6 ideas
The famine was not primarily a product of drought or natural disaster but of state grain procurement set on inflated harvest reports, which extracted grain from the countryside even as villagers starved. Grain exports and requisitions continued through the worst years, so the death toll followed directly from policy choices about how much food to take.
When cadres at each level are rewarded for exceeding targets and punished as 'rightists' for doubt, they report fantastical yields upward. The center then sets new quotas from these fictional figures, so falsehood compounds at every level until real output bears no relation to what the state believes it can seize.
Dikötter argues that beatings, torture, food deprivation as punishment, and killings were systematic tools that local cadres used to extract labor and grain, not isolated excesses. By his estimate a significant share of the excess deaths, a figure he puts in the millions, came from direct violence rather than starvation alone.
In 1959 Defense Minister Peng Dehuai sent Mao a private letter cautiously criticizing the Great Leap's excesses. Mao treated it as a factional attack and purged Peng. The ensuing anti-rightist campaign silenced remaining dissent and drove a renewed radicalization that made 1960 the deadliest year of the famine.
Dikötter reconstructs the catastrophe from provincial, county, and public-security files, including internal investigation reports and complaint letters, rather than from official statistics or leaders' speeches. This bottom-up archival view reveals mechanisms such as hoarding, black markets, and cadre abuse that aggregate data conceals.
Communal kitchens, backyard steel furnaces, massive irrigation campaigns, and pseudo-scientific methods such as close planting and deep plowing pulled labor away from harvests and wrecked tools, soil, and forests. Pursuing ideologically driven leaps in output degraded the very resources that production depended on, leaving the famine's effects to outlast the campaign itself.