The Highlands at Mantua

Project description

The subject parcel is Tax Map Parcel 58-2-10-001A1, comprising 2.12 acres. The site currently contains the Mantua Professional Center, described in project documents as underutilized office buildings of one to three levels totaling approximately 21,566 square feet. No current assessed value for the parcel is available in the project record.

The applicant proposes to replace the existing office buildings with up to 14 townhouses. Per the most recent project narrative (Revised April 6, 2026), the units will be three levels with an option for a roof deck, and building height will not exceed 38 feet. The June 2026 Master Development Plan tabulation shows 62 total parking spaces — 28 garage and 34 surface spaces, of which 32 are shared. The April 6, 2026 narrative states approximately 48% open space overall. Tenure (for-sale vs. rental) is not established in the project documents, and the number of affordable units is not specified in the reviewed materials. The project is currently under review by the City of Fairfax.

Note on document history: earlier submissions described four-level buildings at heights of approximately 43 to 47 feet and as many as 81 parking spaces; the most recent submission governs all figures used in this analysis.

Economic effects

New demand. The 14 proposed units, applied against an occupancy rate assumption, yield an estimated 12.6–13.58 new households. Those households, multiplied by persons-per-multifamily-unit multipliers from the Rutgers CUPR residential demographic tables, produce an estimated 18.9–29.9 new residents. Aggregate annual household income for the new residents is estimated using the mean income of the site census tract (tract 51600300300, ACS 5-year 2024, series B19025/B11001, at $163,796) scaled by a new-construction income premium assumption, yielding a range of approximately $2.06 million to $2.78 million per year.

New annual consumer spending by category — estimated by applying BLS Consumer Expenditure Survey 2023 line items to the new households with Engel-curve income scaling per category — breaks down across grocery (food at home), restaurants and bars (food away from home), comparison retail (apparel and services, household furnishings and equipment), convenience retail (personal care products), personal services, and entertainment. Ranges for the top categories are: grocery $75,654–$121,969/yr; restaurants and bars $63,106–$111,237/yr; comparison retail $78,227–$141,003/yr; and entertainment $64,144–$118,228/yr.

Where spending is likely to land — screening estimates. The Huff capture figures below are screening estimates. They are computed by a joint destination-and-mode choice model applied to individual retail points of interest (using the method: P(j,m) ~ A_j × w_m × exp(−b_m t_mj), with CES category spend, aggregated to named clusters for reporting). They rank where new resident spending is likely to land given travel time and destination attractiveness, with sensitivity bounds — they are not predictions of actual sales.

The top five receiving clusters, by central estimated annual capture, are:

Within the city boundary, the in-city capture share is estimated at 0.534–0.581 across all retail categories, and 0.577–0.654 for the food-away-from-home (restaurant and bar) category specifically. Both are screening estimates subject to the same model sensitivities.

Project's own retail capture. The project includes no proposed ground-floor retail square footage. The own-retail capture estimate is therefore $0/yr (low $0, high $0). No on-site retail spending is modeled.

Foot-traffic index. The foot-traffic index change across the 10 nearest commercial street segments is estimated at 0% (relative index). This figure is a screening estimate computed by allocating new residents' modeled walk trips as exact marginal flows against a sampled baseline betweenness index of today's population; the baseline is a sampled index, not calibrated pedestrian counts.

Jobs ledger. The existing approximately 21,566 square feet of office space supports an estimated 43.1–86.3 on-site jobs, computed as existing commercial square footage divided by an assumed square-footage-per-office-job (range 250–500 sqft/job, central 350). Because no retail or office space is proposed, on-site retail jobs added is 0 and office jobs added is 0, computed as proposed retail and office square footage divided by the respective sqft-per-job assumptions. The net on-site job change is therefore estimated at −43.1 to −86.3 jobs — a reduction driven entirely by the removal of the existing office buildings.

Fiscal effects

Current real estate tax. The site's current real estate tax, computed as current assessed value times the applicable rate divided by 100, is $36,490/yr (this figure has no range; it is drawn from the City of Fairfax Real Estate Assessment Database, 2026, for parcel 58-2-10-001A1).

Projected real estate tax. The projected assessed value, computed as proposed units times assessed value per unit (using three apartment-class comparables from the City assessment database: a 400-unit property at $360,982/unit built 2020, a 403-unit property at $359,599/unit built 2022, and a 268-unit property at $458,932/unit built 2022), yields a projected assessed value range of $5,034,390–$6,425,054. Applying the FY2027 real estate tax rate produces a projected real estate tax of $53,994–$68,909/yr. The real estate tax increase — projected minus current — is estimated at $17,504–$32,419/yr.

Other recurring revenue lines. Three additional revenue streams are included:

• Personal property tax on resident vehicles (rough estimate): computed as new households times assumed vehicles per household times assumed vehicle assessed value times the personal property rate divided by 100. No project-specific vehicle data exists; this is treated as an order-of-magnitude figure. Range: $4,683–$20,191/yr.
• Meals tax on captured in-city dining: computed as restaurant spending times the in-city capture share times the meals tax rate, using the Huff model cross-module output. Range: $1,637–$3,273/yr.
• Local sales tax share on captured in-city retail: computed as in-city captured retail spend times the local-option sales tax share (Va. Code § 58.1-605). Range: $894–$1,659/yr.

Service cost — two methods. This analysis estimates service costs using two distinct framings, which produce different results by design:

The naive per-capita method allocates the full citywide General Fund per-capita cost (non-school: ($207,912,496 − $76,429,791) ÷ 25,026 residents = $5,254 per capita, FY2027 adopted) to each new resident, plus estimated students times the net local cost per pupil ($19,961, computed as the $76,429,791 FCPS tuition contract less $14,492,271 in state education revenue, divided by 3,103 students). This method — residents times non-school General Fund per capita plus estimated students times per-pupil tuition — produces an annual service cost of $113,270–$190,498. It is an upper-bound framing because it allocates fixed citywide overhead to new infill residents.

The marginal framing applies a marginal cost factor to the non-school per-capita figure, on the basis that fixed services (roads, administration) do not scale proportionally with new residents in an infill context, while school costs follow the project's own student estimate directly. The method is: non-school per-capita cost times the marginal factor plus students times per-pupil tuition. This produces an annual service cost of $38,797–$119,864. It is a lower-bound framing because it may understate costs if new residents trigger capacity expansions.

Estimated K-12 students. The student estimate, computed as units times students per unit, yields a central estimate of 1.12 students (range 0.70–1.68), driving the school-cost component identically in both framings.

Net annual fiscal impact. The two cost methods produce different net outcomes:

• Naive per-capita method: −$165,778 to −$55,728/yr (net: recurring revenue minus upper-bound service cost).
• Marginal framing: −$95,144 to +$18,745/yr (net: recurring revenue minus marginal service cost).

Spanning both methods, the full net annual fiscal impact range from this analysis is −$165,778 to +$18,745/yr.

Comparison with external (city staff) estimates. Two sets of city staff estimates are recorded in the project file and reported here as published:

• February 25, 2025 City Council Post-Submission Work Session (Bryan D. Foster, City Manager, and Albert Frederick, Senior Planner): "Staff estimate that this proposal would net a positive fiscal impact of between $24,000 and $72,000 annually with an average of $48,000."
• Council packet: Staff Analysis (Z-24-00236), Page 8 (Albert Frederick, Planner III; July 27, 2026 Planning Commission Public Hearing): "Staff estimates that this proposal would net a positive fiscal impact of between $32,000 and $91,000 annually with an average of $61,500."

This analysis's net range (−$165,778 to +$18,745) does not overlap with either staff estimate. The staff estimates fall entirely in positive territory; this analysis's range is predominantly negative and reaches positive values only under the marginal cost framing at its favorable bound. Several methodological differences visible in the metric methods and notes are plausible sources of this divergence: the staff estimates do not publish their cost methodology, so the cost-framing choice (per-capita vs. marginal vs. a different basis entirely) cannot be compared directly; the staff estimates may use different household-size or student-generation multipliers, both of which are top-sensitivity assumptions in this analysis; and the personal property tax rough estimate, which this analysis carries as a wide-range line, may be treated differently. The external figures are reported as published and are not averaged with or otherwise combined with the figures from this analysis.

Not evaluated in this version

Several analyses are deferred from this version of the report and were not computed. Bike-lane corridor effects, trail connectivity, broader multimodal connectivity analysis, environmental and stormwater impacts, and a comparable-places analysis have not been evaluated. These topics may be addressed in future versions as additional data and analysis are completed.