Criminal mistake in reading data

Leah Verghese, Amnesty International India
2 July 2018 8:54 pm

Editor’s Note: This piece was written in response to the article “Do SCs and STs face disproportionately more violent crimes in India?”, which was originally published on on June 9, 2018 and in the Financial Express on June 13, 2018.  The Financial Express published this rejoinder on June 20, 2018. Swarajya refused to publish it.

Nihar Sashittal’s article in this newspaper on June 13 2018 claimed that Amnesty International India had “fudged” and “manipulated” data, and made “alarmist” and “completely false” claims in a press statement on the Supreme Court judgement diluting the SC/ST Atrocities Act.

Sashittal’s article is ludicrously wrong and misleading on several counts. First, Sashittal suggested that Amnesty had “fudged” data by stating that the conviction rates for crimes under the SC/ST Prevention of Atrocities Act was 15.4%. But this figure is cited by the NCRB 2016 report itself on pages XXV and 571. Sashittal is welcome to disbelieve this statistic, or offer an alternative analysis that examines conviction rates in cases where both the Atrocities Act and the IPC are invoked. But to claim that Amnesty’s using of the statistic is “manipulation” is patently absurd.

It is also disingenuous to only look at certain violent crimes while comparing conviction rates between the IPC and the Atrocities Act. The latter aims to tackle specific targeted crimes against Dalits and Adivasis that take myriad forms, such as dispossessing them of their lands, parading them naked, compelling them to perform manual scavenging, and making them eat inedible substances. When cases involving these kinds of crimes don’t lead to convictions, it must be cause for concern.

The article also suggests that the statement that the Supreme Court had recently cherry-picked data to conclude that the Atrocities Act was being misused was an “ad hoc argument”. The Supreme Court in its judgment had relied on NCRB data from 2015 which showed that around 9.6% of cases involving crimes against SCs and STs were found to be ‘false’ by the police. But this number cannot be seen in isolation. The corresponding number of ‘false’ cases of kidnapping or abduction is 9%. The police also claimed that 12% of all forgery cases were false. Not surprisingly, there seems to be no outrage on the ‘misuse’ of kidnapping and forgery laws.

Sashittal also claims that Amnesty’s statement that human rights organisations have found crimes against SCs and STs to be under-reported is unfounded. A cursory internet search would reveal that several organisations have carried out studies and surveys on violence against SCs and STs which point to under-reporting of crimes under the Atrocities Act. For example, a study by Sisters for Change and the Karnataka Dalit Mahila Vedike in five districts of north-west Karnataka in 2015 found that 75% of Dalit women who were surveyed said that when they disclosed that they had faced violence, they were not believed, were blamed, were told the violence was not a crime, that they should not register a case or that they should settle the matter privately. In several cases of abuses against Adivasis, Amnesty has found the police reluctant to register First Information Reports.

In the absence of victimisation surveys, these reports often provide crucial information for policymakers. Atrocities against Dalits and Adivasis are linked to deeply entrenched prejudices and structural inequalities. Sashittal’s dismissal of the work of these organisations in this article as “anecdotal evidence” displays a limited understanding of these realities.

The author also makes several other misleading assertions. He chooses to attack a claim that Amnesty never made: that under-reporting hides the disproportionate levels of violence against SCs and STs. He fallaciously compares rates of crimes under the IPC with rate of crimes under the Atrocities Act for the period 1995-2013, even though the NCRB had calculated the crime rate under the Atrocities Act during this period by using the population of India, rather than the SC/ST population. He also wrongly suggests that the Supreme Court had recently ruled that a preliminary inquiry be conducted before arrests are made. In fact, the order directed that the police conduct a preliminary inquiry before even registering a First Information Report.Where I agree with Sashittal is his statement that data, even where imperfect, should not be substituted by fact-free assertions. Unfortunately, his article does not demonstrate the application of this sage advice.