Alex - thanks for the helpful summary of this exciting new book.
It looks like a useful required textbook for my 'Psychology of Effective Altruism' course (syllabus here), next time I teach it!
Alex - thanks for the helpful summary of this exciting new book.
It looks like a useful required textbook for my 'Psychology of Effective Altruism' course (syllabus here), next time I teach it!
Thanks a lot to you (and to Claude) for this!
I hadn't realized that context windows are now big enough to feed entire chapters.
I got excited by the idea of giving the book out at the EA table at my university club fair, but it looks like the options are $89 from Oxford University Press, a pdf, or an audiobook. That's not going to work.
Why not print the pdf?
Book by STEFAN SCHUBERT & LUCIUS CAVIOLA
Earlier this year, @Stefan_Schubert & @Lucius Caviola published Effective Altruism and the Human Mind: The Clash Between Impact and Intuition the first book-length examination of the psychology of effective altruism. This isn't a review but it's a great book and I expect it to become an indispensable (oft-cited) resource for practitioners in the EA ecosystem.
Assuming that other busy altruists might not have time to read the full 65,000 word version, I created a 5,000 word summary with the help of Claude Sonnet 3.5 that aims to distill the most important ideas and takeaways. Having compared it to the original text and my own notes I think it’s a good summary and could be very useful for some. It’s not perfect—but as the book itself reminds us in Chapter 9—a focus on perfection can be counterproductive!
The full book is available for purchase at Oxford University Press, as a free PDF, and as an audiobook thanks to @Aaron Bergman here: Apple Podcasts, Spotify, RSS.
We owe Schubert & Caviola a great deal for writing such a useful book and owe Caviola even more for doing so much of the supporting research. It all sheds essential light onto an otherwise dark region of our collective understanding.
I am humbled by their contributions (and those documented in the book) and merely seek to multiply the impact of that great work by making it slightly more accessible. Enjoy!
PART I. OBSTACLES
PART II. INTERVENTIONS
The chapter begins by contrasting how people make decisions in different domains of life. For example, when choosing a restaurant, people typically rely on subjective preferences and personal taste. In contrast, when making investment decisions for retirement, people tend to defer to objective data and expert advice. The authors then pose the question: How do people approach decisions when trying to do good in the world?
Research by Jonathan Berman and colleagues is presented, showing that most people approach charitable giving more like choosing a restaurant than making an investment decision. They tend to base their choices on personal feelings and preferences rather than objective measures of effectiveness. This tendency persists even when people are explicitly told that one charity is more effective than another.
The chapter explores several reasons for this approach to charitable giving:
The authors argue that these factors contribute to a norm where emotional appeal takes precedence over effectiveness in charitable giving. This norm is reinforced by societal expectations - most people don't criticize others for prioritizing causes they care about over more effective alternatives.
The chapter then delves into the philosophical distinction between obligatory and supererogatory actions. Charitable giving is generally viewed as supererogatory - praiseworthy but not morally required. This perception contributes to the idea that donors are free to choose any cause they like, rather than being obligated to select the most effective options.
The authors present research showing that people's sense of moral obligation to help effectively is influenced by several factors:
The chapter also discusses philosophical arguments for obligations to donate effectively, including Peter Singer's famous "drowning child" argument and Theron Pummer's case for effectiveness given that one has decided to donate.
An important section of the chapter focuses on social incentives for helping effectively. The authors note that while society generally celebrates the act of giving, there's much less emphasis on the effectiveness of that giving. They explore several reasons for this:
The chapter concludes by discussing the concept of "aversion to waste." While people don't always prioritize effectiveness, they do care about avoiding obvious waste in charitable giving. This explains why many donors are concerned about overhead costs, even though these aren't necessarily good indicators of effectiveness.
The chapter begins by drawing a parallel between investing and charitable giving, pointing out that while investors spend considerable time searching for the most promising companies, the same level of diligence is often not applied to charitable donations. The authors then pose a crucial question: How large are the differences in effectiveness between various ways of doing good?
To answer this question, the authors conducted a series of surveys comparing laypeople's and experts' beliefs about charity effectiveness. They focused on charities addressing global poverty and defined effectiveness in terms of lives saved per given amount of money. The survey results revealed a stark contrast:
This huge discrepancy highlights a critical misconception among the general public about the potential impact of their charitable giving. The authors argue that this underestimation of effectiveness differences is a major reason why people don't prioritize finding the most effective charities or do more research to identify them.
The chapter then delves into the reasons behind these vast differences in effectiveness:
To illustrate this, the authors compare two interventions addressing blindness:
The authors introduce the concept of quality-adjusted life years (QALYs) as a metric for comparing health interventions. They explain that the most effective global health interventions can be thousands of times more cost-effective than the least effective ones. Moreover, they describe how the distribution of intervention effectiveness follows a "heavy-tailed" pattern, with a small number of interventions being orders of magnitude more effective than most others.
The chapter then explores why people underestimate these differences in effectiveness:
The authors discuss several studies demonstrating scope neglect, showing that people's willingness to pay to solve a problem doesn't increase proportionally with the problem's size. They explain this phenomenon as partly due to the difficulty of evaluating options in "separate evaluation" (considering options one at a time) versus "joint evaluation" (comparing options side by side).
The chapter also touches on the "identifiable victim effect," where people tend to feel more strongly about helping a single, identifiable person than a larger group of "statistical" victims. However, the authors note that more research is needed on this topic.
From a normative perspective, the authors argue that most reasonable moral theories agree that it's untenable to prioritize identifiable victims over larger numbers of statistical victims. They emphasize that every life saved is equally valuable, regardless of whether it's the first or the thousandth.
The chapter concludes by reflecting on the nature of scope neglect. While it leads to suboptimal allocation of resources, the authors suggest that it may stem from a positive trait: our strong empathy for individuals. They argue that we should retain this empathy but channel it more effectively to help as many people as possible.
The chapter begins by explaining that most people prefer to help beneficiaries who are close to them in three ways:
However, the authors argue it is often more effective to help more distant beneficiaries. This is because most opportunities to help nearby beneficiaries have already been taken, while opportunities to help distant ones are often neglected.
The chapter then explores each of these biases in depth:
Parochialism:
Presentism:
Speciesism:
The chapter then discusses some philosophical arguments and empirical research related to these biases:
The authors conclude that our altruistic feelings evolved for a very different environment than today's globalized world. While it can be psychologically challenging, expanding our circle of moral concern to include distant beneficiaries is often the most effective way to do good. They suggest this as an important area for further psychological research and moral development.
The chapter begins by emphasizing that effective altruism requires not only supporting the most effective causes but also choosing not to support less effective causes. This concept of prioritization is presented as one of the most challenging aspects of helping effectively.
The authors explain that while many people are positive about effectiveness in abstract terms, they often struggle with its practical implications. They illustrate this with an example from the Rwandan civil war, where triage decisions had to be made about which patients to treat. While such prioritization feels more acceptable in emergency situations, people are less inclined to apply the same logic to charitable giving in non-emergency contexts.
The chapter presents research showing that people often prefer to split their donations between charities of different effectiveness levels, rather than allocating all resources to the most effective option. This tendency persists even when people are explicitly informed about effectiveness differences.
The authors explore several psychological mechanisms behind this resistance to prioritization:
The chapter delves deeper into the concept of sacred values, drawing on research by Philip Tetlock. It explains that while people are comfortable making trade-offs between "secular" values (like consumer goods), they resist trading off "sacred" values (like human lives) against secular ones or even against each other.
The authors argue that this aversion to trade-offs involving sacred values, while understandable, can lead to suboptimal outcomes in altruistic efforts. They emphasize that prioritizing the most effective interventions isn't cold or uncaring, but rather the best way to express care for others on a larger scale.
The chapter then explores the concept of splitting donations in more detail. It presents several reasons why people might prefer to split their donations:
The authors argue that while splitting might feel intuitively fair, it often leads to less effective outcomes. They introduce the concept of "thinking on the margin" to explain why concentrating donations on the most effective charity is typically more impactful.
The chapter also addresses the "Kantian fallacy" - the idea that we shouldn't prioritize the most effective charities because if everyone did so, those charities would be overwhelmed. The authors explain why this reasoning is flawed and how effective donors can coordinate to avoid such issues.
Finally, the chapter discusses the relationship between prioritization and fairness. While many people view strict prioritization as unfair, the authors argue that true fairness involves helping as many individuals as possible, regardless of which groups they belong to.
The chapter begins by recapping various misconceptions about effective altruism discussed in previous chapters, such as underestimating differences in charity effectiveness and neglecting distant beneficiaries. It then focuses on more fundamental misconceptions about the concept of effectiveness itself.
The Overhead Myth: The chapter starts with what the authors call the "overhead myth" - the common belief that low overhead costs (expenses not directly related to programs) indicate high charity effectiveness. The authors argue that this belief is misguided for two main reasons:
The authors present research showing that many donors prioritize charities with low overhead even when explicitly told that a high-overhead charity is more effective. They explore several reasons for this focus on overhead:
Direct vs. Indirect Impact: The chapter then discusses the preference for direct over indirect forms of impact. Many people prefer interventions with a clear, direct path to impact, even when informed that indirect approaches might be more effective. The authors argue that this bias can lead to neglecting highly effective "meta-charities" that have impact through other charities, such as charity evaluators or fundraising organizations.
Comparability of Different Causes: Another key misconception addressed is the belief that we can't compare the impact of work on different causes (e.g., health vs. education). The authors introduce two metrics for making such comparisons:
They acknowledge the challenges in measuring and comparing impact across causes but argue that it's both possible and necessary for effective altruism.
Aversion to Risky Giving: The chapter explores people's tendency to prefer interventions with certain, but smaller, impact over those with uncertain but potentially much larger impact. The authors introduce the concept of expected value and argue that in altruistic contexts, we should typically be risk-neutral rather than risk-averse.
The Growing Effectiveness Gap: The final section of the chapter discusses why misconceptions about effectiveness in altruistic contexts persist, while similar misconceptions are less common in areas like investment and consumption. The authors argue that this "effectiveness gap" has likely grown over time due to several factors:
They suggest that addressing these issues could help close the effectiveness gap over time.
The chapter begins by introducing a relatively modest strategy for overcoming psychological obstacles to effective altruism: working with people's existing values and preferences rather than trying to change them fundamentally. The authors focus on three types of techniques: providing information, nudging, and incentives.
Providing Information: The authors discuss several studies they conducted to test the effectiveness of providing information about charity effectiveness:
The authors note that while providing information can help, it doesn't make everyone help in the most effective way, as incorrect beliefs are not the only obstacle to effective altruism.
Real-World Examples: The chapter discusses real-world organizations that provide information about charity effectiveness, such as GiveWell, Animal Charity Evaluators, and Founders Pledge. It also mentions 80,000 Hours, which provides career advice for those seeking to have a high impact.
The authors emphasize that these recommendations work insofar as people trust the rating organizations and are willing to defer to them. They note that people are often overly inclined to trust their own judgments over expert recommendations.
Nudging: The chapter then explores nudging as a strategy to alter people's behavior without changing their preferences. The authors focus on a specific nudging technique called "donation bundling."
Donation Bundling: This technique involves giving donors the option to split their donations between their favorite charity and a highly effective charity. The authors present research showing that this approach can significantly increase donations to effective charities. They explain that bundling works by:
Incentivizing Effectiveness Through Donation Matching: The authors introduce a technique that combines bundling with donation matching. This approach offers increasing matching rates for donations that allocate larger proportions to effective charities. The research shows that this technique can substantially increase donations to effective charities.
Giving Multiplier: The chapter describes Giving Multiplier, a real-world donation platform that implements these techniques. The platform allows donors to split their donations between their favorite charity and an effective charity, with higher matching rates for larger allocations to effective charities.
Donor Coordination Through Micro-Matching: The authors explain how Giving Multiplier uses a "micro-matching" system where some donors provide matching funds to incentivize others to give more effectively. This system allows for donor coordination and has made the platform financially self-sustaining.
The chapter begins by addressing the apparent contradiction between the formidable psychological obstacles to effective altruism and the growing number of people who identify as effective altruists. The authors explain that this can be reconciled by recognizing the substantial individual differences in people's inclinations toward effective altruism.
The chapter then discusses the potential strategy of targeting outreach efforts specifically at those who are more open to effective altruism. The authors argue that understanding the psychology of these "effective altruism enthusiasts" is both practically and theoretically relevant.
Practical relevance:
Theoretical relevance:
The authors then present their research on identifying the psychological factors that predict endorsement of core effective altruist ideas. Through factor analysis of survey responses, they identified two core moral factors:
The chapter provides detailed explanations of each factor, including example survey items and correlations with various outcome measures related to effective altruism.
Expansive Altruism:
Effectiveness-Focus:
The authors note that these two factors, which broadly correspond to the "A" and "E" in EA, are psychologically distinct. This helps explain why relatively few people immediately find effective altruism wholly convincing.
The chapter also discusses a third factor, Truth-Seeking, which the authors argue is crucial for applying effective altruism in practice. They describe various measures of epistemic attitudes and virtues that might be relevant to effective altruism, such as actively open-minded thinking.
Finally, the authors suggest directions for future research, including:
The chapter begins by contrasting the strategies discussed in previous chapters, which work within people's existing values, with the more ambitious goal of changing society's fundamental values to align more closely with effective altruism. The authors acknowledge that this is a challenging project but argue that even partial success could make a significant difference.
The chapter then explores two main approaches to fundamental value change:
Reason-Based Moral Arguments: The authors discuss the role of philosophical reasoning in the early adoption of effective altruism, citing examples like Peter Singer's "drowning child" argument and Toby Ord's case for cost-effectiveness in global health. They then explore whether such arguments could persuade broader groups.
The chapter reviews several studies on the effectiveness of moral arguments:
The authors note that while some studies show positive effects, the results are often mixed, and the long-term impacts are unclear.
The chapter also compares reason-based and emotion-based arguments, referencing a study by Lindauer et al. that found no significant difference between the two approaches in increasing donations.
Real-World Studies: The authors discuss Schwitzgebel, Cokelet, and Singer's study on the effectiveness of ethics classes in changing student behavior. The results were mixed, with teaching about meat-eating ethics having an effect but teaching about charitable giving potentially having a backfire effect.
Limitations of Moral Arguments: The chapter acknowledges that moral arguments alone are unlikely to sway large portions of society to adopt effective altruism. The authors argue that if such arguments were highly effective, one would expect the effective altruism movement to have grown more substantially given the arguments made over the past decade.
Interaction of Arguments and Norms: The authors propose that while most people won't immediately accept effective altruist ideas, there's potential for longer-term growth through the interaction of arguments and changing social norms. They suggest that:
The chapter draws parallels with other social movements that have seen rapid norm changes, such as attitudes towards same-sex marriage and racial equality. The authors discuss the concept of "behavioral contagion" and how it could potentially apply to effective altruism.
Potential for Effective Altruism's Spread: The authors consider factors that might support or hinder the spread of effective altruism:
Supporting factors:
Potential hindrances:
The chapter begins by addressing the misconception that effective altruism requires extreme sacrifices, such as working constantly and giving away almost all resources. Instead, the authors argue for a more sustainable approach, aiming for "maximum sustainable goodness" rather than perfection.
The Two-Budget Strategy: The authors introduce the "two-budget strategy" as a way to balance effective altruism with personal needs and desires:
They emphasize that individuals must decide for themselves how much to allocate to each budget, considering factors like income, age, and dependents. The authors suggest making this decision explicitly, perhaps annually, to avoid constant internal negotiation.
Avoiding the Half-Measure Fallacy: The chapter explains the "half-measure fallacy," where people attempt to compromise between effectiveness and personal preferences by choosing the most effective charity within a preferred cause area. The authors demonstrate why this is suboptimal due to the vast differences in effectiveness between causes.
Effective Altruist Virtues: The authors discuss key virtues for practicing effective altruism:
They emphasize that these virtues complement, rather than replace, common-sense virtues like honesty and kindness.
Finding the Highest-Impact Causes: The chapter reviews the scale-tractability-neglectedness framework for evaluating causes and discusses three popular cause areas in effective altruism:
The authors also mention "meta causes" such as global priorities research and spreading effective altruist ideas.
Impact Strategies: The chapter discusses two broad classes of strategies for having an impact:
The authors emphasize the importance of cause-neutrality and means-neutrality when choosing strategies.
Engaging with Effective Altruist Ideas: The chapter suggests that deeply engaging with effective altruist ideas can be highly impactful, potentially more so than immediately taking action. It emphasizes the value of:
Future Directions: The authors conclude by suggesting areas for future research on the psychology of effective altruism, including:
First time I'm curating Claude- that I know of.
Thank you, Alex, for putting this together, it's a public service!