LAUSR

Abstract The input-output (IO) table is the core of IO techniques. It is widely applied in discovering economic structure, analyzing international trade and accounting for emissions. However, its timeliness limits the popularity of IO data. The paper provides an effective method to forecast and update national IO tables. We apply the matrix transformation technique (MTT)-based forecasting method on China's 42-sector IO tables in 1992–2020 and build up a database named China Input-Output Database (CIOD). Then, we compare the CIOD with other three main databases that also provide China's IO tables. The result shows the CIOD data match the real data by China's National Bureau of Statistics (NBS) best. It demonstrates the effectiveness of our method.