Intervention policies are a major but underrecognized way of reducing food waste. This study developed a new methodology for assessing the effectiveness of food waste policies, based on the analysis of 874 policies (1955–2021) and eight different types of entire food supply chain waste (1961–2021) in China, further modeling by support vector regression machine to predict the amount of food waste for 2022–2030. It was found that: (1) The number, policy power yearly and cumulative annual policy power of policies increased each year, while the average annual policy power showed a decreasing trend. (2) Voluntary policies had the highest number of releases, cumulative power and average power, followed by coercive, directive and incentive policies. (3) Voluntary policies had the best overall intervention effect on food waste, followed by coercive, incentive and directive policies. (4) Scenario predictions confirmed that a combined policy mix led by voluntary policies works best.