Ministers take the biscuit

What do the Duke of Grafton, Lord Iddesleigh, W. H. Smith and A. J. Balfour have in common? Which 20th-century prime ministers preceded Blair in appointing a minister without portfolio outside the cabinet? How many former barristers, other than Jack Straw, have held the office of home secretary in the 20th century? Could a Jew become lord chancellor? The answers are all in Ministers of the Crown: all four held the post of first lord of the Treasury when that office was separated from that of prime minister; Winston Churchill, Harold Wilson and Edward Heath appointed ministers without portfolio; nine home secretaries were barristers - George Cave, John Simon, Donald Somervell, David Maxwell Fyfe, Frank Soskice, Leon Brittan, David Waddington, Clarke and Howard - not a particularly inspiring collection.

Rodney Brazier's aim, however, is not to assist question setters for University Challenge, but to lay out the law, constitutional conventions and political practice relating to his subject, and in this he is conspicuously successful. Ministers of the Crown comprises six sections. The first deals with the organisation of government and the history of the main governmental offices such as the lord chancellor and the first lord of the Treasury. From it, we learn that the title "prime minister", derives, like "Cabinet", from the French. After the death of Louis XIV, a premier ministre was recognised by royal decree, with other ministers being declared subordinate to him. In Britain, Walpole was given the sobriquet premier ministre by opponents, who sought to mock his dominance over his ministers. Walpole's only formal ministerial title, however, was first lord of the Treasury, and it ranked below that of several other ministers, including the lord chancellor. The title, prime minister, was not used in an official document until 1878 when Disraeli, as Earl of Beaconsfield, signed the Treaty of Berlin as "First Lord of the Treasury and Prime Minister of her Britannic Majesty". But the prime minister was not given official recognition until 1905 when the king accorded the office precedence after the Archbishop of York.

Brazier's second section analyses how politicians prepare for government, distinguishing sharply between the Conservative shadow cabinet, which has never been made subject to formal rules and whose members are chosen at the personal discretion of the leader, from the Labour shadow cabinet, whose membership and method of working are subject to formal party rules.

He then considers the legal and constitutional qualifications for ministerial posts. There seem not to be any other relevant qualifications. On being appointed first lord of the Admiralty in 1916, Sir Edward Carson remarked: "My only qualification for being put in charge of the Navy is that I am all at sea." Then, as now, being a professional politician meant being an amateur minister.

The third section considers ministers at work, how government decisions are made, and conflicts of interest. Then Brazier analyses the impact of the law on ministers, providing a highly technical chapter on the use of the prerogative. The final sections deal with accountability and the consequences and implications of loss of office.

Brazier is thoroughly at home among constitutional minutiae, his understanding buttressed by a knowledge of mountains of precedents. Like Ivor Jennings's Cabinet Government, which it in some respects resembles, Ministers of the Crown will become an authoritative work of reference. Yet, if Brazier has the knowledge of a Jennings, he lacks the sweep of a Bagehot or a Dicey, that is, the ability to peer behind the precedents to see what they tell us of the fundamental principles of the constitution. Thus, while, to change the metaphor, Bagehot's English Constitution resembles a richly layered cake, reading Ministers of the Crown is very much like eating dry biscuits.

Brazier tends to shy away from deeper constitutional issues. His chapter on the civil service code, introduced in 1996, for example, is superficial. He does not consider how the code distinguishes between what is a party political purpose and a governmental purpose. Would it, for example, be appropriate for a minister to ask a civil servant to assist with backbench amendments to bills? Neither the code nor Brazier offers an answer.

Brazier's account of the new appeals mechanism is too charitable. A civil servant who believes that he or she is being asked to do something unethical now has recourse to an external mechanism, an appeal to the civil service commissioners. But this recourse is only possible once the internal machinery - an appeal to the permanent secretary of the department, and then to the head of the civil service - has been exhausted. It would be a bold civil servant who would use this internal machinery. The old machinery was used only once between 1988 and 1996. This could indicate that relations between ministers and civil servants were satisfactory throughout the period. But there is an alternative explanation.

Brazier's discussion of ministerial accountability is also disappointing since it fails to get to grips with recent developments, such as the growth of executive agencies and the Scott report on arms to Iraq, which raise the question of whether the convention of individual ministerial responsibility is a constitutional fiction. Brazier mentions the fashionable distinction between policy for which a minister will accept responsibility, and "operations", which are the responsibility of officials.

This distinction was frequently used by Michael Howard, as home secretary, to explain why he was not responsible for prison escapes. It is a pity, however, that Brazier does not discuss the value of this distinction. Derek Lewis, the former director-general of the prison service who was sacked by Michael Howard, once complained that a decision was operational if it was difficult; for it cannot be a matter of policy that things go wrong. Therefore, when they do go wrong, the fault must be operational and the blame attached to civil servants. The distinction between "policy" and operations, therefore, seems designed to allow a minister to shuffle off responsibility on to his or her officials.

A similar criticism may be made of Sir Robin Butler's distinction between a minister's constitutional accountability for everything that occurs in a department, and his or her responsibility for things that have gone wrong which are in fact the minister's fault. For this distinction, too, may serve to attenuate the responsibility of ministers almost to vanishing point, We used indeed to be assured that ministers are not subject to vicarious responsibility, and so do not have to resign when their officials have made a mistake. It would be good to be equally reassured that civil servants are not subject to vicarious responsibility, so that they do not have to take the blame for the mistakes of ministers.

Despite weaknesses, Ministers of the Crown is an important work of reference in the unfashionable field of constitutional law. It will be of great value to all those seeking to get to grips with our untidy constitution.

Vernon Bogdanor is professor of government, University of Oxford.