Scientific and economic history is full of ideas that seemed obvious before they were tested. Multifamily investing has its own version of this: better schools should lead to better property performance. The reasoning is straightforward. Better schools attract households, stronger demand supports rents, and stronger rents should support net operating income (NOI) growth. It is intuitive; however, a decade of loan-level data says it's wrong, or at least invisible at the NOI line.

What the academic research says

The idea that school quality is priced into housing is one of the most durable assumptions in real estate, but this largely pertains to home sales. The best evidence on the rental side (Gabe et al., 2022) matched roughly 100,000 asking rent observations on 58,000 apartment complexes across 45 metros to Great Schools ratings and ATTOM School Attendance Zones. Three of their findings matter for this exercise.

First, renters do pay up for school quality, but at about half the rate owners do: roughly 1 to 2% of rent per one-point improvement on a 10-point scale, versus ~4% for buyers. Second, renters care most about elementary schools. When all grade levels are modeled together, only the elementary school effect survives. Renter households with kids likely tend to buy before high school. Third, the effect is conditional: stronger in expensive markets, weaker in the less expensive markets.

One important caveat: that study measures levels; better-schooled locations rent for more at a specific point in time. It says nothing about what happens within a given hold period. It also measures asking rents, not signed leases or NOI, which is the focus point for our analysis.

Our experiment

We took a loan-level multifamily dataset from CRED iQ and kept every property we could observe in both 2015 and 2024. We then matched each one to its assigned schools in both years using SABS 2015 attendance boundaries and ATTOM's 2024 updates, and graded each school with SchoolDigger's in-state rankings, converted to percentiles so 2015 and 2024 are comparable.

When the set of assigned schools changed between the two vintages, we called it a redistricting event (795 properties in all), and classified each as favorable or unfavorable based on whether the new school's percentile rank was higher or lower. Then we asked a simple question: did properties redistricted into better schools grow NOI faster than properties whose assignments never changed?

They didn't

From 2015 to 2024, non-redistricted properties (control group) grew NOI by about 37% cumulatively. Properties redistricted into lower-ranked schools grew about 38% while properties redistricted into better schools only grew 32%.

At first glance, the regressions, with metro fixed effects, seem to confirm it. Favorable redistricting is associated with roughly 6% to 7% lower NOI growth over the same period, and the result is statistically significant. Limiting the analysis to elementary and middle schools, exactly where the rental literature says the effect should be strongest, makes it slightly worse, not better.

Before anyone concludes that good schools destroy value: in every specification, redistricting explains well under half a percent of the variation in NOI growth. This implies that school reassignment is not a contributing factor to NOI growth.

Three ways to read the result

Scale. Existing literature (Gabe et al., 2022) implies that even a meaningful improvement in assigned school quality is worth approximately 1% to 2% of rent. NOI stacks rent on top of expenses, taxes, insurance, and capital events. Over a nine-year window in which the average property grew NOI by 37%, a low, single-digit cumulative rent effect is simply too small to see.

Lag. A boundary change does not reprice an in-place rent roll, and plenty of owners do not even know it happened. Whatever premium exists shows up in asking rents at turnover and migrates into NOI, possibly more slowly than our two-snapshot design can catch, and apparently not within a typical 7 or 10-year hold period.

Endogeneity. School boundaries do not get redrawn at random. Redistricting occurs when enrollment is shifting, which usually means there is new housing being delivered, or neighborhoods are transitioning. If favorable redistricting clusters in submarkets absorbing new competitive supply, our negative coefficient is measuring supply pressure on rents, not schools. This is our leading explanation, and it means redistricting may be more useful as a flag for neighborhood change than as a value catalyst.

What it means for underwriting

Any sponsor making a bet on improved NOI growth as a result of improved schools will likely end up telling investors they came up short. Treat redistricting instead as a due diligence signal: boundaries move where something is changing, and that something is more often a supply pipeline or a demographic shift that may or may not imply a rent windfall. If there is a school premium to capture, it lives in asking rents at lease-up and turnover, rather than a decade-apart NOI comparison. Even still, we would likely be tracking down a 1 to 2% difference in rents that may still not be traceable to NOI.

Anatomy of a redistricting event

Acknowledgements

This exemplary data analysis work was conducted by Neha Arun Angadi, Ishraq Khan, Victoria Lin, and Kenza Slaoui as a part of the M.S. in Data Science Capstone in the Fu Foundation School of Engineering and Applied Science at Columbia University. We also want to thank our industry partners, ATTOM, CRED iQ, MetLife Investment Management, and SchoolDigger for generously providing their data and time.

References

Gabe, J., Robinson, S. & Sanderford, A. (2022). The Relationship Between School Quality and U.S. Multi-family Housing Rents. Journal of Real Estate Finance and Economics, 64, 615–645. https://doi.org/10.1007/s11146-020-09814-0