«Edited by ANNE MASON Research Fellow, Centre for Health Economics University of York and ADRIAN TOWSE Director, Ofﬁce of Health Economics Radcliffe ...»
Fiscal equity, traditionally, has focused on the distribution of income across individuals, independent of its sources and uses. Severe inequality has been considered undesirable, with progressive taxation one of the correctives. But equity may also be viewed in ‘categorical terms’, and applied to the availability of goods which society views to be of particular importance (or ‘merit goods’ as they have come to be known). Among them, the availability of health services has been prominent. Publicly provided health services or support of privately provided services are used to correct for inequality in the patient’s capacity to secure provision. Once more, there exists a link between ﬁscal and health economics.
A TRIBUTE TO OUR FRIEND AND COLLEAGUE, ALAN WILLIAMS 3Health services, like other goods, should be produced efﬁciently, and resources be spread across various diseases so as to maximise beneﬁts at the margin; and personal care should be given to reduce the burden of illness. All this matters, but just who should beneﬁt and who should pay? Equity enters throughout the health problem, across both its healing and care functions, but what is most needed is a formulation of the problem in manageable form.
The ethical and philosophical issues, such as entitlement to equal health, the right to ‘fair innings’ and the right to live or die, must also be faced at the end, and this is where Alan’s later work pointed. But the basics of measuring health, essential for the guidance of health policy, must be addressed ﬁrst.
Alan’s choice to pursue health economics was indeed a perfect match for his humanity, good sense and talents, which have been the key to his pathbreaking work.
CHAPTER 2 Citizens, consumers and clients: Alan Williams and the political economy of cost–beneﬁt analysis... Robert Sugden
INTRODUCTIONThis paper focuses on the work of Alan Williams, the cost–beneﬁt analyst, the predecessor of Alan Williams the leading health economist. In particular, it looks at Alan’s involvement in early debates about the proper role of cost– beneﬁt analysis (CBA) in the political process, about the relationships between the analyst, the agency commissioning a CBA and the wider community, and about the role of willingness-to-pay valuations in CBA. The issues that were involved in these debates remain relevant to current controversies about setting priorities in the delivery of health care – controversies that Alan continued to be engaged with until shortly before he died. The last time I met Alan – at a meeting in York in 2003 to discuss the methodology for a research project on ‘the social value of a QALY’ (quality-adjusted life year) – he was passionately defending a concept of ‘social value’ which can be traced back to the position he took in those early debates about CBA.
In trying to elucidate Alan’s position, I will use two main texts. (I am going back to an era when academic economists published much less than they do now.) The ﬁrst is the paper ‘Cost–beneﬁt analysis: bastard science? and/or insidious poison in the body politick?’, which Alan wrote in 1970, and which was published in the ﬁrst volume of Journal of Public Economics in 1972. The second is the book The Principles of Practical Cost–beneﬁt Analysis which he and I wrote together over far too much of the 1970s, and which ﬁnally appeared in 1978 (Sugden and Williams, 1978). I will do my best to reconstruct from a fading memory the discussions through which we arrived at agreed positions, and in this way to separate Alan’s views from mine.
56 THE IDEAS AND INFLUENCE OF ALAN WILLIAMS
COST–BENEFIT ANALYSIS AS MANAGEMENT CONSULTANCY: THE‘BASTARD SCIENCE?’ PAPER In striking contrast to the staid contents of more recent issues of Journal of Public Economics, Alan’s paper (which from now on I will refer to as ‘Bastard science?’) begins with a satirical verse written in the style of WS Gilbert, entitled The Management Consultant’s Lament. Alan reports that this verse was composed by a civil servant, JM Ross, as a response to Alan’s ‘own early activities’ in the ‘booming business of management consultancy’ (Williams, 1972, p. 199). As this starting point suggests, Alan takes it as given that CBA is a form of management consultancy. His aim in the paper is to identify the features which distinguish CBA from other forms of consultancy, and to defend CBA against two charges: that it is a form of pseudoscience, and that, by substituting economic analysis for political deliberation and negotiation, it usurps democratic processes.
Alan starts from the premise that the objective of CBA is ‘to assist choice[,] not to make choice’. He says that CBA is ‘a natural and logical extension’ of other techniques which serve the same purpose, such as systems analysis, operations research (OR) and cost-effectiveness analysis (Williams, 1972, pp. 200–201).
Viewed in this perspective, CBA has two distinguishing features.
First, CBA is ‘prescriptive’. Alan explains that, by this, he means that the analyst plays a ‘consultative role’; the analysis is ‘designed to help a client improve his situation’ (p. 202). It is signiﬁcant that the role of the analyst is deﬁned in relation to that of a client: the client, presumably, is some agency (in the context of CBA, an agency of government) which is seeking assistance in making choices. The analyst’s prescriptions take the form of advice about how the client agency can best achieve its own objective. This does not mean that the analyst’s attitude towards the client has to be completely uncritical. Alan points out that clients often lack a clear sense of the nature of ‘the problem’ that they are asking the analyst to solve: ‘any OR or CBA practitioner who accepts the client’s initial formulation of the problem uncritically is heading for disaster’ (Williams, 1972, p. 204). Still, it is clear that the analyst’s role is to help the client to conceptualise its problem, as it perceives it, and not to prescribe what the client should be aiming to achieve.
The second, and most distinctive, feature of CBA is that it ‘[tries] hard to render inputs and outputs commensurable’. In its ideal form – an ideal that is not currently attainable – ‘CBA will have all inputs and outputs evaluated in money terms’ (Williams, 1972, pp. 200–202). However, Alan is at his most forceful in insisting that this does not mean that valuations should be based on individual willingness-to-pay.
Using Aaron Wildavsky as a spokesman for the claim that CBA is an insidious
poison in the body politick, Alan quotes the following comment on CBA:
Alan sees this as a gross misrepresentation: ‘This is so patently false that if I had not had an opportunity to read and re-read this statement carefully I would not have believed it possible that such a well-informed and normally judicious observer could have made it’ (Williams, 1972, p. 214). Alan accepts that ‘much of the analysis that goes into most cost–beneﬁt studies is based on the assumption that the market does constitute a good guide to the value of costs and beneﬁts’; but he claims that the practice of CBA allows this assumption to be relaxed. Some of the relaxations he describes seem to me to be in the spirit of Wildavsky’s account of CBA. For example, Alan says that CBA ‘may require... valuations inferred from market behaviour to be placed on items which have no market values’; but this is a case in which the government is eliciting preferences, not dictating them. Alan’s crucial objection is that CBA ‘may require... the use of valuations postulated by the policymakers or decision-makers themselves’ (Williams, 1972, p. 214).
The legitimacy of such postulated values is a recurring theme in the paper.
For example, he accepts that if ‘a client (say a Minister of Transport) [says] that for the purposes of this analysis I want variable X valued at £Y per unit (where X might be the intrinsic value of life)’, then ‘£Y represents a precise statement about the client’s policy trade-off between X and the other variables in the analysis’, and so is a legitimate datum for CBA. This kind of postulated value is ‘arbitrary’ in the sense that it is not supported by evidence, but it is not ‘irresponsible’ because ‘it does imply a commitment to allocate resources in a particular way for which the client will have to answer’, and ‘it provides a precise focus for discussion by others (say Cabinet colleagues) who might wish to inﬂuence his judgements on this matter or modify their own’ (Williams, 1972, p. 210). Notice that, in this account, postulated values are presented as internal to the government machine. There is no suggestion of public debate about these values, apart from the hint that the government may ultimately have to answer to its electorate.
Here is how Alan sums up his position on the status of market-based and
postulated values in CBA:
I am doing no more than accepting the well-recognised proposition that you cannot ascribe values without making value judgements. Market prices are acceptable if the value-premises underlying market behaviour are acceptable (e.g. accepting the prevailing distribution of economic power and the relevance of individual market-oriented valuations to the making of social judgements)... Postulated prices are acceptable if it is believed that the value-premises
8 THE IDEAS AND INFLUENCE OF ALAN WILLIAMS
The reader is entitled to ask: Who is making the value judgements here? To whom are postulated prices acceptable or unacceptable? The logic of the paper as a whole seems to allow only one answer: the client.
Summing up: Alan’s position in ‘Bastard science?’ is that CBA is a par ticularly ambitious form of management consultancy. It is scientiﬁc in the sense of being systematic: ‘assumptions are required to be explicit, evidence presented, results communicable and replicable’ (Williams, 1972, p. 203). It does not usurp the political process because it does not claim to make decisions and does not impose any substantive value judgements; it merely assists clients to formulate their own objectives and to pursue them effectively.
TWO OPPOSING VIEWS OF COST–BENEFIT ANALYSIS: THE PRINCIPLES OF
PRACTICAL COST–BENEFIT ANALYSISBefore discussing the text of The Principles of Practical Cost–beneﬁt Analysis (Sugden and Williams, 1978), I need to explain how the book came to be written, and Alan’s and my respective roles as co-authors.
My becoming an economist was very largely due to Alan. I was an undergraduate at York between 1967 and 1970, studying history and economics. The topic in economics that most interested me was CBA, which I learned about from Alan’s lecture course on investment appraisal. For my ﬁnal-year project, I chose to do a retrospective cost–beneﬁt study of the recent closure of the Whitby–Scarborough railway. Alan was my supervisor, and went far beyond the requirements of duty in helping me to make contact with various agencies that were able to provide me with data I needed. He had gently suggested to me that the plan of carrying out a complete cost–beneﬁt study was over-ambitious for one student over a summer vacation, but I wanted to (and, I think, did) prove him wrong. While reading the reports of previous cost–beneﬁt studies of rail closures, I had my ﬁrst genuinely original economic idea: I realised that the methodology that had been used by the Ministry of Transport in its most recent rail-closure study made a theoretical error in its treatment of changes in fare revenue. When I told Alan that I had discovered this error, I sensed that he was sceptical, but he listened to my argument in his characteristically openminded way. After some thought, he agreed that I was right. He encouraged me to write a paper on the subject and send it to the Bulletin of Economic Research; this became my ﬁrst publication (Sugden, 1972). The following summer, I had my ﬁrst paid job as an economist, working for Alan on a project he had persuaded the University of York to ﬁnance, investigating whether
CITIZENS, CONSUMERS AND CLIENTS 9the apparent scarcity of private-sector student accommodation in the city could be reduced by making suitable adjustments to bus routes. (This is a good illustration of Alan’s conception of the creative role of the analyst. The university perceived its problem as a shortage of student accommodation in areas accessible to the campus; Alan saw that the problem could be restated as the poor accessibility of otherwise suitable areas).
After a year studying for a master’s degree in Cardiff, I returned to York in 1971, as by far the economics department’s youngest and least-qualiﬁed lecturer. Shortly after this, Oxford University Press invited Alan to write a textbook on CBA. He had a very clear idea about the kind of book that was needed – an idea expressed in the apparently paradoxical title he chose for it.