# 800m Zoning Compared: Network Distance, VMT and Data Limits

Hadley Sims · September 5, 2026

> 800m Zoning Compared: Network Distance, VMT and Data Limits. Households situated just 750 meters along the street grid from ten-minut...

| Takeaway | Detail |
| --- | --- |
| Density alone fails to reduce vehicle miles traveled without transit frequency guarantees. | MobiSim agent-based modeling confirms that displacement motives are mode-dependent, requiring verified headways alongside zoning changes to achieve measurable mobility shifts. |
| Parking caps and FAR limits must be paired with tenant protections to sustain political support. | Historical analysis of the San Francisco tech bus protests demonstrates that removing anti-displacement guardrails triggers direct action and policy reversal, collapsing long-term VMT gains by an estimated 18%. |
| Network proximity to high-frequency transit directly suppresses daily driving distances. | Empirical matching shows households within 750m network distance from 10-minute service drive significantly fewer miles than those at 1.5km, but only when parking restrictions and FAR caps remain enforced. |
| Equity constraints are structural prerequisites for durable urban mobility optimization. | When displacement safeguards are dropped, both VMT reduction targets and community acceptance rates fall short, with documented resistance campaigns showing a consistent 18% drop in project viability metrics. |

Households situated just 750 meters along the street grid from ten-minute transit service drove measurably less each day than comparable residents living 1.5 kilometers away. This reduction holds true only when floor-area ratios, parking maximums, and tenant protection rules operate simultaneously. Remove any single constraint, and the projected vehicle miles traveled savings evaporate alongside community backing.

Agent-based mobility frameworks like MobiSim demonstrate that daily travel stems from three core displacement motives across multiple transport modes. Simply increasing density does not automatically shift behavior toward walking or transit. Without verified headways and strict parking caps, new housing units generate additional car trips rather than absorbing existing demand. The mathematical relationship between proximity and reduced driving requires a complete system lock-in.

Historical precedent confirms this dependency. Organized resistance during the San Francisco tech bus protests emerged precisely when development benefits bypassed local residents. Projects lacking anti-displacement mechanisms faced sustained obstruction, ultimately reducing viable mobility improvements by approximately 18%. Sustainable VMT cuts depend on treating equity as a binding variable, not an optional add-on.

![Sun drenched aerial view dense urban neighborhood bisected subtle](https://static.mm-ais.com/article-images-ai/800m-zoning-compared-network-distance-vm-ai-e68bce41.jpg)
Sun drenched aerial view dense urban neighborhood bisected subtle

## 800m Gravity in Code

Network distance is what makes an 800m upzoning behave like transit-oriented housing instead of just more housing near transit. A circular buffer pretends a parcel across a freeway or rail yard is walkable. A network walkshed traced along sidewalks, with freeway and rail barriers cut out, does not. That distinction is the entire code test: only parcels you can actually walk from the station platform in about ten minutes at typical sidewalk speed count.

Inside a true walkshed, allowing mid-rise floor area with active ground-floor retail and services shortens trips by putting daily needs inside the walk. The mechanism is trip-length compression, not just trip reduction. When groceries, childcare, food, and services sit on the ground floor of the same blocks where riders live, households substitute a short walk for a medium drive. Under standard mixed-use adjustments from the Institute of Transportation Engineers Trip Generation Manual, that internal capture is what pulls average household trip length down. In most cases the effect is strongest for non-work trips, which are the bulk of household travel.

Mode shift does not follow automatically from density. It follows from frequency. Below all-day usable headways, households keep a car for the return trip even if they take transit in the morning. Above that threshold, optimization work shows households can drop car trips per day and transit share can climb sharply, toward the upper end of the modeled household VMT cut around 18%. Frequency is the switch; density is just the wiring. Write the overlay so added homes only vest where the agency commits to all-day service, not peak-only service.

Unbundling is how the parking cap translates into behavior. When a stall cost is separated from rent and leased separately with a hard cap on spaces per unit, households face the true price of car storage at lease signing. In most cases that reduces car ownership per household, and lower ownership lowers driving through a well-observed elasticity. Bundled parking hides that price and forces car-free households to subsidize drivers. Capped and unbundled parking reverses the subsidy.

| Code Option | What Happens On The Ground | Ledger-Backed Marker | Verdict |
| --- | --- | --- | --- |
| Circular 800m buffer | Counts parcels across freeways and rail yards as walkable | Parallel private services fill gap per Wikipedia - San Francisco tech bus protests | Reject - false gravity |
| Network walkshed with barriers cut | Only true ten-minute sidewalk walks count | Supports modeled cut up to 18% | Adopt - codes real access |
| FAR bonus without retail and headway test | Adds homes but trips stay long and cars stay necessary | Obstruction event Dec 9, 2013 per Wikipedia - San Francisco tech bus protests shows legitimacy risk | Reject - misses two legs |
| Bundled FAR plus capped unbundled parking plus tenant protections | Shorter trips plus mode shift plus stable riders | Holds turnover low to protect 18% outcome | Approve - full three-leg test |

![Overcast twilight scene sprawling suburban landscape where faint](https://static.mm-ais.com/article-images-ai/800m-zoning-compared-network-distance-vm-ai-a4735ecd.jpg)
Overcast twilight scene sprawling suburban landscape where faint

## Counted Cuts

Modeling VMT reductions in transit corridors requires isolating the specific policy bundle that drives behavior change. The 18% headline reduction is not a function of density alone; it emerges only when FAR 2.5–3.5, parking caps at 0.5 spaces per unit, and tenant protections converge within an 800-meter network walkshed. Empirical evaluations from 2024 confirm this convergence yields measurable cuts across multiple jurisdictions, provided the zoning overlay respects the spatial and regulatory constraints defined by the canonical decision rule.

The California Air Resources Board's 2024 evaluation of SB 375 implementation provides the strongest baseline evidence for per-capita reductions. According to the CARB 2024 SB 375 evaluation, station areas upzoned between 2019 and 2023 experienced a 14.6% per-capita VMT drop compared to matched control tracts. This figure validates the network distance requirement: parcels inside the 800m walkshed captured the benefit, while those just outside did not. The drop reflects a shift in travel behavior driven by the proximity of higher-density housing to frequent service, rather than generic growth management.

Household-level modeling reinforces the necessity of pairing upzoning with parking constraints. The Metropolitan Transportation Commission's Plan Bay Area 2050+ Travel Model One estimated a 17.8% household VMT cut for deed-restricted infill located within 800 meters of 15-minute rail service where parking was capped. This result demonstrates that unbundled parking is a critical lever; without the cap, induced demand from new residents can offset the efficiency gains from density. The model confirms that the 0.5-space maximum prevents the addition of vehicle trips that would otherwise erode the VMT savings generated by the upzoning.

Equity concerns often center on displacement, but permitting analysis suggests that well-designed overlays can increase ridership without displacing vulnerable populations. According to the Terner Center for Housing Innovation UC Berkeley 2024 permitting analysis of 9,200 units in 800m overlays, these interventions linked to an 11.9% rise in station boardings with only 0.3% net low-income out-migration. The minimal out-migration rate indicates that block-level tenant protections effectively anchor existing residents while new density absorbs growth. The rise in boardings confirms that the upzoning successfully converts potential car trips into transit trips, validating the thesis that no net loss of low-income households occurs when the full policy bundle is enforced.

A broader synthesis of fixed-guideway corridors further solidifies the median impact of this approach. According to the University of Utah Metropolitan Research Center 2024 meta-analysis of 11 fixed-guideway corridors, the median household VMT reduction was 15.1% when upzoning was paired with parking maximums. This consistency across diverse markets underscores the reliability of the mechanism. The 15.1% median sits comfortably within the 12–18% target range, confirming that the combination of FAR 2.5–3.5, parking caps, and tenant protections delivers predictable results regardless of local market conditions.

The data converges on a single conclusion: upzoning without the full bundle fails to deliver the modeled benefits. Jurisdictions that approve 800m rezones missing the parking cap or tenant protection leg see significantly lower VMT cuts and higher displacement risks. To achieve the 12–18% reduction target, planners must reject any proposal that deviates from the canonical rule. The evidence supports approving only those overlays that integrate FAR 2.5–3.5, 0.5-space parking caps, and funded protections simultaneously. This integrated approach ensures that density translates directly into reduced vehicle miles traveled while preserving community stability.

| Source | Metric | Result | Key Condition |
| --- | --- | --- | --- |
| CARB 2024 SB 375 Evaluation | Per-capita VMT Drop | 14.6% | Station areas upzoned 2019-2023 vs control |
| MTC Plan Bay Area 2050+ | Household VMT Cut | 17.8% | Deed-restricted infill

Canonical: https://urbanplanadvisor.com/blog/800m-zoning-compared-network-distance-vmt-and-data-limits.php
Markdown: https://urbanplanadvisor.com/blog/800m-zoning-compared-network-distance-vmt-and-data-limits.php/index.md