This book deals with administrative reform. The main idea is that the process of change is typically not a process of validation and disproof using hard data, but rather a rhetorical process of persuasion through doctrines. Doctrines are what-todo-ideas come somewhere between 'policy' and 'theory'. Though doctrines are multiple and often contradictory, they are not infinite in number. The commonest them recur with striking regularity in spite of changes in terminology, and they can be collected and catalogued.

The authors discuss six factors essential for acceptance of administrative doctrines: the idea of symmetry means to produce linguistic 'solutions' which are exactly symmetrical to the social 'problems' experienced by the audience. Where final proof is impossible, persuasion needs to be achieved by the correct choice of metaphor. A third key to acceptance is held to be ambiguity, the ability to speak simultaneously

to groups with divergent interests. Fourth, successful persuasion demands that private benefit must be justified in terms of public good. Fifth, a rhetorical argument needs to be highly selective in its choice of maxims and arguments to fit the desired conclusion. The last key to acceptance is suspension of disbelief by those to whom the argument is addressed.

The authors then go on to record administrative doctrines from major textbooks in administration, plus a set of reform documents and historical works. They end up with 99 doctrines categorized as 'who-type doctrines' (getting 'the best sort of person' in office), 'what-type doctrines' (what type of organization to choose) and 'how-type doctrines' (what procedures or methods should be used). All the doctrines are contestable in the sense that a contradictory doctrine can always be found. The doctrine 'prefer experienced hands' and the doctrine 'prefer raw recruits' are for instance coexistent, as are the doctrines 'have inclusive responsibility by vertical integration' and 'have divided responsibility by administrative pluralism'. What remains puzzling is why some of these doctrinal keys open the lock of acceptance so readily in some times and places, and why the door remains firmly shut to them in other circumstances.

Three reports of administrative reform are then analyzed to discover empirically how the acceptance of particular doctrines is to be explained. The six elements identified above as possible clues to the 'acceptance factor' seem to give us some purchase on understanding the three cases.

Special attention is devoted to the use of metaphor and fiction to achieve persuasiveness. The authors hypothesize that the persuasiveness of economics and management science (rather than, say, traditional Public Administration) in administrative argument can be explained by their metaphors rather than their truths.

Without referring explicitly to it, Hood and Jackson's approach seems to have much in common with the 'institutionalized environments approach' in organization theory. The basic idea is that organizational forms are adopted according to whether they fit into a larger historical lock (p. 195), or broader political and social context (p. 151), or reflect social rules, customs and habits. The perspective is clearly in opposition to a design perspective where the basic idea is that institutions are structured deliberately according to explicitly stated collective goals.

Hood and Jackson present their approach as a post-positivist alternative (or additional line of analysis) to the positivist Simonian tradition in politico-administrative In their opinion, more than 40 years of research concentrating on the link between administrative design and administrative performance have had only limited success: 'We seem to be little nearer to "scientific" answers about when to use contradictory principles of administration than was Aristotle two thousand years ago' (p. 20). My first comment is that surprisingly few studies have been made explicitly on the relationship between structure and performance (Egeberg 1992). The reference to the 'contingency theory' literature in this respect (p. 20) is almost irrelevant, since in these studies structure is normally treated as a dependent variable, contingent on characteristics of the environment, organizational technology, age, etc. This impression of the 'state of the art' is shared by Hammond (1990), who suggests that Simon's criticism of Gulick, together with the impression which took root that Simon won the duel in the 19405, can be the reason why there is almost a complete lack of empirical studies on the effects of alternative structural forms.

My second comment is that even if empirical research is able to reveal understandableconsequences
alternative designs, the principles are still contradictory,

in the sense that one necessarily has to choose among the principles, depending on what one wants to achieve. For instance, if research documents that specialization by purpose or by area makes a difference to policy content (as was Gulick's argument), then one's choice of organizational solution should be derived from one's policy objectives.

Hood and Jackson's book is an important contribution to the literature on administrative reform. It is very well organized, and it draws on classic political theory in a fruitful way. The presentation of the 99 doctrines and their justifications is very useful, and may be seen as a set of hypotheses to be tested on the relationships between structure and performance, and between personnel and performance, respectively. Its perspective on reform processes creates a sound scepticism towards the administrative fashion trade of consultocracy and pop management.

REFERENCES

Egeberg, M. 1992. 'Constructive Political Science and Administrative Policy1. Oslo: Department
Political Science Working Paper.

Hammond, T. H. 1990. 'In Defense of Luther Gulick's "Notes on the Theory of
Organization"', Public Administration 68, 143-173.