In chapter one of Per and my upcoming food policy textbook, we discuss the importance of considering the food system as a holistic system. Here are two papers that try to take one just what that means and why it's important
Richard Levins (2000) "The Food System: A Holistic Policy Approach" (working paper)
Levins puts forward a very simple model of a three sector food system with farmers who buy from an input sector and sell to a processing sector. Total profits are divided among the three sectors. Government support for farmers can be readily appropriated by any of the three sectors. To the extent that policies foster a competitive farming sector and non-competitive processing and input sectors, government support is likely to go mostly to the other two sectors. The basic policy implication he draws from this is that governments should either increase the competition faced by and in the other sectors or decrease the competition faced by farmers to strengthen farmers' relative position.
Byerlee, Jayne, and Myers (2006), "Managing food price risks and instability in a liberalizing market environment: Overview and policy options," Food Policy 31, 275-87 (ungated)
Their findings on food price variability in the run up to the recent food price spike are important. They found "no evidence that variability has increased" but warn that food system changes "may induce higher and more unstable prices in the future." Dealing with price variability, they argue, needs to be done holistically - that is, consider the impacts of price policies on economic growth, poverty reduction, and food security and incorporate the political economy of the food system by balancing market failure and government failure.
They argue that the role of government in the food system depends on what the rest of the food system looks like. As the food system changes, policy and policy system need to change with them: "Especially in Asia, conditions that motivated direct government involvement have changed as markets and infrastructure have developed... Many countries seem stuck in a vicious cycle in which under-provision of public goods hinders market development which, in turn, is used to justify continued government intervention. ... Such "half-way" reforms may create the worst of all possible worlds."
They end by detailing five different paths governments might take to manage food price variability: encouraging more movement to market systems, building long-run capacity in private markets, direct intervention, encouraging market-based instrument to manage risk, and safety nets. Overall, transparency, efficient management of public resources, rules-based rather than discretionary policy, and active participation from many participants and stakeholder groups are encouraged and advocated. "Particular attention should be given to sequencing reforms... "second-best" outcomes will be the norm."