Promoting climate considerations within existing high priority areas of work

This post is in response to Open Philanthropy’s call for cause exploration work in global health and wellbeing. This post has now been slightly edited to provide additional context and references.

Rather than suggesting a completely novel cause area, I suggest increased focus to the effects of climate change within the existing high priority OP cause areas. By highlighting different bodies of research within each focus area that indicate we should be including considerations about climate change, I ultimately argue current estimates for the cost-effectiveness of different interventions and priority areas need to be adjusted. More concretely, I suggest that lack of preparedness for climate change may make some interventions less beneficial, and that, conversely, addressing considerations of climate change such as anticipating who is most at risk, may make some interventions more impactful. Further, I suggest that embedding climate considerations within currently prioritized causes may advance climate research and help mitigate climate change associated suffering.

Main points

• Climate changes are now becoming a reality of life across our planet. As a multifaceted, multi-systems problem, uncertainties in our understanding are a key issue in appreciating potential impacts and scale.
• While there is a debate whether climate change should be considered as an existential risk, it can at the very least still be more unanimously viewed as an important suffering risk, associated with large-scale impacts over significant territories, countries, and species. 
• At the same time, scientists such as Luke Kemp and colleagues have provided some indication that climate change could still become a global catastrophe (PNAS article here). In either scenario, x-risk or s-risk, further consideration of climate change, mitigation, and protection are warranted.
• Climate change is sometimes considered as a risk factor for some existential risks (e.g. climate change linked instability as a vehicle to increase the likelihood of nuclear conflict) and as such also deserve attention.
• A currently somewhat under-researched perspective asks what other risks might climate change exacerbate. 
• Here, I suggest the impacts of climate change will likely decrease the cost-effectiveness of many interventions, including those that are currently considered highly cost-effective, and those that are considered high priority. 
• At the same time, I propose that embedding consideration of climate change in the context of intervention evaluation, may allow us to find populations at high risk and high need, whilst also improving or maintaining high cost-effectiveness. (I have written about this idea in the context of water, sanitation, and hygiene interventions as well)
• Currently, most climate change interventions are not in and of themselves highly cost-effective, so including climate considerations to existing streams of work could be a way to improve research and intervention within the field.
• Embedding climate considerations within existing cause areas can enhance the work within those respective cause areas and also help mitigate the effects of climate change at the same time. It can also be logistically sensible and potentially cost-effective.

(I note as a broad comment that including “risk considerations” in general for the delivery of any program can be important, such as considering the distribution of a diseases across sex or regional context. Here, I highlight the impacts of climate change as there is reason to anticipate these impacts will be global or vastly spread and sustained over long periods of time.)

Here, I take in turn some of Open Philanthropy’s focus areas and discuss how we might embed climate consideration in impactful ways.

Global Health & Development

• Many of the GiveWell supported programs operate in low and low-middle income countries. These are also the countries most likely to experience high levels of poverty, nonunified governments, as well as lack prevention, social protection and infrastructure measures that could mitigate climate impacts. In most climate projections, the distribution of impacts will be unequal, and most felt by LMICs. Here, I focus on efforts targeting malaria and deworming.
• Malaria Consortium and Against Malaria Foundation work primarily in Africa (Burkina Faso, Chad, Ethiopia, Mozambique, Nigeria, South Sudan, Uganda), followed by Asia (Cambodia, Myanmar, Thailand; India, Indonesia, Nepal), and a few countries in South America and the Caribbean (Nicaragua, Haiti, Peru). From the Long-Term Climate Risk Index research report, Myanmar, Haiti, Mozambique, Thailand make the top 10 list of countries most affected from climate changes in the last two decades, totalling 474 extreme climate related weather events. India ranks highly as well, as a major victim of extreme heath, floods and sandstorms. 
• It is likely that climate change may impact logistics (distribution of bednets) for those currently at risk. It is likely that people protected from malaria will experience future significant income shocks and health risks related to cliamte change. 
• Further, it is likely that prevalence and spread of malaria may be impacted from climate change. Via mathematical modeling, researchers have estimated that small increases in temperature at low temperatures may increase the risk of transmission substantially. This may mean that vulnerable communities, at previously lowered risk, such as in areas where there was previously no malaria or there was malaria at controlled degrees (‘stable transmission’), may be at increased risks of future outbreaks.   If anti-malaria bed nets are successful at achieving stable or low transmission rates, the implication of modeling research is that these efforts need to be sustained over prolonged periods of time and expanded over greater areas in order to maintain effectiveness.
• Deworming initiatives (Sightsavers’ Deworming Program, Evidence Action’s Deworm the World Initiative, The END Fund’s Deworming Program, SCI Foundation) are another stream of cost-effective interventions. GiveWell already factors in some degree of temporal uncertainty in their estimates that could be occur due to catastrophic civilization changes including potentially those of climate change  (see tab ‘Moral weights and discount rates’ in this spreadsheet) at about 1.4% – this is likely an underestimate, given the most likely projected increases by the IPCC (between 1.5-2 degrees) are associated with significant multisystem disruptions in food and water security, as well as increases in mortality and morbidity particularly within Africa. 
• Recent work focusing on time decay of effects (absent of climate considerations) already suggests cost effectiveness of deworming initiatives drops significantly when temporal discounting is more robustly applied.

Biosecurity and Pandemic Preparedness

• The response to the COVID-19 pandemic has illustrated the persistence of global inequalities and the continued struggles and failures of international cooperation. While LMICs have experienced the majority of the disease burden, the same countries lag behind vaccination rates in HICs (see here for an extended comment on the evidence). Climate change is likely to exacerbate existing inequalities, and thus serve as a risk factor for future pandemics. 
• Research within pandemic preparedness that focuses on logistics, equity, and cooperation can have further beneficial effects for other important issues.
• Outside of climate change as a risk factor due its ability to drive inequalities, climate change is a key driver of the dynamics and spread of a variety of infectious diseases, including many that are or could be vaccine preventable (e.g. cholera, yellow fever; see further Mahmud et al.). For instance, temperature and humidity are important variables that can alter the spread of airborne pathogens and temperature and rain can alter the spread of vector borne diseases. Yet, we still understand little about how these risks cumulate and lack rigorous forecasts about the scale of impact and potential tipping points.
• Research into disease spread is already a stream within this cause area for Open Philanthropy. All models need to consider time effects (particularly time discounting in intervention effects),  but we need increased focus on flexibly forecasting different climate scenarios. Thanks to fundamental work such as the IPCC reports, forecasting varied temperature scenarios can be more streamlined in regards to diseases spread but there are still more significant gaps in our understanding for precipitation changes.
• This might also mean that interventions that improve water security, both in the context of climate change and disease prevention, might be particularly important.

Figure from Mahmud and colleagues (2020).  Top panel: mean projected change in temperature (ΔT) and precipitation (ΔP) in 2100 relative to 2000. Bottom panel: plot shows a summary of climate drivers (temperature or precipitation) for different vaccine-preventable diseases and the “best case” modeling effort reviewed, where we assume the best case is a fully mechanistic model using projection data. Absolute humidity drivers are counted under temperature-driven given the functional dependence of the two variables.

• Beyond potential increases in vulnerabilities to diseases, climate change may also disrupt the immunization delivery system. A review by Guo and colleagues (2015)  indicated that climate change and solar ultraviolet radiation can impact how the human immune response operates, as well as how potent and effective vaccine programs may be. A crucial gap is the limited number of human areas, which may be an important way to advance this research, both in terms of climate change and biosecurity.
• I note in passing that public pandemic responses, such as uptake of vaccines and appropriate social measures (e.g. social distancing and masks in the context of Covid-19) are influenced by government and healthcare communication. Ensuring large-scale public readiness should include considerations on how health messaging is delivered, and here too vaccine hesitancy and climate change risk communication are intersecting areas. Shared research can increase the effectiveness and public reach.

Effective Altruism Community Growth

• Much has already been written on the focus of growing the EA community and ensuring diversity of ideas at the same time – see in particular discussions on objections against strong value-alignment between community members, on ‘value lock in’ in the context of longtermism and representation within EA, on EA and tackling racism, and more broadly on the importance of diversity here and here, and a detailed survey on diversity from 2019. All of these threads point to the importance of diversity of people and ideas. 
• Representation from the Global South (or even more broadly LMICs or non-Western countries) in EA is still lacking. At the same time, many of the regions and countries that are under-represented are the same ones at highest risk of climate change damages and associated risks for extreme poverty, socioeconomic and political instability. Engaging with interested potential community members from these countries may be (1) especially important now and increasingly difficult over time, (2) helpful to promote participatory knowledge sharing on climate change, (3) contribute to the overall aims of increasing diversity of ideas in other key areas.
• At present, Open Philanthropy provides a number of funding opportunities within the ‘Community Growth’ focus area. The amounts tend to be significant (and while there’s a separate discussion of arguments against and in favor of high levels of compensation within EA), it is worth pointing out the perhaps obvious implication that the many of these remnunerations are even more significant in LMICs. More targeted outreach to LMICs may be warranted, and potential consideration of distributing more grants perhaps smaller in value.

Farm Animal Welfare

• Within this focus area, Open Philanthropy aims to prevent some of the worst suffering animals can experience, particularly animals farmed for food, such as chickens and fish. Work that could fall within this stream seems to include the promotion of  meat alternatives and policy reforms to improve animal lives.
• That climate change will impact animals should be clear. Recently, research utilizing a machine learning algorithm suggested that ~56% of a database of 7,699 (i.e. just over 4310) species that are currently with an unestablished conservation status are likely in danger of extinction. This is a problem, as protection is usually targeted for species that have an established conservation status. This is an indication that current scientific practices may be underestimating actually scope of impact.
• Protecting cow’s lives may be particularly important, as they are often bred specifically for meat production and linked to increased production in methane, and thus a target for both animal welfare and climate change. 
• Further, some estimate that the clearing of forest land for cattle now accounts for 70% of the deforestation in the Amazon. Forecasters have suggested that continuing these trends will lead to a “death spiral” of the Amazon, once known as “the lungs of the Earth”. Improving cow’s lives, minimizing the amount of cattle reared for meat production, could likely translate into climate benefits as well. This requires system-level changes, such as taxes on beef and legislative and policy action.
• Like the Amazon, some regions may be of an immediate higher priority. A PNAS commentary suggests that the tropics may also be particularly important, and that tropical deforestation may lead to global biodiversity loss.
• This month, with the increase in temperatures, in the UK millions of factory chickens died as a direct result of heatwaves. Trapped in industrial farms, these animals died a slow death of exhaustion in temperatures reaching 45 degrees Celisus and over in their sheds. This previously happened in the 2019 UK heatwaves as well. The executive director of Animal Equality is quoted in response to this news: “The sad irony is that animal agriculture is a driving force behind these heatwaves and other deadly effects of climate change. The meat industry is a leading cause of land and water overuse, pollution, deforestation, species extinction and antibiotic resistance.” This suggests that current harmful practices for animal welfare are linked to climate change and addressing both in a holistic fashion may be more warranted. 

Potential Risks from (Advanced) AI

• There is likely a bidirectional relationship between AI development and climate change impacts. AI development will likely require increased energy, which will have an impact on climate change (AI already requires high energy consumption). This could make renewable energy a priority. On the other hand, climate change can be a risk factor for safe AI development both through infrastructure, logistics, and cooperation problems, as well more directly linked to temperature change issues. In July 2022, UK companies had to turn of high-powered computers during the heatwaves. Heat can damage computers and may exacerbate needs for cooling technologies (also requiring further energy spend). 
• Overall, however, the total impacts of current developments of AI on climate change and vice versa are not established and are an important topic for exploration. 
• Kaack and colleagues, in their recent paper in Nature Climate Change, propose a three category system for evaluation of impacts that AI may have on the environment, particularly through greenhouse gas emission. These categories are: 1) computing-related impacts (e.g. energy and hardware used to train, develop, run algorithm), 2) immediate impacts of applying machine learning (e.g. optimizing energy use to decrease emissions), and 3) system-level impacts (e.g. advertising systems, self-driving cars).
• AI can impact the environment in many ways, both positive and negative. For instance, AI tracking deforestation or the conservation status of species can be a positive and important contribution. At the same time, AI-based advertising likely increases consumption.
• Since Open Philanthropy has an interest in advanced AI, its worth explicitly pointing out that many of the perceived advantages of AI in the context of climate change - for instance, it’s increasingly accurate prediction of floods, windstorms, and heatwaves - could still be misused. States, companies or actors that have the capacity to use AI for humanitarian good may not do so but instead preferred “pay per” models with fees which may lead to coercive control of other states or regions in risks. While hypothetical, this illustration serves to point out that there are legal gaps in how AI services could be legislated and (mis)used. In more extreme example, AI algorithms can underpin geoengineering and thus be linked to further risks, when proper mitigation and harm analyses are not carried out.

Figure from Kaack et al., 2022. Framework for assessing the greenhouse gas emission impacts of machine learning. Green lines indicate reductions in GHG emission, magenta lines indicate increases, and gray lines indicate uncertainties (or likely negligible effects).

South Asian Air Quality

• Within this focus area, Open Philanthropy works within South Asia, where people experience  some of the worst levels of air pollution anywhere around the world. With 1.8 billion people in the region, air quality is a far-reaching problem. Funded work currently includes, for instance, policy outreach and low-cost air quality monitoring sensors.
• That climate change will decrease the quality of air we breathe is likely almost self-apparent. Higher temperatures are linked to increases in allergens and harmful air pollutants, according to the CDC. There is evidence that pollen season is already becoming longer, which in turn means we can likely expect prolonged periods of allergy suffering, longer asthma episodes, and diminished productivity within education and labor forces. The CDC further warns that climate change linked temperature increases can also bring about increases in the dangerous air pollutant, ozone. 
• The National Climate Assessment warns that climate change will likely increase ground-level ozone (main part of smog) and likely diminish many people’s lung function, particularly in areas such as South Asia where people are already at increased risk due to poor air quality.
• It is also worth potentially considering how climate-impacted air quality might affect other regions. For instance, regions in Uganda, China, Mongolia already rank high on lists of most-polluted cities by particulate matter. Intervening in these areas, or helping with prevention, might be cost-effective and save many lives. 
• Air quality and monitoring initiatives are likely more cost-effective when climate considerations are accounted for: they can allow vulnerable people, such as those with pre-existing conditions, allergies, children to limit outdoor activities on certain days or periods and thus hopefully prevent hospitalizations and poor health outcomes.
• But monitoring and tracking alone will not improve air quality, and there is a need for broader level adaptation and mitigation efforts, such as improvements in transportation and land-planning decisions (as those can be particularly harmful in city areas to air pollution) and even more broadly, limiting fossil fuel usage and investing in renewable and other alternative energy sources. 
• A recent European Commission DG Environment report called “Integrated climate change and air pollution strategies: a winning combination” (based in part on Bollen et al.’s cost-benefit analysis of combined pollution and climate interventions) highlights the cost-effectiveness and importance of combined interventions. Global climate change mitigation and local air pollution policies bring about greater CO2 reductions than either strategy alone. Climate change mitigation work can supplement air pollution work by brining additional welfare (co-)benefits, as renewable energies also reduce particulate matter emissions (whereas the report finds local air policies tend to do little about particulate matter emissions overall). The European Commission's report recommends that, due to related air quality and health benefits, climate change policies should not be postponed, and that combined intervention can create a “win-win” situation where “medium-term efforts to control air pollution will support long-term strategies that aim to curb climate change”.