The nature of presidential leadership, and the question of how to judge its quality, is an ever-present issue, particularly in this age of constant media coverage and discussion of the presidency, whether on talk radio, cable news networks, on the Internet. Most of what passes for analysis, in this cacophony, is rapid-fire gut reactions to the latest presidential action.

Stephen Graubard, retired professor of history at Brown University, attempts a more systematic analysis of the modern presidency in "Command of Office." This book, which offers chapter-length analyses of each president since Theodore Roosevelt (who is arguably the first modern president), tries to demonstrate the massive power that has flowed into the Executive Branch of the federal government in a century dominated by the perpetual threat, and frequent realization, of war.

The resulting effort demonstrates years of research and study into the 18 presidents (it was published during George W. Bush's presidency). The wealth of material consulted is evident from the lengthy, and heavily annotated, endnotes. This frequently results in a presentation of some of the complex currents of public opinion and bureaucratic advice that shaped key presidential decisions.

However, the book is crippled by Graubard's persistent condescension. At every step, Graubard knows better than the men who held office. And while hindsight is usually considered 20/20, which explains how a scholar could comment with some confidence about many mistakes, "Command of Office" offers a tone that differs from the best history. Oddly, in exploring more than 100 years of presidential decisions, the author never seems to be pleasantly surprised by a decision -- in fact, the insights always seem to point to a gamut of actions that run from slightly better than mediocre to incomprehensible.

There is valuable material and insight in the book, which is what kept me slogging through it, but the tone in which the critique is almost always critical makes it a tiresome experience. ( )