The question

A recurring argument in housing policy is whether building more market-rate housing actually brings rents down, or whether new supply just gets absorbed by demand. This is an empirical question about supply elasticity — how much the housing stock responds to price.

This is a sample project. Replace it with your own, but keep the shape: question → data → method → result → caveats.

An interactive chart

You can drop live, hoverable charts straight into a post. Here's permits vs. rent growth for a few illustrative metros (made-up data) — hover the points:

To make one, just write a fenced code block tagged chart containing a Chart.js config as JSON. Colors are added automatically.

A first model

A simple fixed-effects specification:

rent_growth_it = β · permits_per_capita_it + γ · X_it + α_i + δ_t + ε_it

where α_i are metro fixed effects and δ_t are time fixed effects. The coefficient of interest is β.

Identification is the hard part. Permits aren't random: cities permit because demand is high. A credible version needs an instrument (Saiz 2010 geography constraints) or a natural experiment.

Caveats

  • Permits ≠ completions; there's a 12–24 month lag.
  • This is correlational without a real identification strategy.

Next step: pull the data and actually run it.