Food prices spiked sharply in 2007–2008, in 2010–2011 and again in 2021–2022. However, the impacts of these spikes on poverty remain controversial; while food is a large expense for the poor, many poor people also earn income from producing or marketing food, and higher prices should incentivize greater food production. Short-run simulation models assume away production and wage adjustments, and probably underestimate food production by the poor. Here we analyse annual data on poverty rates, real food price changes and food production growth for 33 middle-income countries from 2000 to 2019 based on World Bank poverty measures. Panel regressions show that year-on-year increases in the real price of food predict reductions in the US$3.20-per-day poverty headcount, except in more urban or non-agrarian countries. A plausible explanation is that rising food prices stimulate short-run agricultural supply responses that induce increased demand for unskilled labour and increases in wages.

食品价格在 2007-2008 年、2010-2011 年和 2021-2022 年再次大幅上涨。然而，这些激增对贫困的影响仍然存在争议。虽然粮食对穷人来说是一笔巨额开支，但许多穷人也通过生产或销售粮食获得收入，较高的价格应会刺激更多的粮食生产。短期模拟模型假定不考虑生产和工资调整，并且可能低估穷人的粮食产量。在此，我们根据世界银行贫困衡量标准，分析了 2000 年至 2019 年 33 个中等收入国家的贫困率、实际粮食价格变化和粮食产量增长的年度数据。面板回归显示，食品实际价格的同比上涨预示着每日贫困人口数量将减少 3.20 美元，但城市化或非农业国家除外。一个似乎合理的解释是，食品价格上涨刺激了短期农业供应反应，从而导致对非熟练劳动力的需求增加和工资上涨。