Abstract

Former researchers conceptualized consistency in criminal sentencing as the agreement between different judges on the same criminal cases. A difficulty associated with this definition of consistency is the fact that a single case is seldom judged by multiple judges. Since it is extremely rare to find the cases in which different judges independently deliver sentences, the empirical verification of the consistency in sentencing based upon the former definition is hard to accomplish. The present paper suggests a new operational definition of the consistency in sentencing as the correlation between the sentencing factors codified by statutes and the actual sentencing outcomes. This new definition is systematically related to that by the former researchers. The former definition is based upon the assumption that the agreement between different judges occurs due to the same statutory principles underlying their judgments. If this assumption is met and thus the judgments of different judges agree with one another, the actual sentencing outcomes which are rendered by many different judges should be correlated with the sentencing factors codified by statutes. This new definition of consistency in sentencing do not create any conflict with the principle of discretion regarding criminal sentencing. A data set on sexual assault was reanalysed in order to describe the analytic procedure of verifying the consistency in sentencing by using the new definition of the consistency.