Urbanization, Economic Growth and Electricity Consumption in Uzbekistan: An ARDL and Toda–Yamamoto Approach

This paper examines how urbanization and economic growth affect per capita electricity consumption in Uzbekistan, using annual data for 2000–2024 within an extended STIRPAT framework. Electricity consumption is measured as gross consumption (generation plus net imports), compiled from IEA, national statistics, UN and US EIA data. Because the sample is short, the ARDL bounds test is implemented with simulated finite-sample critical values and a conditional bootstrap, and causality is examined with a bootstrap Toda–Yamamoto test. The income elasticity shifted structurally after 2016: it was close to zero in 2000–2016 (0.03) but rose to 0.86 in 2017–2024. Once this kink is allowed for, the bounds tests confirm a long-run relationship (F = 11.32), and 86% of any disequilibrium is corrected within a year; a constant-elasticity specification conceals this relationship. Urbanization raises electricity use in the short run (elasticity 2.43) but has no significant long-run effect. Causality runs only from income to electricity consumption (bootstrap p = 0.020), supporting the conservation hypothesis. A combination of conditional ARDL-ECM and Bayesian VAR (Giannone–Lenza–Primiceri) forecasts projects consumption of 117.9 TWh in 2030 (90% interval: 108.0–124.8 TWh), close to the official target of 120.8 TWh, which is reached with a probability of about 29%.