First Advisor

Dr. Andrés Holz

Date of Award

Spring 6-2026

Document Type

Thesis

Degree Name

Bachelor of Science (B.S.) in Environmental Science and University Honors

Department

Environmental Science and Management

Language

English

Subjects

Street trees, Hedonic pricing, Housing values, Environmental justice, Spatial analysis, Portland

Abstract

Street trees are positively associated with Portland home sale prices, but exposure and capitalization are uneven by race/ethnic neighborhood context. Urban trees are ecological assets, but their economic and equity implications remain uneven. This thesis examines street-tree exposure and 2023 residential sale prices in Portland, Oregon (n = 6,414) using a mixed-methodology hedonic design. Ordinary least squares models measure inventoried street-tree counts within 20m, 50m, and 100m buffers together with structural controls and census-block racial and ethnic context. Trees are positively and significantly associated with logged prices at all radii, with scale-sensitive magnitudes and modestly better fit as wider buffers. Interaction models indicate that tree-price slopes differ between higher and lower people of color neighborhood contexts while baseline disparities persist. Moran’s I shows strong spatial autocorrelation in OLS residuals; spatial lag and spatial error specifications reduce residual clustering while keeping tree coefficients positive. Complementing those benchmarks, a cross-validated gradient boosting model on the same covariates, using an 80% training / 20% test stratified train-test split, improves out-of-sample log-price prediction relative to linear Model C (Test R² = 0.753 vs. 0.515). Variable importance assigns large weight to structural attributes but retains substantive influence for 100m tree exposure; partial dependence plots summarize flexible tree-price relationships and joint patterns with percent people of color. Descriptive comparisons suggest lower average tree exposure in higher people of color areas across radii. Findings link urban forestry to housing market capitalization in Portland while highlighting environmental justice-relevant heterogeneity and nonlinear prediction beyond global linear slopes.

Download

Share

COinS