Thursday, December 12, 2013

How to think about displacing Good Ventures in funding GiveWell Labs?

Summary: It appears likely that visibly investing funds to donate next year would be a better bet than donating to GiveWell's 2013 picks, based on expected 2014 opportunities. How does this compare to donating to GiveWell's research fundraising, i.e. substituting for dollars that would otherwise be raised by Good Ventures? Since Good Ventures is investing most of its resources (with economies of scale in investment) for use when better information is available, this can be seen as another way to invest for future donation prospects. The value of doing this, compared to a donor-advised fund, should depend on one's beliefs about possible future disagreements, economies of scale, signalling, and incentives. Those who would firmly plan to give from their donor-advised funds based on GiveWell and Good Ventures recommendations have a good case for supporting GiveWell research, even if is not expected to increase research activity. 
[Disclaimer: this is a first-pass analysis and should be taken with a grain of salt.]

In a recent post, I argued that placing funds in a Donor-Advised Fund (DAF) or the Giving What We Can Trust to await better opportunities looked more promising than donating to GiveWell's current recommended charities. Reasons include potential for more room for more funding at AMF, GiveWell Labs recommendations, and others. I also suggested that GiveWell research appears more valuable than the 2013 top picks, especially for those interested in long run outcomes, as GiveWell Labs research appears likely to be more applicable to that perspective than its work so far. But how should we weigh these two options against each other? I review a few considerations below, although without a firm conclusion.

Donation to GiveWell's operations replaces Good Ventures donations in funding GiveWell Labs

GiveWell has begun to raise funds for its own operations raising:

As a result, our projected expenses have risen significantly, and we now have the largest projected funding gap in our history: we project ~$1.2 million in expenses over the next year, against ~$850,000 in revenue. We have about $900,000 of reserves available, so failing to close the gap would not mean insolvency, but it would mean drawing on our reserves, something we seek to avoid. So fundraising will become a significant priority until the gap is closed (and may continue until we also have a comfortable level of reserves on hand, i.e., ~12 months’ worth).
In some ways, our projected budget may be an overestimate: it allows for a scenario in which we retain all current employees and hire several more. On the other hand, since we expect to keep expanding and since we will need to adjust compensation upward over the long run, we expect our expenses to continue to grow over time.
This funding need could be easily met by Good Ventures, which relies on GiveWell to perform cause and charity selection research, and would be if needed. However, Good Ventures and GiveWell say they would like GiveWell to have a more diversified funding base for various reasons, including:

One reason for this is that it would put GiveWell in an overly precarious position. While our interests are currently aligned, it is important to both parties that we would be able to go our separate ways in the case of a strong enough disagreement. If Good Ventures provided too high a proportion of support to GiveWell, the consequences of a split could become enormous for us, because we wouldn’t have a realistic way of dealing with losing Good Ventures’s support without significant disruption and downsizing. That would, in turn, put us in a position such that it would be very difficult to maintain our independence.
Another reason is that raising substantial support from individuals keeps us accountable to individuals, both in terms of perception and reality. If we did not raise a substantial part of our support from individuals, our incentives would not be aligned with our mission of serving large numbers of donors. We have hopes of serving many more individuals and institutions (such as Good Ventures) in the future; drawing too much of our budget from Good Ventures could make this more difficult by leading to a perception that serving Good Ventures is our main mission.  
GiveWell employee Alexander Berger, however, makes the case that since the increased costs are driven by GiveWell Labs, for which Good Ventures may be the biggest short-term user there is less to gain by diversification:

In this particular instance, and based on my current understanding of our finances, I don’t agree with Holden’s argument that it would be inadvisable to seek more funding from Good Ventures than we currently plan to. Given that he and Elie believe that it would be, though, I think it’s good for donors who use our research to support GiveWell...
As it stands, given that [Holden] and Elie are going to take the time to try to raise funding from individuals, I think it would be good for individuals to contribute.
Suppose that this is right, and donations to GiveWell essentially substitute for Good Ventures. Then a dollar of donation to GiveWell increases Good Ventures' vast reserve funds (which are invested to await later, better, opportunities and increased understanding).

Personal DAF vs Good Ventures?

How should one compare the value of money placed in one's own DAF to funds in Good Ventures' reserves?  How does this compare to putting money in one's own DAF?

First one can ask to what extent one should expect to (correctly!) disagree with Good Ventures' future grant decisions. This might be because of conflicting values, or because of conflicting epistemic priors or incommunicable knowledge. So far Good Ventures' charity evaluation and research have been tightly integrated with GiveWell, thus far, and Good Ventures has expressed major interest in long-run outcomes. These factors should give GiveWell fans and EA types much ground for trust. However, uncertainty, noise, and imperfect communication would all tend to increased the expected value of maintaining a separate DAF, which could mimic Good Ventures' donations while keeping an eye out for more information.

Such information could include the track record of judgment calls, e.g. Good Ventures' launch of a 2013 matching challenge despite GiveWell's disagreement, and their results. More detailed information about ethical stances and empirical worldview is also likely to emerge over time. It should be noted though, that while it is possible that Good Ventures may err in the future and not be open to persuasion, the same may be true of other donors (the difference would be that personal knowledge may be more certain than noisy evidence about others, although various biases complicate the issue).  

Second, Good Ventures may enjoy certain economies of scale in investing funds and subsequent re-granting. While private index funds charge fees of a small fraction of a percent of assets, DAF fees can be several times higher. Good Ventures' large size may allow it to reduce such costs, and otherwise improve net return on invested funds. However, I have not seen information on Good Ventures' investment patterns or performance.

Third, retaining separate funds permits diversifying the funding bases of charities at later times when this is important. This may or may not be important for the GiveWell Labs expansion, but could also arise at a later time, when it might be more or less important.

Fourth, concentrating funds reduces the replication of decisionmaking on later disbursement. This can be a positive, especially for smaller donations, in outsourcing work (although this could also be done by using GiveWell recommendations or mimicry of Good Ventures' donations). It may also be a negative in reducing intellectual diversity, or positive externalities of conversation/persuasion between funders.

Signalling/'voting with cash' effects

Donations can be used to spark conversations, inform media coverage, and otherwise act as powerful signals. If one thinks that the GiveWell Labs work is more valuable at the margin than GiveDirectly, or even a donation to the better opportunities GiveWell might recommend next year (more RFMF at AMF, a novel Labs pick), then donating to that effect (and saying so) is a way to signal this.

I have had many conversations with people looking to influence long run outcomes, but concerned about giving to clear 'giving as consumption' charities like GiveDirectly because of their greater popularity (and potential for building the capacities and scale of the effective altruism movement as a whole). One way to deal with this is to do a donation split for signal purposes, e.g. 50% GiveDirectly, 50% GiveWell research.

But I also wonder whether an even larger focus might make sense in signalling terms. GiveWell is trying to raise an amount for its internal research which is much smaller than the money moved to its current top picks. By supporting GiveWell research, one could say "I support this research because of its track record in moving $N million to well-vetted charities alleviate global poverty, and I care a lot about the kind of interventions affecting long-run outcomes that GiveWell Labs research may help us identify and prioritize." I don't have a strong view here, but it seems plausible that this could be a valuable signal. It might be a way to address a problem described by Ben Kuhn:

Even for those looking to use their donations to set an example, a donor-advised fund would have many of the benefits and none of the downsides. And anyway, donating when you believe it’s not (except for example-setting) the best possible course of action, in order to make a point about figuring out the best possible course of action and then doing that thing, seems perverse.
Displacing Good Ventures in GiveWell funding invests for the future (as Good Ventures is mostly saving and investing), signals priorities more accurately, but also rewards a track record of success in global poverty/health.

I don't end up with a strong generally applicable take-away. For those with very high confidence in Good Ventures (so as to not fear future justified disagreements), the advantages of donating to GiveWell Labs seem like they would win out over a DAF or use of the GWWC trust to donate to GiveWell's top charities next year. Those who worry more about divergence with Good Ventures would face a more difficult choice, but the 'voting with cash' benefits could well be worthwhile. And assessment of signalling and incentive effects require additional judgment calls.

As Ben Kuhn notes in the comments, given that GiveWell and Good Ventures have committed to seeking more outside funding, donations may increase research not by increasing total funding, but by reducing fundraising effort. 

Also, so far my discussion has limited itself to GiveWell, Good Ventures, and their current and future recommendations. As GiveWell writes:

  • If you have access to other giving opportunities that you understand well, have a great deal of context on and have high confidence in — whether these consist of supporting an established organization or helping a newer one get off the ground — it may make more sense to take advantage of your unusual position and "fund what others won't," since GiveWell's research is available to (and influences) large numbers of people.
Still, I hope gathering these thoughts in one place will be helpful to some people this giving season.


Ben Kuhn said...

Note that Alexander Berger also wrote, in his comment against GiveWell raising from individuals:

> As it stands, given that [Holden] and Elie are going to take the time to try to raise funding from individuals, I think it would be good for individuals to contribute.

So I don't think supposing that dollars from individuals only displace dollars of Good Ventures' money is a good assumption. It seems fairly likely to me that GiveWell will eventually be able to meet its funding needs through individuals, so Holden and Elie will continue to raise funds until they meet (or close to meet) their. In that case, you're not buying more funds for Good Ventures at all, you're buying time for Holden and Elie to do research.

Carl said...

Good point Ben. It's a bit tricky to account for this, because of the conditionality of search time on choices by GiveWell and Good Ventures.

Ben Kuhn said...

I've commented on GiveWell's blog asking for clarification on their expected fundraising behavior.