How California's Replacement Unit Requirement Discourages Development in Lower-Income Neighborhoods

What the rule requires

The replacement requirement is codified in California Government Code §65915(c)(3) and was significantly strengthened by SB 8 (2021), AB 1218 (2023), and AB 1893 (2024).

When a project proposes to demolish existing dwelling units, the developer must determine whether any of those units were occupied by lower-income households at any point during the previous five years, OR were subject to rent restrictions, recorded affordability covenants, or housing assistance contracts during that period. Where any such units exist, the new project must include at least as many units affordable to the same or lower income levels as the displaced units — and those replacement units must match the bedroom counts of the units being demolished. A two-bedroom demolished requires a two-bedroom replacement at the same affordability level. A three-bedroom requires a three-bedroom.

The requirement applies to any tenant-occupied dwelling unit on the site — not just multifamily buildings. A single-family home that has been rented to a lower-income tenant within the previous five years triggers the obligation just as surely as a 12-unit apartment building. A duplex with one or both units occupied by lower-income renters triggers it for those units. The rule doesn't care about the structure type; it cares about the tenancy history.

San Diego allows the same affordable units to satisfy both obligations

The most important thing to understand about how the replacement requirement works in San Diego: the same affordable units can count toward both density bonus set-asides and replacement obligations. A project committing to 12 Very Low Income units to qualify for a 50% density bonus, on a site that requires 2 replacement units, doesn't need to commit 14 affordable units total. The replacement units count within the density bonus set-asides, and the project's total affordable commitment remains 12.

This is significant. In jurisdictions that don't allow this overlap, replacement obligations stack on top of density bonus commitments, dramatically increasing the project's affordability exposure. San Diego's approach softens that math. A developer building under State Density Bonus Law on a site with replacement obligations doesn't necessarily face additional affordable units — they face a constraint on how the affordable units they were already going to build must be configured.

Where the real constraint lives: bedroom count

The bedroom-matching requirement is where projects break down — and where the surprising result is that single-family and duplex acquisitions can be just as constrained as larger multifamily acquisitions.

Older single-family homes in San Diego's lower-income neighborhoods commonly have two, three, or even four bedrooms. They were built as family homes, and the neighborhoods where they're concentrated — North Park, City Heights, Linda Vista, parts of Mid-City — were originally working-class family neighborhoods. Duplexes in the same neighborhoods follow a similar pattern, with two-bedroom and three-bedroom units that were standard for that era.

New multifamily development in the same neighborhoods is overwhelmingly studios and one-bedrooms. Smaller units are more efficient to build, command higher rent per square foot, and serve the demographic that actually rents in transit-accessible urban infill: single professionals and couples without children. A typical new infill project in 2025 might contain 60% studios, 30% one-bedrooms, and 10% larger units, if any.

When the replacement requirement is applied, this preference becomes a problem. A developer demolishing a three-bedroom rental house is required to build a three-bedroom replacement unit at the same affordability level. The provision allowing that unit to count toward the density bonus set-aside helps — but only if it's configured as a three-bedroom. The developer cannot satisfy the obligation by building affordable studios where an affordable three-bedroom is required.

This forces the project to include unit types the developer would not otherwise build. Two-bedroom and three-bedroom units cost more per unit to construct, generate less rent per square foot than smaller units, and consume more of the building's gross floor area for the same unit count. The financial penalty lands as a unit-mix constraint, not a unit-count addition. The constraint is real, just quieter than the unit-count framing suggests.

Why this disincentivizes development in lower-income neighborhoods

The bedroom-mix constraint creates a sharp distinction between two categories of infill sites. Sites without recent lower-income tenancy — vacant lots, owner-occupied homes, properties with continuously higher-income tenancy — carry no replacement burden. The developer can build whatever unit mix the market and the project's pro forma support. Sites with recent lower-income tenancy — single-family rentals, duplexes, older multifamily buildings — carry the bedroom-mix constraint.

The geographic correlation between these two categories is direct. Tenant-occupied properties with significant two-bedroom and three-bedroom inventory are precisely what exists in San Diego's older lower-income neighborhoods. Wealthier neighborhoods contain less of this housing type — more owner-occupied homes, more recent construction, fewer rental properties with extended lower-income tenancy.

But the asymmetry doesn't stop with the replacement burden. It compounds with a second asymmetry that runs in the same direction: market-rate rents.

Design, permitting, and construction costs for similar projects are essentially uniform across the city. A 50-unit infill apartment building has roughly the same hard costs regardless of which San Diego neighborhood it goes in. Architects charge the same fees. Permits cost the same. Construction labor and materials cost the same.

Market-rate rents, however, vary dramatically by neighborhood. The same studio that rents for $2,000 a month in a wealthier neighborhood might rent for $1,400 in a lower-income inland neighborhood. The same one-bedroom in a wealthier neighborhood might out-rent its lower-income equivalent by 30% or more. This difference in market-rate revenue is what determines whether the project's unrestricted units can cover construction costs, debt service, and the affordability burden the project carries.

The replacement requirement places its heaviest burden — bedroom-mix constraints that force developers to build larger, less revenue-efficient units — precisely in the neighborhoods where the market-rate units generate the least revenue to offset that burden. The cost is the same. The affordability constraint is heavier. The revenue available to absorb both is lower.

A developer evaluating two parcels with similar zoning, similar entitlement complexity, and identical construction costs — one in a lower-income neighborhood with a tenant-occupied existing house, one in a wealthier neighborhood with an owner-occupied house — finds that the lower-income parcel imposes a unit-mix constraint that reduces the project's economic returns AND offers lower market-rate rents to absorb that constraint. Identical projects pencil very differently in the two locations.

This is the opposite of the policy goal that produced the replacement requirement. The rule was created to protect existing lower-income tenants from displacement. In practice, the combination of bedroom-mix mandates and geographically variable market-rate revenue routes new development toward neighborhoods where the constraint doesn't apply — the neighborhoods where market-rate rents can absorb whatever affordable burden the project carries voluntarily. Lower-income neighborhoods see less new development — including less affordable housing — than they would in a regulatory environment that didn't penalize their specific housing stock.

What this means for infill site evaluation

Anyone evaluating an infill development site in San Diego should make the bedroom-count compliance question an explicit early-stage diligence task.

Has the property been tenant-occupied within the past five years, and what were the tenants' income levels? This is the threshold question. Owner-occupied homes and vacant lots have no replacement obligation. Tenant-occupied properties may, depending on tenant income history. This requires actual investigation — rent rolls, property management records, tenant history — not estimation from neighborhood demographics.

What is the bedroom mix of the existing units? Single-family homes and older duplexes commonly have two-, three-, or four-bedroom configurations. These create much more constraining replacement obligations than studios or one-bedrooms.

Does the project pencil with the required unit mix and the neighborhood's market-rate rents? This is the critical question. The project must absorb both the bedroom-mix constraint and the lower revenue per unit that prevails in the neighborhoods where the constraint applies. A project that pencils on a vacant lot in a wealthier neighborhood may not pencil on a tenant-occupied lot of identical zoning in a lower-income one.

For projects on sites without recent lower-income tenancy, the standard density bonus math the calculator produces reflects the actual project economics. For projects on sites with such tenancy, the density bonus math is a starting point, not the final answer.

What this means for housing policy

The replacement requirement is one of the clearest examples of how California's housing policy framework — a framework that includes density bonuses, ministerial approval, builder's remedy, and significant pro-housing reforms — contains internal contradictions that work against the policy's stated goals. The provision allowing affordable units to satisfy both density bonus and replacement obligations was a thoughtful local accommodation. But the bedroom-matching requirement, combined with geographically variable market-rate rents, re-introduces the perversity the dual-counting provision was meant to soften.

The policy alternative would be to allow replacement units to satisfy bedroom-count obligations through equivalent housing assistance — funding equivalent off-site replacement housing at the required bedroom counts, or replacing demolished bedroom counts in aggregate rather than unit-by-unit. Whether such reforms are politically achievable is a separate question. But the current framework's geographic effects are documentable, and they're the opposite of what the policy was designed to produce.

For the developer evaluating a specific parcel, the takeaway is practical: check for replacement obligations early, and model both the bedroom-mix constraint and the neighborhood's market-rate rent profile into pro forma assumptions. For everyone reading this, the takeaway is analytical: the geography of where new housing gets built in California is shaped substantially by which sites carry bedroom-mix obligations, and where market-rate revenue can absorb them. The pattern that emerges is not the one the policy was designed to produce.