GOVERNMENT WITHOUT BUREAUCRACY 53If there was a ‘real’ imperial civil service under the Principate, it was made up
slaves and freedmen of the emperor’s household. It was they who provided continuity,
institutional memory and cumulative experience. On the familia Caesaris , Weaver
(1972) remains fundamental; see also, briefl y, Morley (2011), 281–4; Mouritsen
(2011a and b), 93ff. Weaver thought that imperial slaves and freedmen at clerical
level or above numbered around 2000 (pers. comm.). In contrast, the equestrian
(and senatorial) offi cials who formed the superstructure of the administration were
relatively few in number. Eck (1995–1998), vol. 1, 15–17 estimates c. 160 senatorial
posts and c. 180 equestrian, under M. Aurelius and Septimius Severus, respectively.
Such offi cials served for short terms (usually one to three years). They were selected
on the recommendation of advisers of the emperor or their friends, and on the basis
of their general qualities and abilities rather than any specialist knowledge or
technical skills. The same applies to promotions, where, however, alongside personal
contacts, proven competence and experience would have come into the reckoning.
The army should be treated as a case apart (Eck 2007). For equestrians in the army,
see Devijver (1976–2001, 1992). Adams (2010) regards Egypt as also exceptional.
Eich (2005) argues that the administration developed some of the organizational
characteristics of a bureaucracy in the course of the Principate, and that this
orientation became more marked in the course of ‘the long third century’. Certainly,
through the period of the Principate, the numbers of salaried offi cials crept up,
administrative functions slowly increased, and a more structured hierarchy of offi ce
gradually evolved. It remains to be demonstrated that there was a palpable stepping-
up of these processes already before the close of the Principate, caused by fi nancial
problems arising from the plague of the time of Marcus Aurelius (Zuiderhoek 2009a
and b) and the increased military activity and costs of the Severan period (Eich 2005,
Bang 2013, 446). No exponential increase in the numbers of salaried functionaries
occurred, nor was the administration thoroughly overhauled, until the late third and
early fourth centuries, in response to the endemic warfare and serious dislocation of
the preceding half- century.
Fergus Millar and Werner Eck have made very substantial contributions to
scholarship in the area of imperial government and administration. In addition to
Millar’s books (see Bibliography), a selection of his articles are assembled in Millar
(2004). Eck (2000, in English) is a good recent introduction to the government of the
empire. See also Eck (1995–1998, 2001, 2007). Bang (2013) and Scheidel
(forthcoming) place the development of the imperial civil administration in its wider
setting; their explanatory analyses are the more convincing because of the added
comparative dimension. The emperors did not attempt to control their vast empire
by military occupation, nor to rule it with a large and intrusive civil administration;
neither of these possibilities was a realistic option. Rather, they sought the support
of a selection of aristocrats and major landowners in Rome, Italy and the provinces,
who were allowed to pocket a proportion of the provincial revenues. The penetration
of the state into provincial society and politics was limited, and did not markedly
increase under the Principate (Eck 2000, 290, Burton 1998 and 2001). Crucially, the
collection of the main taxes was left in the hands of the elite in the very numerous
autonomous cities in the provinces, under the supervision of the thinly spread state
offi cials, with the consequent ‘wastage’ of a percentage of the surplus. As a result,
taxes (both assessed and received) were at a relatively low level. And this, despite the
fact that the emperors, in addition to other expenses, had a substantial and costly
army to maintain – as large as any standing army in history until the time of