Abstract: Research objective: To dynamically measure the long-term trend of China's economic growth. Research method: Construct a non homogeneous time-varying mixed frequency dynamic factor model to estimate the long-term trend of China's economic growth, and use time-varying variance decomposition to calculate the contribution of each factor to the long-term trend of economic growth. Research has found that China's economic growth trend has already entered a downward range since 2007, and after 2016, its main trend began to shift from a sustained downward trend to gradually stabilizing; The coordinated decline of demographic dividend and technological dividend has led to the downward trend of this round of economic growth, while the relative stability of capital deepening has provided the trend momentum for medium- and high-speed growth; At present, the economic growth trend has officially entered a triple stable stage, in which labor input is becoming more stable, technological growth is becoming more stable, and capital deepening is becoming more stable. This shows that the downward trend of economic growth rate caused by local shocks such as the COVID-19 is only a short-term phenomenon, and will not easily break the trend foundation of steady state in the medium and high-speed stages. Research innovation: Incorporating time-varying factors and non homogeneous data into the mixed frequency dynamic factor model has improved the model's ability to handle irregular data. Research value: It fully affirms that the current economic fluctuations are mostly caused by temporary shocks, and that medium to high-speed growth has a solid trend foundation.

Keywords: Economic Growth;Long-term Trend;Time-varying Mixed-frequency Dynamic Factor Model with Different Starting Points;Factor Decomposition