RBR's arguments about subsidies distorting natural competitive advantage and reducing efficiencies are largely unassailable for most areas of trade. To my taste, RBR undervalues the economic pain and dislocation caused by changing trade rules in these areas, which should be part of the cost-benefit analysis, timing, etc. Also, RBR and I disagree as to whether reducing subsidies and tariffs should go hand-in-hand with requirements that our trading partners meet labor and environmental standards. RBR discounts the importance of this, as I see it. But the real issue I have is that arguments about free trade made in general do not apply as well to agriculture as one might think.
Agriculture, is a different kettle of wax, a different ball of fish, a horse of a different feather, and lots of other mangled metaphors. Food is more like a public good (like air and water) than other commodities. Food is a necessity of life. The demand curve for food is unlike most other products. Humans need around 2000 calories/day to thrive (broad averages) and even the most incredible athletes cannot consume more than about 7000 calories per day. Most humans live in the 2000-3000 calorie/day range. Until biofuels were discovered, it was impossible for any human to consume more food than that, regardless of wealth (we have discussed this tragedy elsewhere).
Unfortunately, the market price of food is such that about ¼ of the human race cannot afford their 2000 daily calories. This is not too surprising, when you think of what a market is. A market optimizes. It makes the best price for the most people, but it does not ensure a price low enough for each consumer's demand to be satisfied. Think of any product: housing, medical care, gummi bears, and you will acknowledge that the market creates optimal prices, but does not satisfy all needs. This is a problem when applied to food: each consumer's demand must be satisfied. (IMHO, it is a problem for medical care too). Therefore, unless we subsidize food purchases – whether through lowering prices, payments to farmers to reduce prices, or direct payments to the hungry – starvation results. So, at some level, subsidies for agriculture just have to be part of the world food production scheme. A world free market in agriculture simply cannot feed everyone, and it is not designed to do so. Free trade alone won't do; we need some world agricultural policies.
Think of how subsidies work. If it is true that developed countries "dump" food on third world countries at prices below those which their own farmers could afford to produce and sell food, this may drive those farmers out of business, but it also keeps food prices down. It's a mixed blessing. Allowing the price of food to rise so those farmers would be profitable (by ending subsidies and dumping) would help the farmers but hurt the starving neighbors with higher food prices. If we want to help devloping world farmers without raising food prices, it sounds like more subsidies are needed, just in a different place.
Food calories also are not fungible. Quality matters. Taste matters. Calories are not widgets. If for some reason a rupture in the time/space continuum allowed sufficient quantity of calories for the entire human race to annually appear free of charge, like manna, in the Faroe Islands, this would not be the best of all possible worlds. Fruits and vegetables there would spoil before they could reach the rest of the world. There would be large carbon expenditures in shipping it, too. What I mean by this example is that it is simply impossible to imagine that one can import all foodstuffs and achieve the same quality in all products as is possible with local agricultural production. RBR doubts this is the case, believing that we can basically import everything; I disagree. Steps to make local agricultural production more expensive, therefore, would have the negative effect of increasing prices on high quality products and reduce the number of people who could afford the higher quality products that cannot be imported.
Agriculture is also a way of life that people value. Many people prefer the existence of local agriculture to the more efficient world without it. One could argue that subsidies for rural life in Europe are, in fact, the public's way of "purchasing" or "valuing" the existence of local rural life. As a society, are we not free to value quaintness the way we value other things, such as pretty scenery and the arts?
So there are (1) good reasons to support and maintain local agriculture and (2) good reasons to encourage very low food prices in developing countries, even if that means "dumping" foodstuffs there at prices below that which their own farmers could afford to sell the food. I think any agriculture policy must take these into account. We cannot simply end food subsidies in the developed world and expect that the results will be beneficial in the short term or even the long term.