Consider the type of a newly elected president vary: from noncorrupt to corrupt (continuum). In a natural setting, I assume that corruption ofa president reflects the type ofa president in an exact sameway. In other words, when there is no restraints, the more corrupt is a president, the more he or she seeks illegal private gains (See Figure 1). Henceforward, I use a concept CORRUPTION in this model to refer "a president’s BEHAVIOR deviates from the formal duties of a public role because of private-regarding (personal, close family, private clique) pecuniary or status gains; or violates rules against the exercise ofcertain types ofprivate-regarding influence (Nye 1967)". I assume that there is information asymmetry between a first-term president …show more content…
I define it as single-term presidency. Because a reelection incentive, which hinders a corrupt president from perpetrating corruption, does not exist, I expect that a degree ofcorruption by a first-term president will be higher under single-term presidency than under multi-term presidency, in which a first-term president can be reelected (See Figure 2).
My theoretical argument leads me to expect a negative relationship between single-term limit and first-term president’s corruption. I exploit post-testonlynonequivalentsamples design.
Since my level of observation is country, I cannot randomly assign groups to the treatment group and the control group. Also, I cannot conduct a pretest because presidential term-limit always exists. For level of observation, I confine samples to countries within South America in order to match legal-origin, which is also referred as a predictor of corruption (Treisman 2000). All South American countries share British Common law origin and have presidency. Also, I exclude countries in the specific year which are considered as dictatorship based on DemocracyDictatorship(DD) Index.6 Since my hypothesis deals with first-term presidents only, I exclude countries in which the president is under his or her second-term or