The poorest neighborhood in Texas is a gold mine. Your demographic report needs to know why.

There's a census tract in College Station, Texas that looks, on paper, like the most distressed neighborhood in the state: median household income $12,636, poverty rate 76.6%, 100% renters, median age 19.7.

It's the Texas A&M campus. Those "impoverished households" are sophomores.

We pulled every tract in Brazos County to see how deep the distortion runs. Four tracts cluster at the extreme — median ages 19.7 to 21.2, incomes $12,600 to $29,200, renter shares 98 to 100%, poverty rates 63 to 78%. Around them, the county median looks normal-ish: age 31.2, income $59,378. Same county, two unrelated realities, and every one of those campus numbers is technically accurate.

This is the teardown where the data isn't lying — the interpretation is.

Why the numbers break

The ACS counts students where they live during the school year, and a student's "household income" is what the student earns — not the family money, the loans, or the meal plan actually funding their spending. So a campus tract reports poverty-level income attached to a population with reliable, externally-funded discretionary spending and extreme density. Poverty statistics were never designed for people whose rent is paid from three states away.

The failure runs both directions:

• False negative: an income filter screens out the campus tract as too poor, and a QSR, coffee, or late-night concept passes on some of the highest-velocity blocks in the state. Nobody sells more $8 meals per square mile than a college main drag.
• False positive: a concept that actually needs household wealth — home services, premium grocery, family dining — reads the county average, sees a normal market, and doesn't notice that a chunk of the population is 20 years old, transient, and gone every summer.

The corrected read

College-adjacent trade areas need three adjustments. Read density and age as the demand signal, not income — 4,000 people per tract at age 20 with funded spending is a different machine than the same tract at age 40. Check seasonality — a market that loses a third of its population in June needs a summer plan; ask what the town looks like in July before signing a lease priced for October. And separate the student blocks from the town blocks — the professors and staff live in normal tracts a mile away with normal incomes, and the two markets barely overlap even though every area-level average blends them.

The same distortion, milder, shows up around military bases and seasonal resort towns: any place where the census "household" doesn't match the economic household.

The fix isn't better instincts — it's resolution plus context. Tract-level data flags a campus tract instantly (a 19.7 median age is not subtle), which is exactly the kind of thing a 30-second demographic report surfaces before the income filter quietly throws away the best block in town.