Patterns of Deindustrialization: Are Countries Converging?

During the last decades, the share of manufacturing in aggregate output (and employment) has declined in almost all advanced and emerging economies. In this paper, we investigated the patterns of deindustrialization in a sample of 117 (low-, middle-, and high-income) countries from 1995 to 2018. To this aim, we applied the nonlinear time-varying factor model, initially proposed by Phillips and Sul, to identify potential clubs wherein groups of countries converge toward a similar manufacturing share of GDP. Furthermore, we estimated an ordered logit model to assess the impact of economic globalization and technological revolution on the probability of falling into a particular club. Our results did not provide any support for the hypothesis of global convergence. However, the clustering algorithm successfully identified four strong final clubs, where the share of manufacturing on GDP ranges, on average, from 6% to 18%. Finally, the logit model indicated that as the R&D expenditures and the technological content of manufactured goods increased, so did the likelihood of belonging to a club with a high share of manufacturing value-added on GDP.