One constantly highlighted narrative on caste in India by the academia, media, NGO activists, human rights organizations, and intellectuals is the atrocity perpetuated against the ‘low-castes’ in Indian society. It is almost an inherited narrative made strong by repeated affirmations. But do the data or the extensive readings of literature support such claims? The answer is negative.
Dunkin Jalki and Sufiya Pathan, in this extremely important chapter, show clearly that data hardly supports the popular perception of caste atrocities. Most of the caste studies run into problems while dealing with the hard data. Hence, they focus more on the soft data, the ambit of which is wide and accommodative; allowing anything and everything remotely considered discriminative as an instance of caste violence. In other specialties and sciences, similar studies would have a rejection; but somehow in the Indian context, caste studies have acquired the status of unquestioned respectability. The authors elaborate on the constant puffing up on numbers by all and sundry regarding caste atrocities, helping in no small way by giving us a terrible international reputation.
Are There Caste Atrocities in India? What the Data Can and Cannot Tell Us
Dunkin Jalki and Suﬁya Pathan
‘Caste violence’ or ‘caste atrocities’ have acquired an axiomatic status which says that violence and exploitation of all kinds, including physical atrocities, are an integral part of the caste system, and are always on the rise. Yet, available data reveals some startling anomalies in respect of such claims.
Some studies (the hard data) explicitly seek to show that the lower castes routinely face violence, which includes physical assault, rape, arson, abduction, and other such criminal acts, at the hands of the upper castes. These studies document cases and analyse police statistics to make their case.
The wider and more pervasive model (the soft data) for studying caste violence is documenting much wider sets of practices and events within the ambit of studying ‘caste violence’ like the following: ‘untouchability’ practices, land ownership and labour, remuneration for labour patterns, occupational access, economic and social mobility, access to education, access to particular spaces and resources (especially temples and wells) and so on.
On the Statistical Evidence for Caste Atrocities-large Numbers Are Important
There is a clear trend in research on ‘caste violence’ to draw on data made public by the police departments in India in order to show (i) that the lower castes face a great deal of violence; (ii) that this violence requires attention because it is an inordinate amount of violence faced by one particular group of people and it is of an especially gruesome nature; (iii) that this violence is always on the rise; and (iv) that this violence is motivated by attempts to keep up the ‘caste system’ in its original form, maintain its traditional hierarchies and the subjugation of the lower castes.
Citing large numbers plays an important role in these studies. Caste scholars currently depend upon the data provided by government bodies like the National Crime Records Bureau (NCRB), which publishes annual reports on crime that contain separate chapters on caste atrocities. This practice of separate tables on ‘Incidence, Rate and percentage contribution to All India of crimes committed against SCs’ began in 1995.
Caste studies now include the total crime against SCs as well as those lodged under the so-called Special Laws: Protection of Civil Rights Act (PCR) and Prevention of Atrocities Act (PoA) as caste atrocities, without much scholarly justification for such a move. But one may minimally infer from this strategy that these scholars consider all kinds of crime against a specific ‘lower caste’ group as a valid indicator of ‘caste atrocities’ against that group.
According to the NCRB annual crime reports, the total number of ‘Incidence against Scheduled Castes [SCs]’ in India during 2011 was 33,719. This can become sensational by stating ‘every 18 minutes a crime is committed against SCs; every day, 27 atrocities against them’, and so on. To get another perspective, one may juxtapose the 33,719 cases with the two other figures provided by the NCRB:
(a) There were 6,252,729 cognisable crimes reported in total for the year 2011 in India. This implies that the total number of reported crimes against the SCs in 2011 was about 0.53% of the total reported crimes in India in 2011.
(b) In 2011, the SCs comprise about 16.6% of the total population of India. If 16.6% of the population faces 0.53% of the total criminal incidents in India, the remaining 83.4% faces 99.47% of the rest of the criminal incidence. Hence, on average, every percentage of non-SC population faces roughly 1.19% of the incidence of crime, while every percentage of the SC population faces about 0.04% of the crime. If measuring crime against a group is a reliable measure for atrocities against the group, then can we not conclude that SCs face fewer atrocities than the rest of the population?
Problems With Data Generation and Interpretation
NCRB collects data under 10 categories. The cases registered under special acts are those that are committed by a non-SC person against a SC person. If we consider only the cases lodged under the special laws in question, the percentage of atrocities against SCs decreases considerably, 0.18% of the total incidence of crime in 2011.
These crimes, though not statistically significant, may be harmful and heinous, meriting serious attention. Without looking at the numbers, scholars and NGOs surprisingly claim with unanimity that caste violence is extremely widespread and that the lower castes face greater violence in Indian society than any other groups. Where is the data that shows this to be the case?
It is disconcerting that scholars and ‘social activists’ ignore basic rules of statistical analysis to fit the data into their theories. For instance, the data about general crime against SCs/Schedule Tribes (STs) does not document the caste of the perpetrator. The NCRB data is based on complaints or ‘first information reports’ (FIRs) prepared by the police when they receive information about a ‘cognisable offence’ from either the victim or by someone representing the victim.
Thus, the NCRB data for the total crimes against SCs, save the ones recorded under special laws, does not decide whether an offence is a ‘caste offence’ or not. It is simply a record of a crime committed against an SC person. In the whole majority of the ‘crime against SCs’, excluding those recorded under PoA, therefore, the perpetrator could well be a lower caste person. Yet, there is sweeping away of this fundamental consideration in puffing up the figures.
Caste Violence-always on the Rise?
In complete contradiction of the proposition that caste scholars make about the constant rise in caste atrocities, the percentage of crime against SCs has remained rather constant over the last decade and has not, in fact, shown any significant rise. If one considers the rate of crime against SCs (i.e. number of incidences per one lakh of SC population); it has in fact decreased over the years according to some scholars.
In 2013, 39408 cases of cognizable crimes committed against SCs in India equates to a crime rate of 19.57 incidences per one lakh of SC population. Though the number of crimes has increased from 32996 in 1995 to 33501 in 2001 and to 39408 in 2013, the rate has substantially declined – from 23.24 in 1995 to 20.14 in 2001 and further to 19.57 in 2013.
One also needs to show that crime against other castes or communities is significantly less than that faced by the SCs. Yet, there is no data available today for this. The NCRB reports provide separate data only about the SC and ST communities and no other caste communities. Given this scenario, they simply cannot tell us anything that is statistically significant about ‘caste violence’.
Though crime against SCs is undeniable, statistics contradict the claim that lower caste people face greater violence in society than other groups. Minimally one may say that the idea of widespread caste atrocities is not based on the data available.
Underreporting- a Popular Caveat
Scholars counter the challenges that emerge from the hard data on the popular caveat that ‘the data is unreliable because caste violence is under-reported’. What do we make of this oft-repeated caveat?
Crimes are always under-reported to a considerable extent, for various reasons. Is the case with caste atrocities any different from this general situation? Caste scholars must show that in addition to the general tendency to shy away from reporting crime, there are obstacles and hindrances specific to reporting caste crimes.
Even if we accept one scholar’s statement that ‘the number of unregistered cases of atrocities might range between one and one and a half times that of the registered cases’, the total number of crimes against the SCs, say, in 2011 would not be more than 1.32% (0.53 × 2.5 = 1.32), of the total incidence of crime. This would still not put the rate of crimes against SCs at anywhere close to what the rest of the population faces. To bring the 0.04% (percentage of crime per percentage of SC population) closer to the figure 1.19% (percentage of crime per percentage of non-SC population), we must increase it by about 30 times.
Government compensations offered to victims of the PoA acts as an incentive to lodge complaints on non-serious grounds of the PoA act, thereby inflating the number of cases of ‘caste atrocities.’ Hence, when authors and scholars claim that figures are only indicative and do not reflect the actual situation on the ground’, it cannot be a foregone conclusion that the figures for ‘caste atrocities’ ought to be higher. They may as well be lower.
If one takes seriously the claim that most ‘caste atrocities’ go unreported, this renders the data completely faulty, thereby rendering it completely useless. In the absence of credible data on caste violence, how do caste scholars conclude that caste violence is widespread and is constantly on the rise? Yet, this has not proved a deterrent to caste scholars and human rights organisations generating reports on the issue.
The caveat concerning under-reporting does not signal a dissatisfaction regarding statistics since it predates any statistical records of violence against lower caste people. It also persists in all studies about ‘caste violence.’ This caveat stands independent of the credibility or availability of data about caste violence and unfortunately, no amount of statistical data can prove this caveat wrong.
This caveat of under reporting is thus a priori claim, not a conclusion derived from empirical investigation; and we can perhaps understand this as one of the elements of an inherited narrative about the caste system. But none of the studies on caste violence tell us what the premises for this claim are, much less provide any defence of these premises.
To say that Indian society is ‘casteist’ and therefore caste violence must go unreported is not to provide any logical support for the claim that caste violence must be high. Also, remarkably, absence of credible data, which should have raised questions about the plausibility of the research on caste atrocities, is now evidence for, and an indication of, the magnitude of caste violence and the hold of the caste system on Indian society.
Thus, the difficulty of generating satisfactory data on caste violence ends up acting as evidence for the claim of disguised caste violence and, therefore, more intense than it appears at the outset.
Violence Not Defined by Crimes Against Scheduled Castes-the Soft Data
Studies on ‘caste violence’ relying on NCRB data are a relatively small number. The dominant trend in studies tend to look at ‘violence’ in broad terms and establish the violence of the caste system independently of the hard data.
How do we then know what to classify as caste violence if it is in disguise? This is precisely what reports on ‘untouchability’ or ‘caste disabilities’ (the soft data on caste violence) handle with. Where do the scholars locate caste violence? It is not simply in acts of violence against SC individuals or groups; rather, it is everywhere – from mundane acts of limiting social interaction to economic patterns of land ownership to actual acts of violence.
Government records and reports, independent and credible research institutes, present an alarming picture of the situation of the people affected by caste-based discrimination and violence – an increasing trend in the denial of basic livelihood rights, growing numbers of atrocities, high dropout rate of students, unabated land and labour rights violations, disregard to public health, denial of access to any place or service, obstruction of political participation, negligence of law enforcement authorities in filing complaints, undue delays in police investigation and trial of cases and low conviction rate, etc.
This disregard and apprehension about hard data is in parallel to another predominant feature of writings on caste atrocities: exclusive reliance on anecdotal evidence. Most do not take up surveys of any kind but conclude on a few interviews or ‘fact finding’ missions to the sites of atrocities. Providing discrete micro instances (individual instances of violence reported by victims and their relatives and friends) as evidence for an abstract macro claim (about the age-old caste system and atrocities it generates) would be a fallacy in any other field. In caste studies, however, this has been the dominant way of making the argument.
What Exactly is Caste Violence?
Instead of studying crimes against SCs as the basis for their analysis of caste violence, writers have included a wide array of ‘social disabilities’ in the list of ‘caste violence’. Consider some of the disabilities mentioned earlier: lack of access to education and public health facilities, unjust patterns of land distribution, labour remuneration irregularities, discriminatory practices in the kinds of work offered to SCs, etc. Since studies and reports on caste atrocities resort to collection of anecdotes, all kinds of the problems that an individual face and/or relates, becomes evidence for the existence of caste violence in society.
These studies document problems that large sections of human beings across the world have been facing (predefined as ‘social disabilities’) within samples of the SC and Other Backward Classes (OBC) population group. This makes for a staggering number of social problems attributed to a single cause: namely, caste discrimination. None of these studies makes even a perfunctory attempt to prove that it is the so-called caste system that causes these ‘social disabilities.’ This is an assumption that guides their data collection, not a hypothesis that they take up for investigation.
Irrespective of their relative dominance, the SCs still suffer social disabilities according to writers and scholars across the years. For example, in a village study, one author claims that Dalits form 80% of the population and control the socio-political affairs of the village; and yet ‘caste still impacts and shapes’ the lives of the people of the village. The same facts can however argue that the practices constituted as ‘social disabilities’ simply do not obstruct social mobility, nor do they place any obstacles to the so-called lower castes taking on a position of dominance. And if ‘caste-related social disabilities’ continue whether the social status of the community is high or low, then one may logically reach the conclusion that caste does not affect the social status of groups.
One scholar places the fact that different caste groups have different places of worship as a remnant of caste disabilities. Yet, such an argument about different places of worship used by different denominations among Christians or Muslims in India is not a disability of any kind. Clearly, there are assumptions that guide the general instinct to pick out such phenomena as facts about caste disabilities.
Citing these studies as proof that the caste system exists, or to use them to raise any dispute about whether the caste system causes violence is to be disingenuous. The studies that document ‘caste violence’ do not form a defence of the theory that generates this notion.
Reporting Caste Atrocities: Trends and Strategies
An analysis of writings related to caste violence (comprising, letters to the editor, editorials, reports, and research articles) published in the Economic and Political Weekly (EPW) over five decades from 1949 to 2000 show some definite trends in the way scholars have dealt with the issue.
One of the most striking aspects of this literature is that one clearly finds that the idea of ‘caste atrocities’ is largely a late-1960s development, which ‘caught on’ in the 1970s. A review of this literature reveals how scholars have coped with contradictions brought up by the data on caste violence by using ad hoc strategies in their studies of caste violence.
Caste scholarship uses the term ‘caste violence’ or ‘caste atrocity’ to refer to the victimisation of ‘lower caste’ people by a wide variety of means. Surprisingly, in caste violence reportage, it is not necessary that the perpetrator of atrocities or victimiser be an upper caste individual or group. Where upper caste persons are not involved, the ‘system’ – read ‘the caste system’ or/and ‘Brahmanism’ – is responsible for the so-called atrocities. This trend is visible in the most scholarly writings of the last several decades that India has produced.
In 60 relevant articles on ‘caste atrocities’ in 2283 issues of the EPW journal over 50 years, interestingly, Brahmins do not feature anywhere as perpetrators of violence. Only in one case is it alleged that a convict had the support of a Brahmin moneylender. The articles, nevertheless, talk about ‘upper castes’, without always specifying who the upper castes are. In addition, one finds the following words and phrases liberally used in these writings while analysing the issue: ‘Brahmanical caste system’, ‘Brahmanical social setup’, ‘Brahmanism’ and ‘brahmanical model’. These words act as ‘explanations’ for the violence that has occurred. The caste system is violent because it is ‘Brahmanical’.
Do the Brahmins Hold So Much Power?
Thus, the Brahmins seem to have a disembodied power which allows them to wield a destructive influence on Indian society without any concrete personal involvement. None of the articles attempt to show what is ‘Brahminical’ about violence perpetrated by non-Brahmins. In fact, much of this violence comes from castes that are not part of the classical ‘upper castes’ at all, but are part of what the government categorises as ‘other backward classes’ (OBC). What is ‘Brahminical’ about violence perpetrated by the OBCs?
Such studies achieve the following: they explain the violence perpetrated by so-called lower caste groups by attributing the source of the violence to an upper caste group without any empirical evidence to show any connection between the two. In any other sphere of social science research, such work would become ‘a conspiracy theory’ and written off with disdain. In caste studies, however, it gains the status of a respectable knowledge claim, and earns its author degrees, positions, and honours, in India and abroad.
While reporting the actual cases of atrocities, the authors name the castes that are allegedly responsible for the violence. While talking about ‘caste violence’ in abstract and generalised terms, that is, while theorising the violent incidents, they resort to generalised terms like upper castes, caste Hindus, Harijans, backward castes, Dalits and so on. This creates a very peculiar situation. Even when the violent incidents involve two ‘lower castes’, that is, a perpetrator and a victim who are both from ‘lower castes’, which is not uncommon, they get reported as ‘caste violence’ and are theorised as an incident of violence for which upper castes are considered responsible. This is possible because the terms like ‘Harijans’, ‘Dalits’ or ‘lower castes’ are empty categories and comprise of different castes and units of castes, as the case might be.
Obcs-both Upper and Lower Castes as Per Convenience
The laws against caste atrocities, such as the Prevention of Atrocities Act 1989, consider only SCs and STs as ‘lower castes.’ However, in academic articles on caste violence, often even the OBCs are included under categories like ‘Dalits’ or ‘lower castes.’ Ironically, they may also be categorised as ‘upper castes’ if the argument so requires. Not just that, violence amongst OBCs may also get characterised as ‘caste violence’. OBCs feature both as perpetrators as well as victims of the caste atrocities, as and when required.
For example, in his lengthy report, K Balagopal, an acclaimed human rights activist of his time, a lawyer, and a mathematician by training, provides a list of ‘known incidents of murder or large-scale arson perpetrated against Dalits by caste Hindus in AP, ‘post-Karamchedu.’ On July 20, 1987, as the list mentions, ‘One Dalit labourer was killed in a dispute over a small patch of tank-bed land by a mob of backward caste farmers.’ While in the foregoing entry the perpetrators of crime are a bunch of OBC farmers, in the immediate next entry an OBC person is a victim. On August 13, 1987, ‘One person of a backward caste (Golla) killed in a mob attack by Kammas.’ This game of shifting from one to the other, at will, continues in the list. Here is another example from the later part of the list: ‘About 180 houses of fishing community set on fire by a mob organized by prominent … BC [backward caste] leader …’. A few lines later we read, ‘A backward caste (Boya) farm-servant shot dead by his property owner, G Narayan Reddy’.
As per the 2001 census data, the SCs comprise about 16.2% of the total population. There is, however, a lot of confusion about the population of OBCs. It is safe to consider then that the total SC, ST and OBC population in India is about 65% of the total population. That means, the reports on caste violence may document any violence (‘disguised’ or overt, as the case may be) that occurs amongst or against 65% of India’s population as ‘caste violence’, which is strange indeed.
It is unclear why there is such a strong conviction that SCs as a population group are subject to atrocities in India. There are some serious studies which suspect that holding the ‘caste system’ responsible for generating ‘caste violence’ is unjustified.
However, scholars are busy today in presenting wealth- and market-related crimes as ‘caste crimes.’ While ‘caste atrocities’ are a legal fact under the SC and STs (Prevention of Atrocities) Act, the circumstances that led to the perceived need for new legislation have not yet acquired an appropriate explanation. Unfortunately, the study of caste violence has more anomalies than explanations.
The enthusiasm for new laws related to caste atrocities does not match with new insights, and as a result, at the most fundamental level, we are unable to establish how we should define caste violence. How do we establish that some violence has been committed for ‘casteist’ motives rather than out of personal or economic or any other motives? The law solves this problem by simply registering any case of violence against an SC by a non-SC as a ‘caste atrocity’ and promptly providing compensations at various levels for any such crime registered under the PoA.
Expenditures and Consequences
Central government’s expenditure on the PoA has gone from INR 38.31 crores in 2005–2006 to 127.65 crores in 2013–2014. Each of the state governments also match the centre’s expenditure by 50%. For the year 2013–2014, then, the total expenditure would be over INR 200 crores. These expenses go towards compensation, legal aid to victims, covering travel expenses incurred by the victims, the establishment of special courts, special cells in police stations, sensitisation programmes, as well as, surprisingly, incentives for inter-caste marriage. In other words, in less than 10 years, the expenditure related to the PoA has increased fourfold while we have been unable to settle with any clarity what constitutes caste violence and therefore, what measures could curb it.
This is a serious matter to investigate not just because of the economic implications of the waste of public resources. India faces serious criticism from many international agencies based on reports that project a great deal of ‘caste violence’ by generating slippery definitions of the same. Such reports serve not just to generate international outrage but also fuel faulty and damaging legislative moves like the inclusion of caste in the discrimination laws of the United Kingdom. We must hence do more research on the source of the current assumptions guiding caste studies on the one hand and generating new models to understand the Indian reality on the other.
Featured Image:Foundations of Economic Education