Methodology — CA Housing Permitting Friction

Data sources

HCD Annual Progress Reports (APR)

Every California jurisdiction files an Annual Progress Report with the Department of Housing and Community Development. Each report contains project-level records with application dates, entitlement approvals, building permit issuance, and certificate of occupancy dates, along with unit counts by income level and project characteristics.

Coverage: All 539 CA jurisdictions (cities + counties), annually from 2018 onward.

RHNA Progress Data (6th Cycle)

Regional Housing Needs Allocation targets from the 6th RHNA cycle (2023–2031), paired with permitting progress by income category. See the RHNA section below for a full explanation.

U.S. Census Bureau Population Data

ACS 2022 5-year population estimates for all California cities and counties, used to group jurisdictions into population tiers (large 300K+, medium 50K–300K, small under 50K) for apples-to-apples comparison.

Site Check Streamlining Eligibility

Parcel-level signals from the Governor's Office of Planning and Research Site Check tool indicating SB 35 streamlining eligibility, CEQA exemption status, and by-right development potential. Used to identify projects that could have been streamlined but were processed through discretionary review.

Permitting pipeline model

Each housing project passes through a series of stages. We measure the duration between each transition and track drop-off at each stage.

1

Application

Stage A

2

Entitlement

Stage B

3

Permit

Stage C

4

Completion

Stage durations

Application → Entitlement (Stage A): Time from submission to discretionary approval (zoning, design review, environmental clearance). Fully government-controlled.
Entitlement → Building Permit (Stage B): Time from entitlement to permit issuance (plan check, conditions of approval). Fully government-controlled.
Permit → Completion (Stage C): Time from permit to certificate of occupancy (construction period, final inspection). Mostly builder-controlled; includes some government inspection time.

Approval Speed metric

Approval Speed = Stage A + Stage B + 0.25 × Stage C

= (app → entitlement days) + (entitlement → permit days) + 0.25 × (permit → completion days)

The Approval Speed is the primary ranking metric for this system. It measures how fast a jurisdiction processes housing applications through the permitting pipeline, with appropriate weighting for what the jurisdiction actually controls.

Why 100% weight on stages A and B? These are entirely government-controlled processes — discretionary review, plan check, conditions of approval. A slow jurisdiction can directly speed these up.

Why 25% weight on stage C? Construction time (permit to completion) is mostly determined by the builder, project complexity, and market conditions. However, jurisdictions do control inspection scheduling, change-order processing, and final sign-off — roughly a quarter of the variability. The 0.25 weight gives credit to jurisdictions that facilitate fast completions without penalizing them for builder timelines.

Computed per project, aggregated per jurisdiction. Each project gets its own approval speed score. The jurisdiction score is the median across all its projects, which is robust to outliers.

Conversion funnel

The conversion funnel tracks what share of applications reach each milestone and how quickly. For each jurisdiction, we compute:

Within 1 month: Applications that receive a permit in ≤ 30 days (shown in green)
Within 3 months: Applications permitted in ≤ 90 days (yellow)
Within 6 months: Applications permitted in ≤ 180 days (orange)
Within 24 months: Applications permitted in ≤ 730 days (red)

These are cumulative — the 3-month rate includes everything in the 1-month rate. A jurisdiction where 80% of apps get a permit within 3 months is operating much faster than one where only 30% do.

Projected permit rate

Recent applications may not have had time to reach the permit stage yet. To estimate their eventual outcome, we look at historical cohorts of applications that are now 24+ months old and measure what share ultimately received a permit. This "projected permit rate" gives a best estimate of eventual outcomes for the current pipeline.

Scoring methodology

1. Population-based peer groups

Jurisdictions are grouped by Census population into three tiers: large (300,000+), medium (50,000–300,000), and small (under 50,000). This ensures cities are compared against jurisdictions of similar scale rather than the statewide average.

2. Peer frontier

Within each peer group, the 25th percentile of median approval speed defines the attainable frontier — the speed that better-performing peers demonstrate is achievable. This is not the best performer (which may have unique advantages) but a realistic benchmark.

3. Excess delay score

excess_delay = jurisdiction_median_approval_speed - peer_frontier
Measured in days. A positive value means the jurisdiction is slower than what its peers demonstrate is achievable.

4. Units-delayed score

units_delayed = excess_delay_days × annual_project_count / 365
Estimates the annualized project-years of housing production lost to friction. A jurisdiction processing 100 projects/year with 73 days of excess delay effectively delays 20 project-years of housing.

5. Affordable-units-delayed score

Same calculation as units-delayed, but weighted by the count of affordable units (very low, low, and moderate income). This highlights where friction disproportionately affects deed-restricted and income-targeted housing.

6. Streamlining-missed score

Fraction of SB 35-eligible projects that were processed through discretionary review instead of ministerial streamlining. A high score indicates the jurisdiction is not leveraging available streamlining pathways, adding unnecessary process to qualifying projects.

7. Jurisdiction ranking

The overview page ranks jurisdictions within each population tier by approval speed (lower = faster = better). The recoverable impact score (excess_delay × annual_units) is used to prioritize jurisdictions where reducing friction would unlock the most housing.

What is RHNA?

RHNA stands for Regional Housing Needs Allocation. It is California's mechanism for ensuring that every community plans for and permits enough housing to meet the state's growth and affordability needs.

How it works: The California Department of Housing and Community Development (HCD) determines how many housing units each region needs over an 8-year cycle. Regional planning bodies (like ABAG in the Bay Area or SCAG in Southern California) then allocate targets to individual cities and counties.

Income categories: Each jurisdiction's RHNA target is broken down into four income levels:

Very Low Income: Households earning ≤ 50% of Area Median Income
Low Income: 50–80% of AMI
Moderate Income: 80–120% of AMI
Above Moderate: > 120% of AMI (market-rate housing)

The current cycle (6th Cycle, 2023–2031) set dramatically higher targets than previous cycles, particularly for lower-income housing. Many jurisdictions are struggling to keep pace.

Consequences of falling behind: Jurisdictions that don't meet their RHNA targets face state accountability measures, including:

SB 35 ministerial streamlining requirements for qualifying housing projects
"Builder's remedy" provisions that allow developers to bypass local zoning
Loss of certain state funding eligibility
Court-ordered housing element compliance

Connection to permitting friction: This system shows RHNA progress alongside permitting speed because slow permitting directly impedes a jurisdiction's ability to meet its RHNA targets. A jurisdiction with high excess delay and low RHNA progress is likely suppressing housing production through process friction.

Backtesting

Scores are validated with a rolling backtest: the peer-frontier model is trained on data up to year t-1, used to score jurisdictions for year t, then evaluated against actual outcomes. We report Spearman rank correlation between predicted friction rankings and realized durations to measure whether the scoring identifies persistently high-friction jurisdictions.

Limitations

APR data is self-reported by jurisdictions; completeness and accuracy vary.
Not all date fields are populated — only ~41% of projects have application dates, limiting the approval speed calculation to that subset.
SB 35 eligibility is approximated from project characteristics; actual eligibility depends on site-specific conditions.
Construction time (permit → completion) reflects builder behavior as much as regulatory friction, which is why it receives only 25% weight in the approval speed metric.
Peer groups use broad population categories; jurisdictions near bracket boundaries may be compared against less-similar peers.
The model does not account for project complexity within the same type category.
Population data is from ACS 2022 estimates and may not reflect recent changes.