A riff on [this WaPo Monkey Cage post](http://www.washingtonpost.com/blogs/monkey-cage/wp/2015/05/04/israels-appeal-courts-treat\ -arabs-better-when-one-judge-is-arab-israeli/).
We find that Arab defendants are 20 percentage points more likely to be incarcerated by an all-Jewish panel than by a mixed panel, while Jewish defendants saw no difference. Furthermore, all-Jewish panels result in higher prison sentences for both Arab and Jewish defendants, but the magnitude of this effect is small and insignificant for Jewish defendants. By contrast, there is a much larger change in prison sentence for Arab defendants: their sentences are longer by 2.7–4.6 months, or somewhere between 17 and 35 percent increase in the length of the average prison sentence, depending on the statistical model. Strikingly, since all-Jewish panels sentence more Arabs to prison in the first place, these longer prison terms are being given in cases that are less serious on average than those seen by mixed panels.
The original:
We find that Arab defendants are 17 percentage points less likely to be incarcerated by a mixed panel than an all-Jewish panel, while Jewish defendants saw no difference. Furthermore, mixed panels result in lower prison sentences for both Jewish and Arab defendants, but the magnitude of the effect is small and insignificant for Jewish defendants. By contrast, there is a much larger change in prison sentence for Arab defendants: their sentences are shorter by 2.7–4.6 months, or somewhere between 15 and 26 percent reduction in the length of the average prison sentence, depending on the statistical model. Strikingly, since mixed panels sentence fewer Arabs to prison in the first place, these shorter prison terms are being given in cases that are more serious on average than those seen by all-Jewish panels.