How Town Planning Works | TPW-0094 — The Accessory Dwelling Unit: How One Lot Becomes Two Homes Without Rebuilding the Neighbourhood

Some housing reforms begin with towers, rail lines and master plans.

An accessory dwelling unit begins with a side yard, a garage, a basement, an attic or a small piece of land behind a house.

That is why ADUs are so interesting to town planners. They can add a complete home without requiring a neighbourhood to be demolished and rebuilt. The street can look broadly familiar while the number of households it supports changes one lot at a time.

In the United States, the policy signal is now substantial. The American Planning Association reported in 2026 that 18 states limit local zoning authority over accessory dwelling units; 16 of those states require some form of by-right ADU pathway, while Maryland and Nevada require local jurisdictions to provide a legal route. Massachusetts reported on August 12, 2026 that 854 ADUs were approved in the first half of 2026, bringing the statewide total to 2,084 approved units since the start of 2025. Maryland requires local governments with planning and zoning authority to adopt a law authorizing ADUs by October 1, 2026 and has issued a model ordinance to support implementation.

This is not merely a trend about “tiny houses.” It is a planning question about whether one residential lot should be able to serve two households when the site, building and infrastructure can support it.

An ADU is secondary, but it is still a home

An accessory dwelling unit is generally a smaller independent dwelling located on the same lot as a primary home.

It can be inside the existing house, attached as an addition, created from a garage or other accessory structure, or built as a detached backyard cottage.

The exact legal definition varies by jurisdiction.

That variation matters because an ADU is not simply an extra bedroom. A true dwelling unit normally has the facilities needed for independent living and must meet applicable building, fire, sanitation and safety requirements.

The word “accessory” describes its planning relationship to the primary dwelling. It should not reduce expectations of habitability.

There are several ADU geometries

The same policy can produce very different physical outcomes.

An internal ADU can occupy part of an existing house, such as a basement or upper floor. An attached ADU expands the principal building. A detached ADU creates a separate structure in the rear or side yard. A garage or workshop can sometimes be converted into a dwelling where code and site conditions permit.

These types have different planning implications.

Internal conversions may create little visible change but can require fire separation, egress and utility work. Attached units can change building mass and setbacks. Detached units raise questions about rear-yard privacy, access, lot coverage and emergency response.

A good ordinance should therefore regulate the actual form rather than treating every ADU as one identical object.

The main planning idea is incremental density

An ADU adds density without necessarily changing the street into a different building type.

That makes it useful in places where large redevelopment is politically difficult, financially unnecessary or physically inappropriate.

One lot changes. Then another. Over time, the neighbourhood can support more residents, more age groups and more household types without one dramatic redevelopment event.

This is different from the broader housing-diversity problem addressed in TPW-0040 — The Missing Middle. Missing-middle housing includes duplexes, triplexes, courtyard housing and other small multi-unit forms. The ADU has a narrower job: it adds a secondary home while retaining a primary dwelling on the lot.

The distinction matters because zoning implementation, ownership and building relationships are different.

By-right approval changes whether legal permission becomes real production

A town can say it “allows ADUs” and still make them extremely difficult to build.

If every proposal requires a public hearing, discretionary approval, neighbour consent or a variance, time and uncertainty increase. Homeowners who are not professional developers may abandon the process.

That is why by-right or ministerial approval has become central to many state reforms.

The National Zoning Atlas distinguishes clearly between ADUs allowed by right, ADUs subject to hearings, ADUs with occupancy restrictions and places where ADUs are prohibited. Its current work shows that legal permission varies widely and that restrictions can determine whether an ADU is a practical housing option or merely a theoretical one.

By-right does not mean rule-free.

It means an applicant who meets objective rules can obtain approval without asking decision-makers for a case-by-case judgement about whether the household deserves to build.

Objective rules make small projects possible for ordinary owners

Most ADU applicants are not assembling a large development team.

They may be homeowners trying to create a unit for a parent, adult child, caregiver, tenant or future downsizing plan.

The code should therefore be legible.

Maximum size. Setbacks. Height. Lot coverage. Entrances. Parking. Utility rules. Tree protection. Fire access. Short-term rental restrictions. These should be clear enough that an owner can determine basic feasibility before paying for detailed design.

A simple rule is often more equitable than a flexible rule that requires expensive professional negotiation.

The objective is not to remove design quality. It is to put the design judgement where it matters and remove it where it merely creates procedural friction.

Setbacks can quietly prohibit the use

A detached ADU needs somewhere to stand.

If rear and side setbacks are copied from the main house, many ordinary lots can become impossible. This is especially true in older neighbourhoods where garages and sheds already sit close to property lines.

ADU rules therefore often use smaller accessory-structure setbacks, allow conversion of legally existing structures, or create separate standards for rear-yard buildings.

The setback should manage actual impacts: fire separation, privacy, maintenance access, drainage and neighbour interface.

A requirement that consumes most of the buildable yard without improving those outcomes becomes a hidden ban.

Good zoning tests whether the standard preserves a legitimate planning job or merely preserves an old geometry.

Height controls should distinguish one-storey cottages from two-storey pressure

Detached ADUs can affect privacy when upper-floor windows overlook neighbouring yards.

Height limits, upper-storey setbacks, window placement or landscape screening can manage that interface.

A one-storey rear-yard cottage usually creates a different neighbour relationship from a two-storey structure at the same boundary.

The code can recognize that difference instead of relying on one maximum height for every case.

Again, the planning job is not to guarantee that no neighbour can ever see another building.

It is to allow an additional home while keeping the resulting interface reasonable.

Lot coverage is a capacity question

An ADU adds roof and hard surface.

That can affect drainage, trees, usable yard space and heat.

Lot-coverage limits can therefore remain relevant, but they should be calibrated so that an ADU is physically possible on a reasonable share of eligible lots.

Some towns can pair ADU permissions with permeable paving, rain gardens, tree-retention rules or stormwater requirements where local conditions justify them.

What should be avoided is a collection of individually modest rules that combine into impossibility.

Zoning often fails through accumulation rather than one obvious prohibition.

Parking can become the hidden veto

An ADU may house one or two people and still trigger a requirement for one or two additional parking spaces.

On a small lot, the parking can consume more land than the home.

This is why many ADU reforms reduce or remove off-street parking requirements, especially near transit or in walkable areas.

The correct answer is context-dependent.

A rural ADU with no transit and long travel distances may generate more vehicle demand than an ADU beside frequent transit and shops. A neighbourhood with controlled curb parking behaves differently from a place where every household relies on private driveways.

The planning authority should therefore avoid using a universal parking rule as a substitute for actual transport analysis.

TPW-0034 — The Parking Equation owns the broader question of parking minimums, pricing and curb space.

Fire safety should be solved through building and access design

A second dwelling changes the life-safety condition of a property.

Internal and attached ADUs may require fire separation, alarms, safe egress and code-compliant construction. Detached units may need access for emergency response depending on distance, building configuration and local fire rules.

These are real requirements.

But zoning should not improvise technical fire rules where building and fire codes already govern the issue.

The planning role is to make sure the site geometry does not create an impossible access condition. A narrow side passage, long rear lot or steep slope may require a different design response.

The principle is the same as elsewhere in this series: connect specialist safety systems to land-use geometry without making one duplicate the other.

Water and sewer capacity can decide whether the unit is feasible

A dwelling is not only a building.

It consumes water and produces wastewater.

In urban areas with public systems, one additional small unit may be easily absorbed. In constrained sewer catchments, the cumulative effect of many units can matter. In rural areas, septic capacity can become the binding limit.

The code should therefore distinguish infrastructure contexts.

Requiring a full new utility connection in every case can make internal conversions unnecessarily expensive. Allowing unlimited occupancy on inadequate septic systems can create public-health problems.

The rule should match the actual capacity question.

Where utility upgrades are needed, the approval process should tell owners early rather than after architectural design is complete.

Utility metering is partly a housing-management decision

Some ADUs share water, electricity or gas meters with the primary home. Others have separate services.

Separate metering can improve billing clarity but add cost. Shared systems can reduce construction expense but complicate independent tenancy.

Planning codes should avoid mandating expensive infrastructure unless a clear public purpose requires it.

Utility providers may already have technical rules for service, capacity and meter location. Those rules should be coordinated with the ADU permit process so owners do not encounter contradictory requirements.

One small home should not need to navigate four agencies that each assume another one has explained the process.

Addressing sounds trivial until an ambulance needs the unit

A second dwelling needs a clear administrative identity.

Emergency services, postal delivery, utility records, rental documentation and census systems all work better when the ADU can be identified unambiguously.

A town should establish a simple addressing convention and coordinate it with emergency dispatch.

This is a small example of a larger truth in town planning: administrative details become physical infrastructure when people depend on them.

The goal is not bureaucratic complexity. It is the opposite—a clear identity for the second home so every other system knows what it is dealing with.

Owner-occupancy rules change who can build

Some jurisdictions require the owner to live in either the main house or the ADU.

The rationale may be neighbourhood stability or easier management.

The trade-off is that owner-occupancy requirements can reduce financing flexibility, complicate estate planning and prevent properties from being rented as two long-term homes when the owner moves.

The National Zoning Atlas specifically tracks occupancy restrictions because they materially affect whether ADUs can function as scalable housing.

There is no universal answer for every legal system.

But the planner should ask whether the restriction solves a demonstrated land-use problem or simply limits who may benefit from the housing type.

Family-only occupancy rules can defeat the housing function

An ADU is often described as an “in-law suite,” and family use is one of its strengths.

It does not follow that only relatives should be permitted to live there.

A family-only rule reduces the unit’s usefulness when the household structure changes. A parent may die. An adult child may move. A caregiver may be unrelated. A homeowner may need rental income later.

If the building is a lawful independent dwelling, restricting occupancy to a particular family relationship should require a strong policy justification.

Housing flexibility is one of the main reasons ADUs are valuable.

Short-term rental rules should be separated from ADU legality

Communities sometimes oppose ADUs because they fear the units will become holiday rentals.

That is a short-term-rental regulation problem, not necessarily an ADU problem.

A town can allow ADUs as long-term housing and separately regulate whether they may be used for transient accommodation.

Combining the two issues can create a blunt rule that removes useful housing in order to solve a tourism-management concern.

The estate already has a dedicated owner for that subject: TPW-0084 — The Short-Term Rental Map.

Good planning separates mechanisms before regulating them.

ADUs can support ageing without creating an age-restricted building

One of the strongest uses of an ADU is life-course flexibility.

An older resident can move into the smaller unit while a family member occupies the main house. A caregiver can live nearby without sharing the same interior space. An adult child can have independence while remaining close to family.

The arrangement can change again later.

This is different from age-restricted housing because the building remains ordinary housing capable of serving many household types.

Universal-design features can increase that flexibility: step-free entrances, wider doors, accessible bathrooms, manageable circulation and convenient parking or drop-off where appropriate.

The planning system does not need to require every ADU to be fully accessible in every jurisdiction, but permit-ready guidance can encourage layouts that remain useful as occupants age.

The ADU can be a caregiving infrastructure

Housing and care are often planned by different institutions.

Families experience them together.

When a dependent adult, older parent or person with disability can live close to family while retaining a private dwelling, the lot itself becomes a support system.

This can reduce travel between households and provide an alternative to moving everyone into one interior arrangement.

The same unit can later become ordinary rental housing.

That multi-stage usefulness is part of the ADU’s planning value. A flexible home can serve different social functions over decades without requiring demolition each time the household changes.

Financing can become the first weak link after zoning is fixed

Legalizing ADUs does not guarantee construction.

Homeowners still need design, contractor capacity and money.

Construction can be expensive relative to the small number of units produced. Existing homes may contain unknown structural or utility conditions. Appraisals and loan products may not value future rental income consistently.

Massachusetts’ 2026 policy is instructive because the state did not stop at by-right zoning. It launched an ADU financing programme, a design challenge and technical-support initiatives intended to help homeowners move from legal capacity to completed units.

The lesson is general.

Zoning opens the door. Finance, design and construction determine how many people walk through it.

Permit-ready plans can reduce repetitive design cost

ADUs are unusually suitable for repeatable design because many lots need similar small building forms.

A jurisdiction can provide pre-reviewed or permit-ready designs that owners adapt to site conditions. This can reduce uncertainty and professional cost while retaining building-code review for foundations, utilities, drainage and local constraints.

The estate already owns this mechanism at TPW-0082 — The Permit-Ready Pattern Book.

The relationship is simple.

The Accessory Dwelling Unit explains what the housing type needs from zoning and site planning. The Permit-Ready Pattern Book explains how standardized design pathways can reduce repetitive approval work after the use is legal.

Fees should be proportional to the second home

Development fees can quietly determine whether an ADU is viable.

If a 45-square-metre unit pays the same infrastructure or connection charge as a large detached house, the fee can represent a much larger share of project cost.

Some jurisdictions therefore scale fees by floor area, bedrooms, actual service demand or another proportional measure.

This does not mean the unit should pay nothing where it creates real infrastructure demand.

It means cost recovery should match the burden created rather than using a category designed for an entirely different development scale.

Small housing needs small-project economics.

Existing illegal units create a legalization problem

Many communities already contain unpermitted basement apartments, garage conversions or backyard dwellings.

When the law changes to allow ADUs, planners must decide what to do with this inherited stock.

A pure enforcement approach can displace residents and remove housing. A blanket amnesty can leave dangerous conditions uncorrected.

A legalization pathway can separate life-safety upgrades from less critical dimensional nonconformities.

For example, an older unit may need smoke protection, safe egress and electrical correction even if an existing wall sits slightly inside a newly required setback.

The goal is to convert hidden housing into safe, recognized housing where reasonably possible.

Heritage neighbourhoods need adaptation, not automatic exemption

Historic districts can have legitimate concerns about demolition, massing and visible alterations.

That does not mean ADUs are inherently incompatible with heritage.

Older urban neighbourhoods often historically contained carriage houses, rear cottages, servants’ quarters and secondary structures. In some places, the ADU is less a new invention than a return to an older lot pattern.

Heritage review can focus on visible demolition, roof form, material relationship or placement without requiring the second unit to imitate the main house exactly.

The preservation objective should be the significant character of the place, not the permanent freezing of household capacity.

Flood and wildfire constraints still apply

Incremental housing is still housing in a hazard landscape.

A backyard cottage should not be placed in a floodway simply because the zoning code now allows ADUs. A detached unit in wildfire-prone terrain still needs defensible space, access and applicable building standards.

Statewide ADU mandates can therefore coexist with objective hazard controls.

The important distinction is whether the rule responds to a real hazard or is being used as a disguised barrier to the housing type.

Hazard maps should decide hazard treatment. Housing politics should not be hidden inside technical language.

Subdivision and condominium ownership are separate policy choices

An ADU is usually conceived as a secondary unit on one lot.

Whether it can later be sold separately is another question.

Some legal systems permit condominium or strata arrangements that can create separate ownership interests. Others require the ADU and primary home to remain under common ownership.

This choice affects financing, affordability, speculation and the long-term meaning of “accessory.”

The zoning ordinance should therefore be explicit about what it regulates and should not accidentally create or prohibit separate ownership through vague language.

Land division, title and condominium law may belong to different authorities. The homeowner should not need to discover that distinction after construction.

Neighbourhood infrastructure sees the cumulative effect

One ADU rarely changes a school, road or sewer network by itself.

Thousands can.

This is not a reason to prohibit them. It is a reason to monitor production.

Planning departments should know where ADUs are being permitted, what types are being built, how many reach completion and whether infrastructure constraints are emerging in particular areas.

The cumulative effect can often be handled through ordinary capital planning because ADU growth is incremental and geographically dispersed.

That gradual pace is one of the housing type’s advantages. It gives public systems time to see the pattern forming.

The approval-to-completion ratio matters

Political announcements often count permits.

Housing supply depends on completed units.

A strong ADU programme should therefore monitor the conversion chain: inquiry to application, application to approval, approval to construction start, and start to occupancy.

If many owners ask questions and few apply, the code may still be too confusing. If approvals are high and completions are low, financing, contractor capacity or construction cost may be the binding constraint. If units are completed but rarely rented, occupancy rules or household use patterns may explain the difference.

This is exactly the kind of flow analysis owned by TPW-0051 — The Housing Observatory.

Massachusetts shows why implementation matters after reform

Massachusetts provides a useful live example because the state moved beyond a legal change into implementation infrastructure.

ADUs became permitted by right statewide under the Affordable Homes Act framework. By August 12, 2026, state officials reported 2,084 approved ADUs since the start of 2025, including 854 approvals during the first half of 2026.

The state also launched a homeowner assistance programme, financing support and an ADU design challenge.

The useful lesson is not that every place should copy Massachusetts.

It is that regulatory permission is only the first layer of a housing-production system. Technical help, finance, design resources and clear local administration determine whether the permission becomes homes.

Maryland shows how state policy can leave local implementation work

Maryland’s current framework demonstrates a different approach.

The state enacted an ADU policy in 2025 and requires counties and municipalities with planning and zoning authority to adopt local laws authorizing ADU development by October 1, 2026.

The Maryland Department of Planning has published a model ordinance and assessment tools to help local governments implement the law.

This reveals a recurring governance pattern.

The state can establish the minimum legal pathway while local government still needs to translate that policy into definitions, procedures, dimensional standards and administrative practice.

A mandate without implementation capacity can still produce confusion. Model ordinances reduce that friction.

The national trend is really about local certainty

The American Planning Association’s 2026 review of state zoning limits identifies 18 states that now constrain local ADU regulation in some way.

Sixteen require ADUs to be authorized by right under covered conditions. Maryland and Nevada require a legal path but may retain more local discretion.

The underlying policy debate is not simply state versus local control.

It is about certainty.

A household deciding whether to spend savings on a second home benefits from knowing that meeting objective rules will lead to approval. A contractor benefits from repeatable standards. A lender benefits from a recognizable legal product. A planning department benefits from fewer case-by-case hearings.

Predictability can itself become housing infrastructure.

ADUs should not be sold as a complete housing solution

An ADU will not solve every housing shortage.

It produces one unit at a time. Construction costs can be high. Many owners cannot finance a project. Some lots cannot physically support another dwelling. The units are often small. They do not replace the need for apartments, public housing, affordable housing programmes, family-sized homes or major redevelopment where those are needed.

The strength of the ADU is different.

It creates a low-disruption way to add housing capacity across a large existing neighbourhood fabric.

A good housing system uses several production channels at once.

The ADU is one of the smallest channels, but because there are so many existing residential lots, the aggregate opportunity can be significant.

The ordinance should be tested against a real ordinary lot

One of the simplest planning tests is also one of the most revealing.

Take a typical lot in the district.

Draw the required setbacks. Add lot coverage. Add the parking space. Add access. Add utility clearances. Add tree protection. Apply the maximum ADU size and height.

Is there still a plausible home?

If most ordinary lots fail, the ordinance may technically legalize ADUs while functionally prohibiting them.

This is why implementation should be tested geometrically before adoption.

Planning rules should be read on paper and drawn on land.

A practical ADU ordinance structure

Definition: identify internal, attached, detached and conversion forms.
Eligibility: define which lots and principal uses may contain an ADU.
Approval: use an objective process where policy intends by-right construction.
Size: establish a clear maximum appropriate to the primary dwelling and local context.
Setbacks: use standards that actually permit reasonable rear- or side-yard placement.
Height: manage privacy and mass without eliminating viable forms.
Coverage: coordinate building footprint, drainage and open space.
Parking: require only what the transport context reasonably justifies.
Utilities: clarify water, sewer, septic and metering procedures.
Fire and building safety: reference specialist codes and preserve necessary access.
Occupancy: decide explicitly whether owner-occupancy or family-only restrictions serve a demonstrated purpose.
Short-term rental: regulate transient use separately where possible.
Addressing: create a simple emergency and administrative identity.
Fees: keep charges proportionate to project scale and service demand.
Legalization: provide a route for safe existing units to enter the formal system.
Monitoring: track approvals, completions and infrastructure effects.

A planner’s ADU feasibility audit

Type: Is the proposed unit internal, attached, detached or a conversion?
Legal path: Can a compliant project be approved predictably?
Geometry: Do setbacks, height and lot coverage leave a buildable envelope?
Privacy: Are upper-level windows and outdoor spaces reasonably arranged?
Fire: Can applicable life-safety and access requirements be met?
Water: Does the site have sufficient supply?
Wastewater: Can sewer or septic capacity support the additional dwelling?
Parking: Is any requirement proportionate to actual transport context?
Accessibility: Can the design support ageing or mobility needs where desired?
Address: Can emergency services identify the unit clearly?
Finance: Can an ordinary homeowner realistically fund construction?
Fees: Are municipal and utility charges proportionate?
Ownership: Are rules about owner occupancy or separate sale explicit?
Rental use: Are long-term and short-term occupancy rules clearly separated?
Hazards: Do flood, wildfire, slope or other constraints require special treatment?
Delivery: What percentage of approvals actually becomes completed housing?

The smallest planning unit can still change the housing system

An ADU does not redraw a skyline.

It changes who can live on one piece of land.

A parent can live near an adult child. A homeowner can create rental income. An older resident can downsize without leaving the property. A caregiver can live close by. A neighbourhood can absorb new households without waiting for a major redevelopment site.

Those small changes accumulate.

The planning challenge is to make the second home legal enough to build, safe enough to inhabit, simple enough for an ordinary owner to understand and regulated enough that its relationship with neighbours and infrastructure remains workable.

One lot becoming two homes is not dramatic urbanism.

That is exactly why it can matter.