Applications for these positions have now closed.

We’re looking for candidates to join our 1on1 team.

The 1on1 team at 80,000 Hours talks to people who want to have a positive impact in their work and helps them find career paths tackling the world’s most pressing problems. We’re keen to expand our team by hiring people who can help with at least one (and hopefully more!) of the following responsibilities:

  • Advising: talking one-on-one to talented and altruistic applicants in order to help them find high-impact careers.
  • Running our headhunting product: working with hiring managers at the most effective organisations to help them find exceptional employees.
  • Improving our systems: building tech-based systems to support our team members.

If you think you’d be interested in taking on more than one of these duties, and enjoy wearing multiple hats in your job, we strongly encourage you to apply. The start dates of these roles are flexible, although we’re likely to prioritise candidates who can start sooner, all else equal.

These roles have starting salaries from £50,000 to £85,000 (depending on skills and experience) and are ideally London-based. We’re able to sponsor visa applications.

How to apply

To apply for a role as an Advisor, Headhunting Lead, or Systems Analyst, please fill out this application form by 11PM GMT on Sunday, October 8, 2023. We expect the application to take around an hour.

If you have any issues submitting the form, reach out to [email protected].

We’re aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all the suggested attributes, and would especially like to encourage people from under-represented backgrounds to apply, even if you don’t meet all the suggested criteria.

Click here to apply

About 80,000 Hours

Our mission is to get talented people working on the world’s most pressing problems by providing them with excellent support, advice, and resources on how to do so. We’re also one of the largest sources introducing people to the effective altruism community, which we helped found. Since 2011, we’ve had over 10 million visitors to our website (with over 100,000 hours of reading time per year) and thousands of people have told us they’ve significantly changed their career plans as the result of our work.

The 1on1 team at 80,000 Hours takes people from being “interested in the ideas and wanting to help” to “actually working to solve pressing world problems.” In 2022, we had more than 5,500 applications to speak with us from talented and altruistic people who weren’t sure how to use their skills to help the world. Some of the most important ways our conversations help people increase their impact are:

  • Introductions to experts in relevant fields, as well as to hiring managers
  • Talking through and reframing decisions in order to pin down people’s key uncertainties about their careers
  • Suggesting new ideas — whether that’s specific jobs, new paths to explore, or ways of getting funding

We directly help hiring managers at organisations tackling the world’s most pressing problems by using our unique network to find and recommend talented and altruistic candidates, often for their highest priority roles. If you’re interested in examples of people our 1on1 team has helped, you can find their stories on our website. Thanks for considering joining our team!

Open position: Advisor

The role

It’s a great sign you’d enjoy being an 80,000 Hours advisor if you’ve enjoyed managing, mentoring, or teaching. We’ve found that experience with coaching is not necessary — backgrounds in a range of fields like medicine, research, management consulting, and more have helped our advisors become strong candidates for the role.

For example, Anemone Franz joined us after working as a clinical trial physician, Abigail Novick-Hoskin completed her PhD in Psychology, and Matt Reardon was previously a corporate lawyer. But it’s also particularly useful for us to have a broad range of experience on the team, so we’re excited to hear from people with all kinds of backgrounds.

The core of this role is having one-on-one conversations with people to help them plan their careers, although we have a tight-knit, fast-paced team where people take on a variety of things. These include, for example, building networks and expertise in our priority paths, analysing data to improve our services, and writing posts for the 80,000 Hours website or the EA Forum.

We expect this role to be full-time and based out of our London office, although our remote work policy accommodates being away for up to three months of the year, if needed. If this won’t work for you, please let us know in your application.

What we’re looking for

We’re looking for someone who has:

  • A strong interest in effective altruism and longtermism
  • Strong analytical skills, and would enjoy puzzling out the key considerations for a complex problem
  • A deep interest in understanding people, and who would enjoy having large numbers of one-on-one conversations via video call
  • The ability to rapidly learn and follow developments in the talent needs of nascent and fast-changing fields, especially AI safety
  • Excellent professional communication and social skills

Previous experience in one of our priority areas would be a significant advantage in the role, but we encourage you to apply even without that.

What should candidates expect from the hiring process?

To apply, please complete the application form by 11PM GMT on Sunday, October 8, 2023.

We’re reviewing applications on a rolling basis, so we encourage you to apply as soon as you are able to.

The application process will vary depending on the candidate but will likely involve:

  • A short screening call
  • A ~2–4 hour work test
  • An interview
  • A 2–5 day in-person (if possible) work trial, if we think it’s 50% likely we’d offer you the role

If you’re not interested in the Advisor role, but would be interested to join the 1on1 team in some other capacity, we’re also looking to hire a Headhunting Lead and Systems Analyst.

Open Position: Headhunting Lead

The role

As Headhunting Lead, you would:

  • Develop and lead a headhunting process that makes use of the unique network the 1on1 team has built to help fill roles working on some of the world’s most pressing problems (particularly in AI safety and policy, biosecurity, and global priorities research)
  • Build strong working relationships with hiring managers and team leads at key organisations tackling these problems, as well as top candidates for these roles
  • Lead the 1on1 team’s internal data strategy
  • Coordinate with the 1on1 and Job Board teams to fill important openings and highlight these to promising candidates in our network

We expect headhunting positions to be full-time in-person roles based out of our London office, although our remote work policy accommodates being away for up to three months of the year, if needed. If this won’t work for you, please let us know in your application. You would be managed by Michelle Hutchinson, Director of 1on1.

What we’re looking for

We’d be most excited to hear from people who closely match the following:

  • Enjoys working with different people in a variety of contexts, including maintaining relationships with major stakeholders, hiring managers, and promising candidates for top roles
  • Thinks critically and communicates clearly about their plans and strategy, while prioritising what matters most
  • Is comfortable approaching thorny, open-ended questions from multiple perspectives, moving between big picture thinking and getting stuck into the details of system and process design as the problem demands
  • Has a strong understanding of 80,000 Hours’ focus areas, ideally including risks from advanced AI

In addition, we’d be especially excited to find someone who:

  • Has experience in project management, research, or strategy — this could include roles in consulting, product management, or at early-stage startups or nonprofits
  • Is fascinated by people and enjoys developing working models of individuals’ skills and traits to match them to specific roles
  • Has experience with any of the following:
    • Longtermist and or existential risk strategy, particularly around talent allocation
    • Working with CRM and/or database software, such as Salesforce or Airtable

What should candidates expect from the hiring process?

To apply, please complete the application form by 11PM GMT on Sunday, October 8, 2023.

We’re reviewing applications on a rolling basis, so we encourage you to apply as soon as you are able to.

The application process will vary depending on the candidate but will likely involve:

  • An interview (remote)
  • A ~2 hour work test (remote)
  • A 2–5 day trial (in-person if possible; for candidates with >50% chance of an offer)

If you’re not interested in the Headhunting Lead role, but would be interested to join the 1on1 team in some other capacity, we’re also looking to hire Advisors and a Systems Analyst.

Open Position: Systems Analyst

The role

As a Systems Analyst for the 1on1 team, your responsibilities could include:

  • Building systems to enable new ways of delivering value to users
  • Prioritising and implementing changes requested by staff
  • Crafting automations to streamline internal and external processes
  • Providing project management and operations assistance for 1on1 team priorities
  • Analysing data to answer strategically relevant questions
  • Hiring and managing contractors
  • Salesforce administration

We expect to tailor the scope of this role to fit the successful candidate’s interests and strengths. We’re open to hiring someone full-time or part-time to work exclusively on systems, or to hiring someone to work part-time on systems and part-time on either advising or headhunting. A full-time systems role would likely include more project management and operational assistance work (e.g. coordinating with external stakeholders who give feedback on our advising calls), and may also provide systems support to other teams. We’re considering both London-based (preferred) and remote candidates for this role.

What we’re looking for

We’re looking for someone who has:

  • An operations mindset — you’re good at identifying issues, places for improvement, prioritising, generating solutions, and efficiently implementing new ideas
  • Tech savviness and enthusiasm about experimenting with and learning new systems.
    • Programming experience may be helpful, but is not necessary

Experience in operations, IT, Salesforce administration, project management, or other relevant fields would be a bonus. However, a lack of experience shouldn’t discourage you from applying.

What should candidates expect from the hiring process?

To apply, please complete the application form by 11PM GMT on Sunday, October 8, 2023.

We’re reviewing applications on a rolling basis, so we encourage you to apply as soon as you are able to.

The application process may vary but will likely involve:

  • A short screening call
  • A ~1–3 hour work test
  • An interview
  • A 2–5 day in-person (if possible) work trial, if we think it’s 50% likely we’d offer you the role

If you’re not interested in the Systems Analyst role, but would be interested to join the 1on1 team in some other capacity, we’re also looking to hire Advisors and a Headhunting Lead.

Salaries

Salaries will vary based on your skills and experience, but to give a rough sense, a starting salary for a full-time position would range from ~£50,000-85,000 per year. For someone with five years of relevant experience, a starting salary would likely be in excess of £70,000 per year.

Benefits

Our benefits (prorated for part-time staff, where applicable) include:

  • The option to use 10% of your work time for self development
  • 25 days of paid holiday, plus bank holidays
  • Standard UK pension, with 3% contribution from employer
  • £5,000 mental health support allowance
  • Private medical insurance
  • Generous paid parental leave
  • Long-term disability insurance
  • Flexible work hours
  • Gym, shower facilities, and free food provided at our London office.

Our remote work policy also accommodates being away for up to three months of the year, if needed, and we’re able to sponsor visa applications and cover many moving expenses.

The application process

To apply for a role as an Advisor, Headhunting Lead, or Systems Analyst, please fill out this application form by 11PM GMT on Sunday, October 8, 2023.

Click here to apply

The application process will vary depending on the candidate and the roles they’re interested in, but is likely to include a short chat with our staff, a work test, an interview, and a multi-day in-person trial. We offer payment for work samples and trials where possible (that is, conditional on your location and right to work in the UK).

For any questions or difficulties submitting the application form, reach out to [email protected].

    This is Part 1 of an updated version of a classic three-part series of 80,000 Hours blog posts. You can also read updated versions of Part 2 and Part 3. You can still read the original version of the series published in 2012.

    Doctors have a reputation as do-gooders. So when I was a 17-year-old kid wanting to make a difference, it seemed like a natural career path. I wrote this on my medical school application:

    I want to study medicine because of a desire I have to help others, and so the chance of spending a career doing something worthwhile I can’t resist. Of course, Doctors [sic] don’t have a monopoly on altruism, but I believe the attributes I have lend themselves best to medicine, as opposed to all the other work I could do instead.

    They still let me in.

    When I show this to others in medicine, I get a mix of laughs and groans of recognition. Most of them wrote something similar. The impression I get from senior doctors who have to read this stuff is they see it a bit like a toddler zooming around on their new tricycle: a mostly endearing (if occasionally annoying) work in progress. Season them enough with the blood, sweat, and tears of clinical practice, and they’ll generally turn out as wiser, perhaps more cantankerous, but ultimately humane doctors.

    Yet more important than me being earnest — and even me being trite — was that I was wrong. Medicine was not my best option for helping others when compared to all the other work I could do instead. And I think that is not just true for me in particular, but for many able, altruistically minded people considering a medical career.

    This series of posts will explain why.

    • Part 1 covers the impact of medicine on human health and disease. Upshot: medicine as a whole, at least as practiced in most clinical settings, has been only a minor player in the dramatic improvements in human health over the last couple of centuries (relative to things like nutrition, safer jobs, and general scientific knowledge), so one should expect the impact of providing more of it now by working as a doctor to be fairly modest. Unlike what you see in the medical dramas where the protagonists are saving lives every episode, it is more like saving a couple of lives every year. This is better than the direct impact of most jobs and might still make a compelling case for pursuing this career path if not for other considerations covered in parts 2 and 3.
    • Part 2 takes a closer look at the impact of you working as a doctor in particular. Upshot: thanks to issues like diminishing marginal returns and replaceability, adding another doctor (you) to a place like the UK would be expected to have an even more modest impact than that of medical care in general. Instead of saving a couple of lives every year, it is more like saving a few lives per career.
    • Part 3 will look at possible ways medics can have an outsized impact, such as earning to give or working abroad in lower-income countries.

    Is medicine a determinant of health?

    On average, it’s clear that humans live much longer and healthier lives now than we did a century or two ago. Exactly why this has happened — and how much credit medicine can claim for us being healthier — is much more uncertain.

    UK1 life expectancy doubled from 40-ish in the 1800s to 80-ish now. The largest contributor was a reduction in infant and childhood mortality. For example, in the UK, around 15% of those born in 1900 died before their first birthday — now it is ~0.4%. But survival has improved in every age bracket.

    This trend is basically universal. There remains a lot of global inequality in life expectancy, and the upward trend has disruptions, most recently due to COVID-19. But wherever they are born, children today can expect to live (at least) twice as long than those born in the 1800s.

    One story explaining this trend attributes it to advances in medical care. Nineteenth century medicine was much more primitive than 21st century medicine: antibiotics, vaccines, chemotherapy, defibrillators, surgery, and (effective) drugs basically did not exist. So perhaps we live longer thanks to modern medicine fending off the scythe of death.

    Medicine Holding Back Death, a sculpture by Julian Hoke Harris

    Although this is part of the picture, it is probably a small one. The dramatic improvements in health over the last two centuries are attributed less to medical care and much more to the social determinants of health.

    In essence, living standards have improved, humankind got dramatically richer, and we’re better informed than we used to be. These key changes enabled us to do a bunch of things to prevent illness in the first place. Some examples (among many):

    • A population learning the basics of public health — like germ theory or “smoking kills” — means individuals can better avoid disease. If rich enough, they can also afford public works like sanitation systems, curbs on air pollution, and fortifying common foods with micronutrients.
    • Richer populations tend to have a lower proportion working in more dangerous industries. More people work in services like hospitality or finance, while fewer people work in agriculture and manufacturing. At the same time, dangerous occupations and transportation can be made safer.
    • Richer populations can feed their children enough, setting them up for healthier lives as adults. Height used to be an excellent indicator of wealth because poverty led to childhood malnutrition and stunted growth. Maternal and infant malnutrition remains among the biggest risk factors for ill health worldwide, but it is largely absent in high-income countries.2

    Evidence from a few different sources indicates that it is mostly these social determinants rather than medicine that do the heavy lifting for health and longevity:

    1. Mortality trends show that death rates often started decreasing prior to many major healthcare discoveries. The canonical example is tuberculosis: mortality fell ~30% before the bacterium was identified, ~80% before effective drug therapy, and ~90% before vaccination.3
    2. Good ‘per capita’ healthcare spending figures are trickier to find,4 but available evidence shows that the explosion in health spending happened after the skyrocketing of life expectancy. Here are some suggestive figures for the US:
      Life expectancyHealth spending per person (inflation adjusted)5
      189045Uncertain, but surely <$151
      196070$151
      202077$12,600

    3. There have been a couple of studies randomising people to receive more or less access to healthcare in the US. Although interpreting these studies is tricky, it is fair to say better healthcare access resulted in very minor health gains.5
    4. Observational data of intra-national inequality in health outcomes by occupational class, income, education (etc.) are also consistent with the theory that social determinants have driven health improvements.
    5. If you compare different countries by how healthy they are (whether by life expectancy, disability-adjusted or quality-adjusted life expectancy, or aggregate measures of ill health), the best predictors are measures of wealth (e.g. GDP per capita) or education (e.g. years of schooling). Controlling for these factors, measures of healthcare (e.g. doctors per capita, health spending per capita) have negligible impact (much more later).

    Ironically, the picture of the man holding off the scythe of death above may be — inadvertently — right after all. Although the man is meant to represent medicine (the frieze is on the side of a hospital), he is holding the wrong symbol. The rod of Asclepius, god of medicine, only has a single serpent and no wings on the top. Instead, the man is holding a caduceus (two serpents, wings), the staff of Hermes. Although the caduceus is widely used as a symbol of medicine, this practice originated in the US in the 20th century and is generally deemed an anachronistic confusion.

    But if you are going to fend off the scythe of death, maybe the caduceus is the better stick for the job. Hermes is not the god of medicine; he is the god of trade and commerce, and the caduceus has sometimes connoted wisdom and knowledge. So perhaps it is a better stand-in for what has doubled lifespan over the last two centuries — our growing wealth and wisdom — and Hermes has done better than Asclepius so far.

    Medical care (and medical careers) in context

    Although health is mostly a matter of social determinants versus medical interventions, the latter still matter. Some relevant nuances:

    • For some diseases, it is obvious that medical interventions are the decisive factor in reducing mortality. Consider type-1 (insulin dependent) diabetes: prior to insulin therapy, this early-onset disease was inevitably fatal within a couple of years. Today, those with type-1 diabetes still have ~10-year shorter life expectancy than those without, but essentially all the credit for changing this disease from ‘childhood death sentence’ to ‘reliably surviving to adulthood and often old age’ goes to medicine. There’s not much of a social determinant story to tell for improved outcomes from this disease — indeed, its incidence has been steadily increasing.
    • Although most (either by count or prevalence) conditions have mortality graphs that look like tuberculosis, some — such as polio and smallpox — do show dramatic declines following vaccine deployment.
    • Although cardiovascular disease and cancer (the main causes of death in richer countries) have a variety of lifestyle risk factors, medicine has made steady progress in both improving treatment and using drugs to reduce risk. So the balance of expert opinion tends to attribute a significant part of the steady incremental improvements in health in wealthy countries over the last 50 years to these medical interventions.6

    I think the overall picture is shown by the Global Burden of Disease project. Communicable, maternal, neonatal, and nutritional diseases are principally diseases of deprivation, where most of the battle is won through improving the social determinants of health. Winning this battle, as countries like the UK (mostly) did between ~1800–1950, and many others have done since, gets you from a life expectancy of ~30 to a life expectancy of ~70. Climbing from there is mostly a steady slog against non-communicable disease.

    In this regime, medicine has two important roles. First, the social determinants greatly reduce individual risk and the population burden of disease in youth and adulthood, but eliminate neither — medicine can help pick up the pieces for those who get unlucky. Even a hypothetical ideal public health population of affluent, well-educated, never-smoked, BMI 22, teetotaler triathletes will have some who suffer life-threatening accidents, others who develop rare conditions like type-1 diabetes or childhood cancers, and even some who have high blood pressure and cholesterol. Real populations like the UK fall well short of this, even if they are much closer to the ideal than they are to the prevalent poverty of the 1800s.

    In old age, everyone’s luck runs out eventually. The second role of medicine is fighting a rearguard action against the progressive bodily breakdown of ageing. Its successes here are less dramatic and more incremental. Medical science has few solutions to many of the diseases of old age. Treatment is complicated by comorbidity and multiple organ systems developing varying degrees of failure (e.g. life-saving surgery or chemotherapy is much more treacherous when the patient’s heart, kidneys, and liver are not what they once were).

    The steady development of multiple life-threatening diseases means even if a doctor can save their elderly patient’s life, they will likely succumb to another not too long thereafter; curing an 80-year-old’s cancer is unlikely to give them a new 30-year ‘lease on life’ like it might for a 20-year-old.7

    So the answer to “What has modern medicine ever done for us?” is, relatively speaking, not all that much. Although some of us will have our lives or limbs saved by medical care, for most of us the impact is smaller: potentially making some chronic diseases a little better and making our old age somewhat longer.8 So, perhaps joining the medical profession is worthwhile, but maybe not as great in terms of humanitarian impact as tropes would have you believe.

    An initial number for how much good doctors do

    I still haven’t answered the question of how many lives a doctor saves: ‘less than the industrial revolution’ doesn’t narrow it down much — everyone living a bit longer, and no one dying of smallpox, still matters a lot. And that medicine’s aggregate impact on the human condition is relatively modest versus other factors could still mean one’s contribution as a doctor is great in absolute terms.

    It is tricky to work out the precise fraction of health medicine can take credit for, but here’s one attempt from researchers John Bunker, Howard Frazier, and Frederick Mosteller. In essence, they look at the most important medical interventions (gathering enough to cover most of the benefits from medical care), get quantified estimates of their efficacy from trial data, and then estimate the aggregate benefit from how commonly these interventions are made.

    They estimate approximately five of the 30-year gain in life expectancy in their population was owed to medical intervention and another five years ‘relieved’ from poorer health.

    These estimates are rough,9 but we can use them to generate an even rougher back-of-the-envelope figure. We’ll first translate this estimate into something called disability-adjusted life years (DALYs: in essence, you ‘gain’ DALYs for dying soon or being sick, so fewer is better. So, how many DALYs does a doctor save?

    1. A five-year life expectancy increase is easy: five more years of life, so five DALYs averted.
    2. ‘Five years relieved from poorer health’ is trickier. How bad is life with different sorts of ill health — and how this compares to life in good health (e.g. would you rather live x years longer in good health or x+n years with heart failure?) — remain fraught questions.10 Playing very fast and loose, let’s give the ‘poorer health’ being relieved a disability weighting of 0.2511, so relieving this is ‘worth’ 5 x 0.25 = 1.25 DALYs averted.
    3. So the average ‘DALY per person’ effect is living five years longer (5) + 5 years without 0.25 disability (5 x 0.25 = 1.25). Combining the two gives 6.25 DALYs averted by medical care.
    4. Let’s now allocate all of this benefit to doctors for now. The UK has ~3 doctors per 1,000 people. So — again, roughly — we can take the population level impact (6.25 x 1,000 = 6,250 DALYs), then divide by the number of doctors (3) to get the impact of a UK doctor over their career: 6,250/3 ~ 2,100 DALYs averted.

    ‘How many DALYs does a doctor save?’ is not exactly ‘How many lives does a doctor save?’ There’s a facetious answer to the latter question: none, as everyone dies eventually. Perhaps a better one is taking a ‘life saved’ as (very roughly) equivalent to 30 DALYs: if you stop me dying now but I die in my 60s, this seems the sort of impact we have in mind with ‘saving a life’ (contrast this with, for instance, forestalling my death at 85, but I die two years later).

    So 2,100 DALYs averted divided by 30 comes to 70 ‘lives saved.’ Assuming a 40-year or so medical career, the typical doctor saves a couple of lives each year. This is pretty good — even if other careers might be better on this metric.

    However, this figure is not only a rough estimate, but an overestimate. One key reason is that it overestimates by neglecting to account for diminishing marginal returns, which we turn to in Part 2.

    Read next:  How many lives does a doctor save? (Part 2)

    To further refine our estimates of how many lives a doctor can save, we must account for diminishing marginal returns.

    Learn more:

    Notes and references

    1. This article focuses on the UK. But as we’ll see, the UK story generalises well to other high-income countries.

    2. From the Global Burden of Disease 2019 paper:

    3. See McKeown, T (1976) The role of medicine: Dream Mirage or Nemesis? (p 81). For reassurance this canonical example is not a cherry-picked outlier, see (e.g.) this paper of US infectious disease mortality:

      We again see a large secular decline in infectious disease mortality (note the very large aberration due to the 1918 influenza pandemic, and a reversal in trend from 1980 likely owed to the AIDS pandemic), in which a lot of the decline happened before effective medical countermeasures were developed. Penicillin was discovered in 1928, by which time US infectious disease mortality had already dropped by half. Although some vaccines were discovered earlier than this (and smallpox variolation was widely used in the US prior to 1900), most of the routine vaccines — starting with Polio — were discovered and deployed from the 1950s onwards; by 1950, infectious disease mortality had fallen by ~90% from 1900.

    4. The common challenge is that populations in poorer places and times tend not to carefully record data on their experience of disease, death, and deprivation. Although high-quality birth and death data can extend into the 19th century, for many other measures (e.g. death by cause, disaggregated economic activity, primary school completion rate, etc.) ‘records began’ somewhere between 1900 and 1950, and you have to wait until 1990 to get global coverage good enough for efforts like the Global Burden of Disease project. Numbers before then tend to be mixes of rough estimates and expert speculation.

      I still think this gives good support for the social determinants story of disease, even if large parts of it are hypothesised to have happened before data was recorded: whenever you happen to start recording, you see steep declines happening as soon as you start measuring it. A broad decline in ill health which is picked up at various points along the way by the data seems the best explanation.

    5. The two index studies for ‘RCTs’ on healthcare provision’ are the Oregon Medicaid health experiment and the RAND Health Insurance Experiment. The treatment group with cheaper/free access to healthcare generally had marginal improvements in measures of health: in the Oregon study, although measures like blood pressure and cholesterol favoured treatment, none reached statistical significance; in the RAND study there is a signal for those at highest risk given free care, but other indicators (whether for subgroups or not) were mostly flat.

      This data could be argued either way: despite sample sizes in the tens of thousands, these studies would be underpowered for any particular condition (save blood pressure, where most of the significant results are found), so perhaps these studies are still underpowered to detect many of the health benefits of healthcare. But free or discounted healthcare gave the treatment a significant benefit in-kind: for poorer households, the benefit was worth between one-third to two-thirds of their household income — perhaps they would have ended up even healthier if they were given the face value of the insurance in cash instead.

      Still — and notwithstanding how far these US results should be generalised — it seems fair to say these weigh against the conclusion that healthcare has a dramatic impact on health.

    6. See, for instance, “The Contribution of Medical Care to Mortality Decline:
      McKeown Revisited”
      by Johan P. Mackenbach.

    7. This sometimes goes under the charming phrase “competing causes of death.”

    8. We have mostly glossed over issues of quality as well as length of life. A common worry is the “expansion of morbidity”: although we are saving (really, extending) lives, this added lifespan is one with lots of ill health and disability. A morbid old age is a fate better than death, but less than we might hope for.

      The GBD uses DALYS, composed of years of life lost and years lived with a disability: both components are in steady global decline. In terms of ‘healthy’ or ‘disability-free’ life expectancy, these have increased roughly in proportion to life expectancy gains (both worldwide and for wealthy countries like the UK). So, although years lived in ill health have increased in absolute terms, one can expect to live a similar proportion of one’s life free of disability than in 1990. Although data is scarce and interpretation tricky, I think the picture is one of qualified reassurance.

    9. They are also likely overestimates. First, the additional life expectancy credited to medical interventions are unlikely to be at full health. Second, the technique for the estimate uses clinical trial data, and trials typically overestimate ‘real world’ performance of interventions when used in routine practice.

    10. DALYs (and similar metrics) are principally a tool for health economics to compare interventions between diseases: all else equal, a treatment that mildly improves dementia may be better than a complete cure for knee arthritis, because dementia could be much more important to someone’s overall health. DALYs let you quantify these differences with numerical disability weights (from 0 = full health, 1 = no better than death), and so prioritise health spending. So — for example — (moderate) knee (osteo)arthritis gets 0.079; (moderate) dementia 0.377. These also give trade-offs between length and quality of life: so ~ 12 years without mild osteoarthritis is as good as 1 extra year of healthy life; for dementia it is ~3.

      Such numbers abbreviate many controversies:

      • It is unclear what these weights should — or should not — be capturing: subjective well-being? Economic productivity? Impact on relatives/carers? A poll on a population’s preferences?
      • It is also unclear what these weights in fact capture: typically, those with a given condition rate it as less severe than the general public (cf.).
      • Application is often controversial: DALYs and similar measures can imply those with permanent disabilities should be deprioritised for medical care versus those without them (e.g. if I have dementia and you do not, extending your life by five years is ‘worth’ 5 DALYs, but extending mine by the same is only worth ~3 DALYs).

      I can’t answer these problems in a footnote — not least because I don’t know the answers myself. Fortunately, for our purposes, DALYs do not need to be perfect, but only good enough for a rough estimate to capture ‘quality of life’ improvements. Given social determinants are more ‘preventative’ than ‘disease modifying,’ rejecting such measures and only looking at lifespan and mortality would further discount the impact of medicine.

    11. Any number between 0.1 and 0.5 is defensible. I got to 0.25 by comparing the conditions Bunker highlights as ‘medical intervention success stories’ and eyeballing their (untreated) disability weights here.

    We are no longer accepting expressions of interest for this role. Please check our work with us page or job board to learn about future opportunities at 80,000 Hours.

    80,000 Hours is considering hiring a headhunting lead to build out the headhunting service we provide to other organisations. They will work with the Director of 1-on-1 to set and execute a strategy which uses our team of advisors’ unique network to find and recommend talented and altruistic candidates for high impact roles.

    We’re looking for someone who:

    • Has multiple years of experience in project management, research, or strategy, this could include roles in consulting, product management, or at early-stage startups or nonprofits.
    • Enjoys thinking about and working with different people in a variety of contexts, including maintaining relationships with major stakeholders, and developing models of people’s strengths to match them to specific roles.
    • Has a strong understanding of 80,000 Hours’ focus areas.

    This role is based in London, UK. The salary will vary based on your skills and experience, but the starting salary for someone with five years of relevant experience would be in excess of £70,000 per year.

    To express interest in this role, please complete this form.

    Note: This announcement is for an expression of interest, rather than a job opening. We’ll likely launch a formal hiring round within the next month or two, but aren’t sure exactly when. By filling out this form you’re giving us more information about who might be interested in this role. This information will help us decide whether to run a hiring round and how to structure it.

    If we do run a formal hiring round, we’ll email everyone who filled out this expression of interest, inviting them to fill out an application for the role.

    If you don’t hear back from us, we probably decided not to run a hiring round; please don’t take it as a rejection! You should feel very welcome to apply to future ads for 80,000 Hours positions (which we’ll list here and on our job board).

    About 80,000 Hours

    80,000 Hours’ mission is to get talented people working on the world’s most pressing problems. The effective altruism community, which we are part of, is growing in reach. But how do we make sure people are pursuing the right kinds of work in order to turn all those resources into long-term impact? This is the problem 80,000 Hours is trying to solve.

    We’ve had over eight million visitors to our website (with over 100,000 hours of reading time per year), and more than 3,000 people have now told us that they’ve significantly changed their career plans due to our work. 80,000 Hours is also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Survey.

    About the 1-on-1 team

    The 1-on-1 team at 80,000 Hours provides free, personalised career advice via video calls with our advisors. In 2022 we spoke to well over a thousand people who were looking to make a positive impact with their career, and we think offering tailored job recommendations is a key way we can help. Over the last two years we’ve grown the advising team and maintained a small headhunting service, providing shortlists of great candidates for a few especially exciting roles. We are now looking to build headhunting out into a key part of the value that the 1-on-1 team provides. You can learn more about our work and see some examples of people the 1-on-1 team has helped on our website.

    The role

    As headhunting lead, you would:

    • Design and lead a headhunting process that makes use of the unique network the 1-on-1 team has built to help fill the world’s most impactful roles.
    • Develop and maintain strong working relationships with hiring managers and team leads at key organisations.
    • Lead on the 1-on-1 team’s internal data strategy, finding ways to make the most of both our internal tracking and the large amount of data people choose to share with us.
    • Work closely with the 1-on-1 and Job Board teams to prioritise between organisations and roles.
      You would be managed by Michelle Hutchinson, Director of 1-on-1.

    Who we’re looking for

    We’d be most excited to hear from people who closely match the following:

    • Has multiple years of experience in project management, research, or strategy – this could include roles in consulting, product management, or at early-stage startups or nonprofits.
    • Enjoys working with different people in a variety of contexts, including maintaining relationships with major stakeholders, and developing working models of people’s strengths tomatch them to specific roles.
    • Has a strong understanding of 80,000 Hours’ focus areas.

    In addition, we’d be especially excited to find someone who:

    • Thinks critically, and communicates clearly about their team’s strategy, consistently prioritising well in order to maintain focus on the most important things.
    • Is comfortable approaching thorny, open-ended questions from multiple perspectives, moving between big picture thinking and getting stuck into the details of data analysis as the problem demands.
    • Is an excellent manager, and is excited to coach and develop the skills of those on their team.
    • Has experience with any of the following:
      • Longtermist and or existential risk strategy, particularly around talent allocation
      • CRM software, especially Salesforce
      • Hiring

    We’re aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all the suggested attributes. We’d especially like to encourage people from underrepresented backgrounds to express interest in this role, even if you don’t meet all the suggested criteria.

    Salary and benefits

    This is a full-time role, based at our London office (but you can work remotely for up to one third of the year if needed). We are able to support visa applications. The salary will vary based on your skills and experience, but the starting salary for someone with five years of relevant experience would be in excess of £70,000 per year.

    Our benefits include:

    • The option to use 10% of your time for self development
    • 25 days of paid holiday, plus bank holidays
    • Standard UK pension with 3% contribution from employer
    • Private medical insurance
    • Long-term disability insurance
    • Shower facilities, a small gym, and free food provided at our London office

    We are no longer accepting expressions of interest for this role. Please check our work with us page or job board to learn about future opportunities at 80,000 Hours.

      In a 2013 paper, Dr Toby Ord reviewed data compiled in the second edition of the World Bank’s Disease Control Priorities in Developing Countries,1 which compared about 100 health interventions in developing countries in terms of how many years of illness they prevent per dollar. He discovered some striking facts about the data:

      • The best interventions were around 10,000 times more cost effective than the worst, and around 50 times more cost effective than the median.
      • If you picked two interventions at random, on average the better one would be 100 times more cost effective than the other.
      • The distribution was heavy-tailed, and roughly lognormal. In fact, it almost exactly followed the 80/20 rule — that is, implementing the top 20% of interventions would do about 80% as much good as implementing all of them.
      • The differences between the very best interventions were larger than the differences between the typical ones, so it’s more important to go from ‘very good’ to ‘very best’ than from ‘so-so’ to ‘very good.’

      He published these results in The Moral Imperative towards Cost-Effectiveness in Global Health,2 which became one of the papers that started the effective altruism movement. (Note that Ord is an advisor to 80,000 Hours.)

      This data appears to have radical implications for people interested in doing good in the world; namely, by working on one of the best interventions in global health, you could achieve about as much as 50 people working on typical interventions in that area.

      In some earlier research, I showed that many charitable interventions don’t seem to work at all. But the DCP2 data showed that even among interventions that work, there are still huge differences in impact, suggesting it would be worth going to great efforts to find the most effective ones.

      So it’s crucial to know the extent to which this is true, and whether the results extend beyond global health.

      At the time, it was widely assumed these patterns would hold, but this wasn’t carefully checked.

      In this article, I’ve attempted to check these claims — I would really welcome further research into these questions, ideally by someone trained in social science.

      In the first section, I list all the datasets I’ve seen comparing cost-effectiveness, and compare them to Ord’s findings in global health — finding that the 80/20 pattern basically holds up. (There is more technical information — log standard deviations, and log-binned histograms showing distribution shapes — in the additional data appendix.)

      In the second section, I explore what we can learn from this data about how much solutions differ in effectiveness within cause areas, all things considered.

      I’ll argue the true forward-looking differences between interventions within cause areas are not as large or decision-relevant as these results make them seem; though they’re still far more important than most realise. In other words, they’re underrated by the world in general, but may be overrated by fans of effective altruism.

      In the third section, I speculate about the implications for how to choose interventions within a cause, arguing that it shows that the edge you gain from having a data-driven approach is less than it first seems.

      Overall, I roughly estimate that the most effective measurable interventions in an area are usually around 3–10 times more cost effective than the mean of measurable interventions (where the mean is the expected effectiveness you’d get from picking randomly). If you also include interventions whose effectiveness can’t be measured in advance, then I’d expect the spread to be larger by another factor of 2–10, though it’s hard to say how the results would generalise to areas without data.

      1. Data on how much solutions differ in effectiveness within cause areas

      The original dataset: Disease Control Priorities in Developing Countries (second edition)

      I’ll start with the dataset used in Ord’s original paper as our point of comparison.

      The DCP2 was published in 2006. It compared 107 interventions within global health in poor countries, ranging from surgery to treat Kaposi’s sarcoma, to public health programmes like distributing free condoms to prevent AIDS.

      For each intervention, there’s an estimate of how much illness it prevents — measured in disability-adjusted life years (DALYs) — and how much it costs. The ratio of the two is the cost effectiveness.

      If we line up the interventions in order of cost effectiveness (shown on the Y-axis), we get the following graph:

      DCP2 cost effectiveness graph

      We can see that the first 60 interventions are near the zero line, and so aren’t very effective. But the top 20 or so achieve a huge amount per dollar.

      MeasureDALYs averted per US $1,000
      Mean cost effectiveness23
      Median cost effectiveness5
      Mean cost effectiveness of the 2.5% most cost-effective interventions250 (52x median, 11x mean)3
      Mean cost effectiveness of the 25% most cost-effective interventions794

      Other studies of global health

      In the blog post GiveWell’s top charities are increasingly hard to beat, Alexander Berger, co-CEO of Open Philanthropy3, found three surveys of cost-benefit analyses for health interventions in the developing world: the DCP2, the more current third edition of the same report (DCP3)4, and a WHO-CHOICE review (which in turn provides two datasets: one for the average costs of the interventions, and one for the incremental costs).

      This allows us to compare the DCP2 to some alternative and more current analyses. They turn out to show a similar pattern.

      Disease Control Priorities (third edition)

      DCP3 cost effectiveness graph

      MeasureDALYs averted per US $1,000
      Mean cost effectiveness17
      Median cost effectiveness4
      Mean cost effectiveness of the 2.5% most cost-effective interventions170 (38x median, 10x mean)
      Mean cost effectiveness of the 25% most cost-effective interventions56

      WHO-CHOICE (using average cost effectiveness)

      WHO CHOICE average cost effectiveness graph

      MeasureDALYs averted per Intl$1,000
      Mean cost effectiveness29
      Median cost effectiveness12
      Mean cost effectiveness of the 2.5% most cost-effective interventions310 (25x median, 10x mean)
      Mean cost effectiveness of the 25% most cost-effective interventions85

      WHO-CHOICE (using incremental cost effectiveness)

      WHO CHOICE incremental cost effectiveness graph

      MeasureDALYs averted per Intl$1,000
      Mean cost effectiveness41
      Median cost effectiveness7
      Mean cost effectiveness of the 2.5% most cost-effective interventions670 (93x median, 16x mean)
      Mean cost effectiveness of the 25% most cost-effective interventions150

      Health in high-income countries: public health interventions in the UK (NICE)

      Berger also found a dataset for the UK — National Institute for Health and Care Excellence (NICE)5 — which enables us to extend the analysis to a high-income country.

      The data is related to public health, and covers about 200 interventions focused on things like helping people stop smoking, reducing traffic accidents, improving dental health, and increasing testing for sexually transmitted infections. (The analysis was done in terms of quality-adjusted life years (QALYs) added instead of DALYs avoided, but for our purposes we can take these as equivalent.)

      Here we find a similar pattern to health interventions in poor countries. Overall, the degree of spread seems similar or slightly larger than the DCP2.

      However, in the DCP2, the mean, median, and top 2.5% are respectively 25x, 38x, and 16x higher, so the whole distribution is shifted upwards — in other words, interventions are way more cost effective in developing vs developed countries.

      In fact, the difference in cost effectiveness of health interventions for rich and poor countries is so significant that even the top 2.5% of interventions in the NICE data creates fewer extra years of healthy life per pound spent than the mean in the developing world DCP2 data — and note the mean is the effectiveness you’d expect if you picked randomly. (Health in poor countries and health in rich countries are usually considered different cause areas by people interested in effective altruism in part for this reason.)

      Interestingly, this roughly lines up with the difference in income between the UK and these countries, which makes sense, since richer people will generally be able to pay a lot more to protect their health (and logarithmic returns to health spending will mean the cost-effectiveness difference is proportional to the difference in income).

      UK NICE UK public health interventions graph

      MeasureQALYs created per £1,000
      Mean cost effectiveness1.0
      Median cost effectiveness0.1
      Mean cost effectiveness of the 2.5% most cost-effective interventions15.4 (120x median, 15x mean)
      Mean cost effectiveness of the 25% most cost-effective interventions3.7

      One additional source of data we didn’t have a chance to review is the CEVR CEA registry, which contains over 10,000 cost-effectiveness analyses on health interventions — this would be worth checking in future work.

      US social interventions: Washington State Institute for Public Policy Benefit-Cost Results database

      Now, can we extend the analysis beyond health?

      Alexander Berger also found a database of cost-benefit analyses for about 370 US social policies, compiled by the Washington State Institute for Public Policy.6 The data spans issues from substance use disorders, to criminal justice reform, to higher education and public health, so gives us information on multiple cause areas.

      The studies aimed to account for a variety of benefits from the programmes (rather than just health), which were then converted into dollars and compared to the costs (also measured in dollars).

      First, looking at positive cost-benefit interventions (i.e. where interventions were worth the money spent), we again find a similar pattern:

      US Social policies, positives only graph

      MeasureRatio
      Mean cost-benefit ratio22
      Median cost-benefit ratio5
      Mean cost-benefit ratio of the 2.5% most cost-effective interventions360 (68x median, 16x mean)
      Mean cost-benefit ratio of the 25% most cost-effective interventions75

      Note: there are no units because cost-benefit ratios are unitless.7

      Interestingly, and unlike within health, about 70 (19%) of the interventions had negative benefits — i.e. they made people worse off overall.

      Though, they were distributed in a similar way. This is evidence in favour of a recent paper about ‘negative tails’ in doing good.

      US Social policies, negatives only graph

      MeasureRatio
      Mean cost-benefit ratio-8
      Median cost-benefit ratio-0.8
      Mean cost-benefit ratio of the 2.5% least cost-effective interventions-140 (172x median, 18x mean)
      Mean cost-benefit ratio of the 25% least cost-effective interventions-29

      But before we get too pessimistic, it’s important to remember that only a minority of interventions had a negative impact. The mean over the entire dataset was still positive. This means that on average people in the field are doing good — it’s just important to choose carefully.

      Criminal justice reform

      Since these interventions span many different issues, we might ask what would happen if we further break down the data.

      First, let’s focus only on criminal justice reform — just one of the causes in the full dataset.

      The overall degree of spread is somewhat reduced, though still significant. This is what we’d expect since it’s a narrower domain, since one source of variation — the differences between cause areas — has been eliminated (though it could also be caused by a small sample size meaning there were no outliers in the sample).

      US criminal justice reforms, positives only graph

      Summary statistics (ignoring the interventions with negative cost effectiveness):

      MeasureRatio
      Mean cost-benefit ratio6.8
      Median cost-benefit ratio4.8
      Mean cost-benefit ratio of the 2.5% most cost-effective interventions19.8 (6.0x median, 3,7x medan)
      Mean cost-benefit ratio of the 25% most cost-effective interventions14.9

      Pre-K to 12 education

      We find a similar pattern of reduced but still significant spread if we focus only on education interventions.

      It’s also interesting to note that there seem to be significant differences between the issues: the education interventions come out about four times more cost effective on average compared to criminal justice reform, even though both are in the same broad area of US social interventions.

      US education interventions graph

      Summary statistics (ignoring the interventions with negative cost effectiveness):

      MeasureRatio
      Mean cost-benefit ratio27
      Median cost-benefit ratio7.8
      Mean cost-benefit ratio of the 2.5% most cost-effective interventions160 (21x median, 6x mean)
      Mean cost-benefit ratio of the 25% most cost-effective interventions85

      Climate change: Gillingham and Stock

      Kenneth Gillingham and James Stock assessed the cost effectiveness of about 20 interventions to reduce greenhouse gas emissions.8 They compared the interventions in terms of tonnes of CO2-equivalent (tCO2e) greenhouse gas emissions avoided per dollar.

      The pattern here was very similar to the DCP2.

      Gillingham and Stock graph of interventions to reduce GHG

      Summary statistics (ignoring the interventions with negative cost effectiveness):

      MeasuretCO2e avoided per US $1,000
      Mean cost effectiveness23
      Median cost effectiveness10
      Mean cost effectiveness of the 2.5% most cost-effective interventions180 (18x median, 8x mean)
      Mean cost effectiveness of the 25% most cost-effective interventions65

      Note: the authors gave a range of cost-effectiveness estimates for each intervention — for our analysis we used the middle of these ranges.

      Education in the UK: The Education Endowment Foundation

      The UK’s Education Endowment Foundation provides a toolkit that summarises the evidence on different UK education interventions.

      Danielle Mason, Head of Research at the organisation, told me that the toolkit attempts to include all relevant, high-quality quantitative studies.

      Each type of intervention is assessed based on (i) strength of evidence; (ii) effect size, measured in ‘months of additional schooling equivalent’; and (iii) cost. See how these scores are assessed here. We only considered studies with a “strength of evidence” score greater than 1.

      This data roughly follows a similar distribution to the survey of US social interventions above, though the degree of spread is smaller. This could be because the variety of interventions studied is smaller. It might also be because most of the figures are based on meta-analyses, rather than single estimates. Single estimates tend to be more noisy, increasing spread. Meta-analyses are also more likely to be positive because they combine lots of smaller studies, so are less likely to be underpowered.

      It’s unclear whether this data follows the same sort of heavy-tailed or lognormal distribution as the interventions discussed above. While the mean of the data is only slightly higher than the median, few interventions were studied, meaning it’s hard to determine the overall distribution.

      UK education interventions graph

      MeasureMonths of additional schooling equivalent per £100
      Mean cost effectiveness3.9
      Median cost effectiveness3.75
      Mean cost effectiveness of the 2.5% most cost-effective interventions8.8 (2.3x median, 2.2x mean)
      Mean cost effectiveness of the 25% most cost-effective interventions7.5

      Education in low-income countries: Education Global Practice and Development Research Group Study

      The Education Global Practice and Development Research Group conducted a study of 41 education interventions9 in low- and middle-income countries, published in 2020.

      They compared interventions in terms of how many additional years of schooling they produced per $1,000, aiming to take into account both the number and the quality of the years. They called their metric “learning-adjusted years of schooling” (LAYS).

      This education data is much more clearly heavy-tailed — even more so than the DCP2 — and likely follows a lognormal distribution.

      Education in low income countries graph

      MeasureLAYS per US $100
      Mean cost effectiveness7.1
      Median cost effectiveness0.64
      Mean cost effectiveness of the 2.5% most cost-effective interventions140 (220x median, 20x mean)
      Mean cost effectiveness of the 25% most cost-effective interventions27

      Some other datasets

      The following studies are less comprehensive and rigorous than those above, but also help to check the general pattern in some different areas. I’ve included them to be comprehensive in what’s covered, and to avoid cherry picking studies that back up my findings. I’d be interested to learn about more studies of this kind.

      Note that all of these are much narrower sets of interventions than those covered above, which will likely reduce spread. They also don’t explicitly account for costs, which I’ll argue in the AidGrade section means they probably understate spread.

      AidGrade’s dataset of international development interventions — a potential counterexample?

      We’ve seen some arguments against interventions being heavy-tailed based on AidGrade‘s dataset. They have a large dataset of interventions within international development, going beyond health.

      They found that the distribution of effect sizes for interventions in their data was roughly normal rather than lognormal, though had a slightly heavier-than-normal positive tail.

      However, this doesn’t mean that the distribution of cost effectiveness is normal, because there are two further factors to consider:

      First, effect size is measured relative to many different types of outcome. This means that, roughly speaking, an intervention that cured 20% of people of the common cold would be given the same value as an intervention that cures 20% of people of cancer. Ideally there would be an attempt to weigh up the value of different outcomes across different studies (such as with DALYs, LAYS, or tonnes of CO2). This would add a source of variation, increasing spread.

      Perhaps more importantly, costs also need to be considered. The costs of different interventions often differ by orders of magnitude, and so dividing by cost could increase spread a lot.

      This would not be the case if costs and impact were closely correlated (which is what we’d hope to see if resources are allocated efficiently); however, empirically there seems to be a weak relationship between the two.

      For instance, in the DCP2 data, dividing the average impact in DALYs by costs increases the degree of spread.

      I’ve observed the same pattern in the other datasets I’ve checked. For instance, the Education Endowment Foundation dataset includes both average impact (measured in months of extra schooling equivalent) and costs, and the distribution of impact per cost is wider than average impact alone.

      Without accounting for these two additional factors, we can’t draw conclusions about the shape of the distribution of cost effectiveness.

      And my expectation is that if we did consider them, the distribution would become much wider, and would most likely be lognormal like the others.

      You can see more explanation of the issue here.

      Personal actions to fight climate change

      Founders Pledge produced estimates of the effectiveness of various personal lifestyle decisions for fighting climate change. They produced two sets of estimates: one accounts for government climate targets and policies, and the other does not.

      This data shows somewhat less spread than in other datasets; however, they only compare “actions,” without fully correcting for how some of these actions are probably much more costly than others. If we added this additional source of variation, and calculated “CO2 averted per unit of effort,” then I would expect a significantly wider spread.

      Founders Pledge also estimated that a US $1,000 donation to their top choice climate charity — which seems somewhat comparable to the costs of the interventions here — would avert 100 tonnes of CO2. If we included this in the dataset, then the top intervention would be 250 times the median, and 140 times the mean.

      (The below figures do not include this intervention.)
      personal actions to reduce CO2 emissions, bar chart

      personal actions to reduce CO2 emissions, line graph

      MeasuretCO2e avoided
      Mean effectiveness0.67
      Median effectiveness0.4
      Mean effectiveness of the 25% most effective interventions1.7
      Effectiveness of the best intervention2.4 (6x median, 3.6x mean)

      Units of tonnes of CO2-equivalent (tCO2e) greenhouse gas emissions avoided, accounting for policy.

      Get Out the Vote tactics

      In Get Out the Vote: How to Increase Voter Turnout, the authors reviewed strategies for political parties in the US to encourage people to vote. They looked at 19 strategies — including various kinds of direct mailing, leafleting, phoning, and door-to-door knocking — first estimating how much they increased turnout as a percentage10:

      Then they made estimates of cost effectiveness:

      Cost effectiveness of voter turnout interventions

      This dataset is too small to properly measure its shape, but we can see that four of the interventions didn’t have measurable effects, while the top clearly stood out.

      MeasureVotes per US $1,000
      Mean cost effectiveness26
      Median cost effectiveness22
      Mean cost effectiveness of the top 25% most cost-effective interventions51
      Cost effectiveness of the best intervention71 (3.2x median, 2.7x mean)

      Patterns in the data overall

      Focusing mainly on the large datasets (>50 interventions), here are the key summary stats:

      MedianMeanMean of top 2.5%
      Disease Control Priorities in Developing Countries (2nd edition)4 DALYs averted per US$1,00017 DALYs averted per US$1,000170 DALYs averted per US$1,000
      WHO-CHOICE (using average cost effectiveness)12 DALYs averted per Intl$1,00029 DALYs averted per Intl$1,000310 DALYs averted per Intl$1,000
      WHO-CHOICE (using incremental cost effectiveness)7 DALYs averted per Intl$1,00041 DALYs averted per Intl$1,000670 DALYs averted per Intl$1,000
      NICE Cost-effectiveness estimates0.1 QALY created per £1,0001.0 QALY created per £1,00015.4 QALYs created per £1,000
      Washington State Institute for Public Policy Benefit-Costs Results Database (positive interventions)522360
      Education Global Practice and Development Research Group Study0.64 LAYS per US$1007.1 LAYS per US$100140 LAYS per US$100
      Gillingham and Stock (climate change interventions)10 tCO2e avoided per US$123 tCO2e avoided per US$1180 tCO2e avoided per US$1

      What patterns do we see?

      There appears to be a surprising amount of consistency in the shape of the distributions.

      The distributions also appear to be closer to lognormal than normal — i.e. they are heavy-tailed, in agreement with Berger’s findings. However, they may also be some other heavy-tailed distribution (such as a power law), since these are hard to distinguish statistically.

      Interventions were rarely negative within health (and the miscellaneous datasets), but often negative within social and education interventions (10–20%) — though not enough to make the mean and median negative. When interventions were negative, they seemed to also be heavy-tailed in negative cost effectiveness.

      One way to quantify the interventions’ spread is to look at the ratio of between the mean of the top 2.5% and the overall mean and median. Roughly, we can say:

      • The top 2.5% were around 20–200 times more cost effective than the median.
      • The top 2.5% were around 8–20 times more cost effective than the mean.

      Overall, the patterns found by Ord in the DCP2 seem to hold to a surprising degree in the other areas where we’ve found data.

      Ratio of top 2.5% to median13Ratio of top 2.5% to mean
      Disease Control Priorities in Developing Countries (2nd edition)5211
      WHO-CHOICE (using average cost effectiveness)257
      WHO-CHOICE (using incremental cost effectiveness)9316
      NICE Cost-effectiveness estimates12015
      Washington State Institute for Public Policy Benefit-Costs Results Database (positive interventions)6816
      Education Global Practice and Development Research Group Study22020
      Gillingham and Stock (climate change interventions)188

      2. Given this data, how much do solutions within a cause area actually differ in effectiveness?

      In the DCP2, the top 2.5% of interventions were measured to be on average about 50 times more cost effective than the median. Does that mean you can actually have 50 times the impact?

      It’s unclear, and I think it’s probably hard.

      For one thing, the data we’ve covered are mostly backward-looking, and may not be a good reflection of realistic forward-looking estimates that take account of all sources of error, including model error.

      Here’s an extreme example of the difference. Imagine 1,000 people buy lottery tickets, and one wins. The measured backward-looking distribution of payoffs is extreme — one person won a huge amount and everyone else won nothing. But beforehand, everyone had the same chance of winning, so there was no difference in the forward-looking value of the lottery tickets to each person.

      Something similar could be happening in our studies. Perhaps many interventions looked similarly promising ahead of time, but only a handful succeeded — so it’s only when we look back that we see a large spread.

      In this section, I list some ways that the data might overstate the degree of spread that’s looking forward, and some ways it might understate it. Overall, my guess is that the data overstates the true differences, but there is still a lot of spread.

      (Note there’s nothing new about what I’m saying here (e.g. see this post by GiveWell from 2011). However, I often don’t see these points appreciated, so I thought it would be useful to relist them. These points are based on conversations I’ve had with people who have done research on these topics. I’m not a statistician and would love to see a more rigorous analysis.)

      Ways the data might overstate the true degree of spread

      Regression to the mean

      There’s a huge degree of error in the estimates. Even if the estimates of DALYs averted per dollar were correct, DALYs don’t perfectly reflect improvements in health, and improvements in health aren’t all that matter.

      Studies also often fail to generalise to different future contexts. Eva Vivalt found:

      The typical study result differs from the average effect found in similar studies so far by almost 100%. That is to say, if all existing studies of an education program find that it improves test scores by 0.5 standard deviations — the next result is as likely to be negative or greater than 1 standard deviation, as it is to be between 0-1 standard deviations.

      The median absolute amount by which a predicted effect size differs from the true value given in the next study is 99%. In standardised values, the average absolute value of the error is 0.18, compared to an average effect size of 0.12.

      So, colloquially, if you say that your naive prediction was X, well, it could easily be 0 or 2*X — that’s how badly this estimate was off on average. In fact it’s as likely to be outside the range of between 0 and 2x, as inside it.

      Finally, studies often have incorrect findings. In the replication crisis, it’s been found that perhaps 20–50% of studies don’t replicate, depending on the field and methodology.

      All this error in the estimates means that the interventions that appear to be best have probably benefitted from positive luck, and are not as good as they seem — a phenomenon called regression to the mean.

      In other words, measured impact is given by true impact and noise or random variation. If an intervention seems really good, it might be due to its true impact being high, or because the noise happened to be positive.

      Going forward, noise is as likely to be negative as positive. This means that future measurements of the best interventions will probably look worse.

      How large is this effect?

      As a rough guide, researchers I’ve spoken to seem to think that effectiveness of the better interventions should be reduced by at least twofold, though the reduction could be tenfold or more.

      Regression to the mean can also change the order of the interventions, because the effect is stronger for more error-prone estimates.

      Technical aside on estimating regression to the mean

      Ideally, we could start with a prior distribution, and then perform a Bayesian update using our measurements (with an assumption about how noisy they are). This would give us a posterior distribution of cost effectiveness, which could be compared to the original.

      GiveWell did a quantitative analysis along these lines (and also see the comment thread), showing that when your estimate is highly uncertain, you don’t update much from your prior estimate of effectiveness, and vice versa.

      However, this analysis was performed for normal distributions (rather than lognormal) and with hypothetical values, so I’m not easily able to adapt it to our purposes here. If your prior distribution is lognormal (which mine is), then the reduction in spread will be significantly reduced.

      The researcher Greg Lewis used a different method to quantitatively correct for regression to the mean. He estimates that if we assume our cost-effectiveness estimates are 0.9 correlated with the true value, the real cost effectiveness of the top interventions is about half as much as the original figure.

      I expect that the raw estimates in the DCP2 are much more noisy than a correlation of 0.9 would imply. If I repeat Lewis’s process but with a correlation of 0.5, I get a factor of 50 reduction in true cost effectiveness compared to the initial estimate. This is at least proof of concept that regression to the mean can be a very large effect.

      I’m aware of attempts to do quantitative analyses for lognormal distributions by several others. I’d be keen to see someone try to combine all these analyses and apply them to the question of how much interventions can be expected to differ in cost effectiveness.

      Interventions may no longer be available

      The existence of research on an intervention doesn’t mean that it’s practical for a philanthropist or government to carry it out, and this is especially true of the best interventions.

      If 1% of actors in the field are sensitive to evidence, then they will focus on the most promising 1% of interventions, ‘cutting off’ the tail of the distribution. This means that the best available interventions are often worse than the best that have been studied.

      We’ve seen this play out in global health. One of the most cost-effective interventions in the data was vaccinating children, but these opportunities are almost all taken by the Gates Foundation and other international aid agencies.

      Non-primary outcomes might be important too

      All of my remarks apply only to the primary outcome studied (e.g. DALYs), but we also need to consider that most interventions have multiple outcomes that might matter.

      For instance, many investments in health benefit the patients (as measured by DALYs) but might also have positive or negative effects on health infrastructure in the country, such as through training medical professionals, or discouraging government investment.

      If these effects are small compared to the primary outcome, they can be safely ignored. They can also be ignored if they correlate closely with the primary outcome, because then we can use the primary outcome as a proxy for them.

      For instance, many health programmes will also boost the income of the recipients (because if you’re healthy, you can earn more), but we should expect income benefits to correlate with health benefits, so the effects on income will partially factor out when we compare cost-effectiveness ratios.

      However, if these other outcomes are large and positive (or if they anticorrelate with the primary outcome), then accounting for them could reduce the apparent difference in effectiveness between interventions measured on the primary outcome.

      For instance, if one version of a programme spends more time training local healthcare providers, it might cost more to implement (reducing its effectiveness measured with DALYs in the short term) while doing more to improve health infrastructure and having a longer-lasting impact.

      If the non-primary outcomes are large and negative, then they could completely reverse which interventions seem best.

      Ways the data could understate differences between the best and typical interventions

      Differences in execution and location

      Some organisations will implement the same intervention better than others. Accounting for this difference will increase the spread in effectiveness between best and worst organisations you might support.

      Once we’ve excluded organisations that seem obviously incompetent (which perhaps have zero impact), my impression is that the degree of variation on this factor is relatively small — perhaps around a factor of two between plausibly good organisations.

      However, this is not guaranteed. For highly complex interventions, there are multiple steps that all need to be completed successfully, and if any step fails, the whole intervention fails. In economics this is called an ‘O-ring’ production process. In such a process, a small difference in the chance of successfully implementing each step adds up to a large difference in the chance of completing the whole process.

      In addition, great organisations seem to produce more positive non-primary outcomes. For example, GiveDirectly has carried out several studies of its work, helping to create data on different ways of doing cash transfers that inform international development efforts more broadly.

      In a similar vein, implementing the same intervention in different locations can have a big effect on cost effectiveness. For instance, malaria deaths in Burkina Faso are about five times the rate in Kenya, and about 50 times the rate in India. For a preventative intervention like nets, the benefit is proportional to the chance of infection, which makes a proportional difference to cost effectiveness.

      Selection effects in which interventions were chosen

      Which interventions are studied are not chosen at random; instead, they are chosen because they are unusually interesting and more likely to have especially large positive effects. Normally the point of running a trial is to find something better than what people currently focus on.

      Running a trial is also expensive, so any intervention that has made it to that point must have a serious backer, which is probably also evidence that it’s better than average.

      This is a reason to expect interventions that have been measured to be better than the full set of interventions that could be measured — i.e. there will be lots of hopeless interventions that no one wanted to research. This (combined with ignoring non-primary outcomes) could explain why so few of the interventions have negative effectiveness, even though it seems likely that some non-negligible fraction of international development interventions were negative.

      The findings are probably better interpreted as the spread of effectiveness among interventions that were ‘plausibly good,’ rather than all interventions in the area — which will show more spread.

      One improvement that could be made in future work would be to weight each intervention by the amount invested in it.

      Difficult-to-measure programmes are not included

      When we look at empirical studies of effectiveness, they often only cover interventions that can be measured in trials, and nearly always exclude interventions like funding research and lobbying government.

      If we look at the history of philanthropy, many of the highest-impact interventions seem to be much less measurable, and have involved advocacy, policy change, and basic research.

      This means that we should expect some of the best interventions in an area to be missing from these reviews. If these were added, then it would increase the potential spread of effectiveness among everything that’s available (though not among the interventions that have been studied).

      I expect that the field of global health provides a best-case scenario for using data to select cost-effective interventions. This is because global health interventions are relatively easy to measure, which makes regression to the mean less pressing. In an area with much weaker estimates, like criminal justice reform, I expect the true degree of spread is more overstated than the data suggests.

      A case study: GiveWell and the DCP2 data

      It’s instructive to look at a real attempt to apply these corrections to see how they turned out.

      When Giving What We Can started recommending global health interventions in 2009, it started with the most cost-effective interventions in the DCP2, and then looked for charities that seemed to competently implement those interventions.

      This led Giving What We Can to recommend the SCI Foundation, Against Malaria Foundation (AMF), and Deworm the World (DtW) as donation opportunities. GiveWell also started to recommend the same charities.

      In the DCP2 data, deworming was one of the most cost-effective interventions measured, at 333 DALYs avoided per US $1,000. Insecticide-treated bednets were the eighth most cost-effective intervention, at 90 DALYs avoided per US $1,000.

      Since 2009, these interventions and charities have been subject to much additional scrutiny by GiveWell.

      How well did those figures turn out to project forward, taking account of all of the factors above?

      This story has both a positive and a negative side.

      On the positive side, GiveWell still recommends AMF as among its most cost-effective charities, which is impressive 10 years later.

      On the negative side, GiveWell’s best estimate is that AMF is much less cost effective than the DCP2 data would naively suggest. The latest versions of GiveWell’s cost-effectiveness sheets (as of 2022) give an estimate of under $5,000 per life saved in some countries. Saving a life is often equated to avoiding 30 DALYs, so this would be equivalent to a cost effectiveness of 6 DALYs avoided per $1,000. In the original DCP2 data, insecticide-treated bednets were estimated to avoid 90 DALYs per $1,000, so GiveWell’s 2022 estimate is about 15 times lower. Some of this is due to the best opportunities having been used up in the last 10 years, but I think most is due to regression from less accurate estimates.

      So, we have an empirical estimate that the cost effectiveness of the best interventions in DCP2 were overstated by about an order of magnitude (though they were still very high).11

      The picture with deworming is more complicated. The initial estimates were a great example of regression to the mean, and found to be full of errors. Then further doubts were cast on the most important studies in the so-called “worm wars”. GiveWell now believes deworming most likely doesn’t have much impact, but there’s a small chance it greatly increases income in later life, and because it’s so cheap, the expected value of deworming is still high.

      The latest version of their cost-effectiveness model (version 4 from August 2022) shows they think deworming is similarly cost effective to malaria nets. This would mean that deworming’s effectiveness has also regressed by about a factor of 10 compared to the DCP2 data, but is also still among the most effective health interventions.

      Though, it’s worth noting that this effectiveness is mainly driven by effects on income rather than health. You could see this as showing that the health effects have regressed to the mean far more than tenfold, or as an example of how considering multiple outcomes increased spread.)

      It’s also worth noting that GiveWell recently started to use robustness of impact as a criterion for its top charities, so has removed deworming from their list of top charities (though they might continue to make grants from their new All Grants Fund). Learn more about these changes in GiveWell’s blog post.

      Coming to an overall estimate of forward-looking spread

      To come to an overall estimate of the degree of spread, you need to consider your priors,12 the strength of the empirical evidence, and the significance of the factors above.

      I don’t expect the ‘market’ for charitable interventions to be especially efficient, which means there is scope for large differences. And since effectiveness is given by a product of factors, there’s potential for a heavy tail.

      Moreover, if we’re unsure between an efficient world with small differences and an inefficient world with big differences, then our expected distribution has a lot of spread.13

      My overall view is that there’s a lot of spread, though not as much as naively going with the data would suggest.

      Perhaps the top 2.5% of measurable interventions within a cause area are actually 3–10 times better than the mean of measurable interventions, rather than the 8–20 times better we see in the data (and the lower end seems more likely than the upper end to me).

      If we were to expand this to also include non-measurable interventions, I would estimate the spread is somewhat larger, perhaps another 2–10 fold. This is mostly based on my impression of cost-effectiveness estimates that have been made of these interventions — it can’t (by definition) be based on actual data. So, it’s certainly possible that non-measurable interventions could vary by much more or much less.

      Overall, I think it’s defensible to say that the best of all interventions in an area are about 10 times more effective than the mean, and perhaps as much as 100 times.

      Response: is this consistent with smallpox eradication?

      Toby Ord roughly estimated that eradicating smallpox has saved lives for $25 per life (so far). Is the existence of interventions as cost effective as that consistent with my estimates?

      $25 per life is around 1,000 DALYs averted per $1,000. This would place it in roughly the top 1% of the original DCP2 data.

      If the true degree of variation is a factor of 10 less than the DCP2 data suggests, but we hold the cost-effectiveness estimate for smallpox eradication fixed, then it might mean that smallpox eradication is actually in the top 0.1% of interventions.

      This doesn’t seem unreasonable, given that it was arguably the best buy in global health in the whole of the 20th century.

      Moreover, smallpox eradication was not guaranteed to succeed. Its expected cost effectiveness at the time would have been lower than the cost effectiveness we measured after we knew it was successful.

      So I think my estimates are consistent with the existence of smallpox eradication.

      Response: is this consistent with expert estimates?

      A recent survey of experts in global health found that the median expert estimated the difference between the best charity in the area and the average in terms of cost effectiveness is around 100 times.14

      This is a larger degree of spread than I estimate — what explains the difference?

      One factor is that this survey question was for ‘the best’ charity, whereas my estimate is for the top 2.5%.

      Another factor is that the survey only asked for the difference between the ‘average’ and the best, but didn’t specify whether that meant the median or the mean. Interpreting it as the median seems more natural to me, in which case a difference of around a hundredfold is plausible.

      It also seems plausible that many experts interpreted the question as being about backward-looking estimates, rather than a truly forward-looking estimate that fully adjusts for regression to the mean and the other issues I’ve noted.

      That said, they are experts in global health and I’m not, so I think it would be reasonable to use their estimate rather than mine.

      3. How much can we gain from being data-driven?

      People in effective altruism sometimes say things like “the best charities achieve 10,000 times more than the worst” — suggesting it might be possible to have 10,000 times as much impact if we only focus on the best interventions — often citing the DCP2 data as evidence for that.

      This is true in the sense that the differences across all cause areas can be that large. But it would be misleading if someone was talking about a specific cause area in two important ways.

      First, as we’ve just seen, the data most likely overstates the true, forward-looking differences between the best and worst interventions.

      Second, it often seems fairer to compare the best with the mean intervention, rather than the worst intervention.

      One reason is that as the effectiveness of an intervention approaches zero, the ratio between it and the best intervention approaches infinity. So by picking from among the worst interventions, you can make the ratio between it and the best arbitrarily high. This is a real problem, because the worst interventions do often have zero (or even negative) cost effectiveness. (Though it does also say something about the world that such ineffective interventions are being implemented!)

      What if we compare the best interventions to the median rather than the worst?

      If someone has already chosen a particular intervention that you know is near the median, then you could point out that the backward-looking difference in cost effectiveness is often over 100 times.

      But if we don’t know anything about what they’ve chosen, then it seems more accurate to model them as picking randomly.15 That means they might pick one of the best interventions by chance. A random guess gives you the mean of the distribution rather than the median.

      In a distribution with a heavy positive tail,16 the mean tends to be a lot higher than the median. For instance, in the DCP2 the mean was 22 DALYs averted per $1,000, compared to five DALYs averted for the median — about four times higher.

      Moreover, if it’s possible to use common sense to screen out the obviously bad interventions, then they may effectively be picking randomly from the top 50% of interventions, and their expected impact would be twice the mean.

      So, comparing the best to the mean, rather than to the worst or median, will tend to reduce the degree of spread.

      If we also consider the difficult-to-measure interventions that are missing from the datasets, but make up the positive tail, the difference between the mean and the median will be even larger.

      Overall, my guess is that, in an at least somewhat data-rich area, using data to identify the best interventions can perhaps boost your impact in the area by 3–10 times compared to picking randomly, depending on the quality of your data.

      This is still a big boost, and hugely underappreciated by the world at large. However, it’s far less than I’ve heard some people in the effective altruism community claim.

      In addition, there are downsides to being data-driven in this way — by insisting on a data-driven approach, you might be ruling out many of the interventions in the tail (which are often hard to measure, and so will be missing).

      This is why we advocate for first aiming to take a ‘hits-based’ approach, rather than a data-driven one.

      Another important implication is that I think intervention selection is less important than cause selection. I think the difference between interventions in a single problem area is much smaller than the difference in effectiveness between problem areas (e.g. climate change vs education) — which I think are often a hundredfold or a thousandfold, even after accounting for the issues mentioned here (such as regression to the mean). I go through the argument in the linked article — but one quick way to see this is that comparing across causes introduces another huge source of variation in how much good an intervention does.

      This means, in terms of effectiveness, it’s more important to choose the right broad area to work in than it is to identify the best solution within a given area.

      This is one reason why the effective altruism community focuses so much on deciding which problem to focus on, rather than trying to improve the effectiveness of efforts within a wide range of causes.

      Though of course it’s ideal to both find a pressing problem and an effective solution. Since the impact of each step is multiplicative, the combined spread in effectiveness could be 1,000 or even 10,000 fold.

      Thank you especially to Benjamin Hilton for doing most of the data analysis in this post, and for Toby Ord’s initial comments on the draft. All mistakes are my own.

      Discuss this article on the EA Forum. Or ask me a question about it on Twitter here.

      You might also be interested in

      Speak to our team one-on-one

      If you’ve made it through this giant article, you’re probably the kind of person our advising team would like to speak to! They can help you consider your options, make connections with others working on our top problem areas, and possibly even help you find jobs or funding opportunities. (It’s free.)

      SPEAK WITH OUR TEAM

      Appendix: Additional data

      Standard deviation log10 cost effectiveness

      Another way of looking at the spread of these distributions is by looking at the standard deviation of the log cost effectiveness.

      This is because these are heavy-tailed distributions, so the regular standard deviation isn’t meaningful. Instead, we can take the log of cost effectiveness.

      Many heavy-tailed distributions are lognormal; the log of a lognormal distribution is a normal distribution, so then we can look at the standard deviation of this normal distribution as usual.

      This shows that the health interventions were indeed the most heavy-tailed, though there is still a lot of spread within the other areas.

      DataStandard deviation log10 cost-effectiveness
      Disease Control Priorities Project (2nd edition)0.96
      Disease Control Priorities Project (3rd edition)0.73
      WHO-CHOICE (using average cost-effectiveness)0.85
      WHO-CHOICE (using incremental cost-effectiveness)1.03
      NICE Cost-effectiveness estimates1.13
      Washington State Institute for Public Policy Benefit-Costs Results Database (positive interventions)0.7
      Washington State Institute for Public Policy Benefit-Costs Results Database (negative interventions)0.8
      Washington State Institute for Public Policy Benefit-Costs Results Database (criminal justice reform)0.5
      Washington State Institute for Public Policy Benefit-Costs Results Database (pre-K to 12 education)0.7
      Gillingham and Stock (climate change interventions)0.6
      Education Endowment Foundation Toolkit0.5
      Education Global Practice and Development Research Group Study0.5
      Founders Pledge (personal actions to fight climate change)0.8
      Get-Out-The-Vote tactics0.4

      Log-binned histograms showing distribution shape and full datasets

      We can use histograms to group the data into sections and give an intuitive idea of the distribution shape for each set of data. Because this data spans many orders of magnitude, these histograms are binned such that each bar has equal width on a log scale. In general, these histograms confirm the hypothesis that these distributions have heavy tails — most look qualitatively similar to power law distributions.

      DCP2

      Logarithmic bin histogram of the cost effectiveness of health interventions in developing countries in terms of how many years of illness they prevent, according to data from the DCP2

      Logarithmic binned histogram of the cost effectiveness of health interventions in developing countries in terms of how many years of illness they prevent, according to data from the DCP2.
      Get the data

      DCP3

      Logarithmic binned histogram of the cost effectiveness of health interventions in developing countries in terms of how many years of illness they prevent, according to data from the DCP3

      Logarithmic binned histogram of the cost effectiveness of health interventions in developing countries in terms of how many years of illness they prevent, according to data from the DCP3.

      Get the data

      WHO-CHOICE

      Logarithmic binned histogram of the incremental cost effectiveness of health interventions in developing countries in terms of how many years of illness they prevent, according to data from WHO-CHOICE 2019

      Logarithmic binned histogram of the incremental cost effectiveness of health interventions in developing countries in terms of how many years of illness they prevent, according to data from WHO-CHOICE 2019
      Get the data

      Health in high-income countries: public health interventions in the UK (NICE)

      Logarithmic binned histogram showing the cost-effectiveness of UK public health interventions

      Logarithmic binned histogram showing the cost-effectiveness of UK public health interventions
      Get the data

      Washington State Institute for Public Policy Benefit-Costs Results database

      Positive interventions:
      Log binned histogram showing positive US social policy interventions
      Negative interventions:
      Log binned histogram showing negative US social policy interventions
      Criminal justice reform:
      Log binned histogram showing US criminal justice interventions
      Pre-K to 12 education:
      Log binned histogram showing US education interventions

      Get the data

      Climate change: Gillingham and Stock

      Logarithmic binned histogram of the cost effectiveness of interventions to reduce greenhouse gas emissions, according to data from Gillingham and Stock

      Logarithmic binned histogram of the cost effectiveness of interventions to reduce greenhouse gas emissions, according to data from Gillingham and Stock.

      Get the data

      Education Endowment Foundation Toolkit

      Logarithmic binned histogram of the cost effectiveness of interventions on UK education

      Get the data

      Education Global Practice and Development Research Group Study

      Log binned histogram showing cost effectiveness of interventions on global education

      Get the data

      Founders Pledge: personal actions to fight climate change

      log binned histogram showing cost effectiveness of personal actions to fight climate change

      Get the data

      Get Out the Vote tactics

      log binned histogram showing cost effectiveness of get out the vote tactics

      Get the data

      Read next

      This is a supporting article in our advanced series. Read the next article in the series.

      Notes and references

      1. Jamison, Dean T. et al. Disease Control Priorities In Developing Countries, Second Edition. World Bank And Oxford University Press, 2006, https://openknowledge.worldbank.org/handle/10986/7242.

      2. Ord, Toby, The Moral Imperative towards Cost-Effectiveness in Global Health, Center for Global Development, 2013, https://www.cgdev.org/publication/moral-imperative-toward-cost-effectiveness-global-health.

      3. Open Philanthropy, which was spun out of GiveWell, is 80,000 Hours’ largest funder.

      4. Jamison, Dean T. et al. Disease Control Priorities, Third Edition. World Bank, 2017, https://openknowledge.worldbank.org/handle/10986/28877.

      5. Owen, L. et al. “The Cost-Effectiveness Of Public Health Interventions”. Journal Of Public Health, vol 34, no. 1, 2011, pp. 37-45., https://doi.org/10.1093/pubmed/fdr075.

      6. Benefit-Cost Results. 2019, Washington State Institute for Public Policy, https://www.wsipp.wa.gov/BenefitCost.

      7. If you would like to compare to the figures for health from earlier, you’d need to convert the value of a DALY into dollars. In development economics, it’s common to use figures of $1,000–5,000. The mean cost effectiveness of the DCP2 was 23 DALYs averted per $1,000. So if a DALY is worth $2,500, it would imply its cost-benefit ratio is 57.5.

      8. Gillingham, Kenneth, and James H. Stock. “The Cost Of Reducing Greenhouse Gas Emissions”. Journal Of Economic Perspectives, vol 32, no. 4, 2018, pp. 53-72., https://www.aeaweb.org/articles?id=10.1257/jep.32.4.53.

      9. Angrist, Noam et al. “How To Improve Education Outcomes Most Efficiently? A Comparison Of 150 Interventions Using The New Learning-Adjusted Years Of Schooling Metric”. Policy Research Working Paper; No. 9450. World Bank, 2020, https://openknowledge.worldbank.org/handle/10986/34658.

      10. Green, Donald P., and Alan S. Gerber. Get Out The Vote: How To Increase Voter Turnout. 4th ed., Brookings Institution Press, 2019.

      11. Unfortunately we don’t know how much the mean and median would have been reduced if they were given a similar treatment. The old median was 5 DALYs averted per $1,000, so this would suggest that AMF is now only 2x better than the median, compared to 18x before.
        One thing we do know is that GiveWell estimates their top charities are around 20 times more effective than GiveDirectly, but we don’t know how GiveDirectly compares to the median health intervention, and I wouldn’t be surprised if it were worse.

      12. If we think we can identify the top 10% in expectation, it would imply that by picking carefully we can identify interventions that are at most 10 times better than the mean (which is roughly given by the effectiveness of the top 10% divided by 10, plus a bit more from the rest of the distribution). If we think we can identify the top 1%, then a factor of 100 gain should be possible. Picking the top 10% seems optimistic to me in most areas, especially if we consider the existence of difficult-to-measure interventions, so I think getting more than a factor of 10 boost from intervention selection seems optimistic. (Note that here I’m comparing the mean to the best we’re able to select, rather than the overall ratio between best and worst.)

      13. If you average together a normal distribution and a lognormal distribution, it still has a heavy tail, and extreme events are only reduced in probability by a factor of two compared to a pure lognormal.

      14. Caviola, Lucius, et al. “Donors vastly underestimate differences in charities’ effectiveness.” Judgment and Decision Making, vol. 15, no. 4, 2020, pp. 509-516. Link

        We selected experts in areas relevant to the estimation of global poverty charity effectiveness, in areas such as health economics, international development and charity measurement and evaluation. The experts were identified through searches in published academic literature on global poverty intervention effectiveness and among professional organizations working in charity evaluation.

        We found that their median response was a cost-effectiveness ratio of 100 (see Table 1).

      15. In reality, they’ll use other factors to pick. If we assume these other factors are uncorrelated with measured effectiveness, then that ends up being similar to picking randomly. In real life, the draw is across the interventions that are actually being funded, and it’s unclear how much that distribution reflects the interventions that have been measured — ideally we could weight each intervention by their funding capacity.

      16. If there is a heavy negative tail, this will work in the opposite direction. However, in all the distributions, the negative tail was much smaller than the positive ones, so the positive tail dominates.

      80,000 Hours is looking for a content associate to help us improve and grow the impact of the 80,000 Hours website, which provides free research and advice to help people use their careers to address the world’s most pressing problems.

      As a content associate, you would review 80,000 Hours articles to ensure they are up to date, well presented, and error free, help us collect user feedback, provide analytical support, and generate ideas for the team (more on responsibilities below).

      We’d like to support the person in this role to take on more responsibility over time; one possibility would be to move to become the primary author / researcher of new articles.

      About the 80,000 Hours web team

      80,000 Hours provides free research and support to help people find careers tackling the world’s most pressing problems.

      We’ve had over 10 million visitors to our website (with over 100,000 hours of reading time per year), and more than 3,000 people have told us that they’ve significantly changed their career plans due to our work. We’re also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Community Survey.

      Our articles are read by thousands, and are among the most important ways we help people shift their careers towards higher-impact options.

      The role

      As a content associate, you would:

      • Support the 80,000 Hours web team flexibly across a range of articles and projects.
      • Proofread 80,000 Hours articles before release, suggest style improvements, and check for errors.
      • Upload new articles and make changes to the site.
      • Ensure that our newsletters are sent out error-free and on time to the over 250,000 people on our mailing list.
      • Provide analytical support for the team, improving our ability to use data to measure and increase our impact.
      • Manage the gathering of feedback on our website from both readers and subject matter experts.
      • Generate ideas for new pieces.
      • Generally help grow the impact of the site.

      Some of the types of pieces you could work on include:

      As some indication of what success in the role might look like, over the next year, you might do things like:

      • Provide substantive comments and improvements for new research on various under-explored problem areas — e.g. whole brain emulation, wild animal suffering, or potential existential risks from totalitarianism.
      • Restructure our systems for measuring the impact of the 80,000 Hours website, including both data-driven analysis and interviews of users and experts.
      • Help us streamline our writing and editing workflows.
      • Work on additional projects. We’re excited to help you double down on your strengths, based on 80,000 Hours’ needs and your personal fit. Examples of additional projects might include writing a blog post or short career review in an area of interest to you, analysing the results of our user survey, or supporting the annual review, impact evaluation, or planning processes for the web team.

      Who we’re looking for

      We’re looking for someone who has:

      • An interest in effective altruism and/or longtermism.
      • An aptitude for editing and writing.
      • The ability to learn quickly and independently.
      • Excitement to work on whatever kinds of projects are highest priority.
      • The ability to be organised, detail-oriented, and able to keep track of a large number of tasks at once.
      • A good fit with our cultural values.

      We’re aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all the suggested attributes. We’d especially like to encourage people from underrepresented backgrounds to apply!

      You don’t need any previous experience to apply. In fact, we’d encourage you to apply even if you’re not sure you meet all of the above criteria – we’d much prefer to hear from you than not!

      Details of the role

      This is a full-time role. The salary will vary based on experience, but to give a rough sense, the starting salary for someone with no relevant experience would be approximately £58,000 per year.

      We generally prefer people to work in-person in our London office if possible, but are open to remote work in this case. We can sponsor visas.

      The start date of the role is flexible, but we would expect you to start during the first half of 2023 and prefer you to start as soon as you’re available.

      Our benefits include:

      • The option to use 10% of your time for self development
      • 25 days of paid holiday, plus bank holidays
      • Standard UK pension with 3% contribution from employer
      • Private medical insurance
      • Generous parental leave
      • Long-term disability insurance
      • Flexible work hours
      • Gym, shower facilities, and free food provided at our London office

      How to apply

      To apply, please fill out this application form by 9am GMT on Tuesday, 14 February, 2023.

      We expect it’ll take most people under 45 minutes. If you have any problems submitting the form, please send your CV to [email protected].

      The application process will vary depending on the candidate, but is likely to include a short work test, an interview, and a multi-day in-person trial. The work test and trial will be paid.

      Click here to apply

        Applications for this position have now closed.

        Summary

        We’re hiring a recruiter to help us grow the 80,000 Hours team.

        Not being able to hire fast enough is one of our biggest bottlenecks as an organisation. The person in this role will directly address this by helping us to source candidates, run hiring rounds, and scale our recruitment processes as we grow. They’ll be key to increasing 80,000 Hours’ impact over the coming years.

        You might be a great fit if you:

        • Have a strong understanding of effective altruism / longtermism.
        • Have excellent organisational / project management skills.
        • Are a clear communicator, both in writing and in person.
        • Would enjoy interacting with and evaluating people.

        This is a full-time role. Ideally, you’d be based at our London office, but we’re open to remote candidates (as long as you’re able to work within UK working hours).

        We’re willing to consider candidates who are available to start anytime between December 2022 and August 2023 — so please consider applying even if you’re not available immediately.

        The starting salary for someone with one year of highly relevant experience is £59,200 per year.

        The role

        You’ll be managed by Sashika Coxhead, our Head of Recruiting, and will have the opportunity to work closely with hiring managers from other teams.

        Initial responsibilities will include:

        • Project management of active recruiting rounds. For example, overseeing the candidate pipeline and logistics of hiring rounds, making decisions on initial applications, and managing candidate communications.
        • Sourcing potential candidates. This might include generating leads for specific roles, publicising new positions, reaching out to potential candidates, and answering any questions they have about working at 80,000 Hours.
        • Taking on special projects to improve our recruiting systems. For example, you might help to build an excellent applicant tracking system, test ways to improve our ability to generate leads, or introduce strategies to make our hiring rounds more efficient.

        Depending on your skills and interests, you might also:

        • Take ownership of a particular area of our recruiting process, e.g. proactive outreach to potential candidates, our applicant tracking system, or metrics for the recruiting team’s success.
        • Conduct screening interviews where needed, to assess applicants’ fit for particular roles at 80,000 Hours.

        After some time in the role, we’d hope for you to sit on internal hiring committees. This involves forming an inside view on candidates’ performance; discussing uncertainties with the hiring manager and committee; and, with the other committee members, giving final approval on who to make offers to.

        Who we’re looking for

        You might be a great fit if you:

        • Have a strong interest in and understanding of effective altruism and longtermism. Ideally, you have evidence of previous involvement with EA (e.g. being a member or organiser of a local or university group, attending a previous EAG or EAGx conference, volunteering or working with an EA organisation) and have a deep understanding of key EA concepts.
        • Are highly conscientious and organised. You take pride in your attention to detail, can keep track of multiple moving parts within a project, and don’t let things fall through the cracks. Ideally, you have previous experience successfully managing projects (e.g. running events or doing local/university group organising).
        • Are a clear communicator, both in writing and in person. You’re able to discuss uncertainties with hiring managers efficiently and can communicate clearly but sensitively with candidates.
        • Are interested in thinking about people. You have strong interpersonal skills, love trying to develop accurate models of others, and would enjoy thinking about their fit for roles at 80,000 Hours.

        You don’t need any previous experience with recruiting to apply. In fact, we’d encourage you to apply even if you’re not sure you meet all of the above criteria – we’d much prefer to hear from you than not!

        We’re aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all of the suggested attributes. We’d especially like to encourage people from underrepresented backgrounds to express interest in this role, including if you don’t meet all the suggested criteria.

        Role details

        This is a full-time role. Ideally, you’d be based at our London office, but we’re open to remote candidates (as long as you’re able to work within UK working hours). We are able to sponsor visa applications if required.

        We’re willing to consider candidates who are available to start anytime between December 2022 and August 2023 — so please consider applying even if you’re not available immediately.

        The salary will depend on your previous experience, but to give a rough sense, the starting salary for someone with one year of highly relevant experience would be £59,200 per year.

        Our benefits include:

        • The option to use 10% of your time for self development.
        • 25 days of paid holiday, plus bank holidays.
        • Standard UK pension with 3% contribution from employer.
        • £5,000 mental health support allowance.
        • Private medical insurance.
        • Generous paid parental leave.
        • Long-term disability insurance.
        • Gym, shower facilities, and free food provided at our London office.

        Application process

        To apply, please complete this application form by 11pm GMT on Wednesday, November 2 2022.

        We’re reviewing applications on a rolling basis, so we encourage you to apply as soon as you are able to.

        The application process will vary depending on the candidate, but will likely involve:

        • 1–3 (paid) written work samples
        • A ~45-minute interview
        • A ~3-day in-person trial

          Applications to this position are now closed.

          Summary

          80,000 Hours is considering hiring a Head of Operations to oversee our internal operations as we scale. We’re looking for someone who has:

          • At least two years of experience in operations/organisational management-focused roles, including roles in consulting, project management, and at early-stage startups or nonprofits.
          • At least one year of experience with managing others.
          • A strong understanding of effective altruism and/or longtermism.

          If you might be interested, please submit this expression of interest — it only takes about two minutes to complete!

          Because this isn’t a position we’re actively hiring for, we have a higher bar than usual for inviting candidates to the next stage of the process, and won’t be able to respond to all enquiries we receive. If you don’t hear from us, you’re still very welcome to apply for future roles at 80,000 Hours.

          80,000 Hours

          80,000 Hours provides research and support to help people switch into careers that effectively tackle the world’s most pressing problems.

          We’ve had over 8 million visitors to our website, and more than 3,000 people have told us that they’ve significantly changed their career plans due to our work. We’re also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Survey.

          The internal systems team

          This role is on the internal systems team, which is here to build the organisation and systems that support 80,000 Hours to achieve its mission.

          We oversee 80,000 Hours’ office, tech systems, organisation-wide metrics and impact evaluation, as well as HR, recruiting, finances, and much of our fundraising.

          Currently, we have four full-time staff, some part-time staff, and receive support from the Centre for Effective Altruism (our fiscal sponsor).

          The role

          As 80,000 Hours’ Head of Operations, you would:

          • Oversee a wide range of our internal operations, including team-wide processes, much of our fundraising, our office, finances, tech systems, data practices, and external relations.
          • Manage a team of two operations specialists, including investing in their professional development and identifying opportunities for advancement where appropriate.
          • Grow your team to build capacity in the areas you oversee, including identifying 80,000 Hours’ operational needs and designing roles that will address these.
          • Develop our internal operations strategy — in particular, figure out what your team should focus on in order to add the most value to 80,000 Hours as we grow.
          • Take on additional special projects — for example, this year, you might have run the longtermist census (which received ~3,000 responses) or scaled up our book giveaways with Impact Books (who send out over 100 books per day on our behalf).

          You would be managed by Brenton Mayer, our Director of Internal Systems.

          Who we’re looking for

          We’d be most excited to hear from people with all of the following:

          • At least two years of experience in operations/organisational management-focused roles, including roles in consulting, project management, and at early-stage startups or nonprofits.
          • At least one year of experience with managing others.
          • A strong understanding of effective altruism and/or longtermism.

          In addition to this, we’re looking for someone who is:

          • An excellent manager, who’s excited to coach and develop the skills of those on their team.
          • Able to think critically about their team’s strategy, and consistently prioritise well in order to focus on the most important things.
          • Able to generate ambitious plans for addressing 80,000 Hours’ needs as we scale, and execute them successfully.
          • Highly conscientious, with the ability to juggle multiple priorities at once without dropping balls.
          • An optimiser, who constantly seeks out ways to improve our systems and processes so that we can run more efficiently.

          We’re aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all the suggested attributes. We’d especially like to encourage people from underrepresented backgrounds to express interest in this role, including if you don’t meet all the suggested criteria.

          Salary and benefits

          This is a full-time role, based at our London office (but you can work remotely for up to one third of the year if needed). We are able to support visa applications.

          The salary will depend on your skills and experience, but to give a rough sense, the starting salary for someone with three years of highly relevant experience would be £68,814 per year.

          Our benefits include:

          • The option to use 10% of your time for self development.
          • 25 days of paid holiday, plus bank holidays.
          • Standard UK pension with 3% contribution from employer.
          • £5,000 mental health support allowance.
          • Private medical insurance.
          • Generous paid parental leave.
          • Long-term disability insurance.
          • Gym, shower facilities, and free food provided at our London office.

          Application process

          To apply, please submit this expression of interest by 11pm GMT on Wednesday, August 17, 2022. It should only take about two minutes to complete! We’ll reach out if we think you might be a particularly strong fit for the role.

            Applications for this position are now closed.

            We’re looking for a new marketer to help us expand our readership and scale up our marketing channels.

            We’d like to support the person in this role to take on more responsibility over time as we expand our marketing team.

            80,000 Hours provides free research and support to help people find careers tackling the world’s most pressing problems.

            We’ve had over 8 million visitors to our website, and more than 3,000 people have told us that they’ve significantly changed their career plans due to our work. We’re also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Survey.

            Even so, about 90% of US college graduates have never heard of effective altruism, and we estimate that just 0.5% of students at top colleges are highly engaged in EA. As a marketer with 80,000 Hours, you would help us achieve our goal of reaching all students and recent graduates who might be interested in our work. We anticipate that the right person in this role could help us grow our readership to 5–10 times its current size, and lead to hundreds or thousands of additional people pursuing high-impact careers.

            We’re looking for a marketing generalist who will:

            • Start managing (and eventually own) our two largest existing marketing channels:
              • Sponsorships with people who have large audiences, primarily on YouTube (influencer marketing).
              • Paid advertisements on Facebook and Instagram (digital marketing).
            • Take on management of some other existing marketing efforts, such as promoting The 80,000 Hours Podcast on podcast listening platforms, and managing the promotion of our book giveaway.
            • Be responsible for evaluation and analytics of your marketing efforts, such as comparing conversion rates, cost per acquisition, and analysing the engagement levels of new readers.
            • Write for and design relevant pages on our website, such as landing pages for marketing campaigns.

            Depending on your background and interests, we might also like you to:

            • Help us manage our Google Ads nonprofit grant, or expand into paid search and/or YouTube ads on the Google Ads platform.
            • Work on conversion rate optimisation of the website, such as by running A/B tests on alternative calls to action.
            • Expand into other paid digital advertisement platforms, such as on LinkedIn, Twitter, TikTok, or Snap.
            • Manage our social media accounts.
            • Help the content team with search engine optimisation of our website content.

            Your focus in this role will be on our existing channels to begin with (sponsorships and digital advertisements). However, we would be excited to support you towards eventually taking on more responsibility, including:

            • Providing input on our marketing strategy — for example, generating ideas for major marketing initiatives, discussing which to pursue, and figuring out which metrics we should optimise to most effectively achieve our goals.
            • Investigating and launching new marketing initiatives or partnerships as a primary decision maker.

            Bella Forristal would be your manager. You would be the second marketer hired to the team, which we intend to grow rapidly.

            As some indication of what success in the role might look like, over the next three years you might have:

            • Cost-effectively deployed >$5 million reaching people from our target audience.
            • Worked with some of the largest and most respected YouTube creators (for instance, we have existing contacts with Tom Scott, SciShow (Hank Green), and Wendover Productions).
            • Managed Facebook and Instagram ad campaigns that reached tens of millions of people.
            • Driven tens or hundreds of thousands of additional newsletter subscriptions, leading to hundreds or thousands of people changing to a more impactful career.
            • Expanded your responsibility to include other marketing channels.

            We’re looking for someone who has:

            • A strong interest in effective altruism and longtermism, ideally with experience in EA strategy.
            • An enthusiastic approach to the role; you’re excited about 80,000 Hours’ mission and growing our reach.
            • Excellent written communication (in particular, you’re comfortable discussing decisions and uncertainties with the rest of the team in writing).
            • An interest in thinking carefully about what will drive engagement with our work from people who might make especially high-impact career changes, and what this means for our marketing strategy.

            Ideally, you’d also have the following traits — but we encourage you to apply even if they don’t describe you!

            • Previous experience in marketing, especially influencer or digital marketing, or a related field (this might include things like product management, software engineering, data science, operations, or communications; or maybe you’ve worked on a side project that attracted a large number of users).

            Since we are a nonprofit, and we aren’t selling a product, this is a fairly nontraditional marketing role. We’d therefore encourage you to apply, even if you aren’t otherwise looking for roles in marketing.

            This is a full-time role. We would prefer for you to work in-person, based in London (we can support UK visa applications if needed). You can work remotely for up to three months of the year if needed.

            The salary will vary based on your skills and experience, but to give a rough sense, the starting salary for someone with no relevant prior experience would be approximately £58,000 per year; for someone with four years of relevant prior experience it would be approximately £70,000 per year.

            Our benefits include:

            • The option to use 10% of your time for self development
            • 25 days of paid holiday, plus bank holidays
            • Standard UK pension, with 3% contribution from employer
            • Flexible work hours and location
            • Private medical insurance
            • Long-term disability insurance
            • Gym, shower facilities, and free food provided at our London office

            We have a really awesome team and are excited for more people to join us in our mission to help people use their careers to solve the world’s most pressing problems.

            We’re aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all the suggested attributes. We’d especially like to encourage people from under-represented backgrounds to apply.

            To apply, please fill in this application form. If you have any problems submitting the form, please send your CV to [email protected]. Applications are due by 23 August 2022.

            Note: This is a role which does not require prior experience. However, if you’re a more senior marketer who is enthusiastic about 80,000 Hours’ mission, we’d still be really excited to talk to you about whether you might be able to help. Please apply above or email [email protected].

              Applications to this position are now closed.

              80,000 Hours is looking for full-time staff writers to publish well-researched articles to help people use their careers to help solve the world’s most pressing problems.

              About the 80,000 Hours web team

              80,000 Hours provides free research and support to help people find careers tackling the world’s most pressing problems.

              We’ve had over 8 million visitors to our website (with over 100,000 hours of reading time per year), and more than 3,000 people have told us that they’ve significantly changed their career plans due to our work. We’re also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Community Survey.

              Our articles are read by thousands, and are among the most important ways we help people shift their careers towards higher-impact options.

              The role

              As a writer, you would:

              • Research, outline, and write new articles for the 80,000 Hours website — e.g. new career reviews.
              • Rewrite or update older articles with information and resources — e.g. about rapidly evolving global problems.
              • Generate ideas for new pieces.
              • Talk to experts and readers to help prioritise our new articles and updates.
              • Generally help grow the impact of the site.

              Some of the types of pieces you could work on include:

              Which of these you’ll focus on will depend to some extent on your strengths and interests, as well as the needs of our audience. But as some indication of what success in the role might look like, over the next year, you might do things like:

              • Research and write the most comprehensive and widely read assessment of an under-explored problem area — e.g. whole brain emulation, global surveillance, or potential existential risks from totalitarianism.
              • Revamp our career planning course into a MOOC or interactive guide.
              • Write a popular overview of high-impact career advice by major for undergraduates.

              Who we’re looking for

              We’re looking for someone who has:

              • Strong knowledge of and commitment to the ideals, philosophical arguments, and values of effective altruism and longtermism.
              • Great communication skills.
              • An aptitude for research and writing.
              • The ability to learn quickly and independently.
              • Excitement to work on whatever kinds of projects are highest priority.
              • A good fit with our cultural values.
              • Ideally, a track record of producing high-quality writing. This can include things like blog posts, articles, reports, and substantive Facebook posts — especially for broad audiences, and especially on topics related to effective altruism and longtermism. But even if you don’t already have a track record, we’d still encourage you to apply if you think you might be a good fit for this role.

              We’re aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all the suggested attributes. We’d especially like to encourage people from underrepresented backgrounds to apply!

              Details of the role

              This is a full-time role. The salary will vary based on experience, but to give a rough sense, the starting salary for someone with one year of relevant experience would be approximately £61,000 per year.

              We strongly prefer people to work in-person in our London office if possible, but are open to remote work in some cases. We can sponsor visas.

              The start date of the role is flexible, but we would expect you to start during 2022 and prefer you to start as soon as you’re available.

              Our benefits include:

              • The option to use 10% of your time for self development
              • 25 days of paid holiday, plus bank holidays
              • Standard UK pension with 3% contribution from employer
              • Private medical insurance
              • Generous parental leave
              • Long-term disability insurance
              • Flexible working hours
              • Gym, shower facilities, and free food provided at our London office

              How to apply

              To apply, please fill out this application form by 9am GMT on Monday, May 16, 2022.

              We expect it’ll take most people under 45 minutes. If you have any problems submitting the form, please send your CV to [email protected].

              The application process will vary depending on the candidate, but is likely to include 1-3 written work samples, an interview, and a multi-day in-person trial. The work samples and trial will be paid.

              Check out other positions at 80,000 Hours

                Applications to this position are now closed.

                Summary

                80,000 Hours’ mission is to get talented people working on the world’s most pressing problems.

                We expect to increase our staff count by ~50% in 2022, and to continue scaling from there. We’re looking for an operations specialist to help build the organisation and systems that will support this growth.

                They’ll run the team that manages our office, help scale our internal processes as we grow, and oversee a range of other projects according to their personal fit and our needs as an organisation (e.g. helping to analyse our user survey or organise our team retreat).

                Location: London, England.

                Salary: ~£58,400 for someone with one year of relevant experience (higher for candidates with more experience).

                To apply, please complete this application form by 11pm GMT on Sunday, April 3, 2022.

                80,000 Hours

                80,000 Hours provides research and support to help students and graduates switch into careers that effectively tackle the world’s most pressing problems.

                Over one million people visit our website each year, and more than 3,000 people have told us that they’ve significantly changed their career plans due to our work. We’re also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Survey.

                The Internal Systems team

                The Internal Systems team is here to build the organisation and systems that support 80,000 Hours to achieve its mission.

                We oversee 80,000 Hours’ office, finances, and impact evaluation, as well as much of our fundraising, org-wide metrics, tech systems, HR, and recruiting.

                Currently, we have two full-time staff (Brenton Mayer and Sashika Coxhead), some part-time staff, and receive support from CEA (our fiscal sponsor).

                Role

                This role would be excellent experience for someone who wants to build career capital in operations, especially if you could one day see yourself in a more senior operations role (e.g. taking on more management, and perhaps eventually being a Head of Operations or COO).

                Your responsibilities will likely include:

                • Creating an outstanding office environment. You’ll hire and manage the team that oversees our beautiful central London office. Your team will be responsible for all the systems that keep the office running smoothly, as well redesigning them to accommodate larger numbers of staff (e.g. reconfiguring our office layout to allow for a 50% increase in staffing this year).

                • Managing our operations associate. You’ll be responsible for managing, mentoring, and retaining our part-time operations associate. Our associate is currently focused on the day-to-day running of the office, but will be able to work on other responsibilities delegated by you.

                • Scaling various internal systems. You’ll figure out how to scale our internal processes as we grow, as well as notice opportunities for improvement and implement those that are highest priority. Examples might include:

                  • Designing a system that enables the team to quickly find the documents they need.
                  • Improving our use of internal communication channels.
                  • Figuring out how 80,000 Hours can ensure it has accounts on new social media platforms which might become critical for spreading our ideas.
                  • Improving our system for tracking quarterly team goals.

                • Work on additional projects. We’re excited to help you double down on your strengths, based on 80,000 Hours’ needs and your personal fit. Examples of additional projects might include running a search for a larger office, analysing the results of our user survey, or helping to organise our team retreat.

                About you

                We don’t have any specific requirements for this role, but you might be a great fit if you:

                • Are excited about joining a small, fast-growing team and helping us to design and build the systems we need to scale.
                • Are organised, detail-oriented, and able to keep track of a large number of tasks at once.
                • Have excellent written and oral communication skills — in particular, you demonstrate strong reasoning transparency and enable others to easily identify the most important aspects of their communications.
                • Love to build systems that run exceptionally smoothly, and have promising ideas for how we can improve upon our current processes.
                • Have good judgement: you’re able to weigh up tradeoffs between different options and make real-world decisions under uncertainty.
                • Have strong interpersonal skills and are excited to manage others.
                • Have an interest in effective altruism and/or longtermism.

                Experience in operations, project management, or community building would be a bonus. However, we’re more excited about your potential for growth than previous experience, so a lack of experience shouldn’t discourage you from applying.

                We’re also aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all the suggested attributes. We’d especially like to encourage those from underrepresented backgrounds to apply.

                Role details

                This role is full-time and in-person, based in London, England. We are able to sponsor visas for this position.

                You’ll be managed by the Head of Operations (a position we’re currently hiring for).

                The salary will vary based on skills and experience, but to give a rough sense, the starting salary for someone with one year of relevant experience would be £58,400 per year.

                Our benefits include:

                • The option to use 10% of your time for self-development
                • 25 days of paid holiday, plus bank holidays
                • Private medical insurance
                • £5,000 annual mental health support allowance
                • Long-term disability insurance
                • Standard UK pension with 3% contribution from employer
                • Shower facilities, a small gym, and unlimited free food provided at our London office

                Application process

                To apply, please complete this application form by 11pm GMT on Sunday, April 3, 2022.

                The application process will vary depending on the candidate, but is likely to include 1-3 written work samples, an interview, and a multi-day in-person trial. The work samples and trial will be paid.

                  80,000 Hours is considering hiring a writer.

                  • As a writer at 80,000 Hours, you’d research and write articles read by thousands of people trying to do good with their careers.
                  • We’re looking for someone with a preexisting understanding of our organisation’s priorities (including longtermism and effective altruism), plus great communication skills, an aptitude for research and writing, and the ability to learn quickly.
                  • This is a full-time, London-based role, with a starting salary of around £58,000–£68,000.

                  Note: This announcement is an expression of interest, rather than a formal hiring round. We’ll likely launch a formal round soon, but aren’t sure exactly when. We think learning more about who might be interested in this role will help us better plan and expedite our hiring round when we do run it. Because of this, we have a high bar for considering enquiries, and won’t be able to respond to everyone who reaches out.

                  If you don’t hear back from us, please don’t take it as a rejection! You should feel very welcome to respond to future ads for 80,000 Hours positions (which we’ll list here and on our job board).

                  Why 80,000 Hours?

                  80,000 Hours’ mission is to get talented people working on the world’s most pressing problems. The effective altruism community, of which we are a part, is growing in reach and now includes funding bodies with over $40 billion to allocate in total. But how do we make sure people are pursuing the right kinds of work in order to turn all those resources into long-term impact? This is the problem 80,000 Hours is trying to solve.

                  We’ve had over eight million visitors to our website (with over 100,000 hours of reading time per year), and more than 3,000 people have now told us that they’ve significantly changed their career plans due to our work. 80,000 Hours is also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Survey.

                  If you join us as a writer, you’d likely be one of the most widely read writers in effective altruism.

                  The role

                  As a writer at 80,000 Hours, your work would involve:

                  • Framing, researching, outlining, and writing articles
                  • Generating ideas for additional articles
                  • Helping with others’ writing by providing comments
                  • Generally helping grow the impact of the site

                  Some of the types of pieces you’d help work on include:

                  Who we’re looking for

                  We’d be most excited to hear from people with:

                  • Strong knowledge of and commitment to the ideals, philosophical arguments, and values of effective altruism
                  • High conscientiousness, and the ability to learn quickly and independently
                  • A track record of producing high-quality writing (which can include blog posts, articles, reports, etc.) — especially for broad audiences, and especially on topics related to effective altruism and longtermism. (Don’t have this, but think you have the potential to produce writing for a broad audience? We’d still like to hear from you!)

                  We’re aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all the suggested attributes. We’d especially like to encourage people from underrepresented backgrounds to express interest.

                  The details

                  This is a full-time role based in London, though you can work remotely for up to three months of the year if needed. If that won’t work for you, please reach out anyway (and let us know in your email). We can sponsor visas.

                  The salary will vary based on your skills and experience, but to give a rough sense, the starting full-time salary would likely be between £58,000 and £68,000 per year.

                  The start date of the role is flexible, but we would expect you to start during 2022 and prefer you to start as soon as you’re able to.

                  Our benefits include:

                  • The option to use 10% of your work time for self development
                  • 25 days of paid holiday, plus bank holidays
                  • Standard UK pension, with 3% contribution from employer
                  • Private medical insurance and long-term disability insurance
                  • Gym, shower facilities, and free food provided at our London office

                  Interested? Get in touch.

                  If you’re a writer with a track record and an interest in 80,000 Hours’ mission, please submit the following materials:

                  • A brief explanation (under 2,000 characters) of your interest in 80,000 Hours’ mission
                  • One research writeup (broadly defined) written to be accessible to a non-expert (can be unpublished)
                  • One writing sample written for a broad audience (can be research-based, or not)

                  Submit materials

                  While there’s no hard deadline for getting in touch, we encourage you to submit your materials before mid-April. We’ll reach out to discuss a potential role if we think you’re a particularly great fit.

                  Check out other positions at 80,000 Hours

                    Applications to this position are now closed.

                    Summary

                    80,000 Hours is hiring a Head of Job Board to lead the job board. They will be responsible for setting and executing strategy to grow the job board’s impact, as well as managing and hiring the job board team.

                    More than 180,000 users visited the job board in 2021. Over the next few years, we hope to grow the job board to the point where millions of people per year use it to find out about impactful jobs.

                    This role is based in London, UK. The salary will vary based on your skills and experience, but the starting salary for someone with five years of relevant experience would be approximately £72,000 per year.

                    To apply for this role, please complete this application form by 11pm GMT on Sunday, 27 February 2022.

                    We are offering a £1000 referral bonus to anyone outside the Centre for Effective Altruism who suggests a successful candidate we didn’t otherwise have on our radar. Please email [email protected] with your referrals.

                    80,000 Hours

                    80,000 Hours’ mission is to get talented people working on the world’s most pressing problems. We’re a part of the effective altruism community, which is growing in reach and now includes funding bodies with over $40 billion to allocate in total. But how do we turn all those resources into long-term impact? This is the problem 80,000 Hours is trying to solve.

                    Over one million people visit the 80,000 Hours website each year, and more than 3,000 people have told us that they’ve significantly changed their career plans due to our work. We’re also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Survey.

                    Job board

                    The 80,000 Hours job board enables readers to find vacancies that we believe will help them contribute to the world’s most pressing problems. We include the most impactful jobs we know about, as well as positions that we think would help our users build valuable career capital to work on these problems later in their careers.

                    The job board serves two functions. Its primary purpose is to help readers translate our career advice into specific career moves. However, it’s also the second most common way users enter the site (after the homepage).

                    In total, users spend about 670 hours per month browsing the 80,000 Hours job board. Overall, more time is spent on the job board than on any other single page on our website.

                    After reviewing the summaries of roles on our site, users click through to hiring organisations’ original job ads for more information over 10,000 times per month.

                    The job board is growing. In 2021, we drove 51% more clicks through to hiring organisations’ original job ads than we did in 2020. In 2020, we drove 32% more clicks through than in 2019.

                    We are now looking to considerably increase our investment in the job board. We think we can grow the product substantially into the internet’s top job board for people ambitiously focusing their career on doing good. More than 180,000 users visited the job board in 2021. Over the next few years, we hope to grow the job board to the point where millions of people per year use it to find out about impactful jobs.

                    The role

                    The ‘Head of Job Board’ will be responsible for setting and executing strategy to grow the job board’s impact, as well as managing and hiring the job board team.

                    We are looking for someone who can lead all elements of the job board, from setting product strategy to diving into the details of how we collect vacancies from hundreds of sources, condense them into job board updates, and send those updates to over 100,000 weekly newsletter subscribers.

                    The person in this role would manage our Curator, who is responsible for working with our outsource team to coordinate the weekly aggregation and curation of our job board updates and sending our weekly email updates. The Head of Job Board would also work closely with developers on the web team. There is scope to make additional technical and non-technical hires to the job board team to expand our capacity, in collaboration with the relevant Director and our Chief of Staff.

                    This role reports to Niel Bowerman, Director of One-on-One Programme and Job Board.

                    Responsibilities

                    1. Do whatever is needed to make the job board the most impactful product it can be.
                    2. Lead on setting the high-level strategy for the job board, including metrics and annual and quarterly goals, in collaboration with the relevant Director, the Chief of Staff, and the CEO.
                    3. Manage the job board team and hire to the team as appropriate.
                    4. Develop and execute the product strategy you set for the job board, including setting and implementing policies on what jobs should be listed, what features to build in what order, UI and UX design, and new content to produce.
                    5. Alongside the marketing team, promote the job board externally and on the 80,000 Hours site to grow its engagement and impact.
                    6. Talk with users, run experiments, and do analysis and activities required to understand how the job board impacts our users and the wider world.
                    7. Oversee internal processes to ensure that the job board is updated regularly, including troubleshooting and taking on temporary responsibilities when necessary.
                    8. Stay in sync with 80,000 Hours’ Directors and be the primary point of contact internally and externally on the job board.

                    Experience and skills

                    Essential

                    1. A strong interest in effective altruism and longtermism.
                    2. Ability to develop and execute data- and user-driven product strategy.
                    3. Ability to think carefully about how to use the job board to enable our target audience to make especially high-impact role changes, and apply this to our product strategy.
                    4. An ambitious approach to the role, with enthusiasm for generating new ideas for how we can increase the impact of the job board.
                    5. Reasoning transparency: the ability to explain your reasoning clearly to others on the team, both in writing and orally.
                    6. Comfort using modern cloud software such as Google Docs, Google Sheets, Airtable, Asana, Mixpanel, and Google Analytics.

                    Experience with any of the following could also be valuable for someone in this role, but are not required:

                    1. Building and managing a team.
                    2. Product management.
                    3. Experience in any of the following relevant areas:
                      1. Growth marketing
                      2. User experience
                      3. Web design/development
                      4. Product design
                      5. Data analytics
                      6. Managing outsource teams and designing robust, scalable processes
                    4. Thinking about EA movement-building strategy questions, such as which roles to direct talent towards on the current margin.
                    5. Analysing web product data and a quantitative approach to problem solving.

                    We don’t expect our eventual hire to have most of these attributes, so please don’t let that discourage you from applying. We’d especially like to encourage people from under-represented backgrounds to apply.

                    Role details

                    This is a full-time, in-person role, based in London. You can work remotely for up to three months of the year if needed.

                    The salary will vary based on your skills and experience, but to give a rough sense, the starting salary for someone with five years of relevant experience would be approximately £72,000 per year.

                    Our benefits include:

                    • The option to use 10% of your time for self development
                    • 25 days of paid holiday, plus bank holidays
                    • Standard UK pension with 3% contribution from employer
                    • Private medical insurance
                    • Long-term disability insurance
                    • Shower facilities, a small gym, and free food provided at our London office

                    Referrals

                    We are offering a £1000 referral bonus to anyone outside the Centre for Effective Altruism who suggests a successful candidate we didn’t otherwise have on our radar. Please email [email protected] with your referrals.

                    Application process

                    To apply for this role, please complete this application form by 11pm GMT on Sunday, 27 February 2022.

                    Applicants selected for further consideration will progress through the following stages:

                    • Work test (1–4 hours, remote)
                    • Interview
                    • Trial (only for candidates with a >50% chance of an offer to; 2–5 days; ideally in person)

                    The work test and trial will be paid.

                    If you have any questions, please email [email protected].

                    APPLY

                      Applications to this position are now closed.

                      We’re looking for new colleagues to join our team of advisors.

                      • Our advisors talk one-on-one to talented and altruistic applicants in order to help them find the highest impact career they can.
                      • We’ve found that experience with coaching is not necessary – everything from management consulting to global priorities research has helped someone be a good fit.
                      • London-based role with starting salary around £65,000.

                      80,000 Hours’ mission

                      80,000 Hours’ mission is to get talented people working on the world’s most pressing problems. The effective altruism community, of which we are a part, is growing in reach and now includes funding bodies with over $40 billion to allocate in total. But how do we turn all those resources into long-term impact? This is the problem 80,000 Hours is trying to solve.

                      We’ve had over 8 million visitors to our website, and more than 3,000 people have now told us that they’ve significantly changed their career plans due to our work. 80,000 Hours is also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Survey.

                      The 1on1 team

                      The 1on1 team at 80,000 Hours takes people from “interested in the ideas and want to help” to “actually working to solve pressing world problems.” For example, Sophie Rose applied for advising in 2019. We helped her decide to focus on biosecurity and start working in the field. She co-founded One Day Sooner and is now working at the Johns Hopkins Center for Health Security – doing research to help prevent catastrophic pandemics. In 2021 we had more than 2,000 applications to speak with us from talented and altruistic people who aren’t sure how to use their skills to help the world.

                      Some of the most important ways our conversations help people increase their impact are:

                      • Introductions to experts in relevant fields as well as hiring managers.
                      • Talking through and reframing decisions in order to pin down people’s key uncertainties.
                      • Suggesting new ideas — whether that’s specific jobs, new paths to explore, or ways of getting funding.

                      We also help with recruitment for high impact organisations, sending them lists of people we think might be a great fit for them.

                      While our impact per call is high, the world needs far more people in total working on the world’s most pressing problems. We therefore want to greatly increase the number of calls we have by doubling our team of advisors over the next year. Our aim is to build out an excellent system for finding the best candidates for the most impactful jobs as they are created.

                      What we’re looking for

                      It’s a great sign you’d enjoy being an 80,000 Hours advisor if you’ve enjoyed managing, mentoring or teaching. For example, Habiba Islam joined us from being a manager at the Future of Humanity Institute, Alex Lawsen was a Maths teacher, and Matt Reardon was previously a corporate lawyer.

                      But it’s also particularly useful for us to have a broad range of experience on the team for comparing different types of roles, so we’re excited to hear about people with all kinds of backgrounds.

                      The core of this role is having one-on-one conversations with people to help them plan their careers. We have a tight-knit, fast-paced team, though, so people take on a variety of things. They include, for example, reviewing applications to speak with us, pieces of analysis to improve the service, and writing articles for the 80,000 Hours site or the EA Forum.

                      We’re looking for someone who has:

                      • A strong interest in effective altruism and longtermism, ideally with some experience in EA strategy.
                      • Strong analytical skills and enjoys puzzling to figure out the key considerations in complex problems.
                      • A deep interest in understanding people, and who would enjoy having large numbers of one on one conversations via video call.

                      Previous experience in one of our priority areas would be a significant advantage in the role, but we encourage you to apply even without that.

                      We’re aware that factors like gender, race, and socioeconomic background can affect people’s willingness to apply for roles for which they meet many but not all the suggested attributes. We’d especially like to encourage people from under-represented backgrounds to apply.

                      Details of the job

                      We expect this to be a full-time, in-person role, based in London. We can sponsor visas. You can work remotely for up to three months of the year if needed. If that won’t work for you, please let us know in your application.

                      The salary will vary based on your skills and experience, but to give a rough sense, the starting salary for someone with five years of relevant experience would be approximately £65,000 per year.

                      Start date of the role is flexible, but we would expect you to start during 2022 and prefer you to start as soon as you’re able to.

                      Our benefits include:

                      • The option to use 10% of your work time for self development
                      • 25 days of paid holiday, plus bank holidays
                      • Standard U.K. pension, with 3% contribution from employer
                      • Private medical insurance and long-term disability insurance
                      • Gym, shower facilities, and free food provided at our London office

                      Where to from here?

                      If you’re not interested in the Advisor role, but would be interested to join the 1on1 team in some other capacity (for example technical support), please do get in touch.

                      Candidates will progress through the following:

                      • Interview
                      • ~2 hour work test
                      • 2-5 day in-person (if possible) trial
                      • Job offer

                      The work test and trial will be paid.

                        Scared Straight was a government programme that received billions of dollars of funding, and was profiled in an award-winning documentary. The idea was to take kids who committed crimes, show them life in jail, and scare them into embracing the straight and narrow.

                        The only problem? One meta-analysis found the programme made the kids more likely to commit crimes, and another more recent meta-analysis found it had no effect.1

                        Causing this much harm is rare, but when social programmes are rigorously tested, a large fraction of them don’t work.2

                        So, even if you’ve chosen a pressing issue, it would be easy to end up working on a solution to it that has very little impact.

                        Meanwhile, research finds that among solutions that do have positive effects, the best interventions within an area are far more cost effective than average — often achieving over 10 and sometimes 100 times as much for a given unit of resources.

                        Taking a career that lets you work on more effective solutions is one way to find a greater opportunity to contribute.

                        In this article, we explain what we think the current research implies about how much solutions differ in effectiveness, why this should change how we approach making a difference, and how to find the best solutions within an area in practice.

                        How much do solutions differ in how well they work?

                        In recent years there’s been a wave of advocacy to stop the use of plastic bags. However, convincing someone to entirely give up plastic bags for the rest of their life (about 10,000 bags) would avoid about 0.1 tonnes of CO2 emissions. In contrast, convincing someone to take just one fewer transatlantic flight would reduce CO2 emissions by over one tonne — more than 10 times as much.3

                        And rather than trying to change personal consumption in the first place, we’d argue you could do even more to reduce emissions by advocating for greater funding of neglected green technology.

                        This pattern doesn’t just hold within climate change. Its significance was first pointed out in the field of global health, by Toby Ord’s article “The Moral Imperative toward Cost-Effectiveness in Global Health.”

                        He found data that compared different health interventions in poor countries (e.g. malaria bed nets, vaccines, types of surgery) in terms of how many years of healthy life they produce per $1,000 invested.

                        This data showed that the most cost-effective interventions were around 50 times as cost effective as the median, 23 times the mean, and almost exactly obeyed the 80/20 rule.

                        Intervention cost effectiveness in global health in order of disability-adjusted life years (DALYs) per $1,000 on the y-axis, from the DCP2.

                        This is an incredible finding, because it suggests that one person working on the most effective interventions within global health could achieve as much as 50 people working on a typical intervention.

                        We’ve since seem similar patterns among:

                        • Large US social programmes
                        • Elsewhere in global health, as well as in health in rich countries
                        • Education both in the developing and developed world
                        • Policies to reduce CO2 emissions
                        • Different ways to get out the vote

                        We’ve basically found this pattern wherever data is available.

                        We’ve done a more comprehensive review of the data in a separate research piece:

                        There are some reasons to think that this data overstates the true differences in effectiveness between different solutions — especially those that you can actually work on going forward.

                        One reason is that often the most effective solutions in an area are already being done by someone else.

                        A more subtle reason is regression to the mean. All the estimates involve a lot of uncertainty and error. Some interventions will ‘get lucky’ and end up with errors that make them look better than they are, and others will be unlucky and look worse.

                        In fact, even if all the solutions were equally effective, random errors would make some look above average, and others below average.

                        If we compare the interventions that appear best compared to the average, it’s more likely that they benefited from positive errors. This means that if we investigate them more, they’ll probably turn out worse than they seem.

                        This is what seems to have happened in practice. For instance, the data that Toby used found that distributing insecticide-treated malaria nets was among the most cost-effective solutions in the dataset. However, the latest estimates from the charity evaluator GiveWell are that it’s about 15 times less cost effective than the original estimate.

                        So, the original estimates were too optimistic, and overstated the spread from best to typical, but the expected cost effectiveness of malaria nets still appears to be much higher than average — likely still over 10 times better.

                        In fact, global health experts still believe that the best ways of saving lives in poor countries are around 100 times cheaper than the average.4

                        On the other hand, the data in these studies could also understate the true differences in effectiveness between solutions. One reason is that the data only covers solutions with easy-to-measure results that can be studied in trials, but the highest-impact ways of doing good in the past most often involved research or advocacy, rather than measurable interventions. This would mean the very best solutions are missing from the datasets.

                        For instance, a comparatively small number of people worked on the development of oral rehydration therapy, which now saves around 1 million lives per year. This research was likely extremely cost effective, but we couldn’t directly measure its effectiveness before it was done.

                        Looking forward, we think there’s a good case that medical research aimed at helping the global poor will ultimately be more cost effective than spending on direct treatments, increasing the overall degree of spread.

                        There’s a lot more to say about how much solutions differ in effectiveness, and we’d like to see more research on it. However, our overall judgement is that it’s often possible to find solutions that bring about 10 times as much progress per year of effort than other commonly supported solutions in an area, and it’s sometimes possible to find solutions that achieve 100 times as much.

                        We go into this question in a lot more detail in the research article mentioned earlier.

                        Technical aside: theoretical arguments about how much solutions differ

                        You might think that it’s surprising that such large differences exist. But there are some theoretical arguments that it’s what we should expect:

                        • There isn’t much reason to expect the world of doing good to be ‘efficient’ in the same way that financial markets are, because there are only very weak feedback loops between having an impact and gaining more resources. The main reward people get from doing good is often praise and a sense of satisfaction, but these don’t track the actions that are most effective. We don’t expect it to be entirely inefficient either — even a small number of effectiveness-minded actors can take the best opportunities — but we shouldn’t be surprised to find large differences.

                        • A relatively simple model can give a large spread. For instance, cost effectiveness is produced by the multiple of two factors; we’d expect it to have a log-normal distribution, which is heavy-tailed.

                        • Heavy-tailed distributions seem the norm in many similar cases, for instance how much different experts produce in a field, which means we shouldn’t be surprised if they come up in the world of doing good.

                        • If we think there’s some chance the distribution is heavy-tailed and some chance it’s normally distributed, then in expectation it will be heavy-tailed, and we should act as if it is.

                        What do these findings imply?

                        If you’re trying to tackle a problem, it’s vital to look for the very best solutions in an area, rather than those that are just ‘good.’

                        This contrasts with the common attitude that what matters is trying to ‘make a difference,’ or that ‘every little bit helps.’ If some solutions achieve 100 times more per year of effort, then it really matters that we try to find those that make the most difference to the problem.

                        So, when comparing how much different careers let you contribute to pressing problems, one key thing to consider is the effectiveness of solutions that you’re able to support. (The other factor is how much leverage the career gives you, which we discuss here.)

                        This is especially important later in your career when the focus moves from exploring and investing in your skills to deploying your skills and resources on the most pressing problems.

                        So, how can we find the most effective solutions in an area?

                        Hits-based vs data-driven approaches

                        There are two broad approaches among our readers:

                        1. The data-driven approach: look for data about how much progress per dollar different solutions achieve, ideally randomised trials, and focus on the best ones.
                        2. The hits-based approach: look for rules of thumb that make a solution more likely to be among the very best in an area, while accepting that most of the solutions will be duds.

                        We generally favour the hits-based approach, especially for individuals rather than large institutions, and for people who are able to stay motivated despite a high chance of failure.

                        Why? As noted, the best solutions typically can’t be measured with trials, and so will be automatically excluded if you take the data-driven approach. This is a serious problem — if the best solutions are far more effective than typical, it could be better to pick randomly among solutions that might be the very best, rather than to pick something that’s very likely to be better than average but definitely not among the very best. And we can probably do significantly better than picking randomly — as covered below.

                        Another argument is that many institutions with social missions seem overly risk-averse. For instance, government grant agencies get criticised heavily for funding failures, but the employees at such agencies don’t get much reward when they back winners. This suggests that as individuals who are willing to take risks, you can get an edge by supporting solutions that have a high chance of not working. (You can read more about the arguments for a hits-based approach.)

                        In practice, when it comes to the problem areas we’re most familiar with, the most promising hits-based solutions seem more cost-effective than the most promising data-driven ones.

                        We don’t believe the hits-based approach is always best. Sometimes it might not be possible to find a hits-based solution that seems better than the best data-driven ones.

                        For instance, within their global health and wellbeing team, Open Philanthropy believes that while the best hits-based interventions are indeed more effective than the best data-driven ones, they’re not able to find enough of these to fill up all their grantmaking capacity, so in practice they support both data-driven and hits-based approaches.5

                        Similarly, we’d encourage readers to start by looking for promising solutions with a hits-based approach, and focus on a data-driven one if you’re not able to.

                        Furthermore, if we take a community perspective, we wouldn’t want 100% of our readers to take a hits-based approach. Collectively we’ll have more impact if a minority specialise in a data-driven approach. (Learn more about worldview diversification.)

                        One response to the hits-based approach is that it relies on deeply uncertain judgement calls, instead of objective evidence. That’s true, but we contend that you can’t escape relying on judgement calls. All our actions lead to ripple effects lasting long into the future. By taking a data-driven approach, even if we suppose the data is fully reliable, at best it can capture some of the short-term effects. But you’ll need to rely on judgement calls about the longer-term effects — and if we place some value on future generations, the longer-term effects comprise the majority of all the effects.

                        Given that judgement calls are unavoidable, the best we can do is to try to make the best judgement calls possible, using the best available techniques.

                        Learn more about this argument and the idea of ‘cluelessness’.

                        What does taking a hits-based approach involve in practice?

                        In short, we need to seek rules of thumb that make a solution more likely to be among the very best in an area (and unlikely to be very negative). This could involve methods like the following:

                        • Upside/downside analysis: look for solutions with potentially very high upsides (and limited downsides).
                        • Apply the importance, neglectedness, and tractability framework, but at the level of solutions rather than problems. Here’s an example of this kind of analysis within climate change by Founders Pledge.
                        • Make rough back-of-the-envelope estimates of progress per year of effort invested. Even when the estimates are very uncertain, they can still help us spot very large differences in effectiveness. You can further improve your estimates by applying best practices in forecasting. See some examples from Open Philanthropy.
                        • Bottleneck analysis: try to identify the key step that would unlock progress in an area.6 For instance, in new areas, it’s often most important to do direction-setting research and demonstrate progress is possible; then the priority becomes building a movement around the issue and scaling up the best solutions.
                        • Develop a theory of change, and use it to spot key intervention points.
                        • Taking a data-driven approach can be one input into a hit-based approach: the data-driven approach can help us rule out some interventions that don’t work at all. If we can, say, remove the bottom half of interventions, that can enable us to roughly double our impact, so it’s a useful heuristic – though it’s important to focus more on upsides.

                        In practice, this often ends up with a focus on research, movement building, policy change, or social advocacy, and on solutions that are unfairly neglected or seem unconventional, or that might have especially high upsides.

                        In applying these frameworks, you can either try to do this analysis yourself, or find experts in the area who understand the need to prioritise and can do the analysis on your behalf (or a mix of both).

                        We’re generalists rather than experts in the areas we recommend, so we mainly try to identify good experts — such as those on our podcast — and synthesise their views about how to tackle the problems we write about in our problem profiles.

                        If you aim your career at tackling a specific issue, however, then you’ll probably end up knowing more about it than us, and so should put more weight on your own analysis.

                        Many of the areas we recommend are also small, so not much is known about how best to tackle them. This makes it easier than it seems to become an expert. It also means that your input on which solutions are best is especially valuable — one extra person’s views can add significantly to the collective wisdom on the question.

                        Following expert views also doesn’t necessarily mean choosing ‘consensus’ picks, because those might fall into the trap of being pretty good but not best. Rather, if a minority of experts strongly supports an intervention (and the others don’t think it’s harmful), that might be enough to make it worth betting on. In brief, we aim to consider both the strength of views and how good the interventions seem, and are willing to bet on something contrarian if the upsides might be high enough.

                        By looking for the best solutions, you can dramatically increase your chances of having a big impact on your chosen problem.

                        Now go on to read about some career paths that can set you up to get a lot of leverage on the most effective solutions.

                        Learn more

                        Top recommendations

                        Read next

                        This article is part of our advanced series. See the full series, or keep reading:

                        Plus, join our newsletter and we’ll mail you a free book

                        Join our newsletter and we’ll send you a free copy of The Precipice — a book by philosopher Toby Ord about how to tackle the greatest threats facing humanity. T&Cs here.

                        Notes and references

                        1. van der Put, Claudia E., et al. “Effects of awareness programs on juvenile delinquency: A three-level meta-analysis.” International Journal of Offender Therapy and Comparative Criminology, vol. 65, no. 1, 1996, pp. 68-91. Archived link

                          Petrosino, Anthony, et al. “Scared Straight and other juvenile awareness programs for preventing juvenile delinquency: A systematic review.” Campbell Systematic Reviews, vol. 9, no.1, (2013), pp. 1-55. Archived link

                        2. The percentage that work or don’t work depends a lot on how you define it, but it’s likely that a majority don’t have statistically significant effects.

                        3. According to the 2020 Founders Pledge Climate & Lifestyle Report, just one round trip transatlantic flight contributes 1.6 tonnes of CO2. Figure 2 of the same report shows the comparatively negligible effect of reusing plastic bags.

                        4. Caviola, Lucius, et al. “Donors vastly underestimate differences in charities’ effectiveness.” Judgment and Decision Making, vol. 15, no. 4, 2020, pp. 509-516. Link

                          We selected experts in areas relevant to the estimation of global poverty charity effectiveness, in areas such as health economics, international development and charity measurement and evaluation. The experts were identified through searches in published academic literature on global poverty intervention effectiveness and among professional organizations working in charity evaluation.

                          We found that their median response was a cost-effectiveness ratio of 100 (see Table 1).

                        5. Open Philanthropy is 80,000 Hours’ biggest funder as of 2023.

                        6. Some common types of bottleneck:

                          • Funding: additional financial resources from donations or fundraising.
                          • Insights: new ideas about how to solve the problem.
                          • Awareness and support: how many people know and care about the issue, and how influential they are.
                          • Political capital: the amount of political power that’s available for the issue.
                          • Coordination: the extent to which existing resources effectively work together.
                          • Community building: finding other people who want to work on the issue.
                          • Logistics and operations: the extent to which programmes can be delivered at scale.
                          • Leadership and management: the extent to which concrete plans can be formed and executed using the resources already available.

                        Applications for this position have now closed.

                        We’re looking for a Head of Marketing to help us expand our readership and be the founding member of our marketing team.

                        We’re hoping to find someone who could take on the Head of Marketing position immediately. However, we’re also open to hiring a candidate with less experience who we could support to take on the responsibilities of a Head of Marketing over time. To apply for the more junior position instead, please see our Marketer job description.

                        80,000 Hours provides free research and support to help people find careers tackling the world’s most pressing problems.

                        We’ve had over 8 million visitors to our website, and more than 3,000 people have told us that they’ve significantly changed their career plans due to our work. We’re also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Survey.

                        Even so, about 90% of U.S. college graduates have never heard of effective altruism, and just 0.5% of students at top colleges seem highly engaged in EA. As Head of Marketing, your aim would be to help us reach all students and recent graduates who might be interested in our work. We anticipate this could increase our readership up to five times, and lead to hundreds more people pursuing high-impact careers.

                        We’re looking for a senior marketing generalist who will:

                        • Develop our marketing strategy. For example, you’d generate ideas for major marketing initiatives, decide which to pursue, and figure out which metrics we should optimise to most effectively achieve our goals.

                        • Run experiments in new ways to reach users. For example, you could set up a referral campaign with a book giveaway, coordinate a media campaign to launch the second version of our ‘key ideas’ series, or test other marketing ideas that you come up with.

                        • Own and improve our current approach to reaching users. For example, you might promote our work on social media and other platforms, write newsletters, and run A/B tests to maximise engagement.

                        • Build and manage a marketing team to execute on the above.

                        Which of these you’ll focus on will depend on your strengths and interests, as well as what you decide to prioritise. But as some indication of what success in the role might look like, over the next three years you might have:

                        • Doubled our newsletter subscriptions from 150,000 to 300,000
                        • Doubled our reach, increasing our unique website visitors from 1.5 to 3 million per year
                        • Increased engagement with our job board by 5x, taking the number of times users click through to a job ad they find on our job board from 10,000 to 50,000 per month
                        • Found effective ways to spend a six-figure (or even seven-figure) marketing budget
                        • Launched one or more major new marketing initiatives

                        We’re looking for someone who has:

                        • A strong interest in effective altruism and longtermism, ideally with experience in EA strategy
                        • Previous experience in any area of marketing, or a related field (This might be — but isn’t limited to — product management, software engineering, data science or communication; or maybe you’ve worked on a side project that attracted a large number of users)
                        • Experience building and managing a team
                        • An interest in thinking carefully about what will drive engagement with our work from people who might make especially high-impact career changes, and what this means for our marketing strategy
                        • An ambitious approach to the role, with enthusiasm for generating new ideas for how we might appeal to our audience
                        • Excellent written communication (In particular, you’re comfortable discussing decisions and uncertainties with the rest of the team in writing)

                        Ideally, you’d also have the ability to write engaging marketing copy that both appeals to our target audience and communicates the nature and promise of 80,000 Hours’ programmes.

                        This is a full-time, in-person role, based in London. You can work remotely for up to three months of the year if needed.

                        The salary will vary based on your skills and experience, but to give a rough sense, the starting salary for someone with five years of relevant experience would be approximately £81,000 per year.

                        Our benefits include:

                        • The option to use 10% of your time for self development
                        • 25 days of paid holiday, plus bank holidays
                        • Standard U.K. pension with 3% contribution from employer
                        • Private medical insurance
                        • Long-term disability insurance
                        • Gym, shower facilities, and free food provided at our London office

                        To apply, please fill in this application form. If you have any problems submitting the form, please send your CV to [email protected].

                        Applications for this position have now closed.

                        We’re looking for a Marketer to help us expand our readership and be the founding member of our marketing team.

                        We’d like to support the person in this role to take on more responsibility over time and eventually become our Head of Marketing.

                        We’re also open to hiring someone more senior, who could take on the Head of Marketing role immediately. To apply for the Head of Marketing position instead, please see the job description here.

                        80,000 Hours provides free research and support to help people find careers tackling the world’s most pressing problems.

                        We’ve had over 8 million visitors to our website, and more than 3,000 people have told us that they’ve significantly changed their career plans due to our work. We’re also the largest single source of people getting involved in the effective altruism community, according to the most recent EA Survey.

                        Even so, about 90% of U.S. college graduates have never heard of effective altruism, and just 0.5% of students at top colleges seem highly engaged in EA. As 80,000 Hours’ Marketer, your aim would be to help us reach all students and recent graduates who might be interested in our work. We anticipate this could increase our readership up to five times, and lead to hundreds more people pursuing high-impact careers.

                        We’re looking for a marketing generalist who will:

                        • Run experiments in new ways to reach readers. For example, you could set up a referral campaign with a book giveaway, coordinate a media campaign to launch the second version of our ‘key ideas’ series, or test other marketing ideas that you come up with.

                        • Own and improve our current approach to reaching users. For example, you might promote our work on social media and other platforms, write newsletters, and run A/B tests to maximise engagement.

                        We would also be excited to support you towards eventually:

                        • Developing our marketing strategy. For example, generating ideas for major marketing initiatives, deciding which to pursue, and figuring out which metrics we should optimise to most effectively achieve our goals.

                        • Building and managing a marketing team to execute on the above.

                        We have some in-house marketing expertise you could learn from, and we’d also be open to you working with an agency or coach to develop your skills.

                        Your focus in this role will depend on your strengths and interests, as well as what you decide to prioritise. But as some indication of what success in the role might look like, over the next three years you might have:

                        • Doubled our newsletter subscriptions from 150,000 to 300,000
                        • Doubled our reach, increasing our unique website visitors from 1.5 to 3 million per year
                        • Increased engagement with our job board by 5x, taking the number of times users click through to a job ad they find on our job board from 10,000 to 50,000 per month
                        • Found effective ways to spend a six-figure (or even seven-figure) marketing budget
                        • Launched one or more major new marketing initiatives

                        We’re looking for someone who has:

                        • A strong interest in effective altruism and longtermism, ideally with experience in EA strategy
                        • An ambitious approach to the role, with enthusiasm for generating new ideas for how we might appeal to our audience
                        • Excellent written communication (In particular, you’re comfortable discussing decisions and uncertainties with the rest of the team in writing)
                        • An interest in thinking carefully about what will drive engagement with our work from people who might make especially high-impact career changes, and what this means for our marketing strategy

                        Ideally, you’d also have the following traits, but we encourage you to apply even if they don’t describe you:

                        • Previous experience in any area of marketing, or a related field (This might be — but isn’t limited to — product management, software engineering, data science, or communication; or maybe you’ve worked on a side project that attracted a large number of users)
                        • The ability to write engaging marketing copy that both appeals to our target audience and communicates the nature and promise of 80,000 Hours’ programmes

                        This is a full-time, in-person role, based in London. You can work remotely for up to three months of the year if needed.

                        The salary will vary based on your skills and experience, but to give a rough sense, the starting salary for someone with one year of relevant experience would be approximately £60,000 per year.

                        Our benefits include:

                        • The option to use 10% of your time for self development
                        • 25 days of paid holiday, plus bank holidays
                        • Standard U.K. pension, with 3% contribution from employer
                        • Private medical insurance
                        • Long-term disability insurance
                        • Gym, shower facilities, and free food provided at our London office

                        To apply, please fill in this application form. If you have any problems submitting the form, please send your CV to [email protected].

                        Part 2: Can one person make a difference? What the evidence says.

                        It’s easy to feel like one person can’t make a difference. The world has so many big problems, and they often seem impossible to solve.

                        So when we started 80,000 Hours — with the aim of helping people do good with their careers — one of the first questions we asked was, “How much difference can one person really make?”

                        We learned that while many common ways to do good (such as becoming a doctor) have less impact than you might first think, others have allowed certain people to achieve an extraordinary impact.

                        In other words, one person can make a difference — but you might have to do something a little unconventional.

                        In this article, we start by estimating how much good you could do by becoming a doctor. Then, we share some stories of the highest-impact people in history, and consider what they mean for your career.

                        Reading time: 12 minutes

                        How much impact do doctors have?

                        Many people who want to help others become doctors. One of our early readers, Dr Greg Lewis, did exactly that. “I want to study medicine because of a desire I have to help others,” he wrote on his university application, “and so the chance of spending a career doing something worthwhile I can’t resist.”

                        So, we wondered: how much difference does becoming a doctor really make? We teamed up with Greg to find out.

                        Since a doctor’s primary purpose is to improve health, we tried to figure out how much extra “health” one doctor actually adds to humanity. We found that, over the course of their career, an average doctor in the UK will enable their patients to live about an extra combined 100 years of healthy life, either by extending their lifespans or by improving their overall health. There is, of course, a huge amount of uncertainty in this figure, but the real figure is unlikely to be more than 10 times higher.

                        Using a standard conversion rate (used by the World Bank, among other institutions1) of 30 extra years of healthy life to one “life saved,” 100 years of healthy life is equivalent to about three lives saved. This is clearly a significant impact; however, it’s less of an impact than many people expect doctors to have over their entire career.

                        There are three main reasons this impact is lower than you might expect:

                        1. Researchers largely agree that medicine has only increased average life expectancy by a few years. Most gains in life expectancy over the last 100 years have instead occurred due to better nutrition, improved sanitation, increased wealth, and other factors.

                        2. Doctors are only one part of the medical system, which also relies on nurses and hospital staff, as well as overhead and equipment. The impact of medical interventions is shared between all of these elements.

                        3. Most importantly, there are already a lot of doctors in the developed world, so if you don’t become a doctor, someone else will be available to perform the most critical procedures. Additional doctors therefore only enable us to carry out procedures that deliver less significant and less certain results.

                        This last point is illustrated by the chart below, which compares the impact of doctors in different countries. The y-axis shows the amount of ill health in the population, measured in disability-adjusted life years (DALYs) per 100,000 people, where one DALY equals one year of life lost due to ill health. The x-axis shows the number of doctors per 100,000 people.

                        DALYs compared to doctors
                        DALYs per 100,000 people versus doctors per 100,000 people. We used WHO data from 2004. Line is the best fitting hyperbola determined by nonlinear least square regression. Full explanation in this draft paper.

                        You can see that the curve goes nearly flat once you have more than 150 doctors per 100,000 people. After this point (which almost all developed countries meet), additional doctors only achieve a small impact on average.

                        So if you become a doctor in a rich country like the US or UK, you may well do more good than you would in many other jobs, and if you are an exceptional doctor, then you’ll have a bigger impact than these averages. But it probably won’t be a huge impact.

                        In fact, in the next article, we’ll show how almost any college graduate can do more to save lives than a typical doctor. And in the rest of the career guide, we’ll cover many other examples of common but ineffective attempts to do good.

                        These findings motivated Greg to switch from clinical medicine into biosecurity, for reasons we’ll explain over the rest of the guide.

                        Who were the highest-impact people in history?

                        Despite this uninspiring statistic about how many lives a doctor saves, some doctors have had much more impact than this. Let’s look at some examples of the highest-impact careers in history, and see what we might learn from them. First, let’s turn to medical research.

                        By 1968, it had been shown that a solution of glucose and salt, administered via feeding tube or intravenous drip, could prevent death due to cholera. But millions of people were still dying every year from the disease. While working in a refugee camp on the border of Bangladesh and Burma, Dr David Nalin sought to turn this insight into a therapy that could be used in poor rural areas. He showed in a study that simply drinking a solution made at the right concentration and consumed at the right rate could be almost as effective as delivery via feeding tube or IV.

                        This meant the treatment could be delivered with no equipment, and using extremely cheap and widely available ingredients.

                        Dr Nalin helped to invent oral rehydration therapy
                        Dr Nalin helped to save millions of lives with a simple innovation: giving diarrhoea patients water mixed with salt and sugar.

                        Since then, this astonishingly simple treatment has been used all over the world, and the annual rate of child deaths from diarrhoea has plummeted from around 5 million to 1.5 million.2 Researchers estimate that the therapy has saved over 50 million lives to date, mostly children’s.3

                        If Dr Nalin had not been around, someone else would, no doubt, have discovered this treatment eventually. However, even if we imagine that he sped up the roll-out of the treatment by only five months, his work alone would have saved about 500,000 lives. This is a very approximate estimate, but it makes his impact more than 100,000 times greater than that of an ordinary doctor:

                        Lives saved by Dr Nalin

                        But even just within medical research, Dr Nalin is far from the most extreme example of a high-impact career. For example, one estimate puts Karl Landsteiner’s discovery of blood groups as saving tens of millions of lives by enabling transfusions.4

                        Lives saved by Dr Landsteiner

                        Beyond the medical field, later in the guide we’ll cover the stories of a hugely impactful mathematician, Alan Turing, and bureaucrat, Viktor Zhdanov.

                        Or, let’s think even more broadly. Roger Bacon and Galileo pioneered the scientific method — without which none of the discoveries we covered above would have been possible, along with other major technological breakthroughs like the Industrial Revolution. These individuals were able to do vastly more good than even outstanding medical practitioners.

                        The unknown Soviet Lieutenant Colonel who saved your life

                        Stanislav Petrov probably saved your life

                        Or consider the story of Stanislav Petrov, a Lieutenant Colonel in the Soviet Army during the Cold War. In 1983, Petrov was on duty in a Soviet missile base when early warning systems apparently detected an incoming missile strike from the United States. Protocol dictated that the Soviets order a return strike.

                        But Petrov didn’t push the button. He reasoned that the number of missiles was too small to warrant a counterattack, thereby disobeying protocol.

                        If he had ordered a strike, there’s at least a reasonable chance hundreds of millions would have died. The two countries may have even ended up engaged in an all-out nuclear war, leading to billions of deaths and, potentially, the end of civilisation. If we’re being conservative, we might quantify his impact by saying he saved a billion lives. But that’s almost certainly an underestimate, because a nuclear war would also have devastated scientific, artistic, economic, and all other forms of progress, leading to a huge loss of life and wellbeing over the long run.

                        Later in the guide we’ll discuss why we think these long-run effects could be vastly more important than “just” saving a billion lives from nuclear catastrophe.

                        Yet even with the lower estimate, Petrov’s impact likely dwarfs that of Nalin and Landsteiner.

                        Lives saved by Petrov

                        What do these differences in impact mean for your career?

                        We’ve seen that some careers have had huge positive effects, and some have vastly more than others.

                        Some component of this is due to luck — the people mentioned above were in the right place at the right time, giving them the opportunity to have an impact that they might not have otherwise received. You can’t guarantee you’ll make an important medical discovery.

                        But it wasn’t all luck: Landsteiner and Nalin chose to use their medical knowledge to solve some of the most harmful health problems of their day, and it was foreseeable that someone high up in the Soviet military might have the opportunity to have a large impact by preventing conflict during the Cold War.

                        So, what does this mean for you?

                        People often wonder how they can “make a difference,” but if some careers can result in thousands of times more impact than others, this isn’t the right question. Two different career options can both “make a difference,” but one could be dramatically better than the other.

                        Instead, the key question is: What are some of the best ways to make a difference? In other words, what can you do to give yourself a chance of having one of the highest-impact careers? Because the highest-impact careers achieve so much, a small increase in your chances means a great deal.

                        The examples above also show that the highest-impact paths might not be the most obvious ones. Being an officer in the Soviet military doesn’t sound like the best career for a would-be altruist, but Petrov probably did more good than our most celebrated leaders, not to mention our most talented doctors. Having a big impact might require doing something a little unconventional.

                        So how much impact can you have if you try, while still doing something personally rewarding? It’s not easy to have a big impact, but there’s a lot you can do to increase your chances. That’s what we’ll cover in the next couple of articles.

                        But first, let’s clarify what we mean by “making a difference.” We’ve been talking about lives saved so far, but that’s not the only way to do good in the world.

                        What does it mean to “make a difference”?

                        Everyone talks about “making a difference” or “changing the world” or “doing good,” but few ever define what they mean.

                        So here’s a definition. Your social impact is given by:

                        The number of people5 whose lives you improve, and how much you improve them, over the long term.6

                        This means you can increase your social impact in three ways:

                        1. By helping more people.
                        2. By helping the same number of people to a greater extent (pictured below).
                        3. By doing something which has benefits that last for a longer time.

                        We think the last option is especially important, because many of our actions affect future generations. For example, if you improve the quality of government decision-making, you might not see many quantifiable short-term results, but you will have solved lots of other problems over the long term.

                        Social impact - how to change the world - help more people, or help people more

                        Why did we choose this definition?

                        We have a separate article about our definition, but here are some brief points:

                        Many people disagree about what it means to make the world a better place. But most agree that it’s good if people have happier, more fulfilled lives, in which they reach their potential. So, our definition is narrow enough that it captures this idea.

                        Moreover, as we’ll show, some careers do far more to improve lives than others, so it captures a really important difference between options. If some paths can do good equivalent to saving hundreds of lives, while others have little impact at all, that’s an important difference.

                        But the definition is also broad enough to cover many different ways to make the world a better place. It’s even broad enough to cover environmental protection, since if we let the environment degrade, the future of civilisation might be threatened. In that way, protecting the environment improves lives.

                        Importantly, having a broad scope also allows us to include nonhuman animals, as well as potential future sentient beings that might be entirely digital — which is why we have profiles on factory farming, wild animal welfare, and artificial sentience.

                        That said, the definition doesn’t include everything that might matter. You might think the environment deserves protection even if it doesn’t make people better off. Similarly, you might value things like justice and aesthetic beauty for their own sakes.

                        In practice, our readers value many different things. Our approach is to focus on how to improve lives, and then let people independently take account of what else they value. To make this easier, we try to highlight the main value judgements behind our work. It turns out there’s a lot we can say about how to do good in general, despite all these differences.

                        How can you measure social impact?

                        We are always uncertain about how much impact different actions will have — but that’s OK, because we can use probabilities to make comparisons. For instance, a 90% chance of helping 100 people is roughly equivalent to a 100% chance of helping 90 people. Though we’re uncertain, we can quantify our uncertainty and make progress.

                        Moreover, even in the face of uncertainty, we can use rules of thumb to compare different courses of action. For instance, later in this career guide we argue that, all else equal, it’s higher impact to work on neglected areas. So, even if we can’t precisely measure social impact, we can still be strategic by picking neglected areas. We’ll cover many more rules of thumb for increasing your impact in the upcoming articles.

                        Is social impact all that matters?

                        No.

                        We don’t know the ultimate truths of moral philosophy, but in the real world we think it’s really important not to only focus on impact.

                        In particular, it’s normally better — even from the perspective of social impact — to always act with good character, respect the rights and values of others, and to pay attention to your other personal values.

                        We don’t endorse doing something that seems very wrong from a common-sense perspective, even if it seems like it might let you have a greater impact.

                        Read more about our definition of social impact.

                        So how can you improve lives with your career?

                        In the next article, we’ll cover how any college graduate can make a big impact in any job. After that we’ll cover how to choose a job in which you can fulfil your potential for impact.

                        Read next: Part 3: No matter your job, here’s 3 evidence-based ways anyone can have a real impact

                        Continue →

                        Or see an overview of the whole career guide.

                        Want to come back later? Get the guide as a free book

                        Sign up to our newsletter, and we’ll mail you the entire career guide as a book.

                        You’ll be joining over 350,000 people who receive weekly updates on our research and job opportunities. T&Cs here. You can unsubscribe in one click.

                        Notes and references

                        1. Source: World Bank, p. 402, retrieved 31-March-2016

                        2. Institute for Health Metrics and Evaluation, Global Burden of Disease Study 2019, 2019.https://vizhub.healthdata.org/gbd-results/

                        3. Since the adoption of this inexpensive and easily applied intervention, the worldwide mortality rate for children with acute infectious diarrhoea has plummeted from around 5 million to about 1.5 million deaths per year. Lives Saved: Over 57,500,000.

                          Source: Science Heroes. Archived link, retrieved 4-December-2022.

                          Very roughly, this means 50/40 = 1.25 million lives have been saved per year. So if Dr Nalin sped up the discovery by five months (just a guess), that means that (5/12)*1.25 = 0.52 million extra lives were saved by his actions. This is a highly approximate estimate and could easily be off by an order of magnitude. See more comments in the next footnote.

                        4. Superman of Science Makes Landmark Discovery – Over 1 Billion Lives Saved So Far

                          Every source quoted an amazing number of transfusions and potential lives saved in countries and regions worldwide. High impact years began around 1955 and calculations are loosely based on 1 life saved per 2.7 units of blood transfused. In the USA alone an estimated 4.5 million lives are saved each year. From these data I determined that 1.5% of the population was saved annually by blood transfusions and I applied this percentage on population data from 1950-2008 for North America, Europe, Australia, New Zealand, and parts of Asia and Africa. This rate may inflate the effectiveness of transfusions in the early decades but excludes the developing world entirely.

                          Source: Science Heroes. Archived link, retrieved 4-March-2016.

                          If we assume a constant number of lives saved per year, then that’s about 10 million lives per year. If he sped up the discovery by two years, then that’s 20 million lives saved.

                          This is a highly approximate estimate and could easily be off by an order of magnitude in either direction, and seem more likely to be too high than too low. We’re a bit sceptical of the Science Heroes figures. Moreover, our attempt at modelling the speed-up is very simple. Since most of the lives were saved in the modern era once a large number of people had medical care, it’s possible that speeding up the discovery wouldn’t have had much impact at all. On the other hand, the discovery of blood groups probably made other scientific advances possible, and we’re ignoring their impact. Nevertheless, the basic point stands: Landsteiner’s impact was likely vastly greater than a typical doctor.

                        5. We often say “helping people” here for simplicity and brevity, but we don’t mean just humans — we mean anyone with experience that matters morally — e.g. nonhuman animals that can suffer or feel happiness, or even conscious machines if they ever exist.

                        6. This definition is enough to help you figure out what to aim at in many situations — e.g. by roughly comparing the number of people affected by different issues. But sometimes you need a more precise definition.

                          The more rigorous working definition of social impact used by 80,000 Hours is:

                          “Social impact” or “making a difference” is (tentatively) about promoting total expected wellbeing — considered impartially, over the long term — without sacrificing anything that might be of comparable moral importance.

                          You can read about why we use this definition in our article on defining social impact.

                        How can doctors do the most good? An interview with Dr Gregory Lewis

                        26WAITING-superJumbo

                        Gregory Lewis is public health doctor training in the east of England. He studied medicine at Cambridge, where he volunteered for Giving What We Can and 80,000 Hours. He blogs at The Polemical Medic. This interview was conducted as part of the research I did for Will MacAskill’s book, Doing Good Better: How Effective Altruism Can Help You Make a Difference. Greg’s story is discussed in the fourth and fifth chapters of that book.

                        Pablo Stafforini: To get us started, can you tell us a bit about your background, and in particular about your reasons for deciding to become a doctor?

                        Gregory Lewis: Sure. I guess I found myself at the age of 14 or so being fairly good at science and not really having any idea of what to do with myself. I had some sort of vague idea of wanting to try to make the world a better place, in some slightly naive way. So I sort of thought, “What am I going to do with myself?” And my thoughts were pretty much verbatim, “Well, I’m good at science and want to do good. Doctors are good at science and they want to do good. Therefore, I want to be a doctor.” So based on that simple argument, I applied to medical school, got in, spent the following six years of my life in medical school qualifying as a doctor, and here I am today.

                        Pablo: I suppose that at some point between the age of fourteen and the present you changed your mind to some degree about the kind and amount of good that doctors could do. Can you elaborate on that?

                        Greg: Yes. One of my interests outside medicine was philosophy – I almost studied philosophy at university, but I thought I could do more good as a doctor. It was through this I read Peter Unger’s book Living High and Letting Die. This book opened my eyes to the importance and moral significance of giving substantially to charity, and I took this message to heart. But I didn’t really link it up with what I’d planned in my career: I thought I would heal the sick (if you’ll excuse the expression) in my day job, and the good I would do by giving a lot to charity would be an added bonus.

                        It took a couple of years, and coming into contact with people like Toby Ord and Will MacAskill, to begin to put these things together, and look again at my plans to be a doctor. How much good did doctors really do and (more importantly) how did it stack up in comparison with all the other things I could do instead? So I began to look at this question and found (somewhat to my disappointment) that working as a doctor doesn’t fare well in this sort of comparison.

                        Pablo: You mention Unger’s book, and I recall that in that book the argument for earning to give is briefly sketched. Did you notice that argument when reading the book, or did you just focus on the message that you should be donating a big chunk of the money you’d expect to earn in your current career (as opposed to switching to an even more lucrative career)?

                        Greg: Yes, I remember reading those couple of paragraphs where he suggests that philosophers should consider moving out of academia into corporate law or more lucrative fields, so that they would have more money to give away. That sounded right to me back then, but I didn’t really see medicine at the time as an ‘earning to give’ career—I thought the direct impacts of medicine were substantial, so a medical career got the ‘best of both worlds’, and the money I would give away would be an ‘added bonus’ to the direct work as a medical doctor. It took getting more involved with the effective altruism community to think I should try and combine these two worlds, and that I should try and weigh up how much good doctors do versus how much good donations do, and plan my career accordingly.

                        Pablo: So when you started to think more systematically about the amount of good that doctors could do, do you think you encountered any internal resistance to the possible conclusion that this amount might not be as high as you had assumed initially?

                        Greg: The seeds of scepticism were sown fairly early in my training. Doctors themselves generally are fairly cynical of the good they do, and when they talk about ‘healing the sick’, it is with tongue firmly in cheek. One conversation I remember clearly (and with retrospect I wish I paid attention to more) was talking to a doctor in paediatrics, who said something along the lines of, “I don’t ever feel like I’m saving lives or making a big difference, because although I might be the guy giving the life-saving treatments, if I wasn’t there they would have called the doctor just down the hall, who would have done exactly the same as I.”

                        I gradually internalized this more realistic view on how much good I could do as a doctor. This was somewhat disappointing to me, but I wasn’t that phased by it. Maybe the world is just set up that it’s really hard to make a big difference, and if the best I could hope for was to make a more modest contribution, that is still definitely ‘worth it’, and (like many other doctors) I decided my prior zeal to heal the sick and save the world was quixotic, immature, and naive: “The mark of an immature man is the desire to die nobly for a cause, whilst the mark of the mature man is the desire to live humbly for one.” I flirted with the idea of working for Médecins Sans Frontières (MSF) or abroad, but I wasn’t thinking systematically.

                        So one of the major upsides of reading Living High and Letting Die was finding out my 17 year old self wasn’t so unrealistic in hoping to save hundreds or thousands of lives—things that good are within our reach. The downside was this would happen through a very different channel. Rather than 17-year-old me’s vainglorious visage of (thousands of times over!) striding in, white coat billowing, and saving some stricken patient with my cleverness, it would be me posting a cheque or clicking a bank transfer: I’d know abstractly that this would do so much good, but I wouldn’t be able to point to the person it was that I helped. As it turns out, that’s no big deal — especially compared to the sheer magnitude of good done.

                        Pablo: Your conclusion that you wouldn’t in your capacity as a doctor be doing as much good as alternative paths to impact appears to involve both a premise about the amount of good doctors typically do and a premise about the amount of good that such people can do in other ways. That is, you seem to be claiming that doctors do less good directly than people assume, but also that people, including doctors, can do much more good than they think by donating to the right causes. Is that correct?

                        Greg: Yes. The major upshot of the work I’ve done into how much good a doctor does is that the average doctor probably saves around a handful of lives over their career. So that’s bad news for medics. By contrast, giving fairly small amounts to charity can save hundreds of lives (or maybe more) over your working life, and that’s good news for everyone!

                        Pablo: Let’s zoom in on your work about the good doctors do. Insofar as it’s possible to discuss these issues in an informal interview, without having all the relevant figures in front of us, can you sketch the argument for the conclusion that a doctor saves about 200 DALYs over the course of his or her career?

                        Greg: Sure. I started looking at the research literature expecting there would be a lot of work done on the ‘return’ of having more doctors—I figured this would be important to running a health system, or something more introspective members of the profession would have wanted to find out. As it happened, there was basically no work looking at the question: “How much good does a doctor do?”

                        The closest is work by an epidemiologist called John Bunker: he and colleagues were looking at the question of how much of the dramatic gains in health and life expectancy in the western world could be attributed to medical treatment. Their strategy was to look at the few hundred or so commonest medical interventions: fixing broken bones, treating heart attacks, stuff like that. For each of these, they looked at clinical trials to see how much good each of those things did, and, by adding them together, work out how much good medicine as a whole does. You can extrapolate from their figures to many healthy life years (a measure of length and quality of life—you can think of 30 health years as ‘one life saved’) are added to the population by the medical profession, and then divide by the number of doctors to get the ‘years added per doctor’ This is about 2250 health-years saved per medical career—that’s pretty good, about 80 lives.

                        There are several reasons to suspect this is an overestimate. One of the big ones is that the difference of a doctor should be on the margin. Although the first few doctors should be able to make a massive difference, subsequent doctors (like being the 170001st in the UK) should make a smaller difference, as all the easy ways to make people live longer and healthier should already be being done. If I were removed from my post, there wouldn’t be a ‘Greg shaped hole’ in the hospital where all my patients are not treated. Rather the remaining doctors will reallocate their tasks so only the least important things don’t get done.

                        So I began to attack this problem from the ‘top down’ rather than from the ‘bottom up’. Instead of compiling an inventory of medical treatment, I looked at aggregate measures of health and physician density, and looked at the relationship between them: looking at all the countries in the globe, did having more doctors per capita correspond to lower burdens of death and disability? The answer was ‘yes’, but there were diminishing returns. What I then did was fit a best line to this curve, and work out, if you were in the UK and you added one more doctor to the population, how much further along the curve do you go, and how much does disability fall? This figure is smaller, of the order of 400-800 health-years averted per medical career: 20 to 30 lives.

                        This figure, however, is also going to be an overestimate, because we implicitly ignore confounding factors—there are fairly obvious things that will increase both health and physician density, like wealth, sanitation, or education. Indeed, it’s received wisdom that these ‘social determinants of health’ are far more important than doctors. Happily, international data on these factors are also available, and one can try and tease apart these interrelationships by a technique called regression analysis. This gives a smaller figure still. The average doctor averts 200 or so DALYs per medical career—six lives or so.

                        There are all sorts of caveats with this sort of work—the data is fair but not great, it is fundamentally an observational study, and there’s always the spectre of unaccounted for confounds. Despite these concerns, I’d be surprised if this figure was off by an order of magnitude or more. If anything, this already fairly low estimate is also over-optimistic: two big factors would be that I’m ignoring counterfactuals and elasticity (if I never went to medical school, there wouldn’t be ‘one fewer doctor’, it would be more like ‘I would be replaced by the marginal candidate who just missed out on med school); even worse, physician density is serving as a proxy for ‘medical professional density’, from nurses, to hospital cleaners, to laboratory scientists. It’s implausible that doctors can take all of the credit, or even a majority of it. So even if doctors have the largest impact out of all the health professions, one is still looking at another adjustment down, by a factor of at least two

                        Pablo: How much could altruistically motivated doctors boost that figure if they targeted their efforts more intelligently, e.g. by working in a less developed country or in a more lucrative specialty?

                        Greg: That was the question I asked myself next: given this is the average impact of a doctor, how could I try to do better than average? This is tricky, as the ‘top down’ technique I used to find the average is too coarse-grained to answer these questions: there isn’t the data, for example, to work out whether the marginal impact of cardiologists is greater than colorectal surgeons, or things like that.

                        One strategy could be to exploit the ‘diminishing returns’ effect and go somewhere where the curve is steeper and so there are increasing benefits to having ‘an additional doctor’—this really crudely models ‘a career spent working in MSF’ or with a similar NGO. This does give a bigger impact, by a factor of 10 or so.

                        However, the chequebook can likely beat the stethoscope, even one wielded by an MSF doctor. The average doctor in the UK will earn around 2.5 million pounds over their lifetime. Giving 10% of this to the right interventions will still ‘beat’ an additional doctor abroad. And one can always give more than 10%, and although that is hard, it may not be as hard as spending one’s career in the developing world.

                        The next question—going back to Unger—is whether there are particularly lucrative medical careers one could target with the aim of giving more away. And there are, at least when working in the Western world. To give the UK as an example, average consultant earnings by specialty vary by a factor of 3 or so, and the main determinant of this variation is the capacity that specialty has for private practice: you can’t really work privately as an emergency physician, but one can work wholly outside the NHS as a plastic surgeon. So medicine is a fairly good earning to give option, although it is worth noting that if earning to give ‘beats’ direct impact by a large margin, it perhaps would be even better to attempt to work in even more lucrative careers outside of medicine.

                        Finally, there are ‘peri-medic’ roles that could be really important but hard to quantify: the chief medical officer for the NHS (or for the WHO), the researcher who makes the breakthrough for a malaria vaccine could have massive impact, so much so that it might be worth attempting even if it is a very long shot and one is likely to achieve something far more modest. It’s pretty hard to quantify these considerations, but they look like career paths that could be even better than earning to give.

                        Perhaps the upshot is that direct work as a doctor is relatively small-fry compared to what you could do instead. Which ‘instead’, though, remains very difficult to work out.

                        Pablo: To wrap up, you mentioned that you are currently giving about 50% of your income to cost-effective charities. Can you elaborate on what motivated you to give away such a big portion of your income and whether you find that difficult on a personal level?

                        Greg: Sure. So, given what I’d read by Unger, and in philosophy more generally, giving a lot to charity seemed a bit of a moral no-brainer. On the one side, several lives in my first year of being employed (and several thousand over my career as my salary grows), and on the other side, not a huge amount. So I committed to give 10% of my earnings whilst I was a medical student.

                        It became even more of a no-brainer when I actually started working. I am privileged in all manner of ways, but not least in that I live without dependents in a modern liberal democracy with almost double the median income of my country, and so among the top few percent of the planet by wealth. I found living similarly (but still better) than I did as a medical student left me with almost half my paycheck. So I started giving 10%, and have steadily increased this month on month until now I’m giving about 50%.

                        It’s only at this larger proportion that there’s any real personal ‘sacrifice’ on my part: I now plan journeys in advance, keep a monthly budget, and don’t reflexively eat out whenever the opportunity presents itself. I also haven’t (as some of my colleagues have) got a BMW on franchise, or regularly holiday across the world. I don’t really miss these luxuries, especially as these sacrifices are made without choice by most people living in the UK (and the globe), including people who work much harder and longer than I do alongside me in hospital.

                        I’m still in the wealthiest 10% of people on the planet. More importantly, I still get to keep the things that really matter: family, friends, literature, music, a career that, even though it might not save the world, is immensely personally fulfilling and interesting. Even better, I am happy I am doing something significant in making the world go better. I think the 17 year old me who wanted to be a doctor would be happy, but surprised, at the doctor he turned into.

                        Pablo: Awesome. Thanks, Greg!

                        The best-selling author, Tim Ferriss, promises to teach you how to cut your working week down to just four hours, using a careful combination of Indian virtual assistants, the 80/20 principle and automatic email responders.

                        But Ferriss has nothing on us. If your goal is to help others, then you can cut your entire working year down to just four hours.

                        Suppose your mission in life is to fight HIV/AIDS. Consider two charities.


                        1. The first prevents deaths due to AIDS by handing out antiretroviral therapy to people infected with the virus.

                        2. The second prevents deaths due to AIDS by educating people about the need for contraception through television shows.


                        Which one should you work for?

                        Both are producing massive benefits for others, but the second measure is cheap and preventative, so intuitively it looks like it might work better. To find out, let’s look at the latest cost-effectiveness research.

                        According to the Disease Control Priorities Second Report, $500 used on antiretroviral therapy allows one person to live an extra year of healthy life. This is a great deal, working out at only 6c per hour. The same resources used on mass media education, however, allows about 500 people to live an extra year of healthy life[^1].

                        If both charities are roughly the same size, and you can make the same kind of contribution at each, then it seems like you can do about 500 times as much to fight AIDS at the mass media charity. For instance, suppose you make the charity more efficient, so that it has lower costs. Or you help them to fundraise more. Or you help them recruit another great employee who does either of these. In each case, the extra resources go 500 times as far at the second charity.

                        A year of work is about 2000 hours. So, an afternoon’s work at the mass education charity does as much to fight HIV/AIDs as one long year of full-time work at the antiretrovial therapy charity.

                        It turns out that both of these charities exist. The American Foundation for Children with AIDS gives out antiretroviral therapy. Development Media International (which we’re proud to say includes a member of 80,000 Hours among its Directors) creates mass media campaigns in the developing world aimed at preventing sexually transmitted disases. We’re faced with choices like these all the time.

                        A four hour work week is 200 hours per year. So there you have it. Since we’re focused on helping others, we’ve managed to be 50 times better than Ferriss.


                        In fact, we can go even further. We’ve focused on a very narrow area: HIV/AIDS. The differences become even larger when we look at the entire range of ways to fight global poverty. Indeed, even in the areas we can’t as easily quantify, we should expect these huge differences will still exist. And finally, notice that although our estimate is very rough, the difference between the two options is so large that it’s good enough.


                        What does this mean?


                        • You can have a huge impact. Raising $100,000 for the mass media charity (an easy amount for a charity fundraiser) could mean preventing an entire large village worth of people dying from AIDS.
                        • But you need to think about how best to go about it. Making some rough estimates could let you make hundreds of times more impact.

                        You might also be interested in:


                        [1]: DCP2, Chapter 18: Using a disability-weighting for HIV of 0.135, it costs $500 to avert one DALY for antiretroviral therapy, but about $1 for mass media education.

                        How many lives does a doctor save? – Part 2 – Diminishing marginal returns

                        In the first post, I worked out an upper bound for the average direct health impact of a doctor in the UK, and found it amounted to producing about 2600 QALYs. We can think of this, very roughly, as saving 90 lives. This doesn’t, however, show how much difference you make by becoming a doctor. Working this out requires a number of adjustments. The first is that we need to work out the impact of additional doctors, instead of the average doctor.

                        There’s already about 200,000 doctors in the UK. By becoming a doctor, let’s suppose I increase the number of doctors to 200,001. And let’s assume that all doctors in the UK are equally skilled (we’ll relax this assumption in the next post). The extra doctor won’t produce a benefit of 2600 QALYs. That’s because doctors perform a huge variety of tasks. Some of these do more for the UK’s health than others. The NHS (to some extent) prioritises its distribution of resources so that the most effective tasks get done first. This is part of the remit of the National Institute of Clinical Excellence. So, if there’s one extra doctor, the tasks they do will be less effective than those that are already being done. So we’d expect an additional doctor to have less impact than the 200,000 people who are already doctors. This is called diminishing marginal returns.

                        How can we take the figure for the average impact of a doctor and work out the impact of an additional doctor? One very rough way of estimating this is to look at the maximum the NHS is prepared to spend to save on QALY, and compare it to the average it spends. We know the maximum the NHS is willing to pay for a QALY is between £20 000 to £30 000. This suggests that extra money given to the NHS produces about 1 QALY per £25,000. Assuming that the NHS can freely spend between salaries and technologies and is fairly good at working out when it is more effective to spend money on either, then the marginal benefit of a doctor per year is their salary (£69 952) divided by the cut off (£25 000), making for 120 QALYs over a career (1). However, this assumption is unlikely to be true. So let’s try some better approaches:

                        Approach 1: Life expectancy versus number of doctors.

                        Plotting the most recent values the world bank has (Source), we get the following plot:

                        We see an ‘r’ shaped sort of trend: the life expectancy initially goes up briskly with an increase in doctors per capita, but this relationship levels off as the doctors per capita increases beyond 1 per thousand or so. So it looks like there are diminishing returns of adding more doctors. A similar picture emerges when other sorts of ‘investment’ in health are considered: see for example the plots of health spending per capita versus life expectancy, or GDP per capita versus life expectancy.

                        From here, we can begin to work out the marginal impact of adding ‘one more doctor’ to the UK: one adds a best-fit line to the data, and see what how much ‘gain’ there is in health when you move ‘one more doctor’ along that line from where the UK is now. This gives a final answer of 950 QALYs per medical career, just over a third of our original estimate for the impact of a doctor. (I’ve included the working out in a footnote for the interested (2).)

                        This method of estimation isn’t perfect. It looks at all countries, whilst we might want to use data only from developed countries to look at the impact of doctors in the UK. Much like our previous post, wealth remains a potential confounder: both life expectancy and doctors per capita correlate with gross domestic product, and it might just be that richer societies are able to buy better education, hygiene, nutrition, and other things that really do the work of making their inhabitants healthier, and they coincidentally buy more doctors too.

                        Approach 2: OECD health data and regression

                        We can try a different tack to try and accommodate these concerns. The OECD is a group of wealthy to fairly wealthy countries which maintain records of themselves along a variety of indicators. Amongst these are life expectancy, healthcare spending per capita, doctors per 1000 people, and gross national income per capita adjusted for purchasing power parity. We can regress these to a combined model to work out how large a contribution each of these factors make to improved life expectancy, and, once again, we can then work out how big an effect changing one of our variables (doctors per capita) by ‘one more doctor’ will change the life expectancy of the population. The final answer here is that the impact of one more doctor is around 670 QALYs. (Again, the full working in a footnote (3).)

                        Approach 3: WHO disability adjusted life years

                        A lot of comments in the previous post were worried about quality of life, and not just length. Although I tried to account for this by Bunker’s estimates of how much good medicine does via removing disability, it would be nice to tackle the issue more explicitly.

                        The WHO keeps data on the burden of disability in a population, as DALYs per 100 000 people (a DALY is a measure of length and quality of life, it is the inverse of a QALY – more DALYs are a bad thing, as well as other differences summarized here). Plotting DALYs per 100 000 against doctors per 100 000 gives the following:

                        This graph looks like a mirror image of the life expectancy versus doctors per capita graph above. Although we cannot directly compare ‘DALYs averted’ to ‘QALYs gained’, using a similar technique to approach 1 (draw a line of best fit, work out how much gain one makes by moving ‘one more doctor’ along, and multiply appropriately) means each doctor averts 645 DALYs per career. This reassures us our figures are on the right track.

                        Conclusion

                        These estimates are necessarily very rough, though it’s reassuring to find our three estimates in the same ball park. Splitting the difference between our three best estimates gives the impact of ‘one more’ medical career in the UK as about 760 QALYs, around a third of our estimate of the average doctor. Looking at the degree of noise in the data, I estimate the 95% confidence interval is about 600 – 920 QALYs.

                        The expected impact of becoming a doctor is now around 25 lives: still pretty good, but giving 10% to effective charities can produce a health benefit 25 times larger than that. This underlines the importance of thinking at the margin for those wanting to make the biggest difference they can. One should try to estimate not how much good a career does in general, but how much more good they can do if you get involved. In the case of first world medicine, it appears most of the highest priority interventions for improving health and wellbeing have already been done, and so the additional impact of one more doctor is not that large.

                        Our estimate, however, is still too generous. By becoming a doctor I won’t increase the number of doctors by one. Rather, it seems I’ll just take the job from someone else. I’ll be replaceable. We’ll look at this adjustment in the last post.

                        See part 1 which finds and upper bound

                        See part 3 on replaceability


                        You might also be interested in:


                        References and Notes

                        (1) Full working: £69 952/year * 43 years / £25 000/QALY = 120 QALYs

                        (2) First, we need to find the best fit relationship between number of doctors and life expectancy. The best candidate for this is a hyperbolic curve: it seems plausible there will be a ceiling on how far life expectancy could rise through adding more doctors, even in the limit case of a population comprised entirely of doctors treating each other.

                        This graph is the same as the first, save we have shifted down 47 units – the amount we estimated earlier would be the baseline of no medicine, and so the values on the Y-axis are ‘added years’ of life expectancy. The hyperbola is given by the dashed blue line. Now we have our trend, we can work out the expected impact of moving the UK from its current doctors per capita (2.743 per 1000) to the value it would have with one extra doctor.(1)

                        The equation of our best fit line is given by:

                        Added life expectancy = 30.79459*(Doctors per capita)/(0.16801+Doctors per capita) 
                        

                        So plugging in the difference between our current doctors per capita and the ‘one more doctor’ case:

                        Marginal change = 30.79459*(2.743016)/(0.16801+2.73016) - 30.79459*(2.743016)/(0.16801+2.743016) = 9.76877 * 10^-6 years.  
                        

                        So the marginal impact of one more doctor in the UK will raise UK life expectancy by just under one ten-thousandth of a year. Putting this change into added years of healthy life requires us to multiply by the population of the UK, as well as a correction factor due to our [prior estimate] that for every 9 years of lifespan medicine adds, it adds another 5 years of healthy life via freedom from disability.

                        Marginal QALY yield per doctor = 9.76877 * 10^-6 * 62,641,000 * 14/9 = 950 (2sf)
                        

                        (3) We can regress these data to a linear model, such that:

                        Life expectancy = k1 + k2*(GNIPPP) + k3*(Doctors per 1000) +k4*(Healthcare spending pc)
                        

                        Where k1, k2, k3, and k4 are constants. The best fitting model (adjusted R-square 0.32, P=0.002)

                        Life expectancy = 75.336 + 0.0000291*(GDIPPP) + 0.433*(doc/1000) + 0.000886*(healthcare spending pc)
                        

                        This model explains about a third of the variance (adjusted R-square = 0.32), suggesting the main determinants of health in wealthier countries are not wealth, nor spending on healthcare, nor number of doctors. However, of these three it is health spending that is the largest factor, and the effects of either GNIPPP or doctors per 1000 population are negligible – neither are statistically significant, and 95% confidence intervals for either variable cross zero. In other words, we cannot be that confident, on the basis of this analysis, that increasing the number of doctors per capita increases life expectancy at all.

                        Our best estimate of changing doctors per capita given by the 0.433 coefficient – our central measure. From this we can work out the marginal impact ‘one extra doctor has’ by the similar procedure to before:

                        0.433*0.0000160 extra doctors per capita = 6.91*10^-6 years in added life expectancy
                        6.91*10^-6 * 62 641 000 * 14/9 = 673 QALYs