Giving Outside the Box - A Disruptive Idea to Increase Donations to Effective Charities

TL;DR

We should set up an organisation to accept donations to highly effective charities which would guarantee to return your donation if you ever really needed it. This would overcome a huge barrier to donation which exists today, namely the feeling that it is irresponsible (especially for parents) not to set aside significant savings for a “rainy day” scenario, which usually never materialises. 

This could massively increase the amount of donations from “non-billionaires” – people who are relatively well-off, but not to the extent that they don’t need to worry about money anymore. Furthermore, this approach is uniquely suitable to EA-type charities, and if done well, could not just increase absolute donations, but also increase the percentage of total donations going to effective charities.

This is a discontinuous idea, and many people will spontaneously dismiss it, often looking at it from the perspective of the donor (“but that’s not charity anymore”) rather than the recipient (“I need help, it doesn’t matter where it comes from”). Very interestingly, this is an objection which is uniquely irrelevant in the case of EA – suggesting that this approach might be particularly useful for EA charities. 

Prior to writing this post, I’ve already gone through the first round of feedback and objections over the past year or two. I have failed to find any killer-objection or show-stopper. I believe it can work. I know it can be designed in such a way that there is zero downside risk. A very rough calculation suggests that spending some money to test this idea could have an expected value of well over $100 for every $1 spent, even with very pessimistic assumptions on the likelihood of success and the potential impact if successful.

I’m posting it here to get a second round of feedback, especially any constructive builds, to see if anyone knows of similar schemes that exist or that have been tried in the past (and how they worked out), and also to see if anyone is interesting in working on something like this. 

Note 1: this full post really is TL, but you don't need to read it all !! Don't be scared by the 42 minutes. You need less than 5 minutes to read enough to understand the idea. In case you want to read more, just use the headings to read the parts you think are important. 

Beyond the "Elevator Pitch", everything is optional reading, feel free to comment and respond without reading further. I've included even more optional (in the sense of "optionaler" if that word existed) material in 4 Appendices. 

Note 2: This is my first EA post. So feel free to feedback not just on the content but also on the writing, the formatting, the clarity – anything I can do better the next time. (and yes, I know it's too long ... sorry :( ) 

Introduction - or How to Innovate in the area of Effective Giving? 

I’ve spent my career working in front-end innovation. Usually they contacted us when everyone else said a problem was impossible. So a fun job, where we weren’t afraid to try crazy ideas that sometimes didn’t work. I tried to look at the challenge of generating more donations to effective charities from that Innovation Mind-set. (sorry for the buzz-words ☹)

The secret to successful innovation is to challenge all the assumptions that are preventing you achieving your goal. But usually, the most challengeable ones aren’t even evident – you don’t even realise you’re making them.

To overcome this, often it’s helpful to frame problem as a contradiction (e.g. I want to make a better product, but also reduce costs. However, better products typically cost more to make.) and then try to find ways to overcome the assumptions that lead to the contradiction (e.g. could I make the product better by removing something?). 

In this case, we want to dramatically increase donations to effective charities, but people have been trying this for years, so why should I think I can do better? I shouldn’t. But I tried regardless, because it would be great to find a solution. 

The three obvious routes to increasing effective donations each immediately hit a big contradiction: 

• Enable people to have more money, so that by donating the same share of their income, they’re effectively donating more. But real wages for most non-CEO’s are increasing slowly. And realistically, this is out of our control. 
• Encourage people to donate a bigger share of their existing money. But that means asking them donate more (and so save less)  in a difficult and uncertain economic time when people are fearful about the impact of AI on their jobs, of climate-change or their property, of spiralling health-care costs, pandemics, etc. People want to save more, not give it away. 
• Encourage people to give a larger share of their donations to effective charities. But why should I believe I can do this better than the many great organisations who have been doing this for years now?

Analysing these, the first seems hopeless, and as for the third, I expect that this donations month will feature many better ideas than I could think up. 

But the second is interesing, because it highlights a major obstacle to donation; fear. People do not donate (as much as they can) because they are afraid that they may need that money at some unspecified future time. 

If we could remove the fear of needing the money later on and not having it, could that lead to increased donations? It still feels like a contradiction, but maybe not impossible. To break through this contradiction, we “just” (!) need to separate the notions of “donation” and “not having the money available in a time of need.” 

Which seems impossible at first. Unless we find a way to enable people to donate but still to have access to the money if they really need it. 

Net, an effective way to increase donations to effective charities may be to redefine the very notion of “giving” – specifically to avoid the absolutist definition of giving as meaning to renounce all ownership of something forever, and to replace it with a more nuanced definition in which you would retain some ownership of the donated money. 

One (extreme) way to do this would be to offer a no-questions-asked, money-back guarantee on all donations. For the purposes of this article, I will mostly use this, which is the simplest case to analyse. However, as I note in the section on Variants, I wouldn’t imagine this being the optimal execution - there are many more subtle variants of this which may me more realistic. But let’s leave the complexity for later. 

Summary of the Idea (“Elevator Pitch”)

Have you ever seen a very worthy cause and thought to yourself “I’d like to donate some money to that” only to conclude that to do so would be irresponsible? Yes, you have the money, and you don’t need it today, but what if you lose your job next year? What if one of your children gets very sick? Or you have to replace your car? So instead of donating it, you invest it in a low-risk account, or just put it in your bank account. Where it sits for the next 30 years “just in case” you need it. 

Especially as we get older, our responsibilities multiply. We have more disposable income, many people earn more than they spend, and in theory could donate much more. But there’s always this fear that we should keep more aside for a rainy day. This is especially true for parents, who know that their money is not theirs to give away, in a sense, it belongs to their children. What if one of them gets very sick and needs expensive care? What if two of them get accepted into great $100K/year universities? 

And there is massive social peer pressure on us to save, to invest in our kids’ college funds, to save for retirement in case Social Security goes bust. There are billions of dollars spent every year on advertising to convince us that we need to be more financially secure. It is a powerful message, and hard to ignore. 

And yet, most of these scenarios that we fear never happen, and if they do, they are often covered by insurance, or can be managed without resource to the rainy-day funds. Usually the only result is that the money sits in a bank account for decades instead of being used for good causes.

I have often wondered if there is an easy way around this dilemma. 

Imagine a large charity which made a promise to donors:

“If you make a donation and due to some unforeseen event in the future, you suddenly need some or all of the money you donated, we will return it to you. No questions asked.”

My hypothesis is that this would remove perhaps the single greatest obstacle to charitable donations. 


 

A utopian scenario just to envision how this might work: 

Tomorrow, the Bill & Melinda Gates Foundation issues the following statement about a very effective (imaginary) charity called Donate4Good:

“Henceforth, any donations above $1000 to Donate4Good will be guaranteed. If you need to get your money back due to an unexpected event, the Gates Foundation will refund it to you. No questions asked!”

Imagine how that might incentivise people to give more to Donate4Good – both to give money they had planned to set aside, and to take donations from other less effective causes and move them to Donate4Good.

In real life, this would need to be a lot more complex in terms of paperwork and so on, but the net effect would be more or less the same. 

I would argue that this could be one of the most impactful things that the Gates Foundation could do with their money. At worst, the impact would be to dramatically increase donations to the most effective charities. In a typical scenario, this good would be achieved at a huge rate of return, since the net cost to the Gates Foundation would be limited to the percentage of funds that had to be returned, which would probably be a very small percentage. (see below some quantitative justification for this). 

Isn’t this just plain wrong? Immoral? Sinful? Will we all go to hell? ☹ 

This seems to go against everything that charity is about. Maybe it does go against some ethical schemes. 

When I’ve shared this idea with people, I’ve heard many “moral” objections, like the rhyme our mothers (in Ireland anyway) used if we asked for something back after giving it away: 

          Give a thing, take it back, 

          God will say, where is that? 

          You will say, I don’t know, 

          And down to hell you sure will go!

(and no, it doesn’t make sense to me either – but it does capture the moral qualms many of us would feel about giving something and then later asking if we could please get it back …) 

But as EAs, we are interested in doing the most good, rather than worrying about whether potential donors will go to heaven. Our objective is not to help people be more charitable according to some ethical guidelines, but rather, to maximise the money going to the most effective, impactful charities, so that we can maximise the good they can do. 

That said, I would be happy to argue the ethical case for this against any theologian.  By using their savings to help people in need rather than to help bankers get rich, I believe people will be doing more good in a moral sense too. And (and this is something that I cannot repeat enough times!) the point is that anything that is donated in this way must be in addition to the normal donations that would anyhow have been made, not instead of them. 

But at least in terms of maximising positive impact for those who need it most, a money-back guarantee has huge potential to absolutely change the game. 

A more important objection is: would it actually work? My hypothesis, and my initial analysis, suggests that it would. (But I’m posting this here to let people shoot at it and tell me why it won’t 😊). Let me briefly explain, starting with some calculations:
 

Let’s do some calculations so we’re back on more comfortable territory.

To see why this can work, ask yourself where the majority of the world’s wealth lies. (Just to simplify the math, let’s focus on the US, but something similar would be true for the EU and other well-off countries. So the net potential global impact might be about double the net US impact.) 

Sure, there are a few billionaires and multi-millionaires for whom donating a large part of their fortune comes with no personal risk. But below them there are many households who have more than enough to survive, but not so much that they don’t have to think about money:

• 1.0% of US households have an income above $500K per year
• An additional 11% of US households have an income above $200K per year
• An additional 22.4% of US households have an income above $100K per year. 

There are 130.6 million US households: I believe it is reasonable to assume that 

• average households earning $100K-$200K could spare 5-10% of their income, so $10K/year,
• average households earning above $200K-$500K could spare 8-20% of their income, so $40K/year, 
• households earning $500K - $2,000K could spare up to 12 - 50% of their income, so $250K/year.

By “spare”, I mean something very precise: this is money that they do not need or use to meet annual expenditures. Most years. This is money they either save or invest in low-risk funds or spend on things they don’t particularly need. Of course there are exceptions. But this is just to get an order of magnitude of the vast amount of potential donations that exist among this segment of the population. 

Quick calculation: 

130.6m x (0.01 x $10K + 0.11 x $40K + 0.224 x $250K) = $1.193Trillion

So every year, there is more than one trillion dollars in the US alone which could potentially be donated to effective charities without majorly impacting the quality of life of the donors. 

The actual total amount of charitable donations from all individuals (vs foundations or companies) in 2022 was $319 billion. Excluding mega-donations of $14 billion by individuals who are clearly not part of this group, and making a (very conservative) assumption that at least another few percent of individual donations came from households with an income above $2m or below $100K, it is reasonable to assume that households earning between $100-2,000K/year contributed less than $300B – i.e. less than one quarter of the potential donations calculated above.  

And, of course, the majority of these existing donations goes to highly ineffective charities, like donations to very rich universities, building concert halls, mega-churches, supporting orchestras and the like. I’m sure there will be many great posts this week on ways to convince more donors to give the money they donate to more effective charities. I want to focus on the other $900 billion, which people currently do not donate, even though they could afford to. 

But in fact, this idea could also drive a higher fraction of these existing donations to more effective charities. 

Why do people not donate as much as they could afford to donate? 

My hypothesis (which is unproven, but testable – see below) is that one significant obstacle to donating is the “save it for a rainy day” phenomenon. The feeling that we need to keep some money in reserve in case something bad happens – like losing a job, incurring unexpected large medical costs, or whatever – or even in case something good happens – like both kids get into very exclusive, expensive colleges far beyond my planned college-fund.

The problem with this is that it is by far the lowest-energy-barrier solution. It’s not as if most people put their money on the table and then weigh the options – should I donate to an effective charity, or should I save? – and carefully evaluate the pros and cons. Instead, if you have potential future needs, it is a very obvious default, even a “responsible” choice, to keep that money in reserve in case you need it. It’s what you do if you don’t really think it through at all. 

I believe many very generous people simply never consider the possibility that they could donate more than a small % of their money unless it is very clearly superfluous to their potential needs. Even if you’re a good person and would like to help charities, there is a gnawing voice asking you if you might regret it later – and this is especially strong if you have to worry about other people and not just yourself – say your kids or your aging parents. Which is a pity, because the majority of families in the high-income brackets I’m considering would be at the age where they would have both dependent children and aging parents. 

And part of the problem is that people tend to give in to this voice rather than, say, challenging it to prove its argument logically and quantitatively. If this initiative achieved nothing more than getting people to challenge their assumptions of how much they could really afford to donate, that could already make a dramatic difference. 

Why might a “guaranteed money-back” scheme can increase total net donations to effective charities? 

There are five strands to my argument, which I will outline here and justify in the Appendix:

1. In the vast majority of cases, offering a guarantee would enable and encourage people to give money.
2. In the vast majority of cases, they would never ask for it back. 
3. Even in the case where some people do ask for (some or all of) the money back, that money will have done good during the time it was donated. 
4. This approach can work better for effective charities than for less effective charities. So it can drive not just more donations but also a higher percentage of donations to effective charities. 
5. Optimisations of this scheme can offer new potential to further drive the magnitude, frequency and reliability of donations to effective charities, and eventually step-change the world of effective giving. 

Some obvious objections:

There are also a few obvious watch-outs. I address each of these briefly in the Appendix. For now, just trust me that I’ve thought through these and many others – but do feel free to challenge my proposed answers and/or to suggest objections I haven’t covered:

1. What if people decide they want this guarantee even on the money they already donate? (spoiler: this can be avoided)
2. What if there’s a downturn / depression and a lot of donors request their money back all at once? (spoiler: this can be managed without risk)
3. Won’t this create a bureaucratic mess or record-keeping and receipts and logistics? (spoiler: no)
4. Will this really make a difference? (spoiler: yes. It might take time, but it could gradually become the norm – as in, we could create a culture where instead of “saving for a rainy day” people put their money in trust of a very effective charity knowing that if a rainy day arrives, they can get it back) 
5. Doesn’t this totally undermine the basic principles of charity as selflessly giving to someone who needs it more? (spoiler: the opposite is true – it enhances it)
6. What about the implications for tax-deductibility? (spoiler: short-term this can be managed easily, long-term, the rules can be changed). 
7. This will confuse people. (spoiler: I think it could. The idea might be very slow to take off, but it could gradually become a standard form of giving to complement the current donations). 

Why does this make financial sense for the charity receiving the donation? 

For the charity, it is a win/win situation. Consider the following example:

1. A person makes a donation of $100 which they might not otherwise have made. The charity gets $100 to use for good causes. 
2. In the vast majority of cases, 95% I’d estimate, this money will never be reclaimed. 
   1. We can argue about the numbers, but remember, we’re talking about people making donations to charities in a context where there is no potential financial gain – see below for more on this, which is critical – for them, so presumably they believe in the charity. 

So if 100 people donate $100, that is already a net gain of $9,500 for the charity. No strings attached. 

1. In the scenario where a donation is reclaimed, the charity has still had $100 to use for a number of years and it has achieved some good with that. OK, one loan of $100 would be messy, but if there are a lot, combined, they can be managed in the way banks manage funds, with a reasonable assumption that only a small % will be required over any given period. One could imagine 90% of these loans being treated like cash, and 10% kept in reserve or invested in relatively liquid ways.
2. The charity will still make a net gain on the transaction for two reasons:
   1. The value of $100 when donated will be larger due to inflation than the value when it is reclaimed. 
   2. If the charity was registered as tax-deductible, the charity only needs to return the net cost to the individual. (e.g. if the $100 donation resulted in $40 less tax being paid, then the charity would only pay back $60, for a net gain of $40). See below for more on the subject of tax-deductibility, which isn’t trivial, but can be managed. 

A critical point here is: this will be set up in such a way that there is no way that a charity can lose money, or even, no way it can fail to gain value from a donation. 

Tangibly, how would it work? 

To reduce the logistics and potential complexity, my first proposal would not be that each charity set up such a scheme, but rather that some overarching group, let’s call it Charity Donations Management (CDM) for want of a more creative name, manage the scheme for one charity (e.g. AMF) or umbrella charity (e.g. GiveWell), or for a group of carefully selected charities – for example those recommended by GWWC.  [note: I have not discussed this idea with any of these].

In this context, a further benefit might be to steer a higher fraction of charitable donations to more effective charities, since this option would only be available there. 

So CDM would manage all the financial details, all the logistics, all the paperwork, and would simply pass money on to organisations like Donate4Good, either in the form of pure donations (like any standard donation) or renewable loans. For loans, after a set period, CDM will either renew the loan or ask (with an agreed advance warning) for (some of) the money back, but beyond that, the money will be like any bank-loan, with no risk of being asked to repay it early. Over time, more and more of the loan will be converted into pure donation, and a small % will have to be returned. 

Over time, if it proves to be effective, more groups could do this, even offering different schemes, different set-points, etc. However, my belief is that EA-type charities would have a much higher affinity for this than typical charities, since we are clearly focused on the effect (on the receiver) rather than on the moral generosity of the donor. I cannot see how a similar scheme could work for “charitable” donations to build concert halls. 

Why would it help?

My hypothesis translates to saying that the existence of an organisation like CDM would majorly increase the total money going to effective charities like those recommended by GWWC or Donate4Good, in at least four ways:

• Removing a major obstacle to donating.
• Providing a tangible motivation to prefer effective charities.
• Providing a strong, tangible motivation for regular and larger donation commitments, since this would ease logistics and paperwork for the donor. This can be enhanced by a well-chosen minimum annual commitment in order to be eligible for the scheme.
• Potentially being a provocative news-story that would bring positive attention to effective charities. (challenge: how to make the idea go viral? How to make it “cool” to be a part of this “movement”?)

But will it make a significant difference? 

One obvious objection might be “yes, it could work, but it probably won’t make a lot of difference.” I believe that it could make a huge difference, but I fully accept that this might happen very slowly. 

To some extent we can overcome this with a very targeted campaign of individual marketing – somewhat analogous to the way charities today challenge people to leave money in their will. 

My argument is not that this is sure to work, but that the cost of trying it is almost zero – maybe one additional line of text to mailing lists, maybe a few electronic forms and a website to create. And the potential is huge. So it’s at least worth testing.

The expected value of testing this idea is very high! 

In terms of net expected return from testing, let’s imagine you are very doubtful and believe that this has only a 1% chance to succeed, and that if it succeeds, it will yield only 1% of what I’m claiming is possible in terms of additional donations to effective charities. 

This means that there is a 1% chance of getting an additional 1% of $900B per year, or a 1% chance of an additional $9B year, so an expected value of $90m / year. 

So if there were a way to test this idea for, say, $1m, that would be an expected return of $90 for every $1 spent. 

However, testing the idea should cost much less than $1m, as I’ve outlined below in the Pilot / Testing section. A first step might be to spend less than $100K to do both some preliminary quick & dirty qualitative and quantitative testing and optimisation of the idea, as outlined below. 

If this were to give positive results, you could then decide to invest $1m in testing this idea. But here’s the thing: even if the idea totally failed (e.g. everyone reclaimed their money, or nobody donated at all) you would also recoup nearly every bit of this $1m (just minus some minor costs, but this may be covered by things like interest earnings). Because the $1m wouldn’t really be to fund the testing, but just to guarantee the donations – and any repayment of donations would mean that a donation at least that large had been received in the first place. So the net cost of the test in the absolute worst-case scenario would be closer to $100K at most, meaning that testing the idea has an expected value of $900 for every $1 spent even with the very negative assumptions about likelihood of success and likely impact.  

Net, there is a strong quantitative case to investigate this idea further. 

Variations on a Theme

Realistically, I don’t expect the Gates Foundation to be convinced … yet. In addition to the need for testing and validation, we also need to majorly refine the very simplistic model I’ve outlined so far. What I’ve described here is the MVP – the “Minimum Viable Prototype” – which is usually seen as the simplest possible execution of an idea which still captures the core of it well enough to get meaningful feedback from consumers or users. So we could go to potential donors and test the idea of a 100%, no questions asked guarantee of your money back. Easy to understand quickly, great for enabling discussion and sharing ideas. 

However, the real scheme could be much more sophisticated, while retaining the same core principle. Here are just a few examples of how this it might be different (feel free to skip or skim) 

• Maybe in any one year, the first $X you donate must be a pure donation before you can get a guarantee on any additional donations. This level should be set at a level that would correspond to what your typical donations would be without this option. (yes, I know that’s not easy to do …) 
• Maybe only X% of any donation can be guaranteed. Say 50%. That might seem very reasonable to many people and still spur a major increase in donations. 
• It doesn’t need to be a “no questions asked” guarantee. It could be that you need to justify why you need the money. 
• It could even be that you can only guarantee it against specific outcomes. E.g. “I will need this back if my kid gets accepted into Harvard, or if my wife loses her partnership at the law-firm” – the point here would be that it would apply to specific, tangible concerns only – so more limited applicability. 
• Maybe the % that can be reclaimed decreases over time (even though inflation already provides some element of this) and/or when you sign up, you commit to making the donation permanent at a specific point in the future. 
• There could be regular reminder mails asking every year how much of your existing guaranteed donation you would be willing to convert into an unconditional donation, highlighting the immediate impact such a commitment from you could have on some beneficiaries of your generosity. 
• There would be a carrot of recognition (certificates, targets, gamification, …) which would be available only for fully committed donations. Imagine if you were just $7,000 short of getting a certificate as a gold-level donor, and you had $31,000 in guaranteed donations, would you be willing to convert $7,000 to confirmed donations? Very likely many people would. 
• There would definitely be a planned time-delay for recovering funds (e.g. 1 year), but CDM could have a deal with a bank that would allow people to borrow against this if their need is sudden and urgent. This would be just to give charities time to recover funds that had been loaned. (again, this seems like a pain, but remember the counterfactual here is that without this scheme they don’t have these funds at all!).
• My in-going assumption is that the donation would become permanent on your death. But there are obviously alternatives to this, as long as they do not involve other people deciding, which would be a mess. (e.g. “if I die before event X – which should be a low-probability event, otherwise what’s the point? -  I want you to give the money to my son.” But not “when I die, the right to the guarantee passes to my son.”)
• (or the previous two combined – either a major, unexpected cost or a pre-identified low-probability scenario that materialises). 
• But recall, in the ideal scenario, the charities themselves don’t have to deal with any of this uncertainty. CDM manages the loans as a portfolio and keeps enough liquidity to pay off reclaimed loans without ever asking the charities for unplanned payments. 

These and many other ideas could be tested with potential donors in a preliminary phase, and over time, we’d learn by trial and error what works best. 

Can we ensure charities don’t lose money by doing this? 

This can be designed in such a way that there is no possibility of a net loss for the charity. For example, admin costs can be subtracted from the reclaimed amount. In this post I don’t want to get into the details, but I’m happy to do so if people want to talk seriously. If this is run by one charity (like the imaginary CDM), then the question does not arise. The charities still receive donations, and additionally may (depending on how it is handled) receive zero-interest loans which may be renewed. There is no downside for them.

Call to Action (and Request for Feedback) 

Let me know what you think. If you hate the idea, tell me why. If you love the idea, tell me what you’d suggest we could do to move it forward. If you’re confused by some aspect, or doubtful about something, just ask. If your comment is too negative and you prefer not to share it publicly (although I won’t mind, really!), feel free to just message me. 

I won’t be offended, I promise. :D 

APPENDIX (= even optionaler reading)

APPENDIX 1: How to Test and/or Pilot the idea?

Obviously, nobody is going to commit to a scheme like this without evidence that it works and reassurance that it doesn’t have major drawbacks. 

A good way to get some evidence would be to run a small pilot – for example in one county or one US state or one country. In a perfect world, the easiest way to make this happen would be to find someone rich or a well-funded foundation who already plans to donate funds to a specific charity, to instead “sponsor” the pilot. By this, I mean that, during the period of the pilot, this donor would personally guarantee any donations. In that way, the donations would be directly passed on to the charity, and nobody would be any worse off. In fact, the sponsor would be significantly better off. Let’s look at the numbers:

Let’s say sponsor S volunteers to sponsor a trial of this idea up to $1m in a small region. Sponsor S has already decided that they will donate this $1m to Donate4Good, but sponsors this pilot instead. Tangibly, they put the money in a secure investment for 1 year. Let’s say the pilot runs for 1 year. After 1 year, let’s assume that $500K has been donated with a no-questions-asked guarantee. Then:

1. Donate4Good will have received the $500K (donations) and at the end of the study. That money is theirs to keep. Also, the sponsor will just donate the remaining $500K which was not needed for the pilot, plus the full interest earned on the $1m over the year. So Donate4Good receives the same $1m they would have received, with just a small delay on some of it.
2. Sponsor S, instead of being out of pocket by $1m, is only immediately out of pocket by $500K, with an additional responsibility to potentially have to pay back up to $500K at some future data, most of which is very unlikely ever to be needed. 
   1. A sufficiently generous sponsor might just donate the remaining $500K to Donate4Good and manage any funds which needed to be repaid out of their own additional funds. 
   2. A very prudent sponsor could re-invest the remaining $500K. At any future date, as the investment grows, they could donate any monies over $500K to Donate4Good, while remaining fully financed to deal with eventual repayments. 

One challenge in piloting this is that the time-scales involved are long, be it the time-scale between a donation and potentially reclaiming the money, or the likely timescale needed for this idea to penetrate and catch on via an appropriately low-key communication approach. A pilot is only valuable if you can actually do something with the results you get, it must be actionable, and it should guide future action. It is perfectly possible that this idea is good but that a pilot would not deliver much evidence even within a year or two. 

Therefore, before even suggesting a pilot (unless someone is already convinced by my arguments …), an alternative, but more importantly, complementary lower-risk testing approach would be targeted qualitative and quantitative research and interviews. 

This is not hard, and could be fast and cheap. Typically it might to as follows (no need to read this in detail unless you’re curious) : 

1. Develop some “concepts” – can be as simple as a few sentences on a sheet of paper or a screen – which capture the key elements of the idea. Get a few people who understand the idea to help, and then run them by a few more people who haven’t heard the idea before to get some initial feedback. 
2. Run some focus groups among likely early-adopters to gauge their reactions. Ask them to create better concepts. 
   1. These need to be well moderated to avoid one dominant view drowning out the others – we don’t need to convince everyone! 
   2. In future rounds, it will be possible to actually show a concept and ask “would this interest you?” and only include people who would be interested in one group, and only people who would hate the idea in another group, and so study both perspectives in more depth. 
3. Also use the focus groups to create a list of FAQ’s which they believe potential donors would want to ask. 
4. Create improved concepts, create the FAQ’s and answers, and repeat the process until it feels right.
5. At this point, we have clear concepts, a set of FAQ’s with clear answers and anything else that might be necessary. 
6. Optionally, we might want to create a realistic website with all this information, representing what a potential donor might see in real life. This would only make sense if we do it well enough that the website itself is not an obstacle. 
7. Once this is all ready, we can run a quantitative test among potential target users. This would be anonymised, but potentially (if people agree) we’d have the option to follow-up with people who make interesting comments. 
   1. The research wouldn’t be just abstract questions, although some parts of it could be that. Rather, we’d maybe position it as a game, we pay the volunteer $50 for an hour of their time (for a fun experiment) and we give them a scenario (which isn’t too far from their own reality) and then show them the website and give them some fake “money,” say $150K, to divide among different buckets – home expenses, savings, investments, capital costs, charitable donations, …. In one leg, there is one additional option (the guaranteed donation one). “Success” would be measured by more donations. And any reduction in “normal” donations would be a concern we’d need to fix. 

This can be a game too. Depending on how the risks play out in the donor’s scenario, they end up making some real donations (really, we give the money to the charity) and taking home maybe less, maybe more, of their $50. The donor loses the game (and some real money) if they end up in a future in which they are unable to cover some major crisis. So this forces them to think critically about it. 

1. These games sound silly, but they actually provide a lot of valuable information. While we definitely cannot just take the numbers from the “game” and assume that’s what donors would do in real life, there are typically correction factors that are reasonably empirically validated – for example, maybe for every $100 extra people say they’d donate, reality shows they donate $23. Or whatever. 
2. Based on the study, we start to get a picture of the potential of the idea to increase donations and of how to communicate about it. At this point, if the data all looks good, you might imagine going to a sponsor and asking if they’d be interested in sponsoring a real pilot in some small, controlled area.  

APPENDIX 2: Getting the Communication / Message Right

Communication would be critical to making this work. 

We need to be aware that public communication is different to communication within a charity or with the EA community. We need to absolutely control any message, and avoid any risk of misinterpretation, intentional or otherwise. 

It must, first and foremost, avoid any risk of damaging donations that charities receive today. This is one reason why it might be better for this not to be run directly by the effective charities, but rather by an independent agency. 

The communication would need to be crisp and clear and focused. I won’t attempt to suggest what would be the best way to do it – there are experts who are good at that, and there are very good ways to ensure that the message is what we want it to be. 

It must be absolutely clear and repeated again and again that this is NOT any kind of tax-dodge. There is no way you can ever get out more money than you put in, or that you can use it to find a way to pay less tax. 

It must be absolutely clear that the objective is to allow you to make more donations. We don’t want anyone to reclaim any donations. The only thing that has changed is that we now consider the possibility that unforeseen circumstances may arise, or a pre-identified low-probability risk may happen, causing you to truly need to recover the money. And that by offering you that possibility, we’re effectively avoiding the absurd situation where the money just sits in your bank account while so many people around the world desperately need it. 

A good start might be a standard letter or even face-to-face presentations, that would appeal to a few likely early-adopters – perhaps the engineers and scientists and accountants – and would be ignored as too complex by most others. Then the vision would be to spread it by word of mouth. 

The idea might be very slow to take off, but it could gradually become a standard form of giving to complement the current donations. If it works, we will come to a point where people understand this the same way they understand tax-deductible gifts today – without necessarily understanding the tax-code, they know that if you fill in a number or tick a box, that means more money for the charity and/or a rebate for you (depending on the country) – they don’t ask to study the calculations.

FROM HERE ON JUST READ SPECIFIC POINTS YOU’RE CURIOUS OR DOUBTFUL ABOUT

(also, I didn’t edit or even write the parts below very carefully). 

APPENDIX 3: The five strands of the argument:

A guarantee would enable and encourage people to give (more) money that they would like to give already, but are afraid that they may need it one day in the future for a foreseeable risk or unforeseen circumstance that is not likely to happen, but is possible.

1. If I haven’t managed to convince you of this with the above text, a few words here probably won’t change your mind … 

In the vast majority of cases, they would never ask for it back. 

1. The communication will be crystal clear that there is no possible way to benefit personally (financially) from this scheme. So we should only get legitimate donations. The target group will be people who sincerely want to do good with their money, want to make a real difference. 
2. We will be clear that this isn’t about loaning money until they need it. This is specifically about money that they very likely will not need, but they cannot be sure, and they are worried that it might be reckless to give it away, or that it’s not truly theirs to give away in case one of their kids unexpectedly needs it, or whatever. 
3. We will communicate with them about the good their money is doing – making them feel good about donating and making them feel very bad about asking for the money back. 
4. The process for donating via this route will be far more complex than just making a normal donation, so we will encourage significant, serious donors who are truly in this for the long run, or who want to make a major one-off donation. 
5. The process for recovery will not be easy. There may (depending on the details) be limits as to how much can be recovered, how fast it can be recovered. Etc. There will at minimum be clear decision points where we show them what their money is doing today, and ask them do they really not want to do that. 
6. We will help advise them about ways to not request the money back – so other financial solutions (not reckless ones). We may work with a bank which can offer loans which we would guarantee, so that we give them a chance to get some money without taking any back, while using their donation as a guarantee.
7. Etc. 

Even if people do ask for (some) money back, that money will have done good during the time it was donated. 

1. At minimum, any donation to this scheme will result in some net transfer of cash to the charity and also a zero-interest loan. The exact details need to be worked out. 
2. At worst, a zero-interest loan would enable a charity to invest the money at some profit. But probably they can do a lot more with it than that. 
3. By setting time-limits on recovery and focusing on the donated money as a guarantee rather than as liquid cash, charities can receive zero-interest loans over significant, guaranteed periods. 
4. These loans would be of a type that probably will never need to be repaid, but might. So the charity just needs to keep track of what might be owed when. This can be compatible with, say, investing in projects which are relatively low-risk but need capital to get started, or for overcoming temporary funding shortages. 
5. There are portfolio management techniques which would enable the money to be used for many purposes which are almost like cash provided there was a good chance that at least X% of them would end up recovering the invested costs. This is a pretty low bar. 
6. We could imagine a guarantor of last resort who might offer to step in in times of absolute crisis and donate (or lend) the money needed to cover a cash-shortage for repaying donations. This would allow even more flexibility. 
7. While all this seems a lot more complicated than handling simple pure donations, the point here is that you get this in addition to all the pure donations. Yesterday you got $1m in cash and that’s it. Today you get $1m in cash and an extra $50K in cash and an interest-free loan of $2m for 3 years that will probably be infinitely renewed. And you can use that money to fix problems that are there today.  

This approach can work better for effective charities than for less effective charities. So it can drive not just more donations but also a higher percentage of donations to effective charities. 

1. The approach itself emphasises effectiveness over “charity” so will have a natural fit with effective charities who focus on impact. 
2. The narrative fits perfectly with effective charities, but jars a bit in other contexts. Imagine telling the pastor at the mega-church that you might need your donation back one day. 
3. If we choose to do this, we’d have the first-mover advantage. If another charity started doing the same, we could either support them (if they were good) or double-down on our emphasis on doing more good. But ultimately, if more charities do it, that would result in an increase in the total donated, with, at worst, no reduction in the fraction going to effective charities.  

Optimisations of this scheme can offer new potential to further drive the magnitude, frequency and reliability of donations to effective charities, and eventually step-change the world of effective giving. 

1. This is basically taking the idea of groups like Donate4Good, which coordinate donations to effective charities, and looking at how to target it at a very specific group facing a very specific obstacle. 
2. While all donations are welcome and no charity can or should reject them, with this scheme we could require certain minima or commitments (e.g. regular donations) to make it worthwhile. These minima would also incentive people to give more – in the same way that in Belgium many people donate €40 rather than €20 to random charity events, because if they donate €40 it becomes tax-deductible.
3. We could have set levels for which the paperwork is much easier – “if you commit to $1000/month, we can handle everything in a 10-minute call because the forms are pre-filled”. Or whatever.  
4. There is a whole science of “nudging” which looks at subtle ways to influence behaviour. With a very clear target group and a very clear objective, it is possible to use this mentality to find win-win solutions which enable people to give what they want to give and help charities. I don’t pretend to be an expert and certainly don’t know more than the charities themselves about this, but we can work with them to find the best approaches, while not necessarily having all the same limitations they would have. 

APPENDIX 4: Outline answers to some obvious objections:

(just read the ones that you’re not convinced can be addressed)

What if people decide they want this guarantee even on the money they already donate?

• This is, IMHO, the biggest watch-out here is that this must not become the standard model for donations. Most donations must continue to be irrevocable, as they are today. So nothing is lost. We just want to add this option for people who really would like to donate, but just are too afraid of what could go wrong in their lives. How might this be achieved? I have some ideas, but they still need to be worked out in detail. For example: 
• One option would be to make this quite paperwork-heavy to ensure it doesn’t become the standard.  
• Another would be to set a minimum bar (e.g. $1000 in a year) for this to be applicable. In this case, it might encourage people to donate that much rather than a smaller donation. Or it might encourage people to choose to donate $1000 of the money they planned to donate anyway to more effective charities which would be the only ones offering this option. 
• The previous two ideas would work together. The idea is that this would be an approach for serious donation commitments. Where the person fills in a form saying that they will give $X,000 per year on a monthly basis under these conditions, sets up a monthly bank-draft, etc. Maybe even meets with a rep of the charity to discuss how they’d like the money to be spent. 
• Other options would limit the amount or the conditions for a refund, or put an effective system in place to encourage people to reconsider requesting refunds unless they really need them badly (e.g. maybe you have to explain why you need it back, maybe the process is quite complex and involves a very tangible demonstration of the good that your money is currently doing, as in “Of course we can return your money, but before we do, can we just show you what we’re currently doing with it? Maybe you will want to reconsider.” 
  • This plays on a psychological consideration that people put a lot more value on a loss than a gain – so to stop something good that’s already happening (a loss) would be valued highly, while to receive the money back (a gain) would tend to be valued lower. So there’s a real chance to convince people to use this only as a last-resort in cases of desperation. 
• I believe the big hurdle is getting people to donate in the first place. Once they do so, they feel great about themselves, and they would feel very bad undoing the good they have done. People who do not donate today often simply do not realise how much could be done with their money, despite all our efforts. 
• That said, I consider this to be one valid objection that needs to be handled carefully. Above, under “Variations on a Theme” I consider a few more approaches which might minimise this risk.

What if we get a major recession or some global tragedy that causes a “run” where more people than expected need their donations back all at once? `

• In that context, we could create a “total limit” rule, in which, just like a company that goes bankrupt, the charity would only pay a certain total amount, to be divided among the claimants. 
• Variants of this could be that it would owe the remainder, to be paid only when the charity had the money to spare, or that it would not owe the remainder above this limit. 
• In another variant, the minimum repayment could be a percentage of the amount donated (say 50%), which would then be money that the charity would need to invest responsibly rather than use – or it could do as banks do, and spend this on the basis that only X% of the money is likely to be recalled within any period.
• You might question why people would agree to this risk of not getting all their money back, which seems to be the whole point of the scheme. But the point is: they are protecting themselves against specific fears, mostly individual things. If their mentality is that they want to be sure to be rich even if everyone else is suddenly poor, they will probably not be the kind of people who would donate anyway.
• This could be set at a level which would prevent the charity from having to interrupt or reduce operations. But it’s important, again, to highlight, that even in this scenario, the charity would have done more, with the refunded donations, than it could ever have done without them. 
• Insurance against such a run might be an option. I haven’t investigated this. This would cost money, but might have a net positive expectation value (consider an analogy of betting against the market). 
• A generous donor might offer to provide cover for such a situation – whether by paying insurance themselves or some other means.  

Won’t this create a bureaucratic mess or record-keeping and receipts and logistics?

• Briefly, no. This is not complex.
• For the organisation itself (the one managing this scheme), there is some bookkeeping needed, but nothing beyond a standard ledger which could be handled in Excel. They need to track how much each donor has donated and, depending on the specifics of the scheme, have the program calculate exactly how much this person might be entitled to reclaim at any given time. Not rocket science.
• For the charity receiving the money, there are two possible scenarios:
  • Preferred scenario: everything the charity receives is received as a pure donation, no strings attached. So no additional complexity at all.
  • Potential scenario: in addition to pure donations, the charity also receives interest-free loans, which will hopefully be renewed, but which may potentially be partially recalled at specific times, and with adequate warning. There are plenty of clever people working in charities who could use such loans very effectively. For example, one could imagine helping to fund a portfolio of small businesses of which it is hoped that a good percentage would eventually be profitable enough to repay an interest-free loan. (Kiva is an analogy here, and an eventual scheme like this could learn from them).

Will this really make a difference?

• My position is that there is a real possibility that it can make a positive difference which could be massive relative to the cost of testing it. (see the numbers above).
• In an ideal scenario, this is one of the few possibilities to step-change people’s donation habits, from “giving what we definitely don’t need” to truly “giving what we can, without fearing the consequences.” There is a chance that in 30 years we’ll look back and wonder why we didn’t always do this.

Doesn’t this totally undermine the basic principles of charity as selflessly giving to someone who needs it more? (And does it matter?)

• A lot of religions (I mention only my own, Catholicism) put a lot of emphasis on sacrifice and suffering. The value of what you donate is measured not by how much good it does, but by how much it costs you to do it. This is a difficult concept for EAs to grasp sometimes, since it seems irrational. But it is real, and important. We think it’s wonderful if a billionaire donates $500m, but church teaching says that this is nothing, while a poor man who can’t afford to eat but still gives his last sandwich to a hungrier man is an example to us all. When I was growing up, we children were encouraged to give up sweets during lent and to collect the pennies we saved to help famine victims in Africa. The fact of giving up sweets was a vital part of the thing – we had to sacrifice something. Logically I reject that thinking, emotionally I still feel it.
• It would be easy, in this context, to argue that having a get-out clause devalues a donation. If I make a donation that I can reclaim, is it morally equivalent to not making a donation at all? Like eating as much chocolate cake as I can without getting sick and only then giving away the rest?
• We could of course reply flippantly that nobody will be forced to have a get-out clause. But we are better than that.
• And it does matter. Because, if, as I believe, this method works better for EA-type charities than for others, there is a real possibility that some very influential people will be out to quash it – think pastors at mega-churches, or university presidents, who find that they are getting a smaller fraction of donations. So it will be tempting for them to suggest that this is cold and calculating and inhumane and not really charity at all. Many of them are very good at being able to twist logic and morality in random ways to deliver the compelling conclusion they want – it’s part of the job description. We need to be able to reply with even more compelling ethical arguments.  
• My response would first downplay and change to the “charity” aspect of a donation. Donating is good and not donating isn’t, and people doing this are donating. A person pondering this situation has already decided to be charitable, and is not selfishly looking for a way out. Rather, they are looking for ways to be as charitable as possible to one group (those served by charities) while at the same time not depriving another group (e.g. their children, or their elderly parents) who depend on them.
• They are also looking to donate more, to make an even greater sacrifice than they did before.
• Furthermore, ultimately nobody can seriously argue that being able to do more good is a bad thing.

What about the implications for tax-deductibility?

• This starts out looking like an obstacle, but could actually turn into a big positive.
• First the bad news: if you donate today to a charity like Kiva (who give micro-loans), in which in theory you are making loans rather than donations, even though you have (I presume) no intention of getting your money back, your donation is typically not tax-deductible.
• In the short-term, we need to find a way to manage this. The preferred solution would be to treat any donation to this scheme just like any other donation, with it being fully tax-deductible, and to worry about the ins and outs of cases where the money is reclaimed using a very precise set of rules, for example:
  • It must be managed in such a way that the maximum a person could reclaim would be the net amount donated (i.e., less the tax-rebate), so that there is no possibility of anyone using this to make a profit.
  • One way to do this would be to insist that the tax-rebate be donated to the charity too. Tangibly, this is just an accounting detail – if the tax-rebate is 30%, and you plan to donate $100 to the scheme, then you must also donate a non-recoverable $30. Or whatever. The details can be worked out.
  • The key (IMHO) is to make life easy for the tax-authorities. Make it clear that there is no way someone can find clever loopholes to abuse the system. At that point, by objecting, they would only be depriving charities (not donors) of money.
• In the rare cases where someone does ask for the money back at some future date, there may be a need to handle those cases carefully. It must be totally transparent and free of any suspicion that someone is benefitting from doing this. (This need for transparency may also disincentivise people from reclaiming donations except in cases of real need, which is a good thing. The goal is to make it possible to reclaim the money, not to make it easy or frictionless)
• The ideal long-term solution would involve an agreement with tax-authorities that the initial donation should be treated as tax-deductible, but managed in such a way that the maximum a person could reclaim would be the net amount donated (i.e., less the tax-rebate). The difference here being that they would actually sit down and write this legislation to handle this specific situation, because there would now be large quantities of money being donated in this way. Once one country does this, it should be helpful in encouraging other countries to follow suit. So it would be very important to manage this well.
• The flip-side of this is that, today, there are many countries in which donations to effective charities (e.g. to Donate4Good) are not tax-deductible. This is, I imagine, because there is a point where the additional logistics and infrastructure needed to create a system for another country is just not justifiable by the additional net funds that could be earned. But if a company which focused on this were to be set up, to collect donations directly (and pass them to, say, Donate4Good), and to manage the complexity of some guaranteed donations too, maybe at that point, the company could also make a point of ensuring tax-deductibility in every developed country. This could increase donations to effective charities in two ways – by encouraging people to donate to a tax-deductible charity, and by ensuring that those who do can also donate their tax-deduction. (in Belgium, it is currently 45% - but happily there is now a way to tax-deductibilily donate to Against Malaria Foundation (only) via a scheme called Transnational Giving Europe (TGE) – this is also possible from several other EU countries. By including this link, I ensure that some value may come from this post!)

This will confuse people.

• I think it could confuse people, at least at first. Remember, most people are not quantitative thinkers like most EAs. However, this can be managed by effective communication (see the Appendix on Communication). This is still a concern, but not a show-stopper unless the confusion has adverse, unintended consequences. 

If you've read this far, my sincere apologies for taking so much of your valuable time ... but thank you!