with less than full democracy. Consider, for example, the reference by
Diamond to political restructuring in Malaysia, which has limited political
competition and restricted freedom of expression: ‘While this restructuring
has levelled parliamentary democracy down to semi-democratic status, it has
also brought considerable ethnic peace, political stability and socioeconomic
prosperity’ (1989, p. 2).
It has also to be recognized that stable democracy might be replaced by
stable authoritarianism or even a stable regime which may not conform to
some definitions of democracy, particularly those entailing multi-partyism,
but which would be regarded by few as authoritarian. The failure of democ-
racy should not be equated with the breakdown of stable government. It may
simply be that a particular definition of democracy determines the scale of
the failure, as in Diamond’s analysis of the ‘failure’ of democracy in Sub-
Saharan Africa where, by the early 1970s, multi-party democracy had disap-
peared in all but Botswana, Gambia and Mauritius (Diamond, 1988, p. 5).
Handling the concept of stability is made difficult by the fact that it is
often used in a way which fails to distinguish between countries that have
experienced the overthrow of a democratic regime and its replacement with
a stable but authoritarian one, and countries where there is constant civil
disorder and change of regime. Much of the concern for the health of
democracy over the past 30 years has been prompted by tendencies to ‘drift
away’ from democratic standards rather than a drift into instability, though
the two trends are usually closely related. Huntington fell into this trap
when he cited ‘going communist’ as evidence that Cuba and the Indian State
of Kerala were unstable (Huntington, 1965, p. 406).
Thirdly, there is the problem of finding a satisfactory operational defini-
tion of ‘stability’. Various indicators have been employed, such as the
turnover rate for chief executives, deaths from internal group violence per
million population, and the total number of violent incidents (Russett,
1964). Attempts have been made to measure the aggression of groups
within the political system, the longevity of governments, and the constitu-
tionality of governmental acts (for reviews see Hurwitz, 1973 and Ake,
1974). None of these completely captures a sense of what is involved in
political instability. Even when political structures change, it may not nec-
essarily be destabilizing. Ake’s formulation of instability in terms of mem-
bers of society deviating from ‘the behaviour patterns that fall within the
limits imposed by political role expectations’ has the advantage of extend-
ing the types of events that are to be considered beyond the élite interactions
which are conventionally regarded as indicators of political instability –
coups, electoral violence, political assassinations and so on – as well as
222 Understanding Third World Politics