The GiveWell Blog

Orphanages

(updated on: ) | by Holden

We haven’t done much work on charities that try to help orphans and vulnerable children, and we intend to do more. Here are some preliminary thoughts, though.

At first glance, this area might seem among the simplest and least controversial. SOS Children’s Villages states, “Our sponsors and donors help children whose parents are not there for them. They may be AIDS orphans, street children, child soldiers or children orphaned by war, poverty or natural disasters. We give these children a mother and a family in a home within an SOS Children’s Village. Donations pay to build the Villages and run them until child sponsors cover the running costs.” Could any sort of “impact evaluation” be helpful here? How can one deny that children without homes should be provided homes if at all possible?

However, the picture becomes far more complex upon reading something like Saundra Schimmelpfennig’s series on orphanages.

  • Donor demand for funding orphanages may be outstripping actual need (we have speculated that cleft surgery and microcredit may face similar issues).
  • Many of the children in orphanages are not actually orphans. Parents may send them there because they find caring for them to be too expensive; orphanages may weaken the incentives for children’s other relatives and community members to take them in.
  • If, in fact, orphanages are one option rather than the only option for care, it becomes much more crucial to determine whether they are providing good conditions for children. Ms. Schimmelpfennig raises questions about this issue.
  • There is an ongoing debate in academia about whether abandoned children are better off in institutions or being cared for by relatives/community. One person with field experience told me he personally saw a situation in which he believed that orphanages were actively making the situation worse, and Ms. Schimmelpfennig’s series also implies that this is a serious possibility.

None of this means that donating to orphanages is a bad idea. What it means is that, as usual, the appealing story you see on a charity’s website has a great deal of complexity and open questions behind it. As usual, it is essential to ask critical questions, and not to let your due diligence end with “That sounds like a clear need.”

Smile Train in its own words

(updated on: ) | by Holden

We recently argued that Smile Train has “more dollars than doctors” for its core program. In that light, yesterday’s Virginian-Pilot article (which quotes me) is interesting:

  • The main story is that Smile Train has been trying to make substantial and unrestricted grants to another major cleft surgery charity, Operation Smile. This despite the fact that Smile Train has in the past raised concerns about Operation Smile (“intentionally fabricated tens of thousand of surgeries… distorted its financial reporting… squandered millions of dollars… provided shoddy medical care….”)
  • The story quotes Smile Train’s President as appearing to explicitly support the “more dollars than doctors” idea: “Smile Train focused on funding operations by doctors in the countries in need … Mullaney concedes, though, that in some countries, such as Somalia and Haiti, the need outstrips the number of surgeons available to do the work.”

It’s hard to make sense of Smile Train’s wish to make unrestricted grants to Operation Smile, except by accepting that Smile Train is out of room for more funding in its core program.

Possibly, Smile Train’s concerns about Operation Smile have been addressed. Arguably, the decision to grant extra funds to other organizations is admirable (we don’t know whether other charities respond to the same situation by simply piling up assets). But it certainly seems difficult to argue that Smile Train’s donors should think of themselves as funding more of the “$250 per surgery” core program.

Going deeper on “room for more funding”: Underfunded grant proposals

(updated on: ) | by Holden

We previously discussed some simple ways to get at the difficult question of “room for more funding.” One approach that has been more difficult than expected is asking charities themselves to help figure out where more funds are likely to go – answers tend to be vague and tend to target what the charities think we want to hear. How to get specific, concrete, and credible info?

One approach can work specifically for grantmaking charities, i.e., charities that pool donations and make large grants to other organizations. Grantmaking charities are more common than you might think in international aid – arguably any large organization like UNICEF or CARE can be thought of as making grants from its central office to its many projects around the world.

The question we ask such charities is: Can you share examples of strong grant proposals that went underfunded – or rejected – because of insufficient funds?

This is a very different request from asking for rejected grant proposals. An organization may be funding a small percentage of grant proposals, while still funding all reasonable ones. However, an organization that can show proposals that are both strong and underfunded is making a strong case that it has room for more funding.

The Against Malaria Foundation, our 3rd-ranked international aid charity, gave a stellar response to this request. At its GiveWell review, you can see both approved and rejected proposals, and it seems fairly clear that funds are indeed the bottleneck to approving distribution of more insecticide-treated nets.

Asking for a grantmaking organization’s “best unfunded grant proposals” is a pretty simple idea. But you won’t see it requested in any tax return, included in any online charity information database, or mentioned in any of the various impact/outcomes frameworks that various nonprofit sector actors are putting together. Evidence that the “room for more funding” question is largely neglected as of today.

Some simple ways to check “room for more funding”

(updated on: ) | by Holden

We have been struggling with the “room for more funding” question since the first days of GiveWell, and we have gradually developed a variety of approaches to it.

The most basic approach, and the one we’ve used for most of our history, consists of the following:

  1. Gain confidence in an entire organization; do not overfocus on one program.
  2. Examine financial data, looking for a few basic patterns and warning signs.
  3. Ask the charity how additional funds will be used.

Gain confidence in an entire organization

We generally seek to scrutinize and examine activities accounting for over 50% (at a minimum) of a charity’s budget. Some organizations are small enough or simple enough that doing so is fairly straightforward. Other organizations are large and complex but very well-documented. Organizations that are both complex and poorly documented generally don’t get past the first stage of our process.

Our decision to evaluate entire organizations instead of individual programs has, arguably, drastically reduced our options for recommended charities. We have found that many large organizations can’t even answer the “What do you do?” question at the organization level.

But as of now, we see no other reasonable choice. We strongly doubt that donating to a particular program is wise or efficacious.

If you gain confidence in a whole organization, you can give there without worrying too much about what specific activity is next on the agenda. As long as you continue to hold the organization accountable over time.

Basic patterns in financial data

Late in the process with a strong organization, we will analyze its financial data. The details of our financial metrics here. Questions we ask include:

  • Is the charity large enough that it can plausibly absorb substantial additional funds? (This question is actually applied at the beginning of our process.)
  • Have the charity’s expenses been growing over time, implying that it is on a general trajectory of expansion?
  • Does the charity have a reasonable level of assets, given its size? If its assets are too low, we worry that it isn’t stable; if they are too high, we worry that it is piling up reserves because it cannot productively spend additional funds.
  • Which programs does the charity spend most of its funds on? Which programs have been expanding in the past, and are projected for future expansion, implying that they are on a general trajectory of expansion? (Note that this question generally requires a different kind of financial data than is provided in audited financials and tax returns. And we have been surprised at some of the charities that cannot/will not share such financial information.)

We don’t have a set formula; no single “No” answer will disqualify a charity from getting recommended. But asking these basic questions has raised serious questions about some charities (examples: Smile Train, The Carter Center), while our top charities have more or less sailed through these basic tests.

Asking charities what their plans are

This approach has turned out to be far more difficult than it sounds.

In our first year of research, we used a grant application with a question asking specifically:

What would a significant increase in funding (including, but not limited to, a Clear Fund grant) allow your organization to do that it could not do otherwise?

You can see answers via the grant applications submitted by U.S. equality of opportunity charities. Overall, responses were not very helpful.

  • Charities have a tendency to try to tell funders what they want to hear.
  • Charities often are very bad at guessing what we want to hear.

In many cases we followed up with charities and tried harder to make our meaning clear and get meaningful answers, but it’s only recently that we’ve really developed questions that seem to get us somewhere. We’ll be discussing these in future posts.

“Room for more funding” continued: Why donation restricting isn’t the easy answer

(updated on: ) | by Holden

Yesterday we discussed the difficult question of “room for more funding”: how can a donor determine how more funding will translate to more activities?

One common practice is to try to “force” your donation to fund the activities that attract you. Charities will formally honor your restriction by allocating your funds to the program in question.

But let’s look at an example of this in practice. Say you want to support Smile Train‘s core program of funding developing-world doctors, so you restrict your donation to that program. This restriction can’t change the fundamental underlying issue that we perceive: Smile Train has more funds available for this program than it has appropriate doctors.

So what happens? Smile Train might refuse your donation, but what seems more likely is that they’ll restrict it to the program you requested – freeing up an equivalent amount of unrestricted funding for their other activities (research, “provid[ing] materials on cleft lip and palate for free to anyone interested in this birth defect,” etc.) While “your” money pays for the core program, the effect of your donation is that more of the other programs get funded.

In other words, the idea that your fund is “restricted” to the core program is a close parallel to the illusion that it is “restricted” to non-overhead costs.

Could there be cases in which a restriction “works”?

We think so. It seems to us that the necessary conditions are that

  • The program you restrict your donation to has no unrestricted funding allocated to it – so your donation can’t “replace” such funding, and the charity must react to your donation by increasing that program’s budget. (More precisely, if the amount of unrestricted funding allocated to the program exceeds the size of your donation, your donation will merely “crowd out” unrestricted funding; if the amount of unrestricted funding is less than the size of your donation, your donation will have to increase the program’s budget size by such amount. Thus, the larger your donation is, the more likely it is to be able to actually affect an organization’s priorities.)
  • The program you restrict your donation to can productively use more funds (not a given).

How often do the conditions above actually hold? We aren’t sure, and have been trying to get a clearer picture. We have been working on analyzing major international charities’ restricted/unrestricted funding situations, and so far have produced the following using their most recent audited financial statements (generally from 2008):


The purple, orange, green, and red all represent revenues that are “restricted” in some way, and the black represents the fundraising and administrative costs that presumably have to be covered with unrestricted revenues. The blue is calculated as unrestricted revenue minus “operating costs” – i.e., revenue that is likely free to be used at the charity’s discretion.

Since several of these lines are distorted by large (and possibly concerning) amounts of in-kind donations (i.e., gifts of goods rather than cash), we reproduce the charts with in-kind donations excluded to get a clearer picture of the restrictions on cash revenue:

What seems clear is that most large charities have a sizable chunk of “free unrestricted” funding available. This doesn’t fully answer the question we posed – there still could be some programs that are funded almost entirely with restricted funding (and that could be expanded further).

More investigation is needed. In the meantime, we note that

  • We would guess that cases fitting the conditions for “meaningful restricted funding” are rare – i.e., when you give to a multiprogram organization, your donation usually will expand what they want to expand, regardless of how you restrict it.
  • We have a general aversion to restricting donations. It seems like “micromanaging” an organization in this way is asking for trouble: the charity may avoid your intentions using technicalities or spend the “extra money” allocated to a program badly, and in any case, you are creating an extra headache for the charity.

Thus, our current rule of thumb is to find an organization whose existing priorities you are comfortable with – and give unrestricted.

Further challenges raised by this rule of thumb

The above rule of thumb can be tricky to apply, because you have to (a) identify what counts as an “organization” (b) identify the organization’s priorities, which (as we previously discussed) can be very difficult. We’ll briefly discuss (a) here, and discuss (b) in future posts.

Cases where it’s not obvious what counts as an “organization”:

  • One of our top charities, the Stop Tuberculosis Partnership, takes donations through the UN Foundation, and requests that they be earmarked for the “Stop Tuberculosis Partnership.” Should such gifts be thought of as restricted gifts to the UN Foundation or as unrestricted gifts to Stop TB?
  • A recent debate on Tactical Philanthropy brings up a cancer research program within a university. Can/should one meaningfully earmark a donation to this program as opposed to the university as a whole?

To us, the ultimate test is “Do the people who exercise discretion over the pool of funds including my money have priorities that I’m comfortable with?” We perceive the UN Foundation as a pass-through that does not exercise discretion over the Stop Tuberculosis Partnership budget. We would guess that the university situation is somewhere in the middle, with different parties exercising different amounts of discretion, but that to a large extent, the cancer research program is responsible for raising its own funding rather than reliant on the discretion of the university. Both of these cases contrast with something like earmarking a CARE donation for a particular country – that donation is going to go into the pool of funds allocated with discretion by the central office.

Future posts will discuss some ideas for tackling the more difficult part (b): getting a handle on an organization’s true priorities, even as it tries to assure you that your funds will pay for the program that appeals to you most.

An essential question that no one is asking charities

(updated on: ) | by Holden

If a charity demonstrates that its core program has changed lives in the past, is likely to change lives in the future, and gets great “bang for your buck,” is this enough reason to donate to it? We say no.

The missing piece: Will more funding lead to more of the good program(s)? We generally call this the “room for more funding” question, and we’ve seen next to no helpful discussion of the issue within academia, within the nonprofit sector, or anywhere else.

Often, when I raise this issue, the response I get is “But is that a real problem? Are there charities that have great programs they can’t or won’t expand with more funding?” The answer is yes. Examples:

  • Our analysis of Smile Train strongly suggests that its core program of directly funding doctors has “more money than doctors.” Thus, over 50% of Smile Train’s funds go to activities far from what their fundraising focuses on, including grants to other organizations, research, and “provid[ing] materials on cleft lip and palate for free to anyone interested in this birth defect.” Perhaps these activities have value, but it would be a mistake to donate to Smile Train just based on their headline program.
  • The Aravind Eye Care System is one of the more impressive humanitarian organizations we have seen, performing vision-restoring surgery extremely cost-effectively. They have been so successful, in fact, that their core program doesn’t need donations – as they have explicitly told us. Revenue from for-pay surgeries subsidizes free surgeries, and donations are used on entirely and substantially different programs such as distribution of spectacles and free food.
  • Today’s Aid Watch post gives an excellent picture of why it’s so important to be wary:

    according to Fred Martin, Communications Director at CHF, “In fact our Food Pak program is a small portion of what we do. We highlight it because it is our flagship program that we’ve seen work very well in building relationships with the poor so that deeper needs can be uncovered and responded to.” I learned from Fred they also provide beds in eastern Europe and medicines in Asia …

    As long as charities can get away with it, their incentive is to advertise the best program they have, even well beyond the point where that’s the program that needs more money.

There can be many bottlenecks to expanding a program besides money (skilled labor, environments that are conducive to the program, etc.) If you want to fund great programs, you have to ask not just “What have you done and has it worked?” but “What will you do with more funds than you’re currently expecting?”

We haven’t identified any easy answer or simple formula for this question. We believe that “restricting” your donation to the program you favor is generally a futile endeavor (more on this in a future post).

We have developed some relevant ideas. In addition to some rules of thumb for avoiding the most tangled cases, we ask the strongest charities for documents that speak to the “room for more funding” issue directly, such as examples of un-funded but strong project proposals and financial “scenario analysis” (details to come in future posts).

However, we have found that requesting such documents is an uphill battle because the request is generally so foreign. You won’t see financial scenario analysis on any standard list of “documents a charity should be sharing” (from the IRS or anyone else).

Foundations arguably don’t need to deal with the challenge discussed here, because they can give money in large enough chunks to dictate which projects get carried out. (The extent to which this practice is wise is another question). This may be why no one else seems to be asking for information on “room for more funding.” Whatever the reason, it’s an issue that needs much more attention than it’s getting.