Original Research · 83 Counties · 15 States · August 2026
What Property Assessment Actually Costs to Run
We pulled the adopted budget of every county assessment office serving more than 300,000 people across the western and midwestern United States (83 offices in 15 states) and set each against its parcel count and the taxes it underpins. The offices deciding what your building is worth run on between $18 and $190 per parcel, and the two largest operations in the study sit at opposite ends of that range.
By Charlie Young, Principal, former Managing Director at one of the nation's largest property tax firms · IPT, ULI & CREDA member · LinkedIn · Published August 2026
The short answer
Across 83 county assessment offices, the median cost of valuing one parcel for one year is $48.67. The range runs from $18.36 in El Paso County, Colorado to $189.70 in San Francisco, a tenfold spread for what is nominally the same statutory job. Measured the other way, every office in the study spends between 0.26% and 2.0% of the property taxes it underpins. Assessment is a rounding error in the tax system it governs, and the amount spent per parcel has almost nothing to do with the amount collected: per-parcel spending explains only about a tenth of the variation in cost per tax dollar (r² = 0.11).
The two largest operations are four times apart
Maricopa County, Arizona and Harris County, Texas run the two biggest assessment rolls in this study. Maricopa values 300,000 more parcels than Harris with 451 fewer staff and about a third of the budget.
By state
State structure matters more than county management. Oregon and Texas run the most expensive assessment per parcel; Arizona and Nebraska the least. The last column is the one that explains most of the gap: how many people an office employs for every 100,000 parcels it values.
Does a two-year revaluation cycle make an office cheaper?
Colorado revalues real property every two years rather than annually, and Iowa does the same. Those are the only two states in this study on a biennial cycle (Colo. Rev. Stat. §39-1-104; Iowa Code §428.4). It is a reasonable hypothesis that producing half as many revaluations should cost less.
The data does not support it. Colorado's median county spends $42.13 per parcel, 9th-cheapest of the fifteen states. Iowa sits at $50.29, 12th. Arizona, which values property annually, is the cheapest state in the study at $24.11 per parcel, roughly 40% below the two biennial states. With only two biennial states in the sample, the cycle cannot be isolated from everything else that differs between them, and what can be measured points elsewhere.
Where Colorado does stand apart is the other measure. At $5.18 per $1,000 of tax levied it ranks second of fifteen, and Denver is the single most efficient office in the entire study on that basis at $2.58. That reflects the size of the tax base each Colorado office supports at least as much as the cost of running it.
The sharper signal is inside Colorado rather than across states. Nine Colorado counties work under identical statutes, the same biennial cycle, and the same assessment ratios, and they range from $18.36 per parcel in El Paso County to $67.01 in Boulder, a 3.6× spread that no statutory difference explains.
Staffing is what actually drives cost
Of everything measured here, one variable explains most of the difference between a cheap assessment office and an expensive one: how many people it employs per parcel. Across the 61 counties that publish position counts, staffing intensity explains about two-thirds of the variation in cost per parcel (r² = 0.66), far more than population, tax base, or which state the county sits in.
Maricopa County values 6,034 parcels per employee. San Francisco values 1,174. Every other difference (wage levels, software, statutory duties, appeal volume) ultimately shows up through that ratio.
Why California costs the most, and why that is counterintuitive
Proposition 13 should, in theory, make California the cheapest place in America to run an assessor's office. Most parcels get a 2% inflation adjustment and nothing else. There is no annual revaluation, no sales-ratio chase, no cyclical reappraisal of the whole roll.
California is instead the most expensive state in this study, at a median of $73.34 per parcel. The reason is not salaries. California's median cost per employee ($159K) is almost identical to Washington's ($153K). What differs is headcount: California staffs assessment at a median of 46.2 people per 100,000 parcels, against 23.1 in Washington and 26.0 in Arizona. San Francisco runs 85 per 100,000, the highest ratio in the study.
That is a workload story, and it points at the parts of California assessment that Proposition 13 created rather than eliminated:
- Assessment became transaction-driven rather than cyclical. Every sale triggers a full change-in-ownership reappraisal, and California is the only state here that then issues a supplemental assessment for the remainder of the tax year, effectively a second roll produced parcel by parcel, all year long.
- Exclusions are a claims-processing business. Parent-child transfers, base-year-value transfers for older owners, new-construction exclusions and the changes brought by Proposition 19 all arrive as claims that must be examined individually.
- Decline-in-value review is an annual obligation. Parcels reduced under Proposition 8 must be re-reviewed and restored year by year, which turns a downturn into a permanent recurring workload.
- California is the only state in the study that mandates business personal property audits. Revenue & Taxation Code §469 requires a full audit of every qualifying account on a four-year cycle whether or not anything looks wrong. That means a standing audit staff no other assessor in this dataset has to carry.
The practical read for owners: Proposition 13 makes the annual roll cheap and makes everything around it (transfers, supplementals, exclusions and audits) expensive. If you own business personal property in California, you are the workload.
What Texas spends defending values
Texas is the only state in this study where the appeal apparatus is funded from the assessment budget. Central appraisal districts pay for the appraisal review board that hears protests, and for the district's own litigation. Most districts publish the review-board line separately:
The two measures disagree
Cost per parcel and cost per dollar of tax collected are both defensible ways to judge an assessment office, and they rank counties very differently. King County, Washington spends $54.30 per parcel but only $4.99 per $1,000 collected, because Seattle-area values are high enough to carry the office. Marion County, Oregon spends $74.34 per parcel and $15.17 per $1,000, not through carelessness: Oregon's Measure 5 and Measure 50 caps hold the levy down while the assessment work continues regardless.
Denver runs the cheapest tax dollar in the study at $2.58 per $1,000 levied (every dollar of its assessment budget administers roughly $388 of tax) and cut that budget 10.6% for 2026 while doing it.
What this means if you own commercial property
No county in this study can study your building the way you can. At $20 to $50 per parcel, a few hours of professional time spread across an entire year and every property in the county, assessment is necessarily a modeling exercise. Mass appraisal produces a defensible county-wide result and an unreliable individual one, and the gap between those two things is where every successful appeal lives.
The number worth remembering is the band: every office here spends between a quarter of a percent and two percent of the taxes it underpins. Billion-dollar tax streams are being set by offices funded like a mid-sized municipal department. If nobody checks the model against the reality of a specific building, the model wins by default.
Every county, ranked
Method
- What is counted. The adopted annual operating budget of the body that performs property assessment: a county assessor or appraiser department in most states, an independent central appraisal district in Texas. Texas county tax assessor-collectors, which only collect, are excluded entirely.
- Tax collection is removed wherever it can be. Assessment and billing are separate offices in most of these states but not all. Where one office does both and publishes a program split, only the assessment share is counted: Multnomah, Washington, Clackamas and Marion counties in Oregon, and the Bell and Lubbock appraisal districts in Texas. Where one office does both and publishes no split (Pierce County, Washington, whose elected assessor-treasurer bills 86 taxing districts, and Lane County, Oregon, whose assessor is also tax collector for 85), no split was invented; both are listed with combined figures and excluded from medians.
- One-time money is removed. Reserve draws, fund-balance appropriations and capital surges are deducted so figures reflect recurring cost. Bernalillo County, New Mexico appears at $12.78M of recurring funding rather than its $19.64M authority, which included a $6.86M reserve appropriation. Collin County, Texas is the reverse case: its $6.0M "buy down from designated funds" reduces what taxing units are charged, not what the district spends, so no deduction was taken.
- Texas parcel counts are real property only. Comptroller operations-survey totals include mineral and business-personal-property accounts; Tarrant County's 1.63M accounts include 880,435 mineral interests. Counts here are derived from Comptroller school self-report category detail (categories A, B, C, D, E, F, O and S, summed across every school district split in the county) and were independently re-derived from the source workbook.
- Dollars per $1,000 of assessed value is not comparable across states and is published in the underlying dataset for in-state use only. Assessed value means 100% of market in Texas, Washington, Idaho and Nebraska; Proposition 13 factored base-year value in California; Measure 50 maximum assessed value in Oregon; 35% of taxable value in Nevada; a class ratio applied to limited value in Arizona; a class-specific ratio in Colorado (the 2025 assessment year used here: 6.25% residential for local levies, 27% for most non-residential; for 2026 the non-residential rates split to 25% commercial-classed and 26% industrial-classed real property); one-third in New Mexico; and net tax capacity in Minnesota. Compare states on dollars per parcel and per $1,000 levied only.
- Verification. Every county was researched and then re-audited against its cited primary source by an independent pass. Budget years span FY2025–27 against denominator years within about two years. Where a figure could not be verified it is flagged in the dataset rather than published as fact: Ada County, Idaho's all-entity levy is a rate-method estimate, and Douglas County, Nebraska's assessment-only figure is an estimate because the county publishes no split of its combined assessor and register-of-deeds office.
Cite this page: Strategic Valuation Advisors,
What Property Assessment Actually Costs to Run (August 2026), svatax.com/what-property-assessment-costs.html. Sources: adopted county and appraisal-district budgets for fiscal years 2025–27; property tax statistics published by the revenue departments of Arizona, California, Colorado, Idaho, Iowa, Kansas, Minnesota, Nebraska, Nevada, New Mexico, Oklahoma, Oregon, Texas, Utah and Washington; and U.S. Census Bureau 2025 county population estimates. Corrections welcome:
charlieyoung@svatax.com.