Fortin 1
Weak States, Strong States: Comparing Postcommunist Democratization
Scholarship on comparative institutions has produced many fertile hypotheses
concerning the role of institutional choices in fostering stable democracies (Lijphart and
Waisman, 1996; Diamond and Plattner, 1996; Merkel, 1996; Crawford and Lijphart,
1995). From this institutional perspective, it is conjectured that some constitutional
arrangements have better records for democratic consolidation than others. One of the
liveliest parts of the institutional debate is framed by the contrast between parliamentary
versus presidential systems of government (Easter, 1997; Frye, 1997; Horowitz, 1990;
Linz, 1996; Lijphart, 1991, 1992; Mainwaring and Shugart, 1997; Przeworski et al.,
1996; Stepan and Skach, 1993; Shugart and Carey, 1992). Even if the correlation
between strong (constitutional) executives and autocracy also is evident in the
postcommunist context, this body of literature was harshly criticized for lingering on
shallow causes and for the interchangeable direction of the causal relation under study.
In fact, a very high probability of endogeneity exists, even spuriousness, if there is a
common factor such as state capacity driving parallel variation in both institutional
choices and democracy. Although the direction of the causal relation needs to be further
theorized, in many scholars view, it is not uncommon to see strong executives emanating
from weaker state structures (Easter, 1997; Migdal, 1988; Young, 1994).
In my view, patterns of state capacity, before institutions, offer a crucial
explanation into the types of regimes—democratic, semi-authoritarian, and
authoritarian—that emerged after the breakup of the Soviet Union. However compelling,
this proposition was never rigorously tested in the context of postcommunist countries.
For this reason, by means of different types of statistical modeling, the present research
proposes to evaluate how well these hypotheses reflect the reality of postcommunist
transformations. First, cross-sectional estimations using both static and dynamic
specifications will be performed. Following this, more ambitious time-series cross-
sectional (TSCS) estimations will be performed on 21 countries from 1989 to 2003.
1
1
The twenty-one countries included in the present study are post-communist states that gained
independence between 1989 and 1991: Armenia, Azerbaijan, Belarus, Bulgaria, Czech Republic, Estonia,
Georgia, Hungary, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, Moldova, Poland, Romania, Russian
Federation, Slovakia, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan. Mongolia was omitted for lack
of valid data. Countries from the ex-Yugoslavia also were omitted because they were not part of the Soviet
empire.
Convention