Internet usage and the shadow economy: a panel threshold regression

Purpose: The paper investigates the non-linear relationship between Internet usage and shadow economy. The empirical literature has shown that there is still no consensus on the sign of this type of relationship, supporting the idea that a linear functional form may not be suitable to model the internet–shadow economy nexus. Design/methodology/approach: A dynamic panel threshold regression model is used to test the threshold effect of Internet usage on shadow economy, for 55 countries, from 2001 to 2020. The first-differenced generalized method of moments estimation is employed to control for potential endogeneity and simultaneity. Findings: The empirical findings provide strong evidence of a threshold effect across all empirical specifications, identifying an Internet usage threshold value ranging from 8.6 to 34.7%. This effect is particularly pronounced in Upper-Middle-Income countries. Research limitations/implications: Due to limited data availability, the study sample includes 55 countries, clustered in Upper- and Lower-Middle income countries, thus excluding High- and Low-income countries. Practical implications: Empirical findings indicate that in the lower regime, Internet usage has a negative impact on the size of the shadow economy. However, in the upper regime, this relationship is reversed, also neutralizing the negative effects of other determinants of the shadow economy. Policymakers and governments should be aware that excessive Internet usage may contribute to the expansion of the shadow economy. Originality/value: This is the first study to analyze the linkage between Internet usage and shadow economy in a dynamic panel threshold framework, showing that as proposed by theoretically literature, the sign of this type of relationship can be both positive and negative.