Recent studies have shown that climate change will impose severe challenges on agriculture with profound implications. Although some hypothetical simulations have suggested that an optimal re-arrangement of the growing season can substantially mitigate yield losses under future climate, no causal estimate has been provided on quantifying the extent to which farmers are adapting through growing-season adjustments. Using a novel microlevel data with detailed crop progress information in China over 1993–2013, we show that both planting dates and growing season lengths significantly respond to contemporaneous temperature and precipitation. Our estimates suggest that, for a median site in our sample, the adaptive behavior in growing season adjustments can lead to a two to six days earlier planting date and another three to six days shorter growing season by the end of this century. These induced adjustments can avoid up to 9% of the crop damages caused by climate change. However, our empirical analysis does not find clear evidence of long-run response or accompanied input adjustments, suggesting potential for developing policies and tools to further aid the adaptive process.
