Monoculture in agriculture is like the Martindale system in roulette: your near-certain, small, steady gains are balanced by the tiny chance of disaster. In agriculture the disaster could take the form of a pest against which an entire region's single variety crop has no resistance. In policy it could take the form of high oil prices, which threaten the road transport system. In both cases government played its part in favouring uniform, global models at the expense of diversity. Most of the ludicrously high levels of subsidy paid to agriculture in the rich countries (US$268 billion in 2006) end up with the largest landowners and large agribusiness corporates. They bid up farmland values, making it necessary to maximise yields per hectare, encouraging the use (and raising the price) of purchased inputs. The result is the dreary and fragile monoculture of field after field planted with oilseed rape or (coming soon) biofuel-yielding crops, in response to the current consensus of Brussels bureaucrats. As the level at which critical decisions gets higher and higher, it's not only the distance between the decision-makers and those affected that increases: it's also the seriousness of the disaster that could strike if the decision-makers get it wrong.
I'm thinking also of policies like Kyoto, which put much of the world's environmentalist resources into one supposed climate-stabilising idea: cutting back anthropogenic greenhouse gas emissions. If that turns out not to be a viable way of solving the problem, then the consequences for all of us could be catastrophic. The Social Policy Bond principle understands that governments can have good, popular intentions, and are often the only bodies that can pay for their achievement. Something like climate change demands action at the very highest level. But a uniform and static approach to a social problem is rarely optimal.