Statistical Analysis of GDP Dynamics During the War Period and Economic Growth Forecast

Abstract
The relevance of the study is determined by the need for statistically cautious assessment of Ukraine’s macroeconomic dynamics under the conditions of full-scale war and substantial uncertainty. The purpose was to compare GDP dynamics in the pre-war and full-scale war periods, assess its associations with selected macroeconomic indicators and derive an indicative medium-term recovery trajectory. The analysis used annual data for 2010–2024 (n = 15) covering GDP, gross fixed capital formation, general government final consumption expenditure, exports and imports. Descriptive statistics, Student’s t-test, Mann–Whitney U-test, effect-size estimation, Pearson correlation analysis and exploratory multiple regression were applied. Mean GDP decreased from 101.96 billion constant 2015 US$ in 2010– 2021 to 75.80 billion in 2022–2024. The between-period difference was statistically significant, t(13) = 5.90, p < 0.001 and substantial in magnitude, with Hedges’ g = 3.59 (95% CI: 1.75–5.37). Exports showed the strongest positive association with GDP (r = 0.939), whereas government consumption expenditure was negatively associated with GDP (r = –0.722); these relationships were interpreted as statistical co-movements rather than causal effects. The exploratory regression model produced R² = 0.966 and adjusted R² = 0.953, but its predictive interpretation is limited by the small sample and 10 residual degrees of freedom. Consequently, the 2025–2027 estimates were framed as scenario-based projections rather than validated forecasts. Under the baseline scenario, GDP shows a gradual but incomplete recovery. The findings provide an exploratory empirical basis for assessing wartime macroeconomic dynamics while explicitly representing estimation and projection uncertainty. Keywords: Economic Growth, Gross Domestic Product, Macroeconomic Indicators, Regression and Correlation Analysis, Statistical Regularity, Time Series.

Author(s): Yurii Pidlypnyi*, Nadiia Vasylieva, Nina Petrukha, Nataliia Tymoshyk, Liudmyla Kyrylova
Volume: 7 Issue: 4 Pages: 1254-1269
DOI: https://doi.org/10.47857/irjms.2026.v07i04.012339