[SoGive created the template below; the board of a foundation can follow the steps herein to come to an agreement about how they want to identify the most effective funding opportunities. Individual donors may adjust this according to their circumstances. As can be seen, this is based on a worldview diversification approach to giving.]
[Who is SoGive? SoGive works with donors, especially major donors, to support them to have more impact with their giving. If you are interested in having a conversation about your giving, please reach out to Spencer (spencer@sogive.org).]
Executive Summary
This strategic document outlines the procedures by which the [XYZ Foundation] will identify, evaluate, and prioritize funding opportunities. Our mission is to use our resources to do the most good. This approach will ensure the thoughtful and transparent allocation of our funding.
Overview of Cause Areas
The [XYZ Foundation] will consider the following cause areas for funding:
Global health and development: To improve global health and economic growth among people around the world at present.
Animal advocacy: To improve wellbeing of non-human animals around the world at present.
Longtermism: To ensure that humanity continues to flourish long into the future.
Impact multipliers/meta: To help actors in other cause areas improve their impact by leveraging talent, funding, and research.
Preparation
Well-researched giving: [XYZ Foundation]’s goal to use our resources to do the most good entails making well-researched, thoughtful giving decisions.
However, identifying, evaluating, and administering well-evidenced grants requires a high time investment, including liaising with expert evaluators and grantmakers, and sometimes even including conducting extensive research projects.
[XYZ Foundation] will determine how much time we can dedicate to these activities.
To the extent that we don’t have time to conduct research ourselves, we will do one of the following:
Lean on the research of a trusted evaluator
Commission research ourselves. Where we spend money on commissioning research, we consider this to be a good use of funds, not only because it makes our own philanthropy more impactful, but because we can put the information in the public domain and help other funders have more impact too.
Activities that might be taken on by members of the [XYZ Foundation], or might be delegated, include attending calls/meetings with charities and designing/executing research.
Risk tolerance: [XYZ Foundation] will determine whether they intend to take a risk-averse, risk-neutral, or risk-friendly approach to funding opportunities with a low chance of great success. For more information on this, read this write-up. SoGive recommends a risk-neutral approach.
Measurability and evidence: [XYZ Foundation] will determine the standard to which they want funding opportunities to be measurable and have a strong evidence base. We may choose to favour funding opportunities with strong evidence (e.g. RCTs) or to accommodate speculative funding opportunities if they otherwise fit with our strategy.
Impact investing: [XYZ Foundation] will determine whether to engage only in philanthropy or also in impact investing. Possible attitudes to impact investing include:
“We believe that impact investing is on an equal footing with our philanthropy, and deserves equal amounts of time and attention”
“We will conduct impact investing, but believe that the impact achieved through impact investing is likely to be secondary to the impact achieved by our philanthropy, and hence we will pay only a small amount of attention to it.”
“We will (largely) not conduct impact investing.”
Give now/give later: [XYZ Foundation] will determine what proportion of its funds to put up for funding opportunities each year and what proportion to hold for later years regardless of the available opportunities. The Foundation will also determine how to hold or invest these funds.
Procedure for Prioritizing Cause Areas
Worldview diversification: [XYZ Foundation] will select a strategy for allocating funding to different cause areas. SoGive recommends sticking with a credence-based strategy. (For the Foundation’s interest, this is a list of other possible worldview diversification strategies.)
For the credence-based strategy, the trustees would spend at least one hour discussing their confidence in the normative, empirical, and methodological assumptions that entail allocating funding to each of the four cause areas listed above.
The following questions form the basis of an online tool that SoGive is currently developing to support this process. Here are some questions the trustees will consider:
Longtermism
Do we believe that future individuals matter morally, perhaps as much as current individuals? (Note that some philosophers argue for “person-affecting views”; i.e. the idea that “there is never a person who could have benefited from being created”, which means that people who don’t currently exist have less or no moral value.) The position that the future is overwhelmingly important is defended, e.g., here.
Do we believe there is any plausible pure discount rate on the moral value of the future? This view is defended in this GPI paper.
Do we believe that a great number of people, perhaps trillions or more, will exist in the future? (This GPI paper suggests it might be 1030.)
Do we believe it is more important not to create more suffering people than it is to create more happy people? Or do we believe that these are of equal priority?
Do we believe that a great number of digital beings will exist in the future? And is it plausible that those digital beings will be conscious, sentient, or able to experience suffering and pleasure?
Animals
Do we believe that animals matter morally, perhaps as much as humans?
Do we believe that there is any number of animals where, if we could choose to save either one human or at least that many animals, it would not make a moral difference?
Do we believe that animals are conscious, sentient, or able to experience suffering and pleasure? Which of these characteristics do we care about morally?
At the end of the discussion, for each funding area, each trustee would give their confidence that the funding area should be supported by the Foundation as a number between 0 and 1.
[SoGive recommends that, where the Foundation’s confidence levels are being set based on the aggregate of the perspectives of multiple trustees, that trustees should consider in advance the aggregation method – i.e. should the Foundation’s moral weights be determined by consensus (which might be hard to reach) or something more mechanical, such as the median or mean of different trustee responses?]
The Foundation would allocate funding in proportion to these confidence levels, where the funding area with the highest average confidence level gets the most funding.
[SoGive recommends that the strategy should allow for the possibility that it may not be practically possible to align the giving with the targets set based on these theoretical considerations, e.g., because there may not be enough giving opportunities which meet the required thresholds. The strategy should set ranges around these target proportions, or otherwise include some allowances for the possibility that sticking to the ideal split may not be possible.]
Moral weights: [XYZ Foundation] will determine what outcomes they will prioritize funding, and for which beneficiaries.
We will do this by determining the acceptable ratios for trading off outcomes against each other. For each question, choose the value of X such that the two outcomes are equivalent:
Doubling spending/consumption for X people for 1 year = 1 child’s life saved. Information about the link between money and happiness can be found here and here.
Curing severe depression for X people for 1 year (bringing their mental state from negative to neutral) = 1 child’s life saved.
Granting someone X years of happiness (bringing their mental state from neutral to positive) = 1 child’s life saved.
Alleviating the suffering of X animals for 1 year = 1 child’s life saved. The Foundation may want to set different moral weights for different animals, e.g., based on their estimated capacity to experience pleasure or suffering. Open Philanthropy’s review of animal consciousness topic highlighted the complexities and research gaps in this area.
Saving 1 life today = Saving X lives in the far future (e.g. 500 years from now).
One type of tool for determining these tradeoffs is the MoralMeter, produced by SoGive, which trustees may use if they wish.
SoGive recommends that, where the Foundation’s moral weights are being set based on the aggregate of the perspectives of multiple trustees, that trustees should consider in advance the aggregation method – i.e. should the Foundation’s moral weights be determined by consensus (which might be hard to reach) or something more mechanical, such as the median or mean of different trustee responses?
Explore/exploit: [XYZ Foundation may wish to determine in advance, within each cause area, how much funding will go directly to interventions (exploit) and how much funding will go to research projects (explore). For the purposes of this policy document, this could once again be specified as a target proportion of annual donations, although again, this should likely allow for some wiggle room. XYZ Foundation may also decide to suspend making a decision on this in advance.]
Projects that impact multiple worldviews: [Suppose XYZ Foundation is considering funding a project that simultaneously impacts multiple cause areas, such as by having a benefit both for present humans and future humans. This would result in the project being a candidate for funding under multiple worldviews.
XYZ Foundation may wish to determine a policy for assessing organizations that have impact on multiple cause areas. The Foundation may also consider this on a case by case basis – likely the best approach if the Foundation is starting its giving journey.
One option for this is evaluating the project’s impact on each relevant cause area separately. This might look like creating a cost-effectiveness analysis for each relevant cause area, and assessing the cost-effectiveness within each cause area against the relevant thresholds. If the project surpasses the threshold in multiple cause areas, it is considered as a candidate for funding in each bucket. (Note that it is rare for a project to be one of the most impactful opportunities for multiple cause areas, even if the work happens to impact multiple cause areas.)
Another option is to arbitrarily treat the organization as though it only impacts one cause area. While this has clear downsides, it has the advantage of simplicity.]
Procedure for Identifying Funding Opportunities
Proactive research: We will proactively commission research projects on potential organizations and initiatives that align with our funding priorities.
Coattailing and funging with trusted evaluators: We will donate to organizations recommended by trusted evaluators such as Open Philanthropy and learn more about their intervention through the relationship. This is described in more detail here.
[XYZ Foundation may consider an open call for proposals, issued annually, inviting nonprofits anywhere in the world to submit project proposals aligned with the cause areas above. SoGive does not recommend this for small-to-medium-sized foundations, because a thorough open call is effort-intensive.]
Evaluation
We will conduct a review of all funding opportunities based on the following criteria:
Alignment with cause areas: The project should align with one or more of the cause areas that [XYZ] is considering for funding.
Measurability and evidence: The project should have the potential to generate an impact that is measurable and evidence-based to the standards that [XYZ Foundation] has set.
Cost-effectiveness: The project should have a cost-per-outcome on par with the most cost-effective interventions in the cause area. [In principle, you may find it helpful to define this using an explicit threshold, as outlined below.]
Theory of change: The organization should have a theory of change, with robust reasoning for how their activities will lead to the desired outcome for the beneficiaries.
Transparency: The organization should have a policy for transparently reporting its impact to grantmakers.
Sustainability: The project should demonstrate how it will be sustainable beyond the funding period.
Budget feasibility: The project’s budget should be realistic.
Organizational capacity: The organization should have the capacity to successfully implement the project. They should have a clear strategy and goals. They should be able to demonstrate that their management has a track record of successfully implementing similar projects.
Funding opportunities will be scored on the above, or we will delegate the assessment to an external party, whom we believe is capable of taking into account all of the above criteria.
Prioritizing Funding Opportunities
Scoring: Once projects are evaluated, they will be ranked based on their score.
Funding order: Funding will be allocated to cause areas according to the worldview diversification strategy. Then, projects within a cause area will be funded in the order of highest score. The project with the highest score will be funded until it can no longer take in funding as effectively as the next project. Then, the next-highest scoring project will be funded.
Threshold: [Ideally, your strategy should require funding opportunities to exceed a certain threshold for cost-effectiveness for each cause area. For example, a threshold-setting exercise for longtermism is here. However, in some cause areas, explicit cost-effectiveness estimates are not available for most giving opportunities, even those which have research indicating that they are high-impact. Your strategy should allow for this reality.]
Decision-Making Process
The final decision on funding allocations will be made by the Foundation’s trustees. All decisions will be made in a meeting that allows for open discussion and consensus-building.
Monitoring and Evaluation
The funded organizations will be required to submit progress reports.
Learning and Improvement
We will continuously strive to learn from our experiences to improve our funding strategy. We will seek feedback from our grantee partners and engage in periodic reviews of our strategy and processes.
Conclusion
The goal of this strategic funding approach is to ensure that [XYZ Foundation] maximizes its impact through thoughtful and transparent decisions. We look forward to collaborating with our partners to create lasting change.
Template: Strategic Funding Approach for An Effective Foundation
[SoGive created the template below; the board of a foundation can follow the steps herein to come to an agreement about how they want to identify the most effective funding opportunities. Individual donors may adjust this according to their circumstances. As can be seen, this is based on a worldview diversification approach to giving.]
[Who is SoGive? SoGive works with donors, especially major donors, to support them to have more impact with their giving. If you are interested in having a conversation about your giving, please reach out to Spencer (spencer@sogive.org).]
Executive Summary
This strategic document outlines the procedures by which the [XYZ Foundation] will identify, evaluate, and prioritize funding opportunities. Our mission is to use our resources to do the most good. This approach will ensure the thoughtful and transparent allocation of our funding.
Overview of Cause Areas
The [XYZ Foundation] will consider the following cause areas for funding:
Global health and development: To improve global health and economic growth among people around the world at present.
Animal advocacy: To improve wellbeing of non-human animals around the world at present.
Longtermism: To ensure that humanity continues to flourish long into the future.
Impact multipliers/meta: To help actors in other cause areas improve their impact by leveraging talent, funding, and research.
Preparation
Well-researched giving: [XYZ Foundation]’s goal to use our resources to do the most good entails making well-researched, thoughtful giving decisions.
However, identifying, evaluating, and administering well-evidenced grants requires a high time investment, including liaising with expert evaluators and grantmakers, and sometimes even including conducting extensive research projects.
[XYZ Foundation] will determine how much time we can dedicate to these activities.
To the extent that we don’t have time to conduct research ourselves, we will do one of the following:
Lean on the research of a trusted evaluator
Commission research ourselves. Where we spend money on commissioning research, we consider this to be a good use of funds, not only because it makes our own philanthropy more impactful, but because we can put the information in the public domain and help other funders have more impact too.
Activities that might be taken on by members of the [XYZ Foundation], or might be delegated, include attending calls/meetings with charities and designing/executing research.
Risk tolerance: [XYZ Foundation] will determine whether they intend to take a risk-averse, risk-neutral, or risk-friendly approach to funding opportunities with a low chance of great success. For more information on this, read this write-up. SoGive recommends a risk-neutral approach.
Measurability and evidence: [XYZ Foundation] will determine the standard to which they want funding opportunities to be measurable and have a strong evidence base. We may choose to favour funding opportunities with strong evidence (e.g. RCTs) or to accommodate speculative funding opportunities if they otherwise fit with our strategy.
Impact investing: [XYZ Foundation] will determine whether to engage only in philanthropy or also in impact investing. Possible attitudes to impact investing include:
“We believe that impact investing is on an equal footing with our philanthropy, and deserves equal amounts of time and attention”
“We will conduct impact investing, but believe that the impact achieved through impact investing is likely to be secondary to the impact achieved by our philanthropy, and hence we will pay only a small amount of attention to it.”
“We will (largely) not conduct impact investing.”
Give now/give later: [XYZ Foundation] will determine what proportion of its funds to put up for funding opportunities each year and what proportion to hold for later years regardless of the available opportunities. The Foundation will also determine how to hold or invest these funds.
Procedure for Prioritizing Cause Areas
Worldview diversification: [XYZ Foundation] will select a strategy for allocating funding to different cause areas. SoGive recommends sticking with a credence-based strategy. (For the Foundation’s interest, this is a list of other possible worldview diversification strategies.)
For the credence-based strategy, the trustees would spend at least one hour discussing their confidence in the normative, empirical, and methodological assumptions that entail allocating funding to each of the four cause areas listed above.
The following questions form the basis of an online tool that SoGive is currently developing to support this process. Here are some questions the trustees will consider:
Longtermism
Do we believe that future individuals matter morally, perhaps as much as current individuals? (Note that some philosophers argue for “person-affecting views”; i.e. the idea that “there is never a person who could have benefited from being created”, which means that people who don’t currently exist have less or no moral value.) The position that the future is overwhelmingly important is defended, e.g., here.
Do we believe there is any plausible pure discount rate on the moral value of the future? This view is defended in this GPI paper.
How much credence do we place on the Vulnerable World Hypothesis?
Do we believe that a great number of people, perhaps trillions or more, will exist in the future? (This GPI paper suggests it might be 1030.)
Do we believe it is more important not to create more suffering people than it is to create more happy people? Or do we believe that these are of equal priority?
Do we believe that a great number of digital beings will exist in the future? And is it plausible that those digital beings will be conscious, sentient, or able to experience suffering and pleasure?
Animals
Do we believe that animals matter morally, perhaps as much as humans?
Do we believe that there is any number of animals where, if we could choose to save either one human or at least that many animals, it would not make a moral difference?
Do we believe that animals are conscious, sentient, or able to experience suffering and pleasure? Which of these characteristics do we care about morally?
At the end of the discussion, for each funding area, each trustee would give their confidence that the funding area should be supported by the Foundation as a number between 0 and 1.
[SoGive recommends that, where the Foundation’s confidence levels are being set based on the aggregate of the perspectives of multiple trustees, that trustees should consider in advance the aggregation method – i.e. should the Foundation’s moral weights be determined by consensus (which might be hard to reach) or something more mechanical, such as the median or mean of different trustee responses?]
The Foundation would allocate funding in proportion to these confidence levels, where the funding area with the highest average confidence level gets the most funding.
[SoGive recommends that the strategy should allow for the possibility that it may not be practically possible to align the giving with the targets set based on these theoretical considerations, e.g., because there may not be enough giving opportunities which meet the required thresholds. The strategy should set ranges around these target proportions, or otherwise include some allowances for the possibility that sticking to the ideal split may not be possible.]
Moral weights: [XYZ Foundation] will determine what outcomes they will prioritize funding, and for which beneficiaries.
We will do this by determining the acceptable ratios for trading off outcomes against each other. For each question, choose the value of X such that the two outcomes are equivalent:
Doubling spending/consumption for X people for 1 year = 1 child’s life saved. Information about the link between money and happiness can be found here and here.
Curing severe depression for X people for 1 year (bringing their mental state from negative to neutral) = 1 child’s life saved.
Granting someone X years of happiness (bringing their mental state from neutral to positive) = 1 child’s life saved.
Alleviating the suffering of X animals for 1 year = 1 child’s life saved. The Foundation may want to set different moral weights for different animals, e.g., based on their estimated capacity to experience pleasure or suffering. Open Philanthropy’s review of animal consciousness topic highlighted the complexities and research gaps in this area.
Saving 1 life today = Saving X lives in the far future (e.g. 500 years from now).
One type of tool for determining these tradeoffs is the MoralMeter, produced by SoGive, which trustees may use if they wish.
SoGive recommends that, where the Foundation’s moral weights are being set based on the aggregate of the perspectives of multiple trustees, that trustees should consider in advance the aggregation method – i.e. should the Foundation’s moral weights be determined by consensus (which might be hard to reach) or something more mechanical, such as the median or mean of different trustee responses?
Explore/exploit: [XYZ Foundation may wish to determine in advance, within each cause area, how much funding will go directly to interventions (exploit) and how much funding will go to research projects (explore). For the purposes of this policy document, this could once again be specified as a target proportion of annual donations, although again, this should likely allow for some wiggle room. XYZ Foundation may also decide to suspend making a decision on this in advance.]
Projects that impact multiple worldviews: [Suppose XYZ Foundation is considering funding a project that simultaneously impacts multiple cause areas, such as by having a benefit both for present humans and future humans. This would result in the project being a candidate for funding under multiple worldviews.
XYZ Foundation may wish to determine a policy for assessing organizations that have impact on multiple cause areas. The Foundation may also consider this on a case by case basis – likely the best approach if the Foundation is starting its giving journey.
One option for this is evaluating the project’s impact on each relevant cause area separately. This might look like creating a cost-effectiveness analysis for each relevant cause area, and assessing the cost-effectiveness within each cause area against the relevant thresholds. If the project surpasses the threshold in multiple cause areas, it is considered as a candidate for funding in each bucket. (Note that it is rare for a project to be one of the most impactful opportunities for multiple cause areas, even if the work happens to impact multiple cause areas.)
Another option is to arbitrarily treat the organization as though it only impacts one cause area. While this has clear downsides, it has the advantage of simplicity.]
Procedure for Identifying Funding Opportunities
Proactive research: We will proactively commission research projects on potential organizations and initiatives that align with our funding priorities.
Coattailing and funging with trusted evaluators: We will donate to organizations recommended by trusted evaluators such as Open Philanthropy and learn more about their intervention through the relationship. This is described in more detail here.
[XYZ Foundation may consider an open call for proposals, issued annually, inviting nonprofits anywhere in the world to submit project proposals aligned with the cause areas above. SoGive does not recommend this for small-to-medium-sized foundations, because a thorough open call is effort-intensive.]
Evaluation
We will conduct a review of all funding opportunities based on the following criteria:
Alignment with cause areas: The project should align with one or more of the cause areas that [XYZ] is considering for funding.
Measurability and evidence: The project should have the potential to generate an impact that is measurable and evidence-based to the standards that [XYZ Foundation] has set.
Cost-effectiveness: The project should have a cost-per-outcome on par with the most cost-effective interventions in the cause area. [In principle, you may find it helpful to define this using an explicit threshold, as outlined below.]
Theory of change: The organization should have a theory of change, with robust reasoning for how their activities will lead to the desired outcome for the beneficiaries.
Transparency: The organization should have a policy for transparently reporting its impact to grantmakers.
Sustainability: The project should demonstrate how it will be sustainable beyond the funding period.
Budget feasibility: The project’s budget should be realistic.
Organizational capacity: The organization should have the capacity to successfully implement the project. They should have a clear strategy and goals. They should be able to demonstrate that their management has a track record of successfully implementing similar projects.
Funding opportunities will be scored on the above, or we will delegate the assessment to an external party, whom we believe is capable of taking into account all of the above criteria.
Prioritizing Funding Opportunities
Scoring: Once projects are evaluated, they will be ranked based on their score.
Funding order: Funding will be allocated to cause areas according to the worldview diversification strategy. Then, projects within a cause area will be funded in the order of highest score. The project with the highest score will be funded until it can no longer take in funding as effectively as the next project. Then, the next-highest scoring project will be funded.
Threshold: [Ideally, your strategy should require funding opportunities to exceed a certain threshold for cost-effectiveness for each cause area. For example, a threshold-setting exercise for longtermism is here. However, in some cause areas, explicit cost-effectiveness estimates are not available for most giving opportunities, even those which have research indicating that they are high-impact. Your strategy should allow for this reality.]
Decision-Making Process
The final decision on funding allocations will be made by the Foundation’s trustees. All decisions will be made in a meeting that allows for open discussion and consensus-building.
Monitoring and Evaluation
The funded organizations will be required to submit progress reports.
Learning and Improvement
We will continuously strive to learn from our experiences to improve our funding strategy. We will seek feedback from our grantee partners and engage in periodic reviews of our strategy and processes.
Conclusion
The goal of this strategic funding approach is to ensure that [XYZ Foundation] maximizes its impact through thoughtful and transparent decisions. We look forward to collaborating with our partners to create lasting change.