8
and five is the most repressive regime.
2
For 1997, I use the Political Terror Scale guidelines to
code Amnesty International and State Department Reports. Like Blanton (1999a), I average the
two scores. The scores are closely related to one another, as they are moderately correlated
(0.736). However, the averaging of these two scores may be problematic for two reasons. First,
the two scores may yield different results due to bias. Poe, Carey and Vazquez found the State
Department and Amnesty International reports, “have clearly converged in their assessments of
human rights violations over time” (2001, p.677). However, researchers should be cautious; Poe,
et al., hypothesize that a new bias towards US trading partners may have emerged in the State
Department reports. Second, “the scale requires ordinal judgments - distances between levels are
not equal but a country at level 1 is doing better than a country judged to be at level 2” (Political
Terror Scale website). Therefore, it may not be appropriate to average the two scores. To avoid
these problems, I also analyze my hypothesis with the two measures of repression individually.
Political and civil rights
To measure civil and political rights, I use Freedom House’s Freedom in the World Country
Rankings.
3
Freedom House ranks countries on their level of civil and political rights. Political
rights and civil liberties are each measured on a one to seven scale; where one is the highest
degree of freedom and seven the lowest. To create an increasing scale, I inverted the scores so
that one is the lowest and seven is the highest degree of freedom.
Social and economic rights
To operationalize my measure of social and economic rights I employ the Physical Quality
of Life index gathered by Morris (1979, 1996) in conjunction with the Overseas Development
Council. Three dimensions are captured in this index; countries’ performances are evaluated on
three dimensions: infant mortality, life expectancy at age one and literacy at age 15. These factors
are placed on a scale from 0-100 and the mean of the three is calculated.
4
Milner, et al., (2000),
based on Morris’s work (1979), argue that infant mortality and life expectancy at age one tap the
Convention