Parking Minimum Repeals: Why Scope Beats Size in Permit Yields

In three years after Buffalo wiped its parking minimums citywide in 2017, 47% of new residential projects built fewer spaces than the old code would have forced. This single metric exposes a structural flaw in how cities measure reform success: narrow geographic carve-outs rarely move the needle on actual development patterns. When Minneapolis adopted its downtown-only repeal in 2015, it created a polished exception zone while leaving arbitrary requirements standing across the rest of the urban fabric. The result was predictable—permit yields barely shifted outside the central core, and developers continued designing around constraints they never intended to follow.

The Binding Constraint

The Minneapolis City Council's 2015 ordinance eliminated minimums only within the downtown overlay district, roughly the CBD bounded by the interstate loops. It left the one-space-per-unit-style minimums intact for the Uptown, corridor, and neighborhood districts where most parcels are smaller and minimums bind hardest. According to Streets.mn, the Minneapolis City Planning Commission has scheduled proposals from 10th Ward Councilmember Lisa Bender to reduce or eliminate off-street parking requirements for residential development, confirming that the 2015 repeal did not resolve the binding constraint across the broader urban fabric. Because the 2015 policy applied to a small share of parcels, it tells you almost nothing about what repeal does on the vast majority of urban land where minimums actually bind. That is the myth: that Minneapolis's 2015 repeal proves parking reform unleashes housing everywhere it is tried. The data shows it only unlocked capacity in the downtown pocket where the constraint was already loose.

Buffalo's Green Code, adopted in April 2017 under Mayor Byron Brown and Planning Commissioner Brendan Mehaffy, deleted minimums citywide across all zoning districts. Buffalo became the first large U.S. city to do so, and simultaneously capped maximums downtown to prevent oversupply. By covering ~100% of parcels, the ordinance attacked the binding constraint at its source. The scope of the repeal determined the yield. A citywide removal forces the pro forma to confront the true cost of parking on every parcel, whereas a district-only pilot leaves the binding constraint intact everywhere else, starving projects outside the overlay of the capital conversion that drives permit yields.

To measure the effect accurately, you must define 'permit yield' operationally using three metrics reported separately. First, units permitted per year in formerly-constrained districts. Second, the share of permits building zero structured parking. Third, the average off-street ratio (spaces per unit) on post-repeal permits. If you collapse these into a single headline number, the comparison between Buffalo and Minneapolis collapses. The share of zero-structured-parking permits reveals how many projects successfully shed the capital lock-in. The average off-street ratio shows the new equilibrium. Reporting them together masks whether the repeal actually converted parking capital into units or merely reduced ratios without increasing supply.

The scope-vs-demand distinction governs the entire comparison. Repeal only raises permit yield where the minimum was binding. Therefore, a repeal's yield is a function of how much land it covers. Buffalo covered the whole city; Minneapolis 2015 covered a fraction. Independent of how hot the market is, the citywide approach captures the full value of converting parking capital into housing. Advocacy must model this geometry. Push for a single citywide ordinance paired with curb pricing, not a district-by-district pilot that leaves the binding constraint intact everywhere else. The data confirms that geographic scope, not city size or market temperature, determines how many projects can shed structured parking and convert that cost into more units per permit.

The Permit Ledger

The Permit Ledger

Buffalo’s 2017 Green Code did not merely delete a floor; it rewired the spatial distribution of parking-light development across an entire municipal grid. According to Hess and Rehler (2021, Transport Findings / University of Toronto), 47% of residential projects permitted after the repeal provided fewer spaces than the old minimums would have mandated, pulling the citywide average down to roughly 0.7 spaces per unit from a pre-repeal baseline near 1.0+. Crucially, those below-minimum permits did not cluster in a single transit corridor or downtown envelope. They materialized across Elmwood Village, Larkinville, and multiple near-district corridors, proving that when the binding constraint is removed citywide, developers respond at the parcel level rather than waiting for centralized infrastructure subsidies.

Minneapolis tells a different ledger. City permit records (Minneapolis CPED) and subsequent academic analyses show downtown residential permitting accelerated after 2015—from a pre-repeal average to higher annual counts during the subsequent years—but the data cannot separate the repeal's effect from the concurrent downtown apartment boom and Green Line/LRT investment. More importantly, the geographic distribution reveals the structural limitation of a district-only pilot: post-2015 parking-light permits concentrated inside the downtown overlay, while neighborhood-district permits still showed ratios at or above the old minimums until later citywide reforms. The ordinance shifted supply into a high-demand zone without altering the underlying cost structure on the vast majority of urban land where minimums actually bind.

Buffalo’s ledger also captures a second-order mechanism that Minneapolis’s 2015 ordinance missed. Because the Green Code imposed parking maximums downtown alongside its minimums, post-repeal Buffalo permits show both fewer oversupplied lots and fewer forced lots—a two-sided outcome that simultaneously eliminated capital waste and unlocked density. Minneapolis’s 2015 ordinance, which only removed minimums, produced neither the downward correction nor the upward flexibility. When developers face only a floor, they build to it; when they face a bounded range, they optimize within it.

Geographic scope dictates permit yield more reliably than municipal scale or cyclical demand. When zoning reform deletes a binding floor, the spatial distribution of freed parcels determines whether developers can actually convert saved parking costs into additional housing units. The 2017 Buffalo Green Code and the 2015 Minneapolis downtown overlay provide a clean natural experiment in how coverage radius shapes development response.

Buffalo's citywide repeal and Minneapolis's downtown-only pilot both generated permit surges, yet the raw yield numbers mask structural confounds that determine when this policy actually transfers. The honest reading requires isolating market noise from zoning effects, recognizing geographic limits in shrinking markets, acknowledging measurement gaps on street supply, accounting for selection bias in permitted samples, and disaggregating bimodal developer responses across project scales.

Scope Beats Size

The Minneapolis counterfactual isolates why the 2015 downtown-only repeal failed to replicate this yield gain outside the overlay. If the identical 60-unit project were sited in a Minneapolis neighborhood beyond the downtown boundary, the 2015 ordinance imposes no relief; the developer still owes ~60 spaces. The pro forma remains unchanged, locking up the same capital and sacrificing the same land. This arithmetic proves the 2015 repeal's permit-yield effect was geographically capped at the overlay line. While Minneapolis offers special program pricing and loan offerings for energy improvements based on neighborhood designation, those incentives cannot overcome the hard constraint of a parking minimum that persists on the vast majority of urban land. The pilot leaves the binding constraint intact everywhere else, rendering the downtown exception irrelevant for the vast majority of infill sites.

Buffalo wins on permit yield per covered parcel because its ordinance targeted the exact land typology where parking minimums choke feasibility: small and mid-size infill lots that cannot absorb structured parking without destroying unit economics. Minneapolis’s 2015 overlay removed floors on large downtown sites where developers were already pricing parking into pro formas as a baseline cost. Deleting a requirement that was never binding does not unlock new supply; it merely removes paperwork. Buffalo’s citywide sweep forced the constraint off the parcels that actually needed relief, allowing developers to reallocate capital directly into square footage rather than concrete stalls.

The data confirms that demand intensity did not drive the divergence. According to post-2015 rental market tracking, Minneapolis experienced sustained boom-era rent growth through the late 2010s, which should have amplified any repeal effect. Yet the below-minimum response remained geographically trapped inside the downtown overlay, proving that market heat alone cannot overcome a spatially truncated policy. Scope, not demand, set the yield ceiling. The later Minneapolis citywide repeal under the city's planning framework serves as the confirming natural experiment: once coverage expanded beyond the central business district, parking-light permits spread neighborhood-by-neighborhood. Those later outcomes belong to a distinct ordinance and must not be back-attributed to the 2015 downtown-only rule.

For planners evaluating their own jurisdictions, the transfer rule is straightforward. Run the coverage row test against your parcel map. If your binding constraints sit on neighborhood-scale infill (the default for most U.S. cities), adopt the Buffalo model: a single citywide repeal paired with dynamic curb pricing to internalize externalities. If your constraint is genuinely downtown-only—a rare configuration in contemporary zoning—the Minneapolis 2015 approach suffices. Measure your coverage radius first; the rest follows mechanically.

What the Data Doesn't Tell You

When evaluating a parking reform proposal, practitioners must ignore municipal prestige and focus on the spatial distribution of binding constraints. A repeal's efficacy is determined by parcel coverage, not the year of enactment or the city's demographic weight. Compute the share of total parcels the ordinance frees from minimums; Minneapolis's 2015 downtown-only overlay fails this test because it leaves the binding constraint intact across the vast majority of urban land, whereas Buffalo's 2017 Green Code passes by applying citywide. If your analysis relies on a district-level pilot without verifying how much constrained supply remains elsewhere, you are measuring a local exception rather than a systemic shift.

The Minneapolis 2015–2019 downtown permit surge cannot be cleanly attributed to the repeal. According to data and developer feedback cited by the Federal Reserve Bank of Minneapolis, regional housing dynamics—including a broader apartment boom and major transit investments—coincided with Buffalo's citywide reform window. No published study isolates the repeal's causal contribution from this demand-side acceleration. Consequently, Minneapolis's yield figures function as an upper bound contaminated by market growth rather than a clean estimate of zoning elasticity. This does not invalidate the repeal's role but demands caution: in markets where demand is already surging, the marginal yield gain from removing parking minimums may be indistinguishable from business-as-usual density increases.

Buffalo's results carry a reverse confound rooted in its shrinking-city profile. Cheap land, low construction costs, and a fabric of small lots mean developers faced minimal economic pressure to build parking even under minimums. The 47% below-minimum provision observed post-repeal likely reflects cost structures unique to low-density, low-cost environments. In high-cost cities where land values justify structured parking regardless of zoning, developers might voluntarily build parking at prior ratios even after repeal. Here, the ordinance changed the legal constraint but not the underlying economics of land; yield gains would depend on whether parking was truly binding or merely habitual. This suggests Buffalo's model transfers poorly to markets where parking remains economically rational independent of regulation.

Neither city's permit ledger captures spillover effects on street supply. Claims that repeal 'frees street space' or 'floods curbs' lack empirical grounding because neither municipality published before-and-after curb occupancy studies. Buffalo's Green Code paired repeal with downtown maximums, which theoretically mitigates street congestion, but without occupancy data, the net effect on curb availability remains speculative. Policymakers should verify local curb utilization rates before assuming repeal automatically improves street access; the permit count says nothing about on-street demand.

Selection and survivorship bias further complicate interpretation. Hess and Rehler's Buffalo sample covers only projects actually permitted post-2017. Proposals that remained infeasible despite repeal—and developers who consistently built below-minimum via variances before 2017—are invisible in the dataset. If many marginal projects were previously blocked, the measured yield increase could be inflated. Conversely, if a significant share of below-minimum development already occurred through variances, the repeal's incremental impact might be deflated. This uncertainty applies broadly: any jurisdiction adopting citywide repeal must assess its own pipeline of stalled proposals and variance-dependent projects to gauge realistic yield potential.

Finally, yield aggregates mask a bimodal response across project types. Below-minimum provision concentrated in small multifamily and adaptive-reuse projects, where parking costs dominate pro formas. Large downtown towers in both cities continued building structured parking near prior ratios, suggesting economies of scale or market expectations override zoning flexibility. A single citywide 'yield' number obscures this divergence. When transferring Buffalo's approach, policymakers must disaggregate by building scale: reforms may unlock infill density while leaving large-scale development patterns unchanged. Saint Paul's mixed results, noted by the Federal Reserve Bank of Minneapolis, underscore that neighboring jurisdictions can diverge sharply based on such typological differences.

Worked Case

Headlines citing aggregate permit counts obscure whether yield gains are broad or concentrated. Demand three distinct metrics before accepting any claim: units permitted per year, the share of new permits falling below the old minimum threshold, and average spaces per unit. A single aggregate number can mask a scenario where growth is driven solely by large projects that still build excessive parking. The signal of success is a rising share of below-minimum provision, indicating developers are actively shedding structured costs to add density. Without this breakdown, you cannot distinguish between genuine parking-light conversion and business-as-usual development.

A repeal without a pricing companion invites spillover externalities and weakens the policy's empirical defense. Buffalo's approach bundled downtown parking maximums with its citywide repeal, managing curb demand while freeing interior parcels. A standalone repeal leaves off-street spillover risk unmanaged, providing opponents with their strongest attack vector regarding neighborhood congestion. Verify that the ordinance includes paired maximums or dynamic curb pricing; otherwise, the policy treats the symptom of excess supply while ignoring the mechanism of street saturation.

Market conditions dictate how you attribute yield to the ordinance. In high-heat markets like Minneapolis (2015–2019), permit growth may correlate with broader economic cycles. Attribute repeal effects only where below-minimum provision appears on parcels the ordinance actually freed. In flatter markets like Buffalo, treat below-minimum provision as the cleaner signal of the policy's mechanical effect. Always discount yield claims that do not isolate the treatment group—parcels released by the specific geographic scope of the reform—from general market trends.

Match the model to your binding parcels. If your constrained supply consists of small neighborhood infill lots, replicate Buffalo's citywide ordinance to unlock dispersed capacity. If the constraint is genuinely a downtown phenomenon, a focused repeal may suffice. However, if you cannot name your binding parcels, you are not ready to draft the ordinance. Saint Paul, for example, adopted an ordinance allowing up to four to five units on lots that previously only permitted single-family detached homes, illustrating a targeted response to specific lot-type constraints rather than a blanket removal of all parking floors. Align the geographic scope with the actual location of the binding constraint.

How to Choose Well

When evaluating a parking reform proposal, practitioners must ignore municipal prestige and focus on the spatial distribution of binding constraints. A repeal's efficacy is determined by parcel coverage, not the year of enactment or the city's demographic weight. Compute the share of total parcels the ordinance frees from minimums; Minneapolis's 2015 downtown-only overlay fails this test because it leaves the binding constraint intact across the vast majority of urban land, whereas Buffalo's 2017 Green Code passes by applying citywide. If your analysis relies on a district-level pilot without verifying how much constrained supply remains elsewhere, you are measuring a local exception rather than a systemic shift.

Headlines citing aggregate permit counts obscure whether yield gains are broad or concentrated. Demand three distinct metrics before accepting any claim: units permitted per year, the share of new permits falling below the old minimum threshold, and average spaces per unit. A single aggregate number can mask a scenario where growth is driven solely by large projects that still build excessive parking. The signal of success is a rising share of below-minimum provision, indicating developers are actively shedding structured costs to add density. Without this breakdown, you cannot distinguish between genuine parking-light conversion and business-as-usual development.

A repeal without a pricing companion invites spillover externalities and weakens the policy's empirical defense. Buffalo's approach bundled downtown parking maximums with its citywide repeal, managing curb demand while freeing interior parcels. A standalone repeal leaves off-street spillover risk unmanaged, providing opponents with their strongest attack vector regarding neighborhood congestion. Verify that the ordinance includes paired maximums or dynamic curb pricing; otherwise, the policy treats the symptom of excess supply while ignoring the mechanism of street saturation.

Market conditions dictate how you attribute

Frequently Asked Questions

What specific percentage of Buffalo's post-2017 residential projects built fewer parking spaces than the old code required?

47% of new residential projects in Buffalo built fewer spaces than the old code would have forced.

Which Minneapolis districts retained their one-space-per-unit minimums after the 2015 downtown repeal?

The Minneapolis City Council's 2015 ordinance left the one-space-per-unit-style minimums intact for the Uptown, corridor, and neighborhood districts.

How did Buffalo's Green Code address the risk of excessive parking construction alongside its citywide minimum repeal?

Buffalo simultaneously capped maximums downtown to prevent oversupply.

What was Buffalo's average off-street parking ratio per unit after implementing its citywide repeal?

The citywide average dropped to roughly 0.7 spaces per unit from a pre-repeal baseline near 1.0+.

Why is it difficult to isolate the exact impact of Minneapolis's 2015 downtown parking repeal on residential permitting?

The data cannot separate the repeal's effect from the concurrent downtown apartment boom and Green Line/LRT investment.

What three operational metrics should cities use to accurately measure permit yield after a parking reform?

You must define 'permit yield' operationally using three metrics reported separately: units permitted per year in formerly-constrained districts, the share of permits building zero structured parking, and the average off-street ratio on post-repeal permits.

Quick answers

Also worth reading: The biggest urban development stories and city planning news from the past week: biggest urban development stories and · How smart city technology transforms the future of urban development: How smart city technology transforms · Smart City Explained Your complete guide to intelligent urban development: Smart City Explained Your complete