The Impact of Digital Services Trade on Economic Growth in Developing Economies: A Machine Learning Approach

Digital services trade (DST) has emerged as an important driver for economic growth and development, which has attracted increasing attention from governments, policymakers, scholars, and industry stakeholders. This paper examines the impact of DST on economic growth in 87 developing countries from 2005 to 2023. Using advanced ML methods, specifically the CrossFit Partialing-out LASSO linear regression (CrossFit POLR), the study shows that the DST has a positive and statistically significant impact on GDP. The robustness of these findings is further validated by Bayesian Model Averaging, Driscoll-Kraay standard error correction, and alternative DST proxy variables. The mechanism analysis reveals that employment and technological innovation serve as important mediators in the relationship between DST and GDP. The heterogeneity analysis indicates that low-income, upper-middle-income and high-income developing economies, as well as countries with strong digital infrastructure, derive greater economic benefits from DST compared to lower-middle-income countries and those with weaker digital infrastructure. Based on these empirical findings, the study proposes policy recommendations aimed at enhancing the developmental benefits associated with DST in developing countries.