In this interview, Rukmini S, a data journalist and author discusses with Ayan Gupta, a Death Penalty Fellow at Project 39A about her book and particularly the chapter titled ‘How India tangles with Cops and Courts’. She analyses data to help us understand how to contextualise and comprehend data relating to crime in India.

  • Ayan Gupta: The book notes that the NCRB’s data about crime in India can be unreliable. It is sourced from FIRs, which often represent the police version and, not necessarily, the actual event. These figures are further distorted as reported crimes are listed only based on the most severe offence alleged in the FIR. The NCRB reports, thus, tend to give an inaccurate image of crime in India. In this context, how do we understand the importance of the data that the NCRB currently puts out? How can NCRB data collection and presentation improve for a more accurate representation of crime in India?

Rukmini S: I think that the best way to make sense of not just the NCRB data, but a lot of Indian data that can be misleading, is by properly understanding how it is produced, and then using that background to add context to the data, rather than using the data alone as a proxy for “the truth”. In the case of the NCRB, some of the key problems with the data are the ones that you have mentioned – police statistics are aggregates of FIRs, which are very far from being factual documents, and the NCRB aggregates them by using the ‘principal offence’ rule which means that only the most serious offence makes it to the statistics, ostensibly to prevent double-counting. There are other problems too – the NCRB only releases data on select offences ostensibly to limit the size of its report. And then of course there’s the fundamental issue, which is that police statistics count reported crimes only, and there’s good reason to worry that the most marginalised are those least likely to report a crime, but also that some types of crimes (violent bodily harm) are less likely to get reported than others (minor property crime) which presents a skewed version of the reality of India’s crime-scape. Once we understand these “known unknowns”, it can help us correct for the gaps in the data, either statistically, or just by making a better mental map.
The NCRB for its part could release disaggregated data for researchers even if not in its main report. It could also be a little more upfront about its methodology, including engaging more closely with journalists and researchers. Additionally, a large sample household crime victimisation survey is in my view essential.

  • Ayan Gupta: The book notes that media coverage of data about crime severely distorts our understanding of crime and the functioning of the criminal justice system in India. Due to high numbers of recorded FIRs, Delhi and Kerala are simplistically characterised as having a high incidence of crime, by the media. This narrow interpretation of the data fails to account for other reasons for the high numbers, such as high reporting of crimes to the police in these regions as compared to other parts of the country. Similarly, low acquittal rates, which are often a result of lack of evidence or unnecessary prosecutions, are characterised as judges being “soft” on crime. Due to such incorrect reporting, structural fault lines in the system fail to get adequate attention. Do you think such a coverage results from NCRB’s failure to provide a detailed or fully-reasoned methodology, inhibiting a complete understanding of the data? Would you classify NCRB data as lacking transparency by design?

Rukmini S: I think there is a systematic lack of literacy around crime statistics in the media, among the general public, and even in some sections of academia. I say in the book that we have turned crime statistics into a spectator sport, and I sometimes compare their reporting to cricket statistics – except that I think we’ve developed much more nuance and literacy around cricket statistics than among crime statistics! I don’t like to pass the onus of communicating transparently onto the NCRB alone – certainly, they could do a better job, but I think the entire community of people engaging with and communicating NCRB data need to do a better job of providing context while talking about these numbers. This necessarily means that you cannot look at the NCRB’s data in isolation – it has to be looked at alongside other datasets including levels of development of states, functioning of the judiciary, as well as a more quantitative understanding of the society in which crimes are taking place.
Ultimately, we have to decide what we want the NCRB’s data to be telling us; if we expect to be getting a full understanding of crime from the data alone – and that’s certainly how media reporting often positions it – we are bound to be disappointed at best, and misled and misinformed at worst. We need to temper expectations around what the data can say and do for it to serve any use.

  • Ayan Gupta: As you mention, the availability of accurate data about crime could potentially assist policing in many ways if used responsibly in its appropriate context. Yet, there is a legitimate fear that, if used irresponsibly and devoid of context, demographic data about crime can be used to justify the over-policing of marginalised communities and localities. It can be used to legitimise the notion that some communities are naturally prone to criminality. Given this concern, how do we ensure the responsible use of data in a criminal justice system ridden with massive inequalities? What kind of best practices must state institutions employ when dealing with such sensitive and powerful data?

Rukmini S: To me these concerns stem right from what we currently allow to be made public – the level of detail in FIRs, chargesheets and court documents that is made available to the public and permitted to be reported in full by the media is deeply problematic, impinges on the rights of both complainants and the accused, and is a huge invasion of privacy. Then, there is the point that you make about the problematic use of demographic data around crime. I think I stand on the side of making data available after privacy has been ensured, and then hoping for openness and transparency to bring to the fore “the truth”.
In the book, for instance, I write about the misuse of IPC Section 328 (causing hurt using a poisonous substance) by the Mumbai police to tackle the spread of the drug mephedrone prior to its inclusion in the NDPS schedule. Now someone could have chosen to look at the data and make the claim that Muslims in Mumbai were accused of an outsize proportion of ‘crimes’ under this section. However a deeper look at the data showed me that there was a 100% acquittal rate for reasons that I detail in the book, and the argument really should have been that Muslims were being charged at a disproportionate – and potentially discriminatory – rate. I stand on the side of making the data available, and then for all “sides” to try to see what the data show.

  • Ayan Gupta: The first chapter illustrates that quantitative data is not enough to fully understand India’s criminal justice system and the intersection of life and crime. The socio-political factors that affect how people and institutions within the system act can only be captured by “more and better qualitative research into crime than currently exists”. What kind of qualitative output do we urgently need to better understand some of the most significant issues plaguing our criminal justice system? Do you think state institutions give qualitative data enough importance?

Rukmini S: Yes, I strongly support a combination of quantitative and qualitative work in reaching an understanding of crime and justice. For one, we urgently need more good crime victimisation surveys – perhaps even one conducted by the state, as is done in many other countries like the UK. Additionally, I believe that more quantitative work needs to come from academia. I have seen a large number of academic papers that use “rate of crime” statistics from the NCRB as a proxy for the level of crime in an Indian state, and such regressions can only lead to problematic outcomes. We’re going to need more small qualitative studies that give us a better understanding of the experience of actual crime, as well as the circumstances that produce and mediate it. This also applies to the workings of the judicial system.