The spy in the Oval Office

Each of the books under review is an ambitious project. Christopher Andrew surveys the entire intelligence history of the United States, and Jeffrey T. Richelson attempts a world overview to supersede Constantine FitzGibbon's Secret Intelligence in the 20th Century (1977).

Richelson has already written books about Russian and American secret intelligence, and has the additional advantage of being a senior fellow of the National Security Archive in Washington DC, a private organisation dedicated to the declassification and publication of hitherto secret US government documents. Perhaps because of his familiarity with one country's documents, he supplies in A Century of Spies a first-and-second-world perspective on the intelligence history of the 20th century. Thus he takes the reader through a familiar agenda, if in greater detail than FitzGibbon. His progress through American and European experiences in the two world wars, the interwar years, the cold war and "the new world of disorder" beyond is nevertheless defensible, in the sense that third-world countries are still too backward to uphold civilisation by uncivilised first-and-second-world means.

Richelson's book has its strong points. For example, it has no truck with the notion that the second world war was attributable to the inadequacies of pre-war secret intelligence - that shaky premise predicated on its false conclusion, the dogma that advanced nations need vast intelligence apparatuses. Instead, while acknowledging the Barbarossa and Pearl Harbor mistakes, he offers a balanced appraisal of British and American intelligence in the 1930s. Another strength of Richelson's book is that he recognises the importance of intelligence in the constructive area of the verification of arms treaties - a subject indefensibly ignored by some popular historians who deal with derring-do and salacious scandals to the exclusion of all else.

Richelson is not, however, incapable of errors of fact and judgement. Both he and his publisher should have realised that Britain's interwar codebreaking supremo Alastair G. Denniston could not have "played in the Scottish Olympic field hockey team" - and it is strange that an American writer should be uncritical of the self-inflating claims made by Herbert O. Yardley, Denniston's 1920s equivalent in the US, an unstable character who sold his Black Chamber's secrets to the Japanese government.

Andrew displays a healthier scepticism: "Yardley does not . . . record any interest in the Black Chamber's work by President Harding. Had there been any, Yardley would surely have mentioned it." Andrew, professor of modern and contemporary history at Cambridge University, has written about Russian and French intelligence history and is the author of Secret Service (1985), the standard general history of British intelligence. For the President's Eyes Only compares with the latter work, and confirms Andrew's position as the leading international authority in his field.

Andrew's book displays the erudition, wit and style we have come to expect from him. As ever, he supplies some thought-provoking insights. For example, he traces Franklin D. Roosevelt's Anglo- philia in intelligence matters to the days when Admiral "Blinker" Hall of Room 40 fame dazzled the young FDR, then assistant secretary of the navy, with tales of British brilliance in the great war.

However, For the President's Eyes Only will have to compete with other general histories of US intelligence by Charles D. Ameringer, Nathan Miller and G. J. A. O'Toole, each with their particular virtues. Does the Andrew volume have any advantages over these apart from being (in spite of the naughty omission of Ameringer and Miller from its bibliography) the most recent and up-to-date contribution in the field?

Coming from a British reviewer this may be a subjective plug for objectivity, but the answer is yes, on the ground that Andrew is sufficiently removed from the American scene to be able to stand above the local fray and coolly offer a distant perspective. As the title of his book suggests, Andrew's governing thesis is that American intelligence is a presidential matter. Not for him the 101 inanities about spy plots that have changed the world, or the discredited but pervasive theme of the CIA as a "rogue elephant rampaging out of control". He grasps and exploits the constitutional point that the president is in charge of foreign policy, including foreign intelligence. The strengths and weaknesses of intelligence policy stem, as Andrew unsparingly demonstrates, from the qualities of the White House's incumbent.

Rhodri Jeffreys-Jones is reader, department of history, University of Edinburgh.