Research Design
In the following paragraphs, I will discuss the methods that will test the hypothesized relationship between judicial independence and human rights in African countries. Next, I will conduct the necessary statistical analysis and provide both a substantive and an empirical discussion of the results. Table 1 displays a summary of the complete data set. A cross sectional analysis is conducted. This analysis comprises an OLS regression of human rights in 2008 on the measures of judicial independence in 2006 in thirty-one countries. I anticipate a lagged relationship between independent and dependent variables. For this reason, the independent variables are lagged for two years.
Personal Integrity Abuse Model
To follow precedence, human rights will be labeled as the “integrity of the person” (Poe and Tate 1994, Cingranelli and Pasquarello 1985; Henderson 1991, 1993; Mitchell and McCormick 1988; Stohl and Carleton 1985). The abuse of an individual’s personal integrity will be referred to as violations of a person’s human rights. Examples of acts that invade a person’s human rights are torture, imprisonment, and murder committed by the state.
Those actions (torture, imprisonment, and murder) used to define human rights are not meant to diminish other elements of human rights. Other components that relate to the broader definition that includes economic rights and political rights are still important. On the other hand, the focus of this paper is on what is believed to be the most “egregious and severe crimes against humanity” (Poe and Tate 1994), and judicial independence is a factor that can effectively decrease the number of severe crimes committed within a society.
The measurement of human rights is a composite of rankings from the Amnesty International Reports and the U.S. State Department. Countries were rated on a five-point ordinal scale for human rights, with measurements of repression ranging from 1 to 5. A country represented by the number 1, like Guinea-Bissau, signifies that it is “under a secure rule of law” (Gibney, Cornett, and Wood 2011). A country represented by the number 5 is considered a society filled with terroristic behavior.
Measurement of Judicial Independence
Countries in Africa tend to share a similar background and culture, yet there is considerable variation in the occurrence of human rights violations. The focus on African populations comes from curiosity about the factors that cause such variations. For my research, I developed a sample of 31 African countries. For the measurement of judicial independence, I use the Bertelsmann Transformation Index (BTI), which is measured on an ordinal scale from 1-10 (1 being the lowest and 10 being the highest). The BTI is an international ranking of 125 developing and transitioning countries. This index is a useful source for this project, because its measurement of judicial independence aligns with the way I analyze judicial independence. Bertelsmann categorizes judicial independence as the freedom from external influence, and ranks developing and transitioning countries in accordance with this definition.
Control Variables
Scholars have found that democracy decreases human rights violations (Henderson 1991). Referring to Bollen’s definition, democracy is defined as “the extent to which the political power of the elite is minimized” (1980, 372). A country’s standard of living and political participation rate are important factors in discussing how a country protects its citizens’ rights. This justifies the use of democracy and GDP per capita as control variables. I used the Polity2 measure in Polity IV (retrieved at http://www.systemicpeace.org/polity/polity4.htm on September 8, 2011) to control for regime type. It is a common belief that economically balanced countries protect the rights of citizens. To control for the economic well-being of a country, GDP per capita (World Bank 2011) is measured.
Methodology
An OLS analysis is conducted, because it is an appropriate method given the measurements of the dependent variable. It is usually recommended to conduct an ordered probit or logit for a dependent variable ranging from 1-5. Instead, I followed the precedent set by Poe and Tate and used regression. The same methodology is used because of the similar measurement of the dependent variable. Due to limitations of data, my empirical research will not be time series. I also limit my analysis to just 31 of 50 African countries. To improve empirical studies on this topic, an advanced systematic measure of judicial independence is needed.