The cost of support

Two factors are making universities (in the UK in particular) consider the costs of their courses like never before. The first is the withdrawal of state funding and reliance on student fees. I guess this was always the intention behind the shift to a pure market driven approach, it certainly makes universities focus on their course costs. Can they do it for cheaper? Can they justify their fees? Can they lower fees?

The second is our new old friend, MOOCs. It seems rather arrogant but many MOOC providers think they've just invented the idea of considering elearning costs. But all the same, the fact they are gaining attention, and that they are providing alternative models of support (largely through automatic assessment) makes universities take a look at their costs and assumptions too. 

None of this is new, I wrote this paper about economic models of large-scale elearning back in 2004, building on the foundational work of Tony Bates (MOOC providers may care to read it, as it may save them a few years of coming to the same conclusions). But that doesn't mean it isn't valid to reconsider it from the perspective of MOOCs. As I conclude in one of my MOOC presentations, one of the really good things about them are the questions they make us ask ourselves.

MOOCs are about no-support teaching, that's their basic model. In the cMOOCs this support is replaced by a peer network, in the xMOOC by automatic feedback. The student fees makes us consider how much learners are paying for this support, how much they are willing to pay and how good this service is over unsupported.

Below is an idealised chart showing typical ratios for course production and presentation costs over 5 presentations (you need 5 to get a clear picture as all the production costs are in year one). The big cream chunk in the second column is tuition costs. The green bit on top is student support costs (generic and specific student support services, eg support for students with disabilities, pastoral support, running regional centres, etc). The other bits are things like IT services. I've removed the actual figures, it's the relative amounts I want to focus on. This shows that by far the biggest cost is that of tuition. Paying people to support learners is where the money goes. 

The other key element is that, of course, production is a fixed cost. So once we've paid it, we've paid it (more or less, we may need to produce new items). Whereas, most of the presentation costs are variable – they increase as student numbers increase. Of course, countering this is the income side of the graph, where your income increases as you get more students too. The two should balance each other out.


Now, of course if you remove those big chunks in presentation you can offer courses cheaply. We've always known that. MOOCs are essentially doing away with the second bar, it's just the fixed costs of production. The good thing with this model is that there are no variable costs, and each presentation doesn't cost you, but you do still get the income side of the bar (although how that income is realise isn't clear). So in theory, you should move from debit to profit quickly as you recoup those production costs, without the hindrance of additional presentation costs.

The issue is that these services are key to long term success for learners. It isn't evenly distributed though. Some learners hardly ever avail themselves of these, don't care about tuition and do very well studying on their own. Other learners require a lot of support for various reasons and probably have more than their 'fair' share of these services (ie more than they've actually paid for). And most are in the middle, they make use of them sometimes, depending on circumstances.

For distance education in particular, this first group, the confident, independent learners probably do okay by MOOCs. They probably represent the 10% or so who complete. Then there are some for whom no amount of support can help through, either study isn't for them, or this is the wrong time. But sitting in the middle is a big group who need varying levels of support to 'survive' a protracted course of study. If MOOC dropout over 7 weeks is 90%, imagine what it'd be like over 3 or 4 years of degree study?

But that doesn't mean universities shouldn't look at ways of reducing the cost of that cream bar, if we can do so without losing students. Maybe not all of the support does need to be done by a human. But I'm pretty sure some of it does.

This highlights the dilemma for universities – many students may not think they need that cream bar. It's like a universal credit, such as a state pension. Some need it more than others, but if you remove the principle of all paying into it, then it becomes prohibitively expensive for those who do need it. So the question that fees and MOOCs make both universities and students address is – how much do I value support? It's a profound question for the future direction of education, but I agree with Christopher Newfield when he argues that dreams of zero marginal costs in MOOCs are fanciful.

It all rests in that second column..

Leave a Reply