Is it time to stick a fork in internal cost recovery?

Apologies in advance to those working in departments or agencies based on cost recovery models or whose work itself is to manage internal cost recovery processes within the public sector; what follows isn’t a discussion about you, your worth or your dedication to public service in so much as a discussion (at the conceptual level) about the cost recovery model within which you are working.

Here’s my understanding of how cost recovery works

Agencies based on cost-recovery models are usually ones that provide common yet specialized services to other departments and agencies; common so much they are in demand by all of the organizations they serve and specialized in so much as they usually pertain to one specific service (e.g. information technology, learning and development, legal services, etc).

The rationale is that centralizing these services and administering them centrally costs less than administering them in a distributed fashion, and intuitively this seems to make sense. What I don’t understand, and what I’ve been thinking about a lot this past week is why organizations bother building infrastructure to recover costs from what is (at the end of the day) just another part of the same organization. Why do organizations insist on moving monies from one area to another knowing full well that they will be passed on again to the organization who will ultimately deliver the service for which the money is allocated. Here’s what I mean (click the image to enlarge):

Imagine that the box above represents the entire enterprise. Within that enterprise Org Alpha is in charge of the monies, Org Bravo is responsible for mandate “Y” and Org Charlie is a supporting service based organization that has to deliver Service “X” to all of the organizations within the enterprise.

In order to Org Bravo to achieve its mission it gets funding from Org Alpha and in turn redistributes a portion of it to Org Charlie in exchange for Service “X”. This results in funds being sent and received twice along the path and likely results in a portion of those resources being burned in the transaction as people need to manage that process.

My question is this: If the enterprise knows that the ultimate destination for a subset of the funds is Org Charlie and knows that those funds are required to deliver the necessary and valuable Service “X”, why not simply send the funds directly?

If enterprise resources are scarce, and there is a common understanding that all of the Orgs are a part of the same enterprise, the thought of burning even a small fraction of the enterprise’s precious resources as those resources are moved should give rise to a cost avoidance strategy whereby the enterprise moves resources as little as possible prior to directing Orgs to expend them. Why doesn’t the enterprise simply declare the importance of these value added services as self-evident? (again, click to enlarge)