quite nascent and not yet public when we spoke in September 2022. Effektiv Spenden’s website suggests it aligns with an approach akin to the ITN framework, with the addition of “urgency” (with mention of “irreversible tipping points”). It is my view that a research-sharing platform and/or roundtable discussion could be helpful to situate its theory of change within the broader EA climate context and inform its approach to advising on effective giving.²⁹

In line with the recommendations of Founders Pledge, it currently recommends giving to the Future Cleantech Architects, the Clean Air Task Force, and Carbon180.³⁰ For reasons similar to those provided by Founders Pledge, it also offers the option of donating flexibly — primarily across these organizations,³¹ but also to support additional initiatives — via its climate protection fund, where the organization makes donation decisions half-yearly. It appears that Effektiv Spenden and Founders Pledge are already in regular communication: “[...] we are in regular contact with [the funded organizations] and coordinate our funding decisions closely with partner organizations such as Founders Pledge” (Effektiv Spenden, n.d.).

Let’s Fund

Let’s Fund is a small research organization with a team of three people (led by Hauke Hillebrandt³²) that works to identify high-risk, high-reward funding opportunities and crowdfunds for those it deems worthwhile. I came across this organization in the 2023 EAGx slides that Luca Righetti of OP shared.

According to Let’s Fund’s website, the organization “roughly use[s] a framework that looks at the scale, tractability, and neglectedness of a problem to assess the cost-effectiveness of a given problem (such as climate change, or the replication crisis) or charitable project (such as climate policy and Registered Reports advocacy). It makes its research reports public here and appears to have investigated three areas — clean energy R&D (Let’s Fund, n.d.), the replication crisis in science, and impact investing — with live campaigns to fund the first two and one additional research area (global catastrophic risks) in progress.

The organization claims to differ from much of the EA movement in its recommendation of “smaller fledgling projects” characterized by high risk but significant social value in the event of success rather than recommending donations to “established charities with considerable room for more funding.” I am otherwise currently unclear on its research process and theory of change and have spent no time engaging directly with this organization to date.

Potential directions for EA climate research

A collective and coherent EA theory of change might lead to different approaches across organizations based on the types of donors (e.g., neartermist vs. longtermist) we target. Another (not necessarily mutually exclusive and perhaps complementary) approach seeks to identify climate change interventions with sizable co-benefits — or global health and development interventions with sizable climate co-benefits — to attract new donors, address additional global priorities, and optimize total impact.

Diversify strategies to satisfy different worldviews

Effective Altruists and those aligned with the principles of effective altruism have myriad ethical perspectives with diverging implications for how to approach the ultimate goal of optimizing our giving for impact. One common distinction within the EA community is the extent to which an individual adopts a primarily neartermist or longtermist worldview. The longtermist view holds that “positively influencing the long-term future is a key moral priority of our time,” which derives from the notion that “people’s lives matter, no matter when in the future they occur” and the sheer number of people who we would expect to be alive in the future in the absence of some species-destroying catastrophic event (Moorehouse, 2021). Contrarily, neartermism places moral priority on affecting sentient beings who exist now or in the foreseeable future. According to Vox, people who care about climate change are likely to be “weak longtermists” (Samuel, 2022).

However, supporting climate-related causes may be justifiable from either neartermist or longtermist perspectives, and one’s worldview will affect how they approach climate change solutions. Ackva and Halstead (2022) suggest that longtermists are likely to care about minimizing the likelihood that humans become extinct — or that there is some other related catastrophe with negative consequences that last for a long time — due to climate change. In other words, this worldview concerns climate impacts in a “worst-case scenario.” They, therefore, focus on tipping points that may affect Earth’s habitability, agricultural impacts that can affect our ability to sustain ourselves, and indirect effects such as geopolitical instability and other ‘threat multipliers.’ Neartermists, they suggest, may put more weight on negative climate impacts such as “rising sea levels, loss of biodiversity, and reduced labor productivity” (Townsend 2022’s summary of Ackva and Halstead 2022).

The authors ultimately find that climate change deserves attention from a longtermist perspective given its relative tractability, but that relatively risk-neutral longtermists should likely prioritize beneficial AI and biosecurity. They are highly uncertain about how climate change fares in comparison to other neartermist interventions, claiming that this question is an “active area of research” in EA and “we are likely to know more in the next few years” (Townsend 2022’s summary of Ackva and Halstead 2022).

The discussion seems to suggest that some longtermists who seek impact measurability and/or solution tractability may prioritize climate change over other conventionally longtermist cause areas.³³ Relatedly, neartermists and longtermists alike may simply be more driven to solve climate change over other issues if they have high environmental preferences or other personal reasons to be concerned about climate change specifically. EA climate researchers could consider diverse worldviews that they believe are morally justifiable and offer recommendations for donors that may align with each of these worldviews. Such tailored recommendations could combine considerations such as (i) neartermist vs. longtermist perspectives, (ii) varying weights placed on non-climate co-benefits, (iii) risk preferences, and (iv) the size and purpose of donation under consideration (e.g., business vs. individual donors).³⁴

For instance, opportunities to give associated with geoengineering are likely more appealing to individuals or donors with (weak) longtermist views (i) with low weight placed on non-climate co-benefits (ii) and medium to high risk tolerance (iii).³⁵ On the other hand, research into alternative proteins may be appealing to to individuals or donors with neartermist views (i) with large value placed on non-climate co-benefits, e.g., from improved farmed animal welfare as well as ecological resilience, indigenous community empowerment, and biodiversity benefits from avoided deforestation (ii). When considering anti-deforestation, the prospect of high-quality and verifiable forest carbon offsets may be more appealing to risk-averse businesses with net-zero targets (iii, iv), whereas a neartermist (i), relatively risk-neutral (iii) and large-scale philanthropist (iv) concerned with improving the environmental integrity of said offsets — and perhaps one who is also attracted by the myriad potential co-benefits of anti-deforestation initiatives (ii) — may want to fund a well-designed research study or pilot program to rigorously identify improvements to programs such as REDD+ that receive significant funding but whose effectiveness may be significantly improved.

Identifying the best opportunities for common combinations of these donor traits may additionally attract further donations from individuals whose worldviews are incompatible with those espoused in the EA climate space to date.³⁶ Note that EA may already provide recommendations that satisfy diverse worldviews to some extent, though, to my knowledge, these distinctions have not been clearly stated, transparent, or intentional.

An EA-aligned donor who contributed funds to Rethink Priorities’ Global Health and Development team in exchange for a rigorous research report on climate has since suggested that we should be “funding both” high-risk interventions with potentially very attractive cost-effectiveness (in the event of success) as well as more “reliable” interventions whose cost-effectiveness may be less attractive but still reasonable (in their example, they compare a high-risk intervention with cost-effectiveness potentially as low as $1/tCO2, or a more reliable intervention with cost-effectiveness of around $70/tCO2e).

Maximize intercausal impacts (i.e., climate co-benefits) to attract new donors to climate change, maximize overall impact of climate recommendations, and ensure positive impact of donations in the context of vast climate uncertainty

From various sources I have come across in EA (and no one in particular), it appears that EA attention to climate change remains somewhat controversial, as partly evidenced by 80,000 Hours’ debating whether climate change poses a significantly sizable risk to extinction (Hilton, 2022)³⁷ to merit its continued attention, and ultimately determining that while “working on this issue seems to be among the best ways of improving the long-term future,” it is “less pressing than [its] highest priority areas” (Hilton, 2022).

Many climate solutions, however, breed significant co-benefits, or they could with relatively small tweaks to their design. For instance, climate solutions such as adaptation policy and implementation, water policy, anti-deforestation initiatives, and clean cookstove deployment may bring benefits in terms of reducing the negative impact of climate change while also improving other health and development indicators.³⁸

Likewise, global health and development (GHD) initiatives — such as fertility interventions that affect female autonomy and population growth (e.g., United Nations), educating girls (e.g., Kwauk, 2021), and (climate-friendly) infrastructural development (e.g., Ferrer et al., 2019) — and animal welfare initiatives³⁹ (see Joy, 2022⁴⁰) often come with climate co-benefits. Highlighting and valuing the climate co-benefits of these GHD and animal welfare interventions, e.g., by including them in EA-style CEAs, may, therefore, draw individuals who are not primarily motivated by the cause area of climate change to tailor their giving to more climate-friendly interventions. For instance, a popular anti-poverty intervention in EA — i.e., distribution of direct cash transfers, in this case, conditional on households’ meeting health- and education-related criteria — was found to reduce deforestation in surrounding areas in Indonesia (Ferraro et al., 2020). All else equal, then, a donor to an organization like GiveDirectly — who may otherwise not be primarily interested in making climate donations — may be inclined to target their donations to impoverished fringe forest dwellers in areas where the results of the above study are likely to generalize.⁴¹ Similarly, Mel Basnak reviewed a proposal on geographically targeted anti-deforestation advocacy in the name of pandemic prevention, a major area of concern for both longtermist and GHD donors. And, of course, non-climate interventions with negative climate impacts should consider those impacts in their assessments.

Personally, my own motivation for becoming involved in climate change research was my conviction that it was (at the time) a highly neglected cause area with vast — and inequitable, given the sources of historic emissions — welfare implications for the world’s poorest. However, I have maintained interest in other cause areas, such as animal welfare and female empowerment. Finding high-impact organizations that satisfy common overlaps in climate and non-climate donors’ interests may lead to increased donations in the same way that product differentiation can attract new buyers to a given product category. Of course, the potential to attract new donors must be traded off with the potential downside of spreading climate-related donations across too many interventions in a way that could dampen their effectiveness.

Finally, maximizing intercausal impacts is not currently an objective of the EA climate community, which seems like a potentially important oversight given its situation within a broader movement that values maximizing positive impact and given the vast variety of potential donor worldviews.

Assess the interplay between GHD and climate change in more depth

Related to the above, one area on which we would like to see some clear and convincing conclusions regards ways in which GHD and climate change researchers can make meaningful comparisons of their benefits, e.g., between the value of DALYs averted and metric tons of CO2-equivalent mitigated or removed. Depending on our deeper exploration, we may have our own ideas of how to standardize an approach to this comparability issue, or we may find an alternative approach, perhaps interviewing or surveying experts in health, climate, and relevant areas of non-market valuation.

Additionally, it may be worthwhile to understand the comparative budgets for climate change and non-climate GHD work, both philanthropically and societally, as well as the extent to which these two budgets (could) overlap and complement each other, or the extent to which each may crowd out the other. We believe that having a broader understanding of the interplay between GHD and climate change research would help our organization understand how to assess tradeoffs and complementarities across these related cause areas, as well as whether and how to prioritize within and across them. For instance, GiveWell commissioned a research report in late 2023 in which our team provided a high-level understanding of the interplay between climate objectives and the GHD cause areas GiveWell supports. Just over two weeks of research resulted in a number of ideas for further research to improve decision-making within and outside of GiveWell, and we imagine further research would unveil paths to leveraged impact across these two fields.

Rethink Priorities’ potential contributions to EA climate going forward

Rethink Priorities remains in communication with many of the major incumbent players in the EA climate space to improve transparency and beneficial cooperation within the community, engage in knowledge and resource sharing, expand prospective stakeholders, and overcome bottlenecks and uncertainties in our collective and individual paths to impact. To this end, we invited the organizations who are active in this space to a series of voluntary roundtable discussions throughout 2023, where we engaged in useful discussions that have led to increased idea and research sharing as well as critical and engaging discussions regarding research approaches and methodologies that we hope will continue going forward.

We also hope to develop our own strategy — based on RP’s comparative advantages and important gaps identified in the EA climate space — for supporting impact in this community. Some ideas based on our preliminary research have included (i) conducting rigorous research on the comparative (or interactive) impacts and giving dynamics across climate change and other global health and development research; (ii) supporting the resolution of potentially high-impact uncertainties or stress testing assumptions that have significant influence on (the ranking of) EA climate recommendations; (iii) growing the pipeline of high-skilled (non-)EA climate researchers with expertise relevant to the questions EA climate seeks to answer; (iv) contributing to the development and implementation of primary research projects or interventions in collaboration with reputable organizations or academics; and (v) assessing, conducting, or supporting research into plausibly high-impact interventions, e.g., by reviewing prominent EA climate recommendations,⁴² reconsidering “file drawered” grant proposals (e.g., in academia), or surveying experts on questions relevant to the potential roles of philanthropy in supporting neglected interventions or otherwise limiting climate damages.

Acknowledgments

This post is a project of Rethink Priorities–a think-and-do tank dedicated to informing decisions made by high-impact organizations and funders across various cause areas. Greer Gosnell researched and wrote this report. Thanks to the following individuals for providing helpful feedback, though they do not necessarily endorse any perspectives or conclusions in the report:

Rethink Priorities: Marcus Davis, Kieran Greig, Tom Hird

External organizations (in order of mention): Johannes Ackva (Founders Pledge), Daniel Stein and Justin Labeille (Giving Green), Luca Righetti (Open Philanthropy), Spencer Ericson (SoGive), Sebastian Schienle (Effektiv Spenden), Mel Basnak (Open Philanthropy)

Appendix

Figure A1: Founders Pledge Climate Change Fund recipients as of January 2023

Note. Founders Pledge Climate Change Fund (2023)

Notes

---

1. As such, this document may eventually also be useful for climate researchers new to EA to get up to speed on the history and current state of climate EA, which I would have found very useful when I first started my role at RP. ↩

2. Note that this roundtable series took place over several months starting in May 2023. We thank all participants for their sincere and thoughtful engagement that led to the fruitful discussions and positive outcomes of these convenings. ↩

3. Founders Pledge established its Climate Change Fund in 2020, while Giving Green established the Giving Green Fund in 2022. ↩

4. In 2022, for example, the FP fund primarily allocated money to organizations with small funding margins to whom they had never granted money before, which they viewed as having an added benefit of attracting attention to these organizations that may otherwise not be on donors’ or investors’ radars. ↩

5. Quotations from the podcast may not be exact, as I was unable to quickly find an online transcript; however, my notes from the podcast are likely very similar to Ackva’s words, so I have included quotes where I quote directly from my notes. ↩

6. Note that this and related lists below are not exhaustive. ↩

7. Ackva joined Founders Pledge in late 2019, and has an academic background in social science with about a five-year career at a think tank, where he worked on “carbon pricing, emissions trading systems, and innovation policy supporting government policymakers.” Halstead, who has a doctorate in political philosophy followed by a series of research roles in EA-aligned organizations since around 2015, was a researcher at FP from 2017-2018 and led its applied research from 2019-2020. ↩

8. FP mentioned that, in general, it aims to publish all research, though research constraints can sometimes make publishing past work a lower priority than ongoing research and grantmaking. FP regularly shares unpublished research with relevant stakeholders, though, such as Open Philanthropy, Giving Green, Effektiv Spenden, SoGive, and others. ↩

9. To explain correlates of impact in the Volts podcast, he invites the listener to “consider how different strategies are related. When strategies are characterized by many uncertainties, it is important to consider whether they are independent of each other or correlated, in which case scenarios need to be considered jointly” (Ackva et al., 2021). Ackva uses the example of avoiding deforestation to explain one of the correlates of impact — “hedginess” — claiming that avoiding deforestation is likely to be cost-effective only in the best of worlds characterized by international cooperation and adherence to high-quality agreements, while damages would not be terribly high in those worlds because temperature change would be relatively low in such a globally cooperative world. ↩

10. When describing one type of additionality — policy additionality — he asserts that “philanthropists should also be sensitive to whether they are funding true policy additionality: they must make sure funded efforts do not simply make it easier to reach policy targets that would be reached anyway (in which case additionality would be zero: the money is wasted)” (Ackva et al., 2021). ↩

11. Ackva mentioned that the lack of RCTs in a number of relevant contexts means that bottom-up CEAs can be arbitrary, though his team still uses them when they might be useful for comparative analysis. They do not use an explicit discount rate, but they “discount for future emissions in the sense that the probability that those emissions occur is lower if, for example, coal collapses in an evaluation of Repowering Coal,” though they would more explicitly discount if climate science and expert consensus dictated that we are nearing a tipping point. They generally just use CEAs to “get a sense of the structure of uncertainties” and understand whether an opportunity is worth further pursuit. They plan to standardize their cost-effectiveness modeling given their recent team expansion, and they will use their “impact multiplier framework as a prior to drive cost-effectiveness expectations.” They said they are happy to publish any work on this front. ↩

12. “We think of it as a ‘shut up and multiply’ tool for grantmaking and research prioritization in a situation of many ‘known unknowns’ (more precisely, knowing many drivers of impact, but their effect sizes being uncertain), funding opportunities that can be meaningfully approximated as conjunctions of impact-related variables, and a decision context of allocating grant making and research time budgets concurrently.” Note that we have not spent the time that would be necessary to fully understand and evaluate their tool so we cannot speak to its usefulness for determining optimal grant allocations nor identifying opportunities with maximal decision relevance. However, our understanding is that they plan to make the model publicly available for public consideration and collaboration within and outside of climate change (“we believe that this tool will have useful applications outside our climate work, both at FP and beyond”; Ackva et al., 2023). ↩

13. The EA Forum post refers to “FP Climate” as a “research-based grantmaking program” (Ackva et al., 2023). ↩

14. The organization emphasizes the extent of societal spending on climate (~$1 trillion/ year) relative to the extent of philanthropic funding (around 1% of that, or ~$10 billion/year) and therefore argues that leveraging societal resources is likely to have an outsized impact (Ackva, 2022). ↩

15. A timeline and description of this funding allocation can be found in the Appendix. Note that this report was originally written in early Q1 2023, so more recent updates are not included. ↩

16. While we originally linked to the website in this report, it appears to no longer be active.  ↩

17. Regularly updated information on Founders Pledge’s Climate Change Fund is available here. ↩

18. “Instead of assuming conventional frameworks like "net-zero" are optimal, we explore the following question: given a set of available resources, how can a business maximize its climate impact? This new framing allows for nuance, creativity, and the reality that, in some instances, carbon accounting may limit the impact of a company’s strategy. While each company will inevitably face its own set of challenges and constraints when devising a climate plan, it will also have a unique set of opportunities. In this white paper, we provide 4 concrete, high-impact strategies that your business can take to maximize its climate impact” (Giving Green, 2023).  ↩

19. We were inspired by this transparency, which has led us to outline our own research approach here. ↩

20. Giving Green’s first “broad sweep” exercise in 2021 began with the list of interventions in Drawdown. The organization realized that Drawdown focused too specifically on technologies rather than policy levers to influence the success of those technologies (e.g., “new window installation” rather than “updates to building codes”). Their team has now developed a “policy matrix” that they regularly update as the policy landscape changes. ↩

21. Giving Green was publicly critiqued for not always undertaking CEAs for organizations they recommended in an EA Forum post by Alex Lawsen, who was a volunteer for SoGive at the time of writing and now works with 80,000 Hours. He claimed that Giving Green had an “unusual approach, especially compared to standard EA practice,” specifically in relation to their lack of quantitative modeling “in favour of a combination of metrics which do not have a simple relationship to cost-effectiveness.” In his response (in the post’s comments), Dan Stein (CEO of Giving Green) argues — quite compellingly, in my view, and in a way that aligns with my own perspective — as to why CEAs may not always be appropriate, in particular if they are highly dependent on fundamentally uncertain inputs. ↩

22. In our conversation, Giving Green mentioned their drive for transparency and kindness, mentioning that they, therefore, struggle with the issue of whether to publish research on organizations that they do not deem worthy of recommendation. Our understanding is that they have conducted research that is and will remain unpublished, though they may be able to share their research with others researching climate change in EA. ↩

23. There is a general distinction between climate change “mitigation” and “adaptation” approaches. Mitigation refers to efforts to reduce the severity of climate change damages by reducing greenhouse gas emissions into the atmosphere (e.g., transitioning from fossil fuel energy sources to renewable ones). Adaptation refers to efforts to reduce the damages people experience from climate change by better preparing for its impacts (e.g., large-scale infrastructure changes, such as building sea walls in coastal cities). ↩

24. Updated recommendations can be found on Giving Green’s website, including top global recommendations, top recommendations for Australia, and recommendations for businesses. ↩

25. Giving Green does not consult companies on emissions reductions, but provides resources for companies to find such guidance: “While strategies for direct emissions reductions fall outside of the scope of this white paper, there are many platforms available to help companies measure and track their emissions. A recent report reviews some of these platforms and evaluates which are best suited for small and medium-sized enterprises (SMEs), start-ups, larger enterprises, and VCs. Watershed is one example of a platform that helps to measure a company’s emissions across scopes and develops strategies for climate action. It offers rigorous accounting, tracking, and reporting of emissions across sectors in addition to guided strategies to reduce these emissions.” ↩

26. In communication since the initial writing of this report in early 2023, SoGive provided an up-to-date description of what this new model will look like from the donor perspective, which is here: “SoGive launches expanded advising and custom research service: Feel more confident in your giving, across cause areas.” ↩

27. “Prior to this write-up, there had been little critical analysis of the quality of Giving Green’s work” (SoGive’s 2023 plans). It is my view that there is relatively little critical analysis of climate EA research altogether, which may also be the case for other cause areas within EA to which I’ve had little or no exposure. ↩

28. I have not read the contents of all of the comments (as there are many and they tend to be quite long). I have read Dan Stein’s initial response as well as some of Johannes Ackva’s initial comments. Both Stein and Ackva provide some information on how their perspectives were formed and are now informed, including Stein’s training in economics and Ackva’s evolution from progressive youth climate activist to adopter of a more moderate (“center-left”) perspective. These multiple perspectives make for interesting debate and can help to make sense of what sometimes appears to resemble a clash of perspectives and approaches within the EA climate community, which at times feels productive to me and at others does not. ↩

29. We convened such a roundtable for four sessions in 2023, and we may organize or encourage others to organize additional sessions in the future. ↩

30. Note that since this piece was written in Q1 2023, Effektiv Spenden has since added The Good Food Institute to its list of recommended organizations. ↩

31. “We also support initiatives or organizations that we consider extremely promising, but which we do not (yet) present as individual recommendations on our platform. This is also important because the fund can also support smaller but highly effective initiatives and projects that, due to their small size or time limit, do not qualify as individual recommendations for effective donations” (Effektiv Spenden, n.d.). ↩

32. Hillebrandt appears to have been an early researcher in the EA climate space and has an academic background in neuroscience. In 2019, he published a piece on the EA Forum (Hillebrandt, 2019) originally claiming that climate change interventions were more effective than global health interventions and has since reversed this take. ↩

33. A similar claim can be made for other cause areas that have appeal from both neartermist and longtermist perspectives, such as biorisk. ↩

34. We may also consider targeting investors or entrepreneurs, for instance to conduct shallows or CEAs for young or budding start-ups or for socially responsible banks or venture capitalists deciding where to invest. Giving Green has done a small amount of work here, citing general “uncertainty about (a) the impact of this analysis and (b) our relative value-add compared to others focused on climate change investments” (Giving Green, 2022). ↩

35. Here, we are solely conjecturing based on loose priors and further research could help to inform the relationship between preferences and donation decisions. ↩

36. Giving Green also touches upon additional reasons (e.g., public perception, tax implications) that donors may prefer more options: “[W]e think donors sometimes have restrictions or self-imposed preferences that cause them to give less (overall and/or to specific initiatives) than they would otherwise give. For example, a business with net-zero goals might donate $10,000 to our Frontier carbon removal recommendation but would never donate to our US policy recommendations due to fears of being perceived by customers as political or partisan. An Australia-based donor may be more likely to give to Australian nonprofits for tax-deductibility reasons” (Giving Green, 2022). ↩

37. I have not yet engaged with this debate, though I imagine that if I were to do so, I would ultimately sympathize with a perspective that Sanjay Joshi shared with me at EAG DC 2022. This perspective posits that climate change may not pose a risk to human extinction on its own but that it interacts with other x-risks (e.g., those that may be correlated with geopolitical conflict) in a way that makes them more likely. I would like to explore research (or read a rigorous synthesis of said research) on the potential interactions between climate change outcomes and identified x-risks. In a conversation with Danny Bressler, he mentioned that climate change may indeed be an x-risk and that academics are so strongly incentivized to avoid speculation and provide evidence that they do not engage in this conversation, so EAs do not find evidence for it in the literature. ↩

38. To be clear, I am not claiming it is the case that what is best for one cause area is often, or even sometimes, best for another (which would resurface discussions around suspicious convergence; Lewis, 2016). Rather, my point is that many of the drivers of climate change — such as coal power production or livestock agriculture — have negative implications for human and/or animal welfare, meaning that more holistic assessments of efforts to address those drivers would lead to a rosier picture of impact than when they are considered from a climate perspective in isolation. ↩

39. For instance, Giving Green (2023) recommends the Good Food Institute (GFI), an innovative plant-based and cultivated meat nonprofit. As my colleague Kieran Grieg noted, GFI is generally viewed as a potential giving opportunity for those interested in minimizing animal suffering. However, given that industrial animal agriculture is a leading source of greenhouse gas emissions, GFI could have immense climate benefits. Climate change donors with concerns for animal welfare may be more inclined to give to an organization like GFI than to a nonprofit with equivalent “climate cost-effectiveness” that lacks such co-benefits, and likewise for animal welfare donors who are concerned about climate change. Moreover, aggregating the animal welfare and climate benefits is likely to improve its relative ranking among other organizations in terms of overall impact compared to their ranking when considering welfare benefits for each cause area separately. ↩

40. “While animal agriculture contributes 20% of GHG emissions, not even 2% of the global budget to mitigate climate change is directed towards reducing the global production and consumption of animal products. A recent study by the Boston Consulting Group has demonstrated that ‘for each dollar, investment in improving and scaling up the production of meat and dairy alternatives resulted in three times more greenhouse gas reductions compared with investment in green cement technology, seven times more than green buildings and 11 times more than zero-emission cars.’ And Project Drawdown, which assesses climate solutions, places plant-based diets in the top three of almost 100 options.” ↩

41. Note that I do not believe this option is currently available on GiveDirectly’s donation page, though it may be worth its consideration if there is a low barrier to the addition of such an option. ↩

42. For instance, we have very briefly reviewed a couple of key CEAs and identified errors and/or questionable assumptions. We have also been informed about ways in which some CEAs may have lost relevance over time or require updates to their structures/parameters, yet they still support EA recommendations in their original form. ↩