X
  1. Have crimes against SCs and STs increased of late? Here are the facts

Have crimes against SCs and STs increased of late? Here are the facts

A website named Indiaspend, which calls itself the “country’s first data journalism initiative”, and aspires to be the “agency of record when it comes to data and facts” recently ran a sensational cover story about the rise in crime rates against Dalits in India.

If we look at the Crime In India statistics published by NCRB which the Indiaspend article picks from, we find that the authors of the article are guilty of this fallacy and similar ones.

By Nihar Sashittal

A website named Indiaspend, which calls itself the “country’s first data journalism initiative”, and aspires to be the “agency of record when it comes to data and facts” recently ran a sensational cover story about the rise in crime rates against Dalits in India. It claimed that its analysis of data from National Crime Records Bureau (NCRB) indicated a sharp rise in crimes against Scheduled Castes (SC) and Scheduled Tribes (ST) and that the crime rates against the communities had jumped eight times (746%) and 12 times (1,160%) respectively in the past decade.

The article was highlighted by many activist groups as well as prominent journalists including Shivam Vij and Tufail Ahmad. Rajdeep Sardesai even called it “solid data”. The claims of the ‘analysis’ seemed truly horrifying, for if true, wouldn’t such atrocities against a part of the Indian society in the last few years drawn international attention?

And then Indiaspend realised it had bungled? The authors of the article, in their enthusiasm to play up caste atrocities, had compared entities with different denominators to make the claim. They regretted their error, attributing it to the change in NCRB’s methodology. What they did not mention in their regret message though was that they had misquoted the report. They changed a few things in the article, importantly the claim of 746% increase in crime rates against Scheduled Castes was changed to 25% increase and 1,160% increase in crime rates against Scheduled Tribes was changed to a 9% decrease after their new calculations. But many errors were left as they were and a new title was chosen, cherry-picking another number so that the original narrative of a steep rise in caste atrocities remained intact and the article was salvaged.

Dr Dunkin Jalki and Dr Sufiya Pathan at the Centre for Interdisciplinary Research in the Humanities and Social Sciences (CIRHS), Ujire, have studied the reports on caste published by media and activist groups for past several years. In their research published in 2017, they identify a clear trend in these reports to cherry-pick data made public by the police department to sensationalise and make exaggerated claims of incidence and rise in caste violence. They note that one of the strategies used by some scholars and activists “to puff-up figures and claim that caste violence is always on the rise is to use absolute total figures.”

If we look at the Crime In India statistics published by NCRB which the Indiaspend article picks from, we find that the authors of the article are guilty of this fallacy and similar ones. For example, in the time period that we are told the crime rate against Scheduled Castes increased by 25% (was 746%), the rate of overall crimes as reported under Indian Penal Code (IPC) increased by more than 39% and the rate of violent crimes against all Indians went up by more than 45%. So, if the crime rates against Scheduled Castes have increased at a lesser rate as compared to the overall increase in crime rates, is it justified to talk of a rise in crime rates against SCs? This is a clear instance of what I would like to call the ‘number mining fallacy’ which is similar to the ‘quote mining fallacy’ employed by many creationists and political propagandists. Numbers are “surgically excised” and transplanted into a preconceived politically convenient narrative without making any effort to explain these numbers in the context of the overall numbers. It is managed in a way that the data, even if contradictory to the assumptions, does not upset the applecart.

Again resorting to some more ‘number mining’, the Indiaspend article tells us (this time also based on erroneous calculations) that “as many as 422,799 crimes against dalits or scheduled castes (SCs) and 81,332 crimes against adivasis [STs] were reported between 2006 and 2016.” The absolute total figures are quoted to give an impression that these numbers somehow are disproportionately more.

But do Scheduled Castes and Scheduled Tribes face disproportionately more crimes than others? Do they experience more violence? Implicit to many reports on caste is the assumption that the answer to these questions is a given and that they are in the affirmative. But what does the crime data published by NCRB tell us about this? This question has been considered by Jalki and Pathan (2017) at considerable length. The NCRB’s 2016 yearly report provides additional data on violent crimes which can help us probe this question with a bit more granularity.

The NCRB’s Crime In India (CII) yearly reports dedicate separate chapters named ‘Crime/Atrocities Against Scheduled Castes’ and ‘Crime/Atrocities Against Scheduled Tribes’ that list the number of crimes committed against SCs and STs. There are several problems with the use of this data which are pointed out by Jalki and Pathan, including the conflation of all crimes listed in these chapters as “caste offences”. Till 2015, these chapters even included crimes where the perpetrator of the crime is also from SC or ST but the 2016 report makes a small change in that it excludes these crimes. We can hope that NCRB continues to improve its methodology to get better and more accurate data in the future. Nevertheless, this data in all its imperfections can give us some indications. If not anything else, it would at the least tell us about the claims that are made on its behalf.

The CII 2016 report is especially helpful in comparison to the 2014 and 2015 reports because it publishes a detailed break-up of crimes against SCs and STs under various headings like murder, kidnapping, rape, dacoity, grievous hurt, etc. Since these numbers for the overall population are also published in other chapters of the report, this provides us with an avenue to compare these numbers, especially the rates of violent crimes to get an indication of whether the extent of violence faced by people belonging to the concerned communities is disproportionately more. In case of widespread violence against Scheduled Castes and Scheduled Tribes, the rate of violent crimes against them should be significantly more than the average rates of crimes against others. Let us consider the data to see if this assumption is indeed confirmed.

The accompanying graphic lists the violent crimes and their rates for SCs, STs and the total Indian population. Most of the values have been directly taken from the NCRB’s 2016 yearly report. A few numbers (in bold, in the table) are arrived at by using trivial calculations that can be easily verified.

If we are to posit that the crime rates in the NCRB’s report indicate widespread caste violence, the rates of crimes against SCs and STs when normalised to the whole population should far exceed the rates of crimes for the overall population. But a comparison of these values does not indicate this. Rather, for every violent crime mentioned, the normalised values of crime rates against SCs and STs appear to be significantly lesser than the rates for the overall population.

If we were to consider the rate of all the violent crimes together we get the values of crime rates as 39.46, 10.9 and 4.36 for the overall population, SCs and STs respectively. The crime rates imply that both SCs and STs face at least several times lesser violent crimes than the average of 39.46 for the overall population. So, people belonging to the Scheduled Caste or Scheduled Tribe in India, the data tells us, face lesser violence and have a far lesser chance of being a victim of violent crimes such as murder, rape, and kidnapping than others.

The implications are very clear. The narrative that Scheduled Castes and Scheduled Tribes face more violence than others that some activist groups have constantly promoted and have even tried to internationalise, is completely unsupported and even contradicted by the very data that they pick from or point to. The data from CII 2016 on violent crimes corroborates the findings of Jalki and Pathan (2017) based on earlier reports “that these statistics contradict the claim that lower caste people face greater violence in society than other groups.” And that “minimally one may say that idea of widespread caste atrocities is not drawn on the basis of the data available about the crimes and atrocities in India.”

But the narrative on caste violence is so entrenched today that many of us who think they do not hold strong views on the subject too could find this data counterintuitive, leading to some amount of cognitive dissonance. But if we choose to be faithful to data rather than to our preconceived notions, deriving other conclusions becomes difficult. One could resort to many ways to counter what the data tells us: with personal incredulity (However, that cannot be true. It is unbelievable), availability heuristics (What about the incidents that we hear of in media then?), auxiliary hypotheses (But aren’t crimes underreported?) or word games (economic backwardness is also violence, who’s counting that?). But the fact remains that none of that can help salvage the claim about crime data indicating excessive violence against Dalits and tribals. If it is a falsifiable claim, then it is rendered falsified by the data that we have at hand. But if it is an unfalsifiable tautology then no amount of data anyways will be of any help.

It is important to note that this does not mean that there are no instances of gruesome violence or hatred or that we should not be concerned about them. In fact, we should be concerned about each and every crime, and especially the ones that affect the poor and the underprivileged and which can be avoided by policy measures or interventions. The crude attempts at manipulating data have nothing to do with these concerns. Such attempts could serve some political propaganda or help in creating atrocity literature against India in some international fora, but they do not qualify as attempts to help the underprivileged communities. Far from it, inaccurate data can lead to skewed policy priorities and decisions that have the potential to hurt the underprivileged and worsen their economic and social conditions. Such propaganda can also create mistrust and hatred among communities and can fuel unrest in the society.

Finally, what do we make of Rajdeep Sardesai’s endorsement of Indiaspend’s claims calling it “solid data”—the very data that has now been discarded as erroneous by Indiaspend itself? When he called it “solid data”, he possibly did not imply that he had checked the data himself or that he understood its implications. What he meant by “solid” was possibly that it was very helpful to him in his political propaganda; that it fits perfectly well within his horizon of expectation; that it ‘solidly’ confirmed his long-held beliefs of widespread and ever-increasing caste violence.

Imagine if this had happened in a field of science. Not only would reputations be at stake, but an anomaly of this sort, exposing a gap between the data and the expectations prevalent in the practitioners of that field, would have set in what Thomas Kuhn would call a state of crisis. But in the field that Sardesai belongs to, all that he may have to do is to brazen it out with silence or shrug it off with a spin or two. Moreover, he can also take comfort in the fact that he is not alone out there in having these strong beliefs that are unsupported by data and possibly impenetrable by the force of evidence.

The author is Contributor, Swarajya (Views are personal)

Reproduced (goo.gl/4bq7Ay), with permission, from Swarajya (swarajyamag.com)

Get live Stock Prices from BSE and NSE and latest NAV, portfolio of Mutual Funds, calculate your tax by Income Tax Calculator, know market’s Top Gainers, Top Losers & Best Equity Funds. Like us on Facebook and follow us on Twitter.

Outbrain
Show comments