The Schistosomiasis Control Initiative (SCI) told us earlier this year that it had received a donation of praziquantel (the more expensive of the drugs it uses for deworming) from World Vision. Since it already has the drugs, donations can pay just for delivery costs. Thus, SCI observed, your gift is “leveraged” – $1 in donations buys more than $1 in health programming, since the drugs are essentially “free.”

The Against Malaria Foundation (AMF) in some ways achieves leverage through the opposite strategy. It doesn’t pay for the costs of distributing nets or even shipping nets – it only pays for the nets themselves. Partners are charged with raising funding elsewhere for shipping and distribution (though they are required to do these things; nets aren’t granted without a distribution plan in place). By adhering to this policy, AMF puts other funders – including some government agencies – in a similar situation to SCI’s donors: the nets are essentially “free,” so $1 from these agencies buys far more than $1 in health programming. Because of this dynamic, these other funders (AMF argues, correctly from what we can see) are happy to provide this funding – so the original donor’s gift is “leveraged” too (by buying just a net, the donor gets “free” shipping/distribution costs).

How far could this logic be extended? What if a charity first raised money for travel expenses, then went to another donor and raised money for field staff arguing that “travel expenses are essentially free,” then went to another donor and raised money for supplies arguing that “field staff and travel expenses are essentially free,” etc.? What if a charity went to 50 million separate donors, asking each to give $1 contingent on each of the other 49,999,999 giving as well, arguing that this created 50,000,000:1 leverage for each of the $1 donors?

We think that much of our discussion of donation matching applies here. If you take claims of “leverage” literally, you’re allowing other funders – the ones who gave pills, or nets, or covered distribution costs – to influence your giving merely through the structure of their gift. That in turn creates incentives for them to take gifts they would have made anyway, and structure them in a way that gets you to give more to the program of their choice. In fact, these other donors may even be taking advantage of you to pay for costs they would have been happy to cover themselves. Perhaps World Vision would pay for distribution, but knows that by giving “only pills” it can get others to pay for distribution. Perhaps AMF would pay for distribution if it had to, and doesn’t only because it believes it can get others to.

Understanding the true nature of leverage – who is the “leverager” and who is the “leveragee” – is difficult. When someone else is giving contingent on your giving, and this fact in turn influences your giving … sometimes this means that you’re getting them to give more (or differently) than they would have otherwise (in which case you’re the “leverager”) and sometimes it means the reverse (in which case you’re the “leveragee”). We don’t know of any easy/reliable way to tell which situation you’re looking at, and when.

With this in mind, our general principles for considering leverage are:

• You can’t take “leverage factors” at face value, for largely the same reasons that you can’t take donation matches at face value.
• It’s probably a good thing when the charity you support is engaging in some sort of “leveraging.” This means it’s putting thought into getting your funds to influence others’ funds. If you believe that you’re supporting the most impactful charity possible, this means that others’ funds may be moving from less impactful activities to more impactful activities. (Though the situation isn’t this simple: we believe that governments and other large funders often have a much better array of options for impactful giving than the individual donors we serve.)
• The exact relationship between the claimed “leverage” and your actual impact is very non-straightforward. It may seem that you have more “leverage” when your $1 causes someone else to spend $9, as opposed to when it gets someone else to spend 10c. However, in the former situation, there’s a much higher probability that you are in fact the “leveragee” – that the other funder would cover the other 10% themselves if it weren’t for you, and you’re just saving them money. In general, it seems to us that the higher your claimed “leverage,” the greater the probability that someone else is in fact leveraging you.

  If anything, we would guess that your true impact is higher when the “leverage” is well under 100% (i.e., the other funder is giving less than $1 for each $1 you give), implying that the other party is likely providing funding because of the “leverage” they attain rather than because they would want to fund the project on their own. Once you have another major funder covering a major chunk of the costs, it becomes more important to ask whether they would be willing to cover all of them if the “leverage” from you and similar donors weren’t available.

When we do cost-effectiveness estimates (e.g., “cost per life saved”) we consider all expenses from all sources, not just funding provided by GiveWell donors. For SCI, we count both drug and delivery costs, even when drugs are donated. (Generally, we try to count all donated goods and services at market value, i.e., the price the donor could have sold them for instead of donating them.) For AMF, we count net costs and distribution costs, even though AMF pays only for the former. In the case of VillageReach, we even count government costs of delivering vaccines, even though VillageReach works exclusively to improve the efficiency of the delivery system.

We consider this approach the simplest approach to dealing with the issues discussed here, and given our limited understanding of how “leverage” works, we believe that this approach minimizes the error in our estimates that might come from misreading the “leverage” situation. As our understanding of “leverage” improves, we may approach our cost-effectiveness estimates differently.

We know that donors love donation matching.* We know that if we could offer donation matching on gifts to our top charities this giving season, our money moved would rise. And we know that we could offer donation matching if we thought it was the right thing to do: there are donors planning six-figure gifts to our top charities this year who would almost certainly be willing to structure their gifts as “matches” if we asked. (It might not be possible to “match” all of our money moved, but we could almost certainly provide “matching” for a short period, which would motivate people to give during that period and would also provide us with some data on the impact of matching on our audience.)

But we’ve decided not to do this because we would feel dishonest. We’d be advertising that you can “double your gift,” but the truth would be that we just restructured a gift from a six-figure donor that was going to happen anyway. We’ve discussed whether we might be able to provide “true” donation matching – finding a donor who would give to our top charities only on condition that others did – but not surprisingly, everyone we could think of who would be open to making a large gift to our top charities would be open to this whether or not we could match them up with smaller donors. Ultimately, the only match we can offer is illusory matching.

I don’t deny that non-illusory matching may exist in some other circumstances. A couple possibilities:

• Coordination matching. A charity needs to raise a specific amount for a specific purpose. A large funder (the “matcher”) is happy to contribute part of the amount needed as long as the specific purpose is achieved; therefore, the matcher makes the gift conditional on other gifts.
• Influence matching. The matcher wishes both to support a particular charity and to encourage others to give to that charity. Therefore, the matcher makes a legitimate commitment to give only if others do, in an attempt to influence their giving.

In both of these cases, it may seem at first glance that a one-to-one match really does “double” your donation, but I don’t think it’s quite that simple.

Regarding coordination matching – I would guess it’s relatively common for a funder to say privately, for example, “I’ll give $100,000 if you can raise the remaining needed $900,000.” But there are a couple of problems when it comes to advertising this situation as a “match.” First, saying “every $9 you give will be matched with $1 from a major donor” wouldn’t be very psychologically compelling – matches rarely go below the 1:1 threshold. Second, even if the funder were providing enough for a 1:1 match, it still wouldn’t be quite true that each $1 was matched with another $1: the match would occur only in the case that the total amount needed was raised. So while “coordination matching” is a possibility, we would guess that it rarely explains the “each $1 you give will be matched by $1” campaigns commonly used in fundraising.

Influence matching is something I think impact-maximizing donors ought to be concerned about. In the short run, influence matching makes it true that your $1 donation results in $2 donated to the charity in question. But it also means that you’ve let the matching funder influence your giving – perhaps pulling you away from the most impactful charity (in your judgment) to a less impactful one – just by the way they structured their gift. By giving, you are rewarding this behavior by the matching funder, and you may be encouraging them to take future unconditional gifts and turn them into conditional gifts, because of the ability to sway other donors.

Perhaps, rather than giving your $1 to the charity the matching funder is pushing, you should fight back by structuring your own influence matching – making a conditional commitment to the highest-impact charity you can find, in order to pull other dollars in toward it.

For the average donation match, it’s unclear to what extent the match represents illusory matching vs. coordination matching vs. influence matching. My guess is that coordination matching is by far the least common (since it requires such a specific set of circumstances to hold) and that illusory matching is the norm (since this is generally the easiest to offer, and since donors don’t tend to distinguish between the different types when they decide where and how much to give).

Corporate matching programs sometimes match only gifts to specific charities; in this case I think it’s best to think of them as “influence matching.” If the company offered matching to any charity (as some companies do) and/or simply made gifts to the charities of its choice, it would no longer be pushing its employees to support specific charities. If you are employed at a company offering matching only on specific charities, I recommend pushing for a change in policy (to unconditional gifts to charities and/or unconditional matching for employees, as other companies do) rather than perpetuating a dynamic where your company’s corporate philanthropy team decides where you give.

In general, I advise donors seeking to maximize their impact to simply support the most impactful charity possible, and not to factor in the presence or absence of donation matching either way. If you support a less impactful charity due to the presence of a match, you may be having more total impact, but you also may be having substantially less (in the case of illusory matching) and/or contributing a dynamic that leads to less effective giving broadly (a risk both for influence matching and illusory matching).

GiveWell may offer donation matching sometime in the future. If so, we will be explicit about whether it is influence matching or coordination matching (we wouldn’t be comfortable offering illusory matching, except perhaps as a joke – i.e., “If you’re thinking of giving to another charity just because of a donation match, let us know and we’ll get your donation to a top-rated charity matched”). If we do implement influence matching, it will be to (a) fully neutralize the effect of other matches on impact-oriented donors, further encouraging them to support the most impactful charities possible; (b) raise money from non-impact-oriented donors who are happy to have their donation “matched” despite the logic above.

*“Donation matching” refers to when a large funder offers to give $X to a particular charity for every $Y other people give – for example, “For every $1 you give to this charity, a large funder will contribute another $1, doubling your impact!” For more, see the 2007 study on donation matching by Dean Karlan.