Which policies buy a country the most happiness for a given budget?
Which policies buy a country the most happiness for a given budget?
with Alessandro Girardi, Julien Wolfensberger and Hongshu Yan
A policymaker has a budget and a menu of policies. Each one costs something and moves a single metric: two hundred thousand to cut unemployment by five percent over five years, and so on. The question nobody can answer from a spreadsheet is which combination of those policies leaves the population happiest.
We built a visual tool that answers it. Models trained on Gapminder and the World Happiness Report estimate how each metric feeds into a country’s happiness score, an optimizer searches for the best affordable combination, and the dashboard shows the consequences: a choropleth map for exploring countries, parallel-coordinate clustering, a predicted happiness curve with confidence bands, and bar charts of which metrics moved.
A React and Nivo frontend over a Python API, both dockerized, built as a team of four for the Interactive Machine Learning course at ETH.