Productivity dynamics in Italy. Learning and selection

This paper investigates the sources of labour productivity dynamics in Italy between 2011 and 2018. Exploiting the FRAME-SBS dataset maintained by Istat, we apply productivity decomposition methods to assess the relative contribution of within firm productivity (“learning” effect) and reallocation of market shares across firms (“market selection” effect) to aggregate productivity. While we cannot measure entry/exit dynamics and thus focus on incumbents, the comprehensive coverage of the Italian economy offered by the data enables us to perform a disaggregated analysis at the level of very narrowly defined industries (at 5-digit level, NACE Rev.2). This provides a significant contribution to the literature, as previous studies looked at aggregate economy or aggregate macro-sectors (e.g. total manufacturing). The general picture emerging from the analysis is that within-firm “learning” prevails over between-firm reallocation and allocative efficiency effects in shaping aggregate productivity dynamics. This finding is robust over time and across both manufacturing and service industries. In addition, allocative efficiency is generally stable and rather weak over the reference period, although somewhat stronger in manufacturing than in services.