16

Can we apply start-up investing principles to non-profits?

How do you find the best non-profits to donate to? This is an important question that is critical to effective altruism. One suggestion comes from Holden Karnofsky at the Open Philanthropy Project, who describes a strategy called “hits-based giving” . In this framework, you make a number of investments, some... Read More
Comment author: Peter_Hurford  (EA Profile) 09 June 2017 06:48:40PM 2 points [-]

It would be nice to see a fund dedicated toward research, especially empirical research, to gather information relevant to EA objectives.

Comment author: MichaelDickens  (EA Profile) 03 June 2017 06:39:32AM 1 point [-]

RE #1, organizations doing cause prioritization and not EA community building: Copenhagen Consensus Center, Foundational Research Institute, Animal Charity Evaluators, arguably Global Priorities Project, Open Philanthropy Project (which would obviously not be a good place to donate, but still fits the criterion).

RE #2: if the point is to do what Nick wants, it should really be a "Nick Beckstead fund", not an EA Community fund.

Comment author: Peter_Hurford  (EA Profile) 09 June 2017 06:25:06PM 0 points [-]

There are also independent EA researchers doing cause prioritization research without community building.

Comment author: Kerry_Vaughan 07 June 2017 04:00:28PM 0 points [-]

RE #2: if the point is to do what Nick wants, it should really be a "Nick Beckstead fund", not an EA Community fund.

The fund is whatever he thinks is best in EA Community building. If he wanted to fund other things the EA Community fund would not be a good option.

Comment author: Peter_Hurford  (EA Profile) 09 June 2017 06:24:25PM 2 points [-]

But how is funding cause prioritization related to EA community building?

Comment author: Peter_Hurford  (EA Profile) 02 June 2017 07:06:09PM 5 points [-]

We’re committed to evaluating our own efforts and changing directions or even disbanding the organization if we determine that we can make a greater impact elsewhere. If our research fails to generate new, actionable insights that make a significant difference to advocates’ decisions, we plan to shift our priorities towards outreach and movement-building, such as the career content described above.

Your philosophy is very admirable. How do you plan to track this, more concretely?

Comment author: arunbharatula 31 May 2017 03:25:11AM -1 points [-]

Thanks for elaborating on the history of the INT framework Peter :)

Which problems are you referring to?

I assume you're asking about equity since you noted the example on positioning in my original post. In PAHO's executive committee document they define equity as:

'differential occurrence of disease or access to health programs according to gender, ethnicity, income level, geographic location (urban vs. rural), etc'.

In their scoring system, the more differential the occurrence of a disease (e.g. if a disease overwhelming affects women, or a particular ethnic group, relative to the general population) that more importance that problem is, independent of other prioritization factors.

This is a good point, though I disagree. I think there is an important separation between a framework for evaluating causes and a framework for making sure your chosen interventions and marketing methods don't alienate people. I'm skeptical combining the two frameworks would make for a better approach than considering the two separately.

If we agree that A, B and C variables should be used to evaluate a cause, and that D and E should evaluate the acceptability of that cause, then combining the two frameworks would look like:

A * B * C * D * E

Considering the two seperately would look like:

(A * B * C) (D * E)

Those are mathematically identical.

Do you mean we should follow along with donor interests and avoid focusing on self-inflicted ill-health in order to maybe get more donations? Or do you mean we should go against donor interests because it means the causes are more likely to be neglected?

No, I just raised the topic of popular perceptions of self-inflicted ill-health to illustrate the relevance of positioning. I expect the relative value of leveraging donor interest v.s. neglectedness would vary on a case-by-case basis.

Comment author: Peter_Hurford  (EA Profile) 02 June 2017 06:59:54PM 0 points [-]

If we agree that A, B and C variables should be used to evaluate a cause, and that D and E should evaluate the acceptability of that cause, then combining the two frameworks would look like:

A * B * C * D * E

Considering the two seperately would look like:

(A * B * C) (D * E)

Those are mathematically identical.

They may be mathematically identical, but they would be implemented differently. Our goal is to get to an optimal intervention within an optimal cause. Ideally we would search the entire cause * intervention space, but this is intractable, so the current strategy is to search the cause space for the best cause and then find the best intervention within that best cause. This does produce errors and bias, though, as it could be the best cause has no good interventions.

I recommend searching the cause space by INT and the intervention space by a wider variety of weighted metrics.

Comment author: arunbharatula 31 May 2017 03:25:11AM -1 points [-]

Thanks for elaborating on the history of the INT framework Peter :)

Which problems are you referring to?

I assume you're asking about equity since you noted the example on positioning in my original post. In PAHO's executive committee document they define equity as:

'differential occurrence of disease or access to health programs according to gender, ethnicity, income level, geographic location (urban vs. rural), etc'.

In their scoring system, the more differential the occurrence of a disease (e.g. if a disease overwhelming affects women, or a particular ethnic group, relative to the general population) that more importance that problem is, independent of other prioritization factors.

This is a good point, though I disagree. I think there is an important separation between a framework for evaluating causes and a framework for making sure your chosen interventions and marketing methods don't alienate people. I'm skeptical combining the two frameworks would make for a better approach than considering the two separately.

If we agree that A, B and C variables should be used to evaluate a cause, and that D and E should evaluate the acceptability of that cause, then combining the two frameworks would look like:

A * B * C * D * E

Considering the two seperately would look like:

(A * B * C) (D * E)

Those are mathematically identical.

Do you mean we should follow along with donor interests and avoid focusing on self-inflicted ill-health in order to maybe get more donations? Or do you mean we should go against donor interests because it means the causes are more likely to be neglected?

No, I just raised the topic of popular perceptions of self-inflicted ill-health to illustrate the relevance of positioning. I expect the relative value of leveraging donor interest v.s. neglectedness would vary on a case-by-case basis.

Comment author: Peter_Hurford  (EA Profile) 02 June 2017 06:57:23PM 0 points [-]

In their scoring system, the more differential the occurrence of a disease (e.g. if a disease overwhelming affects women, or a particular ethnic group, relative to the general population) that more importance that problem is, independent of other prioritization factors.

This is definitely a values judgment call. I personally don't value giving unequal interest to anyone on the basis of gender or race.

Comment author: arunbharatula 31 May 2017 04:30:39AM *  -1 points [-]

My mistake with the economist claim. It seems some senior GiveWell employees are economists.

However, I don't see how those exerpts defend GiveWell's assessment of deworming.

Comment author: Peter_Hurford  (EA Profile) 31 May 2017 06:11:29AM 0 points [-]

I don't see how those exerpts defend GiveWell's assessment of deworming.

What did you think of http://blog.givewell.org/2016/12/06/why-i-mostly-believe-in-worms/ ?

Comment author: Larks 30 May 2017 11:14:03PM 1 point [-]

GiveDirectly now gives on the village and not the individual level.

Do you have a source for that? The linked webpage suggests otherwise:

In Kenya, we incorporate a range of factors into our eligibility criteria, including housing (e.g. house size), assets (e.g. presence of a latrine), vulnerable recipient status (e.g. homelessness), and other criteria.

In Uganda, we enroll families in homes that have thatched roofs and those who are homeless. People who live in homes with thatched roofs have been shown to be substantially poorer than their neighbors.

In response to comment by Larks on Red teaming GiveWell
Comment author: Peter_Hurford  (EA Profile) 31 May 2017 03:56:19AM 0 points [-]

I think I misread the situation.

My key quote is from here:

GiveDirectly notes that even though it has not identified significant evidence of negative effects on non-recipients, it now generally avoids conducting experiments that randomize at the individual level, to avoid situations in which one eligible household receives transfers while a similarly situated neighbor does not.

That's just about randomization within experiments, not about distributing money.

Regardless, my general point still stands that GiveDirectly and GiveWell are very aware of this issue, are monitoring it, and are prepared to change if it proves to be a larger problem. It looks like it isn't a sizable problem.

Comment author: Telofy  (EA Profile) 30 May 2017 04:11:32PM 1 point [-]

I’d still say that the original point stands. Developed countries had no problem taking care of their own malaria problems. They still had a few more chemical that don’t work anymore today, but if we still had malaria in Germany today, the government would surely find a way to eliminate it. The trajectory of developing countries indicates that many will cease to be developing countries within half a century or so, and then the same logic will apply to them.

Cost-effectiveness in this cause area means getting the money in early to speed up positive developments that would likely happen anyway, just a few decades later. That saves a few decades of suffering, which is valuable, but it probably doesn’t compare to trajectory-changing interventions. Rather than increase the speed of a development that is already going into the right direction, trajectory-changing interventions can potentially affect very long periods of the future.

That’s probably the comparison Arunbharatula is drawing where international development doesn’t look as strong as some x-risk or values spreading interventions.

In response to comment by Telofy  (EA Profile) on Red teaming GiveWell
Comment author: Peter_Hurford  (EA Profile) 30 May 2017 04:29:06PM 2 points [-]

That’s probably the comparison Arunbharatula is drawing where international development doesn’t look as strong as some x-risk or values spreading interventions.

I think that's a generous read of Bharatula's writing, especially since the Gates Foundation also spends the majority of their money on things that speed up positive developments that would likely happen anyway.

Regardless, it's an important steelman. Maybe it would be valuable to focus on funding information about what works so that governments know what to roll out when the time comes? Or find more ways to help governments listen to cost-effectiveness and other evidence? Or just fund MIRI? I find this somewhat persuasive, but we would need to actually build the case.

View more: Prev | Next