People often wonder why excess grain is dumped instead of shipped off as aid to people that can't afford it. The reason is simple but complex in its implications: the grain is worth less than the cost to move it somewhere else. In addition to the loss of underutilized grain, someone has to pay even more money to move it that in the net is not a cost efficient way to get food to the people that need it relative to not using dumped grain. So the grain is dumped.
One way to improve the situation is to invest in infrastructure to reduce the logistical cost, allowing the grain to compete across a wider range of market prices. However, that investment in infrastructure has to be offset by return on the grain production it effectively allows and has to account for the unpredictability of a global market that the subsidies are crudely trying to mitigate. There is a diminishing return.
One of the big issues in the global market is that grain trade tends to be highly protectionist, in part for food security reasons. While some countries (like the US) produce grain in vast quantities at low cost, most governments do not want to be dependent on foreign countries for their food. At the same time, these policies increase the price and restrict flow of food in these countries in times when the global market is a much more efficient producer. Subsidies and protectionism tend to be very blunt instruments.
I have seen these sort of 'inversions' as well. As an engineer I've pondered whether or not you could create enough infrastructure to create a non-spoiling product which returned enough capital to move the rest. So in the grain situation you might create a simple distillery which produced grain alcohol, which was sold to cover the cost of moving excess grain to locations where it could be used. Grain alcohol has a number of used and it doesn't spoil, further processing grain into alcohol is a relatively self sufficient process (you can use the alcohol produced to run the distiller and still get net positive output.) It of course presumes free grain.
One of the more curious social effects I've witnessed is that someone will dump grain which is too expensive to move, but they are less willing to simply give it to you for your distillery. They note you're getting an economic benefit from their grain (which is true) but they are unwilling to build their own distillery and capture that value but still won't cede the value to someone who does that original investment.
I think this is because, 'If I'm not gaining anything, you shouldn't either' attitude.
Programmers would be pissed if the code written by them gets used by somebody else to make money, while they don't make any money out of it. Isn't this the whole Java debate these days all about. Sun/Oracle aren't getting much money out of Java(While writing/maintaining/advertising it) while the whole world is using it to make money.
How do we make recycling and efficiency more important than personal gain?
Me: "Give me the grain and I'll put it to productive use."
Them: "What do I get out of it?"
Them: "Then why should I give it to you?"
Me: "Because the resources you used to create it won't be wasted."
Them: "Oh, Okay." <- never happens
Now I'm being humorous there but there might be some social engineering needed. And I wonder if one could promise a share of any future profit (if there is any) which comes from the use of the grain.
I always knew 'reward' vs 'punishment' motivations always exist. But only today I had a blinding flash of obviousness that 'no-reward' 'no-punishments' scenarios are equally demotivating.
And understandably so, I would not put in extra hours at office if not rewarded, yet I'm not getting punished either.
zero sum game
One month i only sold 5, but you asked for the other five. If i give you, there will be more things in circulation and now i will be offered less money for my things. Maybe only next month, but i still lose.
now, you give your exceeding production for free. That person that got it for free makes grain alcohol, sells it cheaper than your paying customer. next month your paying customer will either also want it free, or at a much lower price to remain competitive, lowering your next month profit.
The whole problem is people trying to make more money with the grain by creating false scarcity. if they sold way cheaper when they had more production, the original paying customer would have bought twice the grain and made twice the grain alcohol to begin with. and none of that would have been a problem.
The implication is that even slight excesses result in rock bottom prices, and even slight shortages result in very high prices. This volatility is bad for everyone. Therefore the government steps in with subsidies to even out how much farmers get, and to guarantee a surplus to avoid ruinous food prices for consumers.
But there is simply no good way to do this without perverse consequences somewhere...
That's not really accurate. Yes, the total amount of food may not change, but the type of food changes quite a bit. People will eat less desirable, but cheaper food.
The article mentions that it would be better if some of the farmers started growing vegetables rather than grain, but the subsidies don't pay for vegetables.
Same for wheat - low gluten types are cheaper. (Which is basically why they invented the http://en.wikipedia.org/wiki/Chorleywood_bread_process )
However there is a long lead time for changing how many animals there are in the food pipeline. Thus the amount of meat animals does not increase very rapidly to respond to a bumper crop in wheat. (The meat supply can, however, drop rapidly after a drought or flood...)
And the yearly fluctuations in crop yields tend to be fairly large. Certainly larger than the variation we're likely to see in how much people want to eat.
Yeah it's not like there are mechanisms to deal with this sort of volatility (hint - futures).
It is also a technique mostly for businesses, not consumers. How many consumers do you hear about controlling next month's food bill by buying their produce on a future's market? Right, none.
This ignores a subtlety. If one consumer locks in a large fraction of the supply at a low price, then the supply is smaller than expected, you've made the supply-demand problem more extreme for everyone else. So futures allow entities to lessen their own volatility, but that increases everyone else's volatility.
That whole region had ridiculous yields for years after Mt. St. Helens blew (it was in the ash plume -- great fertilizer) but they had difficulty moving that volume economically during years when the rest of the country had high yields. Or at least that is my recollection as a kid. Once they started dumping grain, the pile only got bigger.
Some of the piles in Indiana and Kansas would make you think they are building a ski slope. First time I saw one I was so confused given that moisture is such a problem and drying is not cheap. Some elevators make a lot of profit managing moisture.
 If you are an awesome engineer and can figure out a cheap way to dry grain / corn, there is one big market for you.