Can we evaluate network brokerage initiatives using data that are byproducts of the network broking process?

Increasingly, development projects list social capital development and network brokerage among their objectives.  How do we quantitatively evaluate such initiatives?  Best practice, diff-in-diff methods may be impossible or too costly.  Here, we try using data that are byproducts of the network broking process to evaluate the Challenge Program for Water and Food along this dimension.  We find that, in accordance with its objectives, the program is associated with bridging linkages between organizations in the water and food sectors and between CGIAR members and their counterparts in government and that, in the case of the former, the association may be causal.