← back to the tool · all reviewer documents

This is a working document from the project, published verbatim for reviewers. File paths in it refer to the project workspace.

Does Development Pay for Itself?

Howard County Fiscal Impact Model, Assumptions Working Document

v0.2 (2026-08-10). Fills Jeremy Dommu's v0.1 draft (2026-08-10) with sourced values. Prepared for The Merriweather Post / Eric DeMenthon collaboration.

Working rules (Eric, 2026-08-10): every value carries a source tag [S#] resolving in the Sources appendix, which states how to verify it manually. No value is invented: if no source was found, the cell says OPEN and stays open for discussion.

Status legend - VERIFIED: we read the primary document ourselves (page cited). - REPORTED: a research pass confirmed it against the cited URL; not yet re-read by a human. Promote to VERIFIED by following the how-to-verify step in the appendix. - DERIVED: computed from VERIFIED/REPORTED inputs; the formula is shown. Judge the formula, not just the inputs. - OPEN: no source yet. Do not use.


1. Scope (decided, per v0.1)

Unchanged from Jeremy's draft: single discrete project; full stabilization in Year 1 (no absorption curve, no phasing); one-time revenues booked in Year 1; annual net held flat across a 30-year window; current law only (SOUP and other pending proposals are future toggles, not defaults).

v0.2 proposes three scope additions, argued in section 9: a discounting decision (D5), a resale-revenue decision (D6), and a fund-scope decision (D7). None are settled here.

2. Student count methodology (Years 1-30)

Assumption Value / Status Source
Year 1 yield (new construction), countywide SFD 0.7581, SFA/TH 0.4112, rental apt 0.1447, condo apt 0.1447 students/unit. By school level and planning area: full table in section 8.7. VERIFIED, but 2015-17 construction vintage; freshest public table that exists [S1] p. 59, Table A-10 (HCPSS Office of School Planning, 2018)
Years 2-30 yield (resale/turnover) OPEN. EFMP confirms the two-stage convention exists (new-construction yields in first occupancy year, then folded into resale/turnover pools) but publishes no numeric table. Ask HCPSS OSP: Daniel Lubeley, Timothy Rogers; records request if needed [S8] pp. 40-41 (convention only)
Apartment treatment v0.1 simplification kept: one flat rate from Year 1. Note the A-10 apartment rate is itself a new-construction (2015-17 units) figure [S1] p. 59
Student trend over 30 yrs Held flat (v0.1 decision). Known bias: county births/enrollment declining, so flat overstates long-run students. See bias ledger [S8]
Yield by rent tier / bedrooms OPEN as a tier split; A-10 has no rent or bedroom breakout. Available proxy: planning-area column (rental apt 0.0508 Columbia vs 0.3400 Ellicott City, 6.7x spread) [S1] p. 59
Age-restricted (55+) yield OPEN for the exact HCPSS/DPZ convention. Supporting fact: county planning uses 1.70 persons/household for age-restricted units [S1] p. 58, Table A-8

3. Revenue, one-time (Year 1)

Assumption Value / Status Source
County transfer tax 1.25% of consideration (since 7/1/2020). REPORTED. Fully earmarked: 25% school land/construction, 25% parks, 20% ag preservation, 15% housing/community development, 15% fire. It is county revenue but never General Fund revenue: see D7. First-time-buyer carve-outs: OPEN [S2][S11]
State transfer tax 0.5% (0.25% first-time buyer). REPORTED. State revenue, so excluded from a county-fisc model; listed to avoid double counting in "2.25% total deed tax" figures [S11]
Recordation tax $2.50 per $500 of value (0.5%). REPORTED. Note: the county's 2019 model treats recordation as RECURRING operating revenue (resales + refinancings), not one-time: decision D6 [S2]; [S1] p. 55
School facility surcharge $8.63/sq ft regular, FY27 schedule effective 7/1/2026. Reduced rates: non-affordable senior $1.32; MIHU on-site outside Downtown Columbia $2.88; Downtown Columbia affordable $4.98; state/county-funded affordable non-senior $2.88. REPORTED. Resolves v0.1 flag: CPI-indexed annually per CB42-2019, resets July 1; Jeremy's $8.15 was the prior-year (FY26) rate. Implication of A-4 sizes: a new countywide-average SFD (5,465 sf) pays ~$47k [S4]; indexing [S4]; sizes [S1] p. 56
Road (multimodal) excise tax $1.96/gross sq ft residential and office/retail; $1.00 distribution/manufacturing and institutional. County Code 20.503, CR 94-2026, FY27 schedule. REPORTED [S5]
Building permit / plan review (DILP) Residential SF new construction $0.25/gross sq ft; multifamily and other new construction $0.30/sq ft; filing fee $100 residential / $200 commercial; +10% technology fee on all permits; $50 guaranty fund fee on MD Home Builder residential applications. FY27 schedule (CR 83-2026), effective 7/1/2026. VERIFIED. Confirms v0.1 lean: minor (a 2,400 sf SFD pays ~$770 vs ~$20.7k surcharge) [S19]
MIHU fee-in-lieu (new line, v0.2) $3.77 per sq ft of residential space for each unit when a development pays in lieu of building on-site MIHUs (Code 13.402C; CR 97-2026; ENR-Baltimore indexed annually; FY27, eff 7/1/2026). VERIFIED. Affordable housing now has three modeled paths: market-rate; on-site MIHU (reduced surcharge $2.88/sf + lower unit values per A-6); or fee-in-lieu (this line + full market values) [S20]
Water/sewer connection charges OPEN on fund status. Known: water/sewer is the Metropolitan District (separate $0.08/$100 ad valorem inside the district), an enterprise-style arrangement, supporting the v0.1 lean to exclude. Confirm connection-charge fund before deciding (D3) [S3]

4. Revenue, recurring (annual)

Assumption Value / Status Source
Real property tax County General Fund $1.044 per $100 AV (VERIFIED) + Fire Fund $0.206 (REPORTED; separate special fund, confirming v0.1's suspicion: see D7) + state $0.112 (REPORTED, state revenue, exclude). Sensitivity: 1 cent on the county rate = ~$6.5M/yr countywide [S2][S3]
Assessment mechanics SDAT triennial reassessment; increases phase in over 3 years, decreases immediate; Homestead cap 5% county / 10% state, owner-occupied only (rentals uncapped); apartments/commercial valued by income approach. REPORTED. Under a constant-dollar model (D5) most of this matters only if real appreciation is assumed [S10]
County income tax 3.2% of Maryland taxable income, flat, no brackets (VERIFIED via county FAQ; v0.1 had it from budget.dementhon.com). Household income by unit type: OPEN as a current value; candidate methodology exists: the 2019 county study derived income from mortgage qualification at the unit's price point (2018 values in section 8.4; renter analog needed). This remains the biggest recurring-revenue input to settle [S2]; method [S1] pp. 57-58
Recordation, recurring portion Decision D6. County 2019 model books recordation per housing unit as an annual operating revenue line [S1] p. 55
State aid to county / highway user OPEN. 2026-08-10 scout pass found nothing citable (these line items live in PDF budget documents scouts cannot parse); next source: the county government FY26 Approved Operating Budget revenue schedules (note: distinct from the HCPSS budget book we hold). Also note for later refinement: Blueprint education aid keys to county wealth per pupil, so high-value development slightly reduces state aid share across all pupils (second-order, candidate footnote toggle, not v1) none yet
Energy tax / franchise fees CLOSED as a per-household line: no county utility, energy, or telephone tax found; the county's own household tax/fee enumeration lists none. Cable franchise fees exist but are paid by the cable operator (Local Franchising Authority arrangement), not billed to households. REPORTED (negative finding; scout pass 2026-08-10) [S21]
Flat household fees Trash $391/yr full service; Watershed Protection Fee $40 (TH/apt/condo) to $115-$225 (SFD by lot). REPORTED. Only include if the matching services are included on the cost side (D3): fee-funded pairs should net out [S3][S13]
General fund context GF is ~47% property tax + ~45% income tax (~92% combined); FY26 GF total $1.63B. REPORTED. Useful as a sanity anchor: the model's revenue lines should reproduce roughly this mix for a representative unit [S9][S17]

5. Expenses, education

Assumption Value / Status Source
HCPSS cost per pupil $21,640 budgeted per-pupil expenditure FY26 ($120.22/day). VERIFIED (budget book p. 14). This is the ALL-funds figure (state + county + federal), not the county's marginal cost [S7] p. 14
County share of HCPSS County appropriation $816.005M FY26 (FY25: $766.0M, +6.53%). VERIFIED (p. 39). State aid $375.4M REPORTED [S7] p. 39; [S6]
County appropriation per pupil DERIVED: $816.005M / 57,031 enrolled = ~$14,308/pupil/yr. Formula shown; whether this average is the right cost base is exactly decision D1 (marginal vs average) [S7] p. 39; enrollment [S8]
Enrollment counting convention New construction counted in first occupancy year at new-construction yields, then folded into resale/turnover/pre-K pools. VERIFIED by Jeremy against EFMP (primary) [S8] pp. 40-41
School capital cost per seat DERIVED benchmark: Guilford Park HS (opened 2023) $129M / 1,650 seats = ~$78k/seat. REPORTED inputs; a subcontractor page says $104M (likely construction contract vs total project): use $129M total-project until primary CIP doc is pulled. County 2019 model charged school debt service PER SEAT BY CAPACITY (capacity-sensitive), other debt per capita [S12]; method [S1] p. 55
State share of school construction State cost share for Howard County: 54% (FY2025), 51% (FY2026) of ELIGIBLE costs per COMAR 14.39.02.06 ("Howard 54% 51%"). Eligible is capped by formula (COMAR 14.39.02.07): lesser of gross-area baseline (approved enrollment x IAC baseline sq ft/student) or actual area, times the statewide average per-sq-ft school cost, plus site development, times the share %. REPORTED (scout pass; promote via official COMAR at dsd.maryland.gov). Note (commentary, not sourced): eligible per-sf baselines typically run below actual HoCo costs, so the effective state share of TOTAL project cost is lower than 51%. Actual Guilford Park split: still OPEN (scout found nothing; likely lives in the county capital budget PDF) [S22]
Age-restricted, revenue side Do not model seniors as pure cost savings: senior surcharge rate is $1.32/sf (vs $8.63) and the county offers a 25% senior (65+, income-limited) credit and 20% Aging-in-Place credit (to $650k AV) on county property tax. REPORTED. Symmetric honesty: senior projects have lower school costs AND lower revenues [S4][S16]

6. Expenses, public safety

Assumption Value / Status Source
Police cost per capita OPEN pending D2 (allocation method). Population denominator now REPORTED: ~339,183 (Census Vintage-2025 estimate, July 1 2025); persons/household 2.76 (ACS 2019-23). Ingredients in hand: FY24 approved budget by department/fund/object (local data, [S18]); FY26 GF total [S17] [S18][S23]
Fire / EMS Fire is a separate special fund with its own $0.206 levy (REPORTED), so v0.1's flag was right: under an all-funds scope (D7), fire costs net against fire-fund revenue rather than loading the General Fund [S2][S3]
Judicial / detention OPEN. v0.1 lean (immaterial per capita) plausible, unverified none yet

7. Expenses, public works and general government

Assumption Value / Status Source
Road maintenance OPEN pending D2. Ingredient: DPW lines in [S18]. Note many arterials are state (SHA) roads and many subdivisions have private/HOA streets: allocation needs care [S18]
Solid waste / recycling OPEN on fund status; trash fee ($391) exists, so leaning fee-funded and excluded net (D3) [S3]
Stormwater OPEN on fund status; Watershed Protection Fee exists, same netting logic (D3) [S13]
General government overhead OPEN pending D2. v0.1 candidate ((GF minus itemized) / population) is an average-cost method; fine for v1 if labeled as such. Local quirk to handle: the Columbia Association privately funds much rec/parks inside Columbia [S18]; CA quirk: discuss

8. Supporting estimate tables (county 2019 study, constant 2018 dollars)

All VERIFIED against [S1] (pages cited), but 2015-2018 vintage: usable as defaults only with an "as-of 2018, refresh pending" label and a slider. Countywide values; per planning-area detail is in the study.

  1. Unit sizes (A-4, p. 56): new SFD 5,465 sf; SFA 2,586; apt 1,458.
  2. Unit values (A-5, p. 57): SFD $802,149; SFA $511,429; rental apt $198,000/unit; condo $340,671.
  3. MIHU values (A-6, p. 57): SFA $311,000; rental $161,000; condo $246,000 (no SFD MIHU). This is the affordable-vs-market input Jeremy's spec wants.
  4. Homeowner income needed (A-7, p. 58): SFD $178,748; SFA $115,081; condo $77,548; (renter figure $47,134 listed). Mortgage-qualification method; the candidate answer to the income-by-unit-type gap.
  5. Household size (A-8, p. 58): SFD 3.16 (2025 col.); SFA 2.63; apt 2.05; age-restricted 1.70.
  6. Occupancy (A-9, p. 59): SFD 98%, SFA 97%, apt 96%.
  7. Student yields (A-10, p. 59): full table, students per unit by school level:
Type Level Countywide Columbia Elkridge Ellicott City Rural West Southeast
SFD ES 0.4863 0.5000 0.4563 0.5493 0.5735 0.3778
SFD MS 0.1769 0.1923 0.1688 0.1910 0.2322 0.1212
SFD HS 0.0949 0.0385 0.1063 0.0806 0.1090 0.0949
SFD Total 0.7581 0.7308 0.7314 0.8209 0.9147 0.5939
SFA Total 0.4112 0.3421 0.3751 0.6401 n/a 0.3334
Rental apt Total 0.1447 0.0508 0.1834 0.3400 n/a 0.1834
Condo apt Total 0.1447 0.0508 0.1834 0.3400 n/a 0.1834

(SFA/apt level splits are in the study; Southeast apartments borrow the Elkridge average per study note. Yields from units built 2015-2017.)

  1. The 2019 study's own per-unit results (Table 2, p. 7), the cross-check target: annual operating net: SFD +$5,047.87, SFA +$3,520.74, rental apt +$941.94, condo +$4,133.70. One-time capital revenue: SFD $21,484.35, SFA $11,378.86, rental $5,588.00, condo $6,957.04. Ratio form: $1.38 / $1.45 / $1.25 / $2.14 of operating revenue per $1.00 of services. Constant 2018 dollars, pre-PAYGO adjustment. v0 of our model should reproduce these numbers from the study's own inputs before we swap in current values.
  2. 2025-26 market anchors (REPORTED, scout pass 2026-08-10). All-transaction figures, NOT new-construction: current new-unit values remain OPEN, these bound them from below. Countywide median sold price $634,500 (May 2026, +5.8% YoY, Howard County Assn of REALTORS); median $675,850 / average $760,154 (June 2026, Maryland REALTORS via secondary cite, weaker). Columbia rents (Zumper, Aug 2026): average apartment $2,255/mo, median $2,538, 1BR $1,905, 2BR ~$2,226-2,260. Compare A-5's 2018 new-SFD value $802k: new construction carries a large premium over the resale median, so do not substitute these medians for new-unit values.
  3. The study's multiplier map (p. 55), i.e. its model skeleton: transfer + surcharge
  4. road excise = one-time capital revenue per unit; property tax per unit; income tax per unit by household income; recordation per unit RECURRING; fire fund per unit; all other revenue per capita. Costs: school debt service per seat by capacity; other debt service per capita; school operating per unit by yields; all other operating per capita.

9. Open methodology decisions

D1-D4 are Jeremy's; D5-D9 added in v0.2. None decided.

10. Known biases ledger (v1 simplifications and their direction)

"Direction" = effect on the modeled answer to "does development pay for itself?"

Simplification Direction Size guess
Flat student count, ignoring countywide enrollment decline Against development (overstates school cost) material over 30 yrs
Year-1-only deed taxes (if D6 stays v0.1) Against development (understates revenue) material for owner units
No absorption curve, full Year-1 stabilization For development (front-loads revenue) small, short-lived
Ignoring senior credits + reduced surcharge on age-restricted For development (overstates senior-project revenue) material for senior projects only
Excluding induced commercial-base growth (economic spillovers) Against development small (~1% order; MD routes sales tax to the state, income tax by residence)
Excluding congestion / service-quality externalities For development unquantified, qualitative
Average-cost school mode where slack capacity exists Against development large; D1 dual-view exists for this
Marginal-cost mode when a project trips a capacity threshold For development large; same remedy
2018-vintage estimate tables used as defaults Direction unknown (values and yields both moved) refresh is the fix, not a sign flip

Sources appendix

How-to-verify is the manual path; local copies live in /workspace/HoCo/refs/.

Changelog