DISCLAIMER: I am not an expert on any of the subjects discussed in this post.
NOTE: Many hyperlinks in this post refer PDF documents.
I am impressed by the amount of information the National Crime Records Bureau is sharing online. It recently released the annual report Crime in India 2010. Main documents Compendium2010 and Statistics2010 together contain about 650 pages. In the future I hope NCRB becomes more user-friendly, in formatting its report and more importantly in giving granular access to drill down every other way. I also hope we will have our own National Crime Victimization Survey (if we don’t already), with at least this level of information sharing. It is useful.
When the latest report was released, various news media outlets dutifully poured outrage and spread FUD. One such attempt that was most circulated among my circles was TOI’s take that rapes are rising fastest among major crimes. My fact-finding showed a different picture, so I wanted to clarify a few things that I wish the reporter (Subodh Varma, TNN) had.
Incidents of rape in the country have increased by a staggering 792% over the past nearly 40 years.
This nugget came from the 2nd page of Snapshots-5310. It is an absolute increase in the number of cases reported.
…compared to all cognizable crimes…
What do you think are cognizable crimes? Definition in Compendium2010 (p15): “A cognizable offence or case is defined as the one which an officer in-charge of a police station may investigate without the order of a magistrate and affect arrest without warrant. The police has a direct responsibility to take immediate action on the receipt of a complaint or of credible information in such crimes, visit the scene of the crime, investigate the facts, apprehend the offender and arraign him before a court of law having jurisdiction over the matter.”
Experts believe that while some of this jaw-dropping rise could be explained by increased reporting as awareness has grown among victims and families, the scale of increase undoubtedly reflects increasing violence against women in society.
The age-old classic trick of inferring an undoubted reflection by referring to non-existent people and data. The experts weren’t specified, so I failed to get any information from them. NCRB statistics don’t throw light on all factors. To be fair that is probably not their job. A few factors that matter:
- Gender: Increasingly there are studies on the genders of rape victims. But the current definition as per IPC probably does not have that complexity, so let me ignore this for a moment. However I request knowledgeable people to clarify the latest definition, outraged ones to make noise about the obsolete definition, and lawmakers to take note.
- Population: India has seen a notable growth in population since 1971. It also has been dealing with serious problems and significant plans related to a balanced sex ratio. So considering women population makes more sense to me (for the moment), even though crime incidence rates are usually calculated as number of reported cases per 100,000 population even for rape cases.
- Records: As mentioned in its Message, NCRB prepares the annual report based on the data sent by “State Governments and UT Administrations and Heads of various law enforcement agencies” within a deadline. I don’t know how the data submitted after this deadline is adjusted for in the next report, and what incentives the local police stations have in prompt submissions. I am guessing that records keeping and submissions are not “automatic”. If these two aspects have improved since 1971, which is a good thing and not an unreasonable possibility, they will have resulted in larger numbers with time. I request knowledgeable people to clarify about records keeping and submissions over the years, outraged ones to make noise about their quality, and bureaucrats to take note.
- Victims per Case: Table-5.3 in Statistics2010 (p395) implies that it is possible that a case can have more than one victim. Considering the number of victims instead of cases makes more sense to me, especially after reading about the infamous Vachathi case.
- Reporting Rates: Think about the last accident you witnessed and the last time you lost some cash, and you will have a basic idea why not all crimes get reported. Crimes of rape and sexual assault are infinitely more complex, and many studies across countries claim that they are one of the most under reported (unreported) crimes. Various sources online suggest that reporting rates have increased over the decades, but I couldn’t find reliable data showing trends over a period in any countries. I request knowledgeable people to point me to any studies, outraged ones to make noise about the need for greater awareness and better victimization studies, and sponsors to take note.
To make indubitable references on how much more or less dangerous a society has become for women, we need actual data along with expert opinions. (That is not to say we should stop asking for the situation to improve.) Taking only population and sex ratio into consideration, I have the below table*:
| Year | Population | Sex Ratio (Females per 1000 Males) | Female Population | Reported Rapes | Incidence Rate (Cases per 100,000 Females) |
| 1971 | 548,159,652 | 930 | 264,139,107 | 2487 | 9.42 |
| 1981 | 683,329,097 | 934 | 330,004,848 | 5409 | 16.39 |
| 1991 | 846,421,039 | 927 | 407,178,154 | 10410 | 25.57 |
| 2001 | 1,028,737,436 | 933 | 496,540,107 | 16075 | 32.37 |
| 2010-11 | 1,210,193,422 | 940 | 586,382,380 | 22172 | 37.91 |
* The population and sex ratio figures are from respective documents shared by Census India and its 2011 dashboard. Reported Rapes are from NCRB’s 1953-2010.
This eight-fold increase is
no longer an eight-fold increase. To show how sensitive ratios can be, let me illustrate with an example using one more factor. Suppose that the reporting rate was x% in 1971 and y% in 2010-11. Then the adjusted incidence rates would become 942/x in 1971 and 3791/y in 2010-11. Now that’s an (4.02x/y)–fold increase. If the reporting rate increased from 10% in 1971 to 30% in 2010-11, that would be a 1.34-fold increase. If the reporting rate increased from 5% in 1971 to 20% in 2010-11, that would be negligible increase. Mind you, this is hypothetical and I am considering just one additional factor.
Moving on.
Maintaining a trend which has existed for several years, almost 97% of the rapes were committed by persons known to the victim with about 7% committed by family members and 35% by neighbours.
I could find no data or even a passing inference to such a trend in the NCRB report. The TOI reporter or his experts may have. The figures about “committing” and the definition of “family members” are imaginative. One paragraph in Compendium2010 (p83), however, contains the following two sentences: “Offenders were known to the victims in as many as 21,566 (97.3%) cases. Parents/close family members were involved in 1.3% (288 out of 21,566) of these cases, neighbours were involved in 36.2% cases (7,816 out of 21,566) and relatives were involved in 6.2% (1,344 out of 21,566) cases.”
Back to how dangerous our society is to women.
Rape is not the one danger to women and India is not one society. There are other violent crimes, including deaths from dowry and sati; other crimes against women, including harassment and importation. Their incidences vary across states, cities, villages, demographic categories. We need more awareness in all aspects of these issues among everybody, including men.
For actions and activism to be effective, we need as clear and complete a picture as possible. Not hyperbole. We are not using all available data, and we don’t have as much data as needed. We should. This is as good a time as any other to make noise about these things.
A RAINN would be good too.
P.S. People might be interested in this UNODC document related to sexual violence. Remember that the incidence rates there are calculated per 100,000 population and not 100,000 females.