Terner Labs has built a Housing Policy Simulator that models how local land values, construction costs, zoning rules, and financing conditions shape development outcomes. The tool estimates the likelihood that a multifamily market-rate project gets built on a given site, then shows how that likelihood changes under different policy scenarios. The accompanying visualization starts with Denver, San Diego, and Tucson.
The useful part is not that policy matters. It is that the same policy can land very differently by city, because the binding constraint is not always the same. That is the right lesson for policymakers. If you want more homes, you need a clearer read on what gets built in a specific market, not a generic answer that assumes every city behaves the same way.
The research belongs to Terner Center at UC Berkeley; the read here is mine.