In a CDDRL seminar series talk, Daniel Chen — Director of Research at the French National Center for Scientific Research and Professor at the Toulouse School of Economics — examined whether data science can improve the functioning of courts and unlock their impact on economic development. Improving courts’ efficiency is paramount to citizens' confidence in legal institutions and proceedings.

In a nationwide experiment in Kenya, Chen and his co-authors employed data science techniques to identify the causes of case backlog in the judicial system. They developed an algorithm to identify major sources of court delays for each of Kenya’s 124 court stations. Based on the algorithm, they compiled a one-page report — specific to the local court and tailored to that month’s proceedings — which provided an analysis of court adjournments, reasons for delay, and tangible action items.

To measure the effect of these one-pagers, Chen established two treatment groups and one control. Those in the first treatment group received a singular one-pager, sent just to the courts. The second received one for the courts and one for a Court User Committee (CUC). The committee, which consists of lawyers, police, and members of civil society, was asked to discuss the one-pagers during their quarterly meetings. 

To measure the relevant effects, the authors examined three primary outcomes, namely: (1) adjournment (or case delay) rates; (2) quality and citizen satisfaction; and (3) measures of economic development, including contracting, investment, and business creation. 

Results showed the intervention was associated with a 22 percent improvement in adjournments, or a decline in trial length by 120 days. They found that there was no effect on either the number of cases filed or the proxies for quality. Citizen satisfaction rates also went up, with a reduction in complaints about speed and quality, and the intervention was associated with an increase in formal written contracts and higher wages.