In the absence of some form of government intervention, knowledge is a classic public good which will be under-produced because of lack of price excludability. Government intervention may take the form of establishing intellectual property rights, or other means of shielding knowledge-based innovations from imitation or copying. Such intervention offers the prospect of 'privatising' the production of knowledge in the sense that a certain level of private knowledge production may become profitable if producers can appropriate at least part of the benefits of R&D. However, publicly funded R&D or extension still can 'crowd out' private knowledge production by charging lower prices. The principal finding of this study is that such 'crowding out' behaviour may be efficient in the sense of being potentially Pareto superior even if it is at the expense of public funding for so called 'orphan' areas of knowledge production which are privately unprofitable. The reason why conventional conclusions about privatisation and 'crowding out' of private goods need not apply to rural research and extension is that private goods are both rival in consumption and price excludable, while knowledge is intrinsically non rival in consumption even if it can be made price excludable.