Abstract
Low fertility is not conducive to healthy population development. The total fertility rate (TFR) is influenced by the education expansion (measured by the proportion of non-student women, NSP), marriage delay (measured by the proportion of married women, MP), and marital fertility rate (MFR). This study decomposes the TFR change into the changes in NSP, MP, and MFR using China's census and 1% population sample survey data. During 1990-2020, the changes in NSP, MP, and MFR contributed - 22%, - 90%, and 12%, respectively, to the changes in TFR. The continuous decline in NSP reduced the TFR, and the intensity continued to increase over time. As the primary negative driving force, the rapid decline in MP also consistently reduced the TFR. The marital fertility rate had a downward effect on the TFR before 2000 and an upward effect after 2000. The effects of NSP, MP, and MFR on the TFR varied with the birth order, age and region (among cities, towns, and villages). In summary, China's TFR has considerably changed in combination with changes in NSP, MP, and MFR. Without effective measures, China's TFR may further decline into the lowest-low fertility trap.