← Progress Portsmouth
← Library HAP Triage Module

Housing Action Plan — Module 02

Missing Middle Housing
in Portsmouth

Portsmouth's current zoning creates a gap between single-family homes and large apartment buildings — a range of housing types that once defined New England neighborhoods but are now largely prohibited from being built new. Three questions frame what the Housing Action Plan addresses and why it matters.

"Missing middle" housing refers to multi-unit or clustered housing types compatible in scale with single-family neighborhoods — duplexes, triplexes, fourplexes, cottage courts, townhouses, and small apartment buildings. They are "missing" not because they don't exist in Portsmouth's older neighborhoods, but because current zoning largely prevents building new ones.

The housing type spectrum — click any type to explore

Key points
Portsmouth's pre-war neighborhoods are a missing middle demonstration project. A walk down any street built before 1940 shows these building types at New England scale — duplexes, mansion apartments, cottage clusters.
The ADU update (February 17, 2026) is the most recent addition to the by-right column. It is the model for how the Housing Action Plan treats other missing middle types: remove the discretionary hurdle, establish clear standards, permit by right.
Ask: how many people live in — or grew up in — one of these building types? The answer almost always grounds the conversation in something concrete rather than abstract.

The barriers to missing middle housing are structural, not cultural. Portsmouth residents broadly support neighborhood character — but current zoning conflates character with housing type, when the evidence from Portsmouth's own historic fabric shows they are separable. Click any row to expand.

Current barriers — click to expand

New duplex construction

Variance required in most residential zones
Variance required
In most Portsmouth residential zones, historic duplexes are legal nonconforming — they exist but cannot be replicated. New duplex construction requires a variance or is simply not a permitted use. The result is that the existing stock ages without replacement, and the most accessible entry point into small-scale infill is effectively closed.

Triplex & fourplex

Not permitted by right in residential zones
Not by-right
Three- and four-unit buildings occupy a gap in Portsmouth's code. Federally, buildings up to four units qualify for conventional mortgage financing — making them one of the most financially accessible development types for small-scale builders. Portsmouth's zoning does not leverage this. Most proposals require a variance, which means most are never proposed.

Cottage court / cluster housing

No enabling use definition in current code
No clear pathway
Cottage courts — small detached units around a shared courtyard — are not a defined use type in Portsmouth's zoning ordinance. Tracy Kozak's co-living approval under Section 10.815 is the closest precedent, but it required a Conditional Use Permit and remains limited to Gateway districts. A property owner who wanted to build four small cottages on a larger lot has no clear path through the current code.

Parking minimums

Cross-cutting barrier to small-scale infill feasibility
Cross-cutting
Parking minimums are the single most common reason missing middle housing fails financial feasibility in built-out urban areas. A surface parking space costs $5,000–$15,000 to construct on urban land. Requiring one or two spaces per unit on a small infill lot frequently eliminates the project entirely — not because the housing type is wrong, but because the parking consumes the buildable area or the pro forma. Reducing minimums near walkable areas is the highest-yield zoning change for small-scale infill. HAP item Z-4 addresses this directly.

Lot size & dimensional standards

Makes missing middle types geometrically impossible
Precondition barrier
Minimum lot sizes, frontage requirements, and setbacks written for single-family development often make missing middle types physically impossible even when the use is technically permitted. A duplex that requires a 10,000 sq ft minimum lot in a neighborhood where most lots are 6,000 sq ft is not actually permitted — the dimensional standard eliminates it in practice. Lot size reform (Z-8) is a sequencing dependency for Z-5 and Z-9 to have full reach.

ZBA discretionary review

Variance as de facto veto point
Structural barrier
When the ordinance requires a variance for housing types that were once by-right, the Zoning Board of Adjustment becomes a de facto veto point. The blocking power flows from ordinances, not from board culture. The fix is ordinance reform — which is what HAP items Z-5, Z-9, and Z-22 address. P-7 addresses the variance process directly: replacing variance requirements with clear by-right standards where possible.

The legal and regulatory frame

Why missing middle is banned

Most of Portsmouth's residential zones permit single-family homes as-of-right but require a variance or special exception for duplexes, triplexes, or small multifamily. These are the same housing types that make up much of Portsmouth's older walkable neighborhoods — built before current zoning existed.

RSA 672:1 III-e — prohibition on unreasonable discouragement of housing

The zoning history

Mid-20th century zoning reforms replaced use-mixed neighborhood patterns with single-use residential zones. Buildings considered normal in Portsmouth's historic core are now nonconforming — legal to occupy, but illegal to replicate — across most of the city's residential land area.

See: HAP items Z-5, Z-9, Z-22 (zoning reform category)
Key points
Residents who oppose "density" are almost never opposing the building types that already define their neighborhoods. They're opposing large-scale apartment complexes. Missing middle — house-scale, contextually integrated — is what most residents already live in and around.
Parking minimums (Z-4) are the fastest win. They can be reduced by ordinance amendment without changing permitted uses, and the political resistance is lower than use-expansion changes.
The co-living precedent (Tracy Kozak, Section 10.815) demonstrates the ZBA can approve these uses when the ordinance allows it. The lesson: fix the ordinance, not the board.
For triage: the question is not "do we want these building types" but "which zoning barriers have the most straightforward removal path — legal, political, and administrative."

The Housing Action Plan addresses the root causes of the missing middle gap directly. The missing middle lens does not add items to the HAP — it reorders the argument for sequencing existing items based on barrier-removal yield relative to implementation complexity.

Housing Action Plan items that address the missing middle

Click any item to see the full description from the adopted Housing Action Plan. Yield ratings reflect barrier-removal impact relative to implementation complexity.

Z-5
Missing middle housing allowances
Allow duplexes, triplexes, fourplexes by-right in single-family zones with clear dimensional standards. Form-based standards rather than use-based restrictions. The core missing middle item — directly addresses the by-right gap for small-scale infill without requiring ZBA discretionary review.
High yield
Z-9
Cottage housing / cluster development provisions
Create zoning for cottage housing clusters: small detached units around shared open space, with reduced lot-per-unit requirements. Addresses the definition gap — cottage courts are not a defined use type in current Portsmouth zoning. Z-9 establishes the standards needed to permit this form without a variance.
High yield
Z-22
Zone specifically for diverse housing
Make diverse housing an explicit zoning purpose in targeted areas so it doesn't require variances or special exceptions. When the ordinance doesn't name housing diversity as a purpose, individual board members can interpret standards narrowly. Z-22 removes that ambiguity by embedding the purpose in the zone itself.
High yield
Z-4
Parking requirement reform
Phase 1: Comprehensive parking study for downtown and mixed-use zones. Phase 2: Implement parking maximums, shared parking, and reduced minimums. Cross-cutting enabler — reducing parking minimums near walkable areas immediately improves feasibility for small-scale infill without changing permitted uses. Can be scoped to specific overlay areas to reduce political surface area.
High yield
P-7
Variance streamlining for housing
Review variance process for housing. Replace variance requirements with clear by-right standards where possible. Directly addresses ZBA as a blocking point — the most effective lever is reducing the universe of housing proposals that require a variance in the first place.
High yield
Z-8
Lot size and dimensional requirement updates
Reduce minimum lot sizes, frontage requirements, and setbacks in targeted zones for infill development. GIS analysis provides the evidence base. A sequencing dependency for Z-5 and Z-9 — missing middle types permitted on paper but blocked by dimensional standards have no real pathway without this item.
Precondition
Z-2 + P-2
ADU ordinance enhancements + administrative approval
By-right ADU status achieved February 17, 2026. Z-2 addresses enhancements beyond state compliance: dimensional standards, design guidance, and ordinance optimization. P-2 enables Planning Director approval with 30-day max review, eliminating board hearings for compliant applications. The HAP's most recent tangible win and the model for how zoning reform produces results.
Achieved

For triage: Missing middle items in the HAP are not about large-scale development — they're about replicating the building patterns already present in Portsmouth's most livable neighborhoods. The question for each item is whether it gives those patterns a legal pathway, and whether that pathway requires discretionary review or is truly by-right. Z-5, Z-9, and Z-22 address use permission. Z-4 and Z-8 address enabling conditions. P-7 addresses the process.

Key points for triage
Z-4 (parking) and Z-8 (lot dimensions) are enabling conditions — other items can be adopted in ordinance but have limited real-world reach without them. Treat them as sequencing dependencies, not separate priorities.
Z-5 is the most consequential single item. It is also the most politically exposed. The committee's triage recommendation can sequence it for Year 1 advocacy even if Year 2 implementation — naming it clearly signals intent to Council.
The missing middle framing gives residents a way to support HAP items without endorsing "density" as an abstract concept. That political cover is worth making explicit in public communications about triage recommendations.
The ADU outcome (Z-2 + P-2) is the proof-of-concept for this entire category. Cite it in every conversation about whether zoning reform can produce results on a reasonable timeline.