Global EV Transition: Sales Intensity, Growth, Readiness and Spatial Patterns

Cross-national comparisons of electric vehicle (EV) transitions depend strongly on how adoption is measured. This study reanalyses annual electric-car sales/registration data for 60 countries over 2010–2025 and distinguishes absolute market scale from population-normalised Electric Vehicle Sales Intensity (EVSI), defined as annual EV sales/registrations per 100,000 residents using same-year population. Of 960 possible country-year EV observations, 759 were observed and 201 were missing; missing values were retained rather than treated as zero or interpolated. EVSI substantially reordered national positions relative to absolute 2025 sales (Spearman ρ = 0.492, p < 0.001): Iceland and Luxembourg rose 44 and 38 ranking places, respectively, whereas India and Indonesia fell 43 and 33. Trajectories were reconstructed over common calendar intervals, separating current intensity, long-term growth, recent growth and acceleration. Feature-based clustering supported a broad two- to three-group structure, with a stable three-cluster solution distinguishing established high-intensity markets, rapid-growth emerging markets and lower-intensity transitional markets. EVSI showed strong spatial clustering in 2025 (Moran’s I = 0.566, p = 0.001), as did recent EVSI growth (I = 0.542, p = 0.001), whereas absolute annual EV sales were not significantly clustered under the primary spatial specification. Twelve national readiness domains were constructed from 121 retained Enterprise Survey/B-READY indicators, but broad readiness measures added little predictive information beyond income. GDP per capita alone achieved LOOCV R2 = 0.598 for ln(1 + EVSI2025) and R2 = 0.412 for recent EVSI growth, outperforming readiness-only and readiness-enhanced Ridge, Elastic Net and partial least squares models. No model predicted acceleration reliably. An observed EVSI–readiness framework consequently replaces residual-based alignment classes and distinguishes supported high uptake, uptake ahead of measured readiness, latent potential and constrained transition. The findings support multidimensional EV monitoring based on market scale, population-normalised uptake and transition momentum rather than a single composite ranking.