5 ms·

I always wondered about the original study whether they controlled for the fact that courts control their case calendars. I can't count the number of times I've

I always wondered about the original study whether they controlled for the fact that courts control their case calendars. I can't count the number of times I've had a client's case shuffled to the back of the line because the judge didn't want to deal with it (or wanted to make me or the client wait longer). Also, Muslim countries commonly offer amnesty or executive clemency at the Eid to people who have been convicted of crimes, so that's something else to correct for/think about. For example, if I am a judge adjudicating a case for theft and I know that the defendant will be pardoned on the Eid, I might acquit or pronounce a light sentence. This is a different issue than the sibling comments RE spiritual reflection, more like a concession to a larger policy imperative. Edit to add that state sponsored leniency at the Eid is likely happening in Pakistan but not India.

hgsgm 4y ago
> if I am a judge adjudicating a case for theft and I know that the defendant will be pardoned on the Eid, I might acquit or pronounce a light sentence. Why? Wouldn't that undermine the meaning of Eid?
- public_defender 4y ago
  Think of it more like gambling for credibility. Theft is still wrong during Ramadan, but if I know that everyone convicted of petty theft will be pardoned in a month, do I impose the standard sentence? If I do, it's spiteful judge vs. generous executive. If I give a light sentence, I steal the march on the pardon and get to be the benevolent judge.
mattkrause 4y ago
No, the original study didn’t—-and that was its fatal flaw. They assumed that cases were randomly ordered and this equally likely to get the same sentence. However, they were actually done in blocks, each containing prisoners from one prison. Judges tried to finish a block before adjourning. Within each block, cases were ordered by attorney and those without an attorney were last. For many reasons, people representing themselves tend to fare more poorly. Boom, there’s the effect! PNAS published a commentary pointing this out but naturally it’s been cited 80x less: https://www.pnas.org/doi/10.1073/pnas.1110910108 https://www.pnas.org/doi/10.1073/pnas.1110910108 Another obvious tell that there was a problem is that ordinal rank (1st, 2nd, etc) was associated with severity but wall time was not (once you also include rank). The mechanism presumably isn’t linked to the number of cases heard (swinging the gavel isn’t *that energetically expensive!) but the passage of time since the meal. Thus, you’d expect the exact opposite effect.
- notfed 4y ago
  > They assumed that cases were randomly IMO making this assumption is inexcusable, and I find it hard to believe it was "accidently overlooked". Asking a single judge for comments on the study before publishing it would have been enough.
FormerBandmate 4y ago
Ramadan should be excluded entirely from the data, it's way too big of a confounding variable

hgsgm 4y ago

> if I am a judge adjudicating a case for theft and I know that the defendant will be pardoned on the Eid, I might acquit or pronounce a light sentence. Why? Wouldn't that undermine the meaning of Eid?

public_defender 4y ago
Think of it more like gambling for credibility. Theft is still wrong during Ramadan, but if I know that everyone convicted of petty theft will be pardoned in a month, do I impose the standard sentence? If I do, it's spiteful judge vs. generous executive. If I give a light sentence, I steal the march on the pardon and get to be the benevolent judge.

mattkrause 4y ago

No, the original study didn’t—-and that was its fatal flaw. They assumed that cases were randomly ordered and this equally likely to get the same sentence. However, they were actually done in blocks, each containing prisoners from one prison. Judges tried to finish a block before adjourning. Within each block, cases were ordered by attorney and those without an attorney were last. For many reasons, people representing themselves tend to fare more poorly. Boom, there’s the effect! PNAS published a commentary pointing this out but naturally it’s been cited 80x less: https://www.pnas.org/doi/10.1073/pnas.1110910108 https://www.pnas.org/doi/10.1073/pnas.1110910108 Another obvious tell that there was a problem is that ordinal rank (1st, 2nd, etc) was associated with severity but wall time was not (once you also include rank). The mechanism presumably isn’t linked to the number of cases heard (swinging the gavel isn’t *that energetically expensive!) but the passage of time since the meal. Thus, you’d expect the exact opposite effect.

notfed 4y ago
> They assumed that cases were randomly IMO making this assumption is inexcusable, and I find it hard to believe it was "accidently overlooked". Asking a single judge for comments on the study before publishing it would have been enough.

FormerBandmate 4y ago

Ramadan should be excluded entirely from the data, it's way too big of a confounding variable