Arrow's Impossibility Theorem

Data Skeptic

Episode | Podcast

Date: Fri, 04 Dec 2020 16:39:22 +0000

<p>Above all, everyone wants voting to be fair. What does fair mean and how can we measure it? Kenneth Arrow posited a simple set of conditions that one would certainly desire in a voting system. For example, unanimity - if everyone picks candidate A, then A should win!</p> <p>Yet surprisingly, under a few basic assumptions, this theorem demonstrates that no voting system exists which can satisfy all the criteria.</p> <p>This episode is a discussion about the structure of the proof and some of its implications.</p> <p><strong>Works Mentioned</strong></p> <div class="p-rich_text_section"><a href="https://www.stat.uchicago.edu/~lekheng/meetings/mathofranking/ref/arrow.pdf"> A Difficulty in the Concept of Social Welfare by Kenneth J. Arrow</a></div> <div class="p-rich_text_section"> </div> <div class="p-rich_text_section"><a href="http://dido.econ.yale.edu/~gean/art/p1116.pdf">Three Brief Proofs of Arrows Impossibility Theorem by John Geanakoplos</a></div> <div class="p-rich_text_section"> </div> <div class="p-rich_text_section"><strong>Thank you to our sponsors!</strong></div> <div class="p-rich_text_section"> </div> <div class="p-rich_text_section"><strong>Better Help</strong> is much more affordable than traditional offline counseling, and financial aid is available! Get started in less than 24 hours. Data Skeptic listeners get 10% off your first month when you visit: <a href="https://betterhelp.com/dataskeptic">betterhelp.com/dataskeptic</a></div> <div class="p-rich_text_section"> </div> <div class="p-rich_text_section">Let <strong>Springboard School of Data</strong> jumpstart your data career! With 100% online and remote schooling, supported by a vast network of professional mentors with a tuition-back guarantee, you can't go wrong. Up to twenty $500 scholarships will be awarded to Data Skeptic listeners. Check them out at <a href="https://springboard.com/dataskeptic">springboard.com/dataskeptic</a> and enroll using code: DATASK</div>