Host-Mass Corrections to Type Ia Supernovae: Quantifying the Contribution to the Hubble Tension

This work presents a physically consistent analysis of host-mass–dependent corrections in Type Ia supernova data and their contribution to the Hubble constant tension. Using pre-registered tests and enforcing a single, unit-consistent mapping between distance modulus and the Hubble constant, we show that the observed supernova host-mass step (Δ ≈ −0.041 mag) is real, correctable, and statistically significant, but accounts for only a modest fraction of the total H₀ discrepancy. Fitting the correction amplitude directly from data fully removes the host-mass residual offset without introducing redshift-dependent artifacts or cross-probe inconsistencies. The implied shift in the inferred Hubble constant is −1.1% for the full sample (−1.9% for high-mass hosts), explaining approximately 14% of the current ≈8.4% Hubble tension. Larger proposed corrections, including exponential or tier-based scalings not derived from the data, are shown to be physically inconsistent, overcorrect supernova residuals, and fail cross-probe validation. The results establish host-mass effects as a genuine but secondary systematic and place quantitative bounds on how much of the Hubble tension they can resolve. The remaining majority of the tension likely arises from a combination of smaller astrophysical systematics and/or new physics, motivating further precision studies rather than single-factor resolutions.