Abstract
This paper investigates the extent and ways in which digital growth is embedded in regional economic and governance contexts that influence juvenile offending in Thailand from 2020 to 2024. It uses a balanced panel of 308 observations from 77 provinces and estimates a dynamic model using the 2-step System Generalized Method of Moments, accounting for endogeneity, unobserved heterogeneity, and dynamic persistence. Strikingly, the extent of the weighted age and shattered path dependence is considerably less for a substantially large portion. At the same time, cases agree that higher internet usage rates trigger more juvenile delinquency, leaving more digitally savvy youngsters alone without parents or guardians. Conversely, education is the most significant contributor to reducing youth crime. Moreover, per capita income has the exact opposite and insignificant effect once dynamic controls are in place. A positive sign indicates that the Integrity and Transparency Assessment is a favorable factor for the governance-reporting hypothesis. This hypothesis argues that bias detection is paramount when strong institutions are the base conditions, rather than incident reporting alone. Validity is a necessary component and includes numerous checks, such as alternative lag structures, one-step estimation, and diagnostic statistics to assess instrument validity. An immediate rallying behind its policies is needed, supported by innovations and by financing its programs to ensure sustainability and realization in the face of widespread, fast-track digitalization in the developing world.