Homes Unlike Their Neighbors Are Assessed Less Uniformly

In August we published a finding from Harris County, Texas: older homes are appraised less uniformly than newer ones, and the effect is large and monotonic. A natural question followed. Is that a Houston problem or a property tax problem? To find out we ran the identical specification on Cook County, Illinois, a different state, a different assessor, a different classification system and a different statute. This is the result, reported the same way we would have reported the opposite.

The building-age finding did not replicate. The mechanism under it did. In both counties, the homes most likely to be assessed well above their comparables are the ones whose construction year is far from the median year of their own neighborhood. In Houston those happen to be the old houses. In Chicago they are the old houses and the brand-new ones, because a 2015 house in Cook County is usually a teardown rebuilt on a block of 1920s homes.

What this means for you. If your home is much older or much newer than the houses around it, it is worth checking your assessment against your neighbors. AppealMyTax compares your assessed value to similar homes on the county's own roll in about 30 seconds, free, no signup. If there is a gap, the $29 state DIY kit organizes the evidence and the $49 Property Assessment Audit delivers the comparables when the data supports a claim. You keep 100% of any savings.

What we ran

Every single-family home on each county's own assessment roll, pulled from the data behind our calculator: 928,839 Cook County parcels in residential classes 202 through 210, 234, 278 and 295 across all 38 townships (918,340 with enough comparables to score), and the Harris County set from our August study. Condominiums, two-to-six flats, mixed-use and vacant parcels were excluded in Cook.

The specification is the one from the Harris study, unchanged. The comparable cohort is the assessor's own neighborhood code, not one we drew. Comparables are within 25 percent of the subject's living area and within 10 years of its construction year, minimum five, closest ten by the same distance weighting, compared at the median. A home is flagged when its assessment sits more than 25 percent above the median of its comparables.

We ran it four ways in Cook to make sure the shape was not an artifact: total assessed value with all classes eligible, total value with matching class only, assessed value per square foot, and a strict per-square-foot version with living area within 10 percent, year within 5 and matching class. All four produce the same curve.

What did not replicate: the age gradient

In Harris County the share of homes assessed more than 25 percent above their comparables falls steadily with construction decade, from 18.0 percent for the oldest housing to 1.1 percent for the newest. It never turns around.

In Cook County it turns around. The flag rate falls from the pre-1940 stock to a floor for homes built in the 1980s, then climbs back for every later decade. Homes built in 2010 or later are flagged at about 2.5 times the rate of 1980s homes, and the local dispersion of their comparables is higher than for any decade built since 1960. Cook's newest housing is nearly as badly served as its oldest.

Why: infill versus subdivisions

The two counties built their new housing in opposite ways, and the assessment model feels it.

A house built in Cook County in 2015 is usually a teardown-and-rebuild dropped onto a block of 1920s houses. A house built in Harris County in 2015 is usually one of four floor plans repeated across a subdivision where everything else is also 2015. A mass appraisal model has a dense, homogeneous comparable set for the second house and almost nothing like the first house nearby. That is not a defect in either assessor. It is a difference in how the two counties grew.

What did replicate: distance from your neighbors

Measure each home instead by how far its construction year sits from the median year of its own neighborhood, and the two counties agree.

Same direction in both counties. Monotonic across all six groups in Cook and across the first five in Harris before easing at the extreme. The local dispersion measure moves the same way in both places. Two states, two assessors, two classification systems, two statutes, one relationship.

Cook County is reported as ratios rather than percentages on purpose. Explained below.

Why Cook's absolute rates are not published

We hold every county to the same test before publishing a flag rate as a number. If more homes sit far below their comparable median than far above it, the cohorts are too heterogeneous or the values too clumped to support the level, and we do not publish it. That test disqualified Tarrant County, Texas in August. Cook County fails it too, more mildly: the ratio of homes far above to homes far below is 0.82 in Cook against 1.85 in Harris.

The most likely cause is that Cook's assessed values are heavily discretized. About three quarters of Cook parcels share an exact assessed value with ten or more other parcels, against about two percent in Harris, and there are only 89,622 distinct assessed values across 927,426 parcels. Round-number mass points let a subject sit on a mode while its comparable median sits above it. We could not fully separate that from cohort heterogeneity, so the honest position is to publish the within-county ordering, which a level shift cannot change, and not the level.

Three more differences a careful reader should know. Cook's property classes partly encode age, so class matching is not the pure property-type match it is in Harris. Cook's assessor neighborhoods are far coarser, a median cohort of 652 parcels against 180 in Harris. And our Cook load pairs 2025 board-certified values with 2026 building characteristics, because the 2026 cycle was mid-reassessment when the roll was pulled, while Harris is a single 2026 snapshot.

What this is and is not

This is a claim about how mass appraisal behaves: the further a home is from the typical house around it, the less reliably its assessment tracks its comparables, in either direction. It is not a claim that any particular assessment is wrong, it predicts no outcome, refund or tax saving, and it is not a criticism of either assessor's office. Our commercial interest is plain: AppealMyTax sells comparable-property data and evidence workbooks. We do not file appeals, do not represent owners and take no percentage of outcomes.

The Cook County study this extends, with its full methodology, per-parcel files and limitations, is at Cook County assessment uniformity 2026. The Illinois appeal calendar, township by township, is in Illinois property tax appeal deadlines 2026. For Texas, the county deadlines and the late-correction route are on our deadlines page.

Is my house one of these?

Two quick tests. Look at the construction years of the six or eight houses nearest you. If yours is more than 15 years from the middle of that group, you are in the part of the curve where assessments scatter most. Then compare your assessed value per square foot to theirs. The free check does the second step from the county's own roll wherever we hold parcel data, including Cook County, every Florida county and 14 Texas counties, and the $29 kit organizes the evidence everywhere else.

Analysis date August 29, 2026. Data: Cook County Assessor 2025 certified values and 2026 characteristics; Harris Central Appraisal District 2026 roll. Free to reproduce with attribution.