exercised by the state on the local systems and found that direct (and not ‘arm’s length’)
mechanisms of control either moderate the impact of the network structure on network
clients’ wellbeing or directly and positively influence the network outcomes. It seems
that direct fiscal control leads to better performance since, according to well-known
agency theory argument (Alchian and Demsetz 1972), it lowers the probability of network
shirking and free riding in using funds.
Similarly, scholars belonging to the so-called community partnerships literature
(Mitchell et al. 2002; Bazzoli et al. 2003) took into consideration the role played by state
regulatory agencies in influencing (without assessing the direction) the overall ability of
the network to achieve their goals. The influence of external stakeholders might even
cause some distributional social effects in the network. O’Toole and Meier (2004a) give
evidence of the so-called ‘dark side of managing networks’, illustrating how superintendents’
generic networking in a Texas school district was of more benefit to influential
and more powerful stakeholders, thus overemphasizing inequalities and discriminations
already present in that area. They also demonstrate how networks, as less formalized and
accountable structures, might function as a mean of hidden policy shift or diversion from
social problems, favouring stronger and better informed external constituencies.