The Canadian Networks of Centres of Excellence (NCE) program links research
ers at many universities across the country to work on joint projects. Each
individual network actively supports collaborative, networked research, wi
th dozens or hundreds of researchers working in a common area. Each network
also formally manages its research, technology transfer, networking, train
ing, and so on. The impacts of the program's unique design on research, tra
ining, and achieving socio-economic impacts were of great interest, resulti
ng in a series of studies being undertaken to evaluate the program's succes
s. One of the methods used to evaluate NCE was partial benefit/cost (B/C) a
nalysis. Although there are often serious difficulties in using economic an
alysis to review research programs, in the case of NCE it was very successf
ul-NCE was found to have a small positive Net Present Value on the basis of
the benefits from just nine 'big winner' projects. The results are conside
red highly defensible and a number of lessons are drawn regarding how B/C c
an be employed. (C) 1999 Elsevier Science Ltd. All rights reserved.