Stay up to date with Giving What We Can's blog. Featuring the latest news about our community, how to donate effectively, charity evaluation, community events and more!

Blog

Giving What We Can's blog, featuring the latest news from our community.

Donate

Pledge

footerNavigation

About us

Contact us

Our team

Our values

Our members

Our events

Transparency

Our mistakes

Our impact

Donate to Giving What We Can

Careers

Leave feedback or make a complaint

About

Our best charities

Take a giving pledge

Get involved

Make a bequest

Videos, books, podcasts

FAQs

Learn more

Press kit

Media coverage

Case studies: People who pledge to give

Media

Footer Navigation

topNavigation

Our research team’s 2026 recommendations

How we choose which charities to recommend

Other places to donate

Create a fundraiser

Best charities

The 10% Pledge

The Trial Pledge

Pledging vs. donating

View all pledgers

Share your pledge to inspire others

Motivations

Stories

Giving pledges

About our research

Browse our guides and resources

Connect with our community

Read our blog

Help others discover effective giving

Birth Lottery

How Rich Am I

Research and resources

Who we are & what we do

Our structure

Mistakes

Top Navigation

Bastian Stern

The $/DALY Illusion: Part II - Averaging

In the first post, I discussed how we can be misled when we compare improvements in cost-effectiveness between organisations whose cost-effectiveness is reported in $/DALY.

The $/DALY-DALY/$-contrast also becomes important when we start talking about averages. For instance, suppose that you want to consider how much better it is donate to, say, charity D, compared to spending equal proportions of one's money on charities C and D(1). In this case, you would want to compare C and D's average cost-effectiveness to the cost-effectiveness of D.

To obtain C and D's average cost-effectiveness, it is tempting to simply take the familiar arithmetic mean(2) (which is what we typically mean by "the" average) of the figures for C and D's cost-effectiveness in $/DALY reported in the table in the previous post (20 $/DALY and 10$/DALY respectively), to obtain (20 $/DALY+10 $/DALY)/2= 15 $/DALY

However, this would deliver the wrong result. To see why, suppose I donate $10 dollar to each of C and D. I would thereby avert $10/(10 $/DALY)+$10/(20 $/DALY)=1.5 DALY . Given that I donated $20 in total, the average cost-effectiveness of my donations would be (15 DALY)/$20=0.075 DALY/$, or, reported in $/DALY, ~13 $/DALY.

The reason for this in general is this: Suppose we want to average ratios (e.g. cost-effectiveness reported in $/DALY) across a number of instances (e.g. different charities, whose cost-effectiveness the ratios measure). In that case, the arithmetic mean must be used to average the ratios only when the denominator is held constant and the numerator variable across the different instances. For instance, this is the case when the ratios are reported in $/DALY, the amount of DALYs averted by each charity is meant to be held constant and the amount of $ donated to each charity variable . Accordingly, when we naively took the arithmetic mean between C's and D's cost-effectiveness reported in $/DALY, we were effectively calculating the effectiveness of splitting one's money between C and D in such a way that the amounts of DALYs that are averted by our donations to each of these charities are held constant (which is clearly not something we are typically interested in). But what we instead wanted to hold constant across instances was the amount of money that flows into the charities .

This problem can be remedied by taking the harmonic mean(3) rather than the arithmetic mean of C's and D's cost-effectiveness reported in $/DALY, i.e. 2/(1/(20 $/DALY)+1/(10 $/DALY))=~13 $/DALY.

In general, when we are averaging ratios across a number of instances, we have to use the harmonic mean exactly when the numerator is held constant and the denominator variable across the instances, e.g. when, as in our case above, the ratios are reported in $/DALY, the amount of $ donated to each charity is meant to be held constant across instances and the amount of DALYs averted by each charity is variable.

Note that effectively, taking the harmonic mean amounts to converting the cost-effectiveness estimates for C and D into DALY/$, taking their arithmetic mean, and then converting back into $/DALY. Accordingly, had we reported C's and D's cost-effectiveness in terms of DALY/$ at the outset, we would not have been misled in using the straightforward arithmetic mean.

What the considerations in post I and this post suggest is that it is usually wise to report cost-effectiveness in terms of DALY/$ rather than $/DALY, because then

the comparison of differences in cost-effectiveness that we are about are more conspicuous, as the differences could be obtained by simple subtraction and;

the most often relevant averages of cost-effectiveness could be reached simply by taking the familiar arithmetic mean, rather than the more cumbersome harmonic mean.

Accordingly, it would be helpful if it became conventional to report cost-effectiveness in terms of DALY/$ rather than $/DALY. Until this is the norm, we must take care not to be misled when averaging and comparing differences in cost-effectiveness estimates.

(1) Or similarly, that you are interested how much better D is then choosing between C and D at random, with an equal probability of choosing either. (2) The arithmetic mean of n values a1, a2, a3, … is given by ((a1+a2+⋯+an))/n. (3) The harmonic mean of n values a1, a2, …, an is given by n/(1/a1+1/a2…+1/an).

Julia Wise

When is your help special?

I've heard the argument that we should "think globally, act locally" because we understand the needs of our own communities best. I'm willing to accept this for some situations.

I think it boils down to where your special help would be useful. If you pass a car accident, yes, your physical presence means you have a unique ability to help.

Likewise, when it comes to personal interactions, people are not interchangeable. Getting a card or letter from a stranger is not so comforting as hearing from someone you love. We evolved to interact with people we know in real life, and this still satisfies us more than some abstract kindness from a stranger.

So recognize the areas where you can be uniquely helpful: being kind with your family and friends. Sudden emergencies where you are physically present. Being a good neighbor.

But here's where I think people go askew with this logic: they feel that financial help should also work this way. After all, don't I understand the needs in my own community better than anyone? So I should fund projects in my own community, and other people should take care of theirs.

But rich people live in communities with other rich people, and poor people live near poor people. Your average American probably has several relatives or neighbors who have a few thousand dollars in their bank accounts. Your average Nigerian does not know any such people. When both rich and poor people give within their own communities, the opera gets a lot more funding than the maternal health clinic in Nigeria.

Of course, lots of first-worlders have given misguided aid because they misunderstood the needs of people in other countries. But you can misunderstand the needs even in your own community.

Last summer when I was studying in Ecuador, several of my fellow travellers volunteered in a local orphanage. The place was in bad shape, and it was hard to see the kids living in deprivation without wanting to do something. So some Americans decided to raise money for the orphanage from their friends at home. After all, they saw these kids every day and had seen the need first-hand.

Except they found out the reason the place was falling to pieces was that the owner was embezzling money. Those donations would probably never have reached the kids. Americans weren't alone in being duped by this guy. Every Christmas he had a big fundraiser and lots of Ecuadorans gave money, clothes, and toys. The locals were just as misinformed, despite it being in their own community, because they hadn't seen the financials.

This can happen in any country. At home in the US, I saw a lot of misguided community help during the summer I worked in a domestic violence shelter. We had a garage overflowing with blankets, because people had this idea that battered women need blankets. What they actually needed was a child's car seat so the children could safely ride to doctors' visits. The fact that donors were local didn't give them special insider knowledge about the residents' needs. It would have been better for them to donate cash so that we could use it for what was most needed (the car seat, not more blankets).

There are good and bad charities working in all parts of the world. Find ones that will use your money well, and that are doing important work. (And if you live in a rich part of the world, the greatest need probably isn't local.) Then donate money, which will help more than your blankets, old clothes, or volunteering.

And then find someone you love and give them the hug or the kind word that only you can give.

Rob Wiblin

Our research goals for the next few months

Giving What We Can recently produced a strategic plan for the next six months, including for its research. This plan has been discussed and improved after meetings with an external advisory committee and our trustees.

We are blessed with a wide range of promising research questions, which means the challenge has been picking out the best ones. Those we considered properly are listed in full at the bottom of this post.

Some highlights for what we hope to achieve over the next two to four months are:

 to coordinate and build our growing research team, and encourage greater autonomy among participants. This will complement a new discussion board for members.

Produce and run several training sessions for volunteers to help them understand the approach we use, including its strengths and weaknesses.

the strengths and weaknesses of existing analyses of climate change

making estimates when you are uncertain about many things

why we don't think we should have any 'pure time preference'

the expected returns on financial investment, and others valuable additions to our advice on 'giving now vs giving later' (yet to be determined)

the kinds of strategies that provide 'leverage', their likely strengths and weaknesses, and how they can be identified.

write on whether we should place less weight on the death of infants relative to older children

finish, or at least start, a survey of the meta-charities, like GWWC, that exist including their methodologies and recommendations

Revise our education, TB, hygiene and sanitation, migration, climate change and maternal health pages, and upload new pages on biomedical research and meta-charity.

Update the data and method used in our 'how rich are you' calculator to accommodate the latest research.

Maintain the flow of at least 1 research focussed blog post each week.

Come to a conclusion on whether to recommend 

, and Hilleman Laboratories, or another two organisations who are willing to participate.

Explore how we could help find funding for WHO-CHOICE, should they require it to complete a major piece of research, and also explore a possible collaboration with 

As this list suggests, GWWC has decided to experiment with fewer in-depth evaluations of charities over the next 3-6 months. There are several considerations behind this:

Such projects are highly time consuming and difficult to do 'part time';

We have identified a number of crucial methodological or theoretical questions which we expect will provide a higher return for each hour we invest in them, and which are more suited to our skills;

Thanks to the extensive work done by GiveWell in investigating two of our current recommendations, it will be challenging to find a charity that is reliably better with our current resources.

Nonetheless, a significant number of charities which use potentially high-leverage approaches have still never been analysed for cost effectiveness or room for more funding, so we intend to continue surveying these options. Having done so, we could either investigate further ourselves should a very compelling option showed up, or encourage another organisation to do so. Within four months we will reassess whether this reorientation is delivering the results we hope.

I hope this provides some insight into where we hope to go in the medium term. These plans remain flexible depending on what seems valuable and possible, so get in touch if you would like to offer feedback.

And with that, I wish you a Merry Christmas and a rambunctious New Year's celebration. I look forward to returning full of energy for further fascinating research and I hope you also look forward to either participating in or reading it!

Some extra details on the projects we considered

We talked through 11 different research areas, and rated them on the following criteria:

How valuable is the resulting information?

How difficult will it be to make progress?

How much will working on this topic help to grow Giving What We Can, and other projects in the effective altruist movement?

How personally interested are we in this topic?

Is this a prerequisite for other research we want to do?

Given the above, do we think we should be working on this topic now?

The topics are below, with those we rated in 

 in neglected areas such as we have attempting with ACTION and others.

simple analysis of previously unconsidered charities

 in order to see if they are worth us or someone else investigating in more detail (similar to Back of the Envelope Guide to Philanthropy)

What spillovers are there from policies in rich countries to poor countries?

Long term impacts and the astronomical waste

Requires thinking about the internal rate of return on time from various kinds of spending (social return, financial return, movement building return, information gathering return)

Short paper: Discount rate issues - write a paper on this with a view to influencing DCP3

Meta: How valuable is meta-charity, or even meta-meta-charity, research? Do any such organisations have room for more funding (RFMF).

Look around for evidence of whether this work is influential

How can we weigh e.g. getting a better education against health; or more money versus education; or having to move due to climate change to health; etc. What do we know about the utility impacts of these things?

Short paper: How to value life at different ages, in particular around infancy?

Meta-meta: How to identify awesome projects

Where should we expect to find groups with RFMF?

Dealing with cumulative uncertainty and regression to the mean (with a view to influencing DCP3)

Pascal's mugging, fat tails and 'weirdness theory'

What forms of leverage exist and are they illusory?

Any other issues associated with how to identify causes out on the far right tail of effectiveness that otherwise wouldn't get done

Meta-meta: Write a report into meta-charities, evaluating their methodology and quality of their recommendations

Market research to increase recruitment of new volunteers/interns/members

What could make the public more receptive to our message?

William MacAskill

Some General Concerns About GiveWell

Note: This article hasn't been updated recently. We're keeping it up for the historic record of the debate, and because people still seem to refer to the arguments.

In a previous post, I explained why I was surprised by GiveWell's recent recommendation of GiveDirectly. In this post, I'll take the opportunity to explain some other general concerns I have with GiveWell. As before, I'll note that these just represent my view, rather than those of GWWC as a whole, and that such criticism should be taken as a sign of respect for an organisation which I generally think very highly of. We are making these suggestions in part because we draw heavily on GiveWell's own research and want it to be as compelling as possible.

I'll also note that many of these concerns can also be applied to GWWC itself. This list therefore also represents a list of improvements we would like to make to our own research and publications in future.

Different countries vary dramatically in their rate of economic growth. In 

, India's economy grew by 7.8%, whereas Namibia's economy grew by just 3.8%. In the long-run, insofar as your donation generates some jump in the productivity of the labour force, compounding effects will cause it to snowball into substantially more income if used in India than in Namibia. On this basis one might argue that donations to India constitute a better 'investment' than donations to Namibia. On the other hand, one might reason that, because we should expect people to be richer in India in the future, they will get less benefit from any extra income in the future. The economic trajectory of the country that you try to help could potentially be a 'crucial consideration' concerning which charity you should give to, but it hasn't yet been addressed by GiveWell.

This is not even considering the magnitude and persistence of an intervention's impact on economic productivity. This is another crucial consideration which we understand GiveWell has looked into and found very difficult to answer.

As far as I know, GiveWell haven't commissioned a thorough 

 of their recommendations. (GiveWell have done some external evaluations in the past, but mainly in 2010 or 2011, and mainly done by volunteer undergraduates or recent graduates.) This surprises me. Whereas businesses have a natural feedback mechanism, namely profit or loss, research often doesn't, hence the need for peer-review within academia. This concern, when it comes to 

, is even greater. If GiveWell's analysis and recommendations had major flaws, or were systematically biased in some way, it would be challenging for outsiders to work this out without a thorough independent evaluation. Fortunately, GiveWell has the resources to, for example, employ two top development economists to each do an independent review of their recommendations and the supporting research. This would make their recommendations more robust at a reasonable cost.

To put it simply: If GiveWell's top recommended charities are so good, why hasn't their room for more funding been used up?

, because the best ones tend to be snapped up quite quickly. But they have provided very public information on their top-recommended causes, some of which have been on the list for years. And we know that major philanthropic foundations (such as the Gates Foundation, and CIFF) are aware of their research. So it's surprising, in their view, that the room for more funding for their top-recommended organisations wasn't used up rather more rapidly. This either suggests that their model of how philanthropy works is wrong, or that their recommendations aren't believed by the big players in the 'philanthropic market'.

Even if we accept that their model of an efficient market in philanthropy is inaccurate, and so could attribute the failure of big foundations to fund their recommended charities to irrationality or different values, there remains cause for concern. GiveWell functions in an advisory capacity to GoodVentures, which has very similar aims and attitudes to those of GiveWell. GoodVentures' endowment could easily use up the room for more funding at GiveWell's top charities. So it raises the question why GoodVentures doesn't donate to GiveWell's top charities: even if not exhausting their room for more funding, at least using up a considerable proportion of it. The most plausible explanation in my mind is that GiveWell thinks that GoodVentures can better spend its money at a later date after further research. In which case, it's hard to see why individual donors shouldn't save and donate later, or even try to give to GoodVentures. Perhaps that's wrong, but it's worthy of discussion.

GiveWell claim that they want to maximise good done per dollar invested, but they haven't explained what they think 'good' consists in and why. Do future people count? If so, how much, and how far into the future? Do they have a discount rate? How much priority should they give to the worst off? Is an infant death worse, or better, than the death of a 20 year old? How do they weigh improvements in health against extensions in health? What do they think is the relationship between economic growth and wellbeing?

These are just of theoretical interest - varying any of the parameters could lead to different charitable recommendations. For example, when they discuss AMF's cost-effectiveness, they focus on under-5 mortality. But they don't discuss the fact that most people think that an infant death is not as great a loss as a young adult death; an intuitive view that has good theoretical grounds. (This is something that will likely be taken into account in the next generation of the DALY metric, in the third edition of the Disease Control Priorities Project.) Depending on how, exactly, one thinks one should weigh saving infant lives against saving adult lives, this issue could make AMF several times less cost-effective relative to SCI than it is currently judged to be. This one methodological issue could therefore dramatically alter their recommendations.

To their credit, GiveWell has moved towards helping people to tailor their choice to their values, by comparing the cost of bed nets delivered, children dewormed, and cash delivered. This can be done using spreadsheets and explanations provided in some recent 

. In its current form this requires more work than most donors would be willing to engage in. Making this easier, and guiding people through discovering the implications of their moral values would be worthwhile improvements. This kind of tailored advice is only possible for the minority of charities GiveWell has looked into in great detail, which means it is still important to be explicit about the value judgements that generated that shortlist.

Once they allow the moral view or 'judgment calls' of the donor to be an input, providing a #1 #2 #3 ranking of charities becomes misleading. For example, I wrote previously that I was surprised by their recommendation of GiveDirectly. Most of my surprise would have been alleviated if, rather than ranking them #2, they gave them a conditional recommendation: "If you have such-and-such moral view, then GiveDirectly is an excellent option." Indeed, there are many reasonable views based on which giving to GiveDirectly could be the best option. Perhaps you do not want to be paternalistic towards the poor, even if being paternalistic would do more to improve their wellbeing in the long run. Or perhaps the reason you give is because you think you should redistribute your wealth as a matter of justice, and it's not your right to say how that money is spent. In either case, it's very useful to know that one has a good option to put wealth directly into the hands of the poor. But, if that's the motivation for recommending GiveDirectly, ranking them above SCI and below AMF disguises the fact that it is being judged on a different set of criteria.

The $/DALY Illusion: Part I - Comparing Improvements

Imagine the following situation: Two of your friends, Jack and Jill, are each about to donate $1000 to two different charities. You know that both of them are on their way to their closest letter box, with the intention of posting their cheques to those charities. Unfortunately, neither of them has heard of the idea of cost-effective giving, and you know that you only have time to call up one of them convince one of them of directing their donation to a more cost-effective charity instead before they have reached the letter box to and it is too late to do that. You know that is costs $1000 to avert a DALY by donating to Jack's preferred charity, A, and that the best charity you could convince him to donate to instead, B, would allow him to avert a DALY for only $50. Jill's preferred charity, C, allows her to avert a DALY for $20, but you know of an even better charity, D, that would allow her to avert a DALY for only $10, and you have reason to believe that you could convince Jill to donate to D instead.

A tempting first guess is that you should call up Jack: The difference in A's and B's cost-effectiveness measured in $/DALY (1000 $/DALY- 50 $/DALY=$950 $/DALY) dwarfs that between C and D (20 $/DALY-10 $/DALY=10 $/DALY). (Call this the "linear method" of evaluating the relevant differences.)

A tempting second guess would be to look at what we might call the proportional improvement in cost-effectiveness of the changes from A to B and from C to D. On the most natural understanding of this approach, the proportional improvement between A's and B' cost-effectiveness measured in $/DALY is (1000 $/DALY - 50 $/DALY)/(1000 $/DALY)=95% and the proportional difference in C's and D's cost-effectiveness measured in $/DALY is (20 $/DALY - 10 $/DALY)/(20 $/DALY)=50%. Again, this would seem to strongly suggest that calling up Jack is much better than calling up Jill. (Call this the "proportional method" of evaluating the relevant differences.)

In fact, simple calculation shows that both lines of reasoning are mistaken. By donating the $1000 to A, Jack would avert 1000/1000=1 DALY, whereas donating to B would allow him to avert 1000/50= 20 DALY. This means that calling him up allows you to avert 19 additional DALY. In contrast, donating her $1000 to C would allow Jill to avert 1000/20=50 DALY, whereas by donating to D she could avert 1000/10=100 DALY. This means that calling her up allows you to avert 50 additional DALYs. Accordingly, calling up Jill allows you to have more than twice the positive impact!

What went wrong when we made our first and second guesses, which turned out to be so radically misleading? The answer is that when you are interested in how much good you can do by redirecting a given amount of money between different causes, you are interested in how many more DALYs you can avert by giving to one cause rather than another, i.e. differences in their cost-effectiveness reported DALY/$.

Where "x" is the cost-effectiveness of charity X reported in $/DALY, and "y" the cost-effectiveness of charity Y reported in $/DALY, the improvement in cost-effectiveness in moving from X to Y reported in terms of DALY/$ is given by (1/x)-(1/y), or, equivalently, by (y-x)/xy (which, as you can check for yourself, delivers the correct result in the case above, namely that the improvement in moving from C and D is 50/19≈2.6 times greater than that between A and B.)

This also shows what was wrong with our first two guesses. Our first guess was to evaluate the improvement from X to Y's cost-effectiveness as y-x (which I called the linear method), our second guess as the proportional improvement between y and x, i.e. as (y-x)/y (which I called the proportional method). However, these latter two functions evidently fail to be related in a strictly increasing way with (y-x)/xy (i.e. fail to be related in such a way that whenever (y-x)/xy increases, these latter functions increase, too), which is why both the linear and the proportional methods give the wrong answer.

The linear method misled us because it simply ignores the denominator of (y-x)/xy, xy, altogether. The proportional method misled us because it omits part of the denominator of (y-x)/xy, namely x. Both methods systematically overvalue improvements in cost-effectiveness between less effective charities (like A and B), and, by the same token, systematically undervalue improvements in cost-effectiveness between especially effective charities (like C and D).

Note that had we reported the cost-effectiveness estimates in DALY/$ from the outset, these complications would not have arisen.

Also note that there are cases in which straightforwardly taking the differences (i.e. using the linear method) does tell us what we are interested in. For instance, imagine that both Jack and Jill want to avert exactly one DALY (rather than spending a fixed amount of $1000), and your primary interest is to advise them so that they save as much money as possible. In that case, it really is the case that x-y is what you should be interested in. By averting 1 DALY by giving to B rather than A, Jack can save $950, and by averting 1 DALY by giving to D rather than C, Jill can only save $10, which would suggest that, in this altered scenario, you should call up Jack.

However, whilst this latter sort of situation is what we typically encounter when making day-to-day purchases (e.g. when you want to buy exactly one chocolate bar, and you are interested in comparing how much money you can save by choosing between various shops), it is clearly not the sort of situation donors are typically faced with in the context of charitable giving. Donors are usually interested in the differences in how much good they can do with a given amount of money, rather than the differences in how much money they can save for a given amount of good they want to do.

In the next post, I will discuss complications the $/DALY metric gives rise to when we talk about cost-effectiveness averages.

[1] The phenomenon I am about to describe is analogous to the well-known MPG Illusion: People tend to be misled about the value of improvements in car fuel efficiency when efficiency is reported in miles per gallon (instead of a unit that allows us to state efficiency in terms of volume per distance traveled, e.g. liters per 100 kilometers). Reporting efficiency in terms of MPG also gives rise to complications associated with evaluating average efficiency analogous to those I discuss below.

[2] The Disability-Adjusted Life Year (DALY) is a metric that is commonly used to quantify and compare the benefits of different interventions. DALYs are a measure of lost health due to illness. One DALY represents the equivalent to losing a year of life at full health. Each illness has a disability weight assigned to it, representing the amount it reduces a person's quality of life. For example, a year of life spent with blindness counts for 0.6 DALYs.

[3] I take this to be "most natural" given that basing one's comparison of two improvements in cost-effectiveness on the comparison of the corresponding reductions in cost per unit effect (which this "proportional method" amounts to) appears to be the best analogue of a mistake some people make when asked to compare the improvements in car fuel efficiency reported in MPG. In the latter case, people sometimes (mistakenly) think that, where the improvement comes about by moving from car A to car B, they need to calculate the "proportional improvement" given by (Efficiency(B)-Efficiency(A))/Efficiency(A), and compare improvements by comparing "proportional improvements". 

The Way It Was

In thinking about problems that currently affect developing nations, I try to remember that the United States 

Malaria once plagued the American south and Midwest. It's the reason English colonists abandoned the Jamestown, Virginia settlement for somewhere with fewer mosquitoes. In 1946, the Centers for Disease Control were formed to fight malaria. Five years later, malaria was eradicated in the United States. (

, Laura Ingalls Wilder describes her family's experience with the disease in 1870: 

Laura tried to get up, but she was too tired. Then she saw Ma's red face looking over the edge of the bed. Mary was all the time crying for water. Ma looked at Mary and then she looked at Laura, and she whispered, "Laura, can you?" "Yes, Ma," Laura said. This time she got out of bed. But when she tried to stand up, the floor rocked and she fell down. . . . She knew she must get water to stop Mary's crying, and she did. She crawled all the way across the floor to the water-bucket. There was only a little water in it. She shook so with cold that she could hardly get hold of the dipper. But she did get hold of it. She dipped up some water, and she set out to cross that enormous floor again.

It drives me crazy when people say child mortality is saving us from environmental disaster. Do they wish that for their own families? Do they feel they have too many siblings, too many cousins, and we really ought to get some contaminated water or intestinal parasites around here to deal with the problem?

No one wants that for their family. We want low child mortality and reasonable family size. Developed nations made that transition, but it took a while.

Hans Rosling does an excellent job at making public health statistics understandable to non-statisticians. I recommend his talk on "

", especially for its comparison of child mortality rates across nations (at 8:40). Rosling notes that his home country of Sweden had a high child mortality rate in 1800, twice as high as anywhere now. Over time, with better education and better public health, the rate declined to its current low level. Most countries are currently on this journey of decreasing family size 

 child mortality, many of them progressing faster than Western countries ever did. (The site 

 lets you play with the charts yourself.)

While it's clear that large family size is correlated with child mortality, I'm not as convinced as Rosling is that lower child mortality is the main 

 of smaller families. I gather it's a combination of fewer child deaths, female education, access to birth control, and urbanization.

The Ingalls family were typical 19th-century Americans - they had five children, one of whom died in infancy. I come from a typical 20th-century American family with two children, both still alive. I'm 

 we made this demographic transition. My hometown no longer has malaria. I'm educated, I drink clean water, and I control my own fertility. When I choose, I will raise one or two children who will probably also be quite healthy.

We've come a long way. Now I want this life for everyone.

Can we evaluate political advocacy?

', which advocates for greater government spending to tackle tuberculosis (TB). Research volunteers Nick Beckstead, Rossa O'keeffe-O'donovan, and Gordon Irlam of the 

Back of the Envelope Guide to Philanthropy

, all deserve a great deal of credit for their help with this work. We also appreciate the assistance of the folks at RESULTS, in particular Jennifer Maurer, for providing detailed feedback on an earlier version. Whether or not RESULTS ultimately beats our current recommended charities, they are working hard to make a difference on crucial issues in global health.

This is the first time we have released our preliminary research on a topic in this way, and quantifying the good done by political advocacy raised novel challenges for us. We not only have to establish whether the advocacy efforts have generated additional spending, but also try to work out what the additional money has ultimately been spent on and whether it is likely to have worked. Unlike when we consider the Against Malaria Foundation, we cannot just request the relevant data from the charity delivering the programme. Nonetheless, I am excited that we are stepping beyond the more familiar bounds of direct health interventions. Many members have requested that we do so, and if we can identify the right campaigns to get behind, political advocacy offers the potential to move large sums towards programmes that really work to help the world's poorest people.

We are hoping that posting this document will encourage people with experience in the following areas to talk with us:

• Evaluating the effectiveness of political advocacy;

• Experience with RESULTS/ACTION or similar organisations, and their influence in Washington;