The second is the more complex one though: there is a 'general' belief that the rigour of scientific research (replicable standardised results) should somehow follow into scientific development. In other words, what is done once, or at first, should be done always. This general belief is false because the range of variables operating on a scientific development is uncontrolled, while the variables in experimental science are strictly controlled.

Essentially it means that we will never be able to say what the outcomes of a scientific development will be. Gene inserts may move, sooner or later. Given that it's in the nature of both transplanted and genetic material to change in structure, the gene inserts will either be rejected or modified by the organism as a whole. This happens in nature all the time, but there seems to be a view that scientists should be able to offer guarantees about stability once science gets outside the laboratory. What we need to do better is understand and manage risk, not manage science.

However, if strict pesticide/herbicide management protocols are not observed, all advantages gained from GM crops are lost, as herbicide and pesticide resistant weeds and pests emerge (for the same reasons as to why antibiotic-resistant bacteria arise – when you kill off the things in a population that aren't resistant to a pesticide, the things remaining will be a) resistant and b) produce resistant offspring).

This is one reason why GM crops are not a magical cure for the ills of the developing world – without strict government regulation, all benefits are lost. Some (many?) third-world countries do not have the infrastructure required for such regulation.

And it's not just a case of each country doing it's own thing, either – all it takes is for a pesticide/herbicide resistant insect/weed to spread across the border and again, the advantage of GM is lost.

An interesting question is how different countries approach the question of GM – some ask "How do we know for sure if the crop is harmful?" others ask "How do we know for sure that it isn't harmful?"

Fact is there are no long-term studies on the health and environmental effects of various GM crops. So some countries choose to err on the side of caution, while others take a more short-term view.