Why Your Property Tax Assessment Went Up: Mass Appraisal in Plain English

Your county assessor sent a Change of Value notice showing your home is now worth $50,000 more than last year. Nobody walked through it. Nobody photographed the kitchen. Nobody knows the roof is 22 years old. The number was generated by a computer model — a regression run across thousands of properties at once — and the model gets specific homes wrong all the time. This is how mass appraisal actually works, where it overvalues yours, and how to decide whether to appeal.

The single most common misconception about property assessments is that an appraiser personally evaluated your home. For roughly 99% of homeowners, that didn't happen this year, and it didn't happen last year either. Once you understand what actually produced the number, the decision to appeal becomes much more straightforward.

Your assessor never visited your house

For roughly 99% of homeowners in Washington, the assessor's office produced this year's value via a computer model that processed thousands of properties at once. No walkthrough. No condition check. No interior photos. The county is statutorily required under RCW 84.41.030 to revalue all real property annually, and the only way to do that across hundreds of thousands of parcels with limited staff is to run a model.

Physical inspection happens — but only once every six years, also under RCW 84.41.030, and "physical inspection" is mostly a curbside drive-by to verify the structure still exists and hasn't visibly changed. It is not an interior walkthrough, not a condition assessment, and not the kind of due diligence a fee appraiser performs when you refinance.

Most homeowners don't know this. They imagine the assessor's office as a deliberative body weighing the merits of each home individually. That's not what's happening. What's happening is a regression model is producing a value, a sales ratio study is checking the model's overall accuracy, and your specific home is one of tens of thousands of outputs that nobody on the assessor's staff has personally evaluated.

What mass appraisal actually is

The International Association of Assessing Officers (IAAO) defines mass appraisal as "the systematic appraisal of groups of properties as of a given date using standardized procedures and statistical testing." Translation: a regression model that takes recent sales data, your home's recorded characteristics (square footage, beds, baths, year built, lot size, location), and outputs a "fair market value" estimate. Done in batch. Once a year. By software.

The model has three core inputs:

• Recent arms-length sales in your county or sub-market — typically the last 12-24 months, with non-representative transactions (foreclosures, family transfers, estate sales) screened out.
• Recorded characteristics of your specific home — pulled from your property record card: living area, lot size, year built, bed/bath count, garage type, basement type, stories, and a coded condition rating that's almost always set to "average" by default.
• Location variables — neighborhood, school district, sub-market, sometimes block-level adjustments.

The model fits a relationship between those inputs and recent sale prices, then applies the fitted relationship to every parcel in the county. Your assessed value is the model's prediction for what your home would sell for as of January 1 of the assessment year — the statutory valuation date in Washington under RCW 84.40.020.

The output of this process is what shows up on your Change of Value notice. The whole pipeline runs in a few weeks of compute time, and most of the assessor's office staff time goes into screening sales, calibrating the model, and quality-checking the output at the aggregate level — not at the individual property level.

The three approaches assessors use (and why almost everyone gets the sales-comparison approach)

IAAO recognizes three approaches to value:

• Sales comparison approach — value derived from recent sales of similar properties. Used for roughly 95% of residential properties in Washington. This is what mass-appraisal regression models implement at scale.
• Cost approach — replacement cost of the structure, less depreciation, plus land value. Used for new construction, unique properties, and property types that rarely sell (special-purpose buildings, schools, churches).
• Income approach — value derived from the income the property generates, capitalized at a market rate. Used for rental properties and commercial real estate.

For a typical owner-occupied single-family home, you got the sales-comparison approach run via mass-appraisal software. Your assessed value is essentially: "what the model thinks a home with your recorded characteristics would have sold for in your sub-market as of January 1." If that prediction is wrong, your assessment is wrong.

Why mass appraisal gets your home wrong

The model is, on average, defensible. IAAO standards require it to be. The problem is that "on average" is a statement about the whole county, not about your specific home. There are six systematic ways the model misses on individual properties.

1. Condition isn't really in the data

Most county systems have a condition field, but it's a low-cardinality code (poor / fair / average / good / excellent) and the default for the vast majority of homes is "average." The model assumes you're average. If your roof is 22 years old, your kitchen is 1985, your HVAC is failing, and your foundation has settling cracks — none of that is in the data the model saw. You're priced as average condition. You're being overvalued.

2. Interior renovations aren't tracked unless permitted

Conversely, if your kitchen was gutted and remodeled in 2023 without permits, the model doesn't know. Your neighbor's kitchen could be original 1985 and yours could be a $90,000 remodel — the model treats you both identically. This cuts both ways: it benefits homeowners with un-permitted upgrades and hurts homeowners whose recorded condition overstates reality.

3. Lot quality isn't captured

Lot size is in the data. Lot quality is not. The model treats two 0.25-acre lots in the same neighborhood as equivalent — even if one fronts a busy arterial, has a power easement bisecting it, slopes 30 degrees toward an unbuildable creek, or backs a commercial parking lot. These factors materially affect actual market value. The model doesn't see them.

4. Micro-location effects get smoothed away

Mass-appraisal models work at the neighborhood or sub-market level. Within a single neighborhood, a home one block off the freeway and a home four blocks off the freeway will be treated as comparable on location. In reality, the freeway-adjacent home trades at a noticeable discount. The model averages those effects out — meaning the home next to the freeway is overvalued relative to its actual market price, and the home four blocks away is slightly undervalued.

5. Sales lag the market

The model uses sales from roughly the prior 6-18 months. In a flat market that's fine. In a market that turned recently — interest rates spiked, a major employer announced layoffs, a neighborhood-specific issue emerged — the model is reflecting prices that are no longer current. If your assessment date is January 1, 2026, the model is mostly seeing 2025 sales, some of which closed in early 2025 against very different conditions than today.

6. Outliers and forced sales pull averages around

IAAO standards require screening out non-arms-length sales, but the screening is imperfect. A few outlier transactions — an inheritance sale priced to move, a divorce-driven quick sale, an off-market deal between neighbors — can pull comp averages around in either direction, especially in low-volume sub-markets where each sale carries more weight in the regression.

Why your neighbor pays less than you on a similar house

A frequent and reasonable question: "My neighbor's house is essentially identical to mine. Why is their assessment $40,000 lower?" Four common reasons.

1. They appealed last year

A successful appeal lowers the assessed value for that tax year, and the new lower value becomes the baseline the next year's mass-appraisal model adjusts from. If your neighbor appealed in 2025 and won a 10% reduction, they're starting 2026 from a lower number than you are — even though both homes faced the same percentage market change.

2. They qualified for a senior or disabled exemption

Washington's senior/disabled property tax exemption (RCW 84.36.381) freezes the assessed value at the year the homeowner qualifies and exempts a portion of the value from taxation. A neighbor who qualified in 2018 has an effective frozen value from that year — meaning their tax bill is materially lower than yours even though the homes are similar. The exemption rules and qualification thresholds are covered here.

3. Their recorded characteristics differ from yours

The model uses the data on the property record card. If the assessor has your neighbor recorded at 2,150 square feet and you recorded at 2,400 square feet — and the homes are actually identical — that's a 12% size difference flowing straight into the model output. Data errors on record cards are common: square footage mismeasurements, miscounted bathrooms, an unfinished basement counted as finished. Pull both record cards and compare. If yours is wrong, file a correction with the assessor's office; many counties fix these administratively without a formal hearing.

4. They had a recent low-priced sale that anchored their value

If your neighbor bought their home in late 2024 at a price below current market, the model may weight their recent purchase price heavily in their valuation — particularly in counties where recent sales of the subject property are given outsized influence. You, having owned for years, are getting valued purely off the model's neighborhood-level prediction, which may be higher.

The ratio study — how counties test their own accuracy

Counties don't just publish assessed values and hope they're right. IAAO standards require a sales ratio study — a statistical comparison of assessed values to actual sale prices for properties that sold during the year. The metric is the assessed-to-sale ratio: what fraction of the sale price the assessment captured.

The IAAO standard for residential mass appraisal is a median sales ratio between 0.90 and 1.10. Translation: the median assessed value should be within 10% of actual sale prices. Inside that band, the assessor is doing acceptable work. Outside it, the assessor has a calibration problem.

• Median ratio below 0.90 — the county is undervaluing on average. If you're being assessed near or above market value while the median is below, your specific home is probably overvalued relative to peers.
• Median ratio above 1.10 — the county is overvaluing on average. Your home is suspect by definition, and a ratio-study citation in your appeal petition is direct evidence.
• Coefficient of dispersion (COD) above 15% — the model's predictions are inconsistent across properties even if the median is acceptable. This means individual homes are likely being missed in both directions.

King, Pierce, and Snohomish counties all publish their annual ratio studies. Smaller counties often do as well, sometimes only on request. Search "[your county] sales ratio study" — if your county's most recent ratio study shows the median is outside 0.90-1.10, or the COD is high in your specific market area, that's a sentence you should put directly into your BOE petition.

When to appeal vs. accept

Not every assessment is worth appealing. The BOE process takes 3-5 hours of homeowner time spread across prep, filing, and the hearing — plus the wait for the written decision. If the underlying merit isn't there, you'll lose, and the time was better spent elsewhere.

Appeal if any of these are true:

• Your assessment rose more than 10% year-over-year while comparable sales in your neighborhood were flat or down. The model is reflecting market changes that didn't actually happen at your price point.
• You bought the home in the last 12 months for less than the assessment. Your purchase price is the strongest possible single-property comp — a recent arms-length sale of the exact property — and the assessor's number is by definition higher than market.
• The recorded characteristics on file are wrong. Square footage, bed/bath count, lot size, basement finish — any data error of more than ~5% is a winnable appeal, often resolvable administratively.
• Your neighborhood has condition issues the model can't see. Foundation problems, an outdated roof, original 1980s mechanical systems, an awkward floor plan, an easement or slope issue.
• The county's ratio study shows over-assessment in your area. Even a one-line citation of the ratio study in your petition gives the BOE a reason to scrutinize the assessor's number.

Don't appeal if any of these are true:

• Your assessment rose modestly and comparable sales actually rose. The model is doing its job. Filing an appeal without comp evidence wastes your time.
• Your home is in better-than-recorded condition. You'd be opening the door to the assessor noticing your unpermitted kitchen remodel or the addition that was never recorded.
• You'll move within 12 months. The BOE decision typically arrives 4-6 months after filing. If you sell before the lower assessment takes effect on a tax bill, you've spent time for no return.

Detailed deadline rules and the full BOE petition process are in the master Washington appeal guide.

The math: how much an appeal saves

The dollar savings from a successful appeal are a function of two numbers: how much your assessed value is reduced, and your local effective tax rate (sometimes called the millage rate, expressed as dollars of tax per $1,000 of assessed value).

In King County, total millage rates typically range from $8 to $11 per $1,000 of assessed value depending on the specific taxing district. Pierce, Snohomish, and Thurston are roughly comparable. Some rural counties run a bit lower; some specific city/school combinations run a bit higher.

Worked example

A homeowner in Renton has an assessed value of $825,000. The total millage rate at their address is $10.20 per $1,000 of assessed value. They file a BOE appeal with three solid comp sales and win a reduction to $750,000 — a $75,000 reduction.

A general rule that's accurate enough for back-of-envelope work: in the Puget Sound counties, every $50,000 of assessed-value reduction is worth roughly $400-$550 per year, every year, until your value catches up — which on a successful appeal often takes 2-4 years because the new lower value becomes the baseline the next mass-appraisal model adjusts from.

The compounding matters more than the first-year number. A 10% reduction held for five years on a $750K home is real money — comfortably four figures, often five depending on the specific district.

What "frozen value" looks like for ineligible homeowners

The senior/disabled exemption under RCW 84.36.381 freezes assessed value for qualifying homeowners. If you don't qualify — most working-age homeowners don't — you can still effectively manufacture something close to a frozen value by appealing in any year your assessment jumps materially.

The mechanic: a successful appeal lowers your assessed value for that tax year, and the new lower value becomes the baseline the next year's mass-appraisal model adjusts from. If you compound 3-5% reductions over 5-10 years, the cumulative gap between your assessed value and the unappealed counterfactual becomes substantial.

Numerically: a $750,000 home with three successful 8% appeals over a decade ends up assessed roughly $130,000-$170,000 below where it would have landed under pure mass-appraisal trajectory. At a $10/$1,000 millage rate, that's $1,300-$1,700 per year in tax savings, indefinitely. The work is not glamorous, but the return per hour is high — well into the hundreds of dollars per hour of preparation time for any reasonably-priced home.

What to do right now

A 4-step playbook you can run this week.

Frequently asked questions

Related reading: The complete WA property tax appeal guide · Evidence-gathering for a BOE appeal · The senior/disabled exemption under RCW 84.36.381 · WA Property Tax Appeal Tool

About this article: Written by the Claim Maximizer team. IAAO standards referenced are from the IAAO Standard on Mass Appraisal of Real Property and Standard on Ratio Studies. Washington statute citations verified against the Revised Code of Washington as of April 2026. Not legal or tax advice — we are not a law firm. For commercial appeals or cases involving litigation, consult a Washington-licensed attorney or property tax consultant.