How much can we trust China’s investment statistics?

Abstract

Building upon recent research into the underestimation of China’s official final consumption expenditure, this paper investigates the quality of China’s investment data. We strictly follow the official method to estimate the annual gross fixed capital formation (GFCF) expenditure from 2004 to 2012, and the resulting figures are significantly different from the official statistics. This implies that the ‘total investment in fixed assets’ data, which are the primary source for the estimation of GFCF, grossly exaggerate actual investments, and that the official GFCF figures are not, strictly speaking, independently estimated, as they are purported to be. We deduce that the official gross capital formation figure is more or less a residual item obtained by subtracting final consumption and net exports from the official GDP figure that is calculated based on the production-cum-income approach. As a result, the underestimation of China’s consumption expenditure automatically translates into overestimation of investment expenditure. We conclude that China’s official consumption and investment statistics cannot be trusted as the basis for policy discussions and academic research.