Human Rights Violations, Corruption, and the Policy of Repression

• Published in: Policy studies journal Vol. 36; no. 1; pp. 1 - 18
• Main Authors: Bohara, Alok K, Mitchell, Neil J, Nepal, Mani, Raheem, Nejem
• Format: Journal Article
• Language: English
• Published: Malden, USA Malden, USA : Blackwell Publishing Inc 01.02.2008; Blackwell Publishing Inc; Policy Studies Organization; Blackwell Publishing Ltd
• Subjects: Bureaucracy; Comparative analysis; Compensation; Corruption; Democracy; Government regulation; Human Rights; Implementation; Influence; Interpretation and construction; Laws, regulations and rules; Motivation; Political Factors; Political persecution; Political systems; Politics; Public officials; Repression; Repression (Political); Social policy; Torture; Violations; Violence; War; United States
• ISSN: 0190-292X; 1541-0072
• DOI: 10.1111/j.1541-0072.2007.00250.x

Quantitative cross-national research on human rights violations and repression has made considerable progress in identifying and eliminating economic and political factors that influence the use of torture and killing by governments. Warfare tends to increase violations, democracy--notably full democracy--and trade tends to inhibit violations. Where motives have been considered, this research has generally assumed a strategic motivation for government use of repression. Repression is employed to counter threats from the opposition as represented by the presence of warfare. Less attention has been given to the effect of implementation on levels of repression. Theory suggests that agents are likely to make a substantial independent contribution to the level of repression, if given the opportunity. In this article we develop this argument and present cross-country comparative evidence that suggests that agents' opportunities for hidden action measured by perceived levels of financial corruption substantially influences the incidence of torture in a political system, after controlling for the strategic motive of governments and the other factors found influential in earlier research. We show that the results are robust and not sensitive to alternative modeling, measurement, and research-design decisions.