The openness of effective altruism

Effective altruism is a global social movement that uses science and reason to do the most good. Perhaps the most striking characteristic of the effective altruism movement, that differentiates it from other social movements, is its high degree of openness. This openness is reflected in three facets.

1.     Openness to the world: scientific research, new opportunities

Effective altruists are constantly looking for new, more effective and important opportunities to help others. In order to find such opportunities, they use the scientific method: critically and impartially looking at the world and letting the data speak. In this sense, the effective altruism community strongly resembles the scientific community. This can be contrasted with other social movements. The environmental movement has an aversion for ‘unnatural’ technological innovations and is not open for ideas such as nuclear power or genetically modified organisms, the feminist-antisexism movement has an aversion for evolutionary biology and is not open for men’s rights issues, the antiracism movement is reluctant to criticize discrimination and rights violations in the Islam, the animal rights-antispeciesism movement rejects a reducetarian strategy and renounces research about interventions in nature to decrease wild animal suffering, and social justice movements are often critical about free trade, economic analyses of social problems and market mechanisms to solve social problems. These movements have good intentions, but they lack the scientific approach (they reject nuclear physics, evolutionary biology, economics,…).

2.     Openness to others: criticism, changing minds, allowing people in the movement

The effective altruism movement also resembles the skeptical movement which values critical thinking. This can be seen in the effective altruism community, which has a culture that facilitates and encourages ‘changing your mind’. It acknowledges that emotions can make us less effective. The atmosphere at an effective altruist conference is different than at e.g. an animal rights conference. At an EA conference, people are very open to criticism by others. Instead of dismissing critique, effective altruists have a reaction rather like: “Very interesting, tell me more…!” When criticism is welcomed (as long as the criticism is evidence based, rational or reasonable), people with different opinions are welcomed in the movement.

3.     Openness towards others: transparency

To facilitate critical thinking and encourage critique by others, it is very important that effective altruist ideas are explained as open as possible. One cannot constructively criticize vague or hidden ideas. The high level of transparency of opinions and analyses in the effective altruism community is striking. See for example analyses by Open Philanthropy Project, GiveWell, Animal Charity Evaluators, 80000 Hours, Effective Altruism Funds, Rethink Priorities and other effective altruism organizations, as well as members from the EA community. Effective altruists frequently communicate about their own confidence levels or state their feelings of uncertainty that they have about their own theories. They present counterarguments why they could be wrong, explain when you should not donate to their preferred charities, discuss why they changed their minds and even explain why they sometimes choose to be less transparent. Such reasoning transparency standards at effective altruist organizations are often higher than in other social movements.

Geplaatst in Artikels, Blog, English texts | Tags: , | Een reactie plaatsen

Wild animal suffering, longtermism and population ethics

Should we prioritize wild animal welfare research? If we take a long-term perspective, the problem of wild animal suffering is likely to become very important and perhaps even dominant. But judging whether we should prioritize far future wild animal welfare depends on crucial probabilities, such as the likelihood of human extinction due to global catastrophes, the likelihood of a population ethical theory being valid, and the likelihood of animals having net-negative lives with more negative than positive experiences. These probabilities are crucial considerations in the sense that they strongly influence our cause prioritization. They are the consequence of our factual (empirical) and moral (ethical) uncertainties: we do not know for sure which empirical facts of the world are true and which moral rules of ethical theories are valid.

In this article, I give my personal estimates (credence levels) of the crucial probabilities that influence the importance of far future wild animal welfare. I will also explain a few crucial factors that are important to influence far future wild animal welfare.

Factual uncertainties and far future probabilities

If we avoid extinction, far-future human and animal lives will vastly outnumber the current human and animal generations. Hence, most of the experiences (of happiness, pain and everything else valued and disvalued by future individuals) will be in the future. The current moment is almost negligible. This is the case for longtermism.

Based on some surveys and estimates by existential risk researchers, I assume that the probability of humanity going extinct this century is less than 25%. If humans do not go extinct this century, I assume it is most likely that humanity continues its past track record: decreasing violence, war, diseases, famines, deaths from natural disasters, and increasing happiness, longevity, wealth, education, arts,… We will likely have more technology, scientific knowledge and wealth to guarantee more and more far future human welfare and further reduce the risks of natural disasters such as climate changes, environmental collapses, global pandemic infectious diseases, asteroid impacts, supervolcano eruptions,…. We are more likely to colonize other planets, which means an even further decrease of the probability of human extinction and an expansion of the natural resource pool. The more centuries we survive, the more likely we are to survive extra centuries, so the extinction probability in future centuries decreases. Therefore, the far future is likely to contain a huge number of humans with (highly) positive welfare lives.

If we do not go extinct, those future humans can use technologies to improve animal welfare. So what about future animals? First, there are the domesticated animals, used by humans for food, clothing, experimentation and entertainment. Almost all of those used animals are used for food. Considering factory farms (including aquaculture), I think it is more than 75% likely that those animals have net-negative lives.

However, with new technologies, based on cell culture and tissue engineering (cellular agriculture), we can replace all animal products (meat, fish, eggs, dairy, leather,…) and develop e.g. human-on-a-chip experimental devices that replace animal testing for new drugs and toxics. Also our concern for animal welfare and the consumption of vegan, animal-free food products increases. So I expect that future animal use will be drastically reduced and livestock farming could be eliminated (outcompeted by the animal-free alternatives). I assume the likelihood of livestock elimination is more than 50% per century. That means in the far future it becomes very likely that we will no longer need to be concerned about animal suffering in factory farms.

Elimination of livestock farming and other uses of animals by humans does not mean elimination of wild animal suffering. Given the dominance of fast reproduction strategies and short lifespans (premature deaths) of wild animals, I think it is more than 50% likely that most wild animals have net-negative lives: short lives with negative experiences of hunger, diseases, injuries, parasites, predators,… That means most sentient beings born in the far future will have lives not worth living, dominated by negative experiences. If the probability of human life going extinct is lower than 50%, the probability of all sentient life going extinct is even lower. And if sentient life does not go extinct, there will be a huge number of sentient beings in the far future.

In summary, these are my probability estimates about factual uncertainties:

  • Human population going extinct: lower than 50%
  • Far future humans having mostly net-positive lives: higher than 50%
  • Livestock population going extinct (i.e. factory farm and aquaculture elimination): higher than 50%
  • Livestock animals having mostly net-positive lives: lower than 50%
  • All wild animal populations going extinction: lower than 50%
  • Wild animals having mostly net-positive lives: lower than 50%.

Moral uncertainties and population ethical probabilities

To fully estimate the importance of far future wild animal suffering, we need to consider population ethics. There are many moral theories in population ethics, so we face moral uncertainty about the validity of those moral theories.

The simplest population ethical theory is total utilitarianism, which aims for maximizing the total sum of everyone’s utilities (including those of future generations), whereby utility is an overall measure of someone’s preference for a certain situation or option. However, this theory has a counterintuitive sadistic-repugnant conclusion. Suppose we face the following dilemma. In option 1: a large group of people has a maximum positive welfare (highest utility levels, with maximum happiness) and no-one else exists. In option 2, that same group of people has a maximum negative welfare (lowest utility levels, with maximum suffering), and a huge extra number of people are born, who will have lives barely worth living (very small but positive utility levels). If this extra population in the second option is large enough, the total welfare or utility in option 2 is larger than in option 1, which means total utilitarianism prefers the second option. That seems highly counterintuitive to me.

To avoid this sadistic-repugnant conclusion, ethicists proposed many other population ethical theories. These theories can be unified or summarized in terms of variable critical level utilitarianism. Existing people can choose their own critical levels in different options. (Non-existing people always have zero utility and zero critical levels). Someone’s relative utility equals the own utility level minus the own chosen critical level. Variable critical level utilitarianism aims for maximizing the total sum of everyone’s relative utilities.

If everyone chooses zero as their critical level, for all options, we get total utilitarianism. But it is possible that in option 2 of the above dilemma, the first group of people chooses a very high positive critical level, which means their relative utilities become so negative that they trump the small positive utilities of the many extra people. In that case, the first option is preferred. Equivalently, it is also possible that the extra people in option 2 choose their actual utility levels as their own critical levels, which means their relative utilities are zero instead of positive and again option 1 is preferred. The latter possibility is known as person-affecting utilitarianism, which says that we should make (existing) people happy instead of making (extra) happy people.

With variable critical level utilitarianism, when some people choose a positive critical level in some situations, we can avoid sadistic-repugnant conclusions such as the above dilemma. When some critical levels are positive, there is a discounting of utilities: if people have positive utilities and they choose a positive critical level, their utilities are discounted (they gain less weight in the overall evaluation). If they have negative utilities, their utilities are extra emphasized (they gain extra weight).

Considering the abovementioned sadistic-repugnant conclusion, I consider the likelihood of total utilitarianism being valid as lower than 50%, and the likelihood of the more person affecting views of utilitarianism (or variable critical level utilitarianism where people choose a maximum safe critical level) being valid as higher than 50%. That means the probability that the utilities of future net-positive lives (but not the future net-negative lives) should be discounted is more than 50%.

In the far future, the moral weight of the net-negative lives is likely to become dominant. That means we should prioritize the welfare of future sentient beings who have net-negative lives, and we should prioritize the avoidance of the births of individuals with net-negative lives. (This does not mean that we should prefer non-existence of all future generations, because the current existing generations can set their critical levels in such a way that such non-existence of future generations gets a lower total relative utility.)

Together with the abovementioned probabilities about our factual uncertainties, the probability of variable critical level being valid implies that far future wild animal suffering becomes dominant: wild animals are most likely to exist (not go extinct), they are most likely to have net-negative lives, and they are most likely to have their utility levels not discounted.

Factors that are important for far future wild animal welfare

What can we do to improve far future wild animal suffering? Which trajectories can our society take to maximize the likelihood that the welfare of far future wild animals is as high as possible? There are a few general factors that determine our future trajectory. Two most important factors relate to an increase in values (concern for wild animals) and in means (technologies for intervention).

  1. Expanding our moral circle: increasing awareness of animal sentience and promoting animal rights and welfare. One can support for example Sentience Institute.
    • As an instrumental goal, one can support veganism. This decreases our use of animals, which decreases our cognitive dissonance that distorts our moral perceptions, which means we become more open to the idea of animal rights.
  2. Doing research in welfare biology: looking for safe and effective interventions in nature that increase wild animal welfare. One can support for example Wild Animal Initiative or Animal Ethics.
    • As an instrumental goal, one could support e.g. economic growth. This may seem counterintuitive, as it is often stated that economic growth harms animals, through environmental destruction or increased meat consumption levels. However, the impact of habitat destruction on overall wild animal welfare is far from clear, more technology driven growth means more growth in technologies that are good for the environment and for animals, and in many highly developed economies we see a decrease in meat consumption. The economic growth concerns are merely short-term. Looking at the long term however, economic growth can have a huge positive impact, for example by facilitating research in welfare biology. If we are poor, we will not be inclined to invest in research how to intervene in nature to improve wild animal welfare in the far future. However, if we are very rich, we can afford to spend a little bit of money on wild animal suffering research. The richer we are, the more likely we spend some money. As far future wild animal suffering is the most neglected area of suffering in the world, any additional resources invested in improving wild animal welfare can do comparatively a lot more good than resources going to smaller and less neglected areas of suffering. Hence, economic growth increases the likelihood that we will do research and invest in technologies that tackle the largest and most neglected area of suffering. Once we invent those technologies, they can help huge numbers of wild animals for millions of years in the future.
    • Another instrumental goal, is avoiding human extinction. If humans go extinct, wild animals lose their only hope, because those animals will not be able to do research in welfare biology and invent technologies to improve wild animal welfare.

Geplaatst in Artikels, Blog, English texts | Tags: , , , | Een reactie plaatsen

Why your neighbor cares more about animal rights than you think

Recently, I listened to a very interesting episode of the You Are Not So Smart podcast, about pluralistic ignorance: a group conformity effect where a majority of the population privately rejects a norm, but publicly accepts that norm because they incorrectly assume that the majority accepts it.

The issue of animal rights and meat consumption offers a good illustration of pluralistic ignorance. I do a lot of ‘deep canvassing’ conversations with people on the streets about animal rights and veganism. As a starter, I ask people to rate on a scale from 0 to 10 how important animal rights are according to them: a 0 means that they believe animals do not deserve any rights, we can do with animals what we want, a 10 means that animals deserve equal rights as for example children, i.e. basic rights to life, welfare and liberty. Answers vary mostly from 7 to 10, with an average close to 8. Then I tell my interlocutors that I asked this question to many other passersby on the street. The interlocutors have to guess the average rating given by those other people. Most of the interlocutors believe that other people give a lower rating than they do, e.g. a 6 instead of an 8. In other words: most people care about animal rights, but they believe that other people care less about animal rights than they do.

At the end of the conversation, I ask people a similar question: how strongly do we as a society have to shift our diets on a scale from 0 to 10? A 0 means that what we eat in our society is fine and we should not change our diets, a 10 means that we should immediately move towards an animal-free agriculture and diet. Or a related, more personal question: how motivated are you to reduce the consumption of animal products on a scale from 0 to 10? A 0 means you will not change your diet, a 10 means you will go vegan right now. Most often, when this point in the conversation is reached, people give an answer higher than 5 (the average could be between 6 and 7). But then they have to answer the question about what they believe other people would respond. Most interlocutors believe that other people would give a score lower than 5. In other words: most people are open to reduce their own meat consumption and try some new vegan products and recipes, but they believe that other people are more stubborn meat eaters who are less open to plant-based food options.

This is of course very preliminary, rudimentary research data, but it confirms the theory of pluralistic ignorance. It has two important implications.

First, people are often motivated by what other people think. People are susceptible for groupthink and peer pressure. If people erroneously believe that other people do not care about animal rights, they also start to care less. And if as a consequence people publicly demonstrate that they care less about animal rights than they privately do, other people might erroneously believe that no-one else cares about animal rights. So those other people also start to care less. This self-reinforcing mechanism generates a suboptimal equilibrium: everyone publicly follows a norm (e.g. to eat much meat or allow animal rights violations) that everyone would privately reject. By informing people that other people also do care as much about animal rights, are open to more new vegan options and prefer a decrease or elimination of meat consumption, they become more inclined to change their behavior and reduce their own meat consumption. So animal rights activists and vegan outreachers should inform people that those people’s pro-animal values and beliefs are shared by many other people. Then we can have a paradigm shift towards a more optimal equilibrium where everyone publicly follows the norm that everyone privately holds.

Second, there is a catch. Consider an animal rights activist giving a classroom lecture about veganism. Such humane education in classrooms is risky: it can be effective if done well, but it can also backfire. Students feel strong peer pressure from other students. One student can privately agree with the animal rights speaker. But it takes only one opposing student to generate a dangerous group dynamic. This one opponent raises his voice and publicly argues against animal rights and veganism. Now the other, agreeing student, who privately agrees with animal rights or veganism, believes that the other students are against it. The other students belong to the ingroup of the agreeing student, whereas the speaker is just a stranger. So this agreeing student is inclined to follow the other students instead of the speaker. It is even possible that the privately agreeing student starts to publicly defend the opposing view, the anti-veganism norm that is publicly held by the other students. Psychologists call this ‘false enforcement’. As a result, the students in the classroom become more convinced opponents of animal rights. This is a backfire effect: the vegan outreach by the speaker is counterproductive.

The same happens in other groups, such as families or partygoers. Imagine a party or family dinner where everyone eats meat, and there is one vegan. The other partygoers might privately agree that the vegan makes the good choice. But if one partygoer is critical and publicly argues against veganism, the other partygoers might associate themselves with this opponent and start to consider the vegan as an outgroup member. They can expel the vegan and identify themselves more strongly as meat eaters. This reinforces the dynamic, generating a stronger pluralistic ignorance where veganism is rejected more strongly by group members.

The You Are Not So Smart podcast episode included a very tragic example of an extreme backfire effect: the mass-suicide of the Jonestown commune led by Jim Jones. Before their decision to commit suicide, there was a public discussion in the meeting hall. One person, Christine Miller, stood up and argued that there were other options than suicide. But in doing so, she was ostracized by the other cult members. First, a few other group members criticized her for defying their leader. And then many other group members followed those critics and also started signaling that they were willing to follow their leader and die in order to show their loyalty to the group.

To counter this backfire effect of pluralistic ignorance, we have to avoid such negative, self-reinforcing dynamics. The best way to do this, is by personal conversations with only a few people at a time. Instead of speaking about animal rights and veganism in large groups (parties, classrooms), deep canvassing on the streets is an example of conversations with one or two people. Those interlocutors will not hear the voice of an opponent. And they can be informed about the hidden, pro-animal values and beliefs held by the majority. The same goes for e.g. media campaigns with undercover investigations or animal rights marches. Most of the time, people are alone when they watch those images on TV. Or they are alone when they read about animal rights in the newspaper. Then they see that many other people (e.g. the protesters) do care about animal rights. This can create a snowball effect, where the public support of animal rights increases. More and more people become aware that more and more people care about animal rights. That means more and more people are prepared to privately change their behavior (e.g. by decreasing meat consumption) or publicly demonstrate their care, e.g. by joining the animal rights protests. Of course, it is important that those protests are done well: the protest method should not create a disapproval but should be in line with the privately held values and norms of the public (e.g. in terms of non-aggressiveness, communication style, dress code,…).

Geplaatst in Artikels, Blog, English texts | Tags: , , , | Een reactie plaatsen

Carbon offsetting versus meat offsetting

A carbon offset is a payment to avoid greenhouse gas emissions by others elsewhere in the world according to the same amount that one has emitted oneself, in order to neutralize one’s own carbon footprint. There are several effective ways to compensate your carbon footprint (although carbon offsetting is focused on avoiding harm instead of doing good, and there are probably more effective ways to do good than to avoid harm, which means carbon offsetting is not the most effectively altruistic thing to do).

If carbon offsetting is permissible, is meat offsetting also permissible and positively effective? Meat offsetting means paying others to reduce their meat consumption at the same amount of ones own meat consumption. If you eat one kilogram of meat, you can financially support a campaign by an animal rights organization that reduces meat consumption by others with one kilogram. From an animal rights perspective, meat offsetting is considered as illegitimate, comparable to murder offsetting: a murderer who kills someone and then pays someone else to save a life (e.g. donate to a GiveWell recommended charity), as if this would neutralize the harm done by the killing.

The question is: is carbon offsetting comparable to meat offsetting or murder offsetting, such that carbon offsetting is also irresponsible or immoral? Here I argue that there are several morally relevant differences between meat offsetting and carbon offsetting, such that the former is immoral but the latter is permissible and good.

1.     The social welfare optimum has a positive emission rate but a zero meat consumption level

The most crucial difference between carbon offsetting and meat offsetting, is that the social welfare optimum of meat consumption is at a zero consumption rate (i.e. no meat consumption) but the social welfare of greenhouse gas emissions is (at this moment, when we do not have enough climate neutral energy alternatives to fossil fuels) at a positive emission rate.

This difference focuses on social welfare. The social welfare function is a function of the activity level (e.g. fossil fuel combustion or meat consumption) and is calculated as the difference between aggregate welfare benefits and aggregate welfare costs to all sentient beings. Sentient beings include humans, livestock animals, wild animals and their future generations. The benefits involve happiness, improved well-being and saved lives, the costs involve suffering, loss of well-being and premature deaths. If the activity level increases, both the benefits and the costs increase. The benefits are usually a concave function (like a hill), the costs are usually convex (like a valley), so the difference between benefits and costs is a concave function with a maximum at a specific activity level (see figure).

carbon vs meat offsetting fig 1

The welfare benefits of carbon emissions: strong initial increase

Consider the benefits of carbon emissions. These are the benefits related to the underlying activities, such as agriculture and fossil fuel combustion for heating and transportation. There are both benefits for the agents (the emitters, those who perform the activity) as well as benefits for the patients (the populations at risk of climate change).

Concerning benefits for the agents: imagine living with a zero carbon footprint, i.e. with zero greenhouse gas emissions. That would be a huge loss of your well-being. Almost all food is produced with a non-zero carbon footprint, so you will probably starve to death. You cannot heat a room, so in some cold regions you could freeze to death. You cannot move to warmer places, because that requires carbon emitting motorized vehicles. Also building strong houses, building dikes, using air-conditioning, using health care, extinguishing fires and irrigating land all require greenhouse gas emissions (at this moment), so you will not be able to protect yourself against extreme weather events, hurricanes, floods, heat waves, infectious diseases, forest fires and droughts. This is basically the extreme climate change scenario we all want to avoid. Completely eliminating all fossil fuels at this moment could be as bad for the global economy and for human flourishing as extreme climate change. That means that fossil fuel use is very beneficial (at this moment, when we do not have enough alternatives).

But also for the patients, the populations at risk of climate change, some positive level of greenhouse gas emissions is beneficial. People in poor Southern countries are most affected by climate change, but these people can also benefit from international trade. No fossil fuels means no transport, means no export, means lower incomes for those poor people. Those poor people can also benefit from medicine and agricultural technologies, but technological research without fossil fuel use will be much less productive. The same goes for future generations: they are also mostly affected by climate change, but they also benefit from new technologies. Fossil fuel use increases economic growth (through facilitating new technologies), and future generations benefit from economic growth. With economic growth, the consumption level increases, so future generations can consume more. Finally, economic growth allows us to invest more in new technologies and scientific research that can also benefit wild animals in the far future. If we eliminate fossil fuels, we will generate less new inventions that can benefit future human and animal generations.

These considerations mean that the benefit curve has a strong slope at low levels of activity: the difference in benefits between zero emissions and a little bit of emissions is large. Society will benefit a lot if emissions increased a little bit from zero to something. But the more we emit, the lower the extra benefits from an extra increase in activity will be, so eventually the benefit curve will flatten. There is a decreasing marginal benefit from carbon emissions. In our current situation, with large amounts of carbon emissions, society will not be harmed much if we decrease emissions a little bit, neither will society be benefitted a lot if we increase emissions a little bit.

The welfare benefits of meat consumption: very small

Now we can consider the welfare benefits of meat consumption. There are no benefits for the patients, the animals who are killed for their meat. There are benefits for the agents, the meat eaters, but these are small. Imagine living without meat consumption. You can still eat healthy and delicious vegan meat alternatives, so you will not die from starvation, and you can still experience taste pleasure. Perhaps you like animal meat a little bit more than vegan alternatives, but this benefit from animal meat is much smaller than the benefit we get from using fossil fuels. Without meat consumption, you can still visit friends, travel, have a warm shower, buy clothes,…

The welfare costs of carbon emissions: slow initial increase

Carbon emissions have an increasing marginal cost: if emission levels are low, an extra ton of CO2 emitted will cause less harm, compared to a situation when emission levels are high. At very low or zero emission levels, an extra emission is basically harmless, because the Earth has a finite CO2 absorption capacity. Imagine there are no greenhouse gas emissions, and you start to emit one ton of CO2. This emission will cause no harm, because that CO2 can be easily absorbed by the soil, the trees or the ocean. But this biosphere absorption capacity is finite, so emissions become harmful above a certain level. There is a planetary safety boundary for the amount of CO2 in the atmosphere, and this boundary is not at zero CO2 but at a positive level. The more we transgress this boundary, the more harmful an extra ton of CO2 becomes. This means that the cost curve is convex (like an upward facing valley slope).

The costs of climate change are borne by the patients: the populations at risk of climate change. These include both humans and wild animals. The impact of climate change on wild animal welfare is far from clear: it can be negative but also positive. For example, if most wild animals have net-negative lives, i.e. lives with dominating experiences of suffering instead of happiness, and if climate change decreases population sizes of wild animals, climate change could in fact decrease total wild animal suffering, because fewer animals with net-negative lives will be born. Due to this highly uncertain impact of climate change on wild animal welfare, I will only focus on scientifically more established impact evaluations on human welfare.

The harm caused to humans by climate change can be measured in economic terms as the social cost of carbon (SCC): the dollar value of human harm caused by the emission of one ton of CO2. There are many estimates of the SCC, with wide ranges of variation. An average level is around 50 dollar per ton CO2, increasing yearly with 3% (because the cost curve is convex and CO2 accumulates in the atmosphere).

This social cost includes mostly economic damage, such as loss in world GDP, but it sometimes also includes the human welfare loss in terms of premature mortality. This can be done using the value of a statistical life which can be estimated using our willingness to pay to avoid a risk of dying. How much are you willing to pay for extra road safety, such that your risk of dying by a car accident decreases with say one in a million? Using these estimates, the value of a statistical life in rich countries is lower than ten million dollars. The important fact is: this is a finite (not infinite) number, because we are not maximally risk averse, we do not take overly extreme precautionary measures at infinite cost, we do not simply abolish all risky activities like walking up a staircase, driving a bike or taking a shower (these are all potentially deadly activities).

Related to the notion of the value of a statistical life, is the notion of disability-adjusted life years (DALYs): the number of years lost due to ill-health, disability or early death. There are some estimates of the DALYs from climate change. The highest estimate I encountered in the literature, was 3 DALYs per 1000 ton CO2[1]. Note that 1000 ton CO2 is roughly the amount emitted by an average person in a developed (rich, western) country throughout his/her life (13 ton per year for 80 years). A premature death means on average a loss of more than 30 healthy life years, i.e. more than 30 DALYs, which means that the lifetime emissions of an average person in a developed country results in less than 0,1 premature deaths due to climate change. Or in other words: 10 emitters kill one person. This is a high estimate, because most estimates are an order of magnitude lower: e.g. 0,6 DALYs (Broome, Climate Matters), 0,2 DALYs[2] or 0,12 DALYs (Giving What We Can) per 1000 ton CO2. That means 100 emitters kill one person. This is the mortality cost of climate change, but note that those 100 emitters also indirectly save lives. For example: by using fossil fuels, they contribute to economic activity, which means paying taxes, which means government income available for health care subsidies. This is one pathway included in the benefits of fossil fuel use.

Including the value of life or DALYs in the calculations of the social cost of carbon, and including potential harms from climate change such as increased murder rates or war casualties from climate conflicts, I would give a rough, higher-end estimate of the SCC of about 100 dollar per ton CO2.

This value assumes a very low (0,1% per year) pure rate of time preference. This low rate means that the welfare of future generations count as much as the welfare of the current generations. The rate is not zero, because there is a very low probability of human extinction (assumed to be 0,1% per year) due to other causes than climate change (e.g. a global nuclear war, an asteroid impact, a supervolcano eruption, a pandemic supervirus, rogue artificially superintelligent machines,…). If the human population goes extinct, there will be no further human harm from climate change, so no further climate costs.

The estimate of the SCC also assumes a continuing economic growth (increased productivity due to new information and technologies) of a few percent per year. Economic growth means that future generations can achieve higher levels of consumption. If we prefer a smoothing of consumption over time (i.e. a kind of intergenerational justice where consumption levels between generations are not wildly different), economic growth means that the consumption related welfare of future generations can be discounted a little bit. Without such discounting, our generation would have to sacrifice almost all consumption (i.e. save everything) in order to increase the consumption levels of future generations.

The pure rate of time preference and the intergenerational consumption smoothing with economic growth are the two terms of Ramsey’s social discount rate. This discount rate translates the dollar value of future climate costs into the present value. A cost of 1 dollar in 100 years counts less than a cost of 1 dollar today. The higher the social discount rate, the lower the present value of future climate costs for future generations become.

One can argue, based on population ethical issues, that the pure rate of time preference is larger than the extinction rate of 0,1% per year. Our society is a highly nonlinear, chaotic dynamic system, which means very small differences in choices have – through a butterfly effect – a large influence on who will be born in the future. Whether a boy or a girl is born depends on the exact timing of fertilization. Suppose we do not take climate measures, so we keep a high carbon footprint, e.g. by taking the car instead of the bike or the train. As a result, future generations will have a lower welfare due to climate change. They do suffer from extra extreme weather events, but suppose that overall their net welfare is still positive, i.e. they still have lives worth living, with more positive than negative experiences. Can they complain against our choice to emit a lot of greenhouse gases? No, because in the alternative universe, where we would have reduced our carbon emissions, these future people would not exist, and other people would be born. A couple takes the train from work, arrives home a bit later, goes to bed and has sex a few seconds later, a sperm cell with a Y instead of an X chromosome fertilizes the egg, and a boy instead of a girl is born. Can we say that we have harmed future people affected by climate change when we did not take sufficient climate measures? In a sense no, because the affected people have net positive lives and the alternative choice (taking climate measures) would mean that those affected people would not exist. In summary: the stronger this butterfly effect on reproduction, the higher the pure rate of time preference becomes. And the higher this time preference rate, the lower the social cost of carbon becomes. So the SCC could be lower than 100 dollar per ton CO2.

The welfare costs of meat consumption: strong constant increase

Meat consumption has huge costs, not only for the hunted or slaughtered animals, but also for humans. Meat consumption contributes to chronic diseases, infectious zoonotic diseases and climate change. In terms of economic welfare costs for humans, eating 100 grams of meat costs 4 dollars, eating one day vegan saves 8 dollars in wealth (measured as our willingness to pay for a good health and environment).

But most of the welfare costs of meat consumption are borne by the killed animals. Here we see a strongly increasing, linear cost function. The function is linear, because killing and eating an extra animal increases the harm done to the animal a lot and is independent from how many other animals were already eaten.

The social welfare of carbon emissions and meat consumption

The figure below shows the benefit and cost functions of carbon emissions and meat consumption, based on the above explanations. The difference between benefits and costs is the social welfare.

carbon vs meat offsetting fig 2

Now we can see the clear distinction between carbon offsetting and meat offsetting. Carbon emissions have a positive social optimum level, meat consumption has a zero social optimum level. With a positive level, a complete prohibition of the activity would not be optimal. To reach the socially optimal activity level, a price mechanism is economically efficient. There is a consensus among economists that a market-based mechanism such as a carbon tax is the most economically efficient way to reach climate targets and avoid dangerous climate change. Carbon offsetting can be considered as a kind of voluntary carbon tax, and the social cost of carbon could serve as an optimal tax rate for carbon emissions, so the general, average carbon offsetting price would be around 100 dollar per ton CO2 emitted, increasing with 3% per year.

As mentioned above, from an animal rights perspective, meat offsetting is like murder offsetting. Do economists propose a murder tax? No: given the high welfare costs and low benefits of murder, the optimal tax rate of murder is basically infinite. The social optimum level of murder is zero.

The fact that the optimal meat consumption level is zero, means that meat offsetting is not universalizable: only if all meat eaters offset their meat consumption, we reach the optimum level. But at this level no-one will eat meat anymore so meat offsetting becomes superfluous. Carbon offsetting is different, because there is a non-zero optimal social welfare level of carbon emissions. People can still emit some amount of greenhouse gases, and offset those emissions to reach the social optimum emission level.

One nuance is important here. As mentioned above, the social cost of carbon, and hence the general average offsetting price is around 100 dollar per ton CO2. However, most offsetting mechanisms are much cheaper. The most cost-effective mechanisms range around 1 dollar per ton CO2. This means there is at this moment low hanging fruit that is not picked by others. If no-one else picks those most cost-effective offsets, you can. But this is not universalizable: not everyone can offset his or her own emissions at such low costs. If more low hanging fruit is picked by others, the remaining higher hanging fruit will be more costly, so the carbon offset price will increase. In the end, if everyone would offset their own emissions in line with the social optimum (according to the climate targets), the offset price will be around the social cost of carbon. If the offset price becomes higher than the SCC, it is no longer economically efficient to offset: the costs of further offsetting become higher than the benefits (the avoided climate change).

There is one offsetting mechanism that is always universalizable, because it shifts the social optimum level: investments in new technologies. Some technologies decrease the cost function and hence shift the social optimum to the right. Consider animal-free, cultured (lab-grown or clean) meat: with new technologies, we can produce and consume meat, without causing animal suffering and without having to kill animals. In this case, the cost function could drop down to zero (or to the production costs of cultured meat), and we can eat as much cultured meat as we want (or with non-zero production costs: as much as where market demand equals production costs). We can also make carbon emissions harmless, by investing in carbon capture and storage techniques. The only cost of carbon emissions becomes the cost of capturing and storing that CO2. If this becomes cheap, we can continue emitting a lot of CO2. Or instead of carbon emissions, we can consider energy use as the activity: we can invent new clean energy technologies that do not have carbon emissions, and again we can use as much of that energy as is available without causing climate harms.

Investment in research and development for clean energy technologies is highly effective, because the rest of the world, for all future generations, can benefit from those new technologies. However, currently only 3 dollar per person per year is spend globally on clean energy technology R&D. Suppose we consider a carbon offset at the social cost of carbon price of 100 dollar per ton, and we invest this in clean energy R&D. If everyone would do that, we would have extra investments worth 700 dollar per person per year (7 ton CO2/person/year x 100 dollar/ton CO2). Total investments in R&D will increase with a factor of 240. With such high levels of technological research, we are very likely to invent technologies that eliminate the climate change problem and shifts the social optimum level of energy use very far to the right. Hence, such a strategy is very universalizable (comparable to planting more fruit trees, in the fruit picking analogy).

2.     Carbon offsetting does not involve basic rights violations, meat offsetting does

The above considerations looked at social welfare. This is based on a welfarist (consequentialist or utilitarian) ethic. But we can also see a distinction between carbon emissions and meat consumption from a rights based (deontological) ethic. There is in fact one basic right (unique in the sense that it does not impose negative externalities on others): the right not to be used as a means to someone else’s ends against your will. This is a right to bodily autonomy: we should not use your body against your will.

Climate change does not violate this basic right: if the victims of climate change were not present, one could still emit greenhouse gases, so the presence of the bodies of climate victims is not necessary to reach our end (our activity to emit gases). But meat consumption does violate this basic right: if the killed animals were not present, they could not be killed and eaten, and hence the goal of meat consumption could not be reached. The presence of the body of the slaughtered animal is necessary for meat consumption. No body means no meat.

The total cost includes not only the loss of welfare but also the violation of rights. If we consider basic rights violations as very serious, the total cost of meat consumption strongly increases, but the total cost of carbon emissions remains the welfarist social cost of carbon.

Meat offsetting involves basic rights violations, because it involves meat consumption. Carbon offsetting is not a basic rights violation and is therefore permissible from a rights based ethic.

3.     Greenhouse gas emissions involve a collective harm that lessens personal accountability, meat consumption does not

The above two arguments looked at two ethical theories: a welfare based and a rights based ethic. Apart from this, we can also consider how to morally judge people who offset their carbon emissions or meat consumption.

Suppose there are only a few humans, and you are the first one to start emitting 1 ton of CO2. This does not cause harm to the other humans, because the Earth’s biosphere can cope with those emissions. But if there are many other emitters, one extra ton of CO2 contributes to extra harm (cfr. the convex curvature of the cost function). In other words: the fact that you cause harm, depends on the presence of other people who perform the activity. With a much smaller human population, your carbon emissions would not be a problem. With the current population size, they do. But you are not responsible for the size of the human population. Your harm is part of a collective harm, collectively caused by many people. The fact that you are not responsible for the presence or absence of other emitters makes it difficult to hold you individually responsible for the harm.

Meat consumption is different: if you eat an animal, this causes extra harm to that animal (a premature death), no matter how many other meat eaters there are. Even if you were the only human, if you kill and eat an animal, you would cause harm. That means you can be held individually accountable for killing that animal.

[1] Egalitarian perspective, used in SimaPro for life cycle analysis, based on Goedkoop M. e.a. (2009). ReCiPe 2008. A life cycle impact assessment method which comprises harmonised category indicators at the midpoint and the endpoint level. Report I: Characterisation. Ministry of Housing, Spatial Planning and Environment, the Netherlands.

[2] Anthony J. McMichael A, Campbell-Lendrum D, Kovats S, Edwards S, et al. Global Climate Change. In: Ezzati M, Lopez A, Roders A et al. Comparative Quantification of Health Risks, Global and Regional Burden of Disease Attributable to Selected Major Risk Factors, Geneva, World Health Organization, 2004, pp. 1543-1650.

Geplaatst in Artikels, Blog, English texts | Tags: , , , , | Een reactie plaatsen

Climate activism, alarmism and effectiveness

In recent years, climate activism showed a regained interest in direct action and civil disobedience. One example is Extinction Rebellion. In their communication, these direct action groups often appeal to alarmism with doom scenarios about apocalyptic climate change. Direct actions could potentially be an effective form of climate activism, but the alarmism strategy is probably problematic. In this article I argue why alarmism is worrisome, and how alarmism can make direct action groups such as Extinction Rebellion ineffective or perhaps counterproductive.

The problems of alarmism

Alarmism is comparable to a badly adjusted accelerator pedal of a car. If you press it too little, the engine will stop. If you press the accelerator pedal slightly harder, the engine will rotate in an overspeed like a sports car from the starting blocks. This lack of control of an engine with a hypersensitive accelerator pedal can be dangerous. Like the hypersensitive engine, there are two possible reactions with alarmism: it creates a feeling of apathy resulting in inaction or a feeling of impatience resulting in overaction.

The alarmism risks for the general public: apathy, distrust and polarization

The first reaction of alarmism is a feeling of apathy. This can lead to fatalism, cynicism, paralysis, hopelessness, resignation and a lack of motivation to take action. Like an engine that stops, potential activists fall silent because of their false belief that it is too late, that the deadline is passed, that the climate change problem will not be solved and is impossible to solve.

This reaction is confirmed in psychological research: using terrifying messages and doom scenarios (“the current system will kill us all”) is anything but convincing (see, for example, “Yes! 50 Scientifically Proven Ways to be Persuasive” by Noah Goldstein, Steve Martin and Robert Cialdini). The alarmist strategy is irrational because it does not really encourage action, so that climate goals are not achieved.

As a consequence of this reaction, the general public might even experience a backfire effect: people who hear an uncomfortable alarming message of a problem that is beyond their control (such as the global political system that causes climate change) get even more convinced of the opposite. For example, they will deny climate change even more strongly. They put the head in the sand and become more convinced climate skeptics.

Apathy not only results in a lack of motivation to take action, but also in distrust. Twenty years ago, climate activists said that we only have ten years left to solve the climate crisis. Ten years later, the climate activists didn’t acknowledge that time is up and further actions are futile. Instead, they postponed the deadline: they gave us another ten years to win the battle. Ten years later, they still say “act now”, indicating that the battle is not yet lost. The alarmism of the activists decreases their credibility. This is comparable to an unreliable fuel meter. Suppose the fuel on the dashboard indicates too often that the fuel tank is empty while it is still full. In the beginning you will go to the gas station unnecessarily often. But after a while you will ignore the fuel meter and just continue driving. Until the fuel tank suddenly really runs out of gas. That happens to doom thinkers: if they sound too many false alarms, it leads to disbelief.

Another societal problem created by alarmism, is increased polarization. The reaction of the climate activists is very different from the non-activists: the activists become impatient and overzealous, proposing more radical direct actions and revolutions. This creates a polarization in society with two camps: the climate activists versus the climate skeptics. This polarization results in group think and ingroup-outgroup biases that can increase other biases such as confirmation bias, worsening judgments and decision making. A polarized society creates more mutual distrust, becomes less effective in making the right decisions and increases the risk of group discrimination and violence.

The alarmism risks for the climate activists: impatience, overzealousness and thoughtlessness

The second possible reaction of alarmism is a feeling of impatience. As a result, activists propose more and more drastic measures to save the climate. The impatience means the activists do not want to take enough time to study the effectiveness and costs and benefits of their actions and proposals. They thoughtlessly jump into a climate campaign, without taking the time to critically think about it and look for scientific evidence of its effectiveness.

One example is the disinvestment campaign. Some Extinction Rebellion climate activists took part in disinvestment actions, for example to demand that universities disinvest in the fossil fuel industry. However, the effectiveness of disinvestment actions is questionable. If the university does not invest in the fossil fuel industry, other people will invest in it (because demand in the share market is very elastic, so other investors will immediately take up the slack of a decrease in share price due to disinvestments). If the rates of return in that industry are lower than in other industries, the market mechanism will automatically push towards disinvestment in the less profitable fossil fuel industry, so disinvestment campaigns become superfluous. If the rates of return of that industry are higher than other industries, the university will lose money if it disinvests, and that means they have less money left to finance research and development of climate-neutral clean energy technologies. Those new technologies will do more good for the climate than a disinvestment (i.e. a decrease in harm when instead of the university a more careless investor reaps the rewards of investments in fossil fuels). Hence, a university disinvestment can even be counterproductive (because the other investors in the fossil fuel industry will not use their high profit rates for important climate and technology research). This critique on disinvestment is related to the idea of mission hedging (for a short presentation of this idea, see this video).

Jumping into a specific campaign without taking enough time for critical reflection, is dangerous. According to preliminary research about cost-effectiveness and cost-benefit analyses of interventions (see here, here, and specific examples in human health, education, animal welfare, climate mitigation options, individual climate actions), we can expect that most campaigns, actions, interventions or policy proposals are ineffective, weakly effective or sometimes counterproductive, and a minority is highly effective. Chances are low that the first climate campaign that you stumble upon is highly effective. If – because of impatience – you immediately jump into that specific campaign and invest time and resources in that campaign, there is a high risk for a sunk cost fallacy or escalation of commitment. If evidence shows that the campaign is less effective, you might be reluctant to let it go and do something different that is more effective, because you will not be inclined to accept that all the time and resources that you invested in that campaign were for nothing.

The lack of critical reflection also translates into the vagueness of many recommendations by impatient climate activists. They propose to fight the system, to overturn the government, to dismantle capitalism, to stop economic growth, but they do not make it clear what those recommendations mean, how to exactly achieve them and what the alternatives are. They basically target the wrong enemies, such as economic markets, neoliberalism, capitalism or economic growth. In their critique of the economic system, they do not follow two strategies: they do not base their critique on a scientific consensus among economists, and they do not offer and discuss empirical economic evidence or theoretical economic modelling that help to make the choice how to change the system into something better. This can be contrasted with effective altruists who use critical thinking and scientific evidence to do the most good. Effective altruists are consistently looking for new reliable evidence to update their beliefs and change their minds if necessary. In this sense, the effective altruist culture is very different from the alarmist climate activist culture.

Alarmist climate activists claim to follow the scientific consensus, but there is no consensus on their alarmist claims (e.g. that we are all going to die because of climate change or that we have no time left) nor on their policy proposals. There is no consensus among climate economists that a revolution and a destruction of capitalism is required or more effective to avoid climate change. On the contrary, there is a strong consensus that liberal market-compatible mechanisms (e.g. a cap-and-trade system of emission permits or a carbon tax), as well as investments in technological research are effective and required. A reform towards a green liberalism or green capitalism is probably more feasible than a revolution towards a green, just and efficient non-capitalist system. For example there is no good evidence that socialist or anticapitalist systems are better than capitalist systems in terms of human well-being. The same goes for sustainability: sustainable choices require good incentives and knowledge, and well-functioning free markets, corrected for market failures, have a price mechanism that generates the correct incentives for consumers and producers and integrates information about preferences and costs. Hence, such markets efficiently allocate resources to make sustainable decisions. Climate change is probably the biggest market failure, and a carbon price can correct the market.

Instead of improving market mechanisms of carbon pricing (as done by e.g. Carbon Market Watch), the antimarket, antiliberal and anticapitalist climate activists criticize those effective policy proposals. This is another example of a counterproductive climate action. The same goes for their criticism of economic growth. Long term technological growth is driven by technological innovations and knowledge. With sufficient investments in climate technology research and development, a decoupling between economic growth and greenhouse gas emissions is possible. The economy can grow, even when harmful economic activities decrease (think about slavery, whaling, horse manure in cities, banned toxics, acid rain, ozone depleting chemicals,…). A long term economic growth in value, knowledge and technology does not require fossil fuel use or greenhouse gas emissions. Fossil fuels are depletable, whereas knowledge is non-rivalrous (my consumption of knowledge does not impede your consumption of the same knowledge). We need economic growth to increase valuable knowledge and technologies that make the economy more sustainable.

Plus, we also have to take into account the welfare of everyone in the far future. Economic growth can drastically improve the living standards of humans and animals in the far future. This is one of the reasons to be worried about climate change: with climate change, the economy can shrink with say 20% (damage from climate change could cost 20% of global GDP). This is a lot of wasted money and wealth, that could be invested in scientific research to improve lives. If we take the longtermist perspective, we see that the number of lives in the far future (if we avoid complete extinction) is much bigger than the number of people alive today, so a difference in wealth of 20% means a lot because it affects a lot of people.

Economic growth creates positive sum games or win-win situations, whereas zero or negative growth returns us to zero sum games (win-lose games) that increases inequality, harm and rent seeking (taking wealth instead of producing wealth). Economic growth becomes even more important if we take non-human animals into account. Technology driven economic growth improves the lives of humans, so we can avoid human suffering. But with technologies such as cellular agriculture (clean meat that replaces animal meat) and human-on-a-chip (that replaces animal testing), we can also drastically reduce human caused animal suffering (livestock farming, animal experimentation,…). But by far the biggest realm of suffering could still exist for a long time, even if we eliminated all human diseases, violence, and animal abuse: wild animal suffering, including insect suffering. The size of far future wild animal suffering is probably orders of magnitude larger than current human and livestock animal suffering. If we value well-being in an impartial way, also the well-being of insects in the far future matters. Even if the well-being of one insect at one day in the far future matters only a little, because there will be so many insects born on so many days in the future, their total well-being becomes very important. The problem is: if we are poor, we will not be inclined to invest in research how to intervene in nature to improve wild animal welfare in the far future. However, if we are very rich, we can afford to spend a little bit of money on wild animal suffering research (e.g. research in welfare biology). The richer we are, the more likely we spend some money. As far future wild animal suffering is the most neglected area of suffering, any additional resources invested in improving wild animal welfare can do comparatively a lot more good than resources going to smaller and less neglected areas of suffering. Hence, economic growth increases the likelihood that we will do research and invest in technologies that tackle the largest and most neglected area of suffering. Once we invent those technologies, they can help huge numbers of wild animals for millions of years in the future. A difference between maximum sustainable economic growth and a 20% lower growth due to climate change or a zero or negative growth as proposed by some climate activists, could mean the difference of huge amounts of suffering of wild animals in the future. Hence, we should not underestimate the importance of economic growth (see also the arguments by Tyler Cowen why we should prioritize maximizing sustainable economic growth).

Not only do a lot of climate activists target the wrong enemies (capitalism, economic growth), but they propose risky alternatives. Doom thinking can lead to an overreaction with more harmful measures. People take too drastic means in a final attempt to save the world. An example is the call for mass civil disobedience and revolution by Extinction Rebellion, claiming that mass civil disobedience is the only option left to avoid a catastrophe. However, those activists do not take into account all the costs of such revolutionary direct actions. They do not think like an economist, considering both direct and indirect (opportunity) costs.

Mass civil disobedience has many risks and opportunity costs that are not fully taken into account by the activists. Some examples of costs:

  • time: the long occupations and jail time of activists preclude other use of time (such as time for scientific research, lobby work,…),
  • direct costs for the activists: fines and court costs from the lawsuits that cannot be spend on other effective climate measures, such as afforestation,
  • direct costs for governments for law enforcement, resulting in a misallocation of government resources (intelligence services,…) towards law breaking climate activists instead of more important issues (e.g. terrorism),
  • indirect costs for the broad population: loss of prosperity due to obstruction of economic activities.

Some examples of risks:

  • risk of loss of goodwill or sympathy among the wider population (the action style can induce antipathy),
  • risk of losing respect for our democratic legal system,
  • risk of uncontrollable evolutions due to weakened governments (as history demonstrates, getting governments on their knees through a revolution can create a power vacuum that is unmanageable and difficult to channel in safe ways),
  • risk of losing time and postponing more effective actions.

The latter concern is striking: Extinction Rebellion claims that we do not have time for other alternatives and mass civil disobedience is the only option left. However, the question is whether we have enough time to follow the four steps of Extinction Rebellion: first look for and gather enough activists for mass civil disobedience, second get the government on its knees, third install citizen assemblies that discuss and decide climate policies and finally implement those climate policies. It might take too much time to convince enough people to become climate activists that break the law. Extinction Rebellion is so impatient, that they claim we need climate neutrality by 2025 (that’s not what the IPCC and climatologists say), but they won’t reach that target if we first have to wait for those civil disobedience actions and citizen assemblies. Extinction Rebellion is not able to give reasonable arguments or empirical evidence that the strategy of mass civil disobedience will be quicker to solve climate change than other climate policy proposals. On the contrary, the proposals of Extinction Rebellion are more revolutionary and their demands are stricter than for example a carbon tax, so it seems less likely that those demands will find more political support in the shorter term. This can be contrasted with an effective altruist who not only takes into account the size of a problem or the impact of a solution, but also the chance of success (the political and economic feasibility) of that solution.

Alternatives: effective climate activism

The climate measure with probably the biggest impact, is support for organizations that lobby for more government investments in research and development of climate technologies (such as clean energy and carbon capture and storage). One of the best analyses that argue for public clean energy R&D as the most effective climate policy, was done by Let’s Fund (also covered in a Vox article). Another interesting analysis of effective climate actions in the effective altruism community was done by Founders Pledge. These analyses are much more thought-out than many of the analyses by alarmist climate activists.

Clean energies facilitate the next most important climate measure, widely supported by economists: carbon pricing. Market mechanisms such as a carbon tax or cap-and-trade become more politically feasible if we have cheap clean energy sources. Organizations such as Citizens Climate Lobby lobby for a carbon dividend (a carbon tax with revenues paid out as citizen dividends). These are more concrete, safe and effective policy measures than the vague proposals of many alarmist climate activists, such as revolution or dismantling capitalism.

Extinction Rebellion also promotes citizen assemblies to improve democratic decision making. But I think there are more promising (simple, clear, concrete) proposals based on economic analyses, such as approval voting, quadratic voting or a futarchy with prediction markets. When it comes to estimating the impacts of climate change, prediction markets and superforecasters are more reliable and accurate than alarmist climate activists, so we can invest more in such prediction markets and superforecasting.

Geplaatst in Artikels, Blog, English texts | Tags: , , , , | 1 reactie

Some solutions to utilitarian problems

I have recently written a series of articles about problems in utilitarian ethics that are relevant for effective altruism.

A first article describes why I became a utilitarian and what kind of utilitarianism I support (i.e. preference, rule, variable critical level, variance normalized,…).

The second article deals with the problems of population ethics and argues for a variable critical level utilitarianism, a kind of critical level utilitarianism where everyone is free to choose his or her own critical level in each different situation. Total, average, critical level, negative and person affecting utilitarianisms are all different special cases of variable critical level utilitarianism. With variable critical level utilitarianism, we can avoid counter-intuitive problems in population ethics. This issue becomes crucial when we have to choose between avoiding actual suffering (e.g. of factory farmed animals today) versus increasing well-being in the long-term future (e.g. avoiding existential risks).

The next two articles deal with the problem of interpersonal comparison of well-being. The first discusses a general method of utility normalization, based on an analogy between measuring utilities and measuring temperatures. This applies to utility functions that have continuous inputs (perceptions or experiences). When inputs are discrete another method is possible that counts the amount of just-noticeable differences in utility. The utility function now looks like a multidimensional staircase where the steps can have different widths. With this method we can compare the utilities of for example insects with humans.

Finally, I deal with the more exotic problem of counting persons and conscious experiences. This problem becomes important when we deal with future conscious artificial intelligence and whole brain emulations, but it is also relevant when we discuss insect sentience or split-brain patients.

Geplaatst in Blog, English texts | Tags: , , | 2 reacties

About insect welfare and how to improve it

Recently, Rethink Priorities (a project of Rethink Charity) did an impressive review of our state of knowledge about invertebrate sentience. Considering the fact that 99,9998% of all animals are invertebrates[1], it is crucial to know whether they are sentient and can have positive and negative experiences.

Sentience implies two properties: having a consciousness (subjective experiences) and having a utility function (an internal goal function or reward system). A plant, a thermostat, an artificial intelligent machine or a robot are examples of things with a utility function but without consciousness. On the other hand, some experiences such as seeing a white wall or touching a table are conscious but neutral (no utility function): there is no reward or objective in seeing, touching, looking for or avoiding these things. Combining a consciousness with a utility function, we get valenced (positive or negative) mental states such as pain and pleasure.

Most of the invertebrates are very small roundworms and ringed worms, and it is not clear whether they are sentient. But as the Rethink Priority review shows, there is increasing evidence that arthropods are sentient and have valenced experiences. Arthropods are animals with an exoskeleton and a segmented body, such as insects, spiders and crustaceans. After reading their review, I have updated my personal guess about the probability of insects (in particular bees, fruit flies and ants) having valenced experiences to more than 50%.

Even if the probability of arthropods being sentient is considered much lower (e.g. less than 10%), the precautionary principle should be applied, because there are huge numbers of arthropods. At this moment, there are about 1018 terrestrial arthropods (mostly ants) and 1020 marine arthropods (mostly very small copepods as zooplankton). This can be compared with about 1010 humans. So if arthropods happen to be sentient and we erroneously believe they are not, we are neglecting huge amounts of welfare and suffering. When a lot is at stake, the precautionary principle is reasonable.

In this article I first present experimental results that are in my opinion the most convincing and amazing facts to demonstrate that insects (in particular bees, ants and flies) are sentient, i.e. they have a consciousness and a utility function. Next, I discuss the issue how to compare insect welfare with human welfare. Finally, I explore how we can avoid wild insect suffering (by bee protection) and cultured insect suffering (by avoiding the use of insects for food).

For more on insect suffering, see the work by Brian Tomasik here, here, here and here, and Wild Animal Initiative here and here.

Indications for a consciousness in insects

Although it is impossible to determine for sure whether an animal is conscious, clever experiments can indicate the presence of consciousness. We can use experiments to test for skills or behavioral responses that are never performed by humans who are believed to be in unconscious states, or that sometimes occur unconsciously in humans. For example there are unconscious learning processes[2], so learning does not necessarily require consciousness. Nociception is the ability to detect harmful stimuli, such as burning your finger. But pulling away your hand after touching a hot stove is an unconscious reflex. After a brief moment, you start to feel a burning pain in your finger. Pain is different from nociception, because pain involves a conscious state whereas nociception is an unconscious perception.

Integrated behavioral control system in the brain

Like the midbrain in vertebrates, insects also have a brain structure that integrates perceptual inputs (e.g. vision and touch) to create a neural simulation of the body position of an insect in space that allows for behavioral control. In other words: an insect can be conscious of its environment and its own body in that environment. This is demonstrated by looking at similarities between the functioning of brain structures between insects and vertebrates[3], and can be explained from an evolutionary perspective: those brain structures are an efficient solution to the basic problems of navigation.

Complex learning

Insects such as bees have complex learning skills that cannot be attributed to mere automatic reflexes or pre-programmed responses. They learn to use objects that they have never encountered before. For example bumblebees can learn to use small balls, rolling them to a target place to obtain a food reward. Bumblebees can learn this behavior from each other, and they can decide to use a ball with another color if that ball is closer to the target.[4]

Selective attention

A property of consciousness is selective attention: the ability to focus on one of several competing stimuli. This enables to ignore some stimuli and respond to the stimulus that is most relevant. Neural correlates of visual attention have been found in honeybees. When a bee has already seen something several times and it was not relevant for behavior, attention for that visual stimulus as well as activity in certain brain regions is reduced.[5]

Meta-cognition through uncertainty monitoring

Another indicator for consciousness is meta-cognition: being aware of one’s own mental states. One interesting mental state is the feeling of knowing something. There are experimental indications that insects such as bees and ants have an awareness of uncertainty. For example honeybees can learn to selectively avoid difficult choices that involve uncertainty.[6] They can learn to solve trials, a correct solution gives a reward (sugar), an incorrect solution gives a punishment (a bitter tasting chemical), but they can also decide to opt-out from the trial and receive neither reward nor punishment. When the difficulty of the trial increased, honeybees opted out more often because the risk of punishment increased. Using this option to opt out, honeybees improved their success-to-failure ratio. Hence, bees can perform rational behavior under uncertainty. In order to do that, bees need to monitor and evaluate their uncertainty. Also ants have the capacity for uncertainty monitoring.[7]

Indications for a utility function in insects

Pain can be distinguished from nociception because the former involves a consciousness. But merely having a consciousness is not enough for sentience, because conscious experiences and sensations can be neutral. For example, people with pain asymbolia or pain dissociation are able to consciously feel pain, but they do not have any negative evaluation or feeling of unpleasantness of that pain. For them, feeling a burned finger is like feeling a wet finger: the wetness of lukewarm water is neither positive nor negative. For positive and negative evaluations, a utility function is required. Some experiments indicate that insects have a utility function.

Making trade-offs

Animals can have a preference to avoid the unpleasantness of pain, but they can also prefer food (absence of hunger), safety (absence of predators) and so on. Sometimes they have to make trade-offs between different preferences. If they are able to make such trade-offs in a consistent way, they not only have a utility function, but their utility function is complex and integrated, involving several dimensions (preferences). Hence, being able to make trade-offs is an indicator for the presence of a utility function.

Fruit flies can make trade-offs. For example they choose to endure electric shocks in order to obtain access to alcohol. This is trading off a punishment for a drug.[8] They also trade-off safety for food: when fruit flies see an overhead shadow that resembles a predator, they disperse and hide to avoid potential predators. But if they are very hungry, they will continue eating from a food source.[9] Hence, fruit flies are able to weigh and evaluate benefits (food) and risks (predators), and that requires a unified utility function that measures both the preference for food and the preference for safety.

Communicating trade-offs

Even stronger evidence, is when insects are able to communicate their trade-offs to other animals. This is seen in the waggle dance of honeybees. When a bee has found a new food source, it faces a trade-off between abundance (how many flowers are there?), distance (how easy is it to get there?) and predation risk (how dangerous is it there?). This information is contained in the waggle dance to inform other bees.[10]

Self-administering drugs

Another behavior that is unlikely to be an unconscious process, is the self-administration of drugs. Insects like bees can get alcohol addictions, and they are able to actively look for alcohol.[11]

Sense of the future

Honeybees can not only make trade-offs between food and safety, but also between immediate versus deferred rewards. For example bees choose a larger droplet of sugar water received over five seconds above a smaller droplet received immediately.[12] These time trade-offs require self-control and a sense of the future.

Mood states

There is evidence that bees and fruit flies have mood states. The most impressive evidence is probably the pessimism bias of agitated honeybees. When humans are anxious, depressed or stressed, they often become more pessimistic in uncertain situations, which means they have increased expectations of negative outcomes. The same can be seen in honeybees who are agitated by shaking them, simulating a predator attack on a bee hive. These bees are learned to associate a reward (water with sweet sugar) with one type of odor and a punishment (water with bitter quinine) with another odor. When bees face a new, third type of odor, they can either expect a reward or a punishment, so they can choose to approach or withdraw from the water. This is a situation of uncertainty. The bees that were shaken became more pessimistic: they more often withdrew from the water, expecting a punishment. And these shaken bees had lower levels of happiness hormones such as dopamine and serotonine.[13] This cognitive bias is a measure of negative emotional states. In another experiment, honeybees became more aggressive when they were in social isolation, as if being in isolation increased their levels of frustration.[14]

Also fruit flies seem to experience mood states such as depression. Due to enduring, uncontrollable stress, they experience anhedonia, a loss of appetite.[15] Antidepressants that work in humans also work in these fruit flies.[16]

Next to anhedonia, learned helplessness is also associated with depression in humans and can be a marker for mood states in insects. Researchers compared fruit flies that were exposed to shaking in an inescapable maze versus flies that were shaken in an escapable maze. Both groups of flies were shaken equally hard. But when these flies entered another, escapable box several hours later, the flies who were shaken in the inescapable maze took much more time to escape from the box.[17] They were less willing to search for an escape route, as if they accepted their bad situation. They learned to be helpless, as if they had pessimistic judgments about their efforts to escape. Also honeybees show evidence of learned helplessness.[18]

Next to depression, fruit flies also appear to have anxiety states, with the physiological and behavioral responses that are similar to anxiety-like states in mammals such as rats.[19] Anxiolytics that work in humans also work in these fruit flies.

Relief learning

Fruit flies can display relief learning. When the flies first experienced an electric shock and then smelled an odor after the shock, the flies learned that the odor predicted relief from the painful stimulus, so they started to like that odor (i.e. they approached it).[20]

Chronic pain

According to a recent study, fruit flies can experience chronic pain. When an injury damages a nerve in one leg of a fruit fly, the fly’s other legs had become more sensitive, even after the wound in the first leg healed.


Comparison of welfare

With the above evidence, we can estimate the likelihood that insects are sentient. So we can say that our confidence level for bees being sentient is say 60%. But even if we take the precautionary principle and assume they are sentient, we have to compare the welfare of a bee with the welfare of other animals such as humans. Interpersonal comparison of well-being is very difficult (I have made some attempts elsewhere).

Welfare has several dimensions, but when it comes to painful and pleasurable experiences, intensity and duration are two important dimensions (other dimensions are the certainty and the order of experiences, for example choosing between first pleasure and then pain or vice versa; see the felicific calculus). How can intensity and duration of for example pain be compared?

Concerning the intensity of a feeling, there are two extremal options. On the one side, we can equate minimum levels of perception (or minimum differences in utility) between individuals. There are indications that sense perception is discrete. For example a change of a subjective experience always requires a minimum (not infinitesimally small) amount of a change of an external stimulus. This is the just-noticeable difference or JND. Therefore, a painful sensation can be decomposed as a sum of just-noticeable differences in pain. The number of JNDs required to go from zero pain to the actual feeling of pain, can be considered as a measure of the level or intensity of the pain. More generally, the utility function can be represented as a multidimensional staircase with discrete steps in several directions. Moving from one situation (e.g. a reference point without pain and hunger) to another situation (e.g. with a certain level of pain and hunger), requires a number of steps of the utility function, and this number measures the overall preference of the new situation compared to the old situation.

Assume, as an example, that a bee is a sentient being and experiences burning pain from hot water. How bad are the burns for the bee? We can count the just-noticeable differences of pain from the burns. Going from zero burns to a just-noticeable burn decreases the utility of the bee with one unit. An extra burning sensation decreases the utility with another unit, and so on. Suppose the bee experiences 100 negative utility units, which means 100 just-noticeable differences of pain are exceeded. Now I want to compare this with my painful experience of hot water. Perhaps, hypothetically, when I put a fingertip (the size of a bee) in hot water, I also experience 100 negative utility units, as much as when the whole bee is submerged in hot water. So what the bee experiences is what I would experience when I burn my fingertip, and is much less painful than what I experience if I’m completely submerged in burning hot water. The underlying reason is the smaller size of the brain of a bee: a bee has fewer pain neurons and a smaller brain processing capacity to produce a feeling of pain. That means a bee would be less sensitive: relatively more external stimulus is required in order to generate a JND. If one JND of a bee is comparable to one JND for me, and if I know the number of JNDs of a bee, I can imagine how painful an experience is for a bee.

However, the second extremal option equates the maximum levels of experience between individuals. The brains of a sentient being are finite in size, so it is unlikely that they can generate infinite levels of pain and pleasure. If there exist a maximum level of pain, we can for example consider a bee who experiences maximum pain from burning in hot water. This can be equated to my maximum pain level when I feel burning hot water all over my body. If we take this comparison, a bee would be highly sensitive, like a human, and there will be huge amounts of insect suffering in the world.

To make it more quantitative: suppose there are 1018 insects and 1010 humans. Suppose an average insect has a brain with 105 to 106 neurons (bees have relatively large brains for insects, with roughly 1 million (106) neurons). A human has roughly 1011 neurons. Suppose that the number of JNDs of an experience with a given perceptual input is proportional to brain size, measured as the number of neurons (other options for brain size are the number of neuronal connections or synapses). Hence, a human is more than 105 times as sensitive as an insect. Suppose all humans and insects experience a similar situation, such as dying (from disease, injuries, coldness, starvation or predation). According to the first approach, equating the minimum levels of utility, this experience generates more than 105 times as much discomfort in a human than in an insect. So we have to discount the experiences of insects with a factor 105 or 106. As there are 108 more insects than humans, and their suffering counts 105 to 106 less, total insect suffering from dying is 100 to 1000 times higher than total human suffering from dying. However, if we take the second approach, equating the maximum levels of utility, insect experiences are not discounted and total insect suffering is 108 times higher than human suffering from dying. Furthermore, the rate of dying of insects is much higher, because their lifespans are much shorter (ranging from a few days to a few years). If an average insect lives for a few weeks, it’s mortality rate is 1000 times higher than the human mortality rate, which means insect suffering from dying can be 106 or 1011 times higher than human suffering.

Matters could be even worse for insect suffering when we account for the subjective duration of an experience. Insects such as flies have faster brain processing speeds. Consider vision: humans can see at most 60 flashes of light per second. Showing flashes at a higher frequency results in seeing a continuous light. The flicker fusion rate measures how fast a light has to be switched on and off before one sees it as a continuous light. A fly has a flicker fusion rate four times higher than a human, which means a fly can see 250 images or flashes per second. This explains why it is so difficult to swat a fly: a fly sees everything in slow motion, four times slower than we do.

Perhaps not only vision, but also conscious experiences have a maximum frequency. What is the smallest time interval that we can experience? Suppose an experience of pain is turned on and off. Suppose at this moment you do not feel pain, a second later you feel pain, another second later the pain is gone. That means every second you can have a different conscious experience. But what if we increase the frequency? At this moment you do not feel pain, a millisecond later there is a pinprick. Another millisecond later the needle is removed, and so on. Now you might feel a slight, continuous pain instead of different pain pulses, which means you cannot consciously distinguish milliseconds.

Suppose the flicker fusion rate of your consciousness is 60 experiences per second, as with vision. This is as if you have an internal clock that has a moving hand rotating full circle in 60 steps per second. Every position of the moving hand corresponds with a different conscious state. You can have at most 60 different conscious experiences per second. But some insects may have faster internal clocks. In one real second, they can have 250 different conscious experiences. If you experience pain for one second, you actually have 60 conscious states of pain. But if insects can feel pain and if they feel pain for one second at a higher brain speed, that corresponds with 250 conscious states of pain. It is as if you would experience 4 seconds of pain.

Perhaps the tiny brains of insects indicate that the intensity of their pain experience is lower than the intensity of pain experienced by animals with larger brains. But if their brains are faster, they experience pain in slow motion, meaning that a second of pain appears to last longer for insects. That means one second of pain for a human should be discounted compared to one second for an insect: one second of insect pain counts a few times more than one second of human pain.

One further complication is our sensitivity for time intervals. We have a decreasing marginal sensitivity for time: the longer the time interval, the less important an extra second becomes. The difference in the preference for 0 seconds of extreme pain above 1 seconds might be bigger than the difference in the preferences for 1000 versus 1001 seconds of extreme pain. In the latter case, the one extra second is less important (you probably won’t notice the extra second). The rate of decreasing marginal sensitivity for time intervals might also depend on brain complexity and size. The bigger the brain, the more memory capacity it has and the more previous seconds are remembered and taken into account. This means that insects with smaller brains could have an almost constant marginal sensitivity for time: no matter how much time they experienced pain, an extra second of pain remains equally bad for them. On the other hand, it is possible that insects are not capable to perceive and judge long time intervals, which means they don’t have a preference between 100 seconds of pain and 1000 seconds of pain.

Avoiding insect suffering: prioritization

Humans are harming insects, by accidentally killing them (when running around, driving cars,…), intentionally killing them (using insects for food and clothing, using insecticides, insect traps, …) or indirectly killing them (by competing for resources, natural habitat destruction, pollution…). So one could argue we need less cars, less agriculture, less pollution, less concrete, less insecticides, less walking on the grass,… But things are not so simple when it comes to wild animal suffering.

First, a lot of insects are parasites or predators that kill other insects. So if we (accidentally, intentionally or indirectly) kill some insects, especially predators, we might save the lives of many other insects. Or stated differently: saving one ladybird might mean killing hundreds of aphids. Second, insects in the wild can have net-negative lives, i.e. short lives with more negative than positive experiences. These are lives not worth living. This is due to their reproductive strategy: a fertile adult insect can lay thousands of eggs. If the insect population does not explode at an extreme exponential rate, it is logically required that almost all of the newborn insects will have to die prematurely. The ways of dying are often extremely negative experiences: coldness, starvation, predation, parasitism,…. If an insect is killed, it prevents the birth of many insects with net-negative lives. So, if most insects face very short lives anyway and die horrible deaths anyway, it is far from clear whether killing insects increases overall future insect suffering. We need much more scientific research to estimate the overall effect of killing insects on global welfare.

Prioritization research involves looking for the most effective methods and interventions to improve insect welfare and reduce insect suffering. For example if we come to the conclusion that we should minimize killing insects, we can look for effective means to reduce the killing of insects. One interesting opportunity is the use of non-lethal methods for insect pest control. There are also many methods to avoid and remove insects in your house, for example catching flies with a transparent glass and releasing them outside.

If killing insects is inevitable, we should look for humane killing methods. Switching to more humane insecticides in agriculture could be a very effective way to minimize suffering. Using natural predators to combat insect pests on the other hand might be as bad as using inhumane insecticides.

In our prioritization research we should not only consider harm to insects caused by humans. Harm caused by nature (e.g. by other animals) counts equally. We can investigate safe and effective methods to intervene in wild nature to improve insect welfare. For example new technologies such as gene editing and gene drives could help reduce insect suffering by controlling insect populations in order to limit predation, parasitism, starvation and other causes of suffering.

Avoiding wild insect suffering: bee protection

Our prioritization research should not only consider different methods or interventions, but also consider which populations or species to target. Probably the clearest case can be made for bee protection. One three levels, bees are special.

First, as we have seen above, of all the studied insects, bees (the clade of antophyla) show probably the most scientific evidence for having consciousness and sentience. If they are not conscious, then other insects are probably not conscious either.

Second, unlike most other insects, bees have no negative externalities. A lot of insects are predators or parasites, which means they harm other animals. Herbivorous insects do not kill others, but they can be a pest in agriculture. Beas are the best: they do not harm other animals (except in self-defense) and they do not destroy food (e.g. crops) for other animals.

Third, bees have huge positive externalities: they improve crop yields by pollination. So bees even help to provide food for other animals (humans, birds) who like to eat fruits.

The bad news is: bees face difficult times. The colony collapse disorder kills many bees. Some neonicotinoid insecticides can make the bees more vulnerable to diseases and parasites, so an effective intervention is to replace those insecticides. However, those insecticides are not the only culprit of the colony collapse disorder. Pests such as pathogens and parasites are the biggest threat. Effective solutions for these threats are not yet known, so research is important.

Other ways to help bees, especially in western Europe, is to combat the Asian hornet. This is an invasive exotic species, and European bees do not recognize this insect as a predator. As a consequence, Asian hornets kill many bees (as well as many other insects). Eradicating a nest of Asian hornets could save many bees and other insects. This measure also finds support amongst environmentalists who are against invasive exotic species.

We also have to be careful not to take ineffective or counterproductive measures to protect bees and other insects. One example of an ineffective measure is the ban on GMOs in Europe. Bt-crops are GMOs that produce an insecticide (Bt) that is normally found in soil bacteria. According to a meta-analysis, Bt-crops are not harmful for honeybees.[21] Another meta-analysis shows that fields with Bt-crops have higher biodiversity levels of nontarget invertebrates (beetles, butterflies, spiders,…) compared to non-GMO fields where Bt-insecticide is sprayed (including organic fields, because Bt-insecticides are allowed in organic farming).[22] Spraying of Bt-insecticides not only kills the pest insects but also many other nontarget insects. Bt-crops can reduce the spraying of Bt-insecticides, and hence reduce the overall killing of insects. Also organic farming could be a counterproductive measure for bee protection: organic farming allows the use of insecticides that are harmful for bees, and the lower crop yields in organic farming means that more agricultural land is required. Hence, land that could serve as flower meadow for bees is sacrificed. We have to be careful however: as mentioned above, reducing insect killing or increasing natural habitat might increase insect suffering. So we first need more scientific research to estimate the overall effects of interventions.

The same goes for the most obvious animal rights issue related to bees: the consumption of honey. The production of honey is in many ways harmful to bees: they are often killed accidentally and in many cases also intentionally by the bee keepers (culling less productive hives, making the bees vulnerable to diseases by taking away their nutritious honey, clipping the queen bees’ wings,…). However, boycotting honey most likely means a replacement by other sweeteners, and most of those sweeteners come from agruculture that harms wild insects. Sugar involves the accidental and intentional killing of insects (using insecticides, machines,…). Perhaps very sweet stevia or some artificial sweeteners such as aspartame are better for the insects because they involve less agriculture. But a cookie or breakfast cereals with a strong sweetener contains less volume of sugar and hence more weight in grains and fats, which means more agriculture. On the other hand, more agriculture means less natural habitat and hence less suffering of insects in wild nature. The situation is very complex; we really don’t know the overall effects of honey. The situation is comparable to fishing: there is direct harm to used animal, the captured fish, but all the indirect effects on aggregate welfare of aquatic animals are unknown. As a rule of thumb one could use a provisional deontological principle and abstain from the consumption of fish and honey, avoiding direct harm associated with the use of animals. The idea is: if the presence of the body (of the bee or fish) is required to obtain your objective (consuming honey or fish), and if that animal is harmed (i.e. treated against his/her will), that harm counts more than all the unknown indirect effects. In the meantime, much more research about those indirect effects is required. We should not underestimate the value of information about indirect consequences of our choices. If the picture about indirect effects become clear, we might come to the conclusion that fishing or honey production are overall not negative for aggregate welfare. The known indirect effects gain more weight in the overall evaluation.


Avoiding cultured insect suffering: insects for food

Another important measure, is the decrease of insect farming. Insects are used for food and clothing (silk). An example is cochineal, a scale insect that produces a red dye carmine that is used as a colorant in food. It takes almost 100.000 insects to produce one kilogram of dye. Replacing cochineal dye with synthetic dyes is very feasible, because synthetic dyes are about four times less expensive.

Another worrying trend is the rise of insect meat consumption. Worms and crickets are used for insect burgers and sausages. The problem is that a lot of insects are required for insect meat. If one insect sausage requires more than 100 insects, whereas one beef sausage requires less than one thousandth of a cow, the number of animals used and killed for insect meat could be almost a million times higher than beef meat. Even discounting for brain size or pain sensitivity, insect meat can involve a lot of suffering.

Insect meat is often promoted as a more sustainable option than livestock meat. Chicken meat has the lowest environmental footprint of all livestock meats (lower than pork and beef), and the footprint of insect meat is about half that of chicken meat. However, the environmental impact (in terms of agricultural land use and greenhouse gas emissions) of insect meat is still 10%-50% higher than plant-based protein sources.[23] Eating insects is usually less efficient than eating plants. In other words, insect meat not only requires the intentional direct killing of insects for meat, but also more indirect and accidental killing of insects in agriculture to produce insect feed, compared to vegan alternatives. As insect meat is on the rise in Western countries but still far from being established, it is still possible to halt insect meat. Hence, campaigning against insect meat is very feasible.

[1] Bar-On, Y. M., Phillips, R., & Milo, R. (2018). The biomass distribution on Earth. Proceedings of the National Academy of Sciences, 115(25), 6506-6511.

[2] Kuldas, S., Ismail, H. N., Hashim, S., & Bakar, Z. A. (2013). Unconscious learning processes: Mental integration of verbal and pictorial instructional materials. SpringerPlus, 2(1), 105.

[3] Barron, A. B., & Klein, C. (2016). What insects can tell us about the origins of consciousness. Proceedings of the National Academy of Sciences, 113(18), 4900-4908.

[4] Loukola, O. J., Perry, C. J., Coscos, L., & Chittka, L. (2017). Bumblebees show cognitive flexibility by improving on an observed complex behavior. Science, 355(6327), 833-836.

[5] Paulk, A. C., Stacey, J. A., Pearson, T. W., Taylor, G. J., Moore, R. J., Srinivasan, M. V., & Van Swinderen, B. (2014). Selective attention in the honeybee optic lobes precedes behavioral choices. Proceedings of the National Academy of Sciences, 111(13), 5006-5011.

[6] Perry, C. J., & Barron, A. B. (2013). Honey bees selectively avoid difficult choices. Proceedings of the National Academy of Sciences, 110(47), 19155-19159.

[7] Czaczkes, T. J., & Heinze, J. (2015). Ants adjust their pheromone deposition to a changing environment and their probability of making errors. Proceedings of the Royal Society B: Biological Sciences, 282(1810), 20150679.

[8] Kaun, K. R., Devineni, A. V., & Heberlein, U. (2012). Drosophila melanogaster as a model to study drug addiction. Human genetics, 131(6), 959-975.

[9] Gibson, W. T., Gonzalez, C. R., Fernandez, C., Ramasamy, L., Tabachnik, T., Du, R. R., … & Anderson, D. J. (2015). Behavioral responses to a repetitive visual threat stimulus express a persistent state of defensive arousal in Drosophila. Current Biology, 25(11), 1401-1415.

[10] Abbott, K. R., & Dukas, R. (2009). Honeybees consider flower danger in their waggle dance. Animal Behaviour, 78(3), 633-635.

[11] Maze, I. S., Wright, G. A., & Mustard, J. A. (2006). Acute ethanol ingestion produces dose-dependent effects on motor behavior in the honey bee (Apis mellifera). Journal of insect physiology, 52(11-12), 1243-1253.

[12] Cheng, K., Peña, J., Porter, M. A., & Irwin, J. D. (2002). Self-control in honeybees. Psychonomic bulletin & review, 9(2), 259-263.

[13] Bateson, M., Desire, S., Gartside, S. E., & Wright, G. A. (2011). Agitated honeybees exhibit pessimistic cognitive biases. Current Biology, 21, 1070e1073.

[14] Breed, M. D. (1983). Correlations between aggressiveness and corpora allata volume, social isolation, age and dietary protein in worker honeybees. Insectes Sociaux, 30(4), 482-495.

[15] Abelaira, H. M., Reus, G. Z., & Quevedo, J. (2013). Animal models as tools to study the pathophysiology of depression. Brazilian Journal of Psychiatry, 35, S112-S120.

[16] Ries, A. S., Hermanns, T., Poeck, B., & Strauss, R. (2017). Serotonin modulates a depression-like state in Drosophila responsive to lithium treatment. Nature communications, 8, 15738.

[17] Brown, G. E., Mitchell, A. L., Peercy, A. M., & Robertson, C. L. (1996). Learned helplessness in Drosophila melanogaster?. Psychological reports, 78(3), 962-962.

[18] Dinges, C. W., Varnon, C. A., Cota, L. D., Slykerman, S., & Abramson, C. I. (2017). Studies of learned helplessness in honey bees (Apis mellifera ligustica). Journal of Experimental Psychology: Animal Learning and Cognition, 43(2), 147.

[19] Mohammad, F., Aryal, S., Ho, J., Stewart, J. C., Norman, N. A., Tan, T. L., … & Claridge-Chang, A. (2016). Ancient anxiety pathways influence Drosophila defense behaviors. Current Biology, 26(7), 981-986.

[20] Yarali, A., Niewalda, T., Chen, Y. C., Tanimoto, H., Duerrnagel, S., & Gerber, B. (2008). ‘Pain relief’learning in fruit flies. Animal Behaviour, 76(4), 1173-1185.

[21] Duan, J. J., Marvier, M., Huesing, J., Dively, G., & Huang, Z. Y. (2008). A meta-analysis of effects of Bt crops on honey bees (Hymenoptera: Apidae). PloS one, 3(1), e1415.

[22] Marvier, M., McCreedy, C., Regetz, J., & Kareiva, P. (2007). A meta-analysis of effects of Bt cotton and maize on nontarget invertebrates. Science, 316(5830), 1475-1477.

[23] Smetana, S., Mathys, A., Knoch, A., & Heinz, V. (2015). Meat alternatives: life cycle assessment of most known meat substitutes. The International Journal of Life Cycle Assessment, 20(9), 1254-1267.

Van Diepen J. e.a. (2018). Eiwit-transitie Vlaanderen. Studie naar de status en het potentieel van (hoog-) technologische oplossingen om vleeseiwitten te vervangen in het dagelijks dieet. Blonk Consultants. Gouda, Nederland.

Broekema R. & van Paassen M. (2017). Milieueffecten van vlees en vleesvervangers. Blonk Consultants. Gouda, Nederland.

Geplaatst in Artikels, Blog, English texts | Tags: , , , , , | Een reactie plaatsen