The economy is not a dominant element in Decline and Fall, and indeed the relevant evidence was almost entirely unavailable in the eighteenth century. What Gibbon chose to emphasize was “Religion and Barbarism,” but he also pointed to “the disorders of military despotism,” “the foundation of Constantinople,” and “the division of the monarchy.” These last three issues can usefully be taken first, before we turn to the two factors on which Gibbon came to place most emphasis.

Gibbon did not share Montesquieu's view that the Roman Empire had been despotic for much of its exis-tence. For him the Roman Empire did indeed become despotic, but not until after the Age of the Antonines, and, as his summary suggested, with regard to the internal problems of the Empire his emphasis was very largely on what he called “the disorders of military despotism.”

There is a great deal to be said for the argument that military matters were central to Rome's problems. The maintenance of the army was at the heart of much imperial policy. It underpinned the tax regime, which was extremely burdensome for most sections of the population. Not so for the aristocracy, which was expected to con-tribute rather more to the Empire through acts of munificence than through the payment of taxes (though they were not exempt from State exactions). To describe such acts the French have long spoken of Roman évergétisme (“evergetism”), borrowing a word from the Greek. Imperial income was in large measure dedicated to paying the army, which employed a huge number of men: by combining the evidence of John Lydus, who talks of an army of 389,704 soldiers and a fleet of 45,562 sailors in the days of Diocletian, with the narrative of the sixth-century Byzantine historian Zosimus, who refers to some 581,000 soldiers in the civil wars of 312, with the late fourth-/early fifth-century evidence for the distribution of officials and their legions to be found in the Notitia Dignitatum, its size has been estimated at between 400,000 and 600,000 men, out of an overall population of perhaps 55 million. None of these figures, however, is any more than a guess, and even if those given by John Lydus and Zosimus are accurate, they refer to a very specific period of time, while any estimate made from the Notitia depends on assumptions about the extent to which legions were kept up to strength.