If your community suddenly realizes it needs more food, there is no way that the food can come to existence in days or even weeks. To some degree, other communities will be more willing to sell more of their food surplus to you if prices go up (after of days or weeks of delay to have the whole thing sorted out), and that would pick up some of the slack. But in the short and middle term, what will happen is that the poor will simply get priced out of the market.
An intelligent policy would be to provide a food buffer, a margin of safety, that gives the price signal enough time to travel through the markets. In this case, you get the benefits of a free market, but prevent the worst side effects of it.
I have read Pollan's Omnivore Dilemma, and found his description of New Deal agriculture policies very interesting. I encourage you to have a look and compare those with the ones in effect today.
But policies do not communicate information very well. They don't respond as fast as markets do, to changes in demand or supply. They are liable to political interference, as different sectors of the economy capture policy-making bodies and extract rents: the people who know most about an industry come from the industry, so there tends to be a two-way door between policy-makers and the industry, warping incentives.
So I would argue that to secure more certainty for food supplies, we would be better off looking at insurance-based mechanisms, rather than direct government involvement in food production, subsidization, etc.