How Algorithmic Mass Appraisal Structurally Over-Taxes Low-Income Homeowners

If federal tax authorities systematically pretended that low-income workers earned more than their actual salaries, while recording high-income earners at a fraction of their true wealth, the resulting wealth transfer would trigger immediate congressional intervention. Yet this exact dynamic operates openly at the municipal level across the United States. Through a structural flaw in how local governments value real estate, the property tax functions as a highly regressive levy.[1]

The mechanism driving this disparity is algorithmic mass appraisal. Because municipal assessors cannot physically inspect every home in a jurisdiction annually, they rely on statistical models to estimate property values. These models use hedonic regression, analyzing recent sales data to assign a dollar value to specific property characteristics like square footage, lot size, and bedroom count.[1][2]

In principle, this approach should produce a proportional tax burden. If a county assesses homes at 50 percent of their market value, a $100,000 home and a $1,000,000 home should both be taxed on half their worth. In practice, the algorithms systematically fail at the margins, overvaluing the cheapest properties and undervaluing the most expensive ones.[1]

The scale of this failure is vast. A comprehensive review by the Center for Municipal Finance at the University of Chicago analyzed 40 million property sales between 2007 and 2017. The researchers found that 97.7 percent of United States counties exhibit regressive property tax assessments.[1]

The disparity is not marginal. Nationwide, the least expensive 10 percent of homes are assessed at a rate 133 percent higher than the rate applied to the most expensive 10 percent of homes. This metric, known as the 10/90 ratio, quantifies a structural wealth transfer from the poorest homeowners to the wealthiest.[1]

The algorithmic error stems from unobserved variables. For low-priced homes, the statistical models often miss structural defects—a failing roof, a cracked foundation, or outdated plumbing. Because the algorithm only sees the square footage and the neighborhood, it assumes the home is in average condition, generating an assessed value that exceeds what a buyer would actually pay.[1][2]

Conversely, high-priced homes possess bespoke features that mass appraisal models struggle to quantify. Custom architectural materials, specific panoramic views, and high-end landscaping are rarely captured in standard municipal datasets. As a result, the algorithm reverts to the mean, undervaluing luxury properties relative to their true market clearing price.[1]

The appraisal industry is aware of this vulnerability. The International Association of Assessing Officers (IAAO) maintains strict standards for vertical equity, measured by the Price-Related Differential (PRD). A neutral assessment system should produce a PRD between 0.98 and 1.03. Any figure above 1.03 indicates a regressive system where lower-valued properties bear a disproportionate burden.[2]

Despite these standards, major municipalities routinely operate outside the acceptable range. In 2024, Philadelphia County recorded a PRD of 1.079, indicating severe regressivity. The algorithms used to generate these assessments consistently fail to meet the industry's own baseline for fairness, yet they remain the foundation of local government finance.[2][5]

The consequences of this algorithmic failure are most visible in post-industrial cities. In Detroit, a 2017 parcel-by-parcel reappraisal was intended to correct historical inequities. Instead, researchers found that the overall regressivity worsened. Following the reappraisal, the bottom 10 percent of homes were assessed at a rate four times higher than the top 10 percent.[1][2]

For the lowest-income residents, this over-assessment translates directly into financial ruin. In Detroit, most properties that sold for less than $19,000 were assessed in excess of the state's 50 percent legal limit. This artificially inflated tax burden has been a primary driver of the city's mass tax foreclosure crisis, stripping wealth from neighborhoods that can least afford it.[1][2]

The appeals process, designed as a safeguard against assessment errors, actually exacerbates the regressivity. Wealthy homeowners and commercial property developers routinely hire specialized tax attorneys to appeal their assessments, successfully driving down their tax liabilities. Low-income homeowners rarely have the resources or the technical knowledge to navigate the appeals process.[1]

Because local governments require a fixed amount of revenue—the millage rate—to fund schools and municipal services, successful appeals by wealthy property owners force the tax rate higher for everyone else. The uncorrected over-assessments at the bottom of the market end up subsidizing the successful appeals at the top.[3]

As researchers in the Harvard Law Review recently noted, "In virtually all jurisdictions in the country, expensive homes are undervalued by property tax assessors—and hence under-taxed—while less expensive homes are over-valued and over-taxed." The legal framework of property taxation relies on a mathematical fiction that systematically favors capital.

Some economists argue that this long-standing regressivity is already capitalized into property values. Under this theory, the artificially high tax burden on cheap homes suppresses their market price, meaning that current buyers are compensated for the higher taxes through a lower purchase price. However, this capitalization offers no relief to long-term residents who have borne the inflated tax burden for decades.[1]

Reforming the system requires more than just tweaking the algorithms. While incorporating higher-quality data and machine learning could theoretically improve mass appraisal accuracy, the fundamental opacity of real estate markets makes perfect valuation impossible. As long as local governments rely on estimating the value of unsold assets, the structural bias against the margins will persist.[1][5]

The alternative proposed by some tax equity advocates is a shift toward a land value tax. By taxing the underlying land rather than the structures built upon it, municipalities could eliminate the need to appraise the condition of individual homes. Land values are inherently more uniform and easier to map algorithmically, removing the specific variables that currently drive assessment regressivity.[4][5]

Until such structural reforms are adopted, the property tax will remain a regressive instrument. The algorithms that power municipal finance are not neutral arbiters of value; they are mathematical models that systematically shift the cost of local government onto the properties least equipped to bear it.[5]