Academic Evidence on the Causal Impact of Institutional Landlords on House Prices and Rents
Update (2026-10-05 cycle): The causal-debate frontier is shifting from SFR acquisitions to build-to-rent neighborhood effects, the new GAO audit gives both camps a neutral descriptive benchmark, and the strongest advocacy-side study now concedes the supply-shortage point. The mature synthesis (Gyourko/Brookings + CRS, from the September cycles) still stands: modest local price effects in high-concentration metros, rent-effect evidence thinner and mixed, national effects small.
1. The new neutral benchmark — GAO-26-108675. The second-in-series GAO audit (ICE property-level data, six metros, 2018–2024) is descriptive, not causal, but it fixes the exposure variable every causal study needs: institutional investors own 1–3% of all single-family homes but 4–22% of single-family rentals in these metros, with growth peaking 2021–2023 and declining in 2024. It also documents that dispositions to owner-occupants never exceeded 8% of holdings — relevant to dynamic models that assume investor stock recirculates to families. Full detail in Institutional SFR Ownership and Market Share: National Footprint vs. Local Geographic Concentration.
2. The advocacy side's strongest metro study — and its concession. AELP's Phoenix case study (Sept 2026, Kilgour & Gross) is a narrative case study built on secondary citations rather than new causal identification, but it is the current state of the tenant-advocacy evidence and it makes two notable concessions: institutional investors "did not create the nation's chronic mismatch of supply and demand," and the ROAD Act is "an important measure" — while arguing investors "are taking advantage of and exacerbating that situation." Its Phoenix quantification: institutional investors hold ~13–14% of Phoenix SFR stock (INVH ~9,200 Valley properties); top-10 homebuilders control ~70% of the Phoenix market (up from 44% in 2004); Phoenix leads the nation with 25,000+ BTR units. The affordability deterioration it emphasizes (first-time-buyer price-to-income ratio 3.3 → 4.1, 2013–2024, AEI data) is a supply-scarcity outcome, not an investor-effect estimate — the study does not isolate the investor causal channel for Phoenix itself.
3. The BTR price-effect frontier. The study's sharpest causal citation is Lei Ma's Boston University working paper, "When institutional investors come to town: The local effects of buy-to-rent properties" (2024): "each additional BTR property within 150 meters increased house price growth by 2 to 3%12." If that estimate holds up, the BTR channel — explicitly exempted from the ROAD Act purchase ban — has a larger per-unit neighborhood price effect than the existing-home acquisitions the ban targets, because BTR clusters hundreds of never-for-sale homes in outer-suburb starter-home locations. This is where future methodological scrutiny should go: the Ma paper is a working paper, and BTR's effect on prices via land consumption vs. the counterfactual (would those lots have been owner-occupied starter homes?) is exactly the kind of identification question the SFR literature spent a decade fighting over.
4. The rent-spillover claim to keep testing. AELP also cites the Philadelphia Fed working paper (Lee & Wylie, WP 24-13): "investors raise rents at 60% higher rates than the average increase when first acquiring the property, and higher investor share in a neighborhood is correlated with faster rent increases for non-investor landlords." Note the mixed evidence structure: the first finding is an investor-specific rent premium; the second is a spillover that, if causal, would mean investor effects are understated by investor-only comparisons — but as stated it is a correlation, and Sun Belt rent data keeps cutting the other way (see the running Arbor/Chandan falsification test in the topic threads: fastest SFR rent growth in low-institutional Northeast/Midwest metros, slowest in high-institutional Sun Belt).
Where claims still outrun evidence, both directions: (a) "Institutional landlords charge 5–15% higher rents" (ShelterScope, cited by AELP) rests on cross-sectional comparisons that selection effects can explain; (b) "investors don't affect prices" (industry side) ignores the consistent finding of positive price effects in high-concentration neighborhoods and the new BTR evidence; (c) the 1–3%-of-stock vs 22%-of-rentals gap means both "investors own almost nothing" and "corporations own the market" are rhetorically available and empirically incomplete3 — the honest statement is: small national stock share, meaningful local rental-stock share, modest but consistently positive local price effects, thin and conflicting rent-effect evidence, and a BTR channel whose price effects may exceed the SFR channel the law just restricted.
-
An instance of Institutional single-family acquisitions drive up local housing costs despite expanding rental supply. — The working paper supplies the sharpest causal estimate yet that institutional rental presence raises neighborhood prices, the core causal claim of this theme. ↩︎
-
An instance of The purchase ban restricts the smaller price channel and exempts the larger one. — This is the early causal estimate underwriting the theme's claim that the exempt build-to-rent channel's per-unit neighborhood price effect exceeds the existing-home acquisitions the ban restricts. ↩︎
-
An instance of The share of homes investors own is three different statistics wearing one number. — The same market yields opposing headline shares depending on which denominator is invoked, letting both camps be simultaneously right and wrong. ↩︎