Housing Policy Lab

Each lever applies a calibrated reduction (in points on the 0–200 Housing Stress scale) to the forecast trajectory, ramping in over 18 months to reflect implementation lag.

Baseline trajectories use state-specific trend growth rates drawn from the 5-year HPI change, wage growth, and rent-burden trend lines in ACS / HUD data. The status quo forecast is:

stress(t) = baseline × (1 + trend)^(t/12)

The confidence band shown on the chart is a ±5% envelope around the status-quo line, widening slightly with horizon length to reflect accumulated uncertainty. This is illustrative, not a formal prediction interval.

Policy effects ramp from 0 to full strength over 18 months via a linear ramp factor — consistent with typical legislative and implementation timelines for housing policy.

Trend dampening. Structural policies (rent caps, zoning reform) don't only subtract stress points; they also bend the counterfactual rent-growth curve downward. The with-policy forecast uses effectiveTrend = trend × (1 − dampen), where the dampen factor rises with the gap between a rent cap and market rent growth and with the scale of supply expansion (capped at 0.70 to leave room for exogenous drivers like wages, migration, and construction costs).

• Mostly additive model. Levers are summed as independent point contributions, with one non-linear adjustment: rent caps and zoning reform dampen the counterfactual rent-growth trend, so their benefit compounds over a multi-year horizon. Cross-lever synergies beyond that (e.g. zoning × LIHTC) are not modeled. Aggressive rent control without supply reform can worsen long-run affordability (Diamond et al., 2019) — the supply penalty in the rent-cap formula reflects this.
• Elasticities are national averages. State-level effects vary with market tightness, labor composition, and existing housing stock. Tight coastal markets respond differently from Rust Belt metros.
• No endogenous migration. The model does not capture inter-state migration responses to policy changes.
• Confidence bands are illustrative. They represent heuristic uncertainty, not a formal statistical prediction interval.
• Not a substitute for local impact analysis. For policy design decisions, commission a state-specific study from an academic institution or state housing agency.

This Lab was developed in partnership with University of Virginia SEED Consulting students as part of FinMango's ongoing collaboration on the affordability crisis. Student contributions included literature review on housing elasticities, calibration of the base model, stakeholder outreach, and user-testing with finance and policy practitioners.

The students' spring 2026 presentations on housing affordability — featuring professional discussants from the investment and consulting communities — anchored the lab's design priorities.