Is helping non-humans the best way to help humans?

Abstract

To my knowledge, the benefit to humans of having less animal product consumption globally has never been quantified, despite there being a number of proposed mechanisms for how this could be the case (e.g. public health, climate change, antibiotic resistance, etc.). This scoping study, the first of its kind, estimates that a donation of $1,999 to The Humane League (the animal charity suggested by Animal Charity Evaluators to be the most effective at reducing animal suffering/animals farmed) today is estimated to result in 0.23 human lives saved in the short term, and 0.0039 human lives saved from 2030 to 2050. By way of comparison, $1,999 is the amount estimated by GiveWell to save a human life if donated to the top rated human charity, the Malaria Consortium’s seasonal malaria chemoprevention.

The level of uncertainty in this analysis is large, and many assumptions were made. I expect and hope for a lot of criticism, and intend to amend the analysis in an ongoing manner. This page represents the latest version (02/11/2018), while a history of amendments can be seen here.

Introduction

Many forms of animal advocacy that focus on reducing the consumption of animal products have the obvious primary benefit of reducing farmed animal suffering. They also have numerous possible secondary benefits that may help reduce human suffering, including reduced greenhouse gas emissions, slower global antibiotic resistance, potentially reduced public health burden, and possibly improved food security and equality.

Given these secondary benefits, I’ve often therefore wondered how much supporting the most effective non-human focused charities (e.g. those that aim to reduce non-human animal product consumption) is at reducing human suffering compared to the most effective human focused charities. If it turned out to be the case that they were more effective at reducing human suffering, it could be a game changer. If this were the case, rationally motivated people should start supporting the best non-human charities rather than the best human charities, even if they didn’t care at all about non-humans. For those of us who care about all sentient minds, this would be a win-win.

Despite the potential for this, I haven’t really seen any attempts to quantify the flow-on benefits. The closest I’ve seen was this ACE post by Ben West in 2012, though it only covers climate change.

One must of course be cautious of motivated reasoning while making this analysis – I might be biased during my assessment towards making the human benefits of supporting non-human charities stronger, or the benefits of supporting human charities weaker. I will do my best to show all of my assumptions, estimates, calculations and rationale in full. I encourage people to recreate my conclusion, either from scratch or using my assumptions, and to let me know what they came up with.

This analysis is intended as a scoping study, and should be read as such. It is not intended to be a final, conclusive report. There are of course many improvements that could be made, and I may address them at a later date. At this time, I aim only to create interest in a comparison which I think is important but seriously neglected.

Base impact of top human and non-human charities

First, we need to look at what the best non-human and human charities are, and what they do. For this, I will use the top recommended (as of 17/09/2018) non-human and human charities from Animal Charity Evaluators and GiveWell respectively.

One may reasonably wonder why I picked these charities rather than those focused on reducing suffering or saving lives in the future or far future (i.e. hundreds to billions of years from now), given that they may be orders of magnitude more effective at achieving our goals if we care equally about future lives. I accept this argument, but stick to short term charities mainly because comparing the best far future charity to help non-humans and the best far future charity to help humans would be extremely difficult and highly (even more so) uncertain.

Non-human charities

ACE’s top rated charities are currently Animal Equality (AE), The Humane League (THL) and the Good Food Institute (GFI). While I am a strong supporter of what GFI does, a full cost-effectiveness analysis has not been performed (e.g. an estimate of how many less animals eaten per dollar donated). The reviews and models of THL and AE suggest that $1* donated to them would spare 15.833 and 0.095 animals from life in industrial agriculture respectively. As THL appears to have a significantly higher average impact for reducing the number of animals farmed, we will use them for the comparison.

The models were made on Guesstimate, which uses Monte Carlo simulations to arrive at the average impact values. I ran the simulations for each of THL and AE 30 times and took the mean. ACE has written up an explanation of their cost-effectiveness estimates here.

The methodology for determining the impact of reducing animal product consumption on each of the cause areas (e.g. public health, climate change) is non-obvious and will be done on a case by case basis. Even allowing for assumptions, there may be room for substantial improvement, and so encourage critiques of these areas specifically.

For simplicity, I will generally be focusing on the consumption and reduction in consumption of land animals. Multiple conflicting estimates of the number of land animals farmed per year exist, and I have chosen the somewhat mid-range estimate of 80 billion land animals farmed per year (70 billion in 2013 and rose from 60 billion in 2008, so I assumed the same trend from 2013 to 2018).

It should be noted here that these reductions in animals farmed are not instantaneous. The reductions are spread across the average time that someone remains vegan (for example). Throughout the analysis, I assume that the reductions are instant and apply to the current year. This is not accurate, but I don’t foresee it significantly affecting the estimates.

Human charities

GiveWell present a list of 9 top rated charities. From these, three are recommended primarily due to their effectiveness at reducing mortality (the others are recommended for their potential to increase income and consumption). These three charities are the Against Malaria Foundation (AMF – $3,753 per life saved, using 2018 figures), Malaria Consortium’s seasonal malaria chemoprevention (SMC – $1,999 per life saved), and Helen Keller International’s vitamin A supplementation program (VAS – $2,692 per life saved). Malaria Consortium’s SMC programs (MC from here for short), appears to be the most effective, and so this value will form the baseline that THL must beat in order to be more effective at reducing human suffering than the best human charity.

In other words, we investigate the effect of donating $1,999 to THL on reducing human deaths ($1,999 to THL = 31,650 less land animals farmed – 0.0000395625% of total land animals farmed).

Public health

Evidence suggests that populations with a high proportion of plant-based food in their diet have lower blood pressure, a lower risk of type 2 diabetes, a lower risk of death resulting from cardiovascular disease, and a lower risk of cancer. Diets containing a high amount of red and processed meats are estimated to have contributed to around 584,000 human deaths globally in 2015. This refers to deaths at all ages, while GiveWell’s analysis refers to the ‘cost per outcome as good as averting the death of an individual under 5’.

It could be argued that averting the death of an individual under 5 would be better than averting the death of someone who is, say, 70, who would be more likely to die sooner anyway. To rectify this in a simple manner, I propose the following. If what we value is the life years increased rather than reducing deaths per se, a person under 5 would have about twice the number of remaining life years as the average person dying due to the causes outlined in the above report (assuming an equal distribution of deaths at given ages, which of course is unlikely). In this case, we would halve the 584,000 to 292,000 for a roughly transferrable statistic.

I take the above value as a lower bound, as it does not include deaths due to diets low in vegetables, whole grains, nuts and seeds, and fruit. I would expect to see these deaths reduced somewhat with less widespread animal product consumption. I didn’t include them as the link was much less clear. However, determining how many fewer deaths might result from less animal product consumption due to these factors should be a high priority, since the total number of deaths caused by them globally is 10.191 million (5.096 million after age of death normalisation).

I didn’t include deaths caused by diets high in sodium, which may or may not be reduced with less animal product consumption. I didn’t include deaths caused by diets low in seafood, since the nutrients, omega-3 in this case, can be sourced elsewhere. I also didn’t include deaths caused by diets low in calcium, as this can be sourced from non-animal products as well as animal products. It is plausible that there will be an increase in deaths due to diets low in omega-3 or calcium as animal product consumption is reduced, but there may also be a decrease as information about the existence of plant-based alternatives is disseminated along with the work of animal charities like THL. Having said that, it is possible that in developing countries, fish and dairy products are just more accessible and cheap than plant-based alternatives such as brussel sprouts or bok choy.

We can now use this to estimate how many fewer human deaths will result per $1,999 donated to THL, or per 31,650 fewer animals farmed (0.0000395625% of total land animals farmed). For public health, we will assume that if we went from 80 billion to zero animals farmed for a given year, it would result in deaths due to diets high in red and processed meat consumption going from 584,000 to zero for that year. 0.0000395625% of 584,000 is 0.12, thus, a donation of $1,999 donated to THL would result in 0.12 human lives saved in a once off. To think of this in another way, donating $1,999 each year would result in another 0.12 human lives saved each year. Note that this could see a up to a potential 20-fold increase if we were to include deaths due to diets low in vegetables etc. that would be mitigated if less animal products were consumed.

Climate change

Climate change is projected to cause 250,000 deaths per year from 2030 to 2050. From the source, I can’t tell whether this refers simply to deaths at all ages, but I assume this is the case. As above, we should alter this value to be in line with the GiveWell metric of ‘cost per outcome as good as averting the death of an individual under 5’. Using the rationale and methodology discussed above, this gives us 125,000 deaths per year from 2030 to 2050.

Animal agriculture is one of the leading causes of anthropogenic climate change, contributing 18% of greenhouse gas emissions (determined in 2006), while a more recent study suggests animal agriculture has contributed 23% of the total warming as of 2010. To determine the number of human deaths averted as a result of reducing the number of animals farmed, we need to determine how a reduction of emissions this year by 23% would affect the number of human deaths from 2030-2050. Ideally, accounting for this would require the following.

  • The proportion of each greenhouse emitted by the livestock industry (e.g. carbon dioxide, methane)
  • The climate forcing effect of each of these greenhouse gases
  • The lifetime of each of these greenhouse gases
  • Using climate models to determine how the climate would differ in 2030-2050 given this once off reduction in emissions
  • Revisiting existing studies which determine the number of human deaths from 2030-2050 given status quo climate change, and revising them using the new climate models

This is significantly beyond the scope of this study. A second best option would be to find an existing study which says something like “A reduction of X emissions this year would result in X fewer deaths from 2030-2050”, and then use it to see what a reduction of 23% of emissions this year would mean for deaths from 2030-2050. I don’t believe such a study exists, but if it does this will make this part of the analysis somewhat more robust, and I would ask that you point me to it. In lieu of any such study, I will make some assumptions to come up with a number.

This is harder to quantify than public health, as some of the emissions which will impact future effects of climate change (from 2030 to 2050) have already been emitted, while some are yet to be emitted. Reducing animal product consumption from 80 billion animals to zero for a given year will result in a reduction of emissions of 23% for that year, but the effect of this on the deaths from 2030-2050 are unclear. If we can assume that the emissions over the 30 years will roughly plateau, perhaps we can assume that half the emissions that will affect deaths from 2030-2050 have already been emitted. Thus reducing animal product consumption to zero for this time period might be expected to reduce the relevant emissions by half of 23% (11.5%).

We might then be able to say that reducing animal product consumption to zero for a given year would reduce the relevant emissions by 1/30th of 11.5% (0.38%). We could then suggest that this would reduce the number of deaths over the period of 2030-2050 by 0.38%. I don’t expect this to be a simple relationship, let alone a linear one, but in lieu of any idea what a better model would look like, I will take the simplest one.

Therefore, I simplistically assume for this calculation that reducing animal product consumption to zero for a given year would reduce the number of deaths from 2030-2050 due to climate change per year by 0.38% of 125,000 (475). We can now use this to predict how many fewer human deaths will result per $1,999 donated to THL, or per 31,650 less animals farmed (0.0000395625% of total land animals farmed). 0.0000395625% of 475 gives 0.00019 lives saved per year, or a total of 0.0039 from 2030 to 2050. This is significantly lower than I expected, but it’s a benefit we might expect to become stronger as we allow for more future time frames (e.g. from 2030 to 2060).

This assessment does not include other environmental impacts of animal agriculture, such as desertification and local impacts, e.g. the spraying of animal waste matter in to the air which is breathed in by local residents. It also doesn’t include deaths due to particulate matter from coal power plants and fossil fuel powered vehicles, which is breathed in by local residents. One study suggests that reducing emissions by 50% by 2050 could prevent 100 million early deaths via this mechanism alone. Given that animal agriculture has a non-trivial use of energy (much of which is coal) and transport, we could expect reducing animal agriculture to play a role in this.

Antibiotic resistance

Currently, 700,000 people die per year due to antimicrobial-resistant diseases. If trends continue without any major difference, this figure could reach 9.5 million dead per year by 2050. As above with public health and climate change, we should alter this value to be in line with the GiveWell metric of ‘cost per outcome as good as averting the death of an individual under 5’. Using the rationale and methodology discussed above, this adjusts the figures to 350,000 per year and 4.75 million per year by 2050.

It is currently hard to say how much of these figures are due to the use of antibiotics in the global livestock industry, however growing evidence does point to factory farming being a major contributor.

In a population of bacteria, there will be some proportion (~< 1 in a million) with a natural resistance to a given antibiotic. Antibiotic use imposes a selection pressure on bacteria. The antibiotic will kill the bacteria, leaving the naturally resistant bacteria, allowing it to selectively reproduce. Increased use increases likelihood of the bacteria exposed becoming resistant to antibiotics, however the relationship between antibiotic use and resistance is not clear and likely not linear. Other studies go further to say that, despite the possibility, it is unclear whether antibiotic use in farming results in any increase in antibiotic resistance, due simply to a lack of data and analysis.

Given this, it seems fair to conclude that it is not yet known whether a reduction in animal farming (and therefore antibiotics use in farming) would result in any meaningful reduction in human deaths due to antibiotic resistance. This is not to say that the evidence points to this, but rather that there is not yet much evidence either way. Despite this limitation, I would like to estimate an upper bound for the benefit, if it were discovered that there was indeed a link. Thus all following calculations in this section should be interpreted as a possible upper bound if a link were confirmed.

In addition, the following is an argument that favours the possibility of a link between the intensification of animal agriculture generally and human deaths (C. Nelson, pers. comms.). Zoonoses (emerging human pathogens with non-human animal origins) comprise 60% of currently known infectious agents. The intensification of animal agriculture has increased the risk of zoonotic diseases, examples of which include SARS, avian influenza and hepatitis E. By providing an ideal environment for pathogens (including a large number of hosts and animal-human contact), they can jump more easily to the human population.

To estimate the impact of fewer animals farmed on preventing these deaths, we first need to determine what proportion of antibiotic use is in animal farming. Around 80% of antibiotic use in the US is on farm animals, and 75% for the EU and US combined. THL appears to operate primarily in the US, and so I will use the 80% figure. In other words, we stand to reduce antibiotics usage and (I am assuming) antibiotic resistance-related deaths by up to 80% (therefore 280,000 per year and up to 3.8 million per year by 2050).

There are some questions I don’t know the answers to. Would a reduction of antibiotic use of 50% today result in half the people currently dying per year due to antimicrobial-resistant diseases, or does the impact of past antibiotic use carry over to this year? The answer is almost certainly somewhere in the middle, and most likely leaning towards the past antibiotic use being the leading contributor. Would a temporary drop of 50% this year alone result in any meaningful difference in the number of deaths in 2050? Again, it would probably make some difference, but not a linearly proportional difference.

For simplicity, I will propose that a reduction of antibiotic use of 50% today will result in 50% less deaths the following year due to antimicrobial-resistant diseases. This doesn’t quite seem right, since one can suppose that if in 2049 antibiotic use is dropped by 50%, it probably wouldn’t reduce the number of deaths the following year by 50%, since many of the diseases will have already become resistant. Unless given strong rationale to propose otherwise, I will use this direct, short-term only relationship as a proxy.

We can now use this to predict how many fewer human deaths will result per $1,999 donated to THL, or per 31,650 less animals farmed (0.0000395625% of total land animals farmed). As with climate change, the effect here is harder to quantify than public health.

Therefore, 0.0000395625% of 280,000 gives 0.11 deaths averted per year for $1,999 donated to THL.

This pathway should be tempered with the knowledge that, in some regions at least, agricultural antibiotic use has fallen (12% from 2011 to 2014 in 24 EU countries) due to increased awareness of the negative implications of antimicrobial resistance. Alternatives to antibiotic use have been successful so far, for example vaccine use and better farming practices. However, total antibiotic use in animal farming is still projected to grow by just under 70% from 2010 to 2030. In particular, the use of antibiotics in BRICS countries is expected to grow by 99% by 2030. Relevant to this, it is not clear whether animal advocacy would be more or less effective in BRICS or developing nations.

Injuries and effects on employees

Using the US as an example, there are over 500,000 slaughterhouse and meat-processing workers. From 2004 to 2013, an average of 15 ‘meat and poultry workers’ died from injuries at work per year. As of 2016, the proportion of workers sustaining a notable injury is “in the 10 percent and below [range]” (higher than that of manufacturing), though there are reasons why the rate of injuries is likely to be greatly under reported. Another study shows that in 2001 14.7 percent of full-time workers sustained an injury, which was the highest rate of any industry in the US. More information about the treatment of slaughterhouse workers in the US can be found here.

Given that slaughterhouse workers would transition from this industry to another, which would have its own injury and death rate (almost certainly lower than slaughterhouse workers), it seems difficult to measure the impact of donations to non-human charities on human deaths via this pathway. It is almost certainly positive, but small (relative to the above pathways).

We can simplistically visualise the impact by assuming that all of the workers would transition to a similar industry, e.g. manufacturing, and examine the difference in injury rate. The US Department of Labor estimates that, in 2016, the nonfatal occupational injury rate of private industry was 2.9 total recordable cases per 100 full-time equivalent workers. Total recordable cases includes minor injuries, and is therefore an upper bound for the above measure of ‘workers sustaining a notable injury’. This would be a substantial reduction in human injuries, and presumably deaths.

There is also a compelling argument that the environment of working in a slaughterhouse or factory farm has ongoing adverse mental health effects on employees. These may include violent tendencies, which would affect their own lives and that of others, possibly resulting in death. Again, the link here is sufficiently unexplored that I will not include it in the numerical assessment.

Global food distribution

A large amount of plant food globally is fed to farmed animals. Producing one kg of beef protein requires seven kg of plant protein as feed stock, not to mention the vastly increased water and energy costs (Australian beef is one notable exception to this, given they are mostly pasture raised, but chickens and pigs in Australia are not exceptions). Given this, there is a compelling argument that if less of this plant food was fed to animals, there would be more to feed to humans, potentially alleviating some of the issues associated with food scarcity, particularly in developing nations.

This pathway to fewer human deaths is significantly more complicated than the others, and so I intend to put it aside. If at a future time I can better analyse this (please send me evidence for or against this pathway being a positive factor), I will revisit the analysis and include it. The below quote is interesting, but seems like exaggeration at best, even for a utilitarian like myself who is happy to admit that my spending money on leisure activities is basically killing others due to opportunity cost.

Every day 40,000 children die in the world for lack of food. We who overeat in the West, who are feeding grains to animals to make meat, are eating the flesh of these children.” Thich Nhat Hanh (2003)

A concern about land use change can be addressed here. Some worry that there are arid regions where livestock are raised that would not be suitable for growing plant food. It should be pointed out here that 90% of animals globally are factory farmed and fed plants, resulting in inefficient resource use. One should expect there to be less farmland required, not more, if there were a global shift to veganism, and there are some plant foods that can be grown in arid regions, such as hemp and almonds. The land could also be used for other purposes, such as cotton farming or carbon sequestration.  A report on further analysis on land use change (focused on Australia) can be found here.

Economic impacts

Animal agriculture is a major economy in most parts of the world. Transitioning away, even slowly, might be expected to have some economic impacts. In particular, some animal farmers and slaughterhouse workers may lose their job as a result of reduced demand for animal products. Given that industries have come in and out of existence all the time when replacements arrive (e.g. horse-and-cart replaced with cars and whale oil replaced with kerosene), I don’t foresee this having any long term negative impacts. I also don’t foresee the short term impacts of lost jobs outweighing the human lives saved (let alone non-humans), but it is important to consider this.

At the very least, work should be done by governments, non-profits and industry bodies to assist farmers in transitioning to other jobs, either plant farming or something else. Also, as stated above, other work can be done with the land used to farm animals. A report on further analysis on economic factors as the result of a shift to veganism (focused on Australia) can be found here.

Total human benefit of top non-human charity

It costs $1,999 to save a human life by donating to Malaria Consortium’s seasonal malaria chemoprevention. Summarising the effects on human lives saved from the public health, climate change and antibiotic resistance pathways, we get the following benefit for $1,999 donated to The Humane League:

  • 0.12 human lives saved in the short term due to improved public health
  • 0.0039 human lives saved from 2030 to 2050 due to reduced climate change effects
  • 0.11 human lives saved in the short term due to reduced antibiotic resistance (an upper bound if a link between antibiotic use in farming and human deaths is indeed discovered)

Or in other words:

  • 0.23 human lives saved in the short term
  • 0.0039 human lives saved from 2030 to 2050

This means that even if you only care about humans, a donation to The Humane League is about one quarter as effective at preventing human deaths than the best GiveWell rated charity, Malaria Consortium’s seasonal malaria chemoprevention. However, the evidence for the former is of course significantly less robust. If you also consider the 31,650 non-humans spared a life of suffering in a factory farm, I believe it becomes increasingly harder to justify supporting the best human-centric charities over the best non-human-centric charities.

Future work

I’ve assumed a lot of things throughout this analysis, some due to necessity (relevant numbers are simply not available) and some due to a lack of time and expertise (more accurate assumptions could potentially be made). Given the startling outcome of this research, I believe improving the accuracy and efficacy of this research to be of utmost priority. In particular, there is the potential for the actual number of lives saved via the public health pathway to be significantly higher, and the number of lives saved via the antibiotic resistance pathway to be lower.

I would also like to see other mechanisms examined in this analysis, such as the reduction of deaths due to more fruit and vegetable consumption and less particulate emissions as the result of fewer animals farmed.

This analysis only examined lives saved, not quality adjusted life years (QALY’s) to also account for improvements in the quality of life of individuals. It is hard to say whether this would make supporting non-human charities more or less attractive from a purely human-centric view. Future research should seek to examine the effect on QALY’s for a more complete and ethically relevant (from a utilitarian perspective) analysis.

I am aware of concepts such as the optimisers curse and the risk of sequence thinking vs cluster thinking, and am further aware that they are likely relevant here, but I have not taken them in to account. Future work should critically examine all possible sources of biases and statistical failures.

I have made little to no attempt to estimate the uncertainty, which would be large, instead only using mean values and carrying them through the analysis. This could be done in Guesstimate, and is something I intend to do in a future version of this work. It would also be valuable to generate upper and lower bounds (say at the commonly used but completely arbitrary 95% confidence intervals) for the final estimates. Using the mean effectiveness of ACE recommended charities and the mean effectiveness of GiveWell recommended charities may be another way to make the analysis more robust.

Despite the potentially high number of human deaths resulting from antibiotic use in animal farming, it seems surprising that there has been so little research in to the existence of a relationship. I am not a pathogen researcher and consider biology to be my weakest science, so I don’t intend on ever researching this further myself, however I would consider this to be of utmost importance. If you are a researcher in this field or are considering entering it, this would be a highly important research question.

Given that governments claim to at least care about their own human population, it is surprising to me that no governments to my knowledge have bothered to examine these pathways, let alone quantify them. I have written previously about the possibility for a government public health campaign around reducing animal product consumption to save money via the public health burden alone. Governments should seriously consider the possibility that reducing the animal product consumption of their citizens is a neglected and impactful way of reducing human deaths and suffering, as well as saving money.

Conclusion

A donation of $1,999 to The Humane League today is estimated to result in 0.23 human lives saved in the short term, and 0.0039 human lives saved from 2030 to 2050. By way of comparison, $1,999 is the amount estimated by GiveWell to save a human life if donated to the top rated human charity, the Malaria Consortium’s seasonal malaria chemoprevention.

This outcome surprised me, and I expect it will surprise you too. Unless they can find fault with my argument or calculations, I strongly encourage the reader to, if they currently support human-centric charities, to consider switching some or all of their charitable giving to non-human-centric charities.


I expect to edit this post significantly in the coming days. I will do this as transparently as possible. It is a certainty that I have missed critical information and/or arguments, some of which may substantially alter the numbers and conclusions. I believe it would be far better to edit the post to reflect these updates rather than leave it an potentially redo the analysis at a later date.

Again, I urge you to engage with this in every way. If you support human charities and cannot find fault with my analysis, you should start supporting the most effective non-human charities.

I believe this shouldn’t just be an area of interest to individual donors, but also to funds, advocacy organisations, and governments.

For the purposes of this argument, I have ignored the ‘rich meat eater problem’ (people tend to eat more animal products as they get wealthier), although it is quite important to be aware of for anyone deciding whether to support a human or non-human centric charity via donations or otherwise. I will explain what this is and why it is important in Appendix A.

Also, in Appendix B I examine the impact of reducing human deaths by being vegan personally.

Appendix A – The rich meat eater problem

The rich meat eater problem is the idea that as populations or individuals move out of poverty and become more affluent, they will on average consume more animal products. This is because animal products are often relatively expensive and resource intensive. There are two key factors operating here in seemingly opposite directions. As a population becomes more affluent and their public health conditions improve (e.g. reduced childhood mortality), they consume more animal products (bad), but the population growth also goes down, leaving less people to consume animal products in the long run (good).

As an example, in 1982, the average Chinese citizen ate 13 kg of meat per year, while in 2016 they ate 63 kg of meat per year, a figure predicted to increase to 93 kg per year if the trend continues uninterrupted. For the good effect to outweigh the bad effect, development in China would have had to result in the population being around 5 times smaller than it otherwise would have been without development. This is simply unrealistic, which is why I conclude that the negative effects (for non-humans) of the rich meat eater problem vastly outweigh the positive.

The rich meat eat problem has been part of my transition from focusing on near term human causes to long term non-human causes. In addition to the (in my opinion) vastly greater scale of suffering of non-humans, the greater neglectedness and the high degree of solvability, I worry that helping humans might harm non-humans.

Appendix B – Impact of not being vegan on increasing human deaths

This wasn’t directly related to the initial question so I put it in an appendix, but obviously if donating to an animal charity designed to reduce animal product consumption can save human lives, it follows that consuming animal products increases human deaths. I wanted to consider this additional question while I had the relevant figures available.

Using elasticity estimates, ACE estimates that one person consuming 30 fewer land animals (the average number of animals killed for their flesh consumed by a US meat eater in a year) results in 1.8-21 fewer animals farmed. This was calculated using Guesstimate, and so using the same methodology above (taking the mean of 30 simulation runs), I get 7.8 fewer animals farmed per year per person not consuming land animals.

This is 0.00000000975% of the total 80 billion land animals farmed per year. We can now multiply this by the number of humans expected to die via each of the above pathways. If we are ignoring the increased risk of your own death, that leaves us with climate change and antibiotic resistance.

As above, given that we expect to be able to reduce deaths due to climate change per year from 2030 to 2050 by 475 per year (by eliminating all animal farming in a given year), multiplying it by the above percentage gives us 0.000000046 humans saved for a single year from 2030-2050, or 0.00000097 in total from 2030-2050 per year you are vegan.

As above, given 280,000 deaths per year that we expect to be able to reduce today due to antimicrobial resistant diseases, multiplying it by the above percentage gives us 0.0000273 humans saved per year you are vegan.

This gives a total of 0.0000283 humans killed per year one isn’t vegan. To put this another way, there is around a 1 in 35,000 chance per year that one not being vegan will result in the death of another human.

For comparison, driving a car for 400 km in Australia increases one’s risk of death by 1 in 1 million. The average Australian drives 15,530 km per year, resulting in an increase in one’s risk of death by ~39 in 1 million, or 1 in 25,757. This makes the risk of dying in a car crash per year a little higher than the risk of killing someone by not being vegan pear year. I don’t know whether knowing this would make someone more or less likely to be vegan (or drive), but it might be useful for your decision making.

Note that this particular figure is almost entirely dependent on the antibiotic resistance pathway being the case, which while theorised, hasn’t yet been proven. If animal agriculture didn’t contribute meaningfully to antibiotic resistance, the benefit to humans of one person being vegan would only be around 1 in 1 million.

* For simplicity, all monetary values will be in US dollars.

Leave a Reply

Your email address will not be published. Required fields are marked *