regime change (including possible future democratization), we must understand the different,
and in some respects new, types of authoritarian rule”. Until recently, however, the utility of
this new shift in focus has not been systematically assessed. There is still a want for evidence
that distinguishing among types of authoritarian regimes really pays any dividends (Brownlee
2004, 4; Geddes 1999, 121; Snyder and Mahoney1999, 108).
In this paper we want to add to this body of knowledge. More precisely, we present a new
typology of authoritarian regime types, based but also considerably improving on Geddes’
(1999) seminal contribution. Second, we explore the extent to which this new typology helps
explain the survival (and breakdown) of non-democratic regimes. Third, we assess the impact
of authoritarian regime types on democratic development. Our results strongly confirm the
expectation that different authoritarian regimes face different propensities to survive and to
develop towards democracy. These results hold even in tests, where a broad number of
control variables are taken into account. Hence, this institutional attribute
the nature of
authoritarian regime
deserves, in our judgement, to be added to the list of essential
preconditions of democracy.
Previous Efforts to Classify Types of Authoritarian Regimes
The classical theories on non-democratic regimes developed during the 1950s and 1960s were
primarily based on a distinction between totalitarianism and authoritarianism. With an
emerging awareness that hardly any regime fit the totalitarian category, whereas the
authoritarian type was too inclusive, this typology however soon grew obsolete (Brooker
2000). To remedy this situation, Linz and Stepan (1996) develop a refined typology that add
the categories “post-totalitarianism” and “sultanism” to the classical distinction. They then
link these categories to a model of the “transition game” played by actors involved in the
4
Convention