Are dictatorships returning? Revisiting the ‘democratic rollback’ hypothesis

Abstract

Since 2007 an increasing number of articles have diagnosed ‘freedom in retreat’ and predicted a ‘return to the authoritarian great powers’. Highly distinguished scholars warn against the ‘democratic rollback’, and articles on the resilience of authoritarian regimes have appeared in the best journals of political science. Is the tide of democratization turning, and do we have to expect a new reverse wave of autocratization? This article argues that there is no hard empirical evidence that we are witnessing a trend towards re-autocratization on a global scale. The optimism of the early 1990s of a seemingly irresistible trend towards democracy is partially due to an empirical artefact caused by inappropriate underlying theoretical concepts. The overestimation of human agency and political crafting on the one side and underestimation of structural impediments for democracy on the other side contributed to this optimism, as did the thin concept of ‘electoral democracy’ or teleological speculation about the end of history. Democratic rollback does not seem to be as widespread as is sometimes claimed.

Keywords:

• theories of regime change
• path dependency
• democratic rollback
• defective democracy
• prerequisites of regime change

Notes

O'Donnell and Schmitter (1986, p. 3) explicitly spoke about ‘transition to something else’, which is all, but not a teleological language.

Particularly ardent proponents of this were Przeworski (1986, 1991) and Di Palma (1990).

In addition, other indices of democracy such as the Bertelsmann Transformation Index show a mirror-reversed image of development, i.e. slight gains for democracy since 2006 (Bertelsmann 2006, 2009).

This saying is attributed to such different personalities as Karl Valentin, a Bavarian comedian, or the Danish physicist, Niels Bohr.

Of course, these are just statistical approaches to the question of democratic stability. They have to be contrasted and compared in each individual case with respect to the case's destabilizing factors (and potentially existing stabilizing factors which may compensate for the former).

It is curious to note that Przeworski (1986, 1991) sometimes appear as a hard core rational choice theoretician and sometimes as a proponent of modernization analyses. As brilliant as both strands of analyses are, they always seem to be rather unconnected.

Geddes uses data that cover the period from 1946 to 1998. Most single-party regimes of this period were communist party dictatorships that lasted, as a rule, from 1946–1948 until 1989–1991. This historical peculiarity must be assumed to strongly influence statistic calculations. Neither before 1946 nor after 1998 can we expect to find similar figures. As a consequence of this selection bias, the generalizability of her persistence-hypothesis is much lower than the author suggests.

Moreover, we have to differentiate more between ‘bureaucratic modernization’ regimes such as those of, e.g. Latin American on the one hand, and pure enrichment-regimes run by gangster militaries in Africa on the other.

A notable example is Brazil where during the last 10 years the inequality of income has decreased as has the still high number of the poor and marginalized.

Additional information

Notes on contributors

Wolfgang Merkel

Wolfgang Merkel, born in 1952, is Director of the Department of Democracy Research at the Social Science Research Center Berlin (WZB). His research focuses on political regimes, democracy, and transformation, parties and party systems, comparative public policy, social justice and reform of the welfare state. He is a member of the Berlin-Brandenburg Academy of Sciences and Humanities. Recent books include: System Transformation (2010); Social Democracy in Europe (2008); Defektive Democracies, 2 vols (2006, with others); Special Issue of Democratization: War and Democratization: Legality, Legitimacy and Effectiveness (2008, with Sonja Grimm).