One of the constant tensions in University and College management is working out which services need to be delivered centrally and which can be decentralized and, if the latter, how they can be provided in a way which has at least some semblance of coherence.

The deal is this: generally it’s cheaper to provide most services centrally. Doesn’t really matter what kinds of services: administrative, research, teaching and learning, internationalization, etc. Economies of scale exist, partially because you get fewer 6-figure director’s salaries, but also partially because larger budgets usually mean a greater scope for implementing savings through investments in IT and automation. On the other hand, there are many good arguments for specialized and tailored services to an individual faculty. Medicine is a world apart from the rest of the university on most matters, and so they often have their own research administration. At physically large campuses it may make sense for more far-flung faculties to have some of their own student services because it is inconvenient for students. (There are also less salubrious reasons: some Deans just want their own services because it increases their scope of control and perceived independence from the rest of the institution. Business faculties are the worst for this.)

So, basically, centralization = efficiency and decentralization = convenience. But there is another factor at work here. One of the things that the Nous Group consulting firm seems to have shown through its work at the University of Alberta is that it’s not simply a convenience/efficiency trade-off. In fact, there’s also a professionalism trade-off. When faculties try to set up services on their own, they sometimes do so on the cheap, and hire generalists rather than specialists for the task. Or they take on some decentralized tasks but don’t full devote resources to it, rather asking someone to do this role part-time or as part of other duties. In other words, the efficiency/convenience trade-off sometimes shows up as lower-quality service.

But there are other casualties of de-centralization as well. One of them, quite often, is decent statistics. Want to get statistics on work-integrated learning experiences at an institution? Well, quite often what you’ll find is that WIL is a decentralized operation, which makes sense because student placement is generally a faculty responsibility since it is the faculty, not the university/college which has the relationships with employers. But of course, each of them will have their own way of collecting and defining data regarding their programs and clients. So, getting a global picture of what is happening is difficult.

Same with internationalization. A few months ago, we at HESA Towers did a bunch of ATI requests at Canadian universities asking for direct international student recruitment expenditures. At smaller institutions, we were charged a nominal fee to gather this information; at bigger ones it crept into the hundreds. U of T wanted over $16,000. I was momentarily outraged until I thought about what we were asking for and how their ATI office would interpret the request. First of all, there were literally dozens of offices on campus. Second of all, not all of them labelled their domestic and international recruitment expenditures separately, so answering this question accurately would mean wading through thousands of invoices. Put simply, decentralization makes accurate knowledge for decision-making more difficult; that $16k cost would have been the same, in person hours, had U of T Meric Gertler ever wanted to know that information. (seriously, I have no idea how anyone runs that university, it’s so decentralized it makes the Holy Roman Empire look like Singapore.)

All of which to say, decentralization is convenient, but it has a lot of costs. However, I don’t think a lot of recentralization is all that easy to engineer: academics get ornery when you make things less convenient. I think there are two ways out of this problem. The first is the route that the University of Alberta took, which is partial re-centralization. If you recall from my blog on this subject a couple of months ago, what U of A has done is to create more common services by grouping them together mainly along tri-council lines and making them share services (the parts about merging their budgets and making them work together more on programming and curriculum was put on hold because that was the part faculty really couldn’t accept). At an institution that size, this strategy will probably work to some extent: consolidation provides some economies of scale, and the ability to replace generalists with specialists will drive more efficiencies, too. But it also has costs in the sense that whatever portion of the work that was being done was genuinely faculty-specific might be done less well in future.

But there is another approach which HESA has been using with clients over the last few months (because, believe me, we get a lot of calls about this kind of problem) is to think about how to use a small number of resources centrally to both improve the effectiveness of the generalists working in the faculties while simultaneously imposing some (light) order and common data standards on them at the same time? The creation of one central officer to act as coach, trainer, point of contact to the outside world, and chief data wrangler doesn’t help reduce costs, obviously. But done correctly it can vastly increase the effectiveness of existing spending, bring significant benefits to data reporting and consistency, and provide a little bit more corporate cohesion to a variety of institutional efforts.

And who among us can say we don’t need more of that?