| Abstract: |
We provide theoretical and empirical evidence on the factors that influence
the willingness of academic scientists to share research results. We
distinguish between two types of sharing, specific sharing in which a
researcher shares her data or materials with another and general sharing in
which scientists report results to the entire community (as in conference
presentations). We present two simple games in which scientists research a
problem of scientific merit (with an associated prize of academic and/or
commercial value). In both cases, the scientists have intermediate research
results but none has solved the entire problem.We test these models using a
unique survey of bio-scientists in the UK and Germany regarding their
willingness to "share." Our results generally support both models. In both,
sharing is negatively related to competition and the importance of patents. In
other respects they differ markedly. For example, large teams are more likely
to share specifically but less likely to share generally. Rank does not matter
for general sharing, but it does for specific sharing, where untenured faculty
are less likely to share. One important implication is that policies designed
to enhance sharing must be tailored to the type of sharing. |