📣 making better decisions together

slides
- https://docs.google.com/presentation/d/1i_1YS1iBjQ44ATMjTyvhVxEvU7FC1-aq2qynlPxSzwY/edit?slide=id.p#slide=id.p
collective decision making
- how to convert collective will to singular will
- ex.
  - in the small (~ where do we go for lunch?)
  - in the medium (~ how should we run our corporation?)
  - in the big (~ how should we run the nation?)
- in academia, these questions fall under...
  - social choice theory, mechanism design (economics)

ex. where should we go for lunch?
- (conceptually, each individual has "true" hidden preferences)
  - (ex. A: 🌮🍕🍣 B: 🍕🌮🍣 C: 🍣🍕🌮)
- poll: single answer, most popular wins ("plurality" voting)
  - NOOOOOO!
- hopefully, can intuit that this sucks
- "what" is wrong?
  - there is a solution that makes the total happier
    - (is that what we want?)
      - ...morality
- "why" is it "bad"?
  - broadly: it is lossy, GIGO
  - (conceptually, each individual has "true" hidden preferences)
  - a good "group decision making system" needs to capture information (people communicating what they want, how badly)
will talk about...
- a few different ways to do voting
- but... before that, I need to address an important caveat
  - voting is usually part of a larger "group decision making –system–"
    - often can't improve things in isolation
    - think holistically
- so, first talk about "systems" and then "mechanisms"

systems


members  ------------------>  decision             (direct democracy)
               vote


members  -----> rep ------->  decision             ((single) representative democracy)
          vote        fiat


members  -----> reps ------>  decision             (representative democracy)
          vote        vote


members ----> reps  -\__~~~~> decision             (real world political systems)
        \---> rep   -/

simplest system: direct democracy
- members use some voting process to make decisions
- see this...
  - small group decisions (our lunch example)
  - in the large, most commonly: referendums
- with internet, could replace congress with a system of continuous digital referendums
  - why might this be "bad"?
    - does the voter have time to be informed?
      - or are they just directly sharing preferences
      - (if you force direct democracy on a group that is unwilling/unable of getting informed, they are prone to external manipulation)
        (politicians often call for referendums on topics to make them fail)
...and so, we often have representatives
one representative to call the shots
- but really, for large groups, it's too much work for one person
- ...so a group of representatives
  - and now we have voting to choose the reps... and the reps voting...
in the real world... much more complicated
- constitutions
- system of checks and balances
- distribution of powers

there is a lot of "design freedom" in these systems
- multi-layer plurality
  - we've got this flawed thing... let's do it twice
  - forced into strategic voting
    - (can't vote for my preferred choice, because they have no chance)
  - (IMO the most disenfranchising system one could design)
- alternatives?
  - could replace the mechanism for voting (but leave other parts the same)
    - sometimes better, sometimes worse; it depends on how it interacts
  - but, can do bigger "re-architectures":
    - ranked votings... w/ two candidate ridings
    - proportional
      - popular vote determines the party proportion in parliament
      - but other mechanism determines the specific politicians
      - (over 85 democratic countries transitioned from the plurality system to proportional, and now overwhelming majority use it)
    - sortition (choose citizens randomly), but, into more smaller committees
      - jury duty
      - ancient greece (multiple committees)
      - citizen's assemblies (ex. for drawing district boundaries)

mechanisms

from the big picture, let's get back to the little picture: voting mechanisms
as we've saw, can be used in...
- choosing directly (particularly in small groups)
- choosing representatives
- our representatives making choices

voting
- plurality voting (first past the post)
  - ballot: mark a single choice
  - integration: highest count wins
  - comments:
    - (strategic voting, often leads to poor outcomes)
    - (we've talked enough about it)
- ranked voting
  - ballot: rank choices
  - integration: many ways, with slightly different outcomes
    - ex.
      - borda
      - run-off
      - schulze
  - (DECIDEDLY PLUG)
  - comments:
    - arrow's theorem: none of these have all the intuitive properties we think are fair
      - A B and C -> A wins
      - if clone A, B now wins :/
      - if remove C, B now wins :/
      - (10 or so others)
- score/approval voting
  - ballot: rate a choice on some scale (from 0 to 10; or could just be 0 or 1)
  - integration: sum it up
  - comments:
    - works well because it captures all the inherent preferences
    - (now my preferred voting mechanism)
    - ...in theory, can still vote strategically
alternatives to voting mechanisms:
markets
- in the large, markets are also a way "for groups to integrate individual preferences over how to manage shared resources"
- "free" markets
  - directly involve individual preferences, and their magnitudes
- auctions
  - many kinds
  - studied, designed to incentive participants to participate truthfully
    - ie. the best strategy is to just reveal their true preference
  - (ROOMMATES CHORE AUCTION)
- quadratic voting
  - score voting, but, with quadratic cost increase (1 = 1, 2 = 4, ...)
  - more accurately matches the micro-economics
  - author suggests idea of referendums where you pay normal money

(QUADVOTE PLUG)
consensus
- point is deliberation
  - deliberation as the way of developing common understanding and discovering preferences
- (gets a bad rap, lack of knowledge how to do well, so facilitated poorly)
- examples
  - canadian native communities
  - sociocracy
  - functioning relationships
sortition
- choose randomly from a population of eligible and willing candidates
- imo, under-explored

conclusion

we talked about systems
- direct decision making
- or, involving representatives
- with lots of potential checks and balances
we talked about mechanisms
- single choice, ranked choice, scoring and quadratic scoring
- ...and alternatives to voting
  - markets
  - consensus
  - sortitition
when something is popular, easy to fall into the trap of thinking it's optimal
- but often it's just an artefact of history that is entrenched (stuck in a local maximum)
"it's the way things have been done"
"easy / simple"
...are excuses for systematic disenfranchisement
there are many ways to make decisions as a group
- which are appropriate depends on the context
  - (except plurality voting - it just sucks)

SCRAPS

even dictators need to incentive others (machiavellian analysis)
are some "systems" "better" than others?
design for good (and design for evil)
voting systems motivate certain behaviour in their candidates
democracy in the large
- countries claiming to be democracies: https://en.wikipedia.org/wiki/Democracy#/media/File:Democracy_claims.svg
- quality of their democracy: https://en.wikipedia.org/wiki/TheEconomistDemocracyIndex#/media/File:EconomistIntelligenceUnitDemocracyIndex2024.svg
meta-politics
- can design systems that oppress
- how limit the tyranny of the majority
in governments, delegate the decision to "decide what is good/right"

2026-03-11