[Seminaire CREM] Dating growth and cycles in China: A sector-based approach

Abstract:

China’s growth has become of systemic importance. The dating of global business cycles has thus become heavily dependent on the timing of Chinese cycles. However we know very little about movements of Chinese economic activity. This paper has two major objectives. The first is to reconstruct, over four decades, at high-frequency (quarterly), China’s industrial output on the basis of (official) sectoral data. The second is to establish the first reference dating of industrial cycles in order to detect classical recessions and episodes of rapid growth. We focus on industrial output because it is the only driver of rapid growth in GDP and the major source of bias in China’s official GDP data. We obtain three major results. First, official over-reporting of industrial growth was widespread over the first two decades after the reforms and resurfaced in the 2010s. Second, while official aggregate data portrays three decades of rapid catching-up growth in China, the reconstructed data implies that rapid- growth only started after WTO entry and stopped with the global financial crisis (GFC). In contrast, the post-2010 period is characterized, not only by a slowdown, but also by recurrent recessions. Third, we find that the diffusion of recessions across sectors which had fallen for two decades went up in the last decade.