In Governing the Hollow State, Milward references T.S. Eliot to highlight the significance of his poetic word choice in describing a government that has delegated the authority to deliver social services outside of itself, mostly to nonprofit agencies. This emptiness has been known to create vacuums in which accountability becomes muddled, funds are allotted haphazardly, and a deafening choir of managers yell over each other in an effort to have their individual voices heard. However, successful collaborative networks have been formed, and their success is dependent on specific prerequisites: clear governance, central integration, cultural harmony and risk-free negotiating. Focusing on the latter point, this memo will analyze the network’s need to view traditional economic theory not as a prescriptive solution but rather a contrasting lens through which organizational behavior in a public-nonprofit network can be understood and possibly inverted.

According to Guy Peters, networks seem to be a logical emergence in the public sector where a multitude of taxpayer services are provided among different specialized agencies. This is a seemingly inevitable consequence for a government like ours that favors decentralization and separation of powers. The need for a wide variety of service providers in a single industry fits in with democratic ideals which have strong roots in neoclassical economic theory, in which success is a consequence of competition and risk. However, as Milward and Peters note, it is impossible to apply these modes of analysis to the public sector. For one, collaboration in a network depends on stability over a long period of time to “agree on a division of labor regarding who should do what”, as well as time for the “principal to learn how to govern”[1]. The risk of, say, losing contracts in a rebidding process may create an environment that fosters competition instead of long-term network relationships. Competition between nonprofits for scarce funds would also increase conflict and fragmentation within the nonprofit network and perhaps damage necessary relationships.

The reason why public-nonprofit networks cannot follow traditional economic theory can be found in the latter agency’s namesake: these organizations are not fueled by profit, but rather by mission. In this way, a monopolization of service provision actually functions best, as competition works against and not for the organization and its clients. Milward’s study found that a well-funded monopolized public-nonprofit network has the most successful outcomes, as this model allowed for a direct stream of funds as well as a clear and direct form of political governance. This kind of approach may seem illogical or irreverent to private-sector cheerleaders like Seymour Sarason and Eliabeth Lorentz, who John Grubbs criticizes in his review of their book “Crossing Boundaries”. Sarason and Lorentz suggest using a single networking coordinator – a sort of public sector social entrepreneur – to help bury bureaucrats out of their pigeon holes in an effort to overcome the “organizational chart mentality” that keeps performance and efficiency within the public sector low. However, this approach would exacerbate specialization issues that hinder collaboration, as well as posit the public as customers, thereby stripping them of the opportunity to participate in bottom-up collaboration[2].

Competition in public-nonprofit networks create a state of flux in which long-term relationships are impossible, accountability of policy enforcement is lost, and legitimacy for organizational rules become discounted. Milward advises strongly against economic competition, noting that one of the two best mental health agencies he studied was also the worst funded. He credited stability and a clear structure of rules, set ex ante and as the groundwork of collaborative negotiations, led to this agency’s success.

To highlight this point, I would like to describe the organizational structure and efficiency of Philadelphia’s victim services network.  The Philadelphia District Attorney’s Office contracts with 10 different nonprofits, most of them designated by their regions within the city and all specializing in victim services. They are all part of a network in Philadelphia called “Philadelphia Coalition for Victim Advocacy”; however, they do not share information and they all compete for local, state, and federal grant money. This has created tension among service providers and an unwillingness to share information or join together for certain causes. Unfortunately, this competition results in already scarce funds being spread around 10 agencies, and clients end up suffering when services like free counseling need to be cut. As Peters predicted, “the logical advocates for the interests of the socially excluded often are among the most likely to segment their clients and to focus on a particular type of problem that effects these clients”[3]The nature of this industry draws a specific kind of person, many times people who were victimized themselves, and so as they compete out fear that their work will become delegitimized, they ironically degrade the legitimacy of their greater mission.

This system requires a stronger principal-agent relationship with the District Attorney’s Office, which manages the funds but little else, and these agencies that provide most of the direct victim services in the city. This relationship allows the government to “steer rather than row” victim services, but the competitive attitudes of individual principal managers within the network dismantle clear hierarchies of accountability and direction. Ideally, these nonprofits would join forces and establish a seamless, monopolized network of service providers, such as the one praised in Milward’s article. By following Arsenault’s suggestion for an incremental anticipative governance, in which “broad-based discussion between participants at all levels of alliance instills ownership and commitment”[4], this network as well as others could establish a centrally integrated collaboration based on a resource-sharing model so progressive that it would make free market advocates blush.

Milward, H. B., and K. G. Provan. “Governing the Hollow State.” Journal of Public Administration Research and Theory 10.2 (2000): 359-80.

[2] Grubbs, John W. “Can Agencies Work Together? Collaboration in Public and Nonprofit Organizations Reviewed Work(s): Forging Nonprofit Alliances by Jane Arsenault; Getting Agencies to Work Together: The Practice and Theory of Managerial Craftsmanship by Eugene Bardach; Crossing Boundaries: Collaboration, Coordination, and the Redefinition of Resources by Seymour B. Sarason and Elizabeth M. Lorentz.” Public Administration Review, Vol. 60, No. 3 (May – Jun., 2000), pp. 275-280 Published by: Wiley on behalf of the American Society for Public Administration


[3]  Peters, Guy. Pursuing horizontal management: the politics of public sector coordination. University of Kansas. 2015.


[4] Grubbs, Johns W.