This paper describes agricultural policy choices and tests some predictions of political economy theories. It begins with three broad stylized facts: governments tend to tax agriculture in poorer countries, and subsidize it in richer ones, tax both imports and exports more than nontradables and tax more and subsidize less where there is more land per capita. We test a variety of political economy explanations, finding results consistent with hypothesized effects of rural and urban constituents' rational ignorance about small per person effects, governance institutions' control of rent seeking by political leaders, governments' revenue motive for taxation, and the role of time consistency in policy making. We also find that larger groups obtain more favorable policies, suggesting that positive group size effects outweigh any negative influence from free ridership, and that demographically driven entry of new farmers is associated with less favorable farm policies, suggesting the arrival of new farmers erodes policy rents and discourages political activity by incumbents. Another new result is that governments achieve very little price stabilization relative to our benchmark estimates of undistorted prices, and governments in the poorest countries actually destabilize domestic prices.