RainbowStats county cross-section

Artificial Lines, Real Housing Markets

How state boundaries appear to shape local housing dynamics—even after county income, listings and real economic growth are taken into account.

An exploratory analysis of 213 counties in eight states · Observation year: 2024

State membership alone explains 44.5% of the variation in county house-price growth. After state differences are removed, the three local economic variables explain only 1.2%.

A state line is an artificial mark on a map. It does not stop workers from commuting, families from moving or businesses from selling across the border. Yet housing markets appear to treat those lines as real.

We assembled county observations from Washington, Oregon, Idaho, Texas, New York, Illinois, Maryland and West Virginia. For every county with complete data, we compared annual growth in the FHFA All-Transactions House Price Index with growth in per capita personal income, total listings and real GDP. FIPS codes kept the county observations aligned. Texas, with 59 complete counties, became the reference state.

The result is not subtle. The counties separate primarily by state—not by the three measured county fundamentals.

The state averages separate

The clearest place to begin is the state summary. Texas, Oregon and Washington recorded the lowest average county house-price growth in this sample. Illinois, New York and West Virginia recorded the highest. Those differences do not line up neatly with income growth, listing growth or real GDP growth.

StateCountiesHouse pricesIncomeListingsReal GDP
TX592.5%3.6%16.0%3.8%
OR183.0%5.2%17.1%2.6%
WA223.6%4.7%15.8%3.4%
ID64.2%5.5%10.0%4.4%
MD185.7%5.2%11.0%3.1%
WV126.1%4.3%6.6%3.7%
NY476.4%4.5%30.1%2.5%
IL316.8%3.6%6.3%1.8%

Table 1. Mean annual log changes for complete county observations. Figures are rounded.

Texas, Oregon and Washington occupy the lower-growth part of this sample. Illinois, New York and West Virginia generally sit higher. Maryland and West Virginia were included deliberately: neighboring states with very different political and economic identities help prevent an easy “red versus blue” or “coast versus interior” explanation.

Local fundamentals do little within states

ModelR²Adjusted R²What it captures
Pooled OLS0.0380.024Income, listings and real GDP
State indicators only0.4450.426Differences among state means
State fixed effects0.4510.424Common slopes plus state intercepts
Within-state regression0.012−0.003Variation after state means are removed
State interactions0.5250.444Different intercepts and slopes

The state fixed-effects test is decisive: F = 21.740 with p < 0.001. The joint slope-interaction test is not: F = 1.337 with p = 0.157. The important difference is therefore primarily a shift in state intercepts, not persuasive evidence that income, listings and GDP have entirely different slopes in every state.

Within the fixed-effects model, per capita income growth has a coefficient of 0.236 (t = 1.299), listing growth is effectively zero (t = −0.038), and real GDP growth has a coefficient of 0.019 (t = 0.302). None explains much of the remaining county variation.

State fixed-effect coefficients relative to Texas
Figure 1. Estimated state intercept differences relative to Texas after controlling for county income, listings and real GDP growth. Texas is the largest group and therefore the reference state.

Relative to Texas, the estimated 2024 state effects are approximately 4.4 percentage points for Illinois, 3.7 for New York, 3.4 for West Virginia and 2.9 for Maryland. Oregon's estimated difference is close to zero; Washington and Idaho lie between the extremes.

What remains after accounting for the state?

The sorted residual chart answers a different question. After the model assigns each state its own intercept, which counties still performed materially above or below their state-conditioned predictions?

Sorted Fixed-Effects ResidualsColor identifies state0.00-0.0842.4-0.0584.8-0.021270.021700.052120.08Observations ordered by residualResidualCortland, NY Actual: -0.007 Predicted: 0.062 Residual: -0.069Cortland, NYHays, TX Actual: -0.035 Predicted: 0.028 Residual: -0.063Hays, TXFulton, NY Actual: 0.019 Predicted: 0.066 Residual: -0.046Fulton, NYTravis, TX Actual: -0.019 Predicted: 0.028 Residual: -0.046Coryell, TX Actual: -0.019 Predicted: 0.026 Residual: -0.045Kings, NY Actual: 0.017 Predicted: 0.062 Residual: -0.044Williamson, TX Actual: -0.015 Predicted: 0.028 Residual: -0.043Adams, IL Actual: 0.028 Predicted: 0.068 Residual: -0.040Chautauqua, NY Actual: 0.028 Predicted: 0.067 Residual: -0.039Warren, NY Actual: 0.027 Predicted: 0.065 Residual: -0.038Hood, TX Actual: -0.012 Predicted: 0.024 Residual: -0.036Whitman, WA Actual: -0.001 Predicted: 0.035 Residual: -0.036Wood, WV Actual: 0.021 Predicted: 0.057 Residual: -0.036Cabell, WV Actual: 0.028 Predicted: 0.061 Residual: -0.034Livingston, NY Actual: 0.035 Predicted: 0.065 Residual: -0.029Raleigh, WV Actual: 0.034 Predicted: 0.063 Residual: -0.028Tioga, NY Actual: 0.036 Predicted: 0.064 Residual: -0.028Queens, NY Actual: 0.034 Predicted: 0.061 Residual: -0.027Jackson, IL Actual: 0.039 Predicted: 0.065 Residual: -0.027Baltimore City, MD Actual: 0.030 Predicted: 0.056 Residual: -0.026Jefferson, NY Actual: 0.036 Predicted: 0.062 Residual: -0.026Harrison, WV Actual: 0.033 Predicted: 0.059 Residual: -0.025Walla Walla, WA Actual: 0.010 Predicted: 0.034 Residual: -0.024Bexar, TX Actual: 0.001 Predicted: 0.025 Residual: -0.023Worcester, MD Actual: 0.037 Predicted: 0.060 Residual: -0.023Bronx, NY Actual: 0.039 Predicted: 0.062 Residual: -0.023Josephine, OR Actual: 0.011 Predicted: 0.033 Residual: -0.022Saratoga, NY Actual: 0.045 Predicted: 0.065 Residual: -0.020Smith, TX Actual: 0.003 Predicted: 0.022 Residual: -0.019Richmond, NY Actual: 0.045 Predicted: 0.062 Residual: -0.018Douglas, OR Actual: 0.015 Predicted: 0.033 Residual: -0.017Potter, TX Actual: 0.007 Predicted: 0.024 Residual: -0.017Multnomah, OR Actual: 0.012 Predicted: 0.029 Residual: -0.016Grundy, IL Actual: 0.050 Predicted: 0.066 Residual: -0.016Prince Georges, MD Actual: 0.041 Predicted: 0.057 Residual: -0.016Putnam, WV Actual: 0.049 Predicted: 0.064 Residual: -0.015Cattaraugus, NY Actual: 0.049 Predicted: 0.064 Residual: -0.015Comal, TX Actual: 0.013 Predicted: 0.027 Residual: -0.014Harford, MD Actual: 0.043 Predicted: 0.057 Residual: -0.014Bastrop, TX Actual: 0.011 Predicted: 0.026 Residual: -0.014Rockwall, TX Actual: 0.010 Predicted: 0.024 Residual: -0.014Spokane, WA Actual: 0.022 Predicted: 0.036 Residual: -0.014McLean, IL Actual: 0.055 Predicted: 0.069 Residual: -0.014Angelina, TX Actual: 0.011 Predicted: 0.023 Residual: -0.013Tazewell, IL Actual: 0.058 Predicted: 0.070 Residual: -0.012Madison, IL Actual: 0.058 Predicted: 0.070 Residual: -0.012Clark, WA Actual: 0.024 Predicted: 0.036 Residual: -0.011Guadalupe, TX Actual: 0.015 Predicted: 0.026 Residual: -0.011Taylor, TX Actual: 0.016 Predicted: 0.027 Residual: -0.011Washington, OR Actual: 0.015 Predicted: 0.026 Residual: -0.011Baltimore, MD Actual: 0.046 Predicted: 0.056 Residual: -0.010Tompkins, NY Actual: 0.054 Predicted: 0.064 Residual: -0.010Benton, WA Actual: 0.025 Predicted: 0.035 Residual: -0.010Lubbock, TX Actual: 0.011 Predicted: 0.021 Residual: -0.010Herkimer, NY Actual: 0.054 Predicted: 0.064 Residual: -0.010Bell, TX Actual: 0.018 Predicted: 0.028 Residual: -0.010Kaufman, TX Actual: 0.015 Predicted: 0.024 Residual: -0.009Franklin, WA Actual: 0.024 Predicted: 0.033 Residual: -0.009Oneida, NY Actual: 0.055 Predicted: 0.064 Residual: -0.009St Clair, IL Actual: 0.060 Predicted: 0.069 Residual: -0.008Umatilla, OR Actual: 0.017 Predicted: 0.025 Residual: -0.008Madison, NY Actual: 0.057 Predicted: 0.065 Residual: -0.008Sangamon, IL Actual: 0.064 Predicted: 0.072 Residual: -0.008McLennan, TX Actual: 0.019 Predicted: 0.027 Residual: -0.008Niagara, NY Actual: 0.055 Predicted: 0.063 Residual: -0.008Orange, NY Actual: 0.055 Predicted: 0.063 Residual: -0.008Kendall, IL Actual: 0.061 Predicted: 0.068 Residual: -0.007Peoria, IL Actual: 0.063 Predicted: 0.070 Residual: -0.007Lane, OR Actual: 0.024 Predicted: 0.031 Residual: -0.007Harrison, TX Actual: 0.014 Predicted: 0.021 Residual: -0.007Ellis, TX Actual: 0.017 Predicted: 0.024 Residual: -0.007Galveston, TX Actual: 0.018 Predicted: 0.024 Residual: -0.007Grant, WA Actual: 0.027 Predicted: 0.033 Residual: -0.007Yamhill, OR Actual: 0.021 Predicted: 0.027 Residual: -0.007Marion, OR Actual: 0.022 Predicted: 0.029 Residual: -0.007Grays Harbor, WA Actual: 0.032 Predicted: 0.038 Residual: -0.006Lamar, TX Actual: 0.016 Predicted: 0.022 Residual: -0.006Tarrant, TX Actual: 0.017 Predicted: 0.023 Residual: -0.006Cook, IL Actual: 0.062 Predicted: 0.068 Residual: -0.006De Kalb, IL Actual: 0.062 Predicted: 0.067 Residual: -0.006Nacogdoches, TX Actual: 0.015 Predicted: 0.021 Residual: -0.006Polk, OR Actual: 0.028 Predicted: 0.033 Residual: -0.005Jefferson, WV Actual: 0.060 Predicted: 0.065 Residual: -0.005Cowlitz, WA Actual: 0.029 Predicted: 0.034 Residual: -0.005Brazoria, TX Actual: 0.018 Predicted: 0.023 Residual: -0.005Deschutes, OR Actual: 0.030 Predicted: 0.034 Residual: -0.004Randall, TX Actual: 0.022 Predicted: 0.026 Residual: -0.004Orange, TX Actual: 0.027 Predicted: 0.031 Residual: -0.004Clackamas, OR Actual: 0.025 Predicted: 0.029 Residual: -0.004Anne Arundel, MD Actual: 0.053 Predicted: 0.057 Residual: -0.004Chelan, WA Actual: 0.029 Predicted: 0.033 Residual: -0.004Kitsap, WA Actual: 0.034 Predicted: 0.037 Residual: -0.004Kankakee, IL Actual: 0.064 Predicted: 0.068 Residual: -0.003Ada, ID Actual: 0.041 Predicted: 0.044 Residual: -0.003Montgomery, MD Actual: 0.051 Predicted: 0.055 Residual: -0.003Wicomico, MD Actual: 0.055 Predicted: 0.058 Residual: -0.003Denton, TX Actual: 0.020 Predicted: 0.023 Residual: -0.003Erie, NY Actual: 0.060 Predicted: 0.063 Residual: -0.003Williamson, IL Actual: 0.066 Predicted: 0.069 Residual: -0.003Island, WA Actual: 0.035 Predicted: 0.037 Residual: -0.003Collin, TX Actual: 0.021 Predicted: 0.023 Residual: -0.002Macon, IL Actual: 0.062 Predicted: 0.065 Residual: -0.002Kootenai, ID Actual: 0.038 Predicted: 0.040 Residual: -0.002Calvert, MD Actual: 0.056 Predicted: 0.058 Residual: -0.002Will, IL Actual: 0.068 Predicted: 0.069 Residual: -0.001Bonneville, ID Actual: 0.041 Predicted: 0.042 Residual: -0.001Rusk, TX Actual: 0.020 Predicted: 0.020 Residual: -0.001Linn, OR Actual: 0.028 Predicted: 0.028 Residual: -0.001Kane, IL Actual: 0.068 Predicted: 0.068 Residual: -0.000Chemung, NY Actual: 0.062 Predicted: 0.062 Residual: -0.000Henderson, TX Actual: 0.023 Predicted: 0.023 Residual: 0.000Champaign, IL Actual: 0.068 Predicted: 0.067 Residual: 0.001Twin Falls, ID Actual: 0.043 Predicted: 0.042 Residual: 0.001Canyon, ID Actual: 0.045 Predicted: 0.044 Residual: 0.001Dutchess, NY Actual: 0.067 Predicted: 0.065 Residual: 0.002Henry, IL Actual: 0.066 Predicted: 0.065 Residual: 0.002Washington, NY Actual: 0.066 Predicted: 0.064 Residual: 0.002Steuben, NY Actual: 0.065 Predicted: 0.063 Residual: 0.002McHenry, IL Actual: 0.072 Predicted: 0.069 Residual: 0.002St Marys, MD Actual: 0.059 Predicted: 0.057 Residual: 0.003Nueces, TX Actual: 0.028 Predicted: 0.025 Residual: 0.003Benton, OR Actual: 0.033 Predicted: 0.030 Residual: 0.003Knox, IL Actual: 0.071 Predicted: 0.067 Residual: 0.003Jackson, OR Actual: 0.035 Predicted: 0.031 Residual: 0.004Pierce, WA Actual: 0.040 Predicted: 0.037 Residual: 0.004Bannock, ID Actual: 0.047 Predicted: 0.043 Residual: 0.004Howard, MD Actual: 0.061 Predicted: 0.057 Residual: 0.004Skagit, WA Actual: 0.043 Predicted: 0.039 Residual: 0.004Victoria, TX Actual: 0.026 Predicted: 0.022 Residual: 0.004Lake, IL Actual: 0.073 Predicted: 0.068 Residual: 0.005Yakima, WA Actual: 0.039 Predicted: 0.033 Residual: 0.006Queen Annes, MD Actual: 0.062 Predicted: 0.056 Residual: 0.006Wise, TX Actual: 0.026 Predicted: 0.020 Residual: 0.006Dallas, TX Actual: 0.030 Predicted: 0.024 Residual: 0.006Fort Bend, TX Actual: 0.030 Predicted: 0.024 Residual: 0.006Berkeley, WV Actual: 0.070 Predicted: 0.064 Residual: 0.006Clinton, NY Actual: 0.067 Predicted: 0.061 Residual: 0.006Johnson, TX Actual: 0.029 Predicted: 0.023 Residual: 0.006Cecil, MD Actual: 0.061 Predicted: 0.055 Residual: 0.007Du Page, IL Actual: 0.075 Predicted: 0.068 Residual: 0.007Thurston, WA Actual: 0.045 Predicted: 0.038 Residual: 0.007Grayson, TX Actual: 0.030 Predicted: 0.023 Residual: 0.007Harris, TX Actual: 0.030 Predicted: 0.022 Residual: 0.008Liberty, TX Actual: 0.029 Predicted: 0.021 Residual: 0.008Broome, NY Actual: 0.073 Predicted: 0.065 Residual: 0.008Kerr, TX Actual: 0.038 Predicted: 0.029 Residual: 0.008Columbia, OR Actual: 0.036 Predicted: 0.028 Residual: 0.009Nassau, NY Actual: 0.071 Predicted: 0.062 Residual: 0.009Jefferson, TX Actual: 0.039 Predicted: 0.030 Residual: 0.010Webb, TX Actual: 0.035 Predicted: 0.024 Residual: 0.011King, WA Actual: 0.047 Predicted: 0.037 Residual: 0.011Schenectady, NY Actual: 0.075 Predicted: 0.064 Residual: 0.011Lewis, WA Actual: 0.048 Predicted: 0.037 Residual: 0.011Macoupin, IL Actual: 0.077 Predicted: 0.066 Residual: 0.011Ohio, WV Actual: 0.071 Predicted: 0.060 Residual: 0.011Rock Island, IL Actual: 0.079 Predicted: 0.067 Residual: 0.012Mason, WA Actual: 0.050 Predicted: 0.038 Residual: 0.012Ontario, NY Actual: 0.077 Predicted: 0.065 Residual: 0.012Albany, NY Actual: 0.077 Predicted: 0.065 Residual: 0.012Whiteside, IL Actual: 0.080 Predicted: 0.068 Residual: 0.012Bowie, TX Actual: 0.038 Predicted: 0.026 Residual: 0.012La Salle, IL Actual: 0.077 Predicted: 0.064 Residual: 0.012Montgomery, NY Actual: 0.079 Predicted: 0.066 Residual: 0.013Allegany, MD Actual: 0.073 Predicted: 0.060 Residual: 0.013Washington, MD Actual: 0.071 Predicted: 0.057 Residual: 0.014Brazos, TX Actual: 0.041 Predicted: 0.027 Residual: 0.014Westchester, NY Actual: 0.077 Predicted: 0.062 Residual: 0.014Parker, TX Actual: 0.039 Predicted: 0.024 Residual: 0.015Frederick, MD Actual: 0.071 Predicted: 0.056 Residual: 0.015Monongalia, WV Actual: 0.072 Predicted: 0.057 Residual: 0.015Clallam, WA Actual: 0.056 Predicted: 0.039 Residual: 0.017Rensselaer, NY Actual: 0.083 Predicted: 0.065 Residual: 0.018Columbia, NY Actual: 0.082 Predicted: 0.064 Residual: 0.018Gregg, TX Actual: 0.040 Predicted: 0.022 Residual: 0.018Genesee, NY Actual: 0.081 Predicted: 0.063 Residual: 0.018Ogle, IL Actual: 0.085 Predicted: 0.066 Residual: 0.018Vermilion, IL Actual: 0.083 Predicted: 0.065 Residual: 0.018Cameron, TX Actual: 0.043 Predicted: 0.024 Residual: 0.019Tom Green, TX Actual: 0.047 Predicted: 0.028 Residual: 0.019Coos, OR Actual: 0.052 Predicted: 0.033 Residual: 0.019Hunt, TX Actual: 0.041 Predicted: 0.021 Residual: 0.020Carroll, MD Actual: 0.077 Predicted: 0.057 Residual: 0.020Lincoln, OR Actual: 0.051 Predicted: 0.031 Residual: 0.020Hidalgo, TX Actual: 0.042 Predicted: 0.022 Residual: 0.020Cayuga, NY Actual: 0.086 Predicted: 0.065 Residual: 0.020Charles, MD Actual: 0.079 Predicted: 0.058 Residual: 0.020Oswego, NY Actual: 0.086 Predicted: 0.065 Residual: 0.021Monroe, NY Actual: 0.086 Predicted: 0.065 Residual: 0.021Franklin, NY Actual: 0.083 Predicted: 0.061 Residual: 0.021Hardin, TX Actual: 0.054 Predicted: 0.031 Residual: 0.022El Paso, TX Actual: 0.051 Predicted: 0.029 Residual: 0.023Kanawha, WV Actual: 0.085 Predicted: 0.060 Residual: 0.024Rockland, NY Actual: 0.087 Predicted: 0.062 Residual: 0.024Midland, TX Actual: 0.045 Predicted: 0.020 Residual: 0.025Suffolk, NY Actual: 0.088 Predicted: 0.063 Residual: 0.025Wayne, NY Actual: 0.089 Predicted: 0.063 Residual: 0.026Snohomish, WA Actual: 0.063 Predicted: 0.036 Residual: 0.026Montgomery, TX Actual: 0.048 Predicted: 0.021 Residual: 0.027Otsego, NY Actual: 0.096 Predicted: 0.069 Residual: 0.027Wichita, TX Actual: 0.052 Predicted: 0.025 Residual: 0.028Winnebago, IL Actual: 0.096 Predicted: 0.068 Residual: 0.028Walker, TX Actual: 0.055 Predicted: 0.025 Residual: 0.030Onondaga, NY Actual: 0.097 Predicted: 0.066 Residual: 0.031Ulster, NY Actual: 0.098 Predicted: 0.065 Residual: 0.033San Patricio, TX Actual: 0.060 Predicted: 0.023 Residual: 0.036Whatcom, WA Actual: 0.072 Predicted: 0.035 Residual: 0.037Putnam, NY Actual: 0.103 Predicted: 0.064 Residual: 0.039Stephenson, IL Actual: 0.112 Predicted: 0.070 Residual: 0.042Marion, WV Actual: 0.102 Predicted: 0.060 Residual: 0.042St Lawrence, NY Actual: 0.109 Predicted: 0.067 Residual: 0.042Mercer, WV Actual: 0.103 Predicted: 0.058 Residual: 0.045Mercer, WVKlamath, OR Actual: 0.089 Predicted: 0.033 Residual: 0.056Klamath, OREctor, TX Actual: 0.079 Predicted: 0.023 Residual: 0.056Ector, TXTXIDILMDNYORWAWV
Figure 2. Sorted fixed-effects residuals, colored by state. State adjustment removes the broad group bands but leaves important local exceptions. Large residuals identify counties for follow-up, not automatic evidence of bad data.

The state summary and coefficient estimates show the clustering; the residual chart shows what the state labels do not explain. That distinction matters. A state boundary appears economically meaningful, but it does not eliminate county-specific housing dynamics.

Why might an artificial boundary become economically real?

A border does not directly change the value of a house. Institutions attached to that border might. States differ in property-tax systems, insurance markets, foreclosure procedures, infrastructure finance and the legal authority delegated to local zoning and permitting. They also entered 2024 after different pandemic-era price cycles, migration patterns and construction responses.

The state coefficient is not a policy coefficient. It is a container for everything shared within a state that the regression has not measured. The result tells us where the unexplained separation occurs; it does not yet tell us which law, institution or historical condition produced it.

The listing result deserves special caution. The series measures active plus pending listings and is not seasonally adjusted before annual conversion. More listings can represent added supply, but listings may also increase because a strong market induces owners to sell. New York combined the largest mean listing increase in the sample—30.1%—with 6.4% house-price growth. Contemporaneous listing growth is therefore not a clean supply shock.

This is not a fifty-state conclusion

Eight selected states are not the United States. This exploratory sample contains 213 complete counties from WA, OR, ID, TX, NY, IL, MD and WV. It does not include all fifty states, and the number of usable counties differs substantially across states. The findings should not be generalized nationally without a larger study.

The next step is to repeat the analysis across all available states and multiple years. A stronger design would compare neighboring counties on opposite sides of state borders, add lagged house-price appreciation, construction permits, housing stock, population or migration, insurance costs and effective property-tax burdens, and test whether the apparent discontinuities persist. A border-county or spatial discontinuity design would move the research closer to causal interpretation.

Conclusion

Our measured county fundamentals barely explain house-price growth within states, while state membership explains nearly half of the total cross-sectional variation. That is too large to dismiss, but too incomplete to declare victory.

State boundaries may be artificial marks on a map. Once governments attach different rules, taxes, risks and development constraints to them—and once housing cycles develop different histories on either side—markets can begin to treat those lines as real.

View the complete RainbowStats replication script

Data and methodology

All calculations were performed in RainbowStats. Growth rates are log differences. County series were joined by five-digit FIPS code, and the cross-section was evaluated at 2024-01-01. Results may change as source data are revised.