Despite a growing number of experimental studies in international political research, experimental approaches are not yet truly appreciated in political science and their value for this field is still underestimated (McDermott 2002; Green/Gerber 2003; Druckman et al. 2006; Margetts a. Stoker 2010). I argue that experiments can greatly contribute to advance political research as they supplement other research methods by providing different insights into the investigated subject matter. Especially in the field of corruption research experiments can – as a complementary mode of investigation – help to identify specific factors that represent the individual determinants of corruptive behavior. Since corruption is secretive by nature and often takes place in hidden and unofficial settings it cannot be measured directly. Frequently used statistics to assess corruptive behavior, such as criminal indictments or perception-based indices (e.g., Corruption Perception Index of Transparency International, Control of Corruption Index by World Bank) are flawed and limited. Either they tend to indicate the effectiveness of the judicial system or include a highly subjective evaluation and measure the perception of corruption itself rather than the actual level of corrupt activity (Dusek et al. 2005). Trying to overcome these limitations Armantier and Boly (2010) successfully conducted field experiments which pointed out that experiments in political science can be externally valid. In their experiments students secretly propose a bribe to the tutor who grades the exams in order to obtain a better grade. The researchers manipulated different experimental factors such as the amount of the bribe, the tutor's wage, the level of monitoring and the amount of punishment. In particular, they show that increasing the graders wage reduces the probability of accepting the bribe but also promotes reciprocation. Using their experimental design, I intend to replicate their study with a German sample at different universities to investigate corruptive behavior with regard to sociodemographic factors such as sex, age, financial situation and course of study. This experimental approach may help to understand mechanism of corruption on the micro-level in a deeper way than traditional methods.
