I presented at the second annual workshop on Mechanism Design for Social Good, held in conjunction with EC2018 at Cornell University. My talk was entitled “Mechanism Design and Marginal Distributions.”
Most social outcomes result from the combination of two factors. Marginal distributions are the set of possible outcomes that are available, while allocation processes determine which people get which outcomes.
Most social scientists and most policy discussions focus their attention on allocation processes, with the goal of making them more fair. But as I described in my talk, many of our pressing social problems cannot be solved by better allocating our existing sets of outcomes. Instead, we need to focus our energy and attention on creating the better sets of outcomes that we want.
I gave examples from my work on upward mobility and racial equality, but similar dynamics exist in issues from health care to education to housing. They often result from poor social decision-making: it’s hard to aggregate from individual preferences to social choices, and the ways we do it right now in the United States are inadequate and vulnerable to pressure from wealthy donors and organized interests. Better mechanism design (ranked choice voting, for instance!) can help!
Here are the slides from my talk (you can download them here):