Predicting the impact of climate change on crop yield is difficult, in part because the production function mapping weather to yield is high dimensional and nonlinear. We compare three approaches to predicting yields: (a) deep neural networks (DNNs), (b) traditional panel-data models, and (c) a new panel-data model that allows for unit and time fixed effects in both intercepts and slopes in the agricultural production function—made feasible by a new estimator called Mean Observation OLS (MO-OLS). Using U.S. county-level corn-yield data from 1950 to 2015, we show that both DNNs and MO-OLS models outperform traditional panel-data models for predicting yield, both in-sample and in a Monte Carlo cross-validation exercise. However, the MO-OLS model substantially outperforms both DNNs and traditional panel-data models in forecasting yield in a 2006–2015 holdout sample. We compare the predictions of all these models for climate change impacts on yields from 2016 to 2100.

中文翻译：

---

比较深层神经网络和计量经济学方法来预测气候变化对农业产量的影响

预测气候变化对农作物产量的影响是困难的，部分原因是将天气映射为产量的生产函数是高维且非线性的。我们比较了三种预测产量的方法：（a）深度神经网络（DNN），（b）传统的面板数据模型和（c）允许在截距和斜率上均采用单位时间固定效果的新面板数据模型农业生产函数中的误差–通过称为均值观测OLS（MO-OLS）的新估算器使之可行。使用1950年至2015年美国县级的玉米单产数据，我们显示DNN和MO-OLS模型在样本内和蒙特卡洛交叉验证中均优于传统的面板数据模型来预测单产。然而，在预测2006–2015年保留样本中的产量时，MO-OLS模型大大优于DNN和传统面板数据模型。我们比较了所有这些模型对2016年至2100年气候变化对单产的影响的预测。