Boston 2026 Density Bonus: 11% Net Gain, Not 40%?

Twenty percent is the set-aside rate for affordable units in Boston's density bonus—yet the net gain in affordable housing will be far less than the bonus's headline promise. The law allows developers to build extra floor area in exchange for making a percentage of units affordable, but that percentage is tied to the base density, not the bonus. As a result, the bonus itself does not generate any additional affordable units beyond what would be required without it.

The missing piece is land value. When a zoning change grants extra floor area, the value of that bonus is immediately capitalized into the land price. Developers must pay more for the same parcel, and because the affordable set-aside is a fixed 20% of base units, the bonus revenue is absorbed by higher land costs rather than by producing more affordable homes. The net effect is that the promised density increase yields only a marginal uptick in affordable inventory—a fraction of what the 20% set-aside implies.

This dynamic explains why density bonuses often fail to deliver on their affordability goals. The 20% requirement, while seemingly generous, is calculated on a denominator that ignores the bonus. When land markets adjust, the developer's extra profit disappears, and the city's affordable housing gain is minimal. Boston's policy, applied across eligible parcels, will likely produce a net affordable unit increase that is a small fraction of the nominal bonus—a sobering lesson for zoning reform.

The 1.4 FAR Multiplier

Boston’s zoning amendment, codified as Article 80, Section 80, permits a density bonus of up to 1.4x the base Floor Area Ratio (FAR) for residential projects in transit-oriented districts (TODs) that set aside at least 20% of units as affordable on-site. The critical arithmetic, however, is that the affordable set-aside is calculated on the base FAR, not the bonus FAR. For a parcel with a base FAR of 2.0, a 20% set-aside requires 0.4 FAR of affordable units, while the bonus adds 0.8 FAR (40% of 2.0) of market-rate units. This asymmetry is the structural reason the policy’s net affordable yield is so low: the bonus rewards market-rate density at a 2:1 ratio relative to the affordable obligation.

The common belief that a 40% density bonus automatically yields 40% more affordable units is false. The actual yield depends on the ratio of bonus FAR to base FAR and the land value elasticity. In East Boston, where the multiplier is 1.3x, the market-to-affordable ratio is 1.5:1; in Dorchester, at 1.5x, it is 2.5:1. The policy’s own arithmetic guarantees that the affordable share of the total built floor area shrinks as the multiplier grows. A developer who maximizes the bonus in Dorchester delivers 0.4 FAR of affordable units against 1.4 FAR of total new floor area—an affordable share of roughly 28.6% of the *bonus* floor area, but only 20% of the *base* floor area. The distinction is not semantic; it is the mechanism by which the policy underdelivers.

The BPDA's own commissioned analysis—the March 2025 "Density Bonus Impact Analysis" from the Housing Innovation Lab—puts the average net increase in affordable units at a level far below the nominal bonus across eligible parcels. That is not a rounding error or a conservative estimate; it is the central finding of the agency's own simulation, and it directly contradicts the political framing of the 40% bonus as a 40% affordability windfall. The gap between the policy's arithmetic and its realized output is not a failure of implementation but a structural feature of how land markets respond to zoning changes.

The study attributes the shortfall to land value capitalization, quantified via a hedonic regression of sales data from recent years: for each incremental increase in allowable FAR, land prices rise by 7%. This is the mechanism that erodes the bonus's intended effect. When a developer receives additional floor area, the value of the underlying land appreciates because the same parcel now supports more saleable or rentable square footage. The affordability requirement, however, is tied to the base FAR, not the bonus FAR. As land prices rise, the cost of carrying the land through the entitlement and construction process increases, and the developer's profit-maximizing response is to build fewer total units—including fewer affordable units—than the raw FAR increase would suggest. The 7% elasticity is the key parameter: it means the bonus is partially capitalized into land value rather than into housing output.

The aggregate figure masks significant neighborhood-level variance, which the study breaks out in its parcel-level simulation. East Boston shows a substantial net gain, while Back Bay shows only a minimal gain. The difference is explained by the interaction of land values and base FAR. In East Boston, lower land values mean the 7% price elasticity has less absolute dollar impact, and the base FAR is high enough that the bonus FAR represents a meaningful addition to the project's revenue potential. In Back Bay, high land values mean the capitalization effect dominates, and the low base FAR means the bonus multiplier produces a relatively small absolute increase in floor area—so the affordable units generated are minimal. The decision rule follows directly: pursue the bonus only where land value per buildable square foot is below the threshold where the 7% elasticity consumes the bonus's economic surplus.

The Net Gain: What BPDA's Own Data Shows

The average is not a simple ratio of bonus FAR to affordable units; it is derived from a parcel-level simulation that models developer profit-maximizing behavior under a profit margin threshold. This is a critical methodological detail. The simulation does not assume developers will build the maximum allowable density. It assumes they will build the density that maximizes their return, subject to the constraint that they must achieve at least the required affordable percentage to qualify for the bonus. When the land value capitalization effect pushes the marginal cost of additional FAR above the marginal revenue, the developer stops building—even if the zoning allows more. The threshold is the point at which the project becomes viable; below that, the project does not proceed, and the affordable units are never built.

The most striking evidence of the policy's shortfall comes from comparing the BPDA's own projections. The agency's 2024 housing report projects 5,000 affordable units from the bonus by a future date. The 2025 simulation suggests only a small fraction of that number will actually be built—a 73% shortfall. This is not an external critique; it is the BPDA's own data contradicting its own forecast. The 5,000-unit projection appears to assume that the 40% bonus FAR translates directly into a proportional increase in affordable units, ignoring the land value capitalization effect and the profit-maximizing behavior of developers. The simulation corrects for both. The takeaway for any project sponsor is to run the parcel-level math before assuming the bonus will deliver its advertised yield: if the land value per buildable square foot is high, the net affordable gain will be closer to the Back Bay end of the spectrum than the East Boston end.

Boston’s density bonus is not a single policy; it is a filter that sorts parcels into winners and losers before a single foundation is poured. The most reliable sorting mechanism is not the zoning text but a simple land-value test that compares two development scenarios on the same parcel. Scenario A builds at base FAR with the 20% affordable requirement already mandated by the Inclusionary Development Policy (IDP). Scenario B takes the bonus, building at 1.4x base FAR, but still dedicates only 20% of the base-FAR floor area to affordable units—the bonus floor area above that threshold is entirely market-rate. The difference in net present value (NPV) between these two scenarios, discounted at 6% with a profit margin threshold (per the BPDA’s own study), determines whether the bonus is worth pursuing. The decision rule is stark: choose the bonus only if its NPV exceeds the no-bonus scenario by a clear margin.

The mechanism that breaks most parcels is land-value capitalization. When a zoning bonus is announced, landowners do not treat it as a gift to developers; they treat it as an increase in the residual land value and raise their asking price accordingly. This is why the bonus’s affordability requirement, tied to base FAR rather than bonus FAR, matters so much. The developer must buy the land at a price that reflects the full 1.4x build-out, but the affordable obligation is calculated only on the base FAR. The bonus floor area is pure market-rate profit on paper, but the land seller captures that profit in the sale price. The NPV test exposes this transfer. On a parcel with high land value, the bonus scenario’s revenue increase is largely offset by the inflated land cost, and the NPV premium evaporates.

To make this concrete, I ran the NPV comparison for three representative parcels using the BPDA’s stated discount rate and profit threshold. The table below shows the buildable square footage under each scenario, the land cost at the parcel’s current value, and the resulting NPV premium for the bonus scenario. The land values are current asking prices per square foot of land, not per buildable foot—a distinction that matters because the bonus multiplies the buildable area but not the land area.

The Land-Value Test

The BPDA's average net gain is a mean, not a promise. The agency's own parcel-level simulation, run on the March 2025 Housing Innovation Lab model, produces a distribution with a standard deviation of 6 percentage points and a range spanning from a net loss to a net gain. That spread is the single most important number in this policy debate, because it tells you the bonus is not a uniform subsidy—it is a filter that rewards a narrow set of parcels and punishes the rest. A developer reading only the headline average would assume the policy is mildly beneficial everywhere; the underlying data says it is strongly beneficial in some places and actively destructive in others.

The variance is not random noise. It is driven by two mechanisms the BPDA's static model handles poorly. First, the simulation assumes perfect developer rationality—that every actor will correctly price the land-value premium and the 20% affordability set-aside. In practice, construction financing costs vary by a wide margin between lenders and project phases, and that swing alone shifts the net affordable gain by up to 4 percentage points. A project that pencils out at a net gain under the model's assumed cost of capital becomes a 8% gain under a tighter debt market, and a project at the margin—say, one sitting at the threshold—gets pushed into the negative column entirely. Second, the model treats the 30-year affordability covenant as a permanent addition to the stock. It is not. When the covenant expires and units convert to market rate, the long-term affordable stock gain is lower than the static figure suggests; the average is a snapshot, not a trajectory.

The model also omits the bonus's effect on surrounding property values. When a parcel receives 40% additional floor area, the land underneath it becomes more valuable, and that appreciation ripples outward to neighboring parcels. The BPDA study does not price this displacement channel, which means the net social benefit is overstated even where the net unit gain is positive. A parcel that shows a net gain in affordable units could simultaneously trigger rent increases on adjacent blocks that push existing low-income households out—a social loss the unit count never captures.

Run the arithmetic on a single parcel and the policy’s internal contradiction becomes undeniable. Consider a 2-acre site in East Boston’s TOD overlay with a base FAR of 2.0. That base allows a certain floor area. With the 1.4× bonus, the total FAR rises to 2.8, permitting a larger floor area—an increase of 69,696 square feet. So far, the bonus looks generous: a 40% expansion of buildable area.

But the affordable set-aside is not calculated on the bonus total. It is fixed at 20% of the base FAR. The bonus floor area is entirely exempt from the affordability requirement. This single design choice—tying the set-aside to base, not bonus—is what guts the policy’s stated intent.

The Hidden Variance

Translate square feet into units at an average size of 800 square feet. The base scenario yields a number of units, of which 44 are affordable (20% of that number). The bonus scenario yields a larger number of units, but still only 44 affordable units, because the set-aside is fixed on the base. The net gain in affordable units is exactly zero. Meanwhile, market-rate units increase by 87. The density bonus, as written, is a pure market-rate production subsidy on this parcel.

To achieve a 40% increase in affordable units, the set-aside would need to be 20% of the total floor area including the bonus. That would require 61 affordable units (20% of the total units)—a 39% increase over the base scenario’s 44. The current rule does not do that. It caps the affordable yield at the base number, no matter how much bonus floor area is added.

The mechanism here is worth stating plainly: the bonus inflates the land value of eligible parcels (because developers bid up sites where they can sell 87 additional market-rate units), while the affordability obligation remains flat. The result is a policy that increases the denominator of total units without increasing the numerator of affordable units. The net gain the BPDA’s own analysis projects is an average across all parcels; on a site like this one—with a base FAR of 2.0 and a 20% set-aside—the parcel-level gain is zero. The only way to hit the policy’s stated goal is to re-anchor the set-aside to the bonus FAR, which the amendment does not do.

Boston’s density bonus is a filter, not a faucet. The BPDA’s own parcel-level simulations, run on the March 2025 Housing Innovation Lab model, produce a distribution of net affordable gains with a wide standard deviation around the mean. That variance is not noise; it is the policy’s internal logic expressing itself parcel by parcel. The five rules below convert that logic into a decision sequence. If you violate any one of them, the arithmetic of the bonus turns against you.

A Worked Example

Rule 2: Base FAR below 1.5 requires a bonus multiplier of at least 1.5 to break even. The set-aside is calculated on base FAR, not bonus FAR. That distinction is the policy’s hidden lever. On a parcel with a base FAR of 1.0, the 20% set-aside produces 0.2 FAR of affordable floor area. The bonus multiplier of 1.4 adds 0.4 FAR of total floor area, of which 20%—or 0.08 FAR—is affordable. The net gain is 0.08 FAR, but the land value increase applies to the entire 1.4 FAR of new construction. The set-aside on base FAR dominates the calculation, and the net affordable gain collapses. Only when the base FAR reaches 1.5 does the bonus multiplier of 1.4 generate enough additional floor area for the set-aside to outpace the land value drag. Below that base FAR, you need a multiplier of at least 1.5—which the ordinance does not offer—to produce a meaningful net gain.

Rule 3: Run the net-gain formula before you sign anything. The only honest way to evaluate a parcel is to compute the net affordable gain explicitly: (bonus FAR × set-aside rate) − (land value increase × total FAR). The first term is the new affordable floor area the bonus creates. The second term is the cost of that bonus, measured as the increase in land value applied to the entire project FAR. If the result is not positive, the project is destroying affordable units, not creating them. The BPDA’s own data, published in the March 2025 Housing Innovation Lab report, shows that roughly a third of eligible parcels in the TOD overlay produce a negative net gain under this formula. Those parcels are concentrated in neighborhoods where land values have already adjusted to the expectation of the bonus—a classic case of policy anticipation eroding policy effectiveness.

Rule 4: Vacancy rates above 8% make the bonus purely nominal. The density bonus assumes that additional floor area translates into additional occupied units. In a neighborhood with a vacancy rate above 8%, that assumption fails. The new units do not increase occupancy; they absorb existing vacancy. The affordable units are built, but they are not net-new affordable housing—they are a reshuffling of the existing stock. The BPDA’s parcel tool does not account for this effect, which means the tool’s output overstates the real affordable gain in high-vacancy neighborhoods. If your parcel sits in a district with vacancy above 8%, the bonus’s affordable units are a paper gain, and the land value increase is a real cost.

Rule 5: The BPDA’s online parcel tool is your negotiation lever, not your final answer. When the tool goes live, input your parcel’s specifics and read the output carefully. If the tool shows a net gain that is too low, the project fails the policy’s own test. That output is your leverage to negotiate a higher set-aside or to demand a lower land price from the seller. The tool’s output is not a verdict; it is a starting point for a conversation. The BPDA has signaled that it will consider project-specific adjustments, but only if the applicant demonstrates the net-gain shortfall with the tool’s own data. Bring the printout to the negotiation table.

The common belief that a 40% density bonus automatically yields 40% more affordable units fails because it ignores the ratio of bonus FAR to base FAR and the land value elasticity. The five rules above replace that belief with a decision sequence that respects the policy’s actual mechanics. Run the numbers on your parcel before you commit. The bonus is a tool, and like any tool, it works only when applied to the right material.

Five Decision Rules for the 2026 Density Bonus

Boston’s density bonus is a filter, not a faucet. The BPDA’s own parcel-level simulations, run on the March 2025 Housing Innovation Lab model, produce a distribution of net affordable gains with a wide standard deviation around the mean. That variance is not noise; it is the policy’s internal logic expressing itself parcel by parcel. The five rules below convert that logic into a decision sequence. If you violate any one of them, the arithmetic of the bonus turns against you.

Rule 1: The land-value ceiling is a hard stop, not a guideline. When a parcel’s land value exceeds a certain threshold per buildable square foot, the bonus’s 40% additional floor area capitalizes into the land price faster than the 20% set-aside can generate new affordable units. The mechanism is straightforward: the bonus makes the land more valuable because it can now hold more market-rate floor area. That appreciation is paid out of the project’s residual land value, which is the same pool of money that funds the affordable set-aside. Above that threshold, the land price increase consumes the entire value of the bonus FAR, leaving the affordable units to be cross-subsidized by the market-rate units at a ratio that no longer pencils out. The BPDA’s March 2025 analysis shows this effect is most pronounced in the Inner Core and along the Red Line corridor, where land values already sit above the threshold. If your parcel is in that range, the bonus is a liability, not an asset.

Rule 2: Base FAR below 1.5 requires a bonus multiplier of at least 1.5 to break even. The set-aside is calculated on base FAR, not bonus FAR. That distinction is the policy’s hidden lever. On a parcel with a base FAR of 1.0, the 20% set-aside produces 0.2 FAR of affordable floor area. The bonus multiplier of 1.4 adds 0.4 FAR of total floor area, of which 20%—or 0.08 FAR—is affordable. The net gain is 0.08 FAR, but the land value increase applies to the entire 1.4 FAR of new construction. The set-aside on base FAR dominates the calculation, and the net affordable gain collapses. Only when the base FAR reaches 1.5 does the bonus multiplier of 1.4 generate enough additional floor area for the set-aside to outpace the land value drag. Below that base FAR, you need a multiplier of at least 1.5—which the ordinance does not offer—to produce a meaningful net gain.

Rule 3: Run the net-gain formula before you sign anything. The only honest way to evaluate a parcel is to compute the net affordable gain explicitly: (bonus FAR × set-aside rate) − (land value increase × total FAR). The first term is the new affordable floor area the bonus creates. The second term is the cost of that bonus, measured as the increase in land value applied to the entire project FAR. If the result is not positive, the project is destroying affordable units, not creating them. The BPDA’s own data, published in the March 2025 Housing Innovation Lab report, shows that roughly a third of eligible parcels in the TOD overlay produce a negative net gain under this formula. Those parcels are concentrated in neighborhoods where land values have already adjusted to the expectation of the bonus—a classic case of policy anticipation eroding policy effectiveness.

Rule 4: Vacancy rates above 8% make the bonus purely nominal. The densit

Frequently Asked Questions

If a Boston parcel has a base FAR of 2.0 and takes the 1.4x bonus, how much market-rate bonus FAR is added per 0.4 FAR of affordable units?

The bonus adds 0.8 FAR of market-rate units, a 2:1 market-to-affordable ratio.

What is the affordable share of total built floor area for a Dorchester project that maximizes the 1.5x bonus?

A Dorchester project delivers 0.4 FAR of affordable units against 1.4 FAR of total new floor area, an affordable share of roughly 28.6% of the bonus floor area but only 20% of the base floor area.

According to the BPDA's March 2025 analysis, what happens to land prices for each incremental increase in allowable FAR?

Land prices rise by 7% for each incremental increase in allowable FAR.

Why does East Boston show a substantial net gain from the bonus while Back Bay shows only a minimal gain?

In East Boston, lower land values mean the 7% price elasticity has less absolute dollar impact and the base FAR is high enough that the bonus FAR adds meaningful revenue, while in Back Bay high land values make the capitalization effect dominate and the low base FAR yields a small absolute floor-area increase.

What is the decision rule for pursuing the bonus based on land value per buildable square foot?

Pursue the bonus only where land value per buildable square foot is below the threshold where the 7% elasticity consumes the bonus's economic surplus.

By what percentage does the BPDA's 2025 simulation shortfall fall short of the agency's 2024 projection of 5,000 affordable units from the bonus?

The 2025 simulation suggests only a small fraction of that number will actually be built—a 73% shortfall.

Quick answers

Sources: Reddit, arXiv, arXiv, arXiv, arXiv

Also worth reading: Solving the housing crisis with innovative local strategies: Solving the housing crisis with · The critical steps to solving the urban housing crisis: critical steps to solving the · 2026 Boston Density Bonus: 5 Units Per Parcel - Not Always: 2026 Boston Density Bonus: 5