This post introduces the issue of ‘high-hanging fruits’ and identifies possible avenues of action. The epistemic status of this post is extremely rough. I would in particular welcome suggestions for more concrete scenarios than what I suggest below.
1. Possible Scenarios
Consider two possible scenarios:
Due to shifts in the political climate, we are reasonably confident that animal advocacy will be unusually tractable in the next four years. Experts believe that the most effective intervention method is mass street-level political advocacy. However, there is one catch: the intervention will be effective only if more than x number of people are mobilised.[1] Otherwise, the intervention will have little beneficial effects. The EA community can help ensure that more than x number of people are mobilised within four years, but only if it diverts significant amounts of resources away from other cause areas during that period.
Compelling evidence suggests that extremely dangerous climate feedback loops will occur when global warming reaches 2°C. However, scientists believe that some extremely effective intervention may become available in three years. Once the intervention becomes mature, y amount of funds needs to be invested extremely quickly for us to realistically avoid hitting 2°C. We can expect funding from non-EA sources to be insufficient. The EA community can ensure enough funding for the intervention only if it starts saving funds now and subsequently spend virtually all saved financial resources on the intervention once it becomes available.
2. What are ‘High-hanging Fruits’?
Both scenarios described above are instances of what we might call ‘high-hanging fruits’. ‘High-hanging fruits’ are special cases of increasing marginal returns.[2] In both above scenarios, a critical threshold of resources is required in order to achieve a payoff. In other ‘high-hanging fruit’ cases, there may not be a discrete threshold for a payoff. Rather, there might be continuous increasing marginal returns until a certain point. See figure 1 for representations of cost/payoff functions for ‘high-hanging fruits’.
Figure 1: Cost/Payoff Functions in High-Hanging Fruit Scenarios
If reaching a ‘high-hanging fruits’ requires actions by multiple parties, then it presents a coordination problem, where reaching the optimal outcome requires multiple agents acting on a joint strategy.
3. Importance of ‘High-hanging Fruits’
There are at least three reasons for paying attention to ‘high-hanging fruits’:
High-hanging fruits are, by definition, high value.
High-hanging fruits may be prevalent.
High-hanging fruits are likely to be neglected.
Since the first is mostly self-explanatory, we will only discuss the other two reasons.
Why might high-hanging fruits be prevalent? Does this not contradict the standard assumption of diminishing marginal returns? Surprisingly, ‘high-hanging fruits’ are compatible with diminishing marginal returns. To see why, we can first observe that many realistic cost/payoff functions have a staircase structure (see figure 2). This staircase structure exists whenever payoffs come in discrete units.[3]
Figure 2: Example Staircase Function
We can see that the staircase function in figure 2 has local increasing marginal returns, but global diminishing marginal returns. When the amount of resource needed for an additional unit of payoff (or a) is large enough to require resources from multiple agents, the payoff constitutes a ‘high-hanging fruit’.
Lastly, ‘high-hanging fruits’ are likely to be neglected in the EA community because of our assumption of decreasing marginal returns and ‘pick the low-hanging fruit’ approach. One of the most common criticisms of EA – that it ignores systematic change – maybe understood in terms of high-hanging fruits. The argument can be understood as saying that EA’s focus on low-hanging fruits leads it to ignore possible high-hanging fruits which require radically high costs.
4. Potential Actions
To act optimally when there are high-hanging fruits, the EA community needs to do two things: 1) identify potential ‘high-hanging fruits’, and 2) coordinate our joint response.
Possible suggestions to achieve these goals include:
Dedicated space on EA Forum to share potential ‘high-hanging fruits’ or other coordination opportunities.
Sessions in EA Global or other events dedicated to sharing information about potential ‘high-hanging fruits’ and coordinating our joint strategy.
More integrated finance between EA organizations to allow for more flexibility in resource allocation.
[1] Past evidence suggests 3.5% of the population as a threshold for significant political change.
[2] For relevant discussion, see: Broi, A. (2019). Effective Altruism and Systemic Change. Utilitas, 31(3), 262-276.
[3] For example, consider the distribution of bed nets, or setting up new EA organizations. In each case, a critical threshold of resources is needed to obtain each additional unit of payoff (i.e., each additional bed net or each additional functional EA organization)
‘High-hanging Fruits’ and Coordination
This post introduces the issue of ‘high-hanging fruits’ and identifies possible avenues of action. The epistemic status of this post is extremely rough. I would in particular welcome suggestions for more concrete scenarios than what I suggest below.
1. Possible Scenarios
Consider two possible scenarios:
Due to shifts in the political climate, we are reasonably confident that animal advocacy will be unusually tractable in the next four years. Experts believe that the most effective intervention method is mass street-level political advocacy. However, there is one catch: the intervention will be effective only if more than x number of people are mobilised.[1] Otherwise, the intervention will have little beneficial effects. The EA community can help ensure that more than x number of people are mobilised within four years, but only if it diverts significant amounts of resources away from other cause areas during that period.
Compelling evidence suggests that extremely dangerous climate feedback loops will occur when global warming reaches 2°C. However, scientists believe that some extremely effective intervention may become available in three years. Once the intervention becomes mature, y amount of funds needs to be invested extremely quickly for us to realistically avoid hitting 2°C. We can expect funding from non-EA sources to be insufficient. The EA community can ensure enough funding for the intervention only if it starts saving funds now and subsequently spend virtually all saved financial resources on the intervention once it becomes available.
2. What are ‘High-hanging Fruits’?
Both scenarios described above are instances of what we might call ‘high-hanging fruits’. ‘High-hanging fruits’ are special cases of increasing marginal returns.[2] In both above scenarios, a critical threshold of resources is required in order to achieve a payoff. In other ‘high-hanging fruit’ cases, there may not be a discrete threshold for a payoff. Rather, there might be continuous increasing marginal returns until a certain point. See figure 1 for representations of cost/payoff functions for ‘high-hanging fruits’.
Figure 1: Cost/Payoff Functions in High-Hanging Fruit Scenarios
If reaching a ‘high-hanging fruits’ requires actions by multiple parties, then it presents a coordination problem, where reaching the optimal outcome requires multiple agents acting on a joint strategy.
3. Importance of ‘High-hanging Fruits’
There are at least three reasons for paying attention to ‘high-hanging fruits’:
High-hanging fruits are, by definition, high value.
High-hanging fruits may be prevalent.
High-hanging fruits are likely to be neglected.
Since the first is mostly self-explanatory, we will only discuss the other two reasons.
Why might high-hanging fruits be prevalent? Does this not contradict the standard assumption of diminishing marginal returns? Surprisingly, ‘high-hanging fruits’ are compatible with diminishing marginal returns. To see why, we can first observe that many realistic cost/payoff functions have a staircase structure (see figure 2). This staircase structure exists whenever payoffs come in discrete units.[3]
Figure 2: Example Staircase Function
We can see that the staircase function in figure 2 has local increasing marginal returns, but global diminishing marginal returns. When the amount of resource needed for an additional unit of payoff (or a) is large enough to require resources from multiple agents, the payoff constitutes a ‘high-hanging fruit’.
Lastly, ‘high-hanging fruits’ are likely to be neglected in the EA community because of our assumption of decreasing marginal returns and ‘pick the low-hanging fruit’ approach. One of the most common criticisms of EA – that it ignores systematic change – maybe understood in terms of high-hanging fruits. The argument can be understood as saying that EA’s focus on low-hanging fruits leads it to ignore possible high-hanging fruits which require radically high costs.
4. Potential Actions
To act optimally when there are high-hanging fruits, the EA community needs to do two things: 1) identify potential ‘high-hanging fruits’, and 2) coordinate our joint response.
Possible suggestions to achieve these goals include:
Funding directed toward identifying ‘high-hanging fruits’
Dedicated space on EA Forum to share potential ‘high-hanging fruits’ or other coordination opportunities.
Sessions in EA Global or other events dedicated to sharing information about potential ‘high-hanging fruits’ and coordinating our joint strategy.
More integrated finance between EA organizations to allow for more flexibility in resource allocation.
[1] Past evidence suggests 3.5% of the population as a threshold for significant political change.
[2] For relevant discussion, see: Broi, A. (2019). Effective Altruism and Systemic Change. Utilitas, 31(3), 262-276.
[3] For example, consider the distribution of bed nets, or setting up new EA organizations. In each case, a critical threshold of resources is needed to obtain each additional unit of payoff (i.e., each additional bed net or each additional functional EA organization)