We consider a proposal for pharmaceutical patenting policy: namely, for society to grant and purchase the patent of the first of a new class of drug, instead of purchasing the drug, and award no further patents to runner-up drugs, producing or licensing production with price set to maximise welfare subject to cover costs. It is often observed that when the first of a new class of drugs is patented, it does not necessarily halt the development of a second and a third drug of the same class. The result may be a number of rugs with similar efficacy at similar prices well above the production costs. Where this happens, society could substantially reduce the cost of duplicated R&D and the price of the drug by buying the first patent. This would benefit more patients and produce larger health gains. Under this policy social welfare is increased, the winner is fully compensated, while the runner-up firm incurs possible losses - but there are viable conditions under which firms would not lose on average. We take a drug life-cycle approach to the welfare gains of a patent purchase policy. The results are generated based upon a number of stylised facts regarding R&D in the pharmaceutical industry.
