Hide table of contents

This is a linkpost to the article Counterproductive Altruism: The Other Heavy Tail from Daniel Kokotajlo and Alexandra Oprea. Some excerpts are below. I also include a section at the end with some hot takes regarding possibly counterproductive altruism.


First, we argue that the appeal of effective altruism (henceforth, EA) depends significantly on a certain empirical premise we call the Heavy Tail Hypothesis (HTH), which characterizes the probability distribution of opportunities for doing good. Roughly, the HTH implies that the best causes, interventions, or charities produce orders of magnitude greater good than the average ones, constituting a substantial portion of the total amount of good caused by altruistic interventions. Next, we canvass arguments EAs have given for the existence of a positive (or “right”) heavy tail and argue that they can also apply in support of a negative (or “left”) heavy tail where counterproductive interventions do orders of magnitude more harm than ineffective or moderately harmful ones. Incorporating the other heavy tail of the distribution has important implications for the core activities of EA: effectiveness research, cause prioritization, and the assessment of altruistic interventions. It also informs the debate surrounding the institutional critique of EA.

IV Implications of the Heavy Right Tail for Altruism

Assume that the probability distribution of charitable interventions has a heavy-right tail (for example, like the power law described in the previous section). This means that your expectation about a possible new or unassessed charitable intervention should include the large values described above with a relatively high probability. It also means that existing charitable interventions whose effectiveness is known (or estimated with a high degree of certainty) will include interventions differing in effectiveness by orders of magnitude. We contend that this assumption justifies well-known aspects of EA practice such as (1) effectiveness research and cause prioritization, (2) “hits-based-giving,” and (3) skepticism about historical averages.

V Implications of the Heavy Left Tail for Altruism

What if the probability distribution of altruistic interventions includes both a left and a right heavy tail? In this case, we cannot assume either that (1) one's altruistic interventions are expected to have at worst a value of zero (i.e. to be bounded on the left side) or (2) that the probability that a charitable intervention is counterproductive or harmful approaches zero very rapidly.

Downside Risk Research

Many catastrophic interventions — whether altruistic or not — generate large amounts of (intentional or unintentional) harm. When someone in the world is engaging in an intervention that is likely to end up in the heavy left tail, there is a corresponding opportunity for us to do good by preventing them. This would itself represent an altruistic intervention in the heavy right tail (i.e. one responsible for enormous benefits). The existence of the heavy-left tail therefore provides even stronger justification for the prioritization research preferred by EAs.

Assessing Types of Interventions Requires Both Tails

Another conclusion we draw from the revised HTH is that the value of a class of interventions should be estimated by considering the worst as well as the best. Following such analysis, a class of interventions could turn out to be net-negative even if there are some very prominent positive examples and indeed even if almost all examples are positive. This sharply contradicts MacAskill's earlier claim that the value of a class of interventions can be approximated by the value of its best member.

The Institutional Critique Reassessed

If we are right about the existence of the left tail, certain interventions (even well-intentioned ones) are or can be expected to be extremely net-negative. Furthermore, even certain classes or subclasses of charitable interventions (e.g. foreign aid, food aid, or billionaire philanthropy) can be net-negative as a whole. In these cases, the most good an effective altruist can do may not be to launch new charitable ventures of her own or even to donate to the most effective charities. As noted above, the most efficient intervention might be to stop oneself or other people from launching massively negative interventions.

VI The Evidence for the Heavy Tail(s) Hypothesis: Existing Arguments

In this section, we begin by reconstructing three arguments in favor of a single right heavy tail that EAs have sketched: (i) the argument from examples of extreme values, (ii) the argument from nore systematic observational studies; and (iii) the argument from inefficient markets. For each of the arguments presented, we note that they should be extended to the existence of a heavy left tail.

VII The Evidence for the Heavy Tail(s) Hypothesis: New Arguments

The Crowding Out Argument

Consider any big goal you wish to achieve—the sort of goal that would put your intervention far out in the right tail if you were to achieve it. There is some chance that the goal will be reached anyway without your effort, due to the effort of someone else. There is also a chance—perhaps a smaller chance, but a chance nonetheless—that your effort will cause an effective intervention not to happen or to be less effective than would have been the case without your action. For example, perhaps if you had not chosen to work towards this goal, someone more competent would have noticed the need and taken up the project in your absence. Thus, your choice to work on the project has a chance of backfiring, and if it does, it is a failure of the same magnitude as your success would have been.

The Data Generating Process Argument

Given that calculating the effectiveness of even a narrow range of philanthropic interventions we are considering [e.g. distributing anti-malaria bednets] typically involves multiplying together a large number of independent variables, we should expect the distribution of philanthropic interventions by effectiveness to be at least log-normal [which "are typical when data points are the product of many independent inputs"].

The Burden of Proof Argument

Without specific research into effectiveness, your uncertainty about how effective they are will range over many orders of magnitude. Lining them up side-by-side in your position of ignorance, they might look something like Figure 4 below.

Details are in the caption following the image
Uncertainty about effectiveness of interventions

Hot takes

Minimising downside is a common theme in effective altruism. However, I still found the article interesting as a reminder that heavy left/harmful tails are often neglected, and hidden behind the status quo. What looks robustly beneficial ignoring left tails might not be so once one accounts for them. In other words, left tails may conceal crucial considerations. Some hot takes (from me, not the article):

  • Decreasing the consumption of factory-farmed animals is pretty good for some of them (birds, fish and arthropods), but harmful to humans in the event of an abrupt sunlight reduction scenario (ASRS, such as a nuclear winter).
    • The smaller the population of animals, the less animal feed could be directed to humans to mitigate the food shocks caused by the lower temperature, light and humidity during an ASRS.
    • Because producing calories from animals is much less efficient than from plants, decreasing the number of animals would tend to result in a smaller area of crops.
    • So the agricultural system would be less oversized (i.e. it would have a smaller safety margin), and scaling up food production to counter the lower yields during an ASRS would be harder.
  • Mitigating global warming decreases the chances of crossing a tipping point which leads to a moist or runaway greenhouse effect, but increases the severity of ASRSs.
    • The major driver for the decrease in yields during an ASRS is the lower temperature, so starting from a higher baseline temperature would be helpful.
    • One might argue the severity of ASRSs is only a function of the temperature reduction, not of the final temperature, on the basis that yields are roughly directly proportional to temperature in ºC. However, this is not the case.
    • The typical base temperature of cool-season plants is 5 ºC. So, based on the heuristic of growing degree-days, a reduction from 10 ºC to 5 ºC leads to a 100 % reduction in yields, not 50 % as suggested by a direct proportionality between temperature in ºC and yields.
  • Saving and extending lives is nice for the people who get to live them, but not for the factory-farmed animals who get to be produced (see meat-eater problem).
    • To illustrate, for the 35.07 G poultry birds and 7.84 G humans in 2020, and Rethink Priorities' (RP's) median moral weight of 0.332 for chickens, one can conclude the total moral weight of chickens is 1.49 (= 35.07*0.332/7.84) times that of humans.
    • Based on data from the Welfare Footprint Project, I also guess the intensity of the experiences of chickens relative to their moral weight is higher than for humans, so the factor of 1.49 may well be an underestimate.
    • In addition, I suppose the mean moral weight, arguably what matters, may be higher than the median moral weight.
  • Forestation (or less deforestation) can be good to decrease greenhouse gas (GHG) emissions, but may actually lead to global warming if the albedo of the forested area is significantly darker (lower albedo).
    • Global warming depends not only on the concentration of GHGs, but also on the amount of light reflected by Earth. If this increases, the Earth gets cooler.
    • Forested areas are generally darker because they need light to grow (producing energy via photosynthesis).
    • So, if the forested area is significantly darker (e.g. foresting an area which would otherwise be covered in snow during winter), forestation can lead to global warming (like in the Rocky Mountains; see Williams 2021).
    • However, forestation ususally contributes to mitigating global warming.
  • Ukraine (and its allies) resisting the invasion of Russia may decrease the chance of future invasions (e.g. China invading Taiwan), and accelerate the democratization of Russia (if opposition to its leaders increases), but is horrible for the people involved (from both Ukraine and Russia), and can increase the risk of nuclear war.
  • Remittances lead to benefits via increased consumption in the nearterm, but may (or not) hinder economic growth. From Cazachevici 2020:
    • "We conduct the first meta-analysis of the effect of remittances on economic growth. Although the macroeconomic importance of remittances has been rising over time, the literature has not reached a consensus and continues to produce estimates that differ widely. We collect a dataset of 95 articles displaying 538 regression equations and observe that around 40% of them report a positive and statistically significant effect of remittances, around 20% report a negative and statistically significant effect, and around 40% do not find any statistically significant impact of remittances on economic growth".
  • Foreign aid has obvious advantages in the nearterm, but may (or not) decelerate democratization. From Kono 2009:
    • "Although many people have argued that foreign aid props up dictators, few have claimed that it props up democrats, and no one has systematically examined whether either assertion is empirically true. We argue, and find, that aid has both effects. Over the long run, sustained aid flows promote autocratic survival because autocrats can stockpile this aid for use in times of crisis. Each disbursement of aid, however, has a larger impact on democratic survival because democrats have fewer alternative resources to fall back on".
  • The effects on wild animals, namely arthropods, might be a major driver for the nearterm effects of global health and development interventions (see here), but we do not know whether they are good or bad:
    • It is hard to estimate the change in the number of arthropods caused by changes in the human population, and trends influenced by human activities, such as deforestation, and global warming.
    • It is unclear whether arthropods have good or bad lives.

To be clear, I am not arguing for factory-farming, global warming, shorter lives, deforestation, and resistance to invasions, nor against remittances, and foreign aid. I am just trying to illustrate what is considered robustly beneficial may have a real chance of being harmful. Relatedly, there is the concept of complex cluelessness.

Sorted by Click to highlight new comments since:

Less theoretical example: FWIW im not sold on 'more than anyone' but the top 2-3 current AI labs are all downstream of AI safety!

Great point! It does look like left tails are everywhere in the AI safety space.

Though you need to consider the counterfactual where the talent currently at OAI, DM, and Anthropic all work at Google or Meta and have way less of a safety culture.

Really enjoyed this piece. It is somewhat painful to read, given that I believe most of my professional life did more harm than good.

I do think that partially rationalizing torturing billions of sentient beings every year for more corn in silos in case of a nuclear winter - that's really a stretch.

Thanks for sharing, Matt!

I started following a plant-based diet roughly 4 years ago mostly due to finding out about the badness of factory-farming (and also because I think it mitigates global warming, and is healthier). Meanwhile, I have gotten confused about the overall impact of a plant-based diet, given the uncertain effects on wild animals and in the longterm. I think I continue plant-based because (descending order of importance):

  • It feels intuitively wrong to be responsible for some visible torture based on unclear overall effects which are quite uncertain.
    • I think I should feel fine about doing something with overall unclear effects even if the most visible effects are bad. 
    • However, I do not, and I suppose it makes sense to avoid conflicts with intuitions to some extent. For example, if in theory eating animals was super good overall, and I could not internalise that, still feeling bad about contributing to factory-farming, it is possible the overall best option for me would be continuing not to eat animals, such that I could remain productive working on other matters.
  • I believe a plant-based diet is healthier, and can extend my life for a few years. Since I think my work is positive, having the chance to do more of it is good!
  • I no longer like the taste of animals. Switching at this point would be hard, especially given the above.
  • I think a plant-based diet is more practical (e.g. generally involves less cooking time, and less cleaning due to less fat). There would typically be factors contributing to it being less practical, but I do not think those affect me much. For example, for (rare) family meals in restaurants, I am fine with just eating soup, rice and lettuce, or whatever is available.

Thanks, Vasco. I find it very difficult to imagine a scenario where I would support the active torture of factory farming chickens for any unknown / theoretical counterpoint. I'd certainly rather be a wild animal than a factory-farmed chicken. 

Take care.

Is there a version of the paper that is freely available?

Yes: the version available from Sci-Hub[1]

  1. ^

    On the topic of Sci-Hub generally, this may be of interest.

Curated and popular this week
Relevant opportunities