What the workshop is & a brief overview

Hi all! This is my first post on LessWrong, but I'll be posting here more often with conversation notes & resources from the Effective Altruism Kansas City meetup group.

This workshop was an experiment to give participants an intellectual & intuitive understanding of how EAs typically prioritize between charities - the QALY (and DALY). Participants were told that they're the board for the Hypothetical Foundation and needed to choose a charity to fund between three options to best serve the residents of Hypothetical Town. Discussion is guided from initial impressions, to how to quantify well-being, to how to estimate DALYs, and finally, how to use DALYs to compare charities.

The workshop was fun & engaging, but I've got a few post-mortem revision notes in case anyone wants to use these materials:

• Have participants draw their own QALY boxes
• Have participants intuit QALY shapes of existing charity models
• Walk through an example calculation before having participants do it
• Give a worksheet that lays out steps nicely and walks participants through an example calculation

Workshop materials

Rough schedule & content/activity outline

• 18:15 TALK: Introduce the foundation board meeting premise and the fake charities we'll be evaluating.
• All charity operations in Hypothetown have room to scale according to the possible additional funding (to simplify the decision).
• 18:20 BREAKOUT: Brief discussion of which charities they'd pick based on information they have
• 18:25 DISCUSS: On what basis would they pick charities? Write ideas on the board.
• Lead them to the essential question: "How many are helped, and by how much?"
• 18:30 TALK: Chief Philosophy Officer [[Josh Rainwater]] presents ways to think about how people are helped and by how much
• Will Josh mention the various ways to calculate utility? Should I ask him to?
• 18:40 BREAKOUT: Have them fill out quality of life forms individually then discuss responses with their partner(s)
• Printout link: https://docs.google.com/document/d/18KpMd2qqg5JbVWatcvfxFTTFWdlKHvrEaO0LPM4NasQ/edit?usp=sharing
• 18:50 DISCUSS: Hand out charity infosheets. Discuss how to incorporate improved life & lengthened life into a single number. Aim for a QALY-like conclusion.
• "Ok, board, we've done some due diligence on these charities - I've prepared decision briefs for each of you." Pass out brief infosheets on the two hypothetical charities. These sheets say how many people are afflicted by the problem and how much it costs to help each person.
• Seize on & develop the idea of a QALY (and really name the term for the first time)
• 18:55 DISCUSS: Draw a QALY rectangle on the whiteboard as an example, then have participants tell me what rectangles to draw for the remaining charities based on their infosheets.
• 19:00 BREAKOUT: Have groups calculate QALY cost estimates.
• "Ok, now we know that charity A improves 5 QALYs per person and charity B does 10. But how much does it cost?"
• Breakout on the cost analysis - make the number require super light calculation
• End the breakout with a review of the numbers - first dollars per person helped, then dollars per QALY or something like it
• 19:10 DISCUSS: Return from breakout and share calculated numbers for each hypothetical charity. Add relevant stats & considerations to the board.
• 19:15 DISCUSS: Hold a board vote on who to fund!
• 19:20 TALK: A brief explanation of DALYs, the differences from QALYs, and when to use each
• 19:25 DISCUSS: Open Q&A
• 19:35 BREAKOUT: Interest surveys & brainstorming sessions!
• 20:00 DISCUSS: Brief return to the group to share insights & potential actions.

Hypothetical charity details

Charity A: Prosthetic arms for people who lost arms in freak woodcutting accidents (lend-a-hand)

• Printout link: https://www.notion.so/Charity-A-Lend-A-Hand-Prosthetic-arms-30694d914bc04c9795f1698bf3b22a57
• Metrics:
• People helped annually: 1,000
• Annual budget: High price of some kind ($2,000,000?)
• Years of life extended per person: 0
• Quality of life improvement: To be estimated by participants
• Arms lost at age 40 on average

Charity B: Sanitation training for citizens (wash-ya-hand)

• Printout link: https://www.notion.so/Charity-B-Wash-Ya-Hand-Sanitation-training-c1a7389bd8e94f11bd629762c7e92f45
• Metrics:
• People helped annually: 100,000
• Annual budget: Low price of some kind ($2,000,000?)
• Average years of life extended per person: 0.15
• Quality of life improvement: To be estimated by participants

Charity C: Free treatment for "Spontaneous Combustion Syndrome", SCS (hand-on-fire)

• Printout link: https://www.notion.so/Charity-C-Hands-On-Fire-Vaccinations-against-SCS-4493cb390b484c6eae36fe493c4997c6
• Metrics:
• People helped annually: 1,000
• Annual budget: Middling price of some kind ($2,000,000?)
• Average years of life extended per person: 15
• Quality of life improvement: 0

Pedagogy notes

• See here for full raw notes & ideas on the workshop: https://roamresearch.com/#/app/sams/page/_jYqDkG2p
• Desired takeaways from the workshop:
• (briefly) why impact evaluations matter
• What a QALY is
• How to estimate QALYs given basic information
• What a DALY is
• Start by presenting the problem and the questions that we want to be able to answer.
• Share that Dan & myself will be around during the breakouts to answer questions/provide help, and just wave us over if they need something
• Apply a clip-on tie to my T-Shirt when we start the board meeting.
• Differences between the charities
• One helps few people a lot, one helps a lot of people a little
• Two are both pretty severe to where which is worse is an open question
• One helps only quality of life while another extends only quantity of life
• Charity requirements
• Need a minimum of three charities to meet the above requirements
• Charity A: Helps a few people a lot
• Charity B: Helps a lot of people a little
• Charity C: Extends life without improving quality. Severity gut reaction similar to A