Northfax Chain Bridge

Project description

The site consists of five parcels (PINs 57 2 02 024B, 57 2 02 024A, 57 2 02 021, 57 2 02 023, and 57 2 02 022) totaling 2.30 acres at the northeast corner of Chain Bridge Road and Fairfax Boulevard. Current uses are described in project documents as a gas station, a retail outlet, and a vacant or former restaurant (Fuddruckers).

The proposal, as described in the September 11, 2023 applicant overview and concept plan, is for 271 multifamily rental dwelling units in an 8-story building with five levels of stick-frame construction above a three-level concrete podium base. The building would include 2,253 square feet of ground-floor commercial space; an applicant document notes that increasing the retail to "as much as 3,000 SF" has been explored, but 2,253 square feet is the figure consistent across the authoritative project documents and is the figure used in this analysis. Total structured parking is stated as 316 spaces within 5 levels; a separate applicant document describes 305 of those spaces as reserved for residents and 11 dedicated to retail, plus 12 off-site parallel spaces along Northfax Street—it is ambiguous from the documents whether the 316 figure includes or excludes the 12 off-site spaces. The applicant states that parking would be unbundled from standard unit rental. Six percent of units—16 units—are committed to affordable housing. The project is in pre-application status; no formal application or plat has been filed.

Economic effects

New demand. Applying a stabilized occupancy rate assumption of 0.95 (range 0.90–0.97) to the 271 proposed units, the project is estimated to add 243.9–262.9 new households, computed as proposed units multiplied by the occupancy rate. Multiplying households by a persons-per-multifamily-unit assumption drawn from Rutgers CUPR residential demographic multipliers (Listokin et al. 2006) yields an estimated 365.9–578.3 new residents.

Spending clusters. New household spending was estimated by category using BLS Consumer Expenditure Survey 2023 line items scaled by household count and an Engel-gradient income adjustment, with each category's elasticity applied to an assumed new-construction income premium (range 1.00–1.25 over the site tract mean income of $195,124 per the Census ACS 5-year 2024 data). The five spending categories modeled are grocery (food at home), restaurants and bars (food away from home), comparison retail (apparel and services plus household furnishings and equipment), convenience retail (personal care products), personal services, and entertainment.

Where spending lands — screening estimates. The Huff model results reported here are screening estimates. They are computed by applying a joint destination-and-mode-choice model over individual retail points of interest, with a walk-time decay parameter of 0.10 per minute (range 0.05–0.15) and zero-impedance walk preferences that vary by category. Clusters are a reporting rollup of individual POI results. The model ranks where new resident spending is likely to land within the sensitivity bounds shown; it is not a prediction.

The five named retail clusters and their estimated annual spending capture from new residents are:

• Fairchester Drive & Hill Street: $2,094,000–$4,042,000/yr
• Chain Bridge Road & North Street: $953,000–$1,969,000/yr
• Fair Lakes Parkway & Fair Oaks Mall: $644,000–$1,002,000/yr
• Blenheim Boulevard & Fairfax Boulevard: $393,000–$667,000/yr
• Mathy Drive & Pickett Road: $340,000–$498,000/yr

The Huff model estimates that 57.7%–65.6% of food-away-from-home spending and 57.7%–65.6% of all-retail spending would be captured at destinations inside the City of Fairfax boundary, computed as the share of Huff-model capture at destinations inside the city.

Project's own ground-floor retail capture — screening estimate. The 2,253-square-foot ground-floor retail space competes as a destination in the same Huff model, sized using an assumed square-footage-per-equivalent-POI of 2,000 (range 1,500–3,000). The model estimates that the project's own retail would capture $5,100–$11,800 per year of the new residents' spending, across restaurant-and-bar, comparison retail, convenience retail, and personal-services categories. Of that, $1,557–$8,013 per year is estimated to arrive on foot, computed as the walk-mode share of the joint Huff choice at the own-retail destination.

Foot-traffic index — screening estimate. The foot-traffic index change for the 10 nearest commercial street segments is estimated at 0.1% (relative index), computed by allocating exact marginal trip flows from the site—using POI-weighted destinations and exponential walk-time decay along shortest paths—against a seeded sampled betweenness index of today's population. The wide low-high range on the underlying walk-trip input (0.5–2.0 trips per resident per day) and the walk-time decay parameter mean this figure carries substantial uncertainty. The percentage comparison uses a sampled index of today's walkers, not calibrated pedestrian counts.

Annual spending arriving on foot at all nearby businesses is estimated at $127,500–$3,858,300 per year, computed as the walk-mode share of the joint destination-and-mode Huff choice summed over businesses. The corresponding estimate for spending arriving by bike is $6,000–$660,200 per year, computed as the bike-mode share carved from the drive remainder under the same model. Both figures carry wide ranges driven by the walk-share and bike-share assumptions.

Jobs. The existing commercial space on the site has no documented square footage in the project records; the on-site jobs removed estimate is therefore 0, computed as existing commercial square footage divided by a square-footage-per-office-job assumption—because the existing square footage input is zero, no displaced jobs are estimated. This should not be read as a finding that no jobs currently exist on site; it reflects a data gap. The proposed 2,253 square feet of retail space is estimated to support 3.2–5.6 on-site retail jobs, computed as proposed retail square footage divided by a square-footage-per-retail-job assumption (range 400–700 square feet per job). The net on-site job change is estimated at 3.2–5.6 jobs, equal to retail jobs added plus office jobs added minus jobs removed; no office square footage is proposed.

Fiscal effects

Current taxes. The five subject parcels currently generate $62,128.85 per year in real estate tax, computed as current assessed value multiplied by the city real estate tax rate, using assessment data from the City of Fairfax Real Estate Assessment Database (Patriot WebPro, 2026 vintage) and the FY2027 adopted rate.

Projected real estate tax. The projected assessed value of $97,902,000–$125,272,000 is computed as units multiplied by an assumed assessed value per unit, plus retail square footage multiplied by an assumed commercial value per square foot, plus office square footage multiplied by an assumed office value per square foot. The residential value per unit assumption is drawn from three apartment-building comparables in the city assessment database, all built since 2011: parcel 48 3 02 011A (2020, 400 units, $360,982/unit), parcel 57 1 02 031C (2022, 403 units, $359,599/unit), and parcel 57 2 20 006A (2022, 268 units, $458,932/unit); the range spans the 25th–75th percentile of those three comps. The commercial value per square foot is a screening range of $200–$400 per square foot. Applying the city real estate tax rate to the projected assessed value produces a projected real estate tax of $1,050,000–$1,343,500 per year, and a real estate tax increase over the current site of $987,900–$1,281,400 per year.

Other recurring revenues. Additional revenue lines estimated are:

• Personal property tax on resident vehicles (rough estimate): $90,700–$390,800 per year, computed as new households multiplied by assumed vehicles per household multiplied by assumed average vehicle assessed value multiplied by the city personal property rate. This is a rough estimate; no project-specific vehicle data exist.
• BPOL business license tax on project retail (rough estimate): $166–$1,343 per year (net-new basis), computed as proposed retail square footage multiplied by assumed gross sales per square foot multiplied by the BPOL retail rate, multiplied by a net-new share assumption (range 0.15–0.50) that adjusts for receipts that would be displaced from existing city businesses.
• Business tangible property tax (rough estimate): $372–$1,396 per year, computed as proposed commercial square footage multiplied by assumed equipment value per square foot multiplied by the city business personal property rate.
• Meals tax on captured in-city dining: $39,200–$81,200 per year, computed as restaurant spending multiplied by the in-city capture share from the Huff model multiplied by the city meals tax rate.
• Meals tax on the project's own restaurants (rough estimate, net-new): $1,123–$21,151 per year, computed as ground-floor square footage multiplied by assumed sales per square foot multiplied by an assumed restaurant share of retail space (range 0.30–0.70) multiplied by the meals tax rate and the net-new share.
• Local sales tax share on captured in-city retail: $21,100–$41,100 per year, computed as in-city captured retail spending multiplied by the local sales tax share under Va. Code § 58.1-605.
• Local sales tax on the project's own retail (rough estimate, net-new): $832–$6,715 per year, computed as ground-floor square footage multiplied by assumed sales per square foot multiplied by the local sales tax share and the net-new share.

Service costs — two methods. Two cost framings are applied, and both are reported because each represents a legitimate but different question:

The naive per-capita method allocates all General Fund costs proportionally: it multiplies new residents by the non-school General Fund per-capita cost ($5,254, derived from FY2027 adopted budget data as general fund expenditures net of the FCPS tuition contract, divided by 25,026 residents) and adds an estimated school cost of $270,500–$649,100 per year (computed as estimated K-12 students multiplied by the net local cost per pupil of $19,961, itself derived as the FCPS tuition contract of $76,429,791 less state education revenue of $14,492,271 divided by 3,103 students). The school cost calculation uses an estimated 13.6–32.5 K-12 students from the project, computed as units multiplied by an assumed student-generation rate of 0.05–0.12 per unit. Total cost under this framing is $2,192,600–$3,687,500 per year. This framing overstates costs for infill development because it allocates fixed citywide costs—roads, administration, and other services that do not scale with incremental new residents—as if they were fully variable.

The marginal-cost method applies a marginal cost factor of 0.25–0.55 (central 0.35) to the non-school per-capita cost, on the premise that fixed services do not scale with infill residents, while school costs continue to follow the student estimate directly. Total cost under this framing is $750,900–$2,320,200 per year. This framing may understate costs if the project triggers capacity expansions in services that are currently at or near capacity.

Net annual fiscal impact. Combining all revenue lines with both cost framings, the net annual fiscal impact ranges from −$2,546,000 to +$1,074,000 per year across both methods. Under the naive per-capita method alone, the range is −$2,546,100 to −$367,500 per year (negative throughout). Under the marginal-cost method alone, the range is −$1,178,900 to +$1,074,100 per year (spanning both sides of zero). The revenue side of both calculations includes the rough-estimate personal property and BPOL lines; leaving them at zero would understate revenue for taxes the city does levy, but both carry wide assumption-driven bounds.

No external applicant or staff fiscal estimates are included in the project documents available to this analysis, so no comparison to external published figures is possible at this time.

Not evaluated in this version

Several analyses have been deferred and are not part of this version of the report. Specifically, a bike-lane corridor analysis, trail connectivity assessment, broader multimodal connectivity evaluation, environmental and stormwater impact analysis, and a comparable-places (precedent project) analysis have not been conducted. These topics may be addressed in subsequent versions as the project advances through the application process and additional documentation becomes available.