Raising the Accuracy of Shadow Economy Measurements

Abstract

This article estimates the size of the shadow economy in a Spanish region (Navarre) for the period 1986- 2016. To this end, we employ indirect macro-econometric methods such as the Currency Demand approach, Electricity Consumption (Physical Input) methods and the multiple indicators multiple causes (MIMIC) approach. A differential feature of our empirical analysis is that we incorporate various methodological innovations (e..g. Bayesian Model Averaging, a Time-Varying Parameter model, normalization of the latent variable) to refine and increase the measurement accuracy of each of the indirect methods considered. The temporal pattern of the shadow economy’s size that emerges from the different approaches is similar, which suggests that the estimates obtained are robust and capture the underlying dynamics of the hidden sector. After quantifying the shadow economy, we analyze its determinants by means of Bayesian Model Averaging techniques. We find that the evolution of the shadow economy in Navarre can be explained by a small and robust set of factors, specifically the tax burden, the share of employment in the construction sector, the inflation rate, euro area membership and the ratio of currency outside the banks to M1.