We use rest-frame optical and near-infrared (NIR) observations of 42 Type Ia supernovae (SNe Ia) from the Carnegie Supernova Project at low-z and 37 from the RAISIN (SNIA in the IR) Survey at high-z to investigate correlations between SN Ia host galaxy dust, host mass, and redshift. This is the first time the SN Ia host galaxy dust extinction law at high-z has been estimated using combined optical and rest-frame NIR data (YJ band). We use the BAYESN hierarchical model to leverage the data’s wide rest-frame wavelength range (extending to ∼1.0–1.2 μm for the RAISIN sample at 0.2 ≲ z ≲ 0.6). By contrasting the RAISIN and Carnegie Supernova Project (CSP) data, we constrain the population distributions of the host dust R_V parameter for both redshift ranges. We place a limit on the difference in population mean R_V between RAISIN and CSP of −1.16 < Δμ(R_V) < 1.38 with 95 per cent posterior probability. For RAISIN we estimate μ(R_V) = 2.58 ± 0.57, and constrain the population standard deviation to σ(R_V) < 0.90 [2.42] at the 68 [95] per cent level. Given that we are only able to constrain the size of the low- to high-z shift in μ(R_V) to ≲1.4 – which could still propagate to a substantial bias in the equation-of-state parameter w – these and other recent results motivate continued effort to obtain rest-frame NIR data at low- and high-redshifts (e.g. using the Roman Space Telescope).