The five percent rule
Rees Calder · 17 April 2026 · 7 min read
Most people trying to be a force for good distribute their effort roughly evenly across dozens of small decisions. Paper or plastic. Organic or not. Round up at the till. Cycle on Thursdays.
The data suggests this is backwards.
Almost every domain where humans try to measure impact shows the same pattern. A handful of choices dominate the outcome, and the rest is noise. The Pareto principle (80/20) understates it. In some cases it's closer to 95/5.
This isn't a new idea. Economists have been charting it since Vilfredo Pareto noticed that 80% of land in Italy was owned by 20% of the population in 1906. What's newer is how much data we now have on where the fat tail actually sits for individual action.
If you only audit five decisions in your life, audit these.
1. Your career
Eighty thousand hours is roughly how many working hours a typical professional career contains. That number is also the name of a research org that has spent a decade trying to quantify career impact.
Their headline finding: the variance in "good done" across careers is enormous. A doctor in a high-income country saves a handful of QALYs over a career. A doctor working on neglected tropical diseases for the same organisation may save thousands. Same degree, different deployment, two orders of magnitude.
Research in effective philanthropy shows a similar pattern. The median charity is roughly one-hundredth as cost-effective as the top few, per GiveWell's cost-effectiveness tables. A career spent raising money for the median charity and one spent raising money for a top charity look identical from the outside and are completely different in impact.
The career choice isn't the only choice, but it's the largest single lever most people control. And most people make it once, in their early twenties, based on advice from people who don't know the data.
2. Where you live
Housing and commute are almost invisible in most personal-impact conversations, which is odd, because together they account for roughly 40 to 50% of an individual's lifetime carbon footprint in rich countries.
Living within 5 km of work cuts emissions more than any diet change you're likely to sustain. Living in a dense city cuts them further, because apartment buildings are wildly more efficient than detached houses per occupant. The International Energy Agency's 2023 report on urban emissions puts the delta at roughly 60% lower per-capita CO2 for dense urban living versus suburban.
The lever isn't "move tomorrow." It's "next time you choose, weight proximity and density harder than you did last time."
3. What you eat, in aggregate
Not every meal. The aggregate pattern over a year.
Poore and Nemecek's 2018 meta-analysis in Science, which looked at more than 38,000 farms in 119 countries, showed food accounts for roughly 26% of global greenhouse gas emissions. Within that, animal products account for over 60% of emissions despite providing roughly 18% of calories.
A person who eats beef twice a week has a food footprint roughly four times that of a person who eats it twice a month. Nobody needs to be vegan. Moving from daily meat to a few times a week is the biggest step. After that, returns diminish.
There's a wrinkle here that's worth flagging. The environmental case for plant shift is strong and well-evidenced. The health case is messier than the media often suggests. Read both before overcorrecting.
4. How much you give, and where
The median American household gave roughly $2,500 to charity in 2022, according to Giving USA. GiveWell estimates its top-recommended charities save a life for roughly $5,000 in expected value terms.
That means a median household can save roughly one life every two years by giving to the most cost-effective charities rather than the ones most familiar. Most people never hit the math, because giving isn't typically optimised. It's directed at whoever asked last, or at causes that feel close.
This isn't a guilt trip about volume. It's an observation about allocation. The first $500 to an effective charity likely does more than the next $5,000 to an average one.
5. Whether you have children, and how you raise them
The hardest one to discuss, and the one with the largest measurable carbon impact by a wide margin.
Wynes and Nicholas (2017) estimate an additional child in a developed country adds roughly 58 tonnes of CO2-equivalent per year, accounting for descendants. That dwarfs every other lifestyle choice.
But the framing of "don't have kids for the climate" is wrong and mostly unhelpful. Fertility is deeply personal, and kids are the domain where the whole point of doing good lives.
The useful frame: if you're having them, the values you instill and the leverage you give them probably matter more than the carbon they'll emit. A child raised with agency, evidence-based thinking, and reasonable wealth has the capacity to do vast amounts of good. That capacity is the investment.
The audit
Once a year, sit down for an hour and answer these five questions honestly.
- Is my work directed at the highest-leverage problem I'm qualified to work on?
- Does where I live support lower-carbon living, or fight it?
- Is my food pattern roughly where I want it, across a year?
- Is my giving directed by evidence or by habit?
- If I'm raising kids, am I giving them the tools to do good with the life they've been handed?
Most weeks you don't need to think about this. The annual check is enough. The rest of the year, get on with your life.
Sources used: Pareto, Cours d'Economie Politique (1906), 80,000 Hours career impact research (2020-2024), GiveWell cost-effectiveness analysis (2023), IEA Urban Energy Outlook (2023), Poore & Nemecek, Science (2018), Giving USA Annual Report (2022), Wynes & Nicholas, Environmental Research Letters (2017).