Looking at the numbers from the NCRB Data for the Atrocities Act
A lot of Twitter controversy has surrounded the Scheduled Castes & Scheduled Tribes (Prevention of Atrocities) Act, 1989 (‘Atrocities Act”) and its enforcement. I myself has caught a lot of heat for doing a thread on conviction rates on the Act.
There is a lot of misconception and misunderstanding regarding the statistics relating to the Criminal Justice System, and the numbers and short explanations I provide maybe helpful to some.
The sources for this are Volume 1, Volume 2 & Volume 3 of the NCRB Crime in India-2020 Report.
Chargesheeting Rate (Higher is Better)
What does Chargesheeting rate mean? It is the percentage value of cases in which the Law Enforcement Organizations (LEO) filed a chargesheet compared to those cases that were reported, investigated and were closed due to various reasons (insufficient evidence, case quashed at investigative stage, non-cognizable, false cases). In essence, the ratio of cases in which the Police (or another LEO) could successfully complete an investigation, identify a crime and have basis to prove the case in Court.
Conviction Rate (Higher is Better)
Conviction Rate is the number of cases in which the court convicted the accused- not including those in which the cases were discharged, acquitted or disposed off without trial.
Pendency (Investigation - Lower is Better)
This ratio is the amount of cases that were pending at the year compared to the number of cases that were in total to be investigated.
Pendency (Court Trials - Lower is Better)
This is the number of cases that were filed in which the court took a decision from the total pending gives us this number. It should be that special courts exist at various levels for Crimes Against Children and against SC/STs. This percentage raises questions about their effectiveness.
Cases that ended as inconclusive (Lower is Better)
These investigations were concluded with various reasons - cases which were not cognizable, cases that were false. Cases that were mistake of fact or of law or civil dispute, cases with insufficient evidence, and cases that were abated.
Let me veer into some commentary here:
It should be noted here that the number of investigations that concluded that the complaint filed was a false report was 4.61% in case of Crimes against Women, but a 8.51% in case of Crimes against Scheduled Castes and 8.83% in case of Scheduled Tribes . This indicates that the Supreme Court's decision in 2018 was not as baseless as one might think. However, this is mitigated by the fact that the overall rate of cases that ended inconclusively overall are in line with other cases, as shown above.
These numbers are a mere peek in understanding the criminal justice system, especially in dealing with the Atrocities Act. In understanding these numbers, one cannot quote just a single number to asses the effectiveness of the system. One can generally use the conviction rate as an assessment for how well the LEOs are performing.
 There are two different rates that are of relevance - the broader conviction rate which includes cases under Atrocities Act and IPC. The overall combined conviction rate is 42.4%. The Atrocities Act is generally imposed along with an IPC offence (almost always) as an aggravated crime. Same for Scheduled Tribes: There are two different rates that are of relevance - the broader conviction rate which includes cases under Atrocities Act and IPC. The overall combined conviction rate for is 28.5%. The Atrocities Act is generally imposed along with an IPC offence (almost always) as an aggravated crime.