Coronavirus Research Ideas for EAs

post by Peter_Hurford · 2020-03-27T21:01:48.181Z · score: 117 (48 votes) · EA · GW · 20 comments

Contents

  Meta-Research
    1.) Just how bad are things right now? How bad might we expect it to get? What is the current state of play and what are various plausible scenarios forward? [PRIORITY]
    2.) What paths to impact are available to us with regard to coronavirus related work? What is our rough guess of how this impact might compare to typical EA work?
    3.) An insane amount of research is being produced at a rapid pace. How can we best consolidate and synthesize this information and get people up to speed quickly? [PRIORITY]
    4.) An insane number of people are working hard on this, both inside and outside the EA movement. How should we best coordinate? [PRIORITY]
    5.) What else could and should EAs do, other than these research questions?
    6.) Are there places EAs should donate that focus on coronavirus response that are particularly promising to donate to, relative to existing charities EAs like?
    7.) Does EA need a rapid response task force to coordinate inevitable future catastrophes? How can we be better prepared to respond to the next one?
    8.) What do our biosecurity and pandemic prevention specialists think the rest of us should be doing right now?
  Policy Response
    9.) What, more precisely, is the “endgame scenario” for getting out of the coronavirus situation with minimal deaths? What kind of timeline might we be looking at?
    10.) Will there be some situations or risks that could arise that make some of the desired endgames impossible or much more difficult to implement?
    11.) How have various countries, states, counties, cities, private companies, non-profits, civil societies, etc. successfully or unsuccessfully handled the coronavirus policy response so far?
    12.) How might we expect COVID to change policy advocacy over 2020-2022?
    13.) What, if anything, can be done to change coronavirus response policy for the better? What about for future pandemics / crises?
    14.) Low- and middle-income countries seem at particular risk of not being able to handle the coronavirus problem. How bad will this situation be? What should be done to mitigate it? Do we have any viable paths to impact?
    15.) What is the actual cost-benefit of various pandemic responses? [PRIORITY]
    16.) What opportunities does this give for us to improve institutions and policies more broadly?
    17.) Are there any ballot initiatives we should consider for 2022 and 2024?
    18.) Polling could be very valuable right now. What questions should we be asking the general public and how can we leverage their responses to effect change?
  EA Community Health
    19.) Does coronavirus pose a risk to the cohesion and stability of effective altruism movement? If so, how should we mitigate this? [PRIORITY]
    20.) Are local groups still able to operate ok? What implications does this have for EA local groups strategy? Are any local groups at risk of collapse?
    21.) A recession is something EA has never faced before. Could an associated fundraising shortfall lead to the collapse of any EA orgs? What can we do to mitigate this? [PRIORITY]
    22.) Much EA-aligned “direct work” is severely disrupted or cannot happen. How does that affect these orgs? What should these orgs do differently? [PRIORITY]
    23.) Many EAs and EA orgs are now suddenly working remotely for the first time. Do they have the resources they need to be effective?
    24.) Are there other risks EAs should be preparing for that we aren’t?
    25.) Could the EA movement use the existing groundswell of altruistic action to form better partnerships with non-EAs?
  Existential Risk Implications
    26.) Is this pandemic a catastrophic risk threat multiplier? How should this change EA’s approach to risk mitigation?
    27.) How does this pandemic affect the balance of great powers? What implications does this have? Is this important to look into right now?
    28.) What are the implications of widespread expansion of surveillance? Are there ways we can mitigate any risks?
    29.) Could massive expansion of largely unrestricted funding to biolabs and other research produce any “dual-use” concerns?
    30.) It seems like the world is now aware, perhaps for the first time in modern history, that catastrophic risks can happen. How should we seize this moment?
    31.) What biosecurity policies should we try to implement?
    32.) What does the COVID outbreak teach us retrospectively about how we should’ve handled this risk? Were the right systems in place?
  Animal Welfare Implications
    33.) It seems like the world is now aware, perhaps for the first time, that factory farming might pose severe health risks. This could be the biggest opportunity for a pro-animal campaign in modern history. However, mistiming this campaign could risk the biggest backlash to the animal welfare movement in modern history. Given this, how should we respond to this moment? [PRIORITY]
    34.) How do we shift animal welfare work to the new online-only world and keep the animal movement from being curtailed?
    35.) COVID has had implications for animal welfare. Does this create any new sources of animal suffering that we should take action on? Does it threaten to create any new risks that we can try to get ahead of?
    36.) What will be the medium-term effect of COVID-19 and associated economic consequences on the demand for and growth of plant-based products?
  Other Implications
    37.) Does having widespread lockdowns change the balance of cause areas or create a “Cause X”?
  More?
None
20 comments

COVID-19 is a tragedy with more everyday social implications in the developed world than anything since World War II. Many EAs are wondering what, if anything, to do about COVID to help the world. To try to investigate further, I am helping articulate possible research ideas for further discussion and consideration.

The kind of research we need to do in this situation is very different from the kind of research EA is used to doing well. We normally spend several months carefully researching a single topic that doesn’t change very much. With COVID, everything about this is reversed—the situation currently requires us to rapidly get up to speed and produce research in a matter of days in a situation that is rapidly changing. There is an exponentially growing “speed premium”—much more than we’ve ever seen. As such, please note that we have waived our normal review and quality check standards to get this out ASAP and there may be considerable mistakes in this article.

On the other hand, I do urge some degree of caution and humility. Let’s not all collectively lose our minds. We should worry at least some about armchair epidemiology from non-experts (though also see this [LW · GW]) and properly recognize what our skills are and aren’t, where we can contribute and where we shouldn’t. We should also be careful that research done at breakneck speed is more likely to be wrong.

I'm a bit worried that many people will want to work on this topic just so they don't feel helpless in the face of the pandemic, or because there’s a lot of attention being paid to it now, or that it feels high-status and urgent, or many other reasons unconnected from EA-related impact. It can be really tough to see your community, friends, family, and self hurting and not feel like there’s much you can do. However, the work EA was doing before this pandemic still remains of importance now. If you can contribute to the anti-COVID effort that is great, but it is also fine (sometimes even preferable!) to continue to research what you were researching before.

Naturally, these research questions were put together rapidly and may continue to evolve rapidly. I numbered the questions to make them easier to reference and discuss. They’re grouped by topic. Note that the numbering system may get a bit weird as I add new questions without wanting to renumber all the other questions. I am trying to keep numbers stable so they can be referred to as shorthand.

Questions vary a lot in their urgency, importance, tractability, and neglectedness. If you’re already up to speed, I think questions 1, 3, 4, 14, 15, 19, 21, 22 and 33 are particularly important and urgent to look into right now.

I’m not sure how best to coordinate around these and other questions... the question of coordination itself should be the subject of active research (see question 3). For now, feel free to comment here, or centralize on LessWrong and the “Effective Altruism Coronavirus Discussion” FB group.

If you intend to dedicate significant effort to any of these questions, it might also be worth you joining our cross-org Slack group—comment here or email me (peter@rethinkpriorities.org) to be added.

I do think we should try to communicate early and often about our progress and aim to produce results quickly and share with others.

Note that this isn’t the only compilation of coronavirus research questions. I also know of LessWrong’s agenda [LW · GW] and this list of questions [EA · GW].

I will try to keep this all up to date as rapidly as I can.

Meta-Research

1.) Just how bad are things right now? How bad might we expect it to get? What is the current state of play and what are various plausible scenarios forward? [PRIORITY]

The COVID-19 pandemic is a rapidly escalating problem. Where are we at and where might we be going? How will this burden be faced on a country-by-country basis? It seems like there is a lot we don’t know and we could be making decisions on unreliable data. On the other hand, we do know enough to act.

This will be relevant for the EA community too. EAs are used to measuring things in terms of importance, neglectedness, and tractability. In some ways, COVID-19 is an unprecedented tragedy, but—to be honest—EAs are sadly used to tragedies and have been triaging efforts. Right now I do not think the EA response to coronavirus is an overreaction but we should consider whether it is. EA work has tremendous opportunity cost. Should we even be working on this?

Furthermore, there is a huge amount of work on this already and I don’t expect most EAs to be well placed to produce original research on this topic. However, getting up to speed and informing ourselves of the problem via synthesis of existing research seems important.

Additionally, I think one of the comparative advantages of EA is the ability to look at long-tail distributions and take seriously the possibility of low-probability, high-impact events. There may be specific low-probability scenarios that aren't getting enough attention and I think EAs would be good at finding these. We should be careful to make sure that we understand the uncertainty in both directions.

Existing work: The LessWrong Coronavirus Database [? · GW] has a bunch of links for getting up to speed. Rethink Priorities has done some existing work focusing on caseload and response by US states and are working on aggregating the same data for countries. There is also this forecasting model by FHI and this other model by some EAs.

The importance / neglectedness / tractability axis requires more than just establishing importance (see question 1) and neglectedness (definitely not neglected overall, but maybe neglected in specific efforts). What can we do, if anything, that is tractable?

The broad EA network is surprisingly well-connected and punches above its weight, but I don’t think we’ve ever had to test this network at such a scale and at such speed as before. Does EA actually have a bunch of political connections we can leverage?

Learning about this might be very valuable not just now but for future efforts to improve institutional decision-making and policy.

3.) An insane amount of research is being produced at a rapid pace. How can we best consolidate and synthesize this information and get people up to speed quickly? [PRIORITY]

There is an awful lot of information on COVID-19—probably more has been written in just two weeks than many topics see in a decade. Staying on top of all the research could easily be the full-time job of several people. However, it seems like the vast majority of this does not need to be read by each potential researcher. Can we carefully get the highest quality work? Can we rapidly keep it up to date?

Information needs will probably differ a lot by researcher and the desired research question, so tags / categories might be particularly helpful. Additionally, we may also want to sort links by question, as has been done informally in this document.

This question seems much more tractable than the others for the average EA. Synthesizing research, especially if it can be done in a catchy infographic, could be a much better option for most people than original research.

Existing work: Right now, the best attempt I know of is the LessWrong links database [? · GW]. We should probably focus our efforts on improving and updating that. It could also be good to keep up to date high volume wikis like the Coronavirus Tech Handbook and Wikipedia that a lot of people are looking to.

4.) An insane number of people are working hard on this, both inside and outside the EA movement. How should we best coordinate? [PRIORITY]

I’ve never seen so many people in the world working at breakneck speed on the same problem. This is very inspiring, but also makes coordination difficult. I’ve seen many groups duplicate effort and I’ve spent close to half my time just trying to help resolve these coordination problems.

Right now, I think existing public EA research efforts should centralize (a) on LessWrong and (b) the “Effective Altruism Coronavirus Discussion” FB group.

While I imagine some of this duplication is inevitable (and may even have some upside, as it lets quasi-market forces identify good work), I think it could and should be the full-time job of one or more people to join as many coronavirus groups as possible (both in EA and not), read all the project aggregators, etc., look for duplication, and make sure everyone is aware of the other group’s effort—it could save a ton of time in the long run! As far as I know, no one is doing this full-time yet.

5.) What else could and should EAs do, other than these research questions?

6.) Are there places EAs should donate that focus on coronavirus response that are particularly promising to donate to, relative to existing charities EAs like?

It seems like there is a significant amount of philanthropic interest to fight COVID-19 both inside and outside EA. Are there any promising donation targets?

Some EA mainstays have reoriented toward COVID impacts. As of 25 Mar, AMF announced that their funding needs are more dire and they claim their impact is higher due to COVID-19 as malaria and COVID have comorbidity effects.

GiveDirectly is helping give cash to Americans hardest hit by COVID-19’s economic effects. While this probably is not as good as GiveDirectly’s normal work, it may be a compelling and strong opportunity for people who wouldn’t otherwise be donating to help the developing world.

It’s also possible that some EAs organizations might be hit unusually hard by an upcoming recession (see question 21) and might be in dire need for donations. Perhaps the EA community should start co-ordinating better on fundraising (i.e. everyone announcing their RFMF transparently and collectively at the same time, rather than a rush of individual requests leading up to giving season).

Existing work: GiveWell is going to be working on this, SoGive compiled a list with some suggestions [EA · GW], and Founders Pledge has recommended the John Hopkins Center for Health Security. 80,000 Hours also compiled a list of giving advice also suggesting the Center for Health Security at Johns Hopkins, in addition to Gates Foundation COVID-19 funds, and the Center for Global Development.

7.) Does EA need a rapid response task force to coordinate inevitable future catastrophes? How can we be better prepared to respond to the next one?

It seems like many EAs were in a good position to identify this crisis 1-8 weeks earlier than the general public and even earlier than the US CDC. This led to a lot of individual preparation, but to the best of my knowledge it did not lead to much preparation on the policy / research / outward-facing side. Is EA well-positioned to be able to respond in a crisis when speed is of the essence?

It seems great that LessWrong was able to rapidly pivot to coronavirus coverage and fund Elizabeth Van Nostrand to be a full-time coordinator. Other organizations, like 80,000 Hours and FHI have been able to pivot quickly as well. Do we have enough funding to stay nimble and do this? Should we be doing anything more to be in a better position?

8.) What do our biosecurity and pandemic prevention specialists think the rest of us should be doing right now?

I’m hoping 80,000 Hours, FHI, and OpenPhil have this covered. I’d be especially curious what we can do as EAs that these experts might not be able to do for reputation preservation reasons.

Policy Response

9.) What, more precisely, is the “endgame scenario” for getting out of the coronavirus situation with minimal deaths? What kind of timeline might we be looking at?

I think a lot of us are familiar with the basic “hammer and dance” strategy that is, more or less, going forward, but what will this actually look like in practice? How long do we expect to be in lockdown? What do we expect will get us out safely? How do we expect “test and trace” to roll out in the US / UK / EU like it has in Singapore, South Korea, and Taiwan? (See also this Vox summary.)

We might be interested in understanding (and forecasting, see question 5) expected timelines of various treatments. This forecast suggests some good data on treatments will arrive in April, artificial antibodies potentially in the fall, new antivirals by early 2021 at the latest, and a vaccine by late 2021 at the earliest. Is that forecast sound? Can we improve upon it?

There’s a lot of work happening rapidly and it is hard to keep track. (Here’s an overview of some current approaches as of 19 March.) Will there be a “Manhattan Project”-level effort? Oxford is doing human trials for vaccine candidates now, but others are skeptical of optimistic vaccine timelines and think they might be awhile. (See also.)

How much is the vaccine timeline hampered by safety regulations (e.g., Phase I trials) rather than actual R&D? Should the safety regulations be modified in some way? Could a human challenge study be helpful? Could we benefit from spurring more prizes for the first vaccine to meet certain requirements? See some discussion here [LW · GW].

Do note that here I think a lot of good research will be done by domain experts and it’s not clear to me that the EA community can improve too much on this question. Instead, EAs may want to focus on synthesis to understand the current “state of play” and inform other questions. It’s also possible EAs might want to explore what ways they can use their resources to speed up existing research by domain experts.

Some of these forecasting questions are probably best put into Metaculus (like this question, but for more, and maybe using Metaculus date ranges).

10.) Will there be some situations or risks that could arise that make some of the desired endgames impossible or much more difficult to implement?

The political will to maintain shutdowns could waiver. Reinfection risk across borders could also be a major issue. How bad are these problems? Are there other issues? What can be done about it?

11.) How have various countries, states, counties, cities, private companies, non-profits, civil societies, etc. successfully or unsuccessfully handled the coronavirus policy response so far?

All countries are faced with coronavirus and need to figure out their response. Many countries have taken different strategies and we can learn from the successes and failures about how to respond better. If we collect this rapidly enough, we may even be able to learn in real time and make our current response policy better.

Also, there seems to understandably be a lot of focus on governments but individuals, orgs, businesses, etc., seem like they were often way ahead. Why was that? For example, is it weird that Apple had more N95 stockpiled (9M) than the US Strategic National Stockpile (5M) (seems to have largely been wildfire prep rather than pandemic prep)?

Existing work: IMF has put together a fairly comprehensive list of current economic policy responses. Some folks at 80,000 Hours, FHI, and Rethink Priorities are looking more into this.

12.) How might we expect COVID to change policy advocacy over 2020-2022?

It seems like COVID could define politics for the next few years, similar to how the 11 September terrorist attacks did (at least for the US), if not even more (though also, possibly, less!). Will there be civil movements advocating for biosecurity? Will there be widespread xenophobia? How will organizations such as ALEC and the Council of State Governments work on policy?

13.) What, if anything, can be done to change coronavirus response policy for the better? What about for future pandemics / crises?

Sure, we may have a lot of good ideas (we think), but what kind of political levers do we have as a movement to actually influence policy? Can those be used? Who can use them? How tractable are they? How do we centralize around those efforts?

We should understand that any campaigns to get governments to act in particular ways will easily be drowned out by the many other people seeking to do the same thing, and our campaigns won’t work unless we carefully target them and work through high-level contacts who can actually make a difference. What can we actually do? What is our “theory of change”?

That being said, I definitely think we will need sustained, well-targeted public pressure and advocacy around maintaining the shutdowns that various governments have put in place or risk coming out of shutdowns too early due to short-sightedness (assuming the shutdown strategy is correct, see question 15). If you have any celebrity contacts, encourage them to also keep the pressure up and emphasize that the shutdowns are worth the considerable cost (again, assuming that is true).

See also question 16 for more we could advocate for.

14.) Low- and middle-income countries seem at particular risk of not being able to handle the coronavirus problem. How bad will this situation be? What should be done to mitigate it? Do we have any viable paths to impact?

At least as of 22 March, COVID cases are closely tied with GDP. I don’t think that the coronavirus mainly targets rich countries (though, with globalization, that might be true to a minor extent). Instead, I think that low- and middle-income countries (LMIC) are currently unaware of the extent of the epidemic and will continue to be hit hard due to poorer governance and health infrastructure (though might also be benefited by a much younger population). For example, some describe India’s attempts as “working”, while others use the term ”looming crisis”. What should we expect?

Many global health think tanks (for example, KFF) are already working on this so it might not be all that neglected. Does EA have anything to add here?

Existing work: Some discussion on FB. The EA India group has also undertaken some coordination for India-specific response… not sure if there is any active EA work in any other LMIC, though the EA Hub lists groups in Lagos (Nigeria), Pakistan, Kisumu (Kenya), Kuala Lumpur, Turkey, and Sao Paulo (Brazil).

15.) What is the actual cost-benefit of various pandemic responses? [PRIORITY]

Extreme social distancing measures could potentially save millions of lives if done right. It seems to be working. But there are high social and economic costs to such measures (see also). We must acknowledge that we are engaging in a massive global social experiment which has never been done before and we are wading into uncharted territory.

One analysis suggests the cost-benefit might be as much as $11M per life saved. Another analysis suggests we have already sacrificed ~10M lives in expectation through current lockdown measures.

In the middle, one analysis finds mixed benefits for different lockdown measures. Another model finds that the outcome hinges a lot on the specific policy and the percent of people that are asymptomatic carriers. Yet another model also echoes policy specifics as mattering a lot.

On the positive side, Kyle Bogosian’s model suggests social distancing creates $7500 in economic benefits per person, which seems good.

Furthermore, Rob Wiblin outlines some reasons to consider going forward with lockdowns even if it is considered quite plausible that they may be overly harmful.

Rob Wiblin also has more at his new Victory over COVID site.

Still, there isn't really a definitive, interpretable cost-effectiveness model that adequately accounts for uncertainty and I think it's within the EA wheelhouse to try to produce one.

Right now, the lack of good cost-benefit research is alarming though understandable—the data is scarce and there hasn’t been that much time to put these kinds of things together. However, this should probably be a priority, as we are currently flying blind and we should be open to the idea that some of the policies we are advocating for might be counterproductive. Do we have anything to add to this discussion?

16.) What opportunities does this give for us to improve institutions and policies more broadly?

I want to be careful not to fall into the mindset of “the pandemic clearly shows why all my preferred policy opinions were correct all along”. On the other hand, this could be a chance to move toward making some important policy changes. What should we try to make happen and what levers do we have to pull?

17.) Are there any ballot initiatives we should consider for 2022 and 2024?

COVID might create a political moment that might create new groundswells of large-scale political support that could be channeled into ballot initiatives [EA · GW], where relevant. What kind of campaigns might we expect to be successful now that we wouldn’t have suspected just a little while ago?

18.) Polling could be very valuable right now. What questions should we be asking the general public and how can we leverage their responses to effect change?

Between an Echelon Insights poll and a YouGov poll (and likely many other polls that have run that I haven’t seen yet) we can understand some of the social costs and see that the lockdowns enjoy broad popularity, at least in the US. What else might we want to know?

Existing work: This EA academic collaboration. Spencer Greenberg has been doing a lot of polling. Rethink Priorities and Faunalytics are doing some polling specific to factory farming (see question 33). Rethink Priorities also has the capacity to rather cheaply do a lot more polling in the US and UK.

EA Community Health

19.) Does coronavirus pose a risk to the cohesion and stability of effective altruism movement? If so, how should we mitigate this? [PRIORITY]

People’s lives are being disrupted en-masse to a level not seen in the industrialized first world since World War II. It’s hard to focus on altruistic pursuits when one’s own livelihood or family is threatened. Also, while the EA movement has been an online movement to a large extent, it was also dependent upon physical offices and physical meet-ups. In tense times, disagreement and heated discussions can be even more hard to have and feelings of elitism and unwelcomeness may be magnified. Lastly, the looming recession could pose a further threat to many individual EAs and EA organizations (see question 21).

On the other hand, the common cause of COVID has seemed to create a rallying effect that might overall enhance social cohesion. Generally, disasters spur more pro-social behavior, not less.

Do we see any of risks posing a threat to the EA movement—curtailing or eliminating it in some way? More importantly, what community health steps can we take now to mitigate damage into the future?

20.) Are local groups still able to operate ok? What implications does this have for EA local groups strategy? Are any local groups at risk of collapse?

Until recently, most local groups and local group outreach had been done in-person. Without the ability to meet up physically and actively recruit from physical contact, will many local groups dry up or die off? Will local groups still be able to hold online events and will these be sufficiently attended?

Existing work: Some smaller local groups are pooling together into coordinated online meetups.

21.) A recession is something EA has never faced before. Could an associated fundraising shortfall lead to the collapse of any EA orgs? What can we do to mitigate this? [PRIORITY]

Effective Altruism mainly came to exist around 2010… this comes after the 2008 Recession and means that EA has largely never existed within a major economic downturn. In the US, the Great Recession generally reduced donations by ~7% for two years. The upcoming COVID Recession may well be worse. Additionally, some donors may waver from supporting their original charities to supporting COVID-focused work instead. On the other hand, EA donors tend to be wealthier and more committed to maintaining donations and sticking to their original targets than the American average. The overall impact on EA is fairly unknown at this point.

I think there will be casualties. As Abraham Rowe notes [EA · GW], a major recession will probably negatively impact smaller EA charities, especially those not funded by OpenPhil. We should try to get out ahead of this as much as possible, figure out best practices, and brace to weather out a storm. Regular donors should immediately notify non-profits if their donation plans are changing and non-profits should proactively work with their donors and build out reserves.

Additionally, economic support provided by the government may be helpful to EA charities. For example, there is also a 10% wage subsidy coming in Canada for small businesses who don’t lay off employees. The new $2T US relief act also contains some provisions that may help non-profits (see here, here, and here). Someone should look into this and figure out what new opportunities are available and disseminate this information to other orgs.

22.) Much EA-aligned “direct work” is severely disrupted or cannot happen. How does that affect these orgs? What should these orgs do differently? [PRIORITY]

A global policy for social distancing has dramatic implications for how organizations interact on the ground. GiveDirectly has stopped working door-to-door and is now also making direct cash transfers for Americans. The Humane League has cancelled all in-person events. AMF has claimed some disruption and finds their fundraising needs even more dire and their work even more important, given the comorbidities of malaria and COVID.

Existing work: GiveWell is looking more into this for global health work. See questions 33 and 34 for animal welfare implications.

23.) Many EAs and EA orgs are now suddenly working remotely for the first time. Do they have the resources they need to be effective?

There are a lot of remote guides out there, including some informal tips from Rethink Charity and Rethink Priorities staff. We should try to distill all of this and disseminate it. I’d like to see a more scientific tips article along the lines of this for burnout [EA · GW], except for working and managing remotely. I also think we should work to directly check-in and coach orgs with implementation of best practices.

24.) Are there other risks EAs should be preparing for that we aren’t?

The EA community, like all other communities, has already faced incredible strain from home quarantine and minorly disrupted supply chains. For the most part, we did a good job of preparing for that and I think we as a movement were clearly 1-8 weeks ahead of most other communities on this. Are there other things we need to be 1-8 weeks ahead on?

For example, it is possible, though currently unlikely, that the pandemic could cause further food shortages and supply chain disruptions. Rob Wiblin finds this unlikely for a variety of reasons, but Metaculus puts the probability of a food shortage in a major US, UK, or EU metropolitan area at 30%! I personally think the odds are closer to 5% (and I put “8%” into Metaculus myself), but even this implies that it would be good for EAs to prepare. I have personally stockpiled an additional 14,000kcal of food per person (mostly soup, peanut butter, and Soylent) beyond our normal household food supply. (I estimate ~42,000kcal per person of food in our house normally.) I don’t want to horde to an anti-social degree, but I also want to make sure my family is personally prepared.

We might also want to be prepared for possible civil unrest and riots.

25.) Could the EA movement use the existing groundswell of altruistic action to form better partnerships with non-EAs?

We want to be really careful about optics, but with a huge groundswell of altruistic action, we might want to increase collaboration and learning between EA groups and non-EAs and also interest people in our way of thinking.

Existential Risk Implications

26.) Is this pandemic a catastrophic risk threat multiplier? How should this change EA’s approach to risk mitigation?

An enumeration of possible threats has been omitted due to infohazard concerns, but you can probably get the idea. While we are all busy responding to a pandemic and also bound by social distancing, might this provide a ground for another natural or artificial crisis to be many magnitudes worse?

27.) How does this pandemic affect the balance of great powers? What implications does this have? Is this important to look into right now?

The Eurasia Group identified “The Great Decoupling” between China and the US to be one of the biggest geopolitical risks in 2020 and found the risk to be dramatically heightened by coronavirus. People in the US are rethinking China-centric global supply chains while China has been seeing widespread distrust of foreigners (and I’m sure there is increased xenophobia on the non-China side as well). China has engaged in a massive disinformation campaign intentionally and falsely blaming the US for starting coronavirus and spreading it to China (see also). Axios says “U.S.-China tensions hit a dangerous new high”. FP Magazine says “Virus Competition Is Wrecking China-U.S. Cooperation Hopes”.

Meanwhile, Russia and Saudi Arabia have started a trade war over oil prices that doesn’t look set to end anytime soon.

We might want to be particularly on the lookout for threats to election integrity / security.

What should we be looking out for? How does this affect national security, as well as the overall balance of democracy vs. authoritarianism?

What impact does this have in warzones, on the willingness of the Great Powers to commit boots on the ground for aggressive, defensive, or humanitarian interventions? Does it increase the incentives to use drones, if human soldiers are social distancing?

28.) What are the implications of widespread expansion of surveillance? Are there ways we can mitigate any risks?

Some of the policy responses to COVID have been widespread increases in surveillance—such as China, South Korea (see also, see also), Taiwan, and Israel. The dilemma has been summarized as choosing between “panopticons or pandemics”.

The Electronic Frontier Foundation has put out some notes on maintaining privacy while still using surveillance.

What are the implications of this? Could this accelerate risk around some negative totalitarian state singleton event?

29.) Could massive expansion of largely unrestricted funding to biolabs and other research produce any “dual-use” concerns?

Expanding our biotechnology might come with acceleration of risks around biosecurity, especially with “dual-use” technologies that can be used for both peaceful and disruptive aims. Details here are again omitted due to infohazard concerns.

30.) It seems like the world is now aware, perhaps for the first time in modern history, that catastrophic risks can happen. How should we seize this moment?

This situation is a good reminder of the necessity of books like Toby Ord’s “The Precipice”. While we should be very careful about optics, there might be a way to politely point out those of us who have been talking about these kinds of risks for decades. Maybe people might be more willing to listen to us?

31.) What biosecurity policies should we try to implement?

COVID-19 is probably a very big window of opportunity for biosecurity policy. Are we prepared for this? I’ve heard that the current International Health Regulations that were adopted in 2005 were made possible by the momentum generated by the 2003 SARS outbreak… can we do something similar? Does EA have a role to play here? (See also questions 8 and 13.)

32.) What does the COVID outbreak teach us retrospectively about how we should’ve handled this risk? Were the right systems in place?

Metaphorically, COVID was not a black swan… it was a grey rhino. It seems like there were a lot of signposts for this catastrophe, and many EAs did notice these. Do we have a general system to detect, alert, and act on these warnings? Did we act in the right way?

There's a lot of commentary about how there was evidence about this coming catastrophe (from Bill Gates to academic research specifically pointing to this SARS-CoV-like potential, see also). How can we make sure that this higher level mistake (not being ready for a disaster) doesn’t happen again?

What can we learn about how COVID was handled to see how we should handle other grey rhino events? See this discussion here on the EA Forum [EA · GW] and this article and question 7.

Animal Welfare Implications

33.) It seems like the world is now aware, perhaps for the first time, that factory farming might pose severe health risks. This could be the biggest opportunity for a pro-animal campaign in modern history. However, mistiming this campaign could risk the biggest backlash to the animal welfare movement in modern history. Given this, how should we respond to this moment? [PRIORITY]

The COVID situation has already inspired hundreds of articles tying it to factory farming in some way. The National Review, a conservative magazine, has run some of the strongest animal welfare articles yet, calling for an end of wild animal trade and an end to factory farming. This could be a big moment and many organizations are running with it.

At the same time, many more Americans are feeling food insecure and the farming industry is touting their status as an essential service that deserves protection in these trying times. We might risk a backlash if we mis-time the message. What are the actual risks here?

Tracking price and consumption trends could also be informative and cheap.

Existing work: Faunalytics and Rethink Priorities are separately trying to rapidly run surveys to understand this situation.

34.) How do we shift animal welfare work to the new online-only world and keep the animal movement from being curtailed?

Much of the activism animal advocates took for granted in February is no longer possible with social distancing. The Humane League has cancelled all in-person events. Some groups are pausing their corporate campaigns and legislative efforts.

It’s obviously much more difficult to do undercover investigations. The oversaturated media makes media-based advocacy harder. The overburdened governments makes political advocacy more difficult. What are the implications? How does the animal movement handle this? What should they do instead? See also questions 22 and 33.

Also, the looming economic recession will likely impact animal welfare organizations’ funding worldwide - and not just in the EA community. Many local organizations, especially in regions where the animal rights movement is relatively young, will be forced to close or to reduce their work significantly, which could be a significant setback to the movement. Is there anything we can do to ameliorate this? (See question 22.)

35.) COVID has had implications for animal welfare. Does this create any new sources of animal suffering that we should take action on? Does it threaten to create any new risks that we can try to get ahead of?

While we focus on factory farming messaging more broadly, there may be many underrated impacts on animals that weren’t true just a month ago. Given the size and scale of animals, these could produce very large new concerns. Hungry monkeys are overrunning parts of Thailand. Similarly, urban animals (e.g., pigeons, rodents) are desperate for food. Are there any other negative animal impacts we should be aware of that weren’t on our radar? Is there anything we can do to ameliorate these?

Existing work: Rethink Priorities researcher Daniela Waldhorn has started collecting various news articles about various animal impacts (see also this FB thread).

36.) What will be the medium-term effect of COVID-19 and associated economic consequences on the demand for and growth of plant-based products?

Demand for eggs and meat are up massively amid panic buying, though demand for plant-based foods is even higher and there are some concerns about what kind of conclusions we can draw from this data. It seems that the effects will be different for the market in the US / Europe than for the Chinese market.

One financial analyst suggested that Beyond Meat’s overexposure to the foodservice sector could be bearish for its stock outlook.

Other Implications

37.) Does having widespread lockdowns change the balance of cause areas or create a “Cause X”?

With a third of the world's population currently under lockdown, the world is very different than it used to be and this changes a lot of things:

~

More?

More updates are available in my LW 2 Apr Consolidated Brief [LW · GW] and my LW 29 March COVID Consolidated Brief [LW · GW].

When I can, I will also aim to update this post with what I find relevant.

~

Thanks to Luisa Rodriguez, Jason Schukraft, Derek Foster, Neil Dullaghan, David Moss, Kieran Greig, Justin Shovelain, Marcus A. Davis, Daniela Waldhorn, Max Daniel, Elizabeth Van Nostrand, Nicholas Joseph, Kim Cuddington, Oliver Hybrka, Linh Chi Nguyen, Siobhan Brenton, and Raymond Arnold for valuable contributions and discussions around these research questions. Notably, a few of them have made comments which I occasionally outright plagiarized in the interests of speed. Their rapid response and tireless work this week is deeply appreciated. Note that including their names here does not necessarily imply their endorsement or support.

20 comments

Comments sorted by top scores.

comment by Sean_o_h · 2020-03-28T09:52:10.920Z · score: 14 (8 votes) · EA(p) · GW(p)

Thank you for writing this up; it's extremely helpful, especially in such a rapidly developing space. A very optional request: might you consider updating this e.g. once a week with significant relevant developments on these ideas/questions? With so many of us involved in many different ways, it could provide a helpful evolving roadmap. Feel free to ignore if too much hassle or redundant with summaries elsewhere.

comment by Peter_Hurford · 2020-03-28T16:45:03.332Z · score: 5 (3 votes) · EA(p) · GW(p)

Yes, definitely, thanks! I will aim to keep it up to date as best and as quickly as possible.

comment by Sean_o_h · 2020-03-28T16:56:30.920Z · score: 2 (1 votes) · EA(p) · GW(p)

Thanks Peter, that's awesome!

comment by RomeoStevens · 2020-03-27T21:57:53.406Z · score: 12 (8 votes) · EA(p) · GW(p)

A lot of people are willing to try new things right now. Rapid prototyping of online EA meetups could lead to better ability to do remote collaboration permanently. This helps cut against a key constraint in matching problems, co-location.

comment by Peter_Hurford · 2020-03-28T16:46:56.381Z · score: 6 (4 votes) · EA(p) · GW(p)

This is a really good point. I'd love to see more people look into this.

One thing I've really liked is Zoom breakout groups... I think if done right they can actually be better than in-person meetups for coordinating discussion among >20 people.

Though apparently Zoom's privacy sucks (see also).

comment by Peterslattery · 2020-03-28T00:36:16.863Z · score: 9 (7 votes) · EA(p) · GW(p)

Thanks for sharing, this is a pretty incredible list. The bar for good EA work just keeps on rising.

I have gratitude fatigue from all the people/efforts I am grateful for :)


For 3 - I think that shared Mendeley databases, are a good choice for synthesising and sharing research. New research is suggested, the web clipper makes it easy to add and the shared group mean that many people can contribute. It can also be combined with research updates from Scopus etc. I do one for Behaviourworks (my employer) for behaviour change research - see here for an update. I am happy to explain how to do it if 3+ people comment that they expect they would probably benefit it.


For 13/18 - Thanks for the mention of our project - maybe we can collaborate on getting UK data?

We just collected our first set of Australian data (n=1000) for an open science project. We aim to help policymakers with decision making about how to allocate resources to tackle Covid-19.

We doing something quite different from other ongoing surveys. Most of these are about country level comparisons and not about understanding behavioural drivers. For example, we can report things by location and demographic, but also why people are not doing the behaviours, e.g.,

  • 75% of males between 30 and 40 are social distancing but only 50% of males between 20-30.
  • At only 55% adherence, the inner west region reports the lowest amount of social distancing. The main capability barriers are commuting and desire to see friends.
  • 97% of those surveyed are always washing their hands, suggesting that this need no longer be a key communication target.
  • Maroubra shows the lowest levels of life satisfaction and the highest levels of anxiety
  • Compared to the previous week, emotional concern and risk perception have increased significantly, the population thinks more about Corona and searches for information more often.
  • Chronically ill people have a higher risk perception and need special support to protect themselves effectively.
  • The four biggest concerns are health system overload, small business filing for bankruptcy, recession and society becoming more selfish. Measures and information by the government can create security here; Reports of projects based on solidarity may encourage imitation and can reduce social fears.

The list of the 40 or so people currently involved is here.

If anyone would like to know more, take the survey, or help, then please read the following and check the links (copied and pasted to save time - hope that is ok - I am near exhaustion :P!):

Invitation letter

Dear X,

I am part of an open science research collaboration creating an international and continuously-updated dataset to help government and public health officials make better decisions to tackle Covid-19.

Existing COVID-19 related surveys are less optimised for providing immediate actionable insights to policy makers, such as the differences in adherence to key protective behaviours, and the barriers preventing people from performing these behaviours, and how these differ by postcode or demographic. Such data is therefore critical for efficient resource allocation but not currently available.

Building on the COSMO project, a WHO/Europe initiative - we will conduct a “living survey” -- with both repeated cross-sectional and longitudinal sampling -- running throughout the pandemic, to track relevant protective behaviours, their variations by demographic and location, and their determinants.

After each wave of data collection, we will generate and disseminate an updated report about the prevalence of protective behaviours, their most important drivers or barriers, and a break-down by demographics. As we collect more waves, we will be able to visualise trends over time.

We will also test the effectiveness of different interventions at encouraging relevant behaviours. Here are some examples of the types of outputs that we will be able to produce.

We have just started collecting the first wave of data.

Please consider joining us in collaboration. You can contribute by:

  • Helping to disseminate the survey via social networks or panel data.
  • Reaching and helping policy makers in your country with the data we collect
  • Helping to modify our report template to provide useful and interesting information to policy makers.
  • Developing reports and doing analysis for policy makers
  • Providing us with feedback based on your discussion data collection
  • Helping with write up and dissemination when we seek to publish this work

If you contribute to this project in any significant way then you will be recognised on all outputs and be an author on any subsequent paper. The bar for recognition will be relatively low (perhaps ~5 hours of work).

The data we collect and share will not only be invaluable to current policy and decision making but will also provide insights that help to prevent suffering and death.

You can take the survey here [Add your personal link]?Firstname_Lastname_EMAIL

You can join the collaboration here

Best wishes,

X

comment by Peter_Hurford · 2020-03-30T18:29:53.975Z · score: 4 (2 votes) · EA(p) · GW(p)

For 3 - I think that shared Mendeley databases, are a good choice for synthesising and sharing research. New research is suggested, the web clipper makes it easy to add and the shared group mean that many people can contribute. It can also be combined with research updates from Scopus etc. I do one for Behaviourworks (my employer) for behaviour change research - see here for an update. I am happy to explain how to do it if 3+ people comment that they expect they would probably benefit it.

I think this is a good idea in the abstract, but I would be especially wary of starting "yet another aggregator" right now... I think we should use the LW links database [? · GW] which already has momentum.

For 13/18 - Thanks for the mention of our project - maybe we can collaborate on getting UK data?

Definitely. I think you have been in touch a bit with Derek and David on that. My understanding is it could definitely help but we are funding constrained on the data collection as are you. I shared your opportunity in my 29 March Brief [LW · GW].

comment by Peterslattery · 2020-03-31T04:21:16.739Z · score: 2 (2 votes) · EA(p) · GW(p)
I think this is a good idea in the abstract, but I would be especially wary of starting "yet another aggregator" right now... I think we should use the LW links database [? · GW] which already has momentum.

Yes, I agree. To be clear though, one issue now (I suspect) is that you can't trust the LW links database [? · GW] to be updated with all the key papers on a topic, nor can you search it to find these paper or use it to easily cite those papers. A Mendeley database could do that and in less than a days work. It would only be of use for 'serious' research though so not sure if needed. I also don't have the capacity to do anything to make such a database nor does anyone who could I suspect!

For now, perhaps the LW links database [? · GW] could benefit from someone with research knowledge setting up something to alert them to the latest research as it comes in? I could probably set something something that emails me the latest results from a search (e.g., (Covid-19 OR...) AND (vaccinat* OR ...)) and then forward that to someone involved. Of course this may already be happening - I haven't been able to pay much attention. If this is useful, then let me know and I will look into it.

Definitely. I think you have been in touch a bit with Derek and David on that. My understanding is it could definitely help but we are funding constrained on the data collection as are you. I shared your opportunity in my 29 March Brief [LW · GW].

Yes, they have been very helpful! Thanks for sharing. We will see if we can get more funding for this and if we do then we could potentially pay for that data collection.

comment by Tsunayoshi · 2020-03-30T15:50:18.162Z · score: 6 (2 votes) · EA(p) · GW(p)

[Epistemic status: Uncertain]

While working on covid-19 is by no means neglected, working on unconventional solutions is. I would be surprised, if there is a serious organization outside of the EA sphere evaluating just one of these proposals:

Robin Hanson has suggested both variolation, and controlled infection of essential workers.

Distribution of medical knowledge to amateurs. As most people in the third world will not have access to medically trained professionals, they will turn to local sources of treatment. Having some people armed with a bunch of youtube videos and DIY respirators as your sole treatment option is terrible, but might be less terrible than no treatment or herbal remedies.

Research into whether these ideas might be valuable or could cause long term harm to the EA community might be really important.

comment by IsabelHasse · 2020-03-31T21:46:04.484Z · score: 5 (4 votes) · EA(p) · GW(p)

Is it possible to contribute to research if I've never worked on a similar research project and have no relevant expertise?

comment by Khorton · 2020-04-01T12:40:46.434Z · score: 6 (3 votes) · EA(p) · GW(p)

I think you could contribute, but some projects will be easier to contribute to than other. Some suggestions:

-Helping to update Wikipedia pages about different countries' policy responses to COVID-19 would be helpful and a good introduction, because you'll have collaborators to help you.

-Many of these questions you could make a start on and ask for feedback from Aaron Gertler or Effective Altruism Editing and Review or Effective Altruism Coronavirus Discussion groups on Facebook. Then you can post your work, making the limitations clear.

-I would be careful about doing surveys or complicated modelling for the first time right now (you can give it a shot it's just really easy to get wrong). I'd also be thoughtful about interviewing people affected because they might be quite busy.

If there's a particular question you're interested in researching, feel free to message me to discuss how you'll go about researching it, or you could even write up your plan on your shortform.

Note: I am not an expert in research methods!

comment by AviNorowitz (AviN) · 2020-03-27T23:42:46.325Z · score: 5 (4 votes) · EA(p) · GW(p)

https://pandemic.metaculus.com/ works for me. The link you have is for https://www.pandemic.metaculus.com/ though, which does not work. Maybe that's the problem?

comment by Peter_Hurford · 2020-03-28T16:45:19.716Z · score: 2 (1 votes) · EA(p) · GW(p)

Fixed - thanks!

comment by Casey Riordan · 2020-04-22T22:13:28.365Z · score: 4 (3 votes) · EA(p) · GW(p)

This is Casey from Faunalytics - Thank you for your hard work putting this list together, it's incredible! In case it's helpful to add or update, we finished our COVID-19 poll and the results are available here: https://faunalytics.org/covid-19-poll/

The results may also provide some insight into question 36, as we found that the proportions of people who said they are more likely to try plant-based eating or reduce meat consumption because of the pandemic were similar to the proportion who said they were less likely to do so, suggesting that average consumption may remain similar. 

comment by Peter_Hurford · 2020-03-30T18:30:48.682Z · score: 4 (3 votes) · EA(p) · GW(p)

Some updates in my LW 29 March COVID Consolidated Brief [LW · GW]. When I can, I will also aim to update this post with what I find relevant.

comment by catherio · 2020-05-06T22:50:31.424Z · score: 1 (1 votes) · EA(p) · GW(p)

Thanks for this Peter. We've done some work on "6.) Are there places EAs should donate that focus on coronavirus response that are particularly promising to donate to, relative to existing charities EAs like?"

Here: https://forum.effectivealtruism.org/posts/opdMXibKjkoL69s96/prioritizing-covid-19-interventions-and-individual-donations [EA · GW]

comment by peter_janicki · 2020-03-30T19:53:07.350Z · score: 1 (1 votes) · EA(p) · GW(p)

as for question 21 (reduced donations due to recession):

i guess, we are all pretty sure: donations will go down. many people are loosing their jobs/ their income, some will get huge medical bills, … other people lost money in the stock market, others see their real estate or their pension savings loosing worth… i guess, that many people will “save” some money via donating less.

that might affect many valuable organizations/ employees of those organizations. their work might/will be disrupted.

i think it is a good idea to have a security net for cases like this, some kind of insurance. my best idea (beside having billionairs-pledges to increase donations in recession-times): put- options in the stock market. those gain in value as markets fall. and I don´t think of inidviduals buying a small amount of put-options (that would be pretty expensive, as I expect some costs per trade per person). i think of one or more ea-associated organizations which bundle some money to invest in put-options. this money would be lost in booming times, but it would multiply in bad ones. to be clear: all in all: it would mean loosing some money (the put-options sellers want to make some money too). let´s say “we” loose 2% of the money. but what could we gain? i guess, that having a smooth(er) running of ea- organizations might be of higher worth, than those organizations running constantly on 98%-funding. i guess, that having all those costs which usually come along with shrinking donations (f. e. loosing valuable employees -maybe forever… in better times: costs of time needed to find new fitting employees, to incorporate them… loosing connections, closing organizations, and so on) is higher.

i am not saying, that individual ea-organizations should spent money on a security net (in the stock market). it might seem weird to the broader public, if altruistic organizations would invest money in the stock market. but It would be nice, if those organization would know: there is a back-up, some kind of security for bad times. if one or two organizations would do so, that might be great. they could call themselves: “anticyclical donating” or something like this.

imagine you (=ea) want to drive as far as you can get with a limited amount of energy (=doing the most good with limited ressources). one option might be simply going full throttle (=like us now?). another one might be, to invest some energy in levelling your track (=like investing in a security net for the “bumbs” of a recession?).

actually, I hope that his post is not of worth, because there are no such described downfalls, or because other people already had the same idea. but as i´ve never heard of it. here it is (see above) ;-)

comment by Liam_Donovan · 2020-03-28T06:46:02.078Z · score: 0 (3 votes) · EA(p) · GW(p)

I'd be interested in joining the Slack group

comment by Peter_Hurford · 2020-03-28T16:45:38.486Z · score: 2 (1 votes) · EA(p) · GW(p)

Email me at peter@rethinkpriorities.org and I'll send an invite back to your email.

comment by Liam_Donovan · 2020-03-29T05:52:53.262Z · score: 1 (1 votes) · EA(p) · GW(p)

sent, thank you