Some facts about charity are so useful that they just have to be shared. Here’s
one from the website of Giving What We Can:
‘It is not even a matter of some charities being 10 or 100 times as
effective: even restricted to the field of health programs in developing
countries, research shows that some are up to 10,000 times as effective as
By reading this, most of us will have gained some motivation to give
effectively, and this will deliver years of healthy life to those in need of
Just as we’re grateful to Giving What We Can for this help, it’s natural to
wonder what we can do to nudge others towards cost-effective philanthropy.
First, I will evaluate whether philanthropic influencing can be effective in
principle. Then, I’ll describe the classes of philanthropic influencing to get
us thinking about what philanthropic influencing means in practice.
To discuss cost-effectiveness, we will use a point of comparison. Let’s put
ourselves into the shoes of Jane. Jane is a professional donor, someone who
earns money and donates it as cost-effectively as possible. If Jane earns a
typical American income for forty years and donates 10%, she can expect to save
44,000 years worth of healthy life .
One day, however, Jane makes a change. She cuts down her hours at her
relatively well-paid job so that she can no longer afford to donate in order to
spend more time persuading other people to donate professionally. As a
philanthropic influencer, will she save more years of life or less?
Well, suppose at the end of one month, she has convinced one person to donate
their life’s income to charity. At first glance, this would seem like something
of a failure. However, if this person earns a similar salary to Jane, and will
donate as much as Jane would have for the same amount of time, then Jane’s
single act of persuasion has saved 44,000 years’ worth of healthy life. She has
already broken even. If she completely convinces three people about effective
philanthropy, she will have saved 132,000 years’ worth of healthy life.
Clearly, her impact across 40 years will be tremendous. In this simple
calculation, although a professional donor can change the world a great deal, a
philanthropic influencer can change the world by many times more .
Our actual situation is slightly more complicated than Jane’s. Our potential
audience will probably not donate as much, as cost-effectively, or for as long
as we will. We probably shouldn’t give up donating either. After all, if we
continue to donate, we will be practicing what we preach, which will make us
In the more complete calculation, there is a large margin of error, but
philanthropic influencing promises to be cost-effective . Note that for an
influencer whose income is significantly larger than their audience’s, the
break-even point is harder to reach .
Philanthropic influencing, then, promises to be highly effective. However, in
general, people are not interested in philanthropy. One cannot simply stand on
a soapbox by a local shopping centre and begin to preach the merit of donating!
We owe it to ourselves, to any potential donors and to their beneficiaries to
be far more deliberate than that. It is time to discuss what classes of
philanthropic influencing there are and how well they might work. I will cover
these in order:
- Organised philanthropy
- Popular philanthropy
- Political philanthropy
- Funding philanthropic influencers
Let’s begin with something the founder of Giving What We Can, Toby Ord said to
an audience of Oxford students last year:
‘I was surprised to find that, because I’d actually set up an organization to
look at [poverty], within a couple of years, we’d got $30 million worth of
pledges, which is more than my lifetime’s earnings. And it’s eminently
possible for people in this room to help me out with this kind of thing –
they could also produce more wealth going to help the poorest people by
convincing others [to donate] than their entire lifetime’s earnings.’ 
This pledged total has since risen to $44 million dollars, an impressive figure
for an organisation founded as recently as November 2009. So how does Giving
What We Can attract this money? Well, it begins with the pledging system.
Pledging is not a legal commitment, but by pledging, people make it less
attractive for themselves to stop giving. Pledging also permits everyone to
make their donations public. This is preferable to a tradition of donating
privately and anonymously because when we perceive that others give, we all
give more .
By generating publicity, Ord further encourages effective giving. So too does
the Giving What We Can community, by informing, supporting and encouraging each
other. As Ord says, we can all help. By making a pledge, by joining the Giving
What We Can community, by volunteering and by founding new branches, we can
potentially add many millions to this pledged total. Giving What We Can is not
the only organised philanthropic effort,
either, and one’s impact may be more or less elsewhere.
Although Peter Singer’s work bears some similarities to Ord’s, it mostly seems
to belong in its own class of popular philanthropy. Peter Singer’s website,
also founded in 2009, is based on his book, The Life You Can Save. There,
Singer has assembled $61 million of pledges. These pledges are yearly, not
lifelong. So if Singer’s pledgers renew their commitments five times on
average, they will donate a sum nearer to $300 million . This adjustment
allows comparison with the Giving What We Can figure and other alternatives, as
in the chart below.
Singer does indeed attract a lot of money, and this is primarily because of his
role as a public intellectual. Moreover, because Singer is a public
intellectual, his influence greatly exceeds this dollar value. Each time he
writes for the New York Times or speaks on the Colbert Report, he addresses
an audience of around a million .
The majority will not take his pledge. But one would imagine that they are
still, on balance, more likely to donate cost-effectively thanks to his
influence, an effect which is not to be underestimated. This massive audience
is the defining characteristic of popular philanthropy.
One’s chances of successfully finding such a large audience are decidedly slim.
If one hopes to emulate Peter Singer, one will require not only a powerful
intellect and a talent for writing, but also good timing. One would have to
find a message that will resonate with today’s population, a different audience
to the one that received Animal Liberation 35 years ago.
It is difficult to know what that message will be and it is not even clear that
writing a book is the best way to reach people any more. Is it better, for
example, than a group blog like Less Wrong or a video
In any case, if Singer had not become a public intellectual, he could have
fallen back on a position as an academic. There, he would still have had some
influence over his students, readers of his papers and so on. The same cannot
be said for aspiring producers of online material.
A life in politics appears promising, if only for the size of the national
budget. For example, the Australian government will give 4.8 billion dollars to
international aid in 2011-12. This sum is controlled by 226 members of
parliament. If power was distributed equally among these members, each would
control $21 million per year.
Compared to non-aid related political decisions, this figure will have colossal
impact. To make as much of a difference as possible, one might set the
following three objectives: increase developmental aid, improve aid
effectiveness and get re-elected. If so, the political philanthropist might toe
the party line and play the Machiavellian game of politics for decades until a
key voting opportunity arose.
Political philanthropy has some serious drawbacks, though. It is not so
compatible with other approaches: being a member of Giving What We Can could
make a career in politics more difficult as no major party in recent Australian
political history has campaigned on foreign aid reform.
A political philanthropist, then, probably ought to donate anonymously. The
larger problem is that the chance of winning an election is thin. On the one
hand, independents are rarely elected. On the other hand, Australia’s two major
political parties have over 35,000 members , many of whom compete fiercely
for leadership positions.
What is one’s chance of getting elected in a country of this size? If we place
this probability at between one in ten and one in a hundred, then the expected
value of a career in political philanthropy will be the lowest of the four
classes. Of course, we may always get some mileage from writing a letter to our
local member of parliament, but actually entering politics appears to be a
Funding philanthropic influencers
Funding philanthropic influencers is a fourth option. This means giving money
to a middle man who will convince others to donate effectively, for example,
- Advertising through mass media to encourage viewers to demand an increase in the foreign aid budget from their local members of parliament.
- Funding Giving What We Can or advertising it.
- Funding Givewell or advertising it .
These warrant further exploration. I will just make the general reminder that
the greater your income, the less worthwhile it will be for you to volunteer
and the more viable it will be to fund philanthropic influencing .
We can now draw some tentative conclusions:
- As an adjunct, philanthropic influencing will increase the influence of almost any professional donor.
- The gold standard for influential philanthropy is pledging to an organization and giving one’s time to support it.
- On average, political philanthropy seems the least effective class of philanthropic influencing.
- Peter Singer’s impact exceeds Toby Ord’s by one or more orders of magnitude. The expected impact of an aspiring popular philanthropist deserves further investigation.
- The possibility of funding influencing philanthropy also needs exploration.
- Toby Ord, http://www.givingwhatwecan.org/resources/recommended-charities.php, drawing on the work of the Disease Control Priorities Project: http://www.dcp2.org/page/main/BrowseInterventions.html
40 years × 1100 QALYs/year = 44000 lives. The latter figure is from Toby Ord, http://www.givingwhatwecan.org/the-problem/how-we-can-help.php
The idea for this paragraph comes from Will MacAskill’s presentation ‘A Different Approach To Ethics’: http://www.careers.ox.ac.uk/index.aspx?o=11590
Here is an example of a break even point for the marginal utility of a philanthropic influencer compared to a professional donor:
|You give 50%||As per Bill Gates’ Giving Pledge, the 50% league, Toby Ord’s own further pledge, etc. 50% would seem a realistic estimate because those considering a switch to philanthropic influencing are highly motivated.|
|Your ‘converts’ give 10%||As per the Giving What We Can pledge|
|You would earn double the salary of your ‘converts’||Again, this is reasonable because those considering a switch to philanthropic influencing are highly motivated.|
|You have the same number of years left in your working life as your audience||This factor can be adjusted according to age|
|Your ‘converts’ give up on the pledge, on average, halfway through their working life, while you would have continued to donate for your whole life.||Because complete ‘conversion’ is unlikely|
|Your ‘converts’ donate half as cost-effectively as you, on aggregate.||It would seem reasonable to expect that most ‘new converts’ would donate to Malaria International or Schistomiasis Control Initiative if suggested to do so by their philanthropic influencer. If the philanthropic influencer believes AI or animal welfare-based charities to be more valuable, then they may have more difficulty directing their ‘converts’’ funds.|
Given these assumptions, to break even, you would have to gather 40 new professional donors in your lifetime. Here is the calculation: Impact of your influencing compared to your donating = (Number of persuaded audience members) × (Their salary) ? (Your salary) × (Their % donated) ? (Your % donated) × (Their years until retirement) ? (Your years until retirement) × (Their adherence to the pledge) ? (Your adherence to the pledge) × (The cost effectiveness of their donations)?(The cost effectiveness of your donations) = 40 × 0.5 × (0.1/0.5) × 1 × 0.5 × 0.5 = 1 That is, the impact of your influencing equals of the impact of your donating. There is a large margin of error on all of these estimates. Note also that that this calculation only includes those who have initiated giving because of you. You may have more success getting people who already donate to direct their funds more cost-effectively.
However, if one earns much more than their potential audience, funding professional influencing may be a viable alternative.
Peter Singer, The Life You Can Save chapter 5, referring to various studies including Jen Shang and Rachel Croson, ‘Field Experiments in Charitable Contribution: The Impact of Social Influence on the Voluntary Provision of Public Goods,’ The Economic Journal.
Using an alternative model, if each year 80% of the previous year’s donors renew their commitment, at the end of 20 years they will have donated a total of $301.
US Census 2008: http://pubdb3.census.gov/macro/032008/perinc/new03_001.htm
The circulation of the New York Times on an average weekday is 1.1 million.
The Colbert Report has attracted 0.6 million viewers with a further 0.3 million on the repeat broadcast.
The Australian Labor Party has around 35,000 members. http://www.alp.org.au/australian-labor/review2010/ (page 10)
The Australian Liberal Party has about 80,000 members. http://www.liberal.org.au/The-Party/Our-Structure.aspx
Theoretically, we could fund any of the four approaches to philanthropic influencing. We could commission a popular nonfiction book, for example, but would it be persuasive? We could fund an electoral campaign, but for which candidate? And would the cost and chance of success be any better than the direct pursuit of an electoral seat?
Impact of funding philanthropic influencing compared to doing philanthropic influencing
= (Your salary) ? (Your Persuasiveness) × (Persuasiveness of whatever is funded) ? (Funds required)
Note, persuasiveness here refers to more than just the personality trait, it also depends on the audience, the context, the message and so on.