GIVING what we can

Fundraising Prospectus 2015 Contents

Our Recommended Charities 5 How effective are we? 6 Our achievements so far 7 About 10 Our Team 11 Our Plans for 2015 12 Our Impact 14 Budget 16 Appendix 1: Lower Bound Impact Calculation 18 Appendix 2: Realistic Impact Calculation 22 Appendix 3: Optimistic Marginal Impact Calculation 28 Previous Donors 30

Giving What We Can 16-17 St Ebbes Street givingwhatwecan.org Oxford OX1 1PT, UK

Part of the Centre for Effective , a registered in England and Wales, Registered Charity Number 1149828. © 2015 Our Recommended Charities

Giving What We Can helps people Giving What We Can recommends the following charities on the basis of extensive research showing them to be among the most effective in the world. donate as effectively as possible

Our members pledge to donate at least 10% of their incomes to the world’s most effective charities kills over half Schistosomiasis is a Treating intestinal Malnutrition can stunt a a million people every devastating disease, but worms is one of the most child’s growth and lead year, but the the Schistosomiasis effective ways to improve to learning difficulties, We evaluate charities and Against Malaria Control Initiative can school retention, and but Project Healthy provide a child with Deworm The World can Children can fortify a recommend the most effective can provide an treatment for deworm children year’s worth of meals insecticide treated less than $2! for just 50 cents. with essential nutrients to our members bednet for under $8! for just 5 cents!

So far, we’ve moved over $7 million, and our members have pledged over $400 million to these charities

Now we need your support

Go to givingwhatwecan.org/fundraising to donate

Child from Orissa, India carry- ing a bednet provided by the Against Malaria Foundation How effective are we? Our achievements so far

we estimate that

for every 1000+ members donating

x6 $1 x60 we spend, between $7 million already moved to effective charities $6 & $60 lower bound estimate realistic estimate will be donated to the world’s

$400+ million pledged most effective in future charities. 2010 2020 2030 2040 2050 2060 $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ we believe this represents $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ $ fantastic return on investment!

Want more info? You can review our calculations on page 14

6 7 Our Vision A world without extreme The Pledge

Our Mission I recognise that I can use part of my income to do Inspire donations to the world’s most effective charities a significant amount of good. Since I can live well enough on a smaller income, Our Mission Statement I pledge that for the rest of my life or until the We believe that people in the developed world have an day I retire, I shall give at least ten percent of amazing opportunity to substantially improve the lives of countless people at the other end of the global income what I earn to whichever organisations can most scale. Our members pledge to give at least 10% of their effectively use it to improve the lives of others, income to the organisations which can do the most now and in the years to come. good with their donations. We use evidence to identify those organisations and share our findings. By forming a I make this pledge freely, openly, and sincerely. global community and making public our commitment to effective giving, we can inspire others and ensure that we make a real impact on global poverty.

To become a Giving What We Can member, you must pledge to donate 10% of your lifetime earnings to effective causes.

So far, over 1000 people have taken the pledge, collectively committing to donate more than $400 million over the course of their lifetimes

8 9 Our About Giving What We Can Prof. Dr Michelle Hutchinson Founder and President Executive Director Team Represents Giving What We Can Sets the strategy for Giving What in the media; provides advice to We Can. Manages team members Who we are What we do WHO, World Bank and the Prime and evaluates progress. Has a PhD Minister’s Office. Research Fellow in from Oxford University and was Giving What We Can is an We want to change the way members to stick with their Ethics at the . recently named a Global Shaper international society dedicated people give money to charity. intentions, and we provide re- to eliminating extreme poverty minders and encouragement in the developing world. Many people are happy to to our members to help them give, but do so sporadically meet their giving targets. To achieve this aim, we work and without much awareness to build and maintain a of the impact their donations We undertake and review community of members, all of will have. research to identify the most them committed to giving at effective poverty-relief chari- least 10% of their income to the Instead, we encourage people ties, which we then promote world’s most effective charities. to commit to set aside and to our members, along with donate a fixed percentage anyone else interested in mak- Dr Andreas Morgensen Jonathan Courtney Alison Woodman We believe that extreme pov- of their income in a targeted ing a difference. We also run Deputy Executive Director Director of Outreach Director of Community erty, along with the problems manner. Giving What We Can the Giving What We Can Trust, Provides strategic advice. Professor Responsible for Giving What We First point of contact for new it causes, is probably the worst members pledge to donate at which provides an easy and at Jesus College, Oxford. Has con- Can’s Outreach strategy. Focus- members; keeps established mem- ongoing disaster in the world least 10% on an ongoing basis. tax-efficient way for people ducted research for Giving What es on supporting chapters and bers engaged. Organises events & today. A billion people lack But just as importantly, they to send donations to effective We Can for four years. encouraging new chapters to manages the Trust. Has worked for access to clean drinking water, donate this to the charity or charities overseas. develop. charities in India & Mongolia. and more than 800 million go charities which they believe to bed hungry each night. to be the most effective in the Since we were set up in 2009, world. over $7 million has been However, it doesn’t have to donated to effective organi- be this way. For those of us Donating 10% won’t always sations thanks to our efforts, living in the developed world, be easy, which is why we focus and over $400 million has our comparative wealth gives on building a community of been pledged for the future. us an amazing opportunity givers: a support network We’re ever-growing, and we’re to make a real difference. But in which effective giving is changing the landscape of this will only happen if people normalised and celebrated. charitable giving. give what they can, and give to Our pledge acts as a form of Dr Hauke Hillebrandt Sam Deere Luke Ilott effective charities and causes. pre-commitment which helps Director of Research Director of Communications Director of Growth Researches the effectiveness of our Handles communications and Incoming, commences mid-year. current top charities and looks for helps to bring new members on Will work alongside Sam to help new interventions to recommend. board. Previously a political advis- people along the pathway to mem- Holds a PhD in Neuroscience from er and comms director for several bership. UCL and was a Harvard fellow. high-profile Australian politicians.

10 11 Our plans for 2015 This year is set to be a big one for . With four major books on the topic set to be published, Effective altruism books including one by influential moral philosopher Peter set for release this year Singer, we anticipate heightened public awareness of the importance of and arguments for effective giving. , April 2015 With this in mind, our plan for 2015 and early 2016 is to capitalise on the media attention. We’ll have a particular How To Be Great At focus on targeting people who are intrigued by the Doing Good idea of effective altruism but don’t know where to start, Nick Cooney, April 2015 converting their interest into action. At the same time, as our community grows it is Will MacAskill, August 2015 increasingly important to maintain personal contact with our members, so that they continue to feel close-knit Strangers Drowning and supported. We’ll also be stepping up our research Larissa MacFarquhar, programme to ensure that we make the best late 2015 possible recommendations. Want to support Giving What We Can? Go to givingwhatwecan.org/fundraising to donate Our priorities for the next year:

Making First Contact of mouth1. We now have a Getting New Members plan to have a dedicated staff Ensuring that lines of commu- ing to see whether there are Effective altruism’s profile Director of Outreach dedicated Just as important as making member as the point of contact nication stay open helps mem- any other possibilities we have is growing, and our job is to to overseeing the chapters and first contact is ensuring that we for people on the pathway, bers to learn from each other overlooked. convert that into new member- their progress. maintain contact with people someone who can answer and to continue meeting their ships and more money moved who do express an interest. We questions and concerns and giving targets. It also allows us Over the next year our research to the world’s best charities. The second method is social are working on improving the help to motivate people who to promote effective altruism team will be re-evaluating our We need to make first contact media. We have recently pathway to membership, max- are interested but have yet to events and spread the word to existing recommendations and with potential members, com- brought on board a social imising the chances of inter- make a commitment. new people through our mem- looking into new possibilities, municating our goals and get- media specialist to allow ested people taking the pledge bers (see above), including by updating and adding to our ting them interested. Based on us to spread the word and, and becoming members. Maintaining Our Community encouraging members to set current list. We will also pro- our previous experience, two more importantly, encourage Our membership has doubled up new chapters. It’s also an vide advice on issues such as methods seem to be particu- existing members to share their Our increased focus on local over the past year, making it important way to ensure that the most tax-effective ways to larly well-suited to finding and experiences in order to pique chapters and social media increasingly important to focus members report their own giv- donate money, and whether persuading new members. the interest of their friends and (see above) will improve our on building and maintaining ing, so that we can accurately it is better to donate now or families. ability to follow up and help the community of members. evaluate our impact. donate later (such as through The first method is local move people towards joining. We now have a staff member a will). We will then commu- chapters. Whether based We have already introduced dedicated to existing members, Improving Our Research nicate this information to our around regions or educational our “try giving” feature, where answering questions and help- We have a shortlist of char- members, and to the public, establishments, these provide (1) 37% of members who told us how people choose how much to ing to introduce them to each ities which we are proud to ensuring people have access to a more tangible, physical they first heard of Giving What We Can give and for how long, to get other. We’ll continue to provide recommend. However we are the best possible information community, and can help to mentioned friends. 50% of those who a taste of how membership high-quality events to draw the constantly striving to improve on how to give effectively. bring people in through word mentioned friends did so exclusively. would work in practice. We also community together. our recommendations, check-

12 13 Our Impact Impact Calculations

We think it is important that people choose We measure our impact in terms of how many We measure our effectiveness conservative estimate, a higher, that a to us will have where to donate based on how they can help dollars we move to highly effective charities. using the amount of money more realistic scenario, and as much positive impact as others the most. Therefore, in order to know We need to be moving more than $1 to effective we’ve caused to be donated to an optimistic marginal impact possible. whether Giving What We Can is a good target for charities (which wouldn’t otherwise be donat- highly effective charities. We’ve calculation. We believe that, donations, we need to know the impact of our ed) for every $1 given to us, otherwise it would set out detailed calculations, even at the lowest estimate The calculations are activities. In order to help people to donate to be better to donate directly to those charities. so you can see how we arrive at of our effectiveness, Giving summarised here. For an in- the charities they believe are the most effective, Measuring impact is difficult and uncertain, so our estimates. What We Can represents an depth look at how we arrived at we’ve described our attempts at quantifying our we have given a range of estimates, using dif- excellent return on investment, these numbers, please see the impact. We’ve also provided a spreadsheet so ferent methods and assumptions. That range is We’ve provided both a low, very and that you can be confident appendices, starting page 18. that you can make the same calculation using between $6 and $300 to top charities for every your own assumptions. $1 donated to Giving What We Can. Lower bound 6:1

The lower bound calculation uses more Under this scenario, for every $1 spent by conservative or pessimistic assumptions, and Giving What We Can, around $6 is moved to top counts only the money that we know we have charities actually moved to charity (as opposed to what people have pledged to donate in future). The full calculation is set out on page 18 We think that this represents the worst-case scenario for our impact.

Realistic 60:1

The realistic scenario uses detailed information Under this scenario, for every $1 spent by Giving about what we expect members to pledge into What We Can, around $60 is moved to top the future, and provides what we consider to be charities the most realistic estimate of our effectiveness. The full calculation is set out on page 22

Optimistic 90-300:1

The optimistic calculation tries to find our the other two calcultions) suggests that it is marginal impact – that is, what is the cost of plausible that for every $1 spent by Giving What signing up another member, compared what We Can, somewhere between $90 and $300 is we expect that member to contribute over their moved to top charities. lifetime. The full calculation is set out on page 28 This calculation considers a range of scenarios, and (while it is perhaps more speculative than 14 15 Budget

This spring, we would like to raise our full budget donate regularly and substantially – at least We have previously had more staff turn-over than improving stability year-to-year. Therefore, it is for 2015, so that for the rest of the year we can £1,000 per year. To achieve this, we are running we think is ideal. An important part of making crucial both to broaden our donor-base and to focus on our core activities, instead of fundraising. a matched fundraiser, in which people who have sure that staff view working for Giving What maintain adequate levels of reserves - preferably We also want to finish the year with at least 12 previously donated less than £1,000 to Giving We Can as their preferred long-term option is between 12 and 18 months. months of reserves. In order to achieve both of What We Can, and are now donating over that these goals, we need to raise £150,000 in this amount, can have their donations matched (up fundraising round. to £5,000). We have previously relied mostly on a 2015 Budget few very generous donors; we would like to be less Our aim is to broaden our base of donors who dependent on any one individual going forward. Item Cost (£)

Employee pay/tax/other and intern support expenses 123,452

Communications and Chapters 7,660

We need to raise at least Stretch 15,123 Funds pledged £50,000 from new donors 2015 target target Administration (members and GWWC Trust) and unassignable to access the pledged funds £50,000 £150,000 £280,000 Research, training and equipment 3,200

We are trying to raise an additional Our stretch target will give us GWWC share of CEA expenses 53,988

£100,000 18 months of reserves 10% contingency 20,342 for 2015

Total 223,765 Expenses Summary

Item Amount (£) Giving What We Can share of CEA expenses Giving What We Can is part of the Centre for Effective Altruism, which allows us to Reserves at the beginning of the year + 207,000 benefit from shared HR, office costs, services, staff benefits etc.

Donations so far this year + 78,000 Item Cost (£)

Expected donations from regular donors over 2015 + 50,000 Employees 20,468

2015 budget - 220,000 Office 11,356

12 months of reserves - 265,000 Services (recruitment, fundraising, intern accommodation) 9,112 12 times the Dec 2015 budget, plus £20,000 which is our share of an extra central staff member, which we think will be required by 2016 given current rates of CEA growth. Sundries (including financial and legal costs, lunches, technical, training) 5,281

Total £ 150,000 Total 53,988

16 17 Top Charities

Appendix 1: Throughout these calculations, we refer to ‘top charities’. This includes additional charities to those we currently Our list of top charities: recommend (see page 5). For the purpose this analysis, • Schistosomiasis Control Lower Bound the charities in this category are listed at the right. Initiative • Deworm the World Initiative The first four are our recommended charities, and the rest Impact Calculation are all charities that we consider to have a high chance of • Project Healthy Children being extremely effective, but that we do not include in • The Against Malaria our current list of recommendations1. While we consider Foundation This calculation is a conservative estimate of complete data. It considers only money that was our parent organisation, the Centre for Effective Altruism, • GiveDirectly to be a highly effective charity, we have excluded it from Giving What We Can’s impact. It only takes into actually given up to the end of 2013, rather than • GiveWell account impact that has already happened and using donations made in 2009-2013 to make these calculations to avoid possible double counting and • RESULTS was concretely measured. In this calculation, assumptions about future donations, and only maintain probity. we only count money that we know with high includes money which went to the most effective • The Abdul Latif Jameel certainty has already been moved to effective charities and would otherwise not have been Poverty Action Lab (J-PAL) charities, and would not have been moved donated at all. It then compares that to our costs • Dispensers for Safe Water without us. It takes in the years 2009-2013, up to the end of 2013. (1) Some were previously recommended and have been supplanted by • Cool Earth which are years for which we have the most more effective charities; some have not been recommended yet because of • Stop TB uncertainties in the information available to us when we evaluated them; we • Innovations for Poverty are uncertain of GiveDirectly’s effectiveness, but defer to GiveWell’s consistent Action recommendation Summary of the calculation Adding these numbers gives the total donations Giving What We Can’s costs up to the end of 2013 to top charities attributable Giving What We Can were $395,000. in 2009-2013: $2,371,000.

The donations to top charities that members This means that for every $1 donated to Giving report their donations direct- This number is sufficiently Can. This allows us to calculate told us about (and which would otherwise not What We Can over 2009-2013, $6 was donated ly), we used yearly surveys to small that we have excluded the amount of money which have been donated) was $872,000. to top charities. This excludes future donations ascertain how much members it from our calculation of how would not have been donated from members who joined over the period, donated, and their incomes. much money would have been if it had not been for Giving From information provided to us by two chari- donations to other charities and impact from For those members who joined given anyway. What We Can, by subtracting ties, and one large donor, we learned about an causing people who already planned to donate before the end of 2013, 67% the product of their income additional $1,499,000 which was donated to money to donate it to more effective charities. told us their income and do- In the future we plan to use and their previous predicted top charities, and would not have been donated nations for one or more years money flowing through the donation percentage from their if not for Giving What We Can. in which they were members. Giving What We Can Trust when actual donations. Of course Member donations recorded preparing impact evaluations. this was only possible for those by this survey up to the end of However, we are unable to use members who gave us both do- 2013 had totalled $3,994,0001. it for the purposes of this eval- nation and income data. Doing Costs teers on running it, which gave the total costs up to the end uation as it was only set up in this gives a donation total of Giving What We Can’s total 11,865. Costing volunteer hours of 2013 were £238,488, which Try Givers (non-members who December 2013. $2,138,000. monetary costs up to the end at £10 per hour (indicative of equates to around $395,000 commit to donating regularly of 2013 were £119,838. Up salary costs were these people (2013 dollars). for a specified time) reported Donations that would have Additional Tracked Donations to the middle of 2012, Giving paid staff)1, this volunteer time donating $11,000 over 2013. been made anyway We also try to get information What We Can was run entirely was worth £118,650. Therefore Total Donations from We didn’t advertise this option When they join, we ask all our from our recommended char- by volunteers. For the previous Members widely until 2014, hence the members what percentage of ities as to how much of their impact evaluation (up to 2012), (1) This is conservative - we expect that Until 2014 (when we introduced small number of participants. their income they think they donations were attributed by we worked out the number our current staff are more productive, the ‘My Giving’ section to our would have donated if they the donor to Giving What We of hours spent by the volun- and therefore have a greater impact on website, allowing members to (1) Numbers have been rounded to the hadn’t joined Giving What We Can. how effective we are, than volunteers. nearest thousand for simplicity. 18 19 Project Healthy Children told A large donor who has given us donated anyway, but which us over 2013 they attributed detailed information on their was given to more effective $74,000 of donations to us. donations and the extent to charities due to Giving What We which those were caused by Can. The Against Malaria Founda- Giving What We Can was not tion got $424,000 in donations a member at the time, though The large donor mentioned from donors who cited Giving they subsequently joined. Over above donated both to a Donor What We Can as their reason for the period 2009-2013 they do- Advised Fund, and to a num- donating. nated $8,560,000, of which they ber of charities. Of their com- estimate $7,123,000 wouldn’t bined donations, $1,346,000 1 We did not manage to obtain have been donated to effective was donated to top charities . this information for the Schis- charities if it weren’t for Giv- Adding this to the information Member donations to tosomiasis Control Initiative or ing What We Can. Therefore, a provided to us from the Against top charities Deworm the World Initiative. counter-factually adjusted low- Malaria Foundation and Project $3,055,000 Healthy Children and dona- er bound for the money Giving – Outgoings To prevent double counting, tions that would not other- What We Can moved to charity Member donations to top £119,838 we subtracted the donations wise have been made to top between 2009 and 2013 in total charities that would have of members who told us they charities from members gives + gave to PHC and AMF in the is $9,414,000. been made anyway $2,365,000. 2009-12 volunteer time giving survey from the amounts -$2,183,000 Donations to Top Charities (if paid as salary) that PHC and AMF gave us. Only Putting it all together + £118,650 Doing this gave an addition- Of the donations our members Given our costs for the period = al $52,000 going to PHC and Additional donations by High Net reported over the 2009-13 ($395,000), this means the low- $95,000 going to AMF. This will Worth member to top charities Total costs period, $3,055,000 was to top er bound for our leverage ratio be an under-estimate, given $1,346,000 £238,488 charities. Excluding money (the number of dollars which that not all members would = which would have otherwise we caused to go to top charities = have mentioned Giving What been donated to charity, and for every dollar donated to us) We Can to the charities when money donated to the Centre over the period 2009-2013 was Total additional Total costs in USD donating. This gives an addi- for Effective Altruism, gives approximately 6 : 1. donations to top charities (2013 dollars) tional $147,000 attributable to $872,000. This does not include (1) Again, excluding the Centre for $2,371,000 $395,000 Giving What We Can. money which would have been Effective Altruism.

divide by total costs

Limitations of this calculation This calculation does not include:

• any impact we had on non-members apart from the one large donor mentioned who joined after Impact 2013 $6 donated for every $1 spent • any impact from causing people to give to more effective charities rather than the ones they otherwise would have • members who did not tell us about their giving • any impact by non-top charities • any donations made in the future.

Consequently, it is likely to undershoot Giving What We Can’s actual impact by a considerable mar- gin.

To try to get a more realistic estimate of our effectiveness, the next section (page 22) outlines a more sophisticated calculation which takes some of these factors into account.. 20 21 Costs least 10 per cent) for the rest of can be assumed not to actu- Total costs for Giving What We their working lives. They pro- ally donate the money they Appendix 2: Can over the 2009-2013 period vide us with information about pledged. Others we lose con- were $395,000. This includes how much they expect their tact with, so that we don’t actual expenses, and volunteer average future salary to be, know whether they donate Realistic Impact Calculation time costed as if it were paid as and their current age. From the the money they pledged. The salary (see page 18 for more latter we work out how many rate of people leaving has so This calculation aims to provide a realistic By contrast, this calculation aims to provide our detail). working years they are likely to far been 1.7% of members per 3 estimate of our actual impact. The previous best guess for how much money will be donated have left, and therefore their year . calculation (page 18) gave a very conservative to top charities from our members for each dollar It is plausible that, in order for total expected income over Other people lose contact with lower bound for our effectiveness – it simply asked spent by Giving What We Can. The assumptions the rate of membership attri- their working lives. We then Giving What We Can. The rate how much our members had already reported are therefore intended to be realistic, not tion to remain at its current multiply this by the propor- of people going silent has been they had donated to our recommended charities pessimistic. If you wish, you can visit level (see below), we need to tion of their income they have 4.7% per year (we have count- which they wouldn’t have otherwise donated, givingwhatwecan.org/impact we have invest time in the future into pledged to donate (in most ed people as silent if we we compared to how much our costs had been. provided a spreadsheet where you can try out the activities promoting member cases, 10 per cent) to derive the haven’t had any contact with calculation using your own assumptions. retention. total donations we expect them to make in the future. them for over 2 years) . It seems It is likely that some of Giving likely that members who go si- What We Can’s general activ- This methodology obviously lent still donate some amount, Summary ities will contribute towards relies on the accuracy of mem- but it is likely to be less than member retention (Giving What bers’ predictions about their the amount they pledged. We In this calculation, we start with the total Donations which we would expect to go to non- We Can continuing to grow future income. In general we have assumed that this will amount of money members who joined between top charities, which would otherwise not be might keep members feeling have found that these predic- be around one-third of their 2009 and 2013 pledged to donate to the most donated total $7 million. Members donating to enthusiastic about the organ- tions seem conservative, as original pledge (for example, effective charities. more effective charities than they would have isation for example). Howev- most people underestimate if a person pledging the stan- otherwise total $9 million. er, it is also likely that other their future earning potential1. dard 10% of their income has We then adjust to account for people leaving or activities would be explicitly If members do not estimate gone silent, we’ve only counted ceasing giving, donating a different amount than This means that the total value donated to top required to retain these mem- their future salary, we use the 3.33% of their pledge in this they pledged and a discount rate. charities by the members we recruited between bers (e.g. events, newsletters, median salary for their country. calculation). 2009 and 2013 is $34 million, time discounted We then calculate our impact from a number support etc). We think that is a fairly pes- and adjusted for donations that would have Given these numbers, the total of different sources, in each case calculated in simistic assumption, as our me- been made anyway. We estimate that for the 391 of those ceasing donations per terms of ‘top-charity equivalent’ dollars. dian member has an expected members in these cohorts, the year is 4. We’ve assumed This equates to $67,000 per member, and sug- earning potential higher than 4.8% work involved would be the 2 that this percentage will remain The donations we would expect to go to top gests that our leverage ratio is around 60:1 – that the median wage . equivalent of around one full constant over time. This means charities which would not be donated if it hadn’t is, for every $1 spent by Giving What We Can, week of work per year for a Accounting for that after, say, 30 years, each been for Giving What We Can total $17 million. around $60 will be donated to top charities. member of staff (which costs membership attrition member has a 23% chance of us in the region of $1,200) over Some members leave Giving still donating, which we believe the next 40 years. This puts our What We Can, and therefore is a plausible estimate. Limitations of this calculation total costs for these cohorts at (1) For example, many members . We have made the simplifying $443,000 estimate their future income will be This analysis does not take into have not chosen to become figures for 2014 yet (the UK tax assumption that the money the same as their current income, even account any impact we have members). We believe that this year has only recently finished, Donations pledged pledged will be donated in aside from the money we move means our estimate is likely to and the Australian one does by members though they are at the beginning of their careers - in reality, income typically (3) This number was reached by to our recommended charities be lower than reality. not end until mid-year). As our The amount of money pledged increases throughout a person’s career separating the data into yearly cohorts, by getting people to join Giv- membership has grown faster by members who joined Furthermore, the calculation working out what percentage of each ing What We Can (for example, than our costs over 2014, this up to the end of 2013 was (2) For example, many members attend only includes the future dona- cohort left each year, and taking the people who are persuaded by also means that our estimate is $146,119,864. prestigious universities and/or are tions of members who joined mean of these. The number going silent the case we make for effective likely to be lower than reality. pursuing careers that have an average up to the end of 2013, because was worked out in a similar way. giving and who donate to our Members pledge to donate a salary much higher than the median we do not have complete recommended charities, but proportion of their income (at wage (4) i.e. 1.7 + (⅔)*4.7 22 23 equal parts over the next 40 expect to be made would be the present and the date of a Accounting for members members what percentage of through by that figure means years (the median expected $62,323,117. future donation, and therefore donating a different amount their income they would have that the amount we would working life of members as at the total future positive effects than they pledged given to charity if it had not expect to be given by members the end of 2013). If no members Discounting future donations will be lower. The average pledge percentage been for Giving What We Can1. which would not otherwise ceased donating, that assump- We discount the effective for our members is 11.03%. be donated is $25,093,429 tion would mean that around value of donations over time, The UK Treasury Green Book, The average percentage of To work out how much money (meaning $22,252,663 would $3,652,997 was donated each to account for the diminished used for appraisal and evalua- incomes people actually do- Giving What We Can caused have been donated to charity year. The assumption that the value of a donation made in tion in the central UK govern- nate is 12.92%. Using the to be donated, we divided the anyway). money will be donated in equal the future, versus a donation ment, recommends using a dis- ratio between these, given the difference between the pledge parts seems optimistic, be- made right now. This assumes count rate of 3.5% per annum amount of money pledged, we percentage and the percentage What charities do the cause people’s incomes typical- that there is some loss of value to find the present values of would expect to be donated is members would have given donations go to? ly increase with age. The as- caused by delaying an inter- benefits and harms accrued in $47,346,092. anyway by the pledge percent- The purpose of this calculation sumption that the percentage vention; in essence, if people the future. (See box below for age for each member, and then is to find our impact in terms of 2 of people ceasing donations is donate now, the people affect- rationale). The percentage of income averaged over these : the dollars we cause to be do- a constant percentage, on the ed by the interventions they that people actually donated nated to top charities. This im- other hand, seems pessimistic, fund will be alive and making Using the Green Book rate to comes from self-reported data pledge percentage pact could come from two main as it seems likely that members ongoing and valuable contribu- discount future donations, contained in our membership — sources. Firstly, from members tions to their community, and we expect future donations records. We considered that the (counterfactual percentage ) donating to top charities when would either drop off within to have a present value of the first few years, or remain this is likely to have a com- actual average might be lower; ÷ they would otherwise not have $40,420,077. engaged long-term. pounding positive effect into however, we cross-checked pledge percentage donated at all. Secondly, from the future. By contrast, these figures attached to members Giving What We Can causing Making these assumptions, effects will not be compound- whose average donation per- This tells us that, on average, members to choose to donate the total donations we would ing during the period between centage was more than 15% 53 per cent of members’ do- to more effective charities than of their income by contacting nations would not have been they would otherwise have these members directly. By made were it not for Giving donated to. only revising figures for high- What We Can. Multiplying er-donating members, the To take this into account, we Choice of discount rate lower-donating members are (1) For people who had put their count- have estimated the amount of er-factual as larger than their pledge, money given to top charities The Green Book discount rate incorporates a rate should be very high. However, as com- effectively weighted higher, we contacted them to check that they which would otherwise not number of factors1. One of these is a pure time pounding benefits do not get reinvested into and therefore bring the average did not think that Giving What We Can have been given to charity at discount, which we consider to be a dubious the same interventions, we cannot assume down. Additionally, the people had resulted in them donating less all. We have also estimated the method of discounting future value; however, that social benefits necessarily lead us to con- who give a greater percentage (none of them thought that). We then amount of money which will be this is only a small component of the rate (less clude we should apply a very high discount than their pledge percentage capped their counter-factual at their donated to more effective char- than 0.5 percentage points). The rate primarily rate. are disproportionately peo- pledge (that is, we assumed we had no ities (weighting to account for tries to capture the effects of potential future ple with larger-than-average Another way we might choose a discount rate counter-factual impact on them). For the fact that the money would catastrophes which would prevent the ben- salaries (as you might expect), is by looking at the savings rate as a proxy for people who took the Further Pledge, have been donated anyway, efits of a policy (or in our case, a poverty-re- therefore the actual weighted the social return that we would lose out on only their pledged percentage was used but to a less effective charity). ducing intervention) from accruing, and the average could be a bit higher by donating later. In other words, if people (for simplicity). This makes the final Combining these figures gives effects of economic growth combined with than the one we use in this cal- pledge to give $100 in the future, what present number more pessimistic, because it the amount Giving What We diminishing of resources. culation. We therefore expect value of money would be required to generate this number to be fairly accu- does not account for people who would Can caused to be moved to top There are many considerations bearing on that future $100 if invested now. The rate of rate. have given their pledge percentage if charities, in 2013 top-charity what the correct discount rate to use is. You return you could get on investments depends it were not for Giving What We Can, but equivalent dollars. might think that interventions such as de- on the risk you’re willing to bear, but a 3.5% How much of this would have would not have given everything above Value of donations that worming have strongly compounding benefits discount rate is higher than the average UK been given anyway? their baseline. savings rate of the last decade, and around Some of the money given by would not have been to a society, and therefore that the discount (2) This prevents any one member’s what you could expect to achieve at the mo- our members would have been made otherwise projected donations skewing the (1) https://www.gov.uk/government/uploads/system/ ment with a balanced investment portfolio. given to charity anyway. To The fraction of money donated number too heavily, and means that uploads/attachment_data/file/220541/green_book_com- work out this figure, we ask by members to top charities the number can be generalised to other plete.pdf people when they become members more easily. 24 25 (versus all other charities) for Value of donations that would the ones we influence them to the years 2009-2013 was 54%. have been made anyway actually donate to. Therefore Total pledged 2009-2013: Our costs: We also need to know the the expected value of chang- $146,119,864 $395,000 54% of the $25,093,429 in impact we could expect in ing the locations of people’s Rate of leavers: 0.017 expected donations that donations, in top-charity + terms of affecting where peo- Rate of going silent, otherwise would not have 0.031 ple donate in cases where they donation equivalent dollars, is and not giving: Total cost of maintaining been donated to charity is would have given to charity $7,232,116. ↓ Accounting for leavers and membership retention 1 Membership silent members: $13,550,452 . anyway, but would have given 0.048 at current levels attrition rate $62,323,117 to less effective charities than Adding it all together $48,000 The remaining 46% of our ex- top charities. Overall, the total value we pected donations ($11,542,977) might expect the members in Discount rate for 0.035 is going to charities which we As above, $22,252,663 of ex- the 2009-2013 cohorts to have, future donations: After discounting: Total costs: consider to be less effective pected donations would have in counter-factually adjust- $40,420,077 than a ‘top charity’ . We still Percentage members $443,000 been made anyway. ed time-discounted dollars 11.03 expect them to be reason- pledged to give: to top charities equivalent is ↓ ably effective – they tend to To determine that, we need to . Percentage members Accounting for how much $26,554,056 12.92 be well-respected develop- know the fraction of people for actually give: people actually donate ing-world charities. We have ↓ compared to their pledge: whom we affected where they This is approximately equiv- Ratio of actual donations Proportion of people who 1.17 $47,346,092 multiplied these donations by a gave. Information is limited, alent to $67,000 per member to pledged donations say we’ve affected their factor which reflects how much but in 2014 we put a question who joined over that period. choice of charity less effective we think those on our pledge form asking 0.65 charities are than our recom- whether we influenced where Our impact Donations that would not have Donations that would have been mended ones. We can use the total amount been made otherwise: made anyway: people gave. The percentage Counter-factual and 0.53 $25,093,429 $22,252,663 of people who said that Giving we expect our 2009-2013 mem- donation rate We have assumed that a non- What We Can affected where bers to donate to find our total top charity is around 50% as they give is 65%. Assuming this impact – that is, how much we Money that would have otherwise Estimate for money that would not have effective as a top charity. Using information is broadly repre- might expect our members to been donated, and that we’ve that factor to weight the value been donated and went to top charities: affected the destination of sentative of all our members, donate to top charities for each Fraction of money 0.54 $13,550,452 of donations expected to go to the amount of money for which dollar we receive. given to top charities $14,464,231 non-recommended charities, we expect to affect the destina- + gives $5,771,489 in top-chari- As above, costs up to the end of tion charity is $14,464,231. Value of us affecting the ty-equivalent dollars. 2013 were $443,000. Top-charity equivalent value of money that would not have been donated, and does not go destination of donations that As this money would have been to top charities: would have been made otherwise Adding these figures, the total given to charity anyway, Giving Dividing the total amount Effectiveness ratio 0.5 $5,771,489 $7,232,116 future value of donations that What We Can’s impact lies in we expect to be donated would not have been donated the difference in effectiveness ($26,554,056) by our costs otherwise is $19,321,940. produces a leverage ratio of between where the money Total value of donations that would not would have been donated, and 60:1. That is, for every dollar have gone to charity otherwise we spend, we expect approx- (1) This calculation assumes that the where it was actually donated. $19,321,940 money that would otherwise not be As above, we have assumed imately $60 extra to be do- donated to charity and money which that the value of the charities nated to effective charities (in otherwise would be are equally distrib- people would have donated present-day, top-charity adjust- ed dollars). Total money we expect to move uted between top and non-top charities. to is about 50% as effective as (in top-charity-equivalent dollars) $26,554,056

divide by total costs ($443,000)

Want to try out your own assumptions? Acknowledgements Impact Use the spreadsheet at Kind thanks to Katja Grace, Nick Beckstead, Josh Goldenberg, Rossa O’Keeffe-O’Donovan, Lloyd Chapman, Greg Lewis, Marek $60 donated for every $1 spent givingwhatwecan.org/impact Duda, Paul Christiano and the Giving What We Can team who provided invaluable feedback. 26 27 Appendix 3: Optimistic Marginal Impact Calculation

A related way to estimate the data we have does offer some ating some level of exponential year regardless of what we do. this project. Over that time $520, per additional member fundraising multiplier for Giving useful clues. growth; e) as we get bigger, we We attracted 120 new members they messaged 680 people, generated. What We Can is to determine can spread the fixed costs of in 2013 and 405 in 2014. To be corresponded with 142, and the cost of attracting an extra The average cost of getting a maintaining an organisation, pessimistic, we could assume ultimately convinced 8 people Of course, all of these estimates member, and take the average member up until the end of such as legal compliance and that all growth in 2013 hap- to take the pledge, and another are just indications, but they amount we expect them to give 2013 was around $900. More strategic planning, over a larger pened organically, and would 3 to take the Try Giving pledge. build a reasonable case for the over the course of their mem- recently, in 2014, Giving What amount of growth. This is some have happened without any This would suggest that at the marginal cost of signing up an bership (in this case taken from we Can’s membership grew by evidence that rather than Giv- extra effort on our part. If we time we could get one person extra member falling between the above), and take the ratio 415 at a direct cost of £85,000 ing What We Can confronting take the 120 who joined in 2013 onto the full pledge with an $300 and $1000. If the cost of of the two. ($130,000). This comes to a increasing marginal returns as as the approximate number extra 13 hours dedicated to attracting a new member is much lower $320 per member one might naturally expect, the of people who join each year the task - 1.5 days of full-time $1,000 and they give an ad- The analysis above estimated on average. marginal cost of getting mem- without any effort on our part, work. From our experience, ditional $90,000 to top chari- that an additional member bers today could be lower than then we could estimate that we this seems intuitively plausible. ties as a result, this implies a would on average give around As we grow it seems that the its average over the lifetime of attracted an extra 285 through Of course, not every hour of fundraising ratio of 90, while if $90,000 to top charities due to average cost of attracting the organisation. This is ac- active effort in 2014. This would staff time can realistically be the cost of an extra member is signing up to Giving What We members, rather than increase tually not so surprising for an suggest each member cost dedicated to such outreach, $300, the multiplier would be Can and the follow-up we offer as we exhaust the available organisation as small as Giving $460. as there are operational and as high as 300. over the course of their careers. opportunities to find members, What We Can, which can enjoy management overheads. It These numbers seem remark- has gone down. Possible rea- increasing economies of scale Another way of estimating the is reasonable to assume that ably high, and it is quite rea- We now need to estimate how sons for this are that a) we have as it grows. marginal cost of attracting each extra hour of outreach re- many more members we would made it easier to sign up on- members is to study the spe- quires approximately one hour sonable to think that they are get if we incrementally grew line, which has made it easier Another clue about the margin- cific activities that you under- of other work to support it. We probably overestimates. How- the organisation, say by adding to attract members; b) we are al cost of getting new members take and look at the number of would then need 26 total hours ever, they emerge naturally one more staff member. We learning how to better con- is provided by pulling out ‘or- members who seem to sign up of staff time to attract a mem- from the data we have collect- will never know this number vince people to join; c) as we ganic’ growth from the growth as a result. In 2014 we conduct- ber. At our current salaries, this ed, and bolster the case that precisely - unlike average costs, hire more people, specialised our staff generate through their ed a trial with two interns con- would cost around £300. The Giving What We Can’s fundrais- marginal costs can’t just be roles within the organisation work. For this approach we tacting prospective members overhead on salaries for office ing multiplier is very likely over read off a spreadsheet and so allow everyone to work more would need an estimate of how through Facebook. They spent rent and other materials would 10-fold, and could plausibly be are hard to pin down - but the efficiently; d) existing members many people would join each around 100 hours together on raise this estimate to £350, or much higher than that. spread the word about us, cre-

28 29 Previous Donors Giving What We Can would like to thank our previous donors, who make what we do possible:

Luke Ding Pears Foundation

Julia Wise and Jeff Kaufman Patrick Brinich-Langlois

Alexander Gordon-Brown and Denise Lyndsey Pickup Melchin Robert Collins Frederick Mulder Softwire Oxford University RAG Shola and Jonathan Asante Ross Reason (First State Investments)

Andrew Sutton and Michael Peyton-Jones Shiwei Jin Diana Fleischman

30 Giving What We Can is an international society of people committed to donating at least 10% of their income to the most effective charities in the world. By donating to Giving What We Can, you aren’t just supporting a charity – you’re supporting donors all over the world to make the biggest difference they possibly can.

Find out more inside...

Want to support Giving What We Can? Go to givingwhatwecan.org/fundraising to donate

GIVING what we can

© 2015