This study hypothesizes that the income levels of households are affected by their different areas enrolled in the Sloping Land Conversion Program, the local economic condition, and the statuses of their previous earnings. We test these relationships by running quantile regressions with data collected from 182 households in the Loess Plateau region covering the period of 1998-2011. We find that the more cropland was retired, the more subsidy was received, and the more labor was set free from farming, which, in turn, led to a larger decrease in ...