Jul 11, 2018 · 1h 6m · knowledge-project
#36 William MacAskill: The Science of Doing Good
William MacAskill · 50m spoken Shane Parrish · 8m spoken Advertiser (Indeed) · 55s spoken Advertiser (Accenture) · 53s spoken
0:00 / 0:00
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
William MacAskill joins Shane Parrish on The Knowledge Project to explain Effective Altruism, demonstrating how rigorous scientific methodology, ethical reasoning, and strategic career planning can maximize positive global impact and safeguard humanity's long-term future.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Shane holds 14.3% of the talking time here. How this is scored →
Shane as informed peer 2.4 Guest teaching 5.7 Guest disagreement 1.5 Shane pushing back 0.4
speaking balance: gold is Shane, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 9:00 Dismissing overhead ratio evaluation models
MacAskill forcefully rejects conventional charity evaluation metrics like Charity Navigator's overhead ratios, arguing that an efficient organization doing useless work is still completely bad.
Hardest push from Shane ▶ 14:50 Pushback on financialized career prioritizationShane challenges MacAskill's advice that graduates should pursue high-earning corporate careers merely to donate money, questioning whether that ignores qualitative vocational fit.
Biggest teaching moment ▶ 48:25 The Scared Straight counter-productivity evidenceMacAskill provides rigorous empirical data showing that the widely praised Scared Straight program actually caused a net increase in crime, resulting in $230 in social costs per dollar spent.
Shane holds their own ▶ 46:35 Introducing Harsanyi's Veil of IgnoranceShane demonstrates philosophical expertise by articulating John Harsanyi's Veil of Ignorance model to reframe nationalistic ethical biases.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Shane as informed peer | Guest teaching | Guest disagreement | Shane pushing back | Why |
|---|---|---|---|---|---|---|
| Effective Altruism and Injecting Science into Doing Good | 1 | 6 | 1 | 0 | Shane opens with broad exploratory questions asking MacAskill to explain effective altruism and his philosophical origins. MacAskill explains Peter Singer's shallow pond thought experiment and details how he founded Giving What We Can. | |
| Timing of Giving: Annual Donations Versus Compounding Wealth | 4 | 5 | 1 | 2 | Shane brings up Warren Buffett's philosophy of letting capital compound over 70 years before giving. MacAskill clarifies that compounding returns inside high-growth charitable causes often exceed standard 6-10% financial returns. | |
| Evaluating Impact: Critiquing Overhead Ratios and Arbitrary Metrics | 1 | 7 | 4 | 0 | MacAskill critiques traditional evaluators like Charity Navigator, rejecting their emphasis on overhead ratios as flawed because zero overhead in a useless charity still yields zero impact. | |
| Expanding Beyond Global Health: Existential Risk and Diverse Careers | 1 | 6 | 2 | 0 | MacAskill pushes back on the misconception that effective altruism is solely about randomized controlled trials in global development, explaining work in existential risk, AI safety, and policy. | |
| Strategic Philanthropic Careers: Earning to Give Versus Direct Action | 3 | 6 | 2 | 2 | Shane challenges MacAskill's 'earning to give' concept, asking if it is too financialized and ignores personal passions. MacAskill responds by reframing 'personal fit' around mastery and scalable talent bottlenecks. | |
| Overcoming Social Pressure in Career Planning | 2 | 5 | 1 | 1 | Shane and MacAskill discuss social pressure when choosing unconventional career paths. MacAskill advises students that dedicating time to choosing a high-impact career pays far greater marginal dividends than marginal study time. | |
| Limited-Life Foundations Versus Perpetual Endowments | 4 | 6 | 2 | 0 | Shane cites Chuck Feeney and Warren Buffett's limited-life foundation approaches. MacAskill expresses strong skepticism of perpetual foundations and university endowments, citing obsolete donor intents like the ScotsCare historical charity. | |
| Prioritizing Causes: Navigating Ethical Uncertainty and Scale | 2 | 7 | 1 | 0 | Shane asks how to compare disparate charitable interventions. MacAskill breaks down expected utility under ethical uncertainty, showing that differences in cost-effectiveness span orders of magnitude rather than minor percentages. | |
| Metrics of Impact: Quality-Adjusted Life Years and Income Doubling | 2 | 7 | 0 | 0 | MacAskill outlines quantitative metrics including Quality-Adjusted Life Years (QALYs) and logarithmic income doubling, before discussing Open Philanthropy's funding of biosecurity and AI safety risks. | |
| The Boundaries of Philanthropy, Markets, and Governance | 3 | 5 | 1 | 0 | Shane asks about market and governmental boundaries versus philanthropy. MacAskill clarifies that philanthropy acts as a lender of last resort for structural market and democratic failures. | |
| Taxes as Civic Reciprocity Versus True Philanthropy | 2 | 7 | 3 | 0 | MacAskill dismantles the common argument that paying taxes equates to charity, explaining taxes as reciprocal fee-for-service infrastructure prerequisites that make income generation possible. | |
| Mid-Roll Sponsorship: Enterprise Ad Operations and Hiring Solutions | 2 | 6 | 1 | 1 | Shane asks what policy changes MacAskill would make in government. MacAskill advocates benchmarking foreign aid programs against direct unconditional cash transfers, which Shane points out is politically difficult. | |
| Challenging Nationalism: The Veil of Ignorance and Global Equity | 5 | 4 | 3 | 0 | MacAskill criticizes 'charity begins at home' nationalism using a Beverly Hills reductio ad absurdum. Shane reinforces the point by introducing John Harsanyi's Veil of Ignorance framework. | |
| The Perils of Well-Intentioned Aid: The Scared Straight Case Study | 2 | 7 | 2 | 0 | Shane asks about well-intentioned aid backfiring. MacAskill walks through the 'Scared Straight' intervention, detailing how randomized controlled trials showed it increased recidivism and caused $230 in social costs per dollar spent. | |
| Rapid Academic Success: Action-Oriented Research in Moral Philosophy | 1 | 5 | 1 | 0 | Shane asks MacAskill about becoming an Oxford tenured associate professor at age 28. MacAskill attributes this to choosing neglected, action-oriented philosophical research topics where little literature existed. | |
| Overcoming Overconfidence: Calibration, Humility, and Better Probabilities | 3 | 6 | 1 | 0 | Shane asks how MacAskill applies uncertainty to general decision-making. MacAskill explains probabilistic overconfidence, calibration training, and adopting epistemic humility when encountering equally smart peers. | |
| Foundational Values: Maximizing Well-Being and Minimizing Suffering | 3 | 4 | 1 | 1 | Shane asks about radical honesty. MacAskill distinguishes between total truthfulness and radical candor, arguing that blunt blurting ignores the symbolic pragmatic meaning of language in human social contexts. | |
| Redefining Success: Impact Over Financial Wealth and Status | 3 | 4 | 0 | 0 | Shane and MacAskill discuss the meaning of success, contrasting superficial careerism and post-retirement isolation with impact on long-term human flourishing. | |
| The Most Common Mistake: Reinventing the Wheel Instead of Researching | 2 | 5 | 1 | 0 | MacAskill names reinventing the wheel as the most common human error, explaining how he consulted past winners to land nine grad scholarships, before concluding on protecting humanity's future. |
Statements from this episode (34)
MacAskill: Charitable impact varies vastly even among programs that work
“You can have the absolute best of intentions. You can be really meaning to do good. But then firstly, you can often just fail to have any sort of impact at all. But then secondly, even within those programs that work well, there's often just a vast discrepancy…”
MacAskill: Saving a child's life with bed nets costs under $10,000
“For example, by buying bed nets and distributing them to those who need them just for the, you know, single digit thousands of dollars, statistically speaking, you can save a child's life.”
MacAskill pledges over half of his lifetime earnings to charity
“I thought about how much money I needed to live on over the course of my life, and I decided, even as an academic, where I'm not going to earn as much as perhaps I could have done, I'm still going to earn much more than I need to be happy, and so what I did wa…”
MacAskill: Donating to nascent causes compounds faster than market returns
“Because often money that you donate compounds in just the same way that you return on investment if you save the money can compound. In particular, if you're investing in a small growing area often the rate of growth of that cause area can be far greater than …”
MacAskill: The best charitable giving opportunities are decreasing over time
“And also just because the world's getting richer, there's more and more smart money in philanthropy. And in general, the best giving opportunities are going down over time.”
MacAskill: Overhead ratio is a poor metric for charity effectiveness
“The overheads ratio. The amount that a charity spends on administration costs rather than program costs. And we, for a long time, have said that that's just a really poor metric if you want to actually try and do as much good as possible. Because imagine if yo…”
MacAskill: Effective Altruism cannot rely solely on RCTs and quantification
“And in those areas, you just often don't have randomized control trials. Often you need to use, you know, less quantitative evidence, but that's okay. That doesn't rule that out. Not everything can be quantified.”
MacAskill: Effective Altruism strongly promotes policy and research careers
“Money's only one, but we're very strongly encouraging people to pursue careers in policy and research and working for non-profits too.”
MacAskill: Charities survive without impact because donors are disconnected from beneficiaries
“In the case of the for-profit world, there's a kind of natural evolutionary mechanism. The companies that aren't able to turn a profit will eventually die. You know, the bad companies will get kind of kicked, you know, will get killed effectively, but that doe…”
MacAskill: The top charities achieve up to 1000x more positive impact
“We have some charities that can do, you know, hundreds of thousands of times more good than others which is an amazing opportunity if you are trying to do the most good but at the same time is indicative of the fact that it's just very hard to tell between the…”
MacAskill: High-earning corporate careers often generate more impact than non-profit work
“So often you can actually have more of an impact by going into A higher earning career and donating a very significant proportion of your earnings than you can by simply working at a nonprofit.”
MacAskill: Some recent graduates donate roughly $1M annually through high-earning careers
“And I know a number of people who've chosen to do this, and actually just some of whom are just a few years out of college and already are able to donate on the order of a million dollars per year.”
MacAskill: Many cause areas are talent-constrained rather than funding-constrained
“And in fact, for many cause areas, I think the More constrained by having really good people, really talented people than by having just, you know, additional money on the margin.”
MacAskill: Passion stems from mastering a subject rather than preceding it
“Passion tends to result from mastery of a subject or mastery of an area rather than vice versa.”
MacAskill: Most people's passions lie in hyper-competitive fields like arts and sports
“When you look at what people are actually passionate about, the vast majority tend to be passionate about things where it's extremely hard to have a promising career. So that's the arts, that's music, that's sport, and that's not a coincidence.”
MacAskill: Strongly favors limited-life foundations over perpetual endowments
“I'm extremely pro limited life foundations and very skeptical of perpetual foundations where within that I'd include universities as well.”
MacAskill: Harvard and Stanford should spend their endowments rather than fundraising
“It kind of baffles me that organizations, that, yeah, organizations like Harvard and Stanford are requesting donations from their alumni while at the same time sitting on an endowment of thirty-five billion dollars that they could be spending down.”
MacAskill: Proportional Income Increases Deliver Equal Welfare Value Across Income Levels
“So in general if you can increase someone's income by the same percentage, that's worth about the same, no matter how much someone's income is.”
MacAskill: Open Philanthropy is the model foundation for strategic giving
“And I just think this is the model foundation. Again, extremely transparent in terms of how they're choosing causes. Being extremely strategic in what cause areas they choose to focus on.”
MacAskill: AI creates severe risk when optimizing improperly specified goals
“There's significant risks that the, you develop artificial intelligences where you haven't specified their goals properly, and they don't do what you want. You give them more power, and that results in very bad outcomes because they optimize for what they've b…”
MacAskill: Most global problems should not be addressed through philanthropy
“In a sense, I think that most problems in the world shouldn't be addressed with philanthropy.”
MacAskill: Philanthropy is a last line of defense for market and democratic failures
“Philanthropy is really that, like, last line of defense, as it were, where you've both got a kind of market failing and democratic failing, and I think that applies for people in very poor countries who, you know, they don't get a vote over who's U.S. Presiden…”
MacAskill: Taxes Differ from Charity by Funding Income-Enabling Public Goods
“The taxes that you're paying is like contributing to a well-functioning society. That is precisely what allows you to make that money in the first place. And therefore, I think that's kind of very different from charity.”
MacAskill: Foreign Aid Counts as Charity, Domestic Tax Spending Does Not
“Especially from countries like Sweden and the UK, the money that they send on overseas development aid. You know, sometimes that's politically motivated, it's political gains, and different countries vary in how how much that's due with their foreign aid spend…”
MacAskill: Global health aid massively outperforms economic development aid
“So if you look at the past 50 years of aid spending, the track record of spending on economic development is really very mixed. You know, sometimes it works, sometimes it doesn't. The track record of spending on global health is just outstanding.”
MacAskill: Direct Cash Transfers Are Extremely Effective and Often Beat Traditional Aid
“The huge amount of spending goes on, and we don't even know whether that spending, such as buying livestock or providing wells or providing textbooks, is actually having more of an impact than simply giving poor people cash, which is the simplest thing to do. …”
MacAskill: Birth country and parental wealth dictate 80% of personal income
“80% of the income that people earn is explained by the country of birth and socioeconomic status of their parents”
MacAskill: The poorest people in the US or Canada are in the top 10-15% globally
“Even the very poor in the United States, for example, or Canada, They're still in the richest 10 to 15% of the world's population”
MacAskill: Typical rich-country earners make ~100x the poorest 700M people globally
“People on a typical income in, you know, a rich country, they're earning something like a hundred times the amount that the poorest seven hundred million people in the world have.”
MacAskill: The 'Scared Straight' juvenile intervention program actively increased crime rates
“And what they discovered was not only was the program ineffective, it was actively harmful. So the juveniles who went through this program would go on to commit more crimes, you know, that's more thefts, more rapes, more murders if they had gone through this p…”
MacAskill: Equal Disagreement With Smart Peers Demands Ambivalence
“If, I believe something, and I meet someone who I think is just as smart and just as well reasoned in this case as me, and they believe the opposite. Well, then I just become ambivalent, because this person's as likely to be right as I am.”
MacAskill: Personal Operating Rule Is Trying Never to Tell a Lie
“In general, I, Have a principle of just trying never to lie. I know that's different, however, from radical honesty”