Law in the Internet Society

View   r11  >  r10  ...
JoseMariaDelajaraFirstEssay 11 - 02 Jan 2020 - Main.JoseMariaDelajara
Line: 1 to 1
 
META TOPICPARENT name="FirstEssay"
Line: 20 to 20
 The main fear of judicial officials opposing open justice is that the data will be run through machine learning software and reflect their actual decision patterns. They are right in the sense that justice is fallible. For example, Chen found out that perceived masculinity of the voice of the attorney predicts court outcomes (i.e. males are more likely to win then they are perceived as less masculine). Emotions can also influence legal decision making. For example, a meta-study analyzing 23 experiments with over 4500 participants determined that gruesome evidence led to harsher sentences in 95% of the cases. Also, judges were found to be influenced by irrelevant sentencing demands, even when the demand was a product of them throwing dice. That's not all. Judges have been found to be influenced also by extraneous factors such as unexpected outcomes of football games in the same week of the decision or the last time they took a food break.
Deleted:
<
<
This realistic reflection of human decision-making has generated some backlash. Recently, France banned the publication of statistical information about judges’ decisions. Anyone who breaks the law could be punished with a sentence of up to five years in prison. On its article 33, the new French Justice Reform Act states that “the identity data of the magistrates and the members of the judiciary may not be reused with the purpose or effect of evaluating, analyzing, comparing or predicting their actual or presumed professional practices”. Hence, the French Government did not intend to stop publishing the data; just to punish its comprehension through data analytics.

The reaction against legal analytics relies on a wrong view of data. Most people think that it serves a crystal ball, enabling them to predict the outcome of every single dispute. That is not the case. For one, legal analytics is data-hungry, so it needs enough volume, variety, velocity and veracity (known as the four V’s) Even so, a decision pattern does not necessarily mean the judge will behave the same way every time. It could be just an influence of an unexpected personal event. Also, judicial analytics depends on the details of the data of the specific court. For example, if the record shows that a court has favoured plaintiffs, it will likely attract meritless cases. At a superficial level, this will hinder the prediction power of the data of that court (i.e. the numbers will revert to the mean by the generated as a reaction to legal analytics).

Judicial analytics does not predict the future. Instead, its output is a probability based on past decision patterns. Hence, lawyers must not forget they are still human, and that they suffer from probably-neglect bias. Most importantly, judges need to be reminded that legal analytics provides a key opportunity for identifying unknown patterns, even to them, and learning from past mistakes. It is a means towards making legal decision less intuitive.

 

The way forward: data commons


Revision 11r11 - 02 Jan 2020 - 23:08:48 - JoseMariaDelajara
Revision 10r10 - 24 Nov 2019 - 16:41:48 - EbenMoglen
This site is powered by the TWiki collaboration platform.
All material on this collaboration platform is the property of the contributing authors.
All material marked as authored by Eben Moglen is available under the license terms CC-BY-SA version 4.
Syndicate this site RSSATOM