Judicial decision making constitutes an important part of the law in every country and judges’ decisions are expected to follow the law in its correct meaning. Do judges act this way or there might be non-legal factors that affect their decisions, and hence the law? The positive research of judicial decision making is expanding, whereas its normative significance is undeniable. This study contributes to this strand of research, as it empirically evaluates if, and to what extent, non-legal factors affect judicial decisions in income tax cases. It examines potential effects of judges’ social background and thus joins, and adds to, the empirical literature on tax litigation which is rather small. It is also the first empirical research regarding the Israeli judiciary that uses econometric methods for estimating the effect of non-legal variables on judicial decision making in tax cases.A central methodological innovation, suggested in this study, stems from the unique coding of the dependent variable – the Case's Outcome – as an ordinal variable, rather than the common coding of judicial decisions as binary. Accordingly, an ordinal regression model is the suitable methodology for estimating an ordinal variable, rather than applying the commonly used logistic regression model. Furthermore, binary coding of judges’ decisions is in many times arbitrary, causes loss of information, and may result in misestimation.This study finds that judges' gender, seniority, age at time of nomination and age at time of decision affect judges’ decisions, and hence the law in Israel.
Read full abstract