Open science has won another powerful advocate in the UK Royal Society (Nature 486, 441; 2012). But freely sharing research results can have social repercussions that may be damaging to science.

By confusing the allocation of scientific merit and potentially undermining authorship conventions (see, for example, T. Rohlfing and J.-B. Poline NeuroImage 59, 4189–4195; 2012), data sharing could work against individual scientists' need for recognition. This is one reason why scientific institutions, from universities to research councils, do not reward data sharing.

Policy-makers need to remember that, right or wrong, competition has always been a strong driver of science.