Lindsay Stirton

Dear Professor Foskett,

I am writing to express my concern and disappointment at the proposed closure of PEAK, the Centre for Professional Ethics at Keele, with the absorption of some of its activities within the Law School. I would like to say first that I understand the pressure that institutions like Keele are under in the current funding envi...ronment, and I recognise that difficult choices may have to be made as to where cuts are to fall. At the same time, PEAK is undoubtedly one of the jewels of your University, and I can't help thinking that your analysis of Keele's priorities and strategy has somehow gone awry, however carefully the decision-making process has been undertaken. I know that you have received many e-mails expressing concern at the decision, and putting forward arguments in PEAK's defence. I doubt I have anything new to add to making this case, although I would like to express my belief that such arguments are ultimately sound. I will concentrate instead on expressing some possible scepticism about the way such decisions are made, and how you might respond to this scepticism.

I understand that the decision-making process has been informed by financial modelling of projected income and staff costs over the next few years. This is of course an essential part of business planning, but the robustness of such modelling exercises depend ultimately not so much on the data speaking for itself as from the judgements made by the analyst. Moreover, when such modelling informs the choice among rival proposals, another distorting factor is at play. Suppose that the modelling exercise produces a projection of (say) a financial position two years down the line, with the estimate including some degree of error (it would be inconceivable that the estimate would be spot-on). Now if the size of the error is large relative to the estimate, the decision that is made is likely to be a product of the error, rather than the estimate. In economics, a version of this is known as the 'winner's curse' and it explains a whole manner of things. Why do mergers rarely achieve their projected synergies? Because the company that succeeds in the take-over is frequently not the one which has made the mean error, whose expected value is zero, but the one which has most extravagantly over-estimated the synergies, and offered the wildest price to the taken-over company's shareholders.

With this scepticism in mind (and I speak not as someone opposed to such methods of decision-making; my current research in fact involves Bayesian statistical modelling of judicial decisions) can I now suggest how you can give confidence to yourself and to others that the decision is based not just on modelling but on sound modelling:

1. You need to give some reassurance about the soundness of your financial modelling generally. As a nice summary, out of all the projections about income, staff costs, student numbers etc., you have made in the last five years, what proportion have been within 10% of the actual realised figures? If this is not the overwhelming majority, then can you see the problem that people might have in having confidence in the decision-making process? If you cannot produce the figures I am asking for, then you are not adequately quality-assuring your own decision-making processes, and your decision-making process might equally fail to inspire confidence.

2. You need to make your projections and financial modelling (just of PEAK, not the wider university's finances which are understandably a matter of commercial confidence) available for scrutiny. You need to be able to state exactly the assumptions that went into them, so that alternative assumptions can be modelled and sensitivity analyses undertaken.

I hope you will undertake these measures forthwith, so that even those who disagree with your decision can have confidence in the decision-making process by which the decision was reached.

Best wishes,

Lindsay Stirton

Lecturer in Law