View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by ePrints@APU

COMMENTARY

legislated on, and constituted anti-corrup- Bill: Lessons from tion agencies. Orissa was the first state to legislate on this matter (1970),1 while the Karnataka Lokayukta’s Maharashtra was the first to constitute an anti-corruption agency (1972).2 These pre- Performance ceding policy debates and existing models in the states have framed our policy choices. But the debate so far has focused exten- Narayana A, Sudhir Krishnaswamy, Vikas Kumar sively on the constitutional status of the Lokpal (L M Singhvi coined the word Lok- The debate on the design of the 1 Introduction pal, which means “Protector of the People”, Lokpal has been premised on the he debate on policy choices and the to indigenise the word (Stand- questionable assumption that institutional design for a new anti- ing Committee 2011, para 3.3) and the ad- Tcorruption agency at the union and ministrative and legal mechanisms nec- what is needed to combat state levels has pervaded every public fora. essary for a strong and effective agency corruption is a powerful national The ubiquitous debate is characterised by without any attempt to learn from an institution to prosecute and passionate disagreement between familiar ­empirical analysis of the performance of convict the corrupt, but the opponents who never tire of restating and existing institutional models. rehashing their adop­ted positions which The best case for a national Lokpal is debate has not drawn on the have been informed by their moral or to show that the existing Lokayukta in experience with the existing political commitments. Even by the stand- the states works. For instance, it is often institutional models, namely the ards of India’s loud and noisy democracy ­argued that the Karnataka Lokayukta, Lokayuktas in the states. An the anti-corruption debate has been char- constituted under the Karnataka Lokayukta acterised by a rancour and extra­ordinary Act, 1984, provides an argument for an empirical analysis of the brinkmanship that threatens to derail anti-corruption agency at the national performance of the Lokayukta in India’s everyday practice of politics. level. So, it is puzzling that the debate has Karnataka between 1995 and 2011 Despite the polarised debate there is referenced prior bureaucratic discourse suggests that any anti-corruption agreement on the core moral imperative and the National Crime Record Bureau to tackle corruption seriously. However, (ncrb) statistics, but there has been no agency, no matter how powerful, the debate has quickly moved from this syste­matic effort to evaluate or assess the that is oriented towards criminal agreed premise to the questionable conclu- experience of existing anti-corruption conviction is bound to fail in the sion that we need a powerful national insti- agencies in the states. The latest Parlia- absence of judicial reforms. tution to prosecute and convict the corrupt mentary Standing Committee Report’s use under the criminal law. To our knowledge of anecdotes as empirical evidence exem- the choice of appropriate legal instruments plifies this approach.3 to deal with corrup- Figure 1: Ratio of Raid to Trap Cases (1995-2011) tion has not been 7 We are grateful to Manavi Belgaumkar, debated. But the idea 6.00 6 Dharmendra Chatur, Tasneem Deo, of an anti-corruption Rajeev Kadambi, Vandana Kamat, and agency is not a new 5 Sumandro C for research assistance, to one. It has survived the P G Babu, Poonam Mehra, V Santhakumar, 4 and Alok Tiwari for helpful discussions, and to scrutiny of the National Azim Premji University for institutional Com­mission to Review 3.00 3 support. The data used in this study was the Working of the Con- 2.50 obtained in June 2011 from the public 2.00 stitution (2000), two 2 information officer, the Karnataka Lokayukta Administrative Reforms under the Right to Information Act, 2005. 1 The usual disclaimer applies. Commissions (1966 0.73 0.27 and 2005), four parlia- 0.07 0.06 0.03 0.00 0.12 0.10 0.08 0.14 0.09 0.04 Narayana A (narayana.gatty@ 0 mentary standing com- 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 June azimpremjifoundation.org), 2011 Sudhir Krishnaswamy (krishnaswamysudhir@ mittees (1996, 1998, gmail.com) and Vikas Kumar (vikas.kumar@ 2001 and 2011), and review of eight anti- In this article we hope to bring new azimpremjifoundation.org) are at the Law, corruption bills (1968, 1971, 1977, 1985, insights to this debate by engaging with Governance and Development Initiative, 1989, 1996, 1998 and 2001). Half of the law in action. Over the last six months, Azim Premji University, Bangalore. states and union territories have already researchers at the Azim Premji University’s

12 January 7, 2012 vol xlvii no 1 EPW Economic & Political Weekly COMMENTARY

Figure 2: Yearly Distribution of Cases Initiated under Karnataka Lokayukta Act, 1984 2 Agency Prosecutions 450 Justice Justice N Santosh Hegde 428 vs Citizen Complaints Rabindranath Justice S A Hakim Justice M N Venkatachala 399 400 Pyne The need to establish a strong Lokpal that 362 350 initiates criminal investigation against 308 corrupt officials has been a key demand 300 269 in the current debate. The Karnataka 250 Lokayukta had the power, under the cases

of 200 194 194 Preven­tion of Corruption Act 1988, to in- Raid vestigate cases of corruption and recently Number 150 Trap 142 139 was endowed with suo motu powers 103 100 97 99 even under the ­Karnataka Lokayukta Act 55 4 50 37 39 to initiate criminal investigation. How- 27 24 27 26 13 2 6 1 6 0 3 2 4 4 10 9 3 0 6 ever, between 1995 and 2011, Karnataka’s 0 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 June Lokayukta carried out only 357 suo motu 2011 raids against individual officials but Figure 3: Departmental Distribution of Cases (1995-2011) Number received and tried to trap 2,681 officials 0 100 200 300 400 500 600 700 800 900 1000 (and 59 private persons) in response to Taluk/district office, 992 2,159 citizen complaints.5 In other words, police, court, tax, land 115 Trap for every six cases investigated in response Raid 634 to citizen complaints only one is initiated Local government 63 by the department suo motu. Also, the 464 share of raid cases has been decreasing Welfare cluster 28 over the years (Figure 1, p 12). 336 So, a comparison between raid and trap Economic activities 60 Department cases suggests one of the most ­active

99 Lokayuktas is primarily private com­plaint Agriculture and irrigation 28 driven. This in turn suggests that the legal

67 power to initiate action does not determine Regulation 40 whether we have a proactive anti-corrup-

43 tion agency. The ­incentives for administra- Forest 12 tive action seem to lie elsewhere. Interest- Figure 4: Distribution of Cases across Designations (1995-2011) ingly, institutional leadership is seen to Number of cases have a significant impact on the agency’s 1,229 12501,250 performance. For instance, in Karnataka more than 66% of the raid cases by the 10001,000 848 Lokayukta were initiated ­between 2006 and 2011, when justice Santosh Hegde was 750750 Trap the Lokayukta ­(Figure 2). Our finding

500500 agrees with ­reports in the media suggesting 338 steep changes in Lokayukta’s case load with 201 leadership changes (Aiyappa 2011). But 250250 Raid 133 112 75 note that during the period covered in our 37 0 0 data, raids were conducted under the Pre- Petty officials officials MidMid-ranking-ranking officials officials ProfessionalsProfessionals HigherHigher officials officials Designation Clusters vention of Corruption Act, 1988. The new Law and Governance Initiative have exam- and largely resolved by existing legislation powers conferred by the recent amend- ined a comprehensive data set of all raid and institutional design of the Lokayukta ments to the Karnataka Lokayukta Act, and trap cases handled by Karnataka’s in, say, Karnataka. It has ­ignored critical 1984 that granted the Lokayukta suo motu Lokayukta between 1995 and 2011. This issues that may have little or nothing to do powers to investigate cases are yet to have data was obtained under the Right to with the design of the Lokpal itself but a signi­ficant impact on the­institutional ca- ­Information Act 2005. While our final ­affect its performance. We conclude that a pacity for proactive ­intervention. analysis and conclusions will follow shortly, bill that does not ­assimilate the experience we hope to contribute to the present of existing anti-corruption agencies in 3 Departmental Distribution ­debates with our preliminary findings states like Karnataka is doomed to fail. The There have been several estimates of the through this article. Our analysis suggests rest of the ­discussion is organised around departmental distribution of corruption in that the policy debate on the Lokpal has discrete themes that have been highlighted India. These studies often rely on survey data ­focused on issues that have been anticipated in the current Lokpal debate. of impressions of the public6 or from self

Economic & Political Weekly EPW January 7, 2012 vol xlvii no 1 13 COMMENTARY Figure 5: Processing Rate of Cases (as Percentage of Total Cases) (1995-2011) includes officers of all categories will be 90 80.5 overwhelmed by cases against the lower 80.080 bureaucracy. While the current debate Trap 65.9 63.9 seeks to emphasise the blameworthiness 60.060 54.4 Raid and legal culpability of petty and grand

43.3 41.4 corruption equally, the institutional impact 40.040 on the allocation of scarce prosecutorial resources will be a serious one.8 Further, 20.02 it may be more useful to analyse the

2.8 percentage of prosecuted officers from a 1.4 0.6 0.003 0.00 particular category from among their to- Investigated Sanctioned Chargesheeted Trial completed Convicted Investigated Sanctioned ChargesheetedStage Trial completed Convicted tal cadre strength but ­unfortunately this disclosure.7 We carried out a department- 4 Petty vs Grand Corruption data is unavailable to us at this point. wise analysis of cases of corruption to The distinction between petty and grand map the focus of the Lokayukta’s work corruption is a well-established one in the 5 Process of Investigation (Figure 3, p 13). academic literature on corruption. It has The creation of a strong investigation Interestingly, more than 80% of the trap emerged as one of the contentious issues and prosecution arm has been central to cases are related to four essential func- in the Lokpal debate with respect to the the Indian debate on the Lokpal. It is tions of government: local government inclusion of Group C and D officials within suggested that refusal of sanction for pros- (24.06%), administration – taluk/district its jurisdiction. Though our data is yet to ecution and the failure of the agency to office, police, court, tax, land, revenue be organised in line with Groups A, B, C complete investigations are the key prob- (37.65%), welfare (17.61%), and regulation and D categorisation it allows us to lems that a Lokpal should be designed to (2.54%). The rest of the cases are divided respond to this issue with greater insight avoid (see, for instance, the Jan Lokpal between agriculture and irrigation (3.76%), than is currently the case (Figure 4, p 13). Bill 2011). In Karnataka, our analysis leads forests (1.63%), and economic activities Nearly half of the officials against whom to the conclusion that neither of these (12.75%). The corresponding depart­ the Karnataka Lokayukta has proceeded problems is a critical hindrance for the mental shares for raid cases are as follows: against are officials in the lower bureau- anti-corruption agency. ­local government (18.21%), administra- cratic scale while about 10% are senior Of all the cases, in 65.9% (43.3%) of the tion (33.24%), welfare (8.09%), regula- ­officials. Only 24 officials out of 3,038 (0.8%) trap (raid) cases sanction for prosecution tion (11.56%), agriculture and irrigation belong to the Indian Administrative Service was granted (Figure 5). But when seen (8.09%), forests (3.47%), and economic (IAS), the Indian Police Service (IPS), ­Indian as a function of investigated cases, the activities (17.34%). Two observations are Forest Service, and the Karnataka Admini­ percentage of cases receiving sanction for in order here. One, the distribution is not strative Service (kas) cadres. The upshot prosecution goes up to 94.3 (90.5)% of determined by the Lokayukta as a bulk of of this discussion is that there is ­little trap (raid) cases (Figure 6, p 15). Further, the cases arise out of citizen complaints. doubt that the case docket of a Lokpal that the trap (raid) cases in which sanction is Two, given the growing importance of the welfare function of the State the overall share of essential functions is likely to Ear to the Ground ­increase rather than decrease. Writings on Class and Caste www.navayana.org Some participants in the debate have suggested that structural reform of gov- K. BALAGOPAL Paperback | 488 pages | 6 x 9.25 in | Rs 550 ernment – the withdrawal of the state from non-core activities – will reduce Balagopal’s writings, from the early 1980s till he died in 2009, offer us a rare insight into the making of modern India. Civil rights work provided Balagopal the cause and context . Our preliminary to engage with history, the public sphere and political change. He wrote through nearly analysis suggests that this is not a quick three tumultuous decades: on encounter deaths; struggles of agricultural labourers; the fix since at this stage, even a complete shifting dynamics of class and caste in the 1980s and thereafter in Andhra Pradesh; the venality and tyranny of the Indian state; on the importance of re-figuring the caste order withdrawal of the State from economic as one that denied the right of civil existence to vast numbers of its constituents; the activities will have only a marginal im- centrality one ought to grant patriarchy in considerations of social injustice; the destruc- tive logic of development that emerged in the India of the 1990s, dishonouring its citizens’ pact on the level of corruption measured ‘As a human rights worker active since right to life, liberty and livelihood. This volume comprises essays largely drawn from EPW, in terms of number of cases of corrup- 1981, and slightly older than Balagopal, I to which he was a regular contributor. tion. So, tackling corruption may require remember him as a magical figure. The writ- Balagopal was too self-effacing to put together his writings into a volume. But it is a more fundamental restructuring of the ings in this volume help interpret the often through his writings that his legacy lives on, giving us a roadmap for future struggles. administrative process as it is unlikely chaotic developments in Andhra Pradesh, Distributed by IPD Alternatives, New Delhi. Ph 011-26492040 [email protected] that reducing the size of government will and provide a model tool for understanding Hyderabad Book Trust. Ph 040-23521849 [email protected] other regional realities of India.’ Buy online: swb.co.in and Flipkart.com have a significant impact on the levels —BINAYAK SEN Navayana Publishing 155, 2nd Floor, Shahpur Jat, New Delhi 110049 Ph 011-26494795 of corruption.

14 January 7, 2012 vol xlvii no 1 EPW Economic & Political Weekly COMMENTARY Figure 6: Actual Processing Rate of Cases (1995-2011, in %) will not tackle the core problem with a 97.1 10 100.0100 94.3 95.4 criminal trial in India: the trial stage. 90.5

80.5 80.080 7 Is Karnataka’s Lokayukta a Trap Successful Model?

60.060 As the above discussion shows, if criminal 54.4 Raid conviction is used as the measure of suc- cess then the best Lokayukta in the coun- 40.040 try is undoubtedly a failure. But a caveat is 20.5 20.0 in order: the Lokayukta does not adminis- 20.020 tratively control the criminal court. Hence, 4.3 3.4 we should attribute this failure to the 0.00 choice of a criminal conviction model as InvestigatedInvestigated SanctionedSanctioned ChargesheetedChargesheeted TrialTrial completed completed ConvictedConvicted Stage the centrepiece of our anti-corruption yet to be granted are on an average 1.63 framework in Karnataka does not impede strategy. Alternatively, the performance of (2.14) years old and the median case in investigation of corruption offences. The the Lokayukta may be assessed by estimat- this category is two (three) years old. We Indian public debate on the Lokpal has ing its political and symbolic impact. While do not have further details about the cases focused extensively on the need to equip we do not have data to systematically in which sanction has not been obtained the institution with extraordinary powers of evaluate the function of the Karnataka so far. So, we cannot say if the delay in investigation. Our analysis leads to the con- Lokayukta, maybe the Lokayukta’s capacity these cases can be attributed to political clusion that much of the Indian debate has to occupy media attention through the and/or bureaucratic interference. Howev- sought to extinguish a problem that does spectacle of the raid and high profile er, we can say that almost all the raid cas- not have a very significant impact on the investigations into politically charged cases es in which sanction is pending involve ­effectiveness of the anti-corruption agency. that have unseated a chief minister, shifted senior officials, includingIAS , IPS, and KAS the public mood and ­enhanced political cadre officials, and chief engineers. But in 6 Criminal Trial: The Core Problem accountability, and could be counted as an overwhelming majority of the cases the On the filing of a charge sheet in a special significant successes. However, we should need for sanction for prosecution has not designated court, every case investigated not overstate these effects because the operated as a significant hindrance to the by the Lokayukta enters the criminal ­justice ruling party’s candidates have routinely functioning of the Lokayukta in Karnataka. system. We have noted that the perfor­ won most by-elections to the Karnataka The public debate has also emphasised mance of the Lokayukta in Karnataka has legislative assembly in the ­recent past. A the failure to complete investigations as been creditable on most parameters dis- third potential measure of success of the one of the key problems to be resolved in cussed above. However, the story changes Lokayukta as an anti-corruption agency the design of the Lokpal. We analysed the after charge sheets are filed. may be its capacity to initiate and affect capacity of the Karnataka Lokayukta to Of the trap (raid) cases in which charge administrative measures against officials, resolve this problem within their existing sheets have been filed, 95.7% (96.6%) are including recovery of damages. At this legal and bureaucratic framework. under trial. The average age of the trap stage we do not have the data to estimate In 80.5% (54.4%) of the trap (raid) cases, (raid) cases under trial is 5.1(8) years old or assess these effects and this will be investigation has been completed and this and the median case in this category in the emphasis for future ­research. A final does not vary significantly depending on four (six) years old. Further, of all the trap assessment of the Lokayukta in Karnataka the rank of official being investigated, ex- (raid) cases investigated and under trial would need a more ­fully developed nor- cept at the very highest level. This high in- only 15 (one) have resulted in convictions. mative framework for assessment. How- vestigation rate is particularly impressive. The conviction rate of 20.5% (20%) in trap ever, even at this stage we may conclude The trap (raid) cases pending investigation (raid) cases is much lower than the rate of that the Lokayukta fails to achieve its are on an average about 1.1 (2.14) years old convictions in criminal prosecutions in anti- primary purpose: the criminal conviction and the median case in this category is one corruption cases in India in recent years, of corrupt officials. (two) year old. Similarly, charge sheets which is between 34% and 40% (NCRB were filed in 97.1% (95.4%) of the trap 2007, 2008, 2009, Table 9.1, Col 23).9 This 8 Concluding Remarks (raid) cases in which investigations have suggests that this is the key problem that The presumption of a criminal conviction been completed and sanction for prosecu- lies at the core of a criminal conviction model is at the core of the Lokpal Bill, tion has been obtained. This processing model for tackling corruption in India. which means that it will come up against rate compares ­favourably with the rate at The Indian debate on the Lokpal has the same environmental limits – the effi- which criminal cases in general are proc- focused extensively on the remedying insti- cacy of the criminal justice system – that essed in India. The success of the Karnataka tutional inefficiencies at the complaint and the Lokayukta in Karnataka confronts. Lokayukta in investigating cases suggests investigation stage in the Lokpal. No matter Without highly contentious legal reforms, that the ­existing legislative and bureaucratic how successful these innovations are, they an extremely powerful agency, which the

Economic & Political Weekly EPW January 7, 2012 vol xlvii no 1 15 COMMENTARY Jan Lokpal Bill promises to establish, can 8 Inadvisability of treating petty and grand corrup- Prevention of Corruption Act (1988): available at tion alike and using the same agencies and legal http://lokayukta.kar.nic.in/preact.htm at best marginally improve investigation instruments to deal with them has been questioned Quah, Jon S T (2003): “Combating Corruption in ­India, rates and filing of charge sheet in corrup- for quite some time (e g, Rowat 1984; Panchu 2011). Some Lessons from Asian Experiences” in Stephn 9 Conviction rate is calculated by dividing the cases Howes et al (ed.), State-Level Reforms in India: To- tion cases without securing more convic- resulting in conviction by the total number of cases wards More Effective Government (New Delhi: tions. The proposal for the Lokpal at the in which trial has been completed and judgment MacMillan). passed. Of all the cases, the share of cases leading Rowat, Donald C (1984): “The Suitability of the Ombuds- moment fails to address this core problem to conviction is close to 0.5%. man Plan for Developing Countries”, International and for that reason is bound to fail to 10 Again, this issue has been discussed for long. See, Review of Administrative Sciences, 50(3), pp 207-11. for instance, Palmer (1985), Alexander (1995), Standing Committee (2011): The 48th Report of the achieve its primary purpose: the criminal and Quah (2003). ­Department Related Standing Committee on Per- sonnel, Public Grievances, Law and Justice on the conviction of corrupt officials. Lokpal Bill, 2011 (presented to the Rajya Sabha on References 9 December, 2011), available at http://www.prs- Aiyappa, Manu (2011): “Complaints at Karnataka india.org/uploads/media/Lokpal/SCR%20Lok- pal%20Bill%202011.pdf Notes Lokayukta Dwindle”, Times of India, 23 Nov­ember, Stark, Joshua J M (2011): “Kerala’s Ombudsman: A 1 Orissa Government (nd). available at http://timesofindia.indiatimes.com/ city/bangalore/Complaints-at-Karnataka- Mismatch of Mission and Capabilities”, Governance: 2 Standing Committee (2011, para 3.8). Lokayukta-dwindle/articleshow/ 10836823.cms An International Journal of Policy, Administration, and Institutions, 24 (2): pp 389-92. 3 Standing Committee (2011, passim) and Stark (2011). Alexander, P C (1995): The Perils of Democracy (Bombay: The Jan Lokpal Bill (2011): available at http://ibnlive. 4 Section 7(1)(b) of the Karnataka Lokayukta Act, Somaiya Publications). in.com/news/full-text-the-jan-Lokpal-bill/148401- 1984 as amended by the Karnataka Lokayukta iPaidaBribe.com (nd): www.IPaidaBribe.com (Amendment) Act, 2010. 53.html Orissa Government (nd), http://rtiorissa.gov.in/dept_ The Karnataka Lokayukta Act (1984): available at http:// 5 In a raid case, the police wing of the Lokayukta home.php?id=24&oid=193&sid=2 www.kar.nic.in/lokayukta/karnataka_ lokayuk- raids the office and/or residential premises of a Palmer, Leslie (1985): The Control of Bureaucratic ta_act.htm public servant to ascertain disproportionate ­assets. Corruption: Case Studies in Asia (New Delhi: The Karnataka Lokayukta (Amendment) Act, (2010): In a trap case, the Lokayukta police lay a trap to Allied Publications). available at http://dpal.kar.nic.in/ ao2010% catch the public servants red-handed in the act of Paul, Samuel and Manubhai Shah (1997): “Corruption 5C25of2010(E).pdf accepting a bribe. The latter is initiated in response in Public Service Delivery” in Guhan and Paul (ed.), The National Crime Records Bureau (2007-09): Crime to complaints, while the former is based on the Corruption in India: Agenda for Action (New Delhi: in India, available at http://ncrb.nic.in/ ­intelligence collected by the Lokayukta police. Vision Books). Transparency International (2005): India Corruption 6 See, for instance, Transparency International (2005). Panchu, Sriram (2011): “Lokpal: Where Do We Stand Study, available at http://www. transparency. org/ 7 See, for instance, iPaidaBribe.com (nd) and Paul Now, and How We Got Here”, Economic & Political regional_pages/asia_pacific/newsroom/news_ar- and Shah (1997). Weekly, XLVI (41), pp 19-21. chive2/india_corruption_study_2005

UN-sponsored social reformist program­ From Parliamentary mes, the Indian government punishes citi- zens who assert their rights under the UN’s to Paramilitary Democracy ­Human Rights charter. The UN Human Rights Council will re- view India’s human rights record during Sumanta Banerjee the last four years when it holds its session in mid-2012. Meanwhile, the Working India’s atrocious record with ll through the politically phoney Group on Human Rights (WGHR), a coali- regard to the human rights of and financially wasteful filibuster- tion of Indian human rights organisations, its citizens suggests that, with A ing on the floors of the has submitted to the UN a report which in the recent winter session, neither the ­exposes the atrocities carried on by the parliamentary complicity, it is treasury benches nor the opposition cared ­Indian state behind the façade of a parlia- degenerating to the status of a to remember one particular date during mentary democracy. It records incidents paramilitary democracy. that session – 10 December, World Human which are already known to us – 789 extra- Rights Day. Yet, all other United Nations- judicial killings in Manipur between 2007 mandated dates – like International Wom- and 2010; the discovery of about 2,700 un- en’s Day on 8 March, World Population Day marked graves of people killed by the In- on 11 July, International Literacy Day on 8 dian security forces in Kashmir. When September – are observed with a lot of fan- speaking of Kashmir, we should also add fare by the Indian government by giving that the ­Association of Parents of Dis­ away awards to functionaries for carrying appeared Persons (APDP) in a recent sub- out the tasks under those respective UN mission to the Jammu and Kashmir State mandates. That New Delhi keeps out 10 De- Human Rights Commission, has acknow­ cember from its list of similar ­official cere- ledged that ­besides Indian security forces, monies is indicative of the ­Indian state’s Islamic militant groups have also whisked Sumanta Banerjee ([email protected]) unease over its atrocious record with regard away their sons (The Hindu, 12 December is best known for his book In the Wake of to the human rights of its citizens. In contrast 2011). This brings us to the other important Naxalbari: A History of the Naxalite Movement with its acts of ­rewarding non-governmental issue of the culpability of extra-state agen- in India (1980). organisations and individuals under other cies in the violation of human rights in

16 January 7, 2012 vol xlvii no 1 EPW Economic & Political Weekly