Baltimore County, Maryland property tax

ACS 5-year median real estate taxes paid (table B25103) divided by ACS 5-year median value of owner-occupied housing units (table B25077), county-wide. This is a ratio of two independently estimated medians — not the median of the ratio, and not a rate any parcel is billed. It reflects the actual mix of overlapping districts, exemptions and assessment caps as experienced by owner-occupants, and it lags: a 5-year sample centres about two and a half years before its end date.

Layer 1 of the county property-tax corpus: a MEASURED baseline from ACS 2020-2024 5-year, not a published rate. No statutory rate, assessment ratio or exemption has been collected for these counties yet, so nothing here should be read as what a particular property would be billed.

THE ASSESSMENT IS NOT THE MARKET VALUE, AND THAT IS THE FIRST THING TO GET RIGHT. §8-103(a)(3) defines "phased in value" as, for the 1st, 2nd or 3rd year of a 3-year cycle, "the prior value of real property increased by one-third, two-thirds, or the full amount by which the value increased over the prior value" — and §8-103(c)(1) then says flatly that "the assessment of real property is its phased in value". So for two years out of every three the taxable assessment sits below the value SDAT actually determined, by construction. The Group 2 accounts reassessed for the 1 January 2026 date of finality rose 12.7% statewide; one third of that increase enters the FY2027 base, two thirds does not, and the rest arrives on a schedule already fixed. Any effective-rate calculation that divides a Maryland bill by a market value will understate the nominal rate, and any model that treats the assessment as this year's value will be wrong in the same direction. THE PHASE-IN IS NOT A CAP, AND NO `Cap` MEMBER IS USED FOR IT. A cap bounds an increase; the phase-in schedules one that has already been determined in full, prints all three years of it on the reassessment notice (§8-401(c)(3)), and never reduces the total. It is also ONE-DIRECTIONAL — the definition's second limb takes "the value determined in the most recent valuation" where value has not increased, so a DECREASE lands in full immediately while an increase waits. Calling it an assessment-growth cap would assert a bound that does not exist and would hide the asymmetry that does. The one thing in this record that IS an assessment-growth cap is the homestead credit, and that is a different instrument entirely. THE STATE ASSESSES EVERY PARCEL IN ALL 24 COUNTY-EQUIVALENTS. §2-202(1) directs the Department to "assess all property that is subject to assessment", §2-105(b)(2) has each county's supervisor of assessments appointed by the Director and placed in the State Personnel Management System, and §8-104(b)(1) puts the triennial valuation duty on "the Department or supervisor". There is no county assessor in Maryland, no county board of equalisation setting a level, and no county-to-county ratio drift of the kind North Carolina measures — the sales-ratio surveys §2-202(12) requires are the Department auditing its own work. The counties do pay for it, and from FY2026 they pay far more: §2-106(b) makes each county and Baltimore City reimburse the State for 90% of real property valuation, 90% of business personal property valuation and 90% of the Department's IT office — raised from 50% by the Budget Reconciliation and Financing Act of 2025, so a comparison against an earlier year is comparing two different regimes. MARYLAND LEVIES A STATE PROPERTY TAX, AND IT IS A DEBT SERVICE TAX. 11.2 cents per $100 on all real property other than public utility operating real property and 28.0 cents on that, determined by the Board of Public Works on 22 April 2026 for FY2027 and unchanged from FY2026. State Finance and Procurement §8-134 is the authority and states the purpose exactly: the Board certifies "the rates of State tax on assessable property needed to meet the debt service requirements during the next taxable year on all the State bonds that the Board anticipates will be outstanding", and each county "immediately shall collect" at those rates. §8-134(b) even lets the Board suspend the levy for a year where the General Assembly has appropriated enough to cover debt service — so this is a residual levy against a bond schedule, not a general revenue rate, and reading a year-over-year change in it as tax policy would be a mistake. It reaches real property only: the State column for personal property is 0.0000 everywhere. THE CONSTANT YIELD TAX RATE NO LONGER DOES WHAT EVERY SECONDARY ACCOUNT SAYS IT DOES, AND THIS IS THE SHARPEST CORRECTION IN THIS RECORD. Justia and FindLaw still head Tax-Property §6-308 "Constant yield tax rate"; the served text of §6-308 contains the phrase nowhere. Subsection (b)(1) now bars a taxing authority from setting a rate "that exceeds the CURRENT TAXABLE YEAR'S real property tax rate" without an advertised notice and public hearing — a nominal rate test, not a revenue test. THE ACT IS CHAPTER 531 OF THE ACTS OF 2023 (Senate Bill 114, "Property Tax – Real Property Tax Assessment and Tax Rate Increases – Notice Requirements"), effective 1 June 2023, which repealed and reenacted §1-402(1)(i), §2-205 and §6-308 together and also deleted the old $25,000 municipal de minimis exemption from the constant yield regime. The year-pinned Code editions bracket it exactly: the 2023RS edition of §6-308 still bars a rate "that exceeds the constant yield tax rate", the 2024RS edition reads as it does today. SDAT dates the change to 2024 because February 2024 was the first certification cycle it described under the new rule — "Prior to 2024, the local taxing authority could not set a tax rate that exceeded that Constant Yield Tax Rate until first advertising its intent to do so. Now, the local taxing authority cannot set a tax rate that exceeds the current taxable year's real property tax rate." The agency's own successive reports show the switch in its own prose: the CYTR 2023 report says a jurisdiction must advertise if it plans "to set a tax rate higher than the constant yield rate"; the CYTR 2024 report says "higher than the current fiscal year's tax rate". Cite the act, not the reputation. WHAT THAT PERMITS IS NOT A HANDFUL OF COUNTIES — IT IS ALL OF THEM. Reading SDAT's FY2027 constant yield rates against the rates actually adopted, EVERY ONE of the 24 county-equivalents set a FY2027 county real property rate above its own constant yield. Baltimore City 2.2480 against a CYTR of 2.2005; Harford 0.9779 against 0.9353; Prince George's 1.0000 against 0.9697; Allegany 0.9750 against 0.9348; Somerset 1.0000 against 0.9266; Worcester 0.8150 against 0.7564; and so on through the whole table, the narrowest being St. Mary's at 0.8478 against 0.8245. TWENTY-TWO of the 24 did it without raising the nominal rate at all — several while CUTTING it, Anne Arundel from 0.9770 to 0.9680, Wicomico from 0.8099 to 0.7799, Garrett from 1.0200 to 1.0000 — so twenty-two owed nobody an advertisement or a hearing. Only Frederick (1.1000 to 1.1100) and Talbot (0.8032 to 0.8180) raised a nominal rate and so came within §6-308 at all. SDAT's own explainer states the principle the amendment abandoned — "Any increase in revenue that is generated by adopting a real property tax rate that is higher than the Constant Yield Tax Rate is a tax increase, even if the adopted tax rate is the same as the previous year's tax rate." Maryland should not be described as having a binding revenue-neutrality requirement the way Tennessee or Missouri can be; for FY2027 it did not bind anywhere. A SECOND, SMALLER DISCREPANCY, RECORDED RATHER THAN RESOLVED — AND IT IS ONE YEAR OLD. §2-205 now makes the constant yield calculation REQUEST-BASED — (c)(1) lets a taxing authority "request that the Department calculate" it by February 1, and (b)(1)(ii) sends it only if requested — while both SDAT pages describe universal certification before February 15 and the FY2027 report says certification sheets were mailed to local governments on 13 February 2026. The change is recent and its date matters: the 2025RS Code edition still reads "The Department SHALL notify each taxing authority of the constant yield tax rate", and the request-based text first appears in the 2026RS edition, so it is a 2025-session amendment that WAS in force for the 13 February 2026 certification. The discrepancy is therefore real rather than an artefact of reading a future text — SDAT appears to be doing more than the statute now requires, or every jurisdiction is requesting. This record cites both and asserts neither over the other. "CLASSIFIED" IS TRUE HERE AND MEANS SOMETHING NARROWER THAN IT DOES IN MINNESOTA. Md. Declaration of Rights Art. 15 grants classification power and requires uniformity only "within each class or sub-class"; §8-101 divides property into eleven real-property and seven personal-property subclasses; §6-302(b)(1)(ii) prices personal property and public utility operating real property at up to 2.5 times the real property rate; and the State levy runs at two rates. But there is NO residential subclass and NO commercial subclass — an ordinary house and an ordinary office building are both §8-101(b)(11), "all other real property", both at an assessment ratio of 1, and §6-302(b)(1)(i) requires "a single county property tax rate for all real property". Every assessment ratio in this record is 1; Maryland classifies by RATE and by BASE, never by ratio. THERE IS NO SUCH THING AS "THE" COUNTY RATE IN MARYLAND, AND THIS IS THE SINGLE MOST DANGEROUS FACT IN THIS RECORD FOR A COUNTY-KEYED DATASET. §6-302(b)(1) opens "Except as provided in subsection (c) of this section AND §§ 6-305 AND 6-306" — and those two sections are the municipal TAX SETOFF, which lets a county charge a lower county rate on property inside a municipality that performs services the county would otherwise perform. §6-305 makes it MANDATORY where the case is demonstrated, in eight named counties: Allegany, Anne Arundel, Baltimore County, Garrett, Harford, Howard, Montgomery and Prince George's. §6-306 makes it permissive in the other sixteen. Both say in terms that the county rate inside a municipality "is not required to be the same as the rate for property located in other municipal corporations in the county" or "the same as the rate set in a prior year". This is not theoretical and it is not small. On SDAT's own FY2027 table the Anne Arundel county rate is 0.9680 outside Annapolis and 0.5770 inside it — a 40% difference in the county component of a bill, across a city line. Talbot charges 0.8180 outside its towns and 0.6850 in Easton; Frederick 1.1100 generally and 1.0125 in Frederick City and 0.9666 in Myersville; Allegany 0.9750 generally and 0.8195 in Cumberland. SDAT's constant yield report is laid out around this fact, with a separate row for "Allegany County Excluding towns", "Allegany County in Barton", "Allegany County in Cumberland" and so on for every municipality in the State. The consequence for this corpus: the single county figure a county record can hold is the rate OUTSIDE the municipalities, it is the highest of that county's rates rather than a representative one, and it overstates the county component for every resident of an incorporated town in a setoff county. That is a different failure from the municipal class rates below — those are a rate this corpus never claimed to hold; this is a rate it does claim to hold and would hold wrongly. It is also invisible in any source that prints one row per county. THE RESIDENTIAL/COMMERCIAL SPLIT THAT DOES EXIST IS MUNICIPAL, AND A COUNTY-LEVEL MODEL CANNOT SEE IT. §6-303(a)(2) lets a town set special rates for any class, with no enumerated classes and no cap on the spread. Mount Rainier prints six real-property rates and taxes vacant developed land at 4.0000 against 0.7500 residential; Upper Marlboro prints five; a dozen more towns print two or three. Nearly all of them are in Prince George's County. THE HOMESTEAD CAP IS A STATEWIDE FLOOR WITH A LOCAL OPTION ON TOP, AND ONE NUMBER WILL NOT DESCRIBE IT. 10% for the State levy and for the bicounty commissions, and anything from 0% to 10% in whole points for each county and each municipality (§9-105(e)(2), (e)(5)). For FY2027 that is genuinely dispersed: Talbot and Worcester at 0%, Anne Arundel and Garrett at 2%, Prince George's and St. Mary's at 3%, Montgomery and Calvert and Somerset at 10% — and Annapolis at 10% inside Anne Arundel's 2%. All 24 county figures and every municipal figure are published in the same table as the rates, which is why the whole state is cheap to collect. Note also that the credit caps the growth in the TAXABLE ASSESSMENT and not in the bill: a rate increase passes straight through. CITATION HAZARDS, FOUR OF THEM. First, mgaleg.maryland.gov's `StatuteText` endpoint serves only the CURRENT edition of the Code and `enactments=true` returns "File Not Found" — but THERE IS A YEAR-PINNED PATH AND IT IS THE THING TO USE: `https://mgaleg.maryland.gov/{YEAR}RS/Statute_Web/{article}/{article}.pdf` publishes each article as a complete dated edition (verified live for 2023RS through 2026RS; the Tax – Property article is `gtp`, State Finance and Procurement is `gsf`). The repo's own `pdfText` extracts these cleanly, so a citation to one can carry a real content assertion. The `StatuteText` endpoint at least fails honestly in the meantime: a nonexistent section answers 200 with the body "File Not Found" rather than a shell echoing the number requested, so a content assertion on that host is real evidence too. Second, and this is the trap inside the mitigation: THE EDITION NUMBERING IS OFF BY ONE SESSION. The `2026RS` edition is the Code as amended through the 2025 session, not through 2026 — verified twice, by §6-308, which still carries the pre-Chapter-531 constant yield text in the 2023RS edition although Chapter 531 is a 2023 act, and by §2-106, which reads 50% through the 2025RS edition and 90% in 2026RS, matching the Budget Reconciliation and Financing Act of 2025. Read the wrong way round, that convention turns a correct citation into a one-year error, which is why the FY2027 record pins to `2026RS`. Using it settled three questions this record would otherwise have had to guess at: §2-205's request-based constant yield, §6-302(a)'s June 20 rate deadline and §6-302(d) are all 2025-session changes that DID govern FY2027, and §2-106's 90% cost share is one too (it was 50% for FY2026 and earlier, and the residual (b)(2) still reading "90% instead of 50%" is the tell). Third, SDAT's annual rate table — the single most useful document for this state — is written with Identity-H CID fonts and yields no extractable text to this repo's PDF reader, so it is cited deliberately without a content assertion and the index page is cited beside it so a re-check has a stable entry point. Fourth, the fiscal-year labelling described in `billingNote` will silently produce a one-year error in any automated collection that trusts the number in a Maryland file name. ONE THING RESEARCHED AND NOT ESTABLISHED, recorded rather than guessed: Maryland STATE law contains no rate limit and no levy limit, but two counties are widely reported to impose their own by CHARTER — Prince George's "TRIM" ceiling on the general county real property rate and Montgomery's §305 limit on real property tax revenue growth. Neither was verified against the charter text for this record, both are county law rather than state law, and both belong on the county records. They are named here only so that their absence from `caps` is not read as a finding that they do not exist.