Antonio Dalla Zuanna (Bank of Italy)
Domenico Depalo and Edoardo Santoni
This paper presents a novel methodology to estimating the prevalence of undeclared (dependent) employment by leveraging publicly available and timely data that are maintained in most countries. The method employs partial identification techniques to derive bounds on the proportion of undeclared workers by comparing aggregate data from the Labor Force Survey with public-use Social Security records. Applied to Italian data under minimal assumptions, this approach yields credible estimates with narrow bounds: undeclared employment remained stable at 1012% until 2020 but declined to 810% in the post-pandemic period. As official estimates indicate, undeclared employment is more prevalent in southern regions, though it has followed a downward trend. Notably, before a recent methodological revision, official estimates may have overstated the true rate in several instances.