79624 India Land Policies for Growth and Poverty Reduction India Land Policies for Growth and Poverty Reduction Agriculture and Rural Development Sector Unit South Asia Region OXFORD UNIVERSITY PRESS THE WORLD BANK OXFORD UNIVERSITY PRESS YMCA Library Building, Jai Singh Road, New Delhi 110 001 Oxford Universiry Press is a department of the Universiry of Oxford. It furthers the Universiry's objective of excellence in research, scholarship, and education by publishing worldwide in Oxford New York Auckland Cape Town Dares Salaam Hong Kong Karachi Kuala Lumpur Madrid Melbourne Mexico Ciry Nairobi New Delhi Shanghai Taipei Toronto With offices in Argentina Austria Brazil Chile Czech Republic France Greece Guatemala Hungary Italy Japan Poland Portugal Singapore South Korea Switzerland Thailand Turkey Ukraine Vietnam Oxford is a registered trade mark of Oxford Universiry Press in the UK and in certain other countries. Published in India by Oxford Universiry Press, New Delhi ©The World Bank 2007 The moral rights of the authors have been asserted Database right Oxford Universiry Press (maker) First published 2007 All rights reserved. No part of this publication may be reproduced, or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without permission in writing from Oxford Universiry Press. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford Universiry Press, at the address above You must not circulate this book in any other binding or cover and you must impose this same condition on any acquirer The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Executive Director of the International Bank for Reconstruction and Development/The World Bank, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colours, denominations, and other information shbwn on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement of acceptance of such boundaries. ISBN-13: 978-0-19-568959-4 ISBN-10: 0-19-568959-3 This report has been discussed with the Government oflndia but does not necessarily bear their approval for all its contents, especially where the World Bank has stated its judgement, opinion, and policy recommendations. Typeset in AGaramond 11.5/15 by Eleven Arts, Keshav Puram, Delhi II 0 035 Printed in India by DeUnique, New Delhi 110 018 Published by Oxford Universiry Press YMCA Library Building, Jai Singh Road, New Delhi 110 001 Contents Tables, Figures, and Boxes Vll Acknowledgements IX Abbreviations XI Executive Summary X Ill 1. Introduction 2. Land Administration: Institutional Structure and Key Challenges 8 3. Improving Textual Records: Examples and Potential 22 4. Improving the Status of Spatial Records 34 5. Towards Greater Tenure Security 45 6. Land Ownership Reform 54 7. Land Lease Markets 70 8. Land Sales Markets 84 9. Towards an Integrated Land Policy 92 Appendices 1. Policy Matrix 97 2. Figures and Tables 102 Reftrences 116 Tables, Figures, and Boxes TABLES 1.1: Number of land parcels in different countries 5 2.1: Land revenue as a percentage oflndian states' own revenue 12 2.2: Summary of status of rural maps in Andhra Pradesh 15 3.1: Status of modernizing land records in selected Indian states 23 3.2: Status of registration in selected Indian states 30 3.3: Global levels of land taxation 31 4.1: Role of government and the private sector in different countries 38 4.2: Status of modernizing surveying in selected Indian states 41 5.1: Summary on status of modernizing land administration in selected states 53 6.1: Share of households and area affected by land reforms in Indian states 57 6.2: Characteristics of households in states with low and high land reform effort 59 6.3: Land reform and human capital accumulation 60 6.4: Land reforms and growth with other variables 62 6.5: Impact of land reform on income and asset growth 63 6.6: Land price determinants 67 7.1: Households participating in land rental markets 71 7.2: Household characteristics by rental market participation in 1999 74 7.3: Determinants of participation in land rental markets 76 7.4: Impact of restrictions on productivity and equity 78 7.5: Gender gaps in agricultural wage rates 79 7.6: Determinants of daily wages for casual labourers in India 80 8.1: Household characteristics by market participation status 85 8.2: Determinants of land sales market participation 86 BLES, FIGURES, AND BOXES A2.1: Status of computerization of land records in selected states 104 A2.2: Status of computerization of registration in selected states 108 A2.3: Status of modernizing survey in selected states 112 FIGURES 1.1: Ratio ofland costlsq m to per capita GOP 3 1.2: Corruption in land administration across different Indian states 4 2.1: Agencies involved in land administration in India 10 2.2: Revenue hierarchy in Punjab 11 2.3: Structure of the survey and settlement department in Tamil Nadu 13 2.4: Structure of the registration and stamps department in Tamil Nadu 17 2.5: Land administration institutions and data flows 19 3.1: Manual system of effecting land transfers 24 3.2: Re-engineered system for land transfers in Karnataka 25 A2.1: Illustration of a tip pan 102 A2.2: Illustration of a village map 103 BOXES 2.1: Different types of settlement 9 4.1: Guidelines for PPPs involving licensing of private surveyors 39 5.1: The English system of title registration 46 5.2: Converting from deeds to titles in Scotland: A long process 49 5.3: Kerala's experience with 'Torrens offices' 51 6.1: The far-reaching long-term impact of colonial settlement 55 6.2: The 2005 Hindu Succession Act Amendment-a far-reaching change 69 8.1: Giving land rights to groups as an alternative to prohibiting land sales 90 8.2: Ingredients of a comprehensive property reform-the example of Mexico 91 Acknowledgements This report was prepared by a team led by Klaus Ministry of Rural Development and the Department Deininger, under the overall guidance of Adolfo ofLand Resources, the National Informatics Centre, Brizzi and Gajanand Pathmanathan, Sector Managers. and the Commission for Agricultural Cost and Prices, Major contributors to the report, which was initiated Government oflndia. In addition, invaluable support by Isabel Lavadenz, are Tony Burns, Edward Cook, was received from the officials of the Departments of Tim Hans tad, Songqing Jin, Kundan Kumar, Karen Revenue, Stamps and Registration, Rural Development, Macours, Hari K. Nagarajan, Kevin Nettle, Susan and Information Technology in numerous states. Nichols, Robin Nielsen, Tore Olsen, Neel Ratan, Key inputs for the report from policymakers at the Gutala V. Subrahmanyam, Sanjay Upadhyay, Sonal national and state level, academia, civil society, and Vats, and Barbara Verardo. Invaluable administrative international organizations were received during a two- assistance was provided by Sarita Rana, Sandra Sousa, day workshop on land policies and administration and Lilac Thomas. The team would like to thank held in New Delhi in January 2006 which was Deepak Ahluwalia, Deepak Bhatia, Rabih Karaky, organized jointly with the Ministry of Rural Martien Van Nieuwkoop, and Michael Carter for Development and Commission for Agricultural helpful comments and the peer reviewers, Robert Cost and Prices, with support from the Food and Buckley, Jonathan Lindsay, Pedro Olinto, and N.C. Agricultural Organization of the United Nations, Saxena for their valuable guidance and advice the Rural Development Institute, and the UK's throughout the preparation of the report. Department for International Development (DFID). We gratefully acknowledge the cooperation and We would also like to thank DFID for financial valuable assistance provided by officials from the support for preparation of background studies. List of Abbreviations BC Backward Caste ILlS Integrated Land Information System BOT Build Operate Transfer IT Information Technology CACP Commission for Agricultural Cost and LAA Land Acquisition Act Prices MIT Ministry of Information Technology CESS Centre for Economic and Social Studies MoRD Ministry of Rural Development CLR Computerization of Land Records NSS National Sample Survey CMLR Comprehensive Modernization of Land NIC National Informatics Centre Records PESA Panchayat Extension to Scheduled Areas DFID Department for International Act (1996) Development ppp Public-Private Partnership EC Encumbrance Certificate RI Revenue Inspector EGS Employee Guarantee Scheme RoR Record of Rights ETS Electronic Total Station RTC Record ofTenancy and Cultivation FIG International Federation of Surveyors SC/ST Scheduled Caste/Tribe FMB Field Measurement Book SDC State Data Centre GO Government Order SHG self-help group Go! Government oflndia SRO Subregistry office GPS Global Positioning System UNECE UN Economic Commission for Europe HSA Hindu Succession Act ( 1956) VA Village Accountant IGR Inspector General of Revenue and Stamps WAN Wide Area Network Unit Measurements ha =hectare kg = kilogramme km = kilometre Executive Summary into neglect as officers had to attend to myriad other INTRODUCTION duties-is in urgent need of improvement to In India, land continues to be of enormous economic, effectively fulfil its functions in the 21st century. The social, and symbolic relevance. The way in which access key question is not 'whether' but 'how' to go about to land can be obtained and its ownership documented this task most effectively and which policy issues is at the core of the livelihood of the large majority need to be addressed to ensure that doing so will be of the poor, especially in rural and tribal areas and sustainable and will have an impact on improved determines the extent to which increasingly scarce land access and greater productivity of land use. natural resources are managed. Land policies and Recent advances by a number of states that have used administration are critical determinants of the information technology (IT) to make significant transaction cost associated with modalities to access progress in modernizing their land administration land for business and residential use and, through system, as well as a nationwide survey evidence, provide the ease of using land as collateral for credit, the an opportunity to revisit this question. The goal of development of the financial sector. Land also this report is to draw on recent empirical evidence continues to be a major source of government revenue to assess the extent of improvement in land-related and a key element for implementing government institutions in individual states as well as the impact programmes. This implies that land policies and of a range ofland-related policies. In the case ofland institutions will have a far-reaching impact on the administration we draw on a review ofland records, ability to sustain India's current high rate of growth, survey and settlement, and land registration in 14 the extent to which such growth reaches the poor, and states while a nationwide survey that spans the period the level and spatial distribution of economic activity. 1982-99 provides the basis for an assessment ofland As land administration-through the revenue policies. Analysis of these allows identification of action department-was the core of colonial structures and areas by policymakers to ensure that opportunities continues to be a key pillar oflocal government today, provided for modern technology and the convergence issues relating to land have long been the subject of of interests-visible in recent initiatives such as the an animated policy debate. There is consensus among National e-Governance Programme-can be most policymakers that land administration-which fell effectively utilized. XECUTIVE SUMMARY INSTITUTIONAL STRUCTURE the survey department's responsibility for maintaining AND CHALLENGES an accurate spatial record of land ownership lapsed and the responsibility was transferred to municipal Most of the institutions and processes for administering corporations which were interested in maintaining land in India were adopted from the British at the time spatial records only for tax purposes. Some have done of independence and have been little modified since. this relatively well, implying that spatial records that Given that their main purpose was to raise resources, were established for tax purposes can, in principle, the land revenue as well as the survey and settlement be used as a basis for ownership records. However, departments-the key institutions responsible for others did not live up to the task, resulting in outdated land administration in India-were focused on map products of inferior quality that appear to be productive rural areas that provided opportunities to one of the reasons underlying the high levels ofland- generate tax revenues. Both urban and 'marginal' rural related conflict (28 per cent of parcels according areas remained outside the system. This distinction, to one study) in peri-urban areas. Clarifying the which was difficult to justify in the first place, has institutional situation with a view towards eliminating increasingly become archaic in three respects. this 'spatial data vacuum' where required and ensuring First, and most important, any piece of rural land that there is one agency with the necessary capacity that had been transacted through sale at any point to maintain spatial records in rural and urban areas after 1882 entered the land registry system, implying is a priority for policy. that its records are maintained not only by the revenue Last, but by no means the least in importance, even but also the stamps and registration department. This revenue lands that had previously been waste, and thus duplication of institutions increases transaction costs were not subject to settlement surveys, have increasingly for landowners without providing commensurate been brought under agricultural cultivation. This benefits. However, more importantly, it introduces implies that one of the key challenges from a poverty a major source of tenure insecurity and even fraud perspective is to extend land administration to areas because, for a variety of reasons, the records maintained that had been left out. Doing so will require clarification by both institutions may well be inconsistent. While of the interface with the forest department and a the most radical solution, which indeed has been broadening of tenure choices that can be accommodated mooted in policy circles, would be to merge the within the land administration system to include, for two departments, doing so may encounter strong example, forms of communal ownership. Given the bureaucratic resistance. Even if it were feasible, it is large number ofland and people involved, and the high an option only if and when the necessary groundwork, concentration of poverty in these areas, settling these much of which will be discussed in this report, has issues is of very high priority for poverty reduction been undertaken. At the same time, steps to clarify and requires major policy decisions. Some of these the legal situation ofland records and ensuring back- decisions are currently under debate in the discussion office integration are essential. of the Forest Rights Bill. Once the broad policy Second, rural areas at the urban fringe have parameters have been set, a serious implementation increasingly been subject to urbanization. Although effort will be needed. While this can build on the this is associated with a significant increase in land general principles identified in this report, specific values that would warrant a more precise survey, in measures will have to be discussed in more detail. many Indian states the result of this process was that EXECUTIVE SUMMA IMPROVING TEXTUAL LAND and is generating large surpluses from user fees in RECORDS states where it has been rolled out and manual records have been abolished. Third, computerization helped A review of the experience of states with initiatives to improve credit access. Finally, a number of states to computerize revenue records and land registries, demonstrated that computerized records can be used spearheaded by the Ministry of Rural Development as a springboard to integrate revenue records with (MoRD) and the Ministry ofinformation Technology the registry and even spatial data by automating the (MIT), respectively, allows identification of best back-end of the process, efforts to have surveys done practice for this process, the benefits from doing so, before mutation, providing registry officials with access and the challenges that remain to be addressed. In to the land records database before registering a both cases, a number of factors that were critical document, etc. All of this points towards considerable for success can be identified. Although supported unexploited potential for future improvements that by central funds, the most innovative solutions should be utilized. came about as a result of innovation by individual The successful experience allows derivation offour states, subsequently adopted by others. Also, in all best practice principles. First, digitization will be useful cases, innovations were designed to draw heavily only if computerized records are routinely used, on India's strengths in IT, extensive use was made something that requires abolition of manual records. of subcontracting and public-private partnerships This, together with the dynamic nature ofland records (PPPs) to overcome weaknesses of the public sector, (and the fact that such a step may face bureaucratic and emphasis was on broad coverage and quick roll- resistance), puts a premium on selectivity in what needs out of viable innovations with a possibility oflearning to be computerized and on speedy completion of from experience and adding more sophistication later, digitization. This seems to be a key reason why, thus rather than designing a 'perfect' solution from the start. far, only states which outsourced digitization to a Computerization of land records, which is now private operator with clear timelines have been fully or partly operational in Karnataka, Gujarat, successful in making the switch. Second, the ability Rajasthan, Madhya Pradesh, Maharashtra, and Tamil of the computerized system to be financially self- Nadu, has yielded four key benefits. First, it has sufficient and use the income generated to sustain and/ simplified the system and significantly reduced petty or improve the system has been critical in reducing corruption that was traditionally involved in getting political interference and outsourcing specific tasks access to land records. While the amounts involved in in cases where public sector capacity was insufficient. individual cases may have been small, aggregate losses Business models to ensure this need to be encouraged. associated with petty corruption can be very large--a Third, to ensure acceptance of, and confidence in, survey from Karnataka estimates that computerization the system, three factors are essential, namely in this state saved Rs 80 crore of bribes and Rs 6.6 (i) transparency of the process, as embodied in a crore in waiting time per year. This is in addition to verification process with active participation of the largely non-quantifiable impact ofcomputerization landowners; (ii) integrity of the data, through state on villagers' attitude to the bureaucracy that may be data centres with appropriate security features and very important in environments where governance audit trails; and (iii) publicity, that is, making data is an issue. Second, computerization improved the available on the internet. Finally, there is considerable quality with which government services are delivered I.CUTIVE SUMMARY demand for access to land records at the village level digitized and a data architecture that can eventually and a number of strategies have been adopted to provide a basis to link revenue and registration records effectively respond to this demand in the interim while in a way that ensures minimum consistency and the necessary connectivity to allow village-level access portability of software across states; (ii) best practices is gradually being put in place. on achieving the integration of rural land records with The impressive success of some leading states in urban property card systems; and (iii) ways to achieve shifting from manual to electronic records should not functional integration not only with registration conceal the fact that, in many instances, large amounts but also with surveys that go beyond a mere of money have been spent on initiatives to computerize computerization of old manual systems and thus will land records with little impact on the ground as yet. be cheaper and more sustainable in the long term. This implies that there is considerable scope to re- Supporting well-defined pilots to provide guidance orient or accelerate ongoing programmes to enhance on some of these issues will be important to reap full their effectiveness. Our review suggests that the benefits from computerization and should be a effectiveness of Government oflndia (Goi) schemes priority for the Goi. for Computerization of Land Records (CLR) and Computerizing registration of deeds which is fully Comprehensive Modernization of Land Records or partly completed in Andhra Pradesh, Karnataka, (CMLR) can be enhanced by incorporating the best Maharashtra, Rajasthan, and Tamil Nadu, has practices (identified above) more systematically in similarly helped to realize large benefits. One is that, design and evaluation to enhance the effectiveness in a number of states, it has been associated with a of programmes to be funded under this scheme. significant increase in the number of registered land Specifically, computerization initiatives should receive transfers helping to increase revenue from stamp support only if the information to be digitized is duties even though duty rates, in some cases, were relevant for landowners, if a time bound plan for substantially reduced. This suggests that more data verification and abolishing manual records exists, transparent processes for registration and property if key elements of the process are oursourced to the valuation increased the usefulness of services to private sector, if issues of data integrity are satisfactorily customers and also that demand for registration is addressed, and if they are based on a business model price elastic. In addition, the fact that in some states that has the potential to fully cover operational costs. encumbrance certificates for a significant length of Complying with such requirements is not difficult time are available helps to increase tenure security. The and the ability to draw on the experience of states ability to obtain these electronically via the Internet that have already completed the process will allow implies a significant reduction in transaction costs 'latecomer' states to realize significant costs savings. for sellers and purchasers as well as banks although Efforts to make this experience more widely and evidence regarding its impact on credit market easily accessible to other states merits support. activity is still limited. Finally, computerizing registry A second issue of relevance for policymakers relates data has created the preconditions for a functional to the fact that, even for the leading states, the benefits integration between registration and records that will from record computerization could be significantly have to be a key element of any effort to make the enhanced if guidance and model regulations were land administration system more conclusive and thus available on a number of key issues that include but reduce the transaction costs and insecurity involved are not limited to (i) the nature of information to be in dealing with land. EXECUTIVE SUMMARY 'Best practice' elements that have facilitated the world. Unless these are reduced, even the best effective computerization of the registry are as technical solutions for improving land records are follows. First, more than in the case of land records, unlikely to be sustainable. Options to partially replace computerization of the registry has been based on stamp duty with higher land taxes-levied on market some re-engineering of the underlying business values and ideally shared between local bodies and practices that involved the standardization and states-would be more in line with international best simplification of deeds (contrary to digitizing virtually practice and need to be explored urgently. Second, it everything that is on record), the development of a is important to ensure completeness and consistency process to automate market valuation, and the setting of revenue and registry records. This will require of clear performance standards (parallel to including regulatory changes to ensure that mutations, for mutation from the start). Second, to ensure financial example, through succession, that did not need to be sustainability, roll-out proceeded from offices with registere.d in the past, will be registered automatically high transaction volume to those with less land market and free of charge, which will not be difficult if the activity, with adaptations made in the process. Third, systems are linked electronically. Third, the lack of a outsourcing to the private sector was a key element consistent reference to spatial parcel identifiers in the from the very beginning and payment schedules were registry is one of the key sources of incompleteness designed to have high-volume offices implicitly and insecurity and needs to be tackled, as will subsidize low-volume ones, thereby ensuring equality be discussed in greater detail later. Finally, there is of user fees throughout the state. From a purely a strong perception about registry officials lacking commercial perspective, computerization of land accountability. This issue can be addressed once the registration is straightforward and the main source officials have easy access to the information needed of resistance is likely to be political, often from people to perform basic checks on the transactions that are in the system whose ability to receive rents would be offered to them for registration. While none of these negatively affected. Similarly to what was the case for issues poses insurmountable difficulties and many revenue records, it is important for the Go I to provide of them will require only small administrative changes states with incentives to overcome such resistance and in regulations, leadership by the Go I to support pilots, draw equal with the more progressive states in terms drive the process and ensure communication across of computerizing registration. To the extent that the states will be important role, especially if and can be ease of transacting land provided by a well-functioning linked to the provision of finances. registry will be an important consideration in investment decisions by firms or individuals who are IMPROVING THE STATUS OF not limited to a given state, failure to do so can over SPATIAL LAND RECORDS time translate into widening disparities between more advanced and backward states. Regarding spatial data, there is consensus on three Also, a number of policy issues need to be addressed issues. First, the establishment of a comprehensive, even in the more advanced states. First, even though reasonably accurate, cost-effective, and affordable some states have moved to reduce high levels of stamp spatial framework will be a key element of any strategy duty that tended to drive transactions into informality, to improve India's land administration system. the taxes levied on property transfers in India, in Second, due to long neglect and gaps in institutional contrast to land taxes, remain among the highest in responsibilities, the reliability of existing spatial data ECUTIVE SUMMARY is much inferior to that of the textual database and attention. Third, even though the need to recover neither simple digitization of existing data nor attempts cost was critical to ensure that solutions for textual to resurrect the tradition of revisional surveys records remained focused on clients' needs and overall established by the British will be sufficient to address affordability, in addition to helping to avoid undue the problem. Third, there is an urgent need to political interference, it was not adopted in the case develop viable and replicable models to improve and of spatial pilots. Finally, although none of the states maintain spatial records, along the lines of that that successfully completed computerization of textual achieved for textual records. This is more difficult records did so without outsourcing key components because of the highly specialized nature of surveying, to the private sector, surveying remains a government the presence of strong vested interests pushing for monopoly in all but three Indian states (Andhra technically sophisticated rather than the economically Pradesh, Karnataka, and Madhya Pradesh). Even in viable options, security restrictions related to access to these states what passes as 'private sector' appears de geodetic coordinates and the use of aerial photography, facto more as an appendix to the government with and the fact that surveying costs tend to increase limited autonomy either in commercial/financial (rates exponentially with precision. A spatially differentiated are set by the government and there is little contact approach will be needed that chooses strategies based between private suppliers and their clients) or on a comparison of the costs of available technical technical matters (technology is narrowly prescribed options to the likely benefits in a given environment and output has to be delivered in paper form and and where cost recovery options are in line with checked by officials). beneficiaries' capacity to pay. Still, there is little reason To address this issue, it is desirable that: (i) public to believe that the principles that helped to achieve sector activity focuses first and foremost on success for textual records will not be applicable in comprehensive coverage with a low precision cadastral the area of surveying as well. index map-something that is a dear public good- In fact, a cursory review of pilot surveying activities rather than establishing islands of high quality spatial suggests that a key reason for the failure to develop data in an ocean that remains largely uncharted. replicable models is that these key success factors have Combining satellite imagery with existing spatial been almost completely neglected. First, instead of data (that is, village maps, tax maps, etc.) offers an drawing on existing information and expanding from appropriate and cost-effective opportunity to do so; there to fill gaps, surveying pilots tried to start from (ii) pilots focus on developing scalable and cost-effective scratch with a complete re-survey. Second, instead ways to generate spatial data and link them to textual of considering a range of technology options and records that are applicable to archetypical situations-- then focusing on cost-effective ones where India has for example, unrecorded subdivisions, complete world class capacity, 'modern technology' was narrowly change in land use patterns, loss of spatial data, or defined in terms of electronic total stations (ETS), an complete lack of survey-and be geared towards expensive option in which India has little comparative producing procedural insights that form the basis for advantage and which provides a level of precision regulations and guidelines; and (iii) a major effort that is not needed for at least 80 per cent of the to expand capacity and increase the role of private country. Cheaper and more appropriate options such surveyors (with structures for accountability) is as photogrammetry and use of handheld global mounted, so that, instead of crowding out the private positioning system (G PS) devices received little sector, public effort is focused on setting a regulatory EXECUTIVE SUMMAR framework and conducting or at least financing surveys facilitating searches by parcel as well as by person by in (marginal and tribal) lands that would otherwise introducing/improv~ng parcel maps and generating not be attractive to private operators. parcel indices in the registers, and ensuring that all documents entered into the registry have undergone PUTTING THE ELEMENTS at least rudimentary checks for validity by the registrar TOGETHER: TOWARDS GREATER have been adopted all around the world to improve TENURE SECURITY AND EASIER deeds systems (for example, in the Netherlands, South TRANSFERABILITY OF LAND Africa, the US). Moreover, improved deeds systems of this nature constitute an essential first step in making From the very start, the goal of efforts to modernize the transition towards a title system. Taking these land administration in India was to increase tenure steps is essential for India irrespective of whether or security and reduce the cost of transferring land. While not an eventual transition to Torrens title is envisaged. modernization of textual and spatial records is a Even in the states where land administration is greatly necessary condition for this, it will be fully effective advanced, considerable further action is required to only if accompanied by an appropriate legal and reach the point where all of the above are implemented. regulatory framework. In this context, a key concern Once this is the case, the decision on whether to that is widely debated in policy circles, is whether, make the transition towards a full title registration and, if yes, when and how, India should make the system will hinge on three factors. First, there needs transition towards a system of title registration, often to be sufficient political support to make the legal and also referred to as a 'Torrens' system. institutional changes needed by a system of title over Both deeds and title registration are intended to and above one of deeds registration. Second, even put rights in land on public record. The key difference if the capacity to run a title registration system is is that, while under title registration the register serves available, a consensus on the desirability of incurring as primary evidence of ownership, a deed only provides these costs needs to be reached. Finally, an important evidence of an isolated transaction that says nothing decision relates to the establishment of a guarantee about the validity of this transaction. Simply put, fund that is essential in most jurisdictions for an under a deeds system potential purchasers will need indefeasible system of title registration in contrast to to expend resources to investigate whether the seller's merely an improved deeds system. This implies that title is genuine whereas under a title system this is a title system is more expensive to run but has lower not needed as the validity of such claims has already costs of conveyancing whereas the cost of operating been checked by the registry system. Under most title a deeds system is lower but the effort required from registration systems, significant emphasis is placed conveyancers is higher. Two key dimensions for on the 'indefeasibility' of title and in many systems analysis of this issue are related to sustainability and this indefeasibility is supported by a state guarantee. distribution of benefits. Concerning the latter, a title In a deeds system the cost of a title search by a potential registration system imposes higher costs on society buyer depends on: (i) the completeness of the as a whole but makes it cheaper to transfer land for information contained in the registry; (ii) the ease of the parties involved. Regarding the former, tide searching it; and (iii) the reliability of the registry systems are vulnerable to informality and can become information. To reduce this cost, steps to make dysfunctional if a large number of users are not willing registration compulsory, computerizing it and or able to pay the cost of participating in such a system. XECUTIVE SUMMARY In the Indian context, an important barrier to allow realization of90 per cent of the benefits from a participation in the formal system is the high level title registration system at a fraction of the cost, while of stamp duty. at the same time providing a more appropriate basis to A better understanding of the magnitudes involved decide whether transition towards full title is desirable. in the Indian context would be critical to reach an informed decision on whether or not to go for a title LAND OWNERSHIP REFORM registration system. A close monitoring of efforts to move towards title, that have been initiated in Andhra Although improving land administration along the Pradesh is likely to yield important insights for other lines discussed above can have a significant impact Indian states faced with similar questions. Comparing on the welfare of those who own land, it will at best the experience of England and Scotland-two provide indirect benefits to the large number of jurisdictions which, starting from a basis that was households that are landless and, without any tangible much superior to what is currently encountered in even assets, may well be caught in a poverty trap from which the most advanced Indian states, made a successful it is difficult to escape. In India's situation, policies transition from a deeds to a title system over a period to strengthen land administration will have be of decades-with that of less successful experiences complemented by considering opportunities for the of developing countries trying to make such transitions poor to access land as a means of improving their without having in place the infrastructure to support asset status and livelihood. such a system or being aware of the complex issues Policies of land reform, through abolition of involved, suggests that trying to fast-track these intermediary interests, tenancy legislation, and land processes carries significant risks. ownership ceilings, have formed the core of Gol A more immediate goal, to be accomplished at least policies to improve the land ownership structure. in the medium-term, is to functionally integrate the We use empirical evidence from a nationwide survey different databases used in land administration to spanning the period 1982-99 to assess whether such provide landowners with a certificate-which can even reforms, proxied by dummies at the state level, have be called a certificate of title-that combines relevant been associated with higher levels of investment in and current information pertaining to a plot (that is, physical and human capital growth, and how their ownership status, other current rights and obligations impact has evolved over time. The results suggest that in the property, under a range of registered instruments transferring land ownership may have been quite that may include mortgages, rights-of-way, easements, effective, although limited variation in land reform caveats, liens, covenants, etc., and a map that implementation has to be taken into account in unambiguously defines the spatial extent of the land interpreting them. Although they were initially worse parcel and its relationship to adjoining land parcels) off, households in states with higher levels of reform irrespective of the government department maintaining effort saw their welfare and investment improve more the information. If combined with the option to than that of households in states where land reform maintain spatial data at low cost and regulations to implementation remained lacklustre. This is supported ensure that any changes in either textual or spatial by econometric evidence which points towards a records are automatically updated in the system as well positive, significant, and quantitatively important as requiring registry officials to perform at least basic contribution ofland reform to investment and growth validity checks before registering a document, it will with an overall effect that compares favourably to EXECUTIVE SUMM other forms of social spending. At the same time, title to such land include a variety of tenurial econometric evidence also suggests that the positive instruments such as transfer of government wasteland impact of land reform legislation has been declining to private parties, long-term lease of occupancy over time and actually risks becoming negative, rights, and formal recognition of communal tenure. something that would be consistent with the notion Evidence from Andhra Pradesh highlights not only that land reform legislation is no longer very effective that there is a considerable amount of land that does in transferring new land to the poor but continues to not have dear title but also that providing a dear affect the efficiency with which land is used by land certificate (patta) can significantly increase land reform beneficiaries as well as landowners targeted values-by 15 per cent to 20 per cent for privately by land reform, due to the fact that the latter may owned and by 30 per cent to 45 per cent for assigned take preventive measures to escape this threat. or occupied land. Moreover, having a well-defined If the above analysis is correct, other mechanisms right also increases the probability of a plot of land to bring land to the poor will have to be identified. being rented out, thereby providing indirect benefits Evidence from different Indian states points towards to the poor and landless. Finally, the considerable a range of options for doing so. One option that has potential to improve the welfare of women who been implemented with some success in Andhra depend on agriculture by giving them land rights or Pradesh is to provide the poor with legal assistance ensuring that they can inherit such land has often to either resolve pending cases or to ensure been neglected in the past. The 2005 amendment to implementation of court orders on the ground. As the Hindu Succession Act (HSA) breaks new ground the large number of pending court cases generates in this respect and sufficient attention is warranted large negative externalities, one might even think of in disseminating its content so as to allow it to become specific incentives to resolve these, possibly in tandem effective on the ground. with a participatory process of improving the land administration system. A second option, currently LAND LEASE MARKETS being implemented in West Bengal, Karnataka, and Andhra Pradesh, is to provide funds to the poor to Although land rental markets in India remain highly acquire very small plots ofland in the market, either regulated, empirical evidence on the functioning of for residential purposes and market gardening or for land lease markets or the impact of such regulation productive purposes, often through self-help groups on poor peoples' ability to access land or use it (SHGs). At the same time, to avoid negative effects productively remains limited. Contrary to other ofland reform legislation on the welfare of the poor, countries such as China and Vietnam where economic it is desirable to: (i) allow subdivision and transfer of growth has prompted a veritable explosion in land land received through land reform, especially in the rental activity, rental market participation in India course of generational change; (ii) systematically has declined continuously since independence. This empower civil society to help resolve land cases; and is surprising because in India land rental provides (iii) explore the scope and effectiveness ofland markets almost the only option for accessing land by a large as a means of providing land access. pool of landless where as (involuntary) landlessness Third, where large amount ofland continue to be is virtually non-existent in either China or Vietnam. held by the government, regularization should be The fact that more households rented land at any point much easier. Options to provide secure access and in time than benefited from land reforms through ECUTIVE SUMMARY India's entire independent history further illustrates To reap the potential benefits from land rental, the magnitude and policy relevance of this issue. it will be important for states which, in the Indian Use of nationally representative data provides context, have the responsibility for land policy to: four important insights. First, land rental markets (i) make land leasing legal in states where it is not; contribute to greater equity and productivity ofland (ii) explore options to allow sitting long-term tenants use by allowing the 'productive poor' and landless to (for example, baragadars in West Bengal) to acquire access land and obtain returns to their labour that full ownership of all or part of the land they occupy; are significantly higher than what they would be able and (iii) replace rent ceilings and other limitations to obtain in casual labour markets. While, in the on land leasing with more enabling legislation such 1980s, land rental was biased in favour of those with as dissemination of standardized long-term contracts higher levels of assets, this bias disappeared in the late or leases, local help in conflict resolution, etc. As one 1990s, presumably as a result of the growth of the would expect the benefits from liberalization of land non-farm economy. Second, contrary to the hypothesis lease markets to be large, the impact of eliminating of' reverse tenancy'-according to which the majority land lease restrictions in states where doing so is of land entering rental markets is supplied by small politically feasible should be carefully monitored and and marginal farmers in a situation of relative duress- results from doing so disseminated to generate political it is households with higher levels of education that momentum for similar steps in other settings. supply land to the rental market, presumably to take up more lucrative employment in the non-farm LAND SALES MARKETS economy. This interpretation is consistent with the finding that activity in land lease markets increases The discussion above suggests that land rental is an with the overall level of economic development in a important mechanism to increasing efficiency ofland village. Third, state-level land rental restrictions use. However, policymakers have often been concerned reduce the ability of the poor to access land and also that in rural areas where, as a result of credit market have a negative impact on productivity, implying that imperfections, households are not able to fully insure elimination of such restrictions could help to improve against shocks, distress sales may have a negative impact both equity and efficiency. Finally, wage rates in casual on both equity and efficiency. In other words, farmers labour markets for both agriculture and non-farm would be forced to sell off their land, often to usurious activity discriminate strongly against women even moneylenders or other unscrupulous person, at though there is no evidence to suggest that women bargain prices that are well below the productive value are less productive than men in agricultural self- of the land just to ensure their survival in the face of a employment. This suggests that the relative benefits shock. As they will not be able to re-acquire the land from being able to access land through rental markets through purchase once prices return to normal, this would be larger for women than for men, implying would leave them permanently landless, an outcome that freeing up land rental markets could have a that is likely to be undesirable both from an equity and dispropertionately positive gender impact. This is of an efficiency point of view. Indeed, historically, distress particular significance in view of the fact that the sales were a major factor that led to the accumulation extent of gender discrimination does not disappear oflarge amounts ofland by powerful, though often or decrease with economic development. not very efficient, landlords and moneylenders. EX E CUT IV E S U M M A R.Y .. While there are few restrictions on land sales question whether a prohibition on land sales and markets in India except general land ownership rentals is needed and whether other policies allowing ceilings, the government maintains laws that restrict tribals to utilize land more productively may be more sale of land from tribals to non-tribals as a means to appropriate. Regarding the first issue, international eliminate as much as possible the undesirable impact experience points towards a number of mechanisms, of landlessness. However, this policy is often not including a right of first refusal, the requirement of adhered to and procedures to ensure recovery ofland obtaining community permission for land transfers, that was sold in contravention of such restrictions are the need for a community decision to allow land sales cumbersome, costly, and largely ineffective. To assess in general, or an involvement of the community in whether such a restriction continues to be justified, land sale negotiations. These can achieve the same goal we use data on land transactions during the period but foreclose fewer options for tribals to benefit from 1982-99 to explore the functioning of land markets productivity-enhancing transfers. As to the second in general, and the extent of tribal land market issue, there is little doubt that establishing and participation in particular. documenting tribals' property rights to the land they A number ofinteresting findings emerge. First, cultivate will be more effective in providing them with even though the land sales markets provides less incentives to manage such land in a sustainable fashion opportunity for land acquisition than land rental rather than a prohibition on land sales. While large- markets and participation requires higher levels of scale initiatives to do so on land that is currently assets, markets constitute a more promising avenue classified as forest will have to await the passage of to land access for the landless than non-market the Tribal Land Rights Bill, there are many situations channels. Second, we find clear evidence of drought where, for a variety of reasons, tribal land rights on shocks increasing activity in land sales markets, revenue land are unclear from a legal perspective or supporting the notion that involuntary distress sales are not well documented in the applicable records. by the poor continue to be an important phenomenon Immediate action to redress this situation is critical despite legislation to the contrary. Third, consistent and should accompany any efforts to prevent with the notion that credit market imperfections are involuntary land alienation of tribals. at the root of involuntary land sales, availability of In fact, there are many examples from other safety net programmes (e.g. employment guarantee countries where emphasis on safety nets, together with schemes, EGS) and access to banks significantly reduce providing an overriding right to the community, was the propensity to sell, especially during droughts a much superior approach to avoiding undesirable when greater inequality contributes to an increase land alienation than prohibitions which cannot in land transactions. Presence of banks helps activate prevent those in dire need from selling their land, land sales markets in general but reduces the tendency but reduce even further what they can get for it. By for land sales to occur in drought years. Finally, re-orienting energies from prohibition of alienation scheduled castes (SCs) and scheduled tribes (STs) are towards regularization of land occupation in tribal less likely to sell or purchase land even after household areas as well as systematic resolution of conflicts from and village characteristics are controlled for. past sales according to well-established criteria, it is While this suggests that restricting land sales by possible to make a more significant contribution to tribals may have had some effect, it leaves open the improving the livelihood of the poorest groups. At ECUTIVE SUMMARY the same time it will help reduce the potential for neutralize the impact of adverse policies that push the land-related conflict and extremist movements whose concerned households into informal systems. Similarly, main rallying cry is land-related injustices. There are land administration provides important tools to many international examples that can be drawn upon implement policies. In fact, one reason why India has and doing so would be fully consistent with the no shortage of bold land policy initiatives for the poor Panchayat Extension to Scheduled Areas Act (PESA) which look very attractive on paper but often cannot and the spirit underlying the tribal land rights bill. be implemented in practice, is the fact that its land Launching a major effort towards implementation administration system is weak in general and often once this legislation is passed will provide an non-existent or dysfunctional in the areas where the opportunity to put land administration truly into poorest live. This implies that there are a number of the service of the poor. priority actions required in administration and policy. Expand computerization, integration, and use of PRIORITY ACTIONS AND NEXT STEPS textual records to ensure full coverage. Even though states that successfully computerized textual records Although land administration and policy in India are benefited significantly from doing so, progress remains complex subjects with a bewildering amount of detail slow in many others. Go I initiatives to accelerate this and variation across states, a number of key messages effort can be improved by: (i) clarifying the overall and priority actions emerge. First, land administration policy and using it to establish clear criteria and is an integrated system that is only as strong at the accountability mechanisms for allocation of central weakest link in the chain. Trying to compartmentalize funds; (ii) publicizing best practice approaches on the it into different 'boxes' that are unconnected to each technical and legal or regulatory side through different other (either registry or records or survey or rural- channels (including websites) as well as promoting urban) will not help to improve the overall performance exchange and communication among technical staff of the system. In this sense, improvement of textual across states; and (iii) putting greater emphasis on records has made a very promising start but has solved complete functional integration between records and only half of the problem. At the same time, improving registry together with a systematic process to establish spatial records is much more complex and trying to consistency and eliminate any ambiguities on the legal fix spatial records in an environment where issues status of specific land parcels. relating to textual records are not sorted out may Provide a basis for statewide spatial coverage. Large be risky. The MoRD's efforts to enhance the land amounts of money have been, and continue to be administration system, and the effectiveness of Go I spent on surveying pilots with ill-defined objectives, support to such efforts, can be improved, if a clear the results of which are rarely subjected to rigorous policy to lay out these issues is made available and is evaluation. On the contrary, pilots often make officials used to guide allocation offunds to, and reporting on lose sight of the forest for the trees. Drawing on the use of these funds from, individual states. Second, India's capabilities to combine satellite imagery with policymakers underestimate the intimate links between existing village maps and other readily available spatial land administration and policy, in both directions, products to generate a basic cadastral index map would at their peril. Although land administration is highly be a lower cost option to provide a comprehensive technical, no amount of technical sophistication will framework, identify gaps, and on this basis establish EXECUTIVE SUMMAl1 criteria to address spatial data problems in an affordable that the high rates of stamp duty currently assessed manner using a range of techniques including ETS, on registration of land transfers push people into Differential Global Positioning System (DGPS), and informality on the one hand, while reducing other less accurate but less costly methodology. government revenue on the other. Reducing these Pilot ways ofimproving textual and spatial records rates, which are very high by international standards, for well-defined situations. In line with the ultimate is necessary to ensure the sustainability of any goal of the land administration system, the purpose improvements made in land administration. To make of piloting should be to establish processes that can such a step revenue-neutral, it may be useful to be scaled up rapidly to improve the overall (that is, combine it with an increase in the land tax for specific textual and spatial) record and formulate regulations groups, possibly sharing proceeds between states and that can help to do so, possibly by subcontracting to local governments. While such a decision will not the private sector. To achieve this, pilots should be be easy politically, it is likely to have a more profound targeted to archetypical situations arising from the impact on India's land administration system than a nature of India's land records (that is, unrecorded headlong transition towards a title registration system. subdivisions; inconsistencies across records, decay/loss Eliminate restrictions on land markets. All over the of maps, change ofland use patterns, unsettled lands). world, land rental markets allow rural dwellers to It would be ideal to set up a technical working group join the rural non-farm economy in a manner that to steer this process with the goal of producing results, provides those who stay back with access to additional to feed into a broad debate on this topic, in a 12-18 productive resources. Indian evidence shows that month time frame. rental restrictions reduce equity as well as efficiency. Allow private sector participation in surveying, It will thus be desirable to: (i) make leasing legal where focusing government on a regulatory role. Given the size it is currently prohibited and replace rent ceilings of the gaps in spatial data and the limitations that with regulations to facilitate rental markets instead make it difficult for the public sector to address them of constraining them; (ii) allow transferability ofland comprehensively, the almost complete prohibition by land reform beneficiaries at least through lease of private participation in survey is surprising and and also explore options to make the gains from such inconsistent with international best practice and reform permanent; (iii) drop restrictions on sale of India's own experience in computerizing textual records. land to non-agriculturalists and on subdivision which Efforts to change this should focus on: (i) providing have little economic justification; and (iv) review a regulatory framework for application of a range legislation on compulsory land acquisition and, subject of survey methods with specified levels of precision; to the prevention of undesirable externalities, allow (ii) strengthening capacity in the private and the farmers or their representatives to negotiate with and, public sectors through a range of strategies including if desired, transfer land directly to investors rather than enhancing of capability and resources in the academic having to go through the government and often receive sector; and (iii) revamping survey processes, for example, only very limited compensation. shifting from paper-based to electronic ones to reduce Complement restrictions on tribal alienation with costs and make the surveys more affordable. flexible mechanisms providing them with property rights. Reduce stamp duty rates and explore the scope for While there is little doubt that alienation of lands replacing these with a land tax. There is little doubt through distress sales is an extremely undesirable ··~ outcome that should be avoided, increasing rates of rights, either individually or as a group would, in the tribal landlessness suggest that regulations are often not long term, make a more important contribution to effective in preventing it. In the short term, the most their productive development and thus help avoid promising way to reduce tribal land alienation is likely distress sales. Therefore, the longer-term goal should to be effective safety nets, something that could possibly be to implement systematic programmes that recognize be combined with mechanisms for communities to tribal land rights and resolve whatever conflicts have have a greater say in whether or not land should be arisen from past alienations in contravention of the transferable, such as a right offirst refusal or community law according to accepted principles of policy. consent for sales. Providing tribals with real property CHAPTER Introduction In India, land continues to be of enormous economic, growing evidence that restricting land rental may do social, and symbolic relevance. The way in which little to help the poor, many observers have lost land can be accessed and its ownership documented confidence in the ability of land institutions to is at the core of the livelihood of the large majority contribute to the welfare of the poor or the potential of the poor, especially in rural and tribal areas and for improving the performance ofland administration. determines the extent to which increasingly scarce In this chapter we first show that land administration natural resources are managed. Land policies and in India does indeed have shortcomings but also administration are critical determinants of the use data from India to show that addressing the transaction cost associated with modalities to access shortcomings of the land administration system is land for productive, residential, and business use and, necessary. We then highlight some of the recent success through the ease of using land as collateral for credit, stories to argue that doing so is entirely feasible but the development of the financial sector. Land is also a only if, in addition to focusing on technical aspects, major source of government revenue and a key element a number of policy issues are addressed as well. for implementing government programmes. This implies that land policies and institutions will have WHY ADDRESSING LAND ISSUES a far-reaching impact on the ability to sustain India's IS IMPORTANT current high rate of growth, the extent to which such growth reaches the poor, and the level and spatial Land is of critical importance for the poorest and most distribution of economic activity. marginalized sections of society whose main livelihood At the same time, the policies put in place by continues to depend on agriculture and who may, for different states and the institutions tasked to implement a variety of reasons outside their control, have very them often fail to live up to the importance of the insecure or non-existent land rights. This makes it issue. In fact, land administration institutions seem difficult for them to fully utilize their productive to impose high costs without generating commensurate potential, thereby reducing their welfare and their benefits and are generally perceived as corrupt, ability to accumulate assets. Even if land has been mismanaged, and lacking transparency. With land surveyed, the complexity and high cost of services by reform policies having largely run their course, and the land administration system may prompt potential D I A- LA N D P 0 L I C I E S F 0 R G R 0 W T H A N D P 0 V E R T Y R E D U C T I Q. users to opt for informal arrangements rather than preventing those cultivating them from obtaining use the official system. In fact, various observers have ownership (Upadhyay 2006). recently drawn attention to the fact that an ill- The challenge ofinformality. Complex and costly functioning system ofland administration, together procedures of land administration together with with a second-best policy regime, that is often both legislation that completely prohibits land leasing or a cause and a consequence of the latter, can also imposes strict conditions on it often have the result of seriously reduce overall economic growth. The main driving the poor into informal arrangements. This not focus of the analysis in this report is on the rural only deprives them of formal recognition but also economy. However, we report some studies to makes it difficult for them to access formal credit and highlight the broader importance of the topic and other government programmes. For example, in Tamil to emphasize that the distinction between rural and Nadu efforts to provide targeted payments to producers urban is to an extent artificial and that, the need in compensation for increases in electricity rates today is for a national land administration system, remained ineffective as producers found that it was based on a legal and institutional framework that is too cumbersome to update their land records which sufficiently flexible to accommodate the needs of both were intended to provide the basis for the scheme. rural and urban areas. Non-agricultural land prices in India are among Lack ofland rights. Given the way in which land the highest in the world. Data suggest that rights were settled by the British, numerous households inefficiencies of the land administration system, in marginal areas remain without any rights to land together with the reduction in supply due to land that they have owned for very long periods of time. policies and land use regulations and the fact that This not only reduces their incentive to invest in the large tracts of land are under litigation drives land land and manage it sustainably but also their ability prices up. All these factors contribute to the fact that, to obtain credit for doing so. In addition to reducing in relative terms, land in India is amongst the most productivity, investment, and welfare, it also implies expensive in the world. As Figure 1.1 illustrates, land that, in case the land is needed for other purposes, in New Delhi is more than ten times as expensive as they may not be legally entitled to any compensation. that in Tokyo and 50 times more expensive than that The areas involved are often very large; for example, in Kuala Lumpur. Within the region, this is likely to many tribals in southern and western Orissa are translate into a significant comparative disadvantage cultivating land on the hill slopes without any title for Indian business. Difficulty in gaining access to because the survey equipment used at the time was land, as recently reported in the press for a number unsuitable for land above a gradient of 10 degrees of Indian cities, will also limit the country's ability which was categorized as state-owned uncultivable to attract new businesses. wasteland, despite the fact that much of it is possessed Transferring land is very costly. Although land and cultivated by tribal households (Saxena 2005). prices are affected by a myriad of factors, India does In fact, 7 4 per cent ofland in scheduled areas of Orissa not fare better if only the cost of transferring land- is categorized as state land, with 26 per cent and 48 which is more directly related to the land per cent being revenue and forest land, respectively administration system-is considered. According to (Kumar et al 2006). Similarly, in Madhya Pradesh, the World Bank's 2004 'Doing Business' study, the the status of a total of 12,395 sq km or more than 3 cost of registering land transfers in India, measured million acres, the so-called 'orange areas', is disputed, as a share of property value, was among the highest 115 Kuala Sydney Bangkok Tokyo Singapore Jakarta Seoul Taipei Bangalore New Mumbai Lumpur Delhi FIGURE 1.1: Ratio ofland cost/sq m to per capita GDP Source: Colliers Jardine Asia and Pacific Property Trends (1999). in the world; India ranks 123 out of a total of 140 processes, together with high land values, provide countries included in the data (World Bank 2004). 1 ample opportunities for corruption. Indeed, an This is due to three elements, namely (i) high independent study confirms that land administration stamp duties; (ii) the need to comply with complex fails to command the respect of citizens; in fact it is regulations; and (iii) the time and money spent on considered the least transparent and second most procedures that are often duplicative or inefficient. corrupt public service in the country, in the same The fact that the first two are largely a result of policy group with the police and the lower judiciary. Of rather than the efficiency of the land administration the households in the study interacting with land system highlights that, in addition to improving land records or registration departments, 48 per cent had administration, policy reform will be a critical to pay a bribe, with the total amount of bribes paid element of any effort to sustainably reduce the cost each year estimated at Rs 3126 crore (Transparency of transacting property. In fact, given the magnitude International India 2005). Almost two-thirds (62 per of stamp duties, increasing the efficiency of land cent and 63 per cent, respectively) of respondents felt administration institutions alone may be insufficient that corruption had increased over time and that to ensure that transactions are brought on record. there was no commitment by government to reduce Corruption in land administration is pervasive. corruption in land administration. Figure 1.2 uses state Overlapping institutional mandates and ill-defined level figures to illustrate that the share of households that had to pay a bribe when interacting with land 1 Given the recent reduction in stamp duties effected in administration institutions was below 50 per cent in Maharashrra and the fact that the World Bank's indicator for only six out of the 20 states included in the sample, the whole country is based only on the major commercial city (in this case Mumbai), this indicator decreased somewhat after Kerala (29 per cent), Gujarat and Maharashtra (39 2004. per cent), and Himachal Pradesh, Punjab, and DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCT! 100 90 80 70 60 50 40 30 20 10 0 AP AS Bl CG GJ HY HP JK JD KA KE MP MA OR PJ RJ TN UP WB FIGURE 1.2: Corruption in land administration across different Indian states Note: The first column indicates the share of respondents who considered land administration institutions as corrupt while the second column represents those who thought corruption in land administration had increased compared to the past. Source: Transparency International India (2005). Chhattisgarh (46--48 per cent). More significantly, between different parts of the system, significantly in all the states with the exception of Tamil Nadu increase the likelihood of land-related conflict. and Kerala-where only 37 per cent and 33 per cent Estimates suggest that of the worryingly high number perceived that corruption was on the increase-more of pending civil cases-20 million in lower courts than half of respondents felt corruption to be on the and 3 million in state high courts (Debroy 2000)- increase rather than stable or on the decline. 2 almost 40 per cent pertain to land. A small pilot study Land-related conflicts tie up huge amounts of from non-tribal areas in Andhra Pradesh points resources. In addition to the cost ofland administration, towards modest levels ofland-related conflict in rural ambiguities in the system, in particular a weak or areas (where 2 per cent of plots were affected), as non-existent spatial framework and inconsistencies compared to very high incidence-affecting 28 per cent of all plots-in peri-urban environments. To the 2 All of the information is based on a sample of 14,405 households extent that these figures can be taken as representative, from 20 states (Transparency International India 2005). In resolving land cases speedily would not only allow interpreting these figures, it is important to note that the sample courts and those currently involved in litigation to is heavily biased in favour of the urban sector. The lack of spend their time on more important and productive information on the sampling procedures in the report raises doubts matters but also make available a huge amount of as to whether proper weights have been applied or whether indeed the households interviewed constitute a representative sample of high-value, peri-urban land for productive investment the population, implying that the figures should be treated as and development. High incidence of conflict is also indicative only. Emails to request clarification and/or access to likely to be found in tribal areas due to inefficiences the micro-data from the organizers remained without a response. related to the prohibition ofland alienation by tribals. INTRODUCTJO WHY THERE IS SCOPE TO ADDRESS TABLE 1.1: Number ofland parcels in different countries LAND ISSUES Country Total area Parcels ('000 sq km) (million) Even though the magnitude and complexity of Australia 7600 11 improving the land administration system is huge, a India 3287 500 number of states have achieved remarkable success Argentina 2792 13 in ventures-which could often be implemented as Indonesia 1919 85 quickly as they did only because of a strong partnership Iran 1648 55 with the private sector-to computerize land records South Africa 1220 19 and registries. This is relevant in three respects. First, Japan 380 200 Germany 357 76 it demonstrates beyond any doubt that India has the Malaysia 330 7 capacity to rapidly improve its land administration Philippines 300 50 system. Second, the models developed can be readily New Zealand 266 2 adopted at low cost by other states which are eager South Korea 222 42 not to be left behind, thus setting in motion a healthy Sri Lanka 66 9 Source: FIG Commission 7 (2006). process of competition. Finally, and most importantly, even though what has been accomplished is at best only a first step towards the goal of a more effective together. When discussing the scope for improvement, land administration system, it serves as a springboard one should not forget that establishing and for further innovation that will help develop a long- maintaining a system to describe, map and track term vision for the sector, especially in an environment transactions involving such a large number of objects where independent efforts to provide the necessary in a way that is accessible at the village level is by infrastructure create favourable conditions for itself a huge accomplishment that distinguishes India achieving further integration. from many other developing countries. The size of the challenge. To appreciate the Initial successes of computerizing textual records. magnitudes involved, note that the number of land Building on India's strengths in IT, a number of plots maintained in the Indian system is very frontrunner states have achieved considerable success conservatively estimated at more than 250 million in improving the quality of maintenance of land- (FIG Commission 7 2006) and may actually be closer related textual information, reducing the cost of to 500 million. This puts India in a class of its own managing it, and making it easier for users to access especially if one considers that the number of land it. Although this has not always been sufficient to parcels in other geographically large countries is ensure that the information on land maintained by surprisingly limited (Table 1.1). For example, in different government departments is consistent, up- Australia, the only country in the sample that is larger to-date, and of comprehensive coverage, studies than India, the number of land parcels is only 11 suggest that it did have an impact on reducing million, Argentina has only 13.5 million, and South transaction cost and improving transparency and Africa 19 million. To put this into perspective, note governance within the existing system. In addition, it that the number of land parcels in Andhra Pradesh has resulted in the development of technical solutions, alone approaches that of these three countries put in particular software packages, that can be readily INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIO adopted by states wishing to make the transition as well as the problems with spatial data-which are towards a computerized system in order to realize generally much worse than those related to textual similar benefits. records-into much sharper relief Finally, the process A wave ofadaptation. Realization of the advantage has clearly highlighted that, even with improved of a computerized land administration system and the textual records, there is a long way to go before the danger of being left behind, together with the basic goal of having a conclusive record of land availability of Goi funding and the fact that the ownership widely available at an affordable cost can required processes are now well defined and key be attained. Achieving this requires integration of technical hurdles to be overcome have been identified, the various systems and overcoming the current has allowed a large number of states to adapt the fragmented nature of land administration institutions processes developed by the early adopters to their own (for example, between rural and urban land records needs and often make considerable progress towards or between land records, registration, and survey) an improved system of managing textual land records. which has often little substantive justification. While Given that proven processes for doing so are readily far-reaching institutional changes are likely to be available, it appears that the main challenge to be highly contentious politically and may thus not be overcome is no longer of a technical but rather of a feasible at this point, a number of states have started political nature. Especially in view of the fact that to move towards 'virtual' integration of processes, funding is not a key constraint, the lack of progress functions, and information across departments as a in some states is either due to lack of information on second best option. Initiatives, such as the national best practices or to political resistance by specific e-govemance project that aims to establish connectivity, groups who may be concerned about possible adverse will provide the basis for doing so. Given the current impacts on their welfare. The key issue is to codify plans for roll-our of the necessary infrastructure, and disseminate best practices and take measures that broad connectivity is likely to be more widely available ensure that innovations are implemented quickly. in a few years and it will be important to utilize the The difficulties associated with moving towards a intervening period to establish processes and make comprehensive vision. Although it facilitated some the regulatory changes needed to implement them remarkable successes and led to the re-definition of once they become available. processes in a number of cases, computerization of In addition to the lessons noted above, the successful textual records also highlighted the shortcomings of computerization of textual records has also helped to just automating existing systems that may have been demonstrate the intimate link between policy issues appropriate in colonial times bur have not seen any and land administration in two ways. On the one major change since then. In doing so, three elements hand, a number of policies, in particular high stamp have been of particular importance. First, there is scope duties and land reform and land use regulations that for review and often simplification of processes, for tend to drive users into informality have been shown example, by separating out some of the information to pose a threat to the sustainability of efforts to on agricultural production that has little relevance modernize the land administration system. On the for ownership buchad been included in original other hand, policymakers may not be aware of the records due to their function as a basis for taxation. potential benefits of having in place a comprehensive Second, successful computerization of textual records and well-functioning system of land administration puts both the shortcomings of existing land records not only for overall economic activity but also to INTRODUCTION implement policies and programme in a more cost- approach to improving textual data (records and effective and often also efficient manner. registration) and, based on a review of states' experience, India's remarkable progress in improving the land identifies the associated benefits. In both cases, the administration system is unlikely to be sustainable, evidence is used to identifY unresolved issues that and prospects for realizing the vision of an integrated should be addressed by pilots and to identifY criteria and effective institution will not be achieved, unless a and indicators of performance that could help make number of key policy issues that have emerged as major centrally sponsored schemes more effective. Chapter constraints are addressed. First, levels of stamp duty 4 reviews the extent to which lessons from improving that are extraordinarily high by international standards textual records could help to give a boost to improve- will continue to drive transactions underground and ment of the spatial database for land administration, thus undermine the completeness of the registers. an area that has thus far been largely neglected (not Second, an outdated regulatory framework that surprising in view of the cost and complexity of the leaves little space for private-sector participation and issue) but that will require urgent attention if initia- precludes the use oflow-cost modern technology for tives to have a system providing higher levels of tenure surveying, makes it impossible to generate spatial security are ever to bear fruit. Chapter 5 concludes information at a cost that will allow land administration the discussion on land administration by assessing to reach out to the poor and marginalized (whose land the scope for title registration to help improve tenure rights have long been neglected). Finally, legislation, security in India and by identifYing pertinent issues. most of which is associated with past land reform The key contribution of our discussion of land effons, seriously limits the incentives for individuals policy is that, for each of the policy issues of land to truthfully declare their land holdings (instead of reform, land leasing, and land sales markets, we pro- say, registering land in the name of non-existent vide household-level evidence, based on a nationally persons) and ways to deal with such situations. representative survey of a panel of about 5000 Modifications to the legislation may be required to households (or their descendants) who were inter- modernize the land administration system in a way viewed in 1982 and then again in 1999. Chapter 6 that is sustainable and allows it to realize its potential. highlights the fact that land reform has helped increase accumulation of physical and human capital PLAN OF THE REPORT but that the impact is declining over time. It, therefore, explores a range of alternative ways to The main purpose of this report is to present new increase land ownership by the poor. Chapter 7 empirical evidence on land administration and land explores the functioning of land lease markets, and policy, as well as the possible interaction between the the extent to which restrictions on land leasing reduce two, to derive policy conclusions. The empirical basis the scope for productivity. It shows that transfers in for the discussion ofland administration is provided such markets enhance equity and could be particu- by a review of land records, survey and settlement larly beneficial for women. Chapter 8 reviews records, and land registration in 14 states. Chapter the operation of land sales markets and suggests 2 describes the origin, nature, and main functions alternative approaches for preventing land loss by of current institutions and the ensuing problems tribal people and Chapter 9 concludes with a series for secure tenure and the easy transferability ofland. of policy recommendations. Chapter 3 identifies elements of a 'best practice' CHAPTER Land Administration Institutional Structure and Key Challenges Land administration deals with the recording, institutions and processes for administering land in processing, and dissemination of information about India were adopted from the British at the time of ownership, value, and use of land and the resources independence and have been hardly modified since. associated with it. It includes the determination of Land administration in India was decisively shaped by property rights and other attributes ofland that relate the fact that the British turned earlier mechanisms of to its value and use, the survey and description of these, land management into a tool for revenue collection. their detailed documentation, and the provision of Most of the institutions and processes to administer relevant information in support of land markets land today still closely resemble those established (UNECE 1996). A land administration system that under British rule when land revenue was the main is affordable, accessible, and efficient will help to: source of government income, amounting to 60 per (i) provide secure land tenure for landowners, thus cent of total British government revenue in the 1840s. encouraging them to invest and manage this asset in The amount of taxes (land revenue) to be paid was a sustainable manner; (ii) facilitate low cost transfers determined through a process called settlement, the that allow land to move from less to more productive nature of which differs for different parts of the producers through rental or sale; (iii) provide a basis country, depending on the time and means of their for the use ofland as a collateral for effective operation incorporation into the empire (Box 2.1). of financial markets; (iv) allow spatial planning for This historical process had far-reaching implications issues ranging from provision of infrastructure such not only on subsequent development through the as roads and utility lines to environmental protection; security of property rights, the nature of political and (v) enable central and local governments to power, and the overall distribution of wealth, but it effectively implement programmes and collect revenue. also affected the nature of land administration institutions and the quality of the records supporting HISTORICAL EVOLUTION AND these institutions, that continue to the present day (see OVERALL STRUCTURE OF LAND Box 6.1). For example, cadastral maps in zamindari RECORDS areas-where revenue was fixed permanently and maintaining records therefore did not offer benefits One of the reasons why land administration in India in terms of higher tax revenue-are generally less well often fails to live up to its goal is that most of the maintained than those in ryotwari areas. LAND ADMINISTRATION BOX 2.1: DIFFERENT TYPES OF SETTLEMENT Depending largely on the political preferences in place at the time when a specific area was brought under colonial administration, the British adopted one of three land revenue systems: landlord- based systems (zamindari or malguzar~, cultivator-based systems (ryotwar~ or village-based systems (mahalwari). Zamindari systems were established in Bengal, Bihar, Orissa, many parts of Uttar Pradesh, the Central Provinces (Madhya PradeshL and some parts of Madras Presidency (Tamil Nadu and Andhra PradeshL generally areas which have good quality soils and adequate rainfall. It effectively provided property rights in land to the landlord (zamindar) who in turn had to deliver a certain amount of revenue to the colonial administration while being allowed to keep any surplus for himself. In some areas, the level of revenue to be delivered was fixed in perpetuity under the 1793 'permanent settlement' while in others temporary settlements were revised after a number of years .. Also, while in most areas land records remained with the government, they were maintained directly by zamindars in Bihar. · Under ryotwari, which was adopted in most areas of the Madras and Bombay Presidencies and in Assam which generally had lower inherent agricultural potentiat revenue settlement was made directly with the cultivator (ryo~. To provide the basis for revenue collection, a cadastral survey, to be updated regularly through revisional surveys, was carried out and a detailed title document, called record of rights, was prepared. Tax liabilities were then calculated as the monetary value of a share of the estimated average annual output from the land. Under the village-based (mahalwari) system, which was adopted in the North-West Provinces and Punjab, village bodies were responsible for the land revenue. Depending on the nature of the village body, the final result could be similar to a zamindari system (if there was one big landlord) or a ryotwari-type arrangement with individual cultivators. The primacy of revenue affected the coverage, by local people for generations, remain de jure under nature, and evolution of the land administration state ownership with the result that even long-standing system in three respects. First, to the extent that the occupants may legally be 'encroachers' who may be main purpose of the system was to enable the state evicted or will not be eligible for compensation in to collect revenue, it focused on use rather than case the land is needed for other purposes. ownership, and incentives for owners to keep records The basic structure ofland administration in any up to date remained limited. Second, there was limited Indian state comprises four main institutions (see attention to the spatial framework; and maps were Figure 2.1). The land revenue department maintains prepared only with reference to the village and were the textual database for land records as well as tax infrequently updated. Finally, coverage was eclectic registers where they exist and continues to collect and land administration focused on agricultural lands land revenue but has historically taken on a host of only, thus excluding urban as well as marginal lands. administrative functions. The survey and settlement For the latter, there remain large tracts of 'unsettled' department is responsible for maintaining spatial data, land which, although they may have been occupied mapping and demarcating boundaries, and executing INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION collector. Below the districts are blocks (taluk, tehsil, manda~, often subdivided into revenue circles which in turn are composed of a number of revenue villages • FMB!Tippan • ROR Register • Register of • Town survey • Village Map • Adangal/ Transactions maps and register (see Figure 2.2 for the revenue structure in Punjab). • Shetwar/RSR Pahani • Layout plans • Khata Register • Property tax Within each revenue village, an official (called patwari, register talathi, village accountant, lekhpal, kanungo, etc.) is FIGURE 2.1: Agencies involved in land administration responsible for the maintenance of textual land records in India through annual reconciliations (jamabandi), the recording in the record of rights (RoR) of any mutations surveys for subdivisions on demand. In some states (transfers of ownership and occupancy rights) that the survey and settlement department undertakes full may have taken place, and issuance of certified copies settlement surveys of specific areas within the state of the RoR for use by cultivators as a proof of their while in some it also establishes and maintains a city rights. Key documents, the originals of which are kept survey system. The department ofstamps and registration at the village level with copies at the block (taluk) or is responsible for registering deeds and for collecting district level, are the following. the stamp duty due on these transactions. Finally, as The RoR is equivalent to a title to the land. It land records are limited to agricultural areas and city records the nature and extent of the respective rights survey coverage is limited, local bodies, for example, and interests of all persons, including their credits and municipal corporations or panchayats, maintain liabilities and in some states encumbrances such as property tax registers and sometimes maps and layout mortgages existing on the land. The RoR constitutes plans for areas that have not been covered by original prima facie proof of ownership and all entries in the surveys. While details and names vary across states, RoR are presumed to be true until proven to the a discussion of the general principles can help contrary. It is thus an important document to prove identify some of the key problem areas that need land ownership that is used routinely by farmers to to be confronted. obtain loans from banks or to take out crop insurance. An authorized copy of the RoR can be obtained from REVENUE DEPARTMENT the village accountant or, where these records have been computerized, from the local kiosk. 1 The key institution for managing land records is the The adangal!pahani contains crop details that department of revenue which has traditionally were originally collected on a seasonal basis to compute constituted the main interface between the state and land taxes. In modern times, the information contained the local population in rural areas. Its importance is in this record is expected to be transferred to the illustrated by the fact that at the district and lower agricultural department to make supply projections levels, revenue officials also assume judicial functions. and plan marketing campaigns. The frequency and accuracy with which this information is collected at Institutional structure and types of records maintained 1 A number of states such as Andhra Pradesh have introduced The revenue department continues to maintain an some form of'passbook' that would record all the plots held by enormous field presence dating back to British times. a household. This has generally not been very successful but Functions at the district level are overseen by a may have contributed to RoRs falling into relative disuse. 0 LAND ADMINISTRATI Financial Commissioner Revenue - - - - - - - - - - - - - - - - - - Director Land Records .-------------------+--------------------. r - - - - - - - - , Divisional Commissioner Divisional Commissioner ~ - - - - - - - - - ~ -----------------------~ r--------, L. Deputy Commissioner _ _ _ _ _ _ _ _ .J ;--------, District Revenue Officer - - - - - - - - -, L--------...1 r-- --, Tehsildar/ 1 1 Naib-Tehsildar L-----..1 I Sadar Kanungo I FIGURE 2.2: Revenue hierarchy in Punjab the local level differs widely across states and the the maintenance ofland records that had once been channels through which it is transmitted to higher its core function. levels, as well as how it is used are not entirely clear. Reduction of revenue and institutional overload. The khata is a tax record showing land revenue, cess, After independence, the share ofland revenue in state water rates, and other government dues to be paid resource inflows declined drastically, from more than by a cultivator and the amount paid in a given year. 30 per cent in the late 1950s to less than 2 per cent in The khata register is prepared once a year during the the late 1980s, as illustrated in Table 2.1. This drop annual reconciliation of accounts by the village official. was particularly pronounced in states such as Bihar, Madhya Pradesh, Orissa, Rajasthan, and Uttar Pradesh where, as a result of rather limited non-agricultural Key challenges development in the earlier period, land revenue in With a gradual decline in the importance of land the 1950s had still been at 30 per cent or above. As a revenue as a source ofgovernment income, the revenue consequence, the revenue department has shifted from department has increasingly been settled with other being a net contributor to the state's budget to being mandates that are unrelated to its original focus on supported from that budget, with the amount of land administration. This made it difficult for officials operational expenses exceeding land revenue collected to keep up the quality of service provision and was by a factor of 5 to 10 in virtually all states. The impact often associated with a significant deterioration in of this reduction in revenue was compounded by the ability of the revenue department to focus on three other factors, namely: (i) an increase in the 11 .INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION TABLE 2.1: Land revenue as a percentage oflndian (iii) placement of the revenue department in the non- states' own revenue 2 Plan category. This implied that, until the launch of State Time period centrally sponsored schemes to modernize land 1957-8 1970-1 1989-90 records in the late 1980s, states got neither funding Andhra Pradesh 28 17 1.1 nor technical advice on land matters from the central Assam 15 14 1.3 government. Bihar 30 11 2.2 Discretionary power of local officials. Conditions Gujarat 20 46 1.3 where demand for public services outstrips supply Haryana 16 4 0.1 Jammu and Kashmir 6 10 0.8 by a wide margin provide a fertile ground for rent Karnataka 23 8 0.8 seeking by officials. Indeed, village officials have been Kerala 8 3 1.6 reported to charge high fees (speed money) for routine Madhya Pradesh 43 9 1.4 services such as the provision of certified copies of Orissa 33 9.6 6.8 the RoR to farmers or conducting mutations in a Punjab 16 33 0.3 Rajasthan 40 16 2.7 timely manner. These fees drive transactions further Tamil Nadu 19 3 0.9 into informality, in addition to marginalizing poor Uttar Pradesh 40 12 1.4 landholders for whom the direct and indirect costs Total 30.03 13.52 1.67 of navigating their way around a complex and often Source: Saxena (2005). non-transparent system may well be too high. The impact of this is exacerbated by the fact that land administrative and judicial duties to be performed by records are rarely in the public domain, that revenue the revenue department; 3 (ii) initiation of new land- codes and procedures are complex to comprehend and related programmes such as distribution of ceiling that, in a context of local power structures that are surplus land, tenancy legislation, and identification of not at all in favour of the poor, even ensuring house-sites to the poor, all of which diverted attention enforcement of court orders becomes a non-trivial from vital functions of maintaining land records; and issue. It is thus not surprising to find reports of!ocal officials using their privileged access for tampering 2 To compute the total, each state's percentage was weighted by with records and extracting bribes. its 2001 population share. Also, note that in Orissa, the share ofland in total revenue had declined to 2.4 per cent by the end of the 1990s. SURVEY AND SETTLEMENT 3 The duties of revenue officials include the conduct of general Even though it originally formed an independent elections, issuance of certificates, implementation and department, the survey and settlement department monitoring development schemes, and provision of relief in in many states is now integrated with the revenue natural calamities. A study in Andhra Pradesh showed that revenue officials spent 32 per cent of their time on department, at least at the higher levels, as is the administration of welfare programmes, 25 per cent on judicial and magisterial functions, 25 per cent on developmental quality of the task performed. More importantly, as this shift activities and implementation of schemes, 10 per cent on general reduced the emphasis on land records in the curriculum for administration, and only 8 per cent on land administration public servants, many may no longer be familiar with the details (Agrawal 2006). As the way in which land records were to be of the associated documents, something that would, over time, maintained remains unchanged, this must have affected the lead to further decline in land record maintenance. 12 LAND ADMINISTRATION i~- .------------------------------------------------, STATE LEVEL I Commissione rof Spl Commissioner of Spl Commissioner of Commissioner of Land Reform s Revenue Survey and Land Administration Administration Settlement ~~~~ ~~~~ t ,- ----- j _____ I District Collector J:I I + Regi onal Deputy* District Revenue Director of Survey I Officer (ORO) J! (4) DISTRICT LEVEL ----1----~ I Asst Director of I I Revenue Division I I Survey I I I I r---- -----, Officer (ROO) DIVISION LEVEL I I I I + I I I I Manager Technical I I ----t----' I I I Inspector of Survey I I I I I I I I Tahsildar I I DISTRICT LEVEL I I I + I I I L---- ---- .J I Head Draughtsman I I I I I + I I ~ I I Deputy Tahsildar I DISTRIC LEVEL I I .----- ----.., L---- ____ .J I I I I Deputy Inspector of ____ f ____ TALUK LEVEL ---- J -----, I I I Survey I .----- ----.., Revenue I + I I .----- I ---- I Inspector (RI) Sub Inspector of I Senior Draughtsman I Town Surveyor I I I I I Survey I FIRKALEVEL I I I I 1 I I + I I + I I I TALUK LEVEL L--------- Village Administrative I Firka Surveyor I Draughtsman I I I I Officer (VAO) I I I REVENUE TALUK LEVEL TALUK LEVEL VILLAGE LEVEL FIGURE 2.3: Structure of the survey and settlement department in Tamil Nadu case in Tamil Nadu (see Figure 2.3 for the structure). Types and origin of spatial records In many cases, however, integration is more limited To provide a spatial record of land ownership, the lower down the hierarchy and there are large colonial administration carried out ground surveys variations across states in the extent to which with basic technology (plane table and crosstab) on a municipal bodies are integrated into the structure of village by village basis, and without link to a national the department. coordinate system. Transfers are reconciled in the 1 lA-LAND POLICIES FOR GROWTH AND POVERTY REDUCT! .. textual records on an annual basis by the local task is typically undertaken by the survey and administrator. However, the main mechanism to settlement department. Where this has been done, update cadastral maps was through periodic re-surveys details of land ownership are maintained in a that were to be undertaken once in a generation, simplified document known as a property card. If that is, every 30 years. By then, population growth city surveys do not exist, property tax receipts kept and inheritances, construction of infrastructure (for by the local authorities assume an important example, irrigation), and possible expansion of function. Although they are not meant to provide cultivation into adjoining public land implied that a legal evidence of ownership, local authorities have revisional survey, combined with re-assessment of put considerable efforts into the establishment of tax land revenue and generation of a new set of record maps which allocate a unique assessment number to of rights was probably cost-effective. The basic spatial each property. records are as follows. Field measurement books. In the process of the Key challenges original surveys by the British, parcel boundaries and The key challenge in the area of survey can be village limits were demarcated on the ground with summarized in terms of three issues, namely: (i) the monuments. Typically, for each parcel a field fact that, for a variety of reasons that include lack of measurement book (FMB or tippan) was prepared on resources and staff, revisional surveys that were a 1: 1000 or 1:2000 scale, recording the plot's relative intended to provide the basis for regular updating of coordinates and distance measurements in chains, the textual and spatial records in the original system, links, or other local units, thereby allowing computation have not been carried out as specified by law; (ii) the of its area (see Figure A2.1). Each parcel was assigned fact that the survey department was neither responsible a survey number unique within the village which nor had established a regulatory framework for allowed identification of parcel boundaries. surveys in urban areas-particularly cities-where Village maps. The information recorded in the circumstances required establishment or updating of field books is used to compile the village map, usually the map base, has led to the proliferation of a large at a somewhat smaller scale (1 :7920). This document number of map products that did not follow common contains boundaries of individual plots as per the standards; and (iii) lack of a systematic effort to original settlement or any revisional survey that may expand the textual and spatial database for land have taken place in the interim. As can be seen in administration into areas that had not been settled Figure A2.2, main physical features, the location of at the time of the original survey or to dose gaps that parcels relative to each other as well as prominent had arisen for various reasons. topographical marks, are also included in the village Physical condition oforiginal survey records. Given maps. In some states, village maps were compiled their age, the physical condition of the FMBs and directly using plane table techniques in the field and village maps is precarious, with many of them being these plane table records constitute the field records torn or already having completely disintegrated. of the settlement surveys. While overall figures are difficult to obtain, City surveys and tax records by local authorities. The investigations in Andhra Pradesh revealed that only Land Revenue Act stipulates that for urban areas about two-thirds of parcel maps are in good with a population of more than 5000 (in some condition while 21 per cent are completely missing. cases 2000), city survey maps may be prepared. This LAND ADMINISTRATl TABLE 2.2: Summary of status of rural maps in to current patterns of ownership and occupation may Andhra Pradesh be tenuous or non-existent. In rural areas, a key reason Village Plot maps for discrepancies lies in transfers and partitions- map Tip pans FMBs Total through inheritance or sale-that were either not Total Number (1000s) 28.4 3717 4767 8485 recorded or not accompanied by a re-survey. 5 Given Share of which are: the decline in the collection ofland revenue, carrying in good condition 62.19 49.39 78.89 65.96 out such re-surveys at the required speed and scale is brittle, faded, torn 8.92 19.22 8.35 13.11 miSSlllg 28.89 31.40 12.76 20.93 no longer an option. A5 a consequence, the spatial Source: Agrawal (2006). framework for land administration even in rural areas has fallen into a state of disrepair. For example, in Andhra Pradesh the average survey is overdue by A5 Table 2.2 shows, the level of preservation varies about 60 years (see Table A2.3), implying that in significantly; in areas of the state that were formerly the average district the last re-survey was carried out under zamindari rule, more than 30 per cent of plot around the turn of the last century and the situation maps (tippans) were missing and less than 50 per cent is quite similar in other states. in good condition whereas in former ryotwari areas of Incompleteness and unclear responsibilities in urban the state (FMB), 13 per cent were missing and 79 areas. Given its original agricultural focus, the spatial per cent in good condition. Also, identification of coverage of the land administration system has boundaries in the field is complicated by the fact that considerable gaps in both urban and rural areas. In a large share of the bol_!-ndary stones emplaced during rural areas, large stretches of land were not included the original survey is missing as well. 4 Pilot experiences in original surveys for various reasons even though also suggest that mosaicing FMBs together to obtain they may long have been under cultiv~tion. In urban a substitute for village maps will be near impossible. areas, many of the cities which would be obliged to Thus, unless alternative options (for example, use of do so have not conducted city surveys, or have failed satellite imagery) can be explored, reconstituting village to update the surveys and the system of property cards maps in the 30-40 per cent of cases where these are that is supposed to be based on them. 6 Moreover, there either missing or not usable will require a massive field is no obligation to have any record of land ownership, campaign. Given the cost involved, it is not surprising either spatial or textual, in habitations with populations that this has rarely been undertaken. ofless than 5000 which are home to the vast majority Lack ofupdating. Even in cases where original survey records are still available, intervening transactions 5 In fact, the British realized that owners would have few that were not surveyed imply that their relationship incentives to either register their transactions or carry out a re- survey, implying that neither is mandatory, for example, in case 4 A recent survey in Andhra Pradesh found that about half of the of inheritance. This system works well as long as the cycle of survey stones (900,000 out of a total of 1.8 million) that were revisional surveys is maintained but breaks down without it. planted during the initial survey are now missing. With the cost of 6 In Bangalore, for example, the city survey only covers about stones at about Rs 100 apiece, the total cost of the stones needed 118 sq km. This leaves about 206 sq km to complete the villages to replace missing boundary marks would alone amount to Rs 9 declared to be part ofBangalore city which by itself is only a part crore. In view of this, whether or not emplacement of boundary of the wider Bangalore Metropolitan Area which covers 1357 sq stones is required will have a major impact on the cost of efforts km. Moreover the city surveys in Bangalore were undertaken in to improve the functioning ofland administration in India. the 1960s and have not been kept up to date (Burns 2004). ~'1!\l, INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION of India's population. 7 This implies that, at least as instruments in a much broader sense, is in most states far as their place ofliving is concerned, most Indians headed by the Inspector-General of Revenue and do not have documentation of land ownership. Stamps (IGR) or equivalent functions under the Lack of quality control and standards for data finance department or under the commercial taxes products. The lack of updating is most serious for urban department (see Figure 2.4 for a summary of the areas that were previously part of a village, something institutional structure). 8 Even in cases where revenue that applies to the large majority of urban areas in and stamps are located on the same compound, links India today. It is not too difficult to imagine that and contacts with the land records section of the revenue in these cases maps bear little relation to current departments are typically rare. Coordination is made settlement patterns. In the process of conversion, more difficult by the fact that in a number of states town development authorities, which normally the administrative structure of the two departments acquire land for extension of towns, prepare layout is different, implying that sub-registry offices (SROs), plans which are then passed on to the municipal the lowest administrative unit for registration of council. However, the latter has neither responsibility documents, are typically located at city!taluk levels, nor the incentives and capacity to maintain detailed whilst the revenue records are typically maintained data on land ownership. Developers or individuals at the village (revenue estate) level. can, of course, take the initiative of getting their plots surveyed but this is not a routine process and no Main land-related functions cadastral maps to provide physical descriptions of plot According to the 1882 Transfer of Property Act, any boundaries are normally used in the process of sale of tangible immovable property worth Rs 100 registration or kept at the registry. Even in cities or or more can only be made by registered instrument. municipal corporations that maintain maps or some Structure and procedures for registration are governed other type of spatial record, the quality and reliability by the 1908 Land Registration Act with state of such information differ widely. Clarification of amendments. The importance of the department of responsibilities and establishment of clear standards, stamps and registration derives from the fact that, upon together with technical capacity to monitor their registration, a fee proportional to the value of the implementation, could greatly enhance the value of property transacted has to be paid and that, with high such maps even for land use and urban planning and rates, revenues from stamp duty have become a major for a system that could help to define property rights. source of revenue-amounting to about one-third of the total in a large number of states. Stamps and registration Registration ofsales and mortgages. Concerning real Administratively, the office of stamps and registration, estate, the key function of the registration department which deals with registration of commercial is the registration of sale and mortgages, thus creating 7 According to the 2001 Census, India has a total of 615,168 8 In addition to registering property-related transactions such as villages with a population ofless than 5000 (236,004, 158,124, sales and mortgages, the department is responsible for registering and 221,040 with populations ofless than 500, 500-999, and a large number of other instruments (wills, etc.) which we will 1000-4999, respectively (www.censusindia.gov.in). While the not consider in the discussion here. This is justified by the fact total population in these is not directly tabulated, a conservative that, in practice, transactions of immovable property account for estimate would imply that the figure is 753.8 million. the bulk of activity and certainly revenues in most SROs. 6 LAND ADMINISTRATION.::~!- Inspector General of Registrations I i i i i i i Addl Inspector Add I Inspector DIG (9) Addl Inspector District Registrar Accounts Officer General (Guidelines) General (Intelligence) (One for each Zone) General (Chits (Inspection) and Societies) 1 District Registrar i Asst IG (5) (Guidelines) (One each in 5 zones) i District Registrar- District Registrars Audit (45) (one for each of 50 districts) L ________ Sub-registrars (one for each of 558 SROs) FIGURE 2.4: Structure of the registration and stamps department in Tamil Nadu public record of the transaction. Documents to be registered dealings for a given parcel over a stipulated registered must be presented in person at the SRO period. These certificates are usually prepared by within four months of execution. The stamp duty manually going through the indexes of registered to be paid for registering deeds ranges between 5 per transactions. Having an encumbrance certificate for cent and 12.5 per cent of the property's market value a long enough period9 is important for purchasers of and is often somewhat higher in cities and municipal land to explore possible defects in the tide which a corporations than in rural areas (Table 3.2 in Chapter potential seller holds to a specific piece of immovable 3). In addition, a registration fee that ranges between property, implying that the ease with which such 0.5 per cent and 2 per cent of the market value also documents can be made available will have a major needs to be paid. Registration of mortgages requires bearing on tenure security. The EC is also relevant for payment of the registration fee plus a share of the financial institutions wanting to use a property as mortgage value (from 0.1 per cent to 2 per cent). Thus, collateral for a loan to ensure that no prior and possibly with the decline in the importance of land revenue, superior lien has been registered on it. It is thus not income from registration fees has evolved as one of surprising that this service is in high demand by the the major sources of income for states. public and the financial sector. At the same time, Supply of encumbrance certificates. The sub- registries also supply, against a nominal processing 9 Typically 13 years or a period greater than the period set out fee, encumbrance certificates (ECs) that contain all the applicable Limitations Act (typically 10 years). lA-LAND POLICIES FOR GROWTH AND POVERTY REDUC frequent difficulties and inconsistencies with property Undervaluation and corruption. To avoid having descriptions on registered deeds and the fact that to pay high stamp duties, many parties resort to registration is not compulsory-and evaded to avoid reporting land values that are significantly below paying stamp duty even in cases required by law- those actually paid. The problem with explicitly or greatly constrains the value of evidence from the implicitly condoning such an approach in manual registration system, irrespective of whether it is systems 10 is that values ofland have to be determined generated manually or via computer. or certified by a government official who will often demand a fee for his 'service'. Key challenges Costly and complex procedures even for In practice, the effectiveness of registration is reduced routine operations by four factors: (i) high stamp duty and often opaque Indexing and retrieval is difficult. Deeds are often methods of valuing properties prompt parties to keep lengthy and complex documents that may run over their transactions informal, with negative consequences 40 or 50 pages, and specialist knowledge is required for the completeness or integrity of the system; (ii) low to decipher them and navigate the system. A plethora efficiency and high cost of even routine operations of state-specific land reform legislations and land use act as further deterrent to formalizing transactions; restrictions add to this complexity. Simplification (iii) lack oflinks to other parts of the property system of these laws could help to standardize deeds- reduces the security which registration can provide something that is at present often not complied with even under optimum conditions; and (iv) the fact even in situations where regulations exist-and thus that registration neither ensures legal validity of a help make the process of registering property more transaction nor provides parties with strong assurance transparent and less costly. The registration system is as to the validity of the transaction, discourages governed by antiquated procedures including manual registration. We discuss these in turn. copying and indexing of documents by person rather than by parcel which increases the cost of obtaining High stamp duties discourage registration information. High costs of obtaining information are High transfer taxes push transactions into informality. particularly important in the case of ECs for Despite recent reductions of stamp duty by some mortgages where they could well result in a large states, rates remain, with an average of only slightly number of otherwise credit-worthy borrowers being below 10 per cent of the value of the transaction, rationed out of the credit market. very high by international standards. This discourages Long delays in performance ofessential tasks. Absence formal land transa,ctions and leads the parties of performance standards that made it possible for involved to avoid registration through various means 10 such as powers of attorney and other side agreements, States deal with this problem in interesting, though slightly thereby reducing government revenue and threatening contradictory ways. For example, in Kerala the stamp duty was to undermine coverage by the land administration increased by 2 points (to 14 per cent) based on the argument thar doing so was needed to maintain government's tax income system. In the case of mortgages, a common practice in the face of pervasive under-declaration of land values. In to avoid payment of stamp duty is to enter into an Karnataka, when computerized market valuation was equitable mortgage by way of deposit of title deeds introduced, the amount calculated was automatically reduced rather than a formally registered mortgage. by one-third to account for previous undervaluation. LAND ADMINISTRATION public servants to keep documents pending for that is induced by institutional overlap and which registration or not return original document for years greatly increases the cost ofland administration. The without a good reason. In Madhya Pradesh, it takes main institutions involved in land administration, 30 days or more for an EC to be issued, an activity together with key data flows associated with dealings that can be performed in less than 5 minutes in in land are illustrated in Figure 2.5. The figure neighbouring states (Price WaterhouseCoopers illustrates the complex linkages between different sub- 2006). Delays of 2 to 3 years in registration are not systems and the fact that, at any moment, the system uncommon where extensive and often unnecessary may contain a large amount of overlapping or quality assurance procedures slow the process, or inconsistent information. Local governments as well where many offices have to be contacted for official as the survey and settlement department are likely papers (Hanstad et al. 2006). 11 Simple improvements to use city surveys and tax maps on the one hand in information management could have far-reaching and village maps on the other for some of the same impacts on these performance parameters. areas. Information on ownership of the same parcels of rural land may be available in the registry (ifland Duplicate and redundant processes was transacted or a mortgage was taken out) or the Institutional overlap. Many of the problems with land RoR and the two will differ if land was transferred administration in India can be traced to duplication but mutation was not effected, although neither of Copies of plans, Property tax Property card receipts, Khata extracts Advice of mutation FIGURE 2.5: Land administration institutions and data flows Source: Burns (2004). II An extreme case is Maharashtra where delays of 10-15 years for microfilming of deeds had essentially broken down in 198 5 have been experienced by parties in getting back the original without being repaired or replaced. documents they had registered because the centralized system 19 - ! t ' • : INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION them may be definitive. A5 a consequence, any person between the different systems are difficult to obtain, interested in entering into a transaction for land must a recent study shows that even in Andhra Pradesh, a conduct searches in the land record and the registry state that is considered to be one of the more advanced offices. In each of them, time has to be spent and- ones, less than 40 per cent of registered transactions official as well as unofficial-fees need to be paid. were followed by a mutation of the RoR and less Complex procedures. The difficulties resulting from than 30 per cent had a survey done (Agrawal2006). institutional fragmentation can be illustrated by While the extent to which this affects the quality of reviewing the processes that an individual transferring records more generally depends on the frequency of a parcel of (rural) land has to go through once he/she land transactions, the cumulative effects can be has tried to ensure that the transaction is legitimate. 12 substantial, considering that in many states it has First, the transaction has to be registered and stamp been between 70 and 100 years since the last survey duty paid. Once this is done, the registration and was undertaken. Coexistence of the revenue and stamps department gives notice to both the revenue registration record systems duplicates effort and creates and the survey and settlement departments by sending costs that do not correspond to any tangible benefit a document (J -slip) to do a mutation, as any registered and can lead to mismatch of information regarding the land transaction 13 must also be recorded in the RoR/ same piece ofland. 14 This generates ample potential record of tenancy and cultivation (RTC) or the for fraud and land-related conflict which are near property card in the case of agricultural or urban land, impossible to resolve and will overburden the judicial respectively. However, to mutate the RoR or to survey system. Responses from the World Bank's 'doing the property, the interested party has to approach the business' study reveal that such faults, together with corresponding departments to take action. For a the reason for inappropriate description ofboundaries variety of reasons, which may include the purchaser's are a key reason for land conflict. Depending on the inability or unwillingness to pay the associated bribes state, these conflicts take between 3 and 12 years to or the corresponding department's lack of capacity, be resolved, a period during which there is strong this may not happen. The result is a discrepancy disincentive for undertaking any investment on the between what is recorded in the registry and the RoR land concerned. or the spatial records. Even if the process is followed through, the amount of time required is substantial, Presumptive nature of registration and certainly not less than three months, and this provides lack of cross-checks considerable potential for abuse. Limited value added by registration. While the law Inconsistencies create potentialfor ftaud and conflict. states that registration of a legally valid deed will result Although estimates of the magnitude of discrepancies 14 A famous example relates to land that was acquired by the 12 In the case of urban land, the process may be simpler if government under the Land Acquisition Act. In this case, the property cards and city surveys are maintained in the same office. registration department, unaware of the change in ownership On the other hand, in a large number of cases where one or status, issued ECs that did not indicate the true ownership status both of these do not exist, the purchaser is left vulnerable to of the land. This was exploited on a large scale in fraudulent potential conflicts. land sales when the Urban Land Ceiling Act was imposed. In 13 There are some types ofland transactions, inheritance being this wav large amounts of land was subject to ongoing the most common of them, that need to be entered in the RoR proceedings under urban ceiling legislation (and rhus likely to but do not need to be registered. become government property in the future) were sold to LAND ADMINISTRATION 'll- in transfer of title to a given property, the registrar system expends a huge effort to ensure the correct has no obligation-and often lacks the ability-to identification of parties to a deed (an issue that is check the validity of any deed that is presented for regulated in minute detail) but virtually no regulation registration. Even in cases when rudimentary exists which is at least as relevant for the accurate checking could have discovered a defect in the geographical description of the land in question. transaction, the registration department cannot currently be held accountable for having registered POLICY IMPLICATIONS a deed in which title was unclear or disputed. In fact, given that the rationale for registration was generation The complexity and overlaps in the land administration of revenue, the registrar will, at least in principle, system put an additional strain on a system that already register any document put to him as long as the formal has difficulty in coping with demand and there is little procedures are met and the appropriate stamp duty disagreement about the need for improvement. In fact, is paid. This means that the perceived benefits from states have launched a large number of promising having a transaction registered may be significantly initiatives aimed at addressing some of these problems. lower than the cost of doing so, implying that Since the late 1980s, the central government, under landowners may opt out of the system. the leadership of the MoRD, the MIT, and the No checking ofspatial data. The registry's inability National Informatics Centre (NIC), has provided to access and/ or easily refer to cadastral maps makes invaluable technical and financial support and it very difficult for registry officials to perform even guidance to make innovations feasible and to transfer basic checks on the validity or legality of instruments successful innovations to other states. Reviewing this delivered to them, making it near impossible to enforce experience will thus be critical to chart out a possible stricter standards of accountability. This is exacerbated way forward for land administration in India. by the fact that the description of land parcels in Reviewing lessons from the efforts in the area of both many deeds may leave a lot to be desired in terms of textual and spatial records can provide some cues for providing an accurate location of the property, ways to address the issues as well as more concrete particularly in urban areas, something that greatly policy recommendations. increases the possibility of future disputes. Thus the unsuspecting buyers who were not aware of the proceedings under the Ceiling Act (Saxena 2005). CHAPTER Improving Textual Records Examples and Potential It has long been realized that, given the large number LAND RECORDS of records involved, computerizing the textual records The RoR is not only an important document to prove used in land administration could bring significant land ownership but it is also required for a variety of advantages and cost savings. In the late 1980s, the administrative purposes, most importantly, taking out MoRD initiated a centrally sponsored scheme aiming crop loans or insurance.ln a nutshell, computerization to computerize land records that was later joined by ofland records involves digitizing of the entire textual an initiative for roll-out of promising e-governance database, designing database software to manage these initiatives by the MIT. Although success was limited, data and enter mutations in the records, and providing partly due to the limitations then still imposed by local access points (for example, touch-screen kiosks) technology, very promising results have been obtained where villagers can either obtain legal copies of these more recently. Review of states' experience with documents or view the records to check their status. initiatives to computerize revenue records and land By reducing the official and unofficial costs of registries allows identification of the best practices obtaining such records, this will reduce the outlays for this process, the benefits from doing so, and the to be incurred by farmers and thus make it easier to, challenges that remain to be addressed. for example, access credit. In both cases, a number of factors were critical for success. All of the innovative solutions are the result Status and benefits from of decentralized innovation that emerged in individual computerization of land records states-although supported with central funding- and subsequently adopted by other states. Also, in Status. Table 3.1 illustrates that 12 out of 14 states in all cases, innovations were designed to draw heavily our sample have managed to digitize more than 85 per on India's strengths in IT, extensive use was made of cent of their textual records. Half of the states in the subcontracting and PPPs to overcome weaknesses of sample actually use the computerized data on a regular the public sector, and the emphasis was on broad basis. Only 5 states, however, have banned manual coverage and quick roll-out of viable innovations with records, if not in the entire state, at least in the districts a possibility oflearning from experience and adding where computerization has been completed. more sophistication later, rather than designing a Economic benefits and credit access. By ensuring 'perfect' solution from the start. availability of records in a kiosk for a nominal fee, IMPROVING TEXTUAL RECO TABLE 3.1: Status of modernizing land records in shifted access away from the village to taluk-level selected Indian states kiosks, it could actually end up making access more Digitization Used Manual difficult. To address this issue, a number of states (%) rec. banned now award licences to private kiosk operators against Andhra Pradesh 100 No No the payment of a fee. Examples include cyber-cafes in Bihar 20 No No Rajasthan, 'e-dhara' in Gujarat, and delivery ofRTCs Gujarat 100 Yes Yes through private kiosks in Karnataka. In Rajasthan, Himachal Pradesh 92 No No Karnataka 100 Yes Yes users are willing to pay Rs 20 at a local cyber-cafe Kerala 95 No No for an RoR that could be obtained for Rs 15 at the Maharashtra 100 Yes No teshil headquarters, illustrating the demand for, and Madhya Pradesh 100 Yes No the commercial potential in, making records available Orissa 85 No No locally. Other states are exploring the scope of using Punjab 5 No No Rajasthan 100 Yes Yes bank branches as outlets for land records, delivering Tamil Nadu 100 Yes Yes land records through mobile vans as seen in two Uttar Pradesh 100 Yes Yes districts (Latur and Nashik) ofMaharashtra or, as in West Bengal 97 No No Gujarat, providing thalathis with a complete set of Source: PriceWaterhouseCoopers (2006). RoRs printouts for their village which they can use to make computer copies available on demand. computerization has helped to eliminate the scope Generation of user foes. In virtually all the states for tampering with records by village accountants and where computerization has been effected, significant eliminated petty corruption arising from demands fee income, which can in turn be used to expand for 'speed money' or other side payments by local system coverage, is being generated. For example, in officials (Ahuja and Singh 2006). An evaluation of Karnataka, total income generated from sale ofRTCs computerization in Karnataka puts the total savings in 2004 amounted to about Rs 19 crore per year, in terms of bribes avoided at Rs 80 crore (or US$ 16 leaving a surplus of about 10 crore for the state even million) per year and the value of the reduction in after payment to private operators for a variety of tasks. waiting time at 6.6 crore (Lobo and Balakrishnan This allows not only to comfortably cover operation 2002). Case studies also point towards other economic but is also sufficient to pay back the initial investment benefits beyond the lower transaction cost of obtaining which, in the case of Karnataka, had amounted to RoRs, for example, through better credit access and about Rs 24 crore (52 per cent on hardware, 32 per associated scope for investment or insurance. More cent for data entry, and 15 per cent for training). The systematic evidence on these benefits remains limited fact that in many cases, the entire investment has and there is scope for further study. One reason for a already been fully repaid illustrates that computerizing more limited impact on credit could be that in a land records can be an attractive proposition even number of the early adopters, the reduction in the from a purely commercial point of view. cost of obtaining an RoR may have been outweighed Platform for integration with registry From the point by the need for farmers to travel to the kiosk in the of view of the overall land administration system, one taluk to obtain the RoR. of the most important benefits of computerization Data access at the village level. Originally, a key is that it can be used as a springboard for back-office concern about computerization was that, since it integration that has the potential over time to eliminate DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIO some of the current duplication of systems. Experience prepare the ground for a gradual transaction-based shows that, once records are computerized, it is not upgrading of spatial data which currently constitute too difficult to electronically initiate mutations and one of weakest elements in India's land administration have a system to monitor that these are actually system. For example, in Karnataka an order requiring followed through. If the registry is computerized as subdivision survey as a prerequisite for making any well, there will no longer be a requirement for two mutation in the records that involves a change in separate processes as all the information needed to start parcel boundaries was passed to this effect (Jaamdar registration of a transfer and mutation of the records 2006). The manual process used earlier is illustrated can be collected once, either in the registry or the in Figure 3.1. One notes that three different institutions records office. All that is then required is that the have to be approached in sequence-with all the necessary regulations and circulars are passed to ensure associated problems-and three sets of documents, that this is indeed feasible from a legal point of view. that is, maps, records, and a registry entry, are In fact, a number of states, including Gujarat, Tamil maintained for the same parcel in almost complete Nadu, Maharashtra, and Karnataka, have moved isolation from each other. towards such integration. While limited connectivity By contrast, under the new process (Figure 3.2), implies that this still has to be done via dial-up or even only one document with all the information contained physical exchange of floppies or CDs, the availability in the RTC and the various maps will be maintained of statewide wide area networks (WANs) will provide and interested parties can approach the system either the basis for automating andre-engineering this process. through the registry or the land records office, Basis for updating the spatial framework. depending on the type of transaction. Except for the Computerization of land records can also help to generation of notices which, by law, requires a period [ MANUAL SYSTEM J • J • + • ( Registration ( Mutation J ( Survey ) ~ • J ( Registration Dept ( • Revenue Dept ( Survey Dept ) + + ~ ( J Slip ) • Mutation Process • Separate application, Phodi (Subdivisional) • Separate application and fees action • Independent action and delays • Delay and corruption • Corruption due to lack of surveyors • Incorrect records • Obsolete and inconsistent records + l RTC • Map + Tippan + Akarbhand l Updated New RTC I ) I FIGURE 3.1 Manual system of effecting land transfers Source: Jaamdar (2006). IMPROVING TEXTUAL RECORD INTEGRATED SYSTEM o Collect stamp duty o Collect registration fee • Collect mutation fee o Collect Hissa sketch o Unregistered rights report to • Pay mutation fee o Pay phodi fee Field Word-RI and VA Field work-Survey RTC with Spatial and Textual Data Govt kiosk!PPP (RTC + Village Map + Tippan + Akarband FIGURE 3o2: Re-engineered system for land transfers in Karnataka Notes: VA-village accountant; Rl-revenue inspector. Source: Jaamdar (2006). of30 days, the whole process would be fully automated sources of possible inconsistencies, errors, and demands and, in case all the information is found in order, for bribes. Given the difficulty of making changes can be completed in a matter of minutes rather than in the current institutional set-up it is important to months. More interestingly, once the system is in place, note that this would be possible without making any a number of functions (indicated in Figure 3.2 by changes in the existing institutional set-up. shaded boxes) can be outsourced to the private sector even under the present regulatory regime with the Best practice scope of adding others as the capacity, reliability, and reputation of the sector and experience in managing Abolition of manual records. Experience illustrates the system increase. For example, if it were possible that, unless regulations are passed to give legal validity to allow well-reputed surveyors-who have established to computerized RoRs and declare manual records their credentials over a period of some years-to take invalid as soon as the former are available (possibly over the process of notice generation, it would be on a district by district basis, once digitization has possible to have a true one-stop shop instead of the been completed), it may be difficult to make the present fragmented manual system with its numerous transition towards computerized records. The reason DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCT! is that running manual and digital records in parallel compared to the amount of informal payments that for any extended period of time, as is being done in had often been required under traditional processes, 1 some states, is costly also, if in doubt, staff will always revenue from fees was critical to make the project tend to use the system they are familiar with. In fact, independent, financially as well as politically. Most in a number of states where manual records were importantly, it provided the funds needed for expansion not abolished, digitization of 100 per cent of the into related areas and created the scope for having textual data and equipment of offices with expensive licensed private operators take over the running of computers could not prevent computerized records individual kiosks. The importance of this point is from falling into disrepair. Having a clear timeline illustrated by the fact that some states, for example, for discontinuation of manual records will not only Madhya Pradesh, that originally started out with eliminate the discrepancies, confusion, and uncertainty revenues from computerization going to the general associated with parallel systems but also create the budget are now in the process of establishing incentives for minimizing errors in digitization, independent societies to use these resources. something that has in many cases been further helped Security features and audit trail. Doubts about the by the decision to outsource this task to a private accuracy of the data used or their integrity could operator. This is important because even in states where easily jeopardize and derail the whole process of digitization is complete, delays in data verification computerization. Thus, issues of data quality were 2 can develop into a bottleneck. given top priority right from the beginning and a Sustainable level of user charges. Rather than highly transparent process was adopted in most states. succumbing to pressure to make RoRs available below Two key elements critical for this are use of a state cost, the need for a financially self-sustaining model data centre where all the copies of the data would be needs to be recognized from the beginning. In held from the very beginning, and clear procedures Karnataka, a user charge of Rs 15 was set and it was for access control, biometric identification, and a agreed that any user charges would be kept in a separate strict audit trail to detect any tampering with the data. account, rather than letting them disappear into the In a number of states, for example, in Gujarat and government's general revenue. Although modest, Karnataka, a transparent process of data verification was implemented. This process drew on the active 1 In Orissa, applications for issue ofRoR are sent to the tehsildar participation of users by giving them a free copy of who then prints out the copy from the computer, sends it to the RoR within a set period (often 3 months) during the record keeper for verification with the manual records and, which any irregularities and inconsistencies could be after this verification has been done, issues a certified copy. 2 Delays in verification of records, which may well be due to reported for correction. As a further precaution against foot-dragging by the bureaucracy, have emerged as a key reason unauthorized alteration of data or the use of records for the fact that, in Bihar, digitization of records has not been that are no longer current, regulation also established completed in a single one of the 22 districts and that there is a that only the original RoR, duly signed, would have serious danger that land records will be already outdated by the legal validity, making photocopies unacceptable. time the process has been completed. One option of dealing with Putting data on the Internet. An increasing number this is to provide computers and allow circles, some of which of states (Madhya Pradesh, Orissa, and Rajasthan) now have completed digitization, to switch over to an automatic system, thus demonstrating the benefits from doing so to others. make land records data, in most cases those from a Similarly, in Orissa, records for 157 out of the 171 tehsils in the year before rather than the current one, publicly state have been computerized but have not yet been verified. available on the Internet for viewing while others IMPROVING TEXTUAL RECORI) (Gujarat and Karnataka) which already use intranet digitization of redundant fields was not so much of facilities could easily follow this example. Doing so a problem for the more advanced states. However, further increases transparency and users' confidence in states with a less favourable policy and institutional in the system and can help remove some of the secrecy environment where the process remains to be that currently surrounds land records. In cases where completed, and given fact that a focus on problems are known to exist, this could be combined computerizing the 'wrong' records can be used as a with a structured process to report and resolve any strategic move to derail the process, guidance on this inconsistencies, possibly through public viewing and issue is extremely important. Even if the cost of verification of records in the field. Enhancing the digitizing these data seems marginal, it will slow down search facilities available on the web would allow the process and tie up resources that could be used more government and non-governmental organizations profitably elsewhere. Identifying a set of fields that (NGOs) to use the database for more systematic are common with the registry and with other states searches that can be beneficial for planning and many so as to ensure inter-operability and portability of other purposes. Moreover, if electronic signatures are software has significant potential downstream benefits. legally recognized, it would provide the basis for Extension to urban areas. Given the low transaction licensing of private kiosks to deliver RoRs, the volume in rural areas, computerization of rural importance of which can only increase with the records is arguably economically less rewarding than advent of greater connectivity in rural areas. computerization of property cards in urban areas where there is a more active land market and land Remaining challenges values and demand for documents certifying property ownership are much higher. Indeed, a number of states Even though different states have made considerable have made considerable advances in this respect. For progress in computerizing land records, the fact that example, Maharashtra, where city surveys are well others are still lagging behind points to the potential maintained, has issued about 5.6 million property to apply some of the lessons learned. cards and all 138 city survey offices have in place a Prioritization ofcontents and integration with other fully computerized system to handle these documents. systems. RoRs in different states contain large amounts Advances in this area are also being made in Gujarat of information (cropping patterns, source of irrigation, where property cards for seven municipal areas have soil quality, etc.) that were historically important for already been computerized and those for other areas revenue purposes but have limited relevance for land are in progress. Guidance regarding issuance and ownership. Given the large number of land records involved, any additional data field to be digitized will have major implications in terms of the costs and less than the expected impact because the document computerized was the basic tax receipt (BTR) which attests to the payment of logistics of the process. Inaccuracies in these fields can land revenue only instead of the thadapar pagarp, i.e. the local also impact the public perception of the overall system. equivalent of the RoR. In Bihar, the decision to computerize not It is, therefore, important to be selective and ensure only register 2, which is used by revenue karamcharis in their day that whatever information is entered into the computer to day work, but also an old survey record (chalu khatiyan) that is actually relevant for property rights in land. 3 The has not been updated since the early 20th century and is thus largely irrelevant, led to major delays in data entry. These were 3 For example, in Kerala, computerization of almost 95 per cent compounded by the fact that one of the documents was available of an estimated 10.6 million records at considerable cost had at the taluk level and the other at the district level. DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIO verification of property cards would be particularly significant cost savings and efforts to make this important for areas, including rural towns and experience more easily available may merit support. settlements, where these do not already exist. Widely disseminating criteria for proposals to be Institutional integration. Although computerization eligible for Go I support up front will not only generate of records and registration has shown some impressive a broader debate that could be very useful in educating results, the current land record system is only the policymakers, but also provide a basis for accountability computerization of a very old system without any of government officials ex post by listing out serious re-engineering of the underlying processes or expectations against which their performance can an examination of the extent to which these may still eventually be compared. Given the demonstrated be appropriate. Re-engineered processes that allow revenue earning potential of projects in this area, it to achieve functional integration with registration and would be prudent to carefully examine the need for surveys that go beyond a mere computerization of central funds and possibly make a more specific old manual systems will be cheaper and more distinction between 'roll-out' activities which will sustainable in the long term. Supporting well-defined eventually generate revenue of their own and 'pilot' pilots to provide guidance on some of these will be activities that generate significant amounts of new important to completely realize the benefits from knowledge and thus may warrant 100 per cent Go I computerization and should be a priority for the Go I. financing. The ways in which these funds are rewarded (for example, through competitive grants mechanisms Policy implications in the case of pilots) as well as the specific outputs to be expected could, in turn, be specified in more detail Our review clearly points towards a number of best to ensure maximum impact. practices that are critical for computerization to be successful. The effectiveness ofGoi schemes to support these can be enhanced by incorporating the best MODERNIZING PROPERTY practices (identified above) more systematically in REGISTRATION: ADVANCES MADE the programme design and evaluation of proposals AND CHALLENGES AHEAD for funding. Specifically, computerization initiatives While starting from the other end of the spectrum, should be supported only if: (i) the information to considerable progress in computerizing land registration be digitized is clearly relevant for landowners; (ii) a has been made in states such as Andhra Pradesh time bound plan for data verification and abolishing (CARD), Maharashtra (SARITA), Tamil Nadu (C- manual records exists; (iii) key elements of the process STAR), Rajasthan (RajCREST), and Karnataka are outsourced to the private sector; (iv) there are plans (KAVERI). Table 3.2 illustrates that, in addition to to address concerns of data integrity in a satisfactory this group of states where by now a computerized way; and (v) they are based on a business model that registration system is fully operational in all or the has the potential to fully cover operational costs. large majority of SROs, there is a second group of Complying with such requirements will not be states, including Bihar, Himachal Pradesh, Kerala, difficult; on the contrary, the ability to draw on the Madhya Pradesh, Orissa, Punjab, Uttar Pradesh, and experience of states that have already completed the West Bengal, where efforts have not gone much beyond process is likely to allow 'latecomer' states to realize data entry or piloting on a limited scale. Reviewing IMPROVING TEXTUAL RECO the accomplishments and some of the obstacles is not enough to establish causality, it suggests that an encountered in this context allows us to draw lessons automated registry can make it easier to implement from and highlight the challenges for the sector. such changes. In all cases, reduced rates were associated with increased total stamp revenue, suggesting that Benefits from modernizing property informality is not insurmountable and that citizens registration are likely to be willing to pay 'reasonable' amounts of taxes in a hassle free and predictable environment Improved delivery ofregistration services. In the states for good service. that have completed the process, service level agreements Transparency through automatic valuation. A key and penalties in the case of non-performance have problem that plagued land registration in the past ensured that standards are vastly improved from those was the high level of discretion by individual officials that existed before, these are regularly met, and the in valuing property which was often seen as an arbitrariness of valuation has been greatly reduced. invitation for corruption. To reduce arbitrariness in The experience shows that computerization can help the process, an automatic mass valuation module is to improve accountability by reducing the scope for embedded in most states' software. The associated demands for 'speed money' as well as discretional reduction in bureaucrats' discretionary power is one valuation of properties. Also, under the new system, of the most visible benefits from computerization of search for ECs takes minutes rather than days and the registry. The experience of different states also can in many cases be done through the Internet rather illustrates that, with increased connectivity, virtually than through physical presence in the SRO. Although all of the advanced states are moving towards putting quantitative evidence thus far is limited, it is easier for market valuation guidelines on the web to enhance banks to check a potential client's credit worthiness transparency. and should thus be associated with higher levels of Encumberance certificates. Operation of a fully credit access, something that is consistent with the automated registry in an environment where recent large increase in mortgage lending that has been registration is compulsory will, over time, lead to observed in India. the accumulation of evidence that would allow ready Increased number oftransactions and tax revenue. issuance of ECs by computer. To make this service In Maharashtra, computerization has been associated available immediately, digitization oflegacy data for with a 50 per cent increase in the number of registered an appropriate time period is necessary. Tamil Nadu transfers, from about 1 million in 1998-9 to more and Andhra Pradesh have done this for 18 years, with than 1.5 million in 2004-5. Stamp duty collected efforts to digitize legacy data and automate supply of during the same period has more than doubled, from ECs ongoing also in Rajasthan and, on a more limited Rs 1624 crore toRs 4137 crore, despite a significant scale, in Karnataka. Although empirical evidence on reduction in the rate of stamp duty (from 13 per cent impact is still limited, the ability to verifY whether a to 8 per cent) that was undertaken gradually before property has a lien in a matter of minutes through the 2004. We must also note that so far reductions in stamp web will not only make it easier to transact property duty have been implemented only in states (that is, but also ease the burden of financial intermediaries Maharashtra, Kama taka, and Andhra Pradesh) where who want to approve loans. the registry is fully computerized. While this correlation DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIO.· Basis for forther integration. Similar to the case of TABLE 3.2: Status of registration in selected land records, states that have computerized the registry Indian states have embarked on a range of innovative ventures Total Digital Stamp duty Registration including: (i) establishment of a link between the (o/o) fee registry and land records systems (on a pilot basis in (SROs) (SROs) Rural Urban Maharashtra and Gujarat); (ii) allowing registration Andhra 387 387 5 10 0.5 Pradesh from anywhere rather than the particular SRO in Bihar 111 0 8.4 15.4 2 whose jurisdiction the property is located in Rajasthan; Gujarat 150 25 6 1.5 (iii) application for ECs through the web rather Himachal 109 10 8 8 2 than via personal presence in Tamil Nadu; and Pradesh (iv) remittance of stamp duty through franking and Karnataka 202 202 8 10 banks. The fact that the impetus for making some Kerala 308 0 10 12.5 2 Maharashtra 359 359 8 8 of these improvements originated in response to the Madhya 226 0 9 10.4 new system illustrates that it may be better to first Pradesh introduce a limited innovation and then expand Orissa 176 0 8 11 2 rather than to develop a full-blown solution only to Punjab 153 0 6 6 find out that it is inappropriate. Rajasthan 346 269 8 11 Tamil Nadu 558 350 8 13 Uttar Pradesh 344 1 8 10 Best practice elements West Bengal 238 4 6 10/8 1.1 Source: PriceWaterhouseCoopers (2006); Aim eta!. (2004). Process re-engineering. Development of SARITA in Maharashtra (which sparked many the other states' initiatives), helped to considerably simplifY the process crore that was difficult to obtain, given the precarious of registration by clearly defining the requirements state of public finances; lack of technically savvy staff for a 'complete' document that can be registered, to maintain and upgrade hardware and a poor record determining upfront the fees and charges, designing of the public sector in running computer centres; a simple forms, accepting only complete documents, diversion of government employees from their core and immediately returning incomplete documents roles; and the danger of strong resistance. To overcome with a written explanation of the reason for rejection. these problems, the project was outsourced to private This facilitated the use of an automated system for vendors on a build-operate-transfer (BOT) basis. The scrutiny, checking of supporting certificates, and the private partner supplies and installs the hardware and immediate scanning and archival ofdata. Furthermore, peripherals needed, keeps the set-up at all locations to reduce other forms of discretion, time limits for functional, ensures data entry and scanning of original every activity were set, which in the end reduced the registered documents, and complies with specified time required for the registration of standard cases performance standards against a share of the user fees. to less than 30 minutes. In case performance standards (for example, minimum Private sector involvement. It was realized that uptime or maximum time for registration of a undertaking the Maharashtra project within the document) are not met, penalties have to be paid. public sector would run into formidable challenges, Fiscal sustainability out ofuser foes. The increase in including a high initial capital cost of about Rs 20 the number of transactions noted earlier dispels the IMPROVING TEXTUAL RECOR notion that users are not willing to pay if they get maintained is critical. Some of the challenges for doing good quality service. For example, in Maharashtra so are discussed here. equity is achieved by cross-subsidization whereby fees Levels of stamp duty. Table 3.2 illustrates that, to be paid remain uniform across the state (currently although rates of stamp duty vary widely across states, Rs 20 per page) but the amount retained by the private from more than 15 per cent in urban Bihar to 6 per operators is higher in outlying areas and lower in the cent in Gujarat, they remain relatively high, even main cities, in line with the cost of service provision. after the recent reductions by some states. Registration With an average Rs 12 of the Rs 20 fee per page going fees range from 0.5 per cent in Andhra Pradesh to 2 to the BOT agency, total inflows in 2004 amounted per cent in Bihar, Himachal Pradesh, Kerala, and toRs 45 crore to BOT vendors and Rs 35 crore surplus Orissa. To put this into perspective, Table 3.3 presents which was used to upgrade the system. levels of transfer taxes and registration fees, as well Starting ftom areas with high transaction volume. as regular land taxes, for a sample of 17 countries. All successful cases of computerizing the registry started As much as possible, figures are for urban land that in locations with high levels of business volume and is normally taxed higher than agricultural land. While then spread out to less advanced localities, building there is considerable variation, the payment to the on the experience, resources, and growing credibility. state required upon transacting land exceeds 5 per Doing so also entailed the possibility of maintaining cent only in one case, Indonesia, where levels of equity by cross-subsidizing the operation of registries informality are notoriously high and where reduction in less advanced areas. The fact that in virtually all the of transfer taxes has been on the policy agenda for a cases where the registry was computerized, the capital cost of doing so could be quickly recovered from TABLE 3.3: G1oballevels of land taxation4 user charges suggests that a strategy of starting in Country Transfer tax Land tax (urban) locations with high levels of demand and of Brazil 4.0 0.6 establishing user charges that allow for cost recovery Colombia 1 1.6 and operation of the system in a self-sustaining manner, Denmark 0.8 1-3 is important. The example of Himachal Pradesh Estonia 0.5 0.5-2.0 where, due to low transaction volumes, the expected Finland 4.0 0.5-1.0 Germany 3.6 1.1 volume of business is not sufficiently high to make Hong Kong 3.75 0.5 operation attractive for a private operator, illustrates Indonesia 10.0 0.1-0.4 that it is better to address these issues upfront. Japan 4.0 1.4-2.1 Malaysia 2.3 0.2-0.8 Mexico 4.6 0.3 Remaining challenges Philippines 3.2 0.1-1.6 Russia 2.4 0.1-2.0 The examples above show that innovations introduced Singapore 2.8 1.0 in the context of computerizing the registry have the Sweden 3.0 1.0-1.5 potential to address many of the issues that may Thailand 3.0 1.3 Turkey 3.0 0.4-0.5 undermine the credibility of the land administration Sources: Bird (2004); Brown and Hepworth (2002). system. Building on these advances to maintain and, where necessary, further improve the reliability, 4 In countries where land tax is levied on rental values, this was credibility, and completeness of the information assumed to amount to one-tenth of market value. long time. In all the other countries in the sample, do not need to be registered, partly because it was the duties and government fees to be paid for land deemed unreasonable to charge stamp duty on these. transactions are much lower, ranging from 1 per Exempting such transfers from duty and ensuring cent or less in Colombia, Denmark, and Estonia, that any change in the revenue records where these to a maximum of3.5 to 4.5 per cent in Hong Kong, are registered will trigger a corresponding change in Japan, Brazil, and Mexico (Brown and Hepworth the registry database are not too difficult in states where 2002; Bird 2004). both these databases are computerized, and can be High rates of transfer taxes imply that those underwritten by the necessary regulatory changes. transferring land in India have the incentive to avoid Due diligence by registry officials and parcel registration and opt for informal processes (Aim et a!. identification. Even though a deeds registry can never 2004). This is important because any measures to provide a full guarantee, the evidentiary value of a improve the land administration system will remain deed can be greatly enhanced if officials are required unsustainable as long as the concerned parties have to exercise due diligence and a minimum level of no incentive for registration, for example, because they scrutiny of deeds, encumbrances, and descriptions perceive the cost of doing so to exceed the benefits. ofland being transacted before accepting documents To avoid this, further reductions in stamp duty should for registration. To avoid boundary disputes and be high on the policy agenda. International evidence conflict that may arise from imprecise descriptions, suggests that one possibility of doing so without losing it is important to ensure that identification of parcels large amounts of revenue is to compensate for low be sufficiently detailed. Provided that the regulatory transfer duties with higher levels of regular land taxes framework exists and private sector surveyors can do which are often earmarked for collection and use by the job, this is possible through a simple administrative local governments. For example, the three countries change in the deeds system; in fact, survey plans that with the lowest state burden on transfer have regular meet certain standards are routinely required in many tax rates ofO.S-2 per cent, 1-3 per cent, and 1.6 per deeds systems around the world. cent, respectively, all of which are assessed on market Political support. Experience illustrates that in states values. 5 The fact that land tax rates in India are low where computerization has made less progress than to start with, with actual tax payments often further desired, the root cause can most often be found in reduced by the fact that they are based on rental values political resistance from within the land administration that are artificially depressed because of rent ceilings, system rather than technical issues that could not be suggests that there may be scope for individual states resolved. 6 Such resistance by vested interests should to negotiate with local bodies to come to an agreement 6 For example, in Kerala, resistance by deeds writers and other that would allow increased land taxes and possibly staff concerned about losing discretionary power forced the some revenue sharing in return for lower stamp duties. registration department to disable most of the functionality of Compulsory registration. Another key threat to the the software used, implying that deeds and valuation are still being integrity of the registration system is that currently a done manually and the system is used only to generate indices of large number ofland transfers, in particular succession, deeds. Experience from land administration projects in other countries, where similar issues were encountered, suggests that 5 Simulations suggest that such increases could actually reduce steps to train such people in IT skills to avoid their displacement the tax burden on the poor, something that would make such a have the benefit of helping to maintain the institutional memory. policy move even more attractive (World Bank 2005a). Of course, clear line on corruption is also necessary. IMPROVING TEXTUAL RECORD not come as too much of a surprise; in fact resistance registration is unlikely to be completed on its own from survey or legal lobbies has accompanied most implies that, similar to the situation for revenue changes in the area ofland administration, from the records, it is important to provide states with incentives introduction of para-legal landbrokers by Robert for the computerization of registration that allow them Torrens in South Australia in the 19th century, to to overcome resistance and draw equal with the more the establishment of a new title register that did not progressive states. To the extent that the ease of require intervention by notaries in Peru in the 1990s. transacting land provided by a well-functioning registry In addition to adhering to the above principles, will be an important consideration in investment political support from the highest level, including decisions by firms or individuals who are not limited pronouncements by central government will, therefore, to a given state, failure to do so can, over time, translate be important in overcoming such resistance. into widening disparities between the more advanced and backward states. Policy implications The fact that, despite the considerable potential commercial benefits, the process of automating CHAPTER Improving the Status of Spatial Records One of the unusual features of India's land area ofsurveying. We then highlight that a combination administration system is that a large part of the of satellite imagery with existing spatial data (village information used on a daily basis is actually based maps) can be used to provide an index map that will on textual rather than spatial data. While this allowed allow states to obtain near-complete spatial coverage India to make much more rapid progress than if at a very reasonable cost. Doing so would allow it had first tried to sort out its graphical database, policymakers to: (i) quickly obtain an overview of the running a land administration system without maps problems encountered in different areas of the country is clearly not sustainable in the long term. Finding that could serve as a basis for assessing the needs in cost-effective ways of updating the spatial records, terms of improving spatial data; (ii) devise strategies, many of which are in much worse condition than based on a wide range of modern technology, to address textual ones, is thus of high importance. Specifically, these problems in a way that uses existing data and it will help to: (i) reduce boundary disputes which India's IT capacity in innovative ways; (iii) get at least are already appearing in increased numbers in peri- a rough idea of the types and possible magnitude of urban areas; (ii) unambiguously identifY land parcels benefits from improving land records, that can help over which rights are recorded and thus allow users to realistically estimate the potential for cost recovery to have more confidence in the information provided and, in turn, for potential private sector involvement; by the system; and (iii) identifY and subsequently and (iv) use all of this information to take decisions devise strategies for eliminating gaps in the coverage on the allocation of public funds for improvement of of textual and/or spatial records and in so doing also spatial infrastructure that would not only fit into a provide a basis for the more sustainable management broader strategic framework but also allow the of public lands. development of private sector capacity to complement In this chapter, we argue that applying the the expertise already available within the government. principles that helped to underwrite the successful computerization of textual records can also provide THESPAT~FRAMFWORKFOR a basis for the improvement of spatial records. To do LAND ADMINISTRATION: so, we show that failure to adhere to these principles RATIONALE AND STATUS IN INDIA is a key reason for the lack of success of the large There is little doubt that a well-defined spatial majority of pilots which India has conducted in the framework is needed to eliminate ambiguity in the IMPROVING THE STATUS OF SPATIAL RECORD system, thereby avoiding boundary disputes and updating of spatial records was expected to be carried increasing tenure security, in addition to providing out through periodic re-surveys using traditional other benefits such as the ability to search by technology. While it is difficult to get an accurate geographical location and facilitate planning for picture, available evidence illustrates that in many infrastructure, and the provision of other services. states, spatial information is seriously out of date. In Moreover, a defining characteristic of well-functioning Andhra Pradesh, Himachal Pradesh, Karnataka, land registration systems, whether based on title or Punjab, and Tamil Nadu, the last re-survey was improved deeds, is that they allow the unambiguous conducted around the year 1900, 1 there are many identification of land parcels. more districts in other states where the last re-survey Having a spatial reference will enable the mapping dates back to the period before independence, and of the the parcels over which rights are recorded in even where re-surveys were declared 'completed' the textual register, thus ensuring completeness, that considerable gaps persist. It is thus not surprising to is, all land parcels are indeed identified, and there are find that the coverage and often also the quality and no gaps or overlaps of the parcels for which rights are reliability of spatial data are weak and, therefore, registered. This also provides a framework to record simple digitization of existing data is not sufficient changes in parcels, for example, through subdivision to establish a workable solution. or consolidation. This framework can either be in At the same time, attempts at implementing graphical or digital formats and can be useful to traditional re-surveys of the type conducted by the validate textual data, identifying, for example, parcels British in virtually all states suggest that resumption where numerical data are not available. Spatial data of the 30-year cycle of revisional surveys as set down can also be used to support common registration by law is not a viable option for improving the spatial queries, such as quickly identifying parties with interests framework. The time and financial requirements of in adjoining parcels for service of notice as well as such ventures are too large. Even the most optimistic other broader spatial queries such as the identification estimates put the cost for a state such as Karnataka at of land parcels impacted by possible future about Rs 1000 crore and a time requirement of about development activity. The cadastral spatial data can 20 years. Moreover, even if this were technically also be integrated with other spatial data such as feasible, a full re-survey is likely to carry high political topographic data, satellite imagery, etc., to support costs. The reputation for corruption which land broader objectives such as the formulation of land administration has acquired and the tendency of such policy and better land management. The spatial an exercise to generate conflict among households framework can support user-friendly graphical means which have coexisted peacefully greatly reduce of accessing the data by plot. Many people will not the political attractiveness and, thus, feasibility of a know their survey numbers, but most will be able to full re-survey. 2 This means that new models to identify their holdings on a map. The use of spatial 1 map data also readily supports a graphical Interface In Tamil Nadu, an attempt to update spatial information was undertaken in 1987 but this is judged as not having been too with land records on computers or the internet. successful. However, even though it is essential for land 2 'The policymakers are acutely aware of the unsatisfactory administration to fulfil its basic mandate, the spatial nature of the output of survey operations and the exploitative framework for land administration in India is much nature that the whole process had acquired over the years. For inferior to the situation for textual records. Originally, instance, the Government of Rajasthan decided to do away with -~5~!': JNDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIONO conduct surveys are needed. In choosing such implies that precision in actual use is well below models, the complexity of surveying and the fact that potential and that the sustainability of the data is costs tend to increase exponentially with precision not assured. 4 It will, therefore, be critical to carefully requirements-which imply that modalities will identifY the strengths and weaknesses of specific have to vary with land values-need to be taken technologies in specific contexts and to apply a judicious into account. mix that is most suitable for specific situations. In view of the well-documented difficulties of A second and more fundamental concern is that traditional methods to improve the spatial framework, technology is embedded in a broader institutional high hopes are generally pinned on expensive 'modern' environment and that before narrowing the discussion equipment to solve the predicament of spatial land down to technology and accuracy of surveys, it will administration data in India. In fact, to respond to be necessary to clarifY: (i) the role and responsibility this challenge, the MoRD has launched an ambitious of the state vs that of individuals in maintaining the series of 'pilots' in all states to modernize the spatial spatial framework; (ii) the scope for private sector data. Many of these pilots are focused on high-tech involvement; and (iii) the way in which public funds ETS. Experience with such technology in India and should be allocated. To do so, we first identifY some elsewhere suggests that the scope for these pilots should of the reasons why past efforts had only limited be broadened for two reasons. First, while ETS has success and then aim to draw some lessons from its place within a broader array of survey technologies, this experience. reliance on ETS only will not allow the scale needed to address India's problems, 3 but also the capacity WHY PILOTS WITH 'MODERN required to use it is effectively beyond what is available TECHNOLOGY' HAD LIMITED in India's public sector for the foreseeable future. This SUCCESS Limited attention to existing information. Contrary to the Department of Settlement altogether and transformed it into a much thinner Department of Land Records with no many developing countries where cadastral databases systems for wholesale periodic revision of records in place. In do not exist and have to be created from scratch, one Madhya Pradesh, settlement was ordered in 26 districts in 1990, of the great advantages of India is that the spatial but had to be suspended in 2000 due to persistent complaints record created by the British, in the form of village by locals .... Many Survey Directorates (as in West Bengal and maps and FMBs, is surprisingly accurate. The main Tamil Nadu) are either under-worked or have been given other problems with the existing spatial database are lack of miscellaneous responsibilities. Politically it is safer to let the updating or wholesale change ofland use. This implies existing record of rights with all their deficiencies continue, that in (rural) areas where land use has not changed rather than expect that government servants would bring any large-scale improvements through survey based on the ground dramatically and old spatial records are still available, situation. State has abdicated its responsibility and correction of efforts to utilize this information may be warranted, records has been left to the individual initiative' (Saxena 2005). 3 Estimates suggest that, even if the speed of surveying were to 4 A few random checks of measurements taken in one of the be tripled from what had obtained in the final phase of a 33- pilots by an independent consultant revealed that the accuracy village pilot in Karnataka that is generally considered to have attained was in the order of magnitude of meters rather than been quite successful, it would take more than rwo centuries to centimeters, a level that could have been attained with much cover the whole state (Burns 2004). cheaper technology. IMPROVING THE STATUS OF SPATIAL RECORD especially in view of the fact that the MoRD is also of which are more compatible with India's skills mix. 5 involved in efforts to preserve these records. It appears, Only under the recently started Integrated Land however, that the large majority of pilots conducted Information System (IUS) project in Andhra Pradesh under past centrally financed schemes was either is aerial photography, complemented by other directed towards areas where these records were no methods, being applied as the primary mapping longer available or did not make extensive use of them. tool. 6 Given India's strong remote sensing capabilities Obviously, a full re-survey will be the most attractive and the obvious advantages of satellite imagery in option from a merely technical point ofview. However, covering large areas at very low cost and with experience from a number of recent pilots where the sufficiently high levels of accuracy, at least in most changes in different households' land areas that rural settings that are not forested and not too hilly, it resulted from the re-survey created significant is surprising that it has not been used more in order upheaval and conflict among neighbours suggests to generate cadastral index maps by geo-referencing that doing so might not be desirable from a social existing village maps and using this information to point of view even if the funds for it were available. identify high priority areas for follow-up with more With limited funds, the case to focus only on sophisticated technologies. incidents where changes have occurred or to employ No cost recovery. The successful computerization methodologies that build on or combine different types of textual records was possible only because it involved of existing data in more innovative ways becomes cost recovery. However, in spite of the fact that the even stronger. While this does not imply that existing higher cost of surveying would call for even greater data can solve all the problems, there is certainly no attention to the design of a sustainable business model, reason to completely neglect it and instead try to create even high precision surveys are at present provided everything from scratch. without charge, making any progress entirely dependent Narrow technology options. Instead of considering on government budget allocations. Although it is a range of technology options and then focusing on debatable whether users should be made to pay for those that are most suited to a given environment, the full cost of an initial ground survey, the fact that 'modern technology' was in most cases narrowly nothing was charged at all clearly led to a situation defined in terms ofETS, an expensive option in which where no economic bounds were placed on the India has little comparative advantage and which technological options to be explored or the need provides a level of precision that is not needed for at to maintain some correspondence between the least 80 per cent of the country. The emphasis on ETS requirements in terms of precision and the actual land even in situations where it is not the most appropriate values. The lack of cost recovery also implies that technique was rooted in the mistaken belief that precision equals security even though in many cases 5 For example, aerial photography was applied in Madhya Pradesh lack of effective supervision led to large errors even in hilly terrain that is not the most appropriate for this technology and was then discarded. in surveys made with high-precision equipment). It 6 The fact that aerial photography has been obtained for a whole led policymakers to prematurely discard cheaper and district constitutes an interesting pilot that will allow observation more appropriate options without fully exploring of the potential and limitations of this technology in the Indian their advantages and disadvantages, for example, context, implying that attention to further developments in this photogrammetry combined with GPS devices, both state are warranted. fA-LAND POLICIES FOR GROWTH AND POVERTY REDUC'tl TABLE 4.1: Role of government and the private sector in different countries Country System Government role Private sector role Argentina Deed Regulatory Comprehensive Australia Title Regulatory Does most cadastral surveys Denmark Title Regulatory Comprehensive Finland Title Comprehensive Input into govt survey Germany Title Regulatory Mandated to do surveys Hong Kong Deed Regulatory Does most surveys since 1996 India Deed Comprehensive None, except in Andhra Pradesh, Karnataka, and Maharashtra Indonesia Title Comprehensive Increasing since 1997 Iran Title Regulatory Comprehensive Japan Title Regulatory Comprehensive South Korea Title Comprehensive No role Malaysia Title Comprehensive 80% of surveys; approval Netherlands Deed Comprehensive Contracted by public sector New Zealand Title Regulatory Comprehensive Philippines Title Subcontracting S. Africa Deed Regulatory Does all the surveys Sri Lanka Deed Regulatory Comprehensive Source: FIG Commission 7 (2006). limited emphasis was placed on sustainability, regular something that has led to large backlogs, as in the updating of the information generated, and its link case ofMaharashtra where the backlog is estimated at to the broader system. This creates a danger that, about 3 million subdivision requests. The need to cope even in situations where boundary stones have been with such backlogs has led some states (Maharashtra, emplaced and measured with millimetre accuracy, Andhra Pradesh, and Karnataka) to allow private households may still be unlikely to register transactions, surveyors to operate on a limited scale. However, even get mutations done, or have subdivisions surveyed. in these states private surveyors are limited in their Some cost recovery-either directly through user choice of technology and have to follow a paper- charges or indirectly through taxes--could have helped based process that is cumbersome, error prone, and to steer specific technologies more directly to situations subject to approval by a government surveyor in every where their use is indeed appropriate (for example, single case. All these factors tend to increase the cost areas of high values for ETS), thus providing increased of a survey, thereby preventing the system from chances of regular maintenance. scaling up quickly, with the end result of discouraging No involvement ofthe private sector. Given the critical use of the system of the public. importance of PPPs in the digitization of textual To put India's regulatory regime into perspective, records in virtually all states, it is surprising that Table 4.1 summarizes the role of the private sector in attempts at establishing spatial frameworks have been surveying in 17 countries. Doing so highlights not and continue to be run essentially as a public sector only India's exceptional position in terms of the monopoly. This is even more so in view of the fact number ofland parcels but also that India is the only that in many states the public sector is unable to cope other country, in addition to South Korea, where, with even with the normal volume of subdivision surveys, the exception of three states, private survey activity is IMPROVING THE STATUS OF SPATIAL RECORDs·· not allowed at all. This is surprising as even countries One of the arguments advanced in favour of where the role of the private sector has historically restricting the role of private surveyors in India is been strongly circumscribed, such as Indonesia, the that maintenance of the spatial framework is a Philippines, and Finland, have significantly increased responsibility of the public sector. In this context, the role of the private sector in the conduct of survey the experience from Eastern European transition either through subcontracting or by allowing private countries, when faced with a similar problem after the sector firms to operate independently, subject to collapse of communism, is relevant. Box 4.1, which is government examination and approval of the resultant taken almost verbatim from the guidelines elaborated survey plans. by the United Nations Economic Commission BOX 4.1: GUIDELINES FOR PPPs INVOLVING LICENSING OF PRIVATE SURVEYORS The issue of how much latitude to give to the private sector has been a concern for governments all over the world. The following passage, which is almost literally taken from the guidelines elaborated by the UNECE WPlA highlights insights from Eastern European countries on this topic. Land administration activities lie at the heart of good government. Nations are defined by their land coverage and land represents the source of wealth and the basis for economic prosperity. It is, therefore, axiomatic that land administration represents a key group of public activities. But with proper safeguards and good management, it is possible for specific activities to be transferred to the private sector under PPP arrangements. There are many examples of this in relation to licences or concessions granted to surveyors. In addition to the general principles guiding PPPs in land administration, there are a number of key principles relating to such arrangements. 1. Governments retain final responsibility for the discharge of all public tasks. Even in licencing or concessionary arrangements, the government will, at least in the public perception, be held ultimately accountable for performance and liable for the consequences of any errors or mistakes. At the same time, private partners should be made fully responsible for their activities, including liability for poor performance or mistakes. 2. If the government wants to assign a public task to a third party, it should specify, ideally by regulation, the requirements to be met to ensure proper implementation. Such regulations should comprise: (i) professional standards (education, training, ethical behaviour); (iiJ exact competencies; (iii) indicators for performance measurement; and (iv) liability, including financial and other penalties (for example, revocation of licence) for failure to meet performance standards or mistakes. These requirements should be included, possibly in a more explicit form, in the licence or concession. 3. The government must make appropriate arrangements for monitoring and auditing performance by the private partner. Such arrangements may be included within the licence, concession, or statutory regulation. The extent to which the government retains control will need to be carefully examined in light of the specific functions to be carried out. While 'hands on' control by government may provide assurance, it must be avoided as it comes at the cost of stifling the innovation and initiative which the government is seeking from the private partner. Source: UNECE, Working Party on land Administration (WPLA) (2005) http://www.unece.org/hlm/wpla/ 4thSession_LA.htm. DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIO for Europe (UNECE) Working Party on Land already been surveyed, as is implicitly being done in Administration, suggests that, if certain key principles the current system, the criteria for surveys will be more are followed, the private sector can perform a very transparent, and it will be possible to bring equity important role in surveying. considerations to bear and at least offer scope to deal As illustrated in Box 4.1, these principles include: with the urgent need of surveying the many marginal (i) full responsibility and liability of the private lands that were not settled in British times; and (iii) it sector for poor performance; (ii) a strong regulatory would reduce the cost of surveying to manageable levels framework that includes professional standards, by simplifying and modernizing the process and responsibilities, indicators for performance setting precision requirements that are affordable, measurement, and clear provisions for liability; and given the existing technology. (iii) sufficient latitude for creativity and innovation by the private sector within this framework. Obviously, Improving the coverage of the a well-regulated private sector cannot emerge overnight spatial framework and capacity-building and training efforts are required. Still, the fact that private sector involvement has Use satellite imagery to establish digital base maps. In allowed most of the transition countries to build a rural areas, the prime mapping tool should be high- well-functioning surveying sector in a relatively short resolution satellite imagery, geo-referenced and time period suggests that this route could also be an preferably orthorectified. Although this imagery has option for India, especially in view of the fact that a lower spatial resolution than ground or aerial these principles are not qualitatively different from mapping methodologies (0.6 to 2.5 m compared to those applied in digitizing textual records. 0.01 to 0.2 m for high-tech solutions such as ETS) it has a number of advantages that are likely to more than compensate. First, India has strong skills in IMPLICATIONS FOR INDIA the generation and processing of satellite imagery. To meet these challenges and develop a sustainable Moreover, contrary to all other applications, and model for spatial data, it is necessary to strengthen especially aerial photography, data are likely to be the regulatory role of the government and take active available immediately at minimal cost. Second, the steps to expand the surveying capacity at the ground fact that the imagery is readily available should allow level while at the same time putting in place a more quick production of statewide base digital mapping comprehensive but low-precision framework that in most states, with the possible exception of hilly forms not only the basis for prioritizing among and heavily vegetated areas. potential areas to improve the spatial framework but Link to existing spatial information to establish a can also be improved upon by private parties, for cadastral index map. There is little disagreement that example, through subdivision surveys. Running pilots the preservation of existing records is a key task for that would explore technology options in specific cases the public sector and many states have already started would accompany both of these purposes. Doing so processes to digitize or vectorize existing village maps would have the following advantages: (i) the limited and FMBs (see Table 4.2). In fact, given the public sector capacity would be focused on establishing magnitude of the challenge, it would be prudent that, the regulatory framework and providing basic public in trying to reconstitute the spatial framework, as much goods; (ii) instead of focusing only on areas that have as possible use is made of the existing data. The digital IMPROVING THE STATUS OF SPATIAL RECOR TABLE 4.2: Status of modernizing surveying in between existing spatial records and actual land use. selected Indian states The first is that there are unsurveyed subdivisions (in per cent) that may or may not be recorded in the land records. Village maps FMBs Second, land records often become outdated as a result Scanned Vectorized Scanned Vectorized of wholesale change in land use patterns, for example, Andhra Pradesh na 10 78 through urbanization or through construction of Bihar 0 0 0 0 irrigation structures. Third, existing maps may have Gujarat 100 10 Pilot been subject to physical decay and loss. Fourth, there Himachal Pradesh 0 0 0 0 Karnataka 100 50 2 0 is a large number of revenue lands where survey was Kerala 0 0 na na never done. At present, little knowledge exists on Maharashtra 100 100 the extent of each of these situations and generation Madhya Pradesh 7 7 of the above map product would provide important Orissa 3 1.5 1 0 new information. Punjab 0.1 0 0 0 Rajasthan Deciding allocation of public survey efforts. TamilNadu 85 0 5 Overlaying the map described above with other Uttar Pradesh 0 0 0 0 layers-that can provide information, for example, West Bengal 2 na about relative deprivation or the potential productivity Note: na-not available. of the land-will make it possible to make inferences Source: PriceWaterhouseCoopers (2006). on the possible benefits of improving the spatial infrastructure and the identity of those who are likely base mapping from satellite imagery described above to benefit. ClarifYing the criteria underlying the could be used to geo-reference existing village maps, selection of areas in which to expand the spatial in addition to providing a base to reference new spatial framework, can help to increase transparency in the data produced from existing techniques. This would selection of pilots. Moreover, to the extent that having allow the generation of a map-that can at the very a survey provides considerable economic benefits, it least be used as a cadastral index map, similar to the should not be too difficult (for either the government English model and contain all the information available or the private sector) to use this information to make in existing village maps. It would be generated at a predications about households' willingness to pay for very low cost and in a time period that is much shorter improved land ownership and spatial records and the than that possible with the exclusive use of ETS. potential for cost recovery implied by these. This, in IdentifY the size ofthe 'spatial gap: As the satellite turn, could be an important input into the design of image will contain information on actual land use, mechanisms to finance the generation of spatial data its overlay with geo-referenced village maps can be (for example, requiring matching contributions), used to identifY areas where no spatial data are available to target public funds more explicitly towards and to set criteria that can help to establish priorities poverty reduction, and to prevent crowding-our of for undertaking work to fill gaps in a much more the private sector. Such comparison between textual systematic way, after taking into account a variety of records and spatial data would also form the basis technical and cost-recovery options, which was not for identifYing areas where there is a need for an effort possible in the past. To illustrate this, note that there to systematically update and validate on the ground are a number of key reasons that result in discrepancy the textual and graphical information contained in the DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIO land administration system where such information achieved with these technologies is well known, the is seriously out of date, for example due to efforts to need to explore this under Indian conditions, along avoid land ceilings and other administrative restrictions. with the rather rapid development of technology in Approximate (and reduce) the cost ofclosing the spatial this field make it mandatory to explore the options in gap. While the numerous redundancies that are built a more detailed fashion under local conditions before into India's land administration system tend to increase drafting any regulations. Given the far-reaching transaction costs and should, therefore, be eliminated, implications of doing so, it is critical to ensure that a they can possibly be very useful in re-establishing wide range of technology options, as well as actual some of the spatial information if used in a systematic field conditions, are taken into account. manner. For example, in situations where maps have Subdivisions. One class of situations is unsurveyed disintegrated, land records (as well as FMBs) may still or un-recorded subdivisions, mainly in rural areas be available, allowing at the very least to ex ante identifY where there is still a 'reasonable' correspondence be- the number of surveys that need to be undertaken tween the map base and current land use. In this case, and the neighbours who have to be notified. Even in piloting of re-survey options with different types of situations where survey has not been conducted at technology can provide an information basis that all, the satellite image could be relied upon to assess would allow to use the information from the patterns ofland use and gain an approximate idea of satellite/village map to determine the extent to which the requirements involved in establishing a spatial actual land use still corresponds to the boundaries frame. Pilots will be necessary to work out the details identified on the map-something that is likely to of such an approach in each of a number of specific depend on the age of the maps, the frequency of cases that are discussed in greater detail below. transactions, and the completeness of their record- ing. This will provide an indication of whether sur- Pilots options for specific situations veys on a sporadic or a systematic basis would be more appropriate. Overlaying this with a map indi- As measurements undertaken in colonial days cating the area's economic potential would also help were surprisingly precise, the main reason for the in deciding about the possibility of cost sharing. maps generated then to become outdated is through Mismatches. A second situation likely to be unrecorded subdivisions or structural change such encountered is that there is considerable mismatch as urbanization. Having a statewide digital base map between the ground situation and the entries in overlaid with existing village maps would not only different types of records, for example, because of allow quantification of the relative incidence of the desire of large landowners to evade land ceiling these situations but also help to devise strategies to legislation. Once the magnitude of this issue has been address them. determined, possibly helped by some simple local Technical options. There is little doubt that, in order inquiries through the village officer, it will be possible to establish a land administration system that is widely to devise strategies to deal with the situation. Key affordable, India needs a range of technology options, issues are whether or not a systematic field verification broader than that available in the past, allowing in exercise, to clarifY textual as well as spatial records, is particular for the extensive use of satellite imagery, required, how such an exercise would be financed, aerial photography, and handheld GPS. However, even and what changes are required in the regulatory though the general range of precision that can be framework to allow it to be effective, that is, to be IMPROVING THE STATUS OF SPATIAL RECORDS '"t- able to resolve at least the majority of conflicts on the land. In fact, in 2000, the Government of Orissa the spot. Pilots to explore these issues in concrete decided to confer ownership to people in scheduled settings will be of immense importance for informing areas cultivating land up to a 30° slope and to carry the policy process. out a special survey to demarcate the lands in question. Complete change in land use patterns. The most Not surprisingly, given the fact that the technology promising way to deal with the non-existence of maps is not the most suitable for this type of circumstances, or the complete lack of correspondence between carrying out this survey with ETS proved unsuccessful. existing village maps and actual patterns of land Mounting a pilot with handheld GPS-a technology use is likely to be through checking the availability that is much cheaper and can be undertaken and quality of alternative spatial products in other with much greater local participation, as it is less government departments, especially layout plans in demanding in terms of manpower-would not only development authorities and tax maps in municipal be an ideal opportunity to broaden the range of corporations. If the resolution of the imagery is choices considered but also to develop procedures sufficient, a first assessment of the quality of these that could make a huge difference to the welfare of maps can help determine the extent of re-survey tribal populations all over the country. needed. If there is a desire to transform these into ownership records, options for a participatory and Redefining and strengthening the· transparent process of systematic ground verification role of the public sector to do so will need to be explored through pilot activity. Expanding availability of capacity. The fact that 5. Decay and loss. In case village maps have been meeting the normal needs of day-to-day re-surveys lost, algorithms to reconstitute spatial information to already stretches current public sector capacity to its create a simple index map from the imagery, together limits suggests that expanding the number of trained with textual records can be explored through pilots surveyors is critical. This will in the first instance and subsequent public verification of the resulting require technical training. The experience of Andhra product. If successful, these procedures could be put Pradesh where a collaborative effort between the survey down in regulations that would allow the process to department and local SHGs trained village youths be used on a wider scale. to become 'barefoot' surveyors and subsequently Complete lack ofsurvey. One of the shortcomings acquiring a licence with the possibility of either being of past pilots was that their focus was on land that attached to local SHG federations or receiving a loan had already been settled. As a result, certain types of to start their own business-suggests that there are land in Andhra Pradesh, Assam, Karnataka, Kerala, plenty of innovative options that can be explored to Madhya Pradesh, Maharashtra, Orissa, and Tamil increase survey capacity not only in urban centres Nadu, especially those that had been part of princely but also at the village level where the lack of surveying states and those above a certain slope (1 0 °) were capacity is a significant bottleneck. Given that excluded from consideration. Although satellite simple handheld GPS devices with 1-2 m precision imagery may not be the most suitable for these lands are available at a cost comparable to that of setting due to hilly terrain, covering these is important for up a computer kiosk, there is considerable scope to equity reasons because it was essentially state action explore business models for entrepreneurial rural that deprived a large number of tribal households of youth that could make use of this opportunity and their rights even though they had occupied and used DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIONO provide surveying capacity at the local level at a ofbusiness processes to suit Indian conditions, similar reasonable cost. to what was done for registration and, to a more limited Regulatory.frameworkfor private sector involvement. extent, land records. For example, if a decision has Expansion of supply alone will not be effective unless been reached that in rural areas regular subdivision it is combined with a regulatory framework that allows surveys on an index map may be conducted using private surveyors greater independence from the public handheld GPS devices, it would be possible to design sector or, in the majority oflndian states, allows private a completely electronic process where the local surveyor sector activity in the first place. Such delegation of downloads existing spatial records either from the responsibility by the state to licensed surveyors may Internet at the local kiosk or from the taluk office, need to be a gradual process that is based on clear goes to the field to record the coordinates for the newly standards of service delivery and penalties in the formed parcels, has the survey duly authenticated case of non-performance. Ideally there will be an by the affected parties, and uploads the modified plan independent self-regulating body that provides a first for archiving. Supporting pilots to explore such lower- level of enforcement. While this will require further cost options not only for sporadic subdivision surveys work, there are plenty of country experiences that can but also for systematic demarcation in situations where be drawn upon and adapted to the Indian situation. records have been lost, or no previous survey has been Updating technical regulations. While expansion of conducted, would be of high priority. They should capacity is important, it can do little to reduce the at least complement the current de facto emphasis cost of surveys and there is a significant danger that, on high-end technology, especially because the Indian unless lower-cost options together with less demanding experience suggests that households are not willing precision requirements are developed and put in place, to pay infinite amounts for surveys. even the promising innovations to make surveys an Focus public support on poor areas. The above is integral part of the process of mutation may not be particularly important because India has large amounts complied with. Rewriting survey regulations-which ofland, especially in tribal areas, where no survey has are often quite archaic-to bring them up-to-date can, been undertaken yet and where, as a result, property in many cases, be combined with efforts to integrate rights to land may be highly insecure or subject to new technology where appropriate. It appears that appropriation by powerful private interests. While recent developments in handheld GPS provide options private surveyors will be able to cater to the needs of that are well suited to the requirements of rural the better-offin areas with higher levels ofland values, areas in India. For example, in Uganda, systematic especially if a regulation is put in place that allows a demarcation using handheld GPS with an accuracy reduction in the cost of doing so, it is the poor groups of 10 em is currently being undertaken at an average which should be the focus of the government's efforts. cost of US$ 0.41 per point, that is, US$ 1.6-3.2 per The benefits from doing so would be enormous in parcel depending on its physical shape. A systematic terms of poverty reduction, environmental protection, investigation of the options available, including pilot and possible local development. In addition, applications to test the potential under field conditions, development of processes to deal with these issues will is thus highly desirable and could provide a basis for be important to prepare the ground and to effectively such decisions in India. implement the provisions of the tribal land rights Re-engineering business processes. The above should bill, once it is passed. be integrated with, and informed by, re-engineering CHAPTER Towards Greater Tenure Security From the very start, efforts to modernize land A deeds registration system is a public repository administration in India were not an end in themselves where documents evidencing transactions with land but intended to help increase tenure security and make can be lodged, numbered and dated, indexed, and it easier to transfer rights. While modernization of each archived. Recording of the document will give public of the subsystems is a necessary condition for such an notice of the transaction, serve as evidence for it, impact, neither textual nor spatial records will be fully and may assign priority to the right claimed in that effective without an appropriate legal and regulatory document. Registered deeds normally take priority framework. A key concern in this context has been over unregistered ones, or any deeds registered whether and, if yes, when and how, India should make subsequently. However, registration of a deed does the transition towards a system of tide registration, not imply anything about the legal validity of the often also referred to as a 'Torrens' system. This chapter transaction or whether the pa~ties were legally entitled briefly clarifies key differences between tide and deed to carry out the transaction. As discussed earlier, the registration systems and, based on a discussion of registration office will, in principle, register anything the options to improve deeds registration systems, and in practice in India, officers often invest more highlights the trade-offs to be considered while time in ensuring the identity of the parties to the deciding whether to make the transition to a title transaction than the physical location and attributes system, as well as the strategies available for doing so. of the land. Under registration of tides, the register itself serves THE PURPOSE OF REGISTRATION as the primary evidence of ownership. It is commonly AND THE DIFFERENCE BETWEEN identified by three attributes, namely (i) the 'mirror' principle indicating that the situation in the registry is DEEDS AND TITLES an exact reflection of reality; (ii) the 'curtain' principle, The key function of a land registration system is to implying that anybody interested in inquiring about facilitate land transactions by placing rights in land the tide status of a given property will not have to on public record. While there are many variations of engage in a lengthy search of documents but can both, it is useful to consider the key elements of deed rely on the evidence from the tide registry as being and tide registration. definitive; and (iii) the 'assurance' principle according -~~ INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION to which the government will indemnifY for damages payment of such compensation. By contrast, under a incurred as a consequence of errors in the registry. deeds system, it is B's responsibility to investigate the To illustrate, if, under title registration, A veracity of Xs ownership claims and C will be abk to fraudulently sells a piece ofland (which actually belongs demand restitution of the property from B, implying to C) to B who purchases in good faith, B becomes that B will incur the loss. In other words, under a the rightful owner and any claims by Care extinguished deeds system, the cost of acquiring information about as soon as the sale is registered. The only recourse open the ownership status of a particular piece ofland has to C would be to demand compensation, but not to be incurred by the purchaser (something that may restitution of the property, from the state which in lead to a less than optimum amount of land market turn has the option to sue A. The need to ensure that activity) while under a title system the state guarantees the responsibility taken up by the state can be met is this information. one of the reasons why title registration systems are One common misconception is that the Torrens normally associated with a guarantee fund to facilitate system of registration of title guarantees boundaries BOX 5.1: THE ENGLISH SYSTEM OF TITLE REGISTRATION In Britain, no system of land administration existed before 1862 when the land Registry Act, which introduced a voluntary system of land registration, became the law. The attempt to introduce land registration, however, failed as the Act's high accuracy requirement for boundary demarcation was too costly to comply with, in addition to prompting a large number of conflicts among adjoining landowners. After successive legislation, it led to the concept of registration with general boundaries in 1875, the use of the National Ordnance Survey Map (a very low precision map) as the basis for mapping land registration in 1897. The land Transfer Act, which was passed in the same year, made registration of title compulsory in dealings with land in the County of london. Compulsory registration of title was extended to cover the rest of the country in a process whereby municipalities were progressively designated as areas for compulsory registration in case of sales. Since 1 December 1990, the whole of England and Wales has been subject to compulsory registration. Once an initial registration had been effected, all other transactions and dealings (for example, mortgages, creation of new rights of way, bankruptcies, leases, etc.) need to be registered. The Chief land Registrar acts as a title insurer, taking responsibility even if the error or omission has not arisen as a result of a mistake by the land registry (for example, they unwittingly registered fraudulent documents) although the size of indemnity will be affected by the level of carelessness by the person involved and the registry has the possibility to sue the person who committed the fraud. This guarantee covers not only the land register but also any search certificates and official copies of registers and plans issued in response to a proper application. With a total of about 20 million registered (and 3 million unregistered) parcels worth about £ 3000 billion, the annual volume of transactions amounts to 4.6 million, plus more than 11 million enquiries. With a total annual income of £ 410 million, indemnity payments in 2003-4 amounted to £ 2.65 million. Source: Based on Manthorpe 12004). TOWARDS GREATER TENURE SECURITY in the sense of guaranteeing lengths of boundaries Having in place certain simple procedures to reduce and areas of parcels. In fact, no title registration these costs can greatly improve the value of a deeds legislation expressly supports that notion and no case system. These include: law supports it. To the contrary, many more modern Compulsory registration. Registration may be title registration statutes expressly provide that made a condition of the validity of the deed by indefeasiblity does not extend to the lengths, bearing, providing in the law that unregistered deeds may not and areas of registered parcels. While boundaries that be admitted in court as evidence of tide. Documents can be identified with high levels of precision may which are not registered can then be safely ignored and be an advantage, they are not essential for a tide searching the deeds register, which can be automated registration system. In consequence, surveying and by having computerized ECs, will enable anybody mapping can be undertaken with cheap, cost-effective dealing in land to make sure that no material factor methods. For example, in England, which operates has been overlooked, thereby affording significant on a tide registration system, attempts to require costly protection against concealed conveyancing. high-precision surveys failed (see Box 5.1). Instead Parcel-based indexing. Basic deeds registration a system is used where many of the tide boundaries systems are normally indexed under the names of have not been measured but are what is known as the parties, rather than a unique parcel identifier, 'general boundaries'. These boundaries are defined with cadastral maps not being used in the system by pre-existing buildings, walls, fences, and hedges, and sometimes not even required as part of the rather than by fixed measurement. Cadastral maps documentation for the transaction. In virtually all then serve essentially as indexes to the land parcels countries where deeds systems operate satisfactorily, and related tides as well as field measurements. They it has been found that an improved system using may or may not have legal status but, in case they unique parcel numbers to identifY documents has less have, they are guaranteed to correctly show only the ambiguity than a system using changing names of relationship of the land to adjoining parcels and to grantees and grantors, making it much easier to review other features such as roads and public land. This has and assess all the documents required to determine the advantage that cadastral index maps can be the validity of a claim to ownership before entering compiled from the best available information, for into a transaction. Thus, while development of plans example, tax maps in some parts ofAustralia, or regular showing parcel boundaries, unique parcel identities, topographical ordnance survey maps in England. and cadastral index maps is a precondition for moving towards tide, it is likely to be required to improve deeds OPTIONS FOR IMPROVING systems irrespectively of whether such a transition is A DEEDS SYSTEM planned or not. As the cost of surveying increases exponentially with the level of precision required, much Before deciding whether or not to make the transition hinges on the extent to which existing information to a Torrens system, it is necessary to explore the can be used and the degree of accuracy required. options for improving the existing system. Based on Standardization and computerization. Computer- our earlier discussion, the cost which a potential buyer ization of tide abstracts as well as the links to cadas- has to incur depends on: (i) the completeness of the tral maps, parcel-based indices, and examination of information contained in the registry; (ii) the ease of documents described earlier can greatly improve the searching it; and (iii) the reliability of such information. quality of information provided by deeds registries · N DIA-L AN D P 0 L I C I E 5 F 0 R G R 0 W T H AN D P 0 V E R T Y R E DUCT I 0 N ·, and reduce the cost of searching them. A number of MAKING A TRANSITION TOWARDS countries, such as the Netherlands, South Africa, and TITLE REGISTRATION: the USA use these techniques to run highly effective OPPORTUNITIES AND CHALLENGES deeds systems that offer levels of protection that are equivalent to, or even higher than, those found in title Qualified or provisional title. QualifYing the title in a systems where these are not effectively implemented number of ways can help to make a transition easier. or where a government guarantee is not feasible. While In situations where immediate issuance of an absolute other jurisdictions, including Scotland and Hong title may not be possible due to constraints, qualified Kong, decided to convert to a title system even from or provisional titles may provide an alternative. Under a well-established deeds system, there are little this system, which has been adopted by a number of inherent advantages of one model vs the other and countries, land is brought into the title registration any decision will have to be based on the merits of the system with limited investigation and a provisional case at hand (Manthorpe 2004). title which is not conclusive, that is, that would still be Examination of documents to ensure compliance subject to any interests existing in the land which may with laws and regulations. The risks of a deed not not be disclosed in the title. A search behind the title being properly drawn can be further minimized by to preceding transactions is thus still necessary. However, requiring officials to check compliance with essential the advantage is that the land can be dealt with using rules and making the registry liable for any damages the same forms and procedures as land in a full-fledged incurred as a consequence of negligence and/or by title and the title is conclusive as to all matters occurring having insurance. Insurance against defects in title for after issue of the qualified title. Furthermore, with the a property being transferred is a common arrangement passage of time, or with the presentation of further in most states in the USA where title insurance evidence, the title can mature into a fully conclusive companies have developed private deeds registers and title through a process that can be defined in advance. which insure purchasers against losses as a result of a Title qualified as to boundaries. Apart from the defective tide. difficulties of ensuring a good tide, one of the factors Once a deeds system meets these conditions, the inhibiting the conversion ofland to title registration decision on whether to make the transition towards a may be the lack of an acceptable plan showing the full title registration system will hinge on three factors. boundaries of the land. The land may, for example, First, there needs to be sufficient political support to only be defined by general description or a sketch or effect the legal and institutional changes required for compiled plan. In such situations in Australia and New a title system (for example, for having only one land Zealand, in order to expedite the conversion of the institution). Second, in addition to the need to ensure land, the title has been issued without the boundaries availability of higher levels of capacity required for of the land having been defined to the usual degree maintaining a title system, a consensus about the of accuracy. The title bears a warning that it is qualified social desirability of incurring higher cost of title or limited as regards the boundary definition. This registration is also needed. Finally, an important warning can be removed from the title by the owner decision relates to the establishment of a guarantee lodging an acceptable plan to properly define the fund without which a titling system will be not boundaries. However, experience with this form of complete. After these issues have been resolved, a title has shown that many owners are satisfied with strategy for conversion has to be agreed upon. their title remaining in that state and do not see the TOWARDS GREATER TENURE SECURIT BOX 5.2: CONVERTING FROM DEEDS TO TITLES IN SCOTLAND: A LONG PROCESS In Scotland, land registers were established more than seven centuries ago to give citizens the power and protection of having their rights recorded and to reduce the scope of fraud. From 1876, so-called search sheets creating a parcel-based index map were introduced as a purely administrative measure that helped to greatly improve the security provided by the register as it was now possible to search by parcel rather than owner. Political pressure to introduce registration of title began shortly after title registration was introduced in England and a series of enquiries took place into the subject. However, 69 years passed between the report of the first Royal Commission on the subject (in 191 0) and the eventual passage of the land Registration (Scotland) Act in 1979 which provided conversion to title registration in a sporadic manner upon sales on a county by county basis, similar to the process adopted much earlier in England. Conversion has been a long drawn-out process and the last county was 'converted' to title registration only in 2003. Properties have to be converted to title only if they are sold or leased according to a well-defined process. 1 In 2004, exactly 25 years after the low had been passed, coverage of the land Register stood, with slightly more than 1 million registered titles, at 38.9 per cent, and 1.6 million titles remained to be registered. This implies that, in addition to the land register (of title), the Sasine register {of deeds) remains of great importance. Source: Registers of Scotland. need to incur the trouble and expense of preparing a land whenever it is the subject of a transaction, for fresh plan to more accurately define their boundaries. example, by requiring that the party lodging the deed Strategies for conversion. There are essentially three for registration be required to produce proof of a options to convert a deeds to a title registration good tide which will be examined by the registry. system, namely voluntary, compulsory transaction The disadvantage is that conversion will be slow and based, systematic, and automatic conversion. Under not all land will be converted, as demonstrated by the voluntary conversion, owners of land may apply to case of England and Scotland (Box 5.2). Systematic convert their land from deeds to title registration. conversion may be necessary if maintaining a dual Experience with this measure alone in countries such system for an extended period of time is considered as Australia has shown that this process will not undesirable. achieve full conversion. 2 Compulsory transaction- The key benefit normally cited in favour of title based conversion offers the advantage of converting registration is that it will cost less to transfer land or access land-related information. While systematic 1 See www. ros.gov. uk for details including the forms to be used evidence is scant, in Australia, conversion to title and a succinct description of the history and the 1979 Land registration is said to have reduced transfer costs by Registration Act. 2 In the states of Victoria and New South Wales (Australia), to go to the cost and trouble of applying for title registration unless voluntary conversion was the only conversion process for around a large-scale development was planned. It was finally concluded 100 years, with the result that there were still large amounts of that there needed to be some element of compulsion or automatic unregistered land. There was insufficient incentive for landowners conversion for the change to take place within a finite time frame. -j);f.o INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION about 50 per cent (Nettle 2006). These benefits need require additional resources from the government or to be compared to the extra costs, at the collective and the landowner. Careful quantification of these one- the individual level, that will not be incurred if a off costs incurred in the process of conversion, as decision is taken in favour of an improved deeds system. well as the increments in regular registration charges Legal issues. While improvement of a deeds system arising in a title system vis-a-vis those incurred in the can often be accomplished by making only minor current system, and the manner in which the burden amendments to an existing law, title registration re- would be distributed between the public and private quires new and more complex laws. The nature of the parties, is crucial before moving towards a title system. legal changes needed will require a more detailed un- Guarantee fund. One key advantage of a title derstanding of the different legal situations which title system for users is that it is usually associated with a registration may trigger, particularly the indefeasibility government guarantee of title that is supported by an provisions which can, for example, create a situation assurance fund. Such a guarantee will, however, require where an innocent landowner can lose title to the land a source of funds, either from levies on registration through a fraudulent sale to an innocent purchaser fees with a government guarantee in case funds for value. Other matters in which title registration may are insufficient, or directly from the government's differ from deeds registration are the extent to which consolidated revenue. possessory title (or acquisition of title through posses- Impact ofinformality. Title registration systems are sion alone) applies, the application of the doctrine of more dependent on registration, and failure to register notice (the binding effect of knowledge of unregis- has more serious implications in a tide registration tered interests in land), and the powers of rectifica- system than under deeds registration. The reason is tion of the register to reinstate pre-existing interests that, in a deeds system, registration can take place in land. The length of time required to introduce major despite a gap in the chain of title caused by the failure legislation is generally underestimated and a number to register preceding dealings. In a title registration of failed efforts at introducing title highlight that system, dealings must generally be made by the existing failure to dear the legal hurdles before other processes registered proprietor or someone claiming ownership are set in motion can easily jeopardize the whole through a registrable dealing executed by the registered enterprise of establishing title registration. proprietor. Informality on a large scale can render a Institutional capacity and status ofthe current system. title registration system ineffective, implying that Tide registration requires more complex registration individuals need to have an incentive to register procedures and more highly trained staff, both for transactions, that is, perceive the benefits to exceed first registration and for registration of subsequent the costs from doing so. Unless these conditions are transactions. As for legal issues, excessively optimistic satisfied, or there is at least a process that makes it projections can easily backfire. For example, Ghana reasonably certain that they will be attained, shifting made provisions for conversion from deeds to title to a title system may prove difficult. registration in 1986. During the first 15 years, lack The above discussion suggests that, compared to of resources and staff capacity, in particular to prepare an improved system of deeds registration, a title system and check survey plans, implied that only 11,000 titles will incur three types of additional costs. The first is were issued and no activity took place in most of the the cost of reaching consensus and the expense oflegal areas gazetted for systematic conversion. Depending drafting and possible delays associated with passing on the status of the system, conversion is likely to a law that would be required if one sticks with an TOWARDS GREATER TENURE SECURIT improved deeds system. The experience of Scotland inappropriate technologies are used, as illustrated by where, even though it started from a basis that was the example ofKerala. As Box 5.3 illustrates, attempts much superior to that encountered even in the most in Kerala to set up 'Torrens offices' that were to achieve advanced Indian states, it took decades and a large full integration between the revenue, registration, and amount oflegal drafting as well as commissions (Box survey systems remained unsuccessful because without 5.2) to make the transition towards titling suggests any improvement in the underlying technology the that these costs should not be underestimated. A cost of compliance remained too high. second cost element is the amount of extra scrutiny to Ensuring that these costs are commensurate to the be applied at conversion. The magnitude and incidence benefits and that their distribution is socially acceptable of these costs will depend on the conversion strategy (that is, the frequency ofland transactions is high and adopted; it can be large under a systematic conversion broadly distributed enough not to make this highly strategy but would be more modest, and distributed regressive) is a key element in any decision to move over a longer period of time, under a transaction-based towards a title system. In cases where the costs are conversion strategy. Finally, there is the incremental too high to be acceptable to the large majority of the recurrent cost associated with more demanding population or if there are too many unknowns, the standards as well as the operation of a guarantee fund fact that title registration is less robust, that is, can that is associated with a system of title as compared to easily break down if there are high levels of informality, deeds registration. While the magnitude of these costs may be an argument in favour of first making depends on the level of technology applied in the improvements in the deeds system to lower the cost system, it can be large in cases where rudimentary or before taking the plunge towards a title system. BOX 5.3: KERALA'S EXPERIENCE WITH 'TORRENS OFFICES' To illustrate that the high cost of title registration can well constitute a deterrent, it is instructive to consider the experience of 'Torrens offices' in Kerala. In two districts of this state, Kottayam and Eranakulam, so-called 'Torrens offices' were established in 1995. In these offices, it is mandatory for applicants to have their land surveyed before registration, irrespective of whether or not a subdivsion is involved. The FMBs needed are prepared by licensed private surveyors and their quality is attested by the survey department. These are then brought by the applicant to the subregistrar and, with the registered deed, proceeds to the village office where the applicable registers, called Pokku Varavu and Thandaper are duly updated. While this has helped to ensure integrity of land records, the response has been muted as the cost is very high and landowners felt that too much of the onus for ensuring accuracy of records was put on them. The fact that, despite containing many of the best practice elements identified earlier, the approach was eventually unsuccessful suggests that, unless the cost of updating land records can be reduced, for example, through back-office integration along the lines being attempted now in Karnataka, or by lowering survey costs via less demanding standards of accuracy and a more competitive survey sector, requiring individuals to make greater efforts to keep their records updated may end up pushing them into informality. Source: PriceWaterhouseCoopers (2006). DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIO the legal and institutional bases, to update both IMPLICATIONS FOR INDIA spatial and textual data, to adjudicate, and to handle The above examples illustrate that it is generally easier conflicts arising from the process. As all of these will to make the transition towards a title registration provide valuable lessons for other Indian states, it is system if the existing system is already well functioning. important to monitor this effort closely to draw Table 5.1 reviews the status of the different elements lessons that will have much broader relevance. of land administration for rural and urban areas in At the same time, the discussion suggests that, in the main states in India. states where there is less scope for taking such far- It suggests that, at present, there are only two reaching decisions, there should be no excuse for states, Maharashtra and Karnataka, that have delaying a determined move to make the necessary completed the integration of registry and records for improvements to the existing deeds system. Such rural areas, with three more (Gujarat, Rajasthan, and improvements will be needed irrespective of whether Tamil Nadu) on the way. In Uttar Pradesh and or not an eventual transition to title is envisaged. The Madhya Pradesh, records are fully computerized but goal should be to achieve the full functional integra- computerization of the registry remains in its early tion of the different textual and spatial databases used stages whereas in Andhra Pradesh, the registry is fully in land administration in the medium term. This computerized but land records continue to be handled would provide landowners with a certificate-which manually. Of course, a land registration system will can even be called a 'certificate of title' -that com- have to serve urban areas as well as rural ones and, in bines relevant and current information pertaining to addition, requires at least a rudimentary framework a plot (that is, ownership status, other current rights, of spatial information. Concerning urban areas, only and obligations in the property, under a range of Maharashtra has completed computerization of registered instruments that may include mortgages, property cards and Gujarat as well as Tamil Nadu rights-of-way, easements, caveats, liens, covenants, etc., are in the process of surveying urban areas and and a map that unambiguously defines the spatial computerizing property cards. As discussed earlier, extent of the land parcel and its relationship to the quality of spatial data in rural areas is most adjoining land parcels) irrespective of the government advanced in Maharashtra, with Karnataka and Tamil department maintaining the information. This Nadu as distant followers and hardly any activity in should be combined with a regulatory framework to: the other states. (i) provide dear, affordable standards for spatial records While this brief review does not imply that and ensure their integration with textual ones across embarking on a 'big bang' conversion to a title the country; (ii) require registry officials to perform registration system, as is being attempted by Andhra at least basic validity checks before registering a Pradesh, is completely impossible, it does suggest that document; and (iii) put in place mechanisms for doing so will be a rather complex, risky, and costly updating that are complete and sufficiently cost exercise. The reason is that, contrary to the case of effective. If this can be achieved, it would allow Scotland, where the necessary infrastructure was fully realization of about 90 per cent of the benefits from in place and had been operating successfully for a long a title system at a fraction of the cost, while at the same time, when the decision to shift towards title was made, time providing a more appropriate basis to decide significant cost will have to be incurred to establish on the desirability of a transition towards full title. TOWARDS GREATER TENURE SECURIT TABLE 5.1: Summary on status of modernizing land administration in selected states State Registry Computerization of Digitization of village Town and habitation computerized land records maps/FMBs surveys/property cards Andhra Completed Not operational < 10% ofVM vectorized; Data not satisfactory, thus Pradesh FMBs scanned little activity Bihar Piloting Data entry No activity No activity Gujarat Roll-out Fully operational, < 10% ofVMs vectorized Roll-out for computerization manual records banned of property cards starting; 0.9 million surveys (out of 2.3 million cards) in progress; GIS support planned Himachal Piloting Roll-out; manual records No activity No information. Pradesh not banned anywhere Karnataka Completed Fully operational, 50% village maps Limited coverage of city manual records banned vectorized, FMB scanning surveys, with survey started in pilot phasen pilot phase in 48 cities; No computerization Kerala Only indexes Data entry stage No activity for village No information operational maps; Piloting digitization ofFMBs Maharashtra Completed Fully operational, All VMs scanned, vectorized, Survey nearing completion manual records not and geo-referenced; All property cards banned Scanning of FMBs ongoing computerized and available via PCIS. Madhya Only indexes Operational, manual VM digitization at pilot Survey for one-third of Pradesh operational records not banned stage, problems with area cities ongoing since 1964; majority completed No information on PCs Orissa Piloting Data entry and Piloting of digitization of piloting VMs and scanning of FMBs Punjab Piloting Data entry Limited piloting No survey for urban centres No property cards Rajasthan Near Fully operational, No activity completion manual records banned Tamil Nadu Roll-out Fully operational in all VMs scanned; not Natham survey almost taluks, manual records vectorized; Piloting complete. Survey done in banned digitization of FMBs corporations; roll-out in municipal towns Uttar Piloting Fully operational in all No activity Pradesh taluks, manual records banned West Bengal Piloting Limited pilots Limited piloting of both No information (only Kolkata is urban) Note: Pilot stage implies that less than 10 per cent has been completed; roll-out means that 10 per cent-75 per cent has been completed; and near complete means that more than 75 per cent has been completed. Source: PriceWaterhouseCoopers (2006). CHAPTER Land Ownership Reform Although improving land administration along the (Bardhan et al. 2000; World Bank 2005). Empirical lines discussed above can have a significant impact studies have demonstrated the far-reaching and long- on the welfare of those who own land, it will at best lasting implications of initial differences in land provide indirect benefits to the large number of holding patterns on productivity, growth, and social households who are landless, without tangible assets, articulation (Nugent and Robinson 2002; Banerjee and may well be caught in a poverty trap. In India, and Iyer 2004) (Box 6.1). A number of channels policies to strengthen land administration will have through which high levels of asset inequality may to be complemented by considering opportunities for reduce growth have already been identified in the the poor to access land as a means of improving their literature. These include credit market imperfections asset status and livelihood. Land reform has long been in the presence of indivisible investments, for example, a key element in the Indian government's strategy. in education (Calor and Zeira 1993; Aghion and Bolton However, it is surprising that, in spite of the large 1997); wealth-induced limitations on households' amount of effort devoted to this policy, the amount ability to articulate their concerns in the political arena of analysis of its impact on a national scale (as compared (Bourguignon and Verdier 2000); high inequality to case study evidence) has been quite limited. We reducing local communities' willingness and ability use a long panel of households that spans the period to provide public goods that benefit all inhabitants 1982-99 to provide at least suggestive evidence on (Cardenas 2003); and a link between inequality and this issue and derive some implications for policy destructive tensions and social strife that can directly which, given that land is a state subject, can then be and indirectly undermine the basis for economic translated into policy by individual states. growth (Conning and Robinson 2002). There is also a large body of empirical literature on JUSTIFICATION AND BACKGROUND agricultural production which has shown that, due to the transaction costs involved in supervising hired Researchers and policymakers are increasingly aware labour (Carter 1984; Feder 1985; Eswaran and Kotwal that the distribution of productive assets and the 1985; Benjamin 1995), a farm structure based on associated economic opportunities will have far- owner-operated units is more efficient than one based reaching implications for long-term development on wage labour (Berry and Cline 1979; Binswanger LAND OWNERSHIP REFORM BOX 6.1: THE FAR-REACHING LONG-TERM IMPACT OF COLONIAL SETTLEMENT In India the impact of the type of settlement was far-reaching as it affected the nature of property rights and incentives to invest in land, the distribution of wealth, and the structure of political power. Differences in the security of property rights: In landlord areas, concentration of power in the hands of landlords made peasant property relatively insecure as any productivity-enhancing investments ran the risk of being expropriated by the landlord. In contrast, in the ryotwari areas, farmers had an explicit, typically written, contract with the colonial state, implying higher investment incentives. Incentives for public investment were reduced because the permanent nature of the settlement, together with the political power of zamindars, would have made it difficult for the state to increase rents. The colonial state had thus more interest in the prosperity of non-landlord areas which could be translated into higher rents, something that is indeed reflected in a much higher level of public investment in irrigation, railways, schools, and other infrastructure. Differences in the distribution of wealth: Under landlord systems, landlords were given the authority to extract as much as they could from the tenants, and, as a result, they were in a position to appropriate most of the gains in productivity-for which they could use judicial and other powers vested in them by the colonial state. As the 19th century witnessed significant productivity growth, the landlord class grew rich and inequality increased. By contrast, in ryotwari areas, the British raised rents frequently to extract surplus from tenants and differentiation within the rural population remained more limited. The distribution of wealth is important because: (i) it determines the size of the group with enough wealth to be able to make lumpy or risky investments to raise productivity; (ii) it affects the balance between owner-cultivation and sharecroppers which in turn has implications for productivity; and (iii) it makes it likely that the political interests of the rural masses would diverge from those of the elite, in particular the support for programmes to expropriate the assets of the rich. Post-independence, this implied that political correctness may have focused more on expropriating from the rich than towards trying to establish public goods (schools, water, electricity) while the rich, who had mainly been absentee landlords, focused on salvaging their wealth or transferring it to urban areas rather than improving productivity and living conditions in rural areas. eta!. 1995). Market mechanisms can, in principle, be that provide land access to the most productive relied upon to equalize the operational distribution producers can have significant social and economic of land holdings and thus help maximize production. benefits (Chau 1998; Carter and Zimmerman 2000). However, high levels of transaction costs limit the In situations where non-market forces have de- scope for rental markets to do so while credit market prived the majority of the population from acquiring imperfections, together with the collateral value of asset ownership, a redistribution of assets, even though land, reduce the number of transactions in land sales it would improve efficiency and equity, is unlikely markets and, even in cases where these have the highest to be brought about through market forces alone shadow value for land, fail to transfer land to the poorest (Zimmerman and Carter 2003). The potential pro- (Binswanger and Elgin 1988). To overcome these ductivity benefits from a more egalitarian distribu- obstacles, appropriate ways ofgovernment intervention tion ofland, together with an argument in favour of OIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTtO historical justice, have in many cases provided a spurious. Before 1972, few land in excess of the ceilings justification for redistributive policies as a means to was expropriated. Increased implementation efforts establish a foundation for an inclusive and broad- after 1972 resulted in redistribution of about 5.4 based pattern of economic development in some of million acres to 5.6 million beneficiary households. the world's most unequal societies. India is a prime Tenancy laws aimed to increase tenure security example of a case where colonial settlement policies of sitting tenants by registering them and often also have had a long-lasting impact on the level of public by establishing limits on the amount of rent to be investment and long-term economic performance paid or the scope for new rental transactions. Legal (Government oflndia 1966). This, together with the provisions across states differ from each other in the expected social benefits from such action, provides definition of a tenant, 1 the conditions for permitting the conceptual basis for the government to engage tenancy, the amount of rent that can be charged, and in land reforms. the type of property rights awarded to tenants. Key issues relate to the circumstances under which landlords can 'resume' rented land for 'personal cultivation' and LAND REFORM IMPLEMENTATION the amount of compensation, if any, that tenants have IN INDIA to pay in return for receipt of partial ownership rights. Given the inequality in the distribution of productive However, in most states, implementation was slow assets, especially land, which the country inherited and landlords were often able to resume self-cultivation at independence, land reform has occupied the centre with wage labour and evict large numbers of tenants stage in the Indian policy debate for a long time. Three to prevent them from gaining more permanent land main policies were used to implement land reforms rights in anticipation of such laws becoming effective (Mearns 1999). (Appu 1997). Abolition of intermediaries immediately after Table 6.1 provides summary statistics by state for independence is considered to have been highly the level of land reform implementation, measured successful. By the end of the 19 50s, almost all states as the share of total population or area that has been had enacted legislation to abolish intermediary transferred as a result of tenancy or ceiling legislation, interests that resulted in some 20-25 million tenants- in addition to the number of ceiling laws and their most of them in the zamindari areas of West Bengal total age. Overall, land reform constituted a major and Uttar Pradesh-becoming landowners. This part effort that resulted in the transfer of almost 10 million ofland reforms, which is essentially complete, shifted hectare-2.5 million ha under programmes to the power structure in rural areas and implied that redistribute of ceiling surplus land, and 7.35 million large tracts of forests and wasteland that had previously ha under tenancy legislation. The amount of land been owned privately came under state ownership. involved is much larger than that redistributed in Ceiling legislation provides a basis for the state to other Asian land reforms such as in Japan (2 million expropriate land held by any given owner in excess ha), South Korea (0.58 million ha), and Taiwan (0.24 of a state-specific ceiling and subsequently transferring 1 In a number of states including Andhra Pradesh, Bihar, it to poor farmers or landless agricultural workers. By Karnataka, Madhya Pradesh, Uttar Pradesh, Haryana, Punjab, contrast, passage and implementation of ceiling laws Rajasthan, and Tamil Nadu, the definition of'tenant' does not was very slow, giving landowners ample time to engage include sharecroppers, implying that the extent to which they in subdivision, much of which is alleged to have been will benefit from protection depends on local discretion. LAND OWNERSHIP REF TABLE 6.1: Share of households and area affected remained considerably below the state average, by land reforms in Indian states implying that the goal was to maximize the number (per cent) ofbeneficiaries. On the other hand, while some states Tenancy Ceiling (for example, Gujarat or Tamil Nadu) provided State Area Population Area Population beneficiaries with plots of about average size, in most Andhra Pradesh 3.48 0.75 8.34 3.81 states the fact that the share of beneficiaries remains Bihar 0.00 0.00 4.42 4.00 significantly below the area share points towards Gujarat 15.00 11.20 1.95 0.31 transfer of above-average plot sizes, as in Maharashtra Haryana 0.51 0.01 1.26 0.26 Himachal Pradesh 0.16 3.19 0.06 0.05 (27 per cent of area distributed to 10.7 per cent of Karnataka 15.38 5.29 1.71 0.30 population), Karnataka (15.4 per cent and 5.3 per Kerala 8.47 12.49 1.30 1.04 cent), Andhra Pradesh and Madhya Pradesh (3.5 per Madhya Pradesh 2.15 0.61 2.69 0.71 cent and 2.2 per cent to 0. 75 per cent and 0.61 per Maharashtra 27.01 10.68 7.74 1.08 cent of population, respectively). Orissa 0.15 1.43 2.24 1.28 Punjab 1.89 0.04 1.50 0.25 Even though tenancy reform also required gov- Rajasthan 0.00 0.16 6.63 0.75 ernment intervention, mainly in registering tenants, Tamil Nadu 3.65 3.23 2.47 1.24 the amount of bureaucratic effort involved is modest Uttar Pradesh 0.00 0.00 5.81 3.68 compared to ceiling laws which require appropriation West Bengal 6.41 10.80 14.91 19.73 ofland by the state and its subsequent redistribution. Total 5.45 5.35 4.41 2.27 It is thus not surprising to find that, with 4.4 per cent Source: Kaushik (2005). and 2.3 per cent, the shares of area and beneficiary households, respectively, remained significantly below million ha). In terms of total area distributed, this the figures for tenancy reform. Although some states puts India on par with Mexico which, in a much such as Rajasthan, Uttar Pradesh, Bihar, and Andhra more land-abundant setting, and over a much longer Pradesh transferred more land (6.6 per cent, 5.8 per period that started in 1917, managed to distribute cent, 4.4 per cent, and 8.3 per cent) under ceiling slightly more than 13 million ha (Deininger et al. than through tenancy legislation, the results seem to 2002). Although the magnitude of accomplishments be biased towards transfer of above-average sized plots is put somewhat into perspective if the share ofland of land, suggesting that even where it was possible and population are considered, land reform in India to acquire above-ceiling land by the state, overcom- was undoubtedly a major historical event, especially ing political pressures in the distribution of such land if the amount ofland distributed under abolition of may have been difficult. Only West Bengal, a state intermediaries is added to these figures. that ranks at or near the top for both measures and Comparing the share of households that benefited that is characterized by a formidable level of from tenancy legislation to that of the area transferred, grassroots-level organization (Rawal200 1; Banerjee points towards large variation across states in the total et al. 2002b) seems to have been able to avoid pres- amount redistributed and relative magnitudes. In sures to transfer above-ceiling lands in larger chunks. some cases, for example, Kerala or West Bengal, 12.5 A more detailed look at the time dimension of per cent and 10.8 per cent of the population benefited these measures allows us to draw a number of from a transfer of8.5 per cent and 6.4 per cent of the conclusions (Kaushik 2005). After a spurt of land land area, respectively and plot sizes for land transferred transfers in the 1970s and 1980s, progress has slowed INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION down considerably; in fact between 1995-6 and line with earlier studies (Lieten 1996; Rawal200 1). 2003-4, that is, for almost a decade, progress in Attention to land reform is widely seen as one of the awarding land rights to tenants had come to a complete key reasons for the remarkable political stability in standstill and the increment in ceiling surplus land West Bengal. Use of state level data over time suggests transferred during the period amounted to only that land reforms had a significant impact on poverty 10,800 ha. The latter represents about one-tenth of reduction, though not on increased productivity the land declared ceiling surplus, with the remainder (Besley and Burgess 2000)-something that is being tied up in litigation. This suggests not only that interpreted as indicating that the poverty reducing further progress in achieving redistribution of ceiling impact ofland reforms will be worth the cost in terms land could be slow-it would take almost 90 years to of productivity (Besley et al. 2004). At the same time, dispose of remaining ceiling surplus cases if the current there is little evidence on three areas, namely: (i) the pace is maintained-but also that, by clogging up the impact ofland reform on investment; (ii) its interaction court system and preventing it from quickly dispensing with other policies; and (iii) the longer-term impact justice in other urgent matters, the ceiling legislation of maintaining land reform legislation. may impose external effects beyond land rental markets (Moog 1997). 2 EMPIRICAL EVIDENCE ON THE The literature on land reforms in India and beyond IMPACT OF LAND REFORM is considerable (Warriner 1969; Thorner 1976; King To explore this issue empirically, we use a nationwide 1977; Haque and Singh 1986) and highlights that, panel survey of about 5000 rural households which where they were effectively implemented, land reforms were interviewed by National Council of Applied yielded positive results. 3 For example, a very effective Economic Research (NCAER) in both 1982 and drive of registering tenants in West Bengal after 1978 1999 to assess the extent to which cumulative land has been shown to have led to significant productivity reform legislation and/or implementation at the state gains, largely as a result of massive efforts at grassroots level affected changes in the accumulation of human mobilization of tenants (Banerjee et al. 2002b) in and physical capital and income levels for the same households over the 17-year period spanned by the 2 In addition to cases related to contestation by landlords, there data. The use of changes for the same households is an unknown number of instances where beneficiaries were washes out any time-invariant effects and we are allocated land but were either unable to establish effective able to control for households' initial conditions as possession or were subsequently evicted. A field survey to explore this issue in Andhra Pradesh pointed to at least 20 per cent of well as state-level spending on social programmes. At beneficiaries who were not able to access the property they had a descriptive level, we find that, even though they received, although the number of those who were able to file had less favourable initial conditions, households in court cases calling for their (re)instatement is much more limited. states with high levels ofland reform effort did better 3 Successive generations of lAS officers at the La! Bahadur Shastri than those in states where little land reform had been National Academy of Administration in Mussoorie have made undertaken. Econometric investigation not only a significant contribution to the area by conducting case studies of land reform implementation in almost all Indian states that supports this but also allows a comparison of the has thus far resulted in the publication of 10 volumes on the impact ofland reform to that of other social spending, subject that are replete with empirical evidence. assessment of the impact oflegislation relative to actual LAND OWNERSHIP REFO implementation, and inferences on how the impact and a slightly higher level of SC/ST population. The of such legislation has evolved over time. only exception is in the level of female education, Initial conditions in high land reform states were which was slightly higher in high effort states. worse than for those with low effort. Table 6.2 reports However, households in the former appear to have descriptive data on initial conditions and subsequent done better. Their disadvantaged initial position economic performance, separately for states with high notwithstanding, households in states with high land and low levels of effort in land reform implementation reform effort achieved significantly higher growth as defined by the share of households which received in the three outcome indicators-asset and income 4 tenancy rights. Doing so suggests that, if anything, growth, and educational attainment by offspring- states with high level ofland reform implementation than those in the low effort group. Rates of asset started the period with conditions that were much accumulation and income growth for households in less favourable than for those where implementation the high effort group (5.04 per cent and 4.47 per cent, effort remained low. In 1982, 'high effort' states had respectively) are significantly higher than those in significantly higher levels oflandless households (21. 7 the low effort group (3.74 per cent and 2.07 per cent). per cent vs 17.5 per cent) compared to 'low effort' The data on dependents ofhouseholds who completed ones, lower levels of per capita land availability for their educational process after the majority of land those who had land (0.37 vs 0.45 ac.), lower levels reforms had been implemented show that, in high- of initial per capita income (Rs 1423 vs Rs 1568), effort states, they completed on average 8.81 years of schooling compared to 8.14 for those in low-effort TABLE 6.2: Characteristics of households in. states. In all cases, t-tests of the mean difference states with low and high land reform effort between the two groups indicate statistical significance Total Land reform effort at 1 per cent. To explore this further, and to check Sample High Low whether land reform indeed had a positive impact Initial Conditions (1982) on economic growth and accumulation of physical Share of landless (per cent) 19.25 21.76 17.47 and human capital, we turn to econometric analysis. Land endowment p.c. (ha) 0.42 0.37 0.45 Per capita income 1508 1423 1568 SC/STs (per cent) 0.17 0.18 0.16 Human capital accumulation OBCs (per cent) 0.24 0.20 0.26 Table 6.3 reports results from regression of the level of Spouse's education 2.23 2.90 1.81 human capital acquired by about 5500 individuals Subsequent performance who completed their schooling after land reform was Growth of income (per cent) 3.07 4.47 2.07 Growth of assets (per cent) 4.27 5.04 3.74 implemented. The indicator for land reform is the Years education of kids 8.40 8.81 8.14 number of tenancy laws (column 1). To explore the Source: Own computation from NCAERARIS-REDS surveys. extent to which such reforms provided disproportion- ate benefits to the poor, we also interact this variable 4The statesin the high effort group include Gujarat, Himachal with initial levels of households' per capita consump- Pradesh, Karnataka, Kerala, Maharashtra, Tamil Nadu and West tion (column 2). 5 Other right-hand side variables Bengal. And those in the low effort group include Andhra Pradesh, Bihar, Haryana, Madhya Pradesh, Orissa, Punjab, Rajasthan, and 5 Substantively similar results can be obtained using actual measures Uttar Pradesh. of implementation or the equivalent variables for ceilings. DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTl include the mean per capita level of public spending TABLE 6.3: Land reform and human capital accumulation for education in the state during the 1980-95 period and initial household characteristics. Dependent variable: Years of education Land reform had a positive impact. Our estimates No. of tenancy 0.169*** suggest that land reform had a significant and positive laws (3.89) Tenancy laws 0.024*** impact on educational attainment by households' initial cons. (3.93) offspring. At the mean level, tenancy and ceiling laws Per capita education 4.752*** 22.010*** helped to increase by 0.34 and 0.23 years, respectively. spending (7.66) (3.08) Comparing this to the coefficient for initial landlessness Educ. spendg -2.513** suggests that land reform efforts were not quite initial cons. (2.42) sufficient to overcome the initial educational SC/ST dummy -0.671 *** -0.678*** (4.16) (4.20) disadvantage oflandlessness. A similarly significant BCdummy -0.324** -0.331 ** impact is obtained for educational spending. In fact, (2.43) (2.48) the point estimate suggests that, as one would expect, Male dummy 1.596*** 1.595*** the impact of educational spending was much higher; (12.07) (12.07) at the mean it is estimated to have raised attainment Head's education 0.332*** 0.331 *** (23.49) (23.36) by 2 to 2.5 years. Spouse's education 0.151*** 0.155*** To test for differences in the distributional impact (7.77) (7.92) of both interventions, for example, the possibility that Landless in 1982 -0.476*** -0.488*** land reforms may have had a particularly beneficial (3.12) (3.20) impact on the poorest groups, we interact the number Per capita 0.853*** 2.083*** consumption (7.59) (3.84) oflaws and the amount of educational spending with 1982 (log) initial per capita consumption. The results, reported -3.199*** -11.625*** Constant in columns 2 and 4, provide strong support to the (3.99) (3.13) hypothesis that educational spending did indeed No. of observations 5486 5486 favour the poor while they do not allow rejection of Rz 0.25 0.25 the hypothesis that land reforms were not targeted to Note: Regional dummies included throughout but not reported. Absolute value oft statistics in parentheses. ** significant at 5 per those with the highest need but instead have provided cent; *** significant at I per cent. slight advantages to the non-poor. This is consistent with the fact that under the threat of ceiling laws, landowners transferred land mainly to relatives and An important role of social status in determining access the better-off while tenancy legislation led landlords to education is suggested by the fact that members to keep those with relatively high levels of ability of an SC or ST are predicated to have educational who, even though they were not the poorest, could attainment that is lower by between 0.6 and 0.8 years make more productive use of the land than the original (0.3 to 0.4 years for backward castes, BCs) than landlord. To illustrate the magnitude of the effects, it comparable non-scheduled, tribal, or backward castes. is useful to compare them to the estimated coefficients There is also significant male bias in the provision of for initial household characteristics most of which education, with male individuals getting between point towards significant differences in educational 1. 56 and 1.61 years of schooling over and above what opportunities that are consistent across specifications. is provided to females. LAND OWNERSHIP REFO The positive coefficients on the head's and spouse's contribution of tenancy (ceiling) legislation is estimated levels of educational attainment points towards to be between 0.41 (0.34) percentage points for persistence rather than equalization of educational consumption and 1.25 (0.68) points for income differences across generations. According to the point growth. This would suggest that land reform estimate, one additional year of educational achieve- accounted for as much as one-third of the growth ment by the head or spouse would be predicated observed during the period. The interaction with to increase the level of their children's educational initial income is in most cases insignificant (not attainment by 0.30 and 0.15 years, respectively. The reported), suggesting that, overall, benefits from land positive coefficients for initial consumption indicate reform in India did not confer a relative advantage that children from rich households tend to receive to the poor? significantly higher levels of education; in fact, the First, both tenancy reform and ceiling legislation difference in per capita consumption between the are estimated to have made significant contributions individual at the 10 per cent and 90 per cent range of to growth ofincome, assets and expenditure. Estimated the distribution (7.56 as compared to 6.53) amounts overall magnitudes are, with the exception of to 1.11 years. Landlessness is also estimated to carry a expenditure in the case of tenancy reform, comparable significant disadvantage in terms of gaining access to those obtained earlier, suggesting that tenancy and to education; our point estimate suggests that ceiling reforms increased annual income growth by educational attainment of children of a household 1.04 and 1.06 percentage points, respectively, with which was landless in the initial period was lower by the figures for asset and expenditure growth being almost half a year. 0.84 and 1.09 and 0.05 and 0.78 points, respectively. One consistent message from the regression is that Growth effects implementation was more important in the case of tenancy as compared to ceiling legislation, something Positive impact on asset accumulation and consumption that would be consistent with the interpretation growth. To test whether land reform affected growth that ceiling laws send a signal to landowners rates of income, consumption, and asset ownership, even if they are not immediately implemented. Table 6.4 reports results from regressing growth of Implementation accounted for all of the impact of assets, income, and expenditure on the number and tenancy legislation on expenditure and roughly 60 per intensity of implementation of tenancy laws (results cent and 40 per cent of its effect on asset and income for ceilings and only the number of laws are very similar and can be found in Deininger et al. 2006). 6 7 The results suggest that household level data support Interestingly, use of household level data fails to support the hypothesis that land reform benefited the poorest groups; the an impact of land reform on consumption growth coefficients on interactions between the number ofland reform found earlier (Besley and Burgess 2000) but also point laws with initial landlessness or with initial levels of consumption towards a strong impact ofland reform on the other are consistently insignificant or negative (not reported). Such a two variables of interest. The magnitude of the failure of the landless to benefit would be consistent with estimated effect is quite large; at the mean (2.09 and difficulties in targeting the distribution of ceiling surplus land 1.14) of tenancy and ceiling laws in the sample, the to the most needy that are widely echoed in the literature or the notion that tenancy reform will benefit those who have already 6 Note that tenancy and ceiling legislation are highly correlated managed to have access to land through leasing while general (p=0.63) and including them together is thus not advisable. equilibrium effects may be limited. OIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIO TABLE 6.4: Land reforms and growth with other variables Growth of per capita Income Consumrtion Non-land assets No. oftenancylaws 0.003*** -0.002*** 0.002* (3.57) (3.25) (1.66) Tenancy reform benefits (share) 0.111 *** 0.112*** 0.139*** (5.09) (8.46) (5.58) Per capita social spending 0.035*** 0.028*** 0.033*** (6.33) (8.31) (5.31) SC/ST dummy -0.008*** -0.007*** -0.0 18*** (4.25) (5.79) (7.97) BC dummy -0.002 -0.000 -0.005** (0.90) (0.35) (2.55) Landless in 1982 -0.004** -0.002 -0.014*** (2.07) (1.60) (6.02) Land owned in 1982 (log) 0.004*** 0.001* 0.004*** (4.52) (1.77) (3.94) Household size (log) 0.064*** 0.001 0.062*** (27.91) (0.58) (23.30) Income in 1982 (log) -0.051 *** 0.005*** 0.011 *** (34.34) (5.75) (6.25) Non-land assets in 1982 (log) 0.001 -0.000 -0.057*** (1.11) ( 1.08) (71.77) Consumption in 1982 (log) 0.018*** -0.048*** 0.025*** (7.70) (34.12) (9 .41) Constant 0.209*** 0.286*** 0.202*** (15.72) (35.37) (13.28) Observations 3762 3795 3795 R-sguared 0.40 0.45 0.62 Note: Regional dummies included throughout but not reported. Absolute value oft-statistics in parentheses. * significant at I 0 per cent; ** significant at 5 per cent; *** significant at I per cent. growth, respectively. By contrast, for ceiling laws, any pre-emptive action by landlords will not be directly actual implementation is estimated to have accounted harmful to the intended beneficiaries of the reforms. for about 12 per cent of the impact for both income Social programmes have a similarly positive impact. and expenditure, and none for asset growth. This is The coefficient on per capita spending on other social consistent with experience, for example, in West programmes (based on data by Besley and Burgess Bengal, where tenancy laws are likely to backfire and 2000) is highly significant and of considerable lead landlords to pre-emptively evict tenants unless magnitude throughout. Comparing its magnitude to they are effectively implemented (Banerjee et al. the predicted impact ofland reforms suggests that the 2002a). On the other hand, the threat of expropriation increase in growth associated with land reforms was created by ceiling laws is likely to be present even in between one-fifth and one-third that of public spending cases where actual implementation is quite limited and during the period. For a more detailed comparison LAND OWNERSHIP REFOR of land reform with social spending, knowledge of TABLE 6.5: Impact ofland reform on income the administrative costs of implementing such a policy, and asset growth as well as the net gain in productivity, that is, a Growth of per capita comparison of the land utilization before and after Income Non-land assets being subject to reform, is needed. While this would No of tenancy laws 0.020*** 0.025*** be of considerable interest, it transcends the scope of (13.99) (15.49) Tenancy laws (or X) -0.001 *** -0.002*** this report and all that can be said is that the evidence age of laws (12.33) (15.27) points toward a positive impact on both income 0.026*** 0.020*** Per capita social spending growth and asset accumulation at the household level. (5.03) (3.33) Coefficients of initial household conditions are SC/ST dummy -0.007*** -0.016*** as expected. In addition to government intervention, (3.48) (7.11) initial individual characteristics were of high relevance BCdummy -0.001 -0.004* (0.29) (1.82) as determinants of subsequent growth performance. Landless in 1982 -0.005** -0.015*** For example, growth of income, consumption, (2.47) (6.62) and assets for SCs and STs were lower by 0.83, Land owned in 1982 (log) 0.004*** 0.004*** 0.68, and 1.79 percentage points, respectively, other (5.03) (4.64) things remaining constant-pointing to a general Household size (log) 0.063*** 0.060*** (27.67) (23.07) disadvantage of accumulating assets for this group. Income in 1982 (log) -0.051 *** 0.010*** This, together with a highly significant and negative (35.02) (6.28) coefficient of the landless dummy for asset Non-land assets 1982 (log) 0.002** -0.056*** accumulation (but not for other variables) could point (2.37) (71.88) towards the presence of credit market imperfections. Consumption 1982 (log) 0.014*** 0.020*** (6.24) (7.70) Although we find signs of conditional convergence, Constant 0.237*** 0.244*** the initial land endowment, as well as household size, (17.92) (16.24) contributed positively to subsequent growth of income Observations 3762 3795 and assets. R2 0.42 0.64 Note: Regional dummies included throughout but not reported. Evolution of land reform effects over time Absolute value oft statistics in parentheses; * significant at 10 per cent; ** significant at 5 per cent; *** significant at 1 per cent. Reform impact declines over time: As in most states the bulk ofland reform implementation was undertaken of insights. First, and consistent with earlier evidence, shortly after the promulgation of the respective laws, the number of laws is estimated to have a very it will be of interest to use our sample to assess whether significant and positive impact on income, and non- the effectiveness of land reforms in bringing about land asset growth. However, the interaction with mean growth and asset accumulation has changed over time. age of the legislation is significant and negative, Table 6.5 reports results from estimating the above pointing towards a tendency for the impact of such equations having the number of (tenancy) laws legislation to decline over time. The point estimate interacted with time. 8 Doing so provides a number for the land reform measure increases significantly 8 Note that very similar results are obtained if ceiling legislation if we allow the effectiveness ofland reform legislation is used instead (Deininger eta!. 2006). to decline with age of the applicable legislation at a DIA-LAND POLICIES FOR GROWTH AND POVERTY REDU rate of 1.5 per cent per annum, implying that the and rural labour markets is seen as a key obstacle to growth-effect of land reform will disappear between achieving a smaller-scale production structure that can 25 to 30 years after it had been first promulgated, in time, and with appropriate policies such as credit suggesting that it could become negative in the not for land purchase also help to address the problem too far future. of latifundismo (Rezende 2005). Similarly, in the An intuitive explanation for the negative time Philippines, a long drawn-out process ofland reforms effect from our regression is that the positive impact that entailed a very low limit on operational land of land reform legislation is contingent on actual holding, together with prohibitions on land leasing, implementation and distribution ofland. Given that is found to have negatively affected land lease as the laws have by now become largely ineffective in well as rural credit markets, thus severely reducing distributing any further land, it should not be too private investment in rural areas up to the point surprising to find that the direct impact from having where it has been accused of trapping many rural these laws on the books will be rather limited. While households in poverty not only because of investment the indirect impact of legislation could go either disincentives but also because of the inability of way, it is likely to be negative because indirect benefits, beneficiaries to increase their holdings even through for example, through increases in wage rates, are linked rental (Fabella 2003). to implementation. On the other hand, restrictions on beneficiaries' ability to use the received land and OTHER OPTIONS TO INCREASE on the willingness oflandlords to supply land to the LAND OWNERSHIP AND ACCESS market through voluntary land transactions will have BY THE POOR an impact even without any actual transfer of land. The magnitude of potentially negative supply effects Our results suggest that providing land access to the could actually be quite large, implying that this effect poor can have a significant and very positive impact, will over time actually come to outweigh the positive thereby vindicating the emphasis on redistributing impacts of land reform. land in post-independence India. They also support The negative coefficient is consistent with the the notion that land has a broader social and cultural experience from other countries in a number of relevance. At the same time, they imply that the legal respects. The most effective land reforms, in particular mechanisms that have been used to redistribute land those in Japan, South Korea, and Taiwan, transferred earlier are no longer very effective and that their assets during a relatively short period of time, after continued maintenance may have to be re-thought. which markets were allowed to operate, at least This suggests that there is scope to explore the potential within broad limits. By comparison, processes ofland of other mechanisms to make land available to the reform that were drawn out for a long time often had poor. In fact, a number of states have recently started counterintuitive effects. For example, in Brazil, labour initiating programmes that aim to go beyond the and land reform policies instituted in the 1960s traditional land reform paradigms in a number of contributed to tenant eviction and large-scale subsidy- ways, namely: (i) making legal assistance available to dependent agricultural mechanization that substituted the poor to help them overcome the shortcomings of capital for labour, giving rise to a highly concentrated past land reform legislation; (ii) providing grants or pattern of production (Binswanger et al. 1995). This loans that allow them to acquire small pieces of! and, has reached a point where deregulation ofland rental possibly in a group, through purchase in the market LAND OWNERSHIP REFO or through devolution of state land; (iii) increasing clarification of rights or, if combined with some women's inheritance rights and disseminating recent incentive compatible mechanism to encourage the legal changes so as to allow effective exercise of rights; parties to strike a deal, in disposing of pending cases (iv) giving recognition and/or full title to occupants for a wide range of issues. of land without proper documentation; and (v) empowering communities to demand more effective Regularization of government land delivery of services from local revenue departments. There is little doubt about the potential to regularize occupation on government land, either by giving Legal assistance outright ownership or through long-term and secure leases, to enhance land access by the poor. The largest Although detailed figures are hard to come by, with amount ofland is without doubt concentrated in the some estimating that land-related conflicts account country's more marginal areas and tackling this would for two-thirds of all pending court cases, the large be a key priority. In fact, in many cases of marginal amount ofland-related litigation in India is proverbial. lands, the state continues to be the largest landowner While part of these cases are related to land ceiling and focusing policy on land alienation between private legislation which has nearly 20 lakh acres tied up in parties without the possibility of giving the people dispute, others, especially at the local level, are more cultivating these lands the option of acquiring more straightforward but never get resolved because courts secure use or ownership rights is likely to have only a are overloaded, options to settle out of court are either limited impact. At the same time, considerable amounts not known or not available, or where court orders of land that have long been occupied but for which are not being implemented because those affected, documentation is either not available or is outdated often tribals, are ignorant about the enforcement often remain in long settled areas. For example, in mechanisms available. A number of states, including Andhra Pradesh, conduct of a systematic inventory Gujarat, Karnataka, and Andhra Pradesh, have in a number of districts during the 2003-5 period recently embarked on programmes to make legal provided information that subsequently allowed the assistance available in a number of forms such as: state to distribute a total of325lakh acres to 2.4lakh (i) systematic identification and pursuit of pending beneficiary households (Raju et al. 2006). It is likely land cases by law students; (ii) courses to familiarize that a systematic review will lead to similar results in members ofSHGs or other community groups with other states. key land-related legislation; and (iii) conduct of a systematic inventory of pending land issues to identifY and subsequently resolve all land-related problems Land purchase in a locality. In tribal areas, the latter will include not One of the key justifications for land reform has been only conflicts between individuals bur also those the insight that, to achieve sustainable poverty relating to lack of regularization and alienation of tribal reduction, government programmes should not only lands, forest boundary issues, and scrutiny of cases provide temporary relief bur transfer or help in the under land transfer regulations that have not been acquisition of assets. Agricultural land is no{ only an enforced. These efforts have elicited a very positive asset that can generate self-employment, utilize unused response, suggesting that similar approaches could family labour in the most productive manner to be of use in cases of systematic field survey and produce for self-consumption as well as marketable D I A- LA N D P 0 Ll C I E S F 0 R G R 0 W T H A N D P 0 V E R T Y R E D U C Tf. surplus in a way that utilizes the skills of the poor, and programme are reported to be in a much better can be transferred if the family wants to shift out of position to acquire land through leasing and thereby agriculture. However, in most states neither the improving their livelihood. While descriptive lowering of the ceiling nor a significant increase in the evidence from Andhra Pradesh suggests that in amount of ceiling land that can be made available is many cases the poor indicate preference for land not fensible. To the extent that distributing state lands as compared to a liquid asset (livestock) even if the is not an option, acquisition ofland through purchase latter is of higher value, the cost and benefits of such by the poor, often in a group, may be considered. a step need to be assessed carefully and compared to Some states, including West Bengal and Karnataka, those of alternative options of transferring (non-land) have decided to improve upon existing rural housing assets to the poor. schemes by acquiring land for subsequent distribution to the poor in small 'homestead and garden' plots. Giving dear tides The rationale for this is to provide households with One can argue that the government's strong emphasis an economically valuable asset that gives them an on numerical targets for redistributive land reform has opportunity to generate surplus and improve their diverted attention away from whether beneficiaries economic condition. Sample budgets from West are able to appropriately utilize lands transferred to Bengal suggest that a three decimal plot including a them under a variety ofschemes or whether they actually small pond would allow households to implement have appropriate title to it that would allow them to an intensive rotation of crops that could keep them do so. While nationwide figures on this phenomenon occupied for 150 days a year and, with an expected are not available, evidence from a large-scale survey annual net income ofRs 16,800, provide a return to in Andhra Pradesh 9 suggests that this is an important labour in excess of Rs 100 per day. Although initial issue; only 66 per cent of the plots included in the expenses are significant, the ability to obtain a return survey had proper title (patta) in the name of the of this magnitude could well make such a scheme an current occupant. The share of households without economically viable proposition. proper documents is significantly higher in the bottom The Government ofAndhra Pradesh, with support quintile of the per capita income distribution (71 per from the World Bank, has recently launched a similar cent) as compared to the top (59 per cent) suggesting scheme that aims to allow the poorest agricultural wage that clarifYing the land records situation has the labourers in rural areas to acquire up to one acre of potential of providing significant direct benefits to productive agricultural land based on direct negotiation the poor, in addition to the indirect benefits which with the landlord, with SHGs providing loan- they can derive. 10 financing and in many cases also significant technical assistance and supervision. While training requirements 9 are high and further empirical evidence on the incomes The survey, which covers more than 7000 rural households which were randomly selected from the state's population, was obtained and the economic feasibility of the scheme carried out by the Centre for Economic and Social Studies in is desirable, it appears that the programme, which is the context of the evaluation of the Andhra Pradesh District integrated into a wider effort to clarifY land issues at Initiatives Project (DPIP) and the Andhra Pradesh Rural Poverty the local level, generates significant and positive Reduction Project (RPRP). 10 external effects. For example, in project areas even Obviously, these benefits could be further increased through those who are unable to enter the land purchase an appropriately designed system of user charges. LAND OWNERSHIP REFORM··n~~- TABLE 6.6: Land price determinants Sales value Rental value OLS Fixed effect OLS Fixed effect Pucca patta 0.235*** 0.147*** 0.354*** 0.213** (7.53) (2.90) (6.19) (2.18) Land purchased 0.074*** 0.011 0.166*** 0.093 (2.88) (0.32) (3.46) (1.43) Assigned land -0.280*** -0.132** -0.333*** -0.304*** (6.33) (2.22) (4.21) (2.70) Quality very good 0.821 *** 0.751*** 1.212*** 0.974*** (13.58) (11.06) (10.75) (7.38) Land quality good 0.688*** 0.547*** 0.960*** 0.876*** (20.64) (14.12) (15.64) (11.77) Land quality fair 0.359*** 0.321*** 0.579*** 0.622*** (11.14) (9.01) (9 .81) (9.13) Has irrigation 0.311 *** 0.348*** 0.507*** 0.439*** (7.73) (8.14) (6.75) (5.33) lrr. functions 0.330*** 0.085* 0.351 *** 0.232*** (7.83) (1.90) (4.47) (2.69) No. of households 3615 3615 3716 3716 No. of plots 5929 5929 6081 6081 R2 0.47 0.31 0.23 0.21 Note: Regional dummies included but not reported. Abs. value oft statistics in parentheses. * sig. at 10 per cent; ** sig. at 5 per cent; *** sig. at 1 per cent. Source: Own computation from CESS2003 DPIP/RPRP survey. To quantify the potential impact of providing up- giving a proper title would, according to the fixed to-date tides, we use the fact that households in the effect estimates, increase land values between 15 per survey have on average 1.6 plots to conduct both OLS cent and 21 per cent. For a mean household in the and household fixed effects regression analysis of the lowest income quintile owning 2.06 acres ofland, this prices respondents think their land can fetch in will translate into a conservatively estimated monetary land sales and land rental markets, respectively. The benefit ofRs 14,000. regressions, reported in Table 6.6, perform very well, More importantly, clarifying the tenure situation pointing towards significantly higher land sales and of the approximately 10 per cent of sample households rental prices for plots that have a functioning source who cultivate assigned land (6 per cent) or land that of irrigation (between 35 per cent and 44 per cent, has been cleared and occupied by them (4 per cent) using the fixed effect estimates) are endowed with could almost double the value of their land assets, very good (75 per cent and 97 per cent) or good land illustrating that potential for systematic regularization quality (55 per cent and 88 per cent). Results also of the land tenure situation. The estimated impact of point towards a large impact of clarifying land records, clarifying the tenure status of assigned land on rental lA-LAND POLICIES FOR GROWTH AND POVERTY REDUC market values is more than double that on sales prices, users clear titles, ideally by systematically clarifYing not surprising in view of the fact that renting out land records and determining the status of assigned assigned land is likely to carry a substantial risk of lands, can greatly increase the asset endowment of the land loss and may thus not be undertaken even in cases predominantly poor households which do not have where it would be advantageous from a productivity clear documentation, presumably because they are point of view. It would imply that, in addition to the not able to afford the transaction cost of dealing with direct benefits to small landowners, a systematic the complex land administration system or due to process of reviewing land records, clarifYing ownership resistance in recognizing tenure of the land that they or use rights to assigned land, and establishing dear have been assigned at the local level. Finally, the much rules for their transferability will also provide greater security of women's inheritance rights afforded significant indirect benefits to the landless who, as a by the recent amendment to the Hindu Succession result, are likely to be in a much better position to Act (HSA) is a huge step forward that provides a lease in land. 11 tremendous opportunity to empower rural women The discussion allows us to draw four policy- in India. Evidence from other countries suggests that relevant conclusions. First, for a number of reasons, dissemination of innovative existing legislation is a that include political feasibility, the scope for very cost-effective way to increase awareness of these transferring much more land to the poor via existing provisions which, in turn, has a significant impact land reform legislation appears limited. Our results on household behaviour (Deininger eta!. 2006). As suggest that in such a situation, keeping land reform women will be able to exercise their rights only if legislation on the books may well have a negative they are aware of them, dissemination efforts should impact. A debate on the rationale for such laws and be high on the list of priorities, ideally as a first step alternative approaches, through market and non- towards a more comprehensive and systematic review market channels, of land access for the poor, is of land records and land ownership by the poor. desirable. This should include a thorough review of the scope for devolving government land outside of Strengthening women's rights forests to those occupying and using it. Second, innovative approaches to provide land access for the One issue which has often been neglected in the past most marginalized and disadvantaged by a number relates to land tenure security, in particular inheritance of states should be monitored closely so as to assess rights, for women. This is surprising because the their costs and benefits and the associated potential negative impacts of limitations on Indian women's for scaling them up. Third, although it will not ability to inherit land have long been recognized provide direct benefits to the landless, giving land (Agarwal 1994). The 2005 HSAA (see Box 6.2) demonstrates that policy change even on such a 11 Ordered pro bit regressions of rental market participation using controversial issue is feasible and the challenge the Andhra Pradesh data, similar to the ones reported below now is to disseminate these rules and ensure their (Chapter 7) suggest that having a proper title significantly implementation. If those affected are aware of legal increases the propensity to supply land to the rental market (not reported). This implies that, by enhancing their ability to access provisions and are confident that they will be land through rental markets, measures to improve land records implemented, the implications of this change, which could provide significant indirect benefits to the poor. is comparable to reductions in stamp duty if land is LAND OWNERSHIP REFOR BOX 6.2: THE 2005 HINDU SUCCESSION ACT AMENDMENT- A FAR-REACHING CHANGE The 2005 Hindu Succession Act Amendment (HSAAJ addresses a number of issues which in the past had negatively affected gender equality. First, it is clarified that all agricultural land is to be inherited according to the HSA, annulling earlier regulations in a number of states-for example, Uttar Pradesh, Delhi, Haryana, Punjab, and Jammu and Kashmir which together are home to more than one-sixth of India's population and gave primacy to male descendants. Second, sons and daughters are now both independent coparceners with shares of the property that cannot be willed away. In particular, daughters have rights equal to those of sons in residing and obtaining a share of the parental dwelling house. Finally, widows of a predeceased son or of a predeceased son of a predeceased son or a brother can get an equal share of the land and non-land inheritance even in cases where they have remarried, something that had not been possible earlier. The HSAA demonstrates that not only is policy change possible even on a contentious topic but also that such change can affect a huge number of individuals and with far-reaching multiplier effects. Even where the income derived from the land in question is small, a woman's ability to access (or the prospect of being able to access) her own regular stream of income will affect her bargaining power within the household and the way in which resources in the household are spent. Although no analysis of the impact of the HSAA is available, evidence from India points towards significantly lower rates (18 per cent vs 48 per cent) of intra-marital violence against women who own property as compared to landless ones (Panda and Agarwal 2005); and having a greater share of household income going to women has been linked to higher spending on children's education and health (Quisumbing and Maluccio 2003). The implications could be far reaching, and possibly affect not only women's ability to start enterprises of their own but also the magnitude of dowry which is traditionally viewed as a substitute for a woman's share in her parental property. A more detailed examination of these benefits and the extent to which they depend on dissemination would be highly desirable. registered in women's name in a number of states, comprise an increased number of non-agricultural are likely to be far reaching. Possible impacts would enterprise start-ups by them. Ensuring that include greater empowerment and bargaining power dissemination of these changes is adequate to make of women within the household but could also women aware of their rights is a high priority. CHAPTER I Land Lease Markets While one justification for tenancy legislation was land landowners to participate in the non-farm economy reform, that is, the desire to award property rights without closing off the possibility of returning to to tenants, a second motivation for adopting such rural areas; and (iii) credit market imperfections and legislation was that policymakers were concerned other restrictions rooted in custom or policy impose that, in an environment where land was virtually the limits on economic agents' ability to adjust through only economic (and social) asset and possible tenants land sales markets. had few alternatives to make a living outside of agriculture, monopolistic landlords would abuse their SOME STYLIZED FACTS ON LAND position to extract the maximum rent from their RENTAL MARKETS IN INDIA tenants. As we have already seen, such legislation no Rental market activity has declined precipitously longer transfers large amounts of land to tenants, therefore the rationale for its maintenance hinges on over time. India is one of the few countries where participation in, and activity of, rental markets has whether or not it provides effective protection to declined sharply since the early 1970s. As Table 7.1 tenants. Exploring whether this is still the case and illustrates, the share of households that reported whether there are any other impact of land leasing leasing in land has declined from about 26 per cent restrictions on the poor are the purpose of this chapter. to less than 12 per cent. 1 The decline in land market Economic theory indicates that well-functioning participation has been particularly marked in Punjab land rental and sales markets are critical to achieve an (from 53 per cent to 15 per cent), Bihar (from 40 per efficient allocation of resources and support financial cent to less than 10 per cent), and Haryana, West systems. They are expected to be of particular Bengal, Orissa, and Uttar Pradesh (all from about one- importance in situations where: (i) differences in third to around 10 per cent). More interestingly, even endowments of different agents are large, for example, though the main drop in participation coincided with high levels of landlessness coexist with large land a strong emphasis on implementation ofland reform concentration so that where land rental can help to redistribute land towards poorer sections of the 1 As only state-level figures were reported, the all-India figure population; (ii) the broader economy is undergoing was constructed using population weights from the 2001 Census rapid structural change and renting out land can allow for all the states. LAND LEASE MARKE .·. TABLE 7.1: Households participating in associated with a significant increase in the extent of land rental markets rental market activity. It is instructive to contrast this (per cent) decline in India with recent increases in such activity Year observed in China and Vietnam, both of which are State 1971 1981 1991 2001 now characterized by much higher levels ofland rental Andhra Pradesh 21.66 16.35 16.50 16.58 market activity than India. In Vietnam, the share of Bihar 39.76 21.25 7.98 9.28 households renting in increased from 3.8 per cent Gujarat 9.32 5.15 4.83 4.67 Haryana 37.35 27.71 17.92 13.10 to 15.8 per cent in the 5-year period between 1993 Himachal Pradesh 12.76 9.48 9.08 and 1998 (Deininger and Jin 2003). In China, the Karnataka 28.82 16.43 10.94 7.99 same variable increased from 2.3 per cent in 1996 Kerala 17.33 7.89 5.94 5.01 to 9.4 in 2001 (Deininger and Jin 2005). This Madhya Pradesh 21.02 15.76 12.58 10.96 Maharashtra 11.97 12.09 8.16 6.29 is surprising as, with much higher inequality and Orissa 32.24 25.60 21.58 19.40 landlessness in India than in these countries, the Punjab 52.95 23.46 17.50 14.64 potential for land rental markets to equalize pre- Rajasthan 8.92 10.07 7.92 6.89 existing inequalities in factor endowments is clearly Tamil Nadu 31.65 28.63 18.42 13.10 Uttar Pradesh 27.82 21.44 17.47 15.57 much greater in India. It has long been known that West Bengal 34.56 26.94 18.64 14.44 in China, land performs an important function as a India 25.81 18.45 13.27 11.57 social safety net (Burgess 2001), thus allowing the Source: NSS, various rounds. government to minimize spending on social safety nets and instead invest in infrastructure construction. In legislation in the decade of the 1970s, land rental addition, high levels of rental market activity in China market activity continued to decrease, though at a played a key role in facilitating growth of the rural slower rate, throughout the 1980s and 1990s. fu a non-farm economy, allowing those who do not have result, land-poor households are either unable to access a comparative advantage in agriculture to rent out land through rental markets or have to do so under their land to those who lack alternative opportunities informality, implying that they lack recognition and (Deininger 2003). all the benefits, from access to institutional sources The number of households affected is large. of credit, controlled rental rates, and protection against Although reported participation in India's rental eviction, that come with it. 2 markets has declined over time and is now low by Such a decline in rental activity is contrary to trends international standards, the number of people who elsewhere. The decline in rental market activity access land through rental markets every year is still observed in India is surprising in view of the fact that, enormous. Even according to the official figures, which virtually all over the world, economic growth has been are likely to suffer from considerable underreporting, the number of those renting in land is much larger 2 To the extent that households may not be willing to disclose than the number of those who have received land their participation in rental markets in violation of existing through land reform throughout the country's legislation, the figures included in this table establish a lower independent history. This has a number ofimplications, bound on actual rates of participation and land rental market activity. As it is reasonable to assume that the tendency to conceal namely: (i) in view of the discrepancy between actual rental market activity has been constant over time, the evidence and observed rental market activity, especially keeping on changes still remains credible. in mind the situation observed in other countries, a A-LAND POLICIES FOR GROWTH AND POVERTY REDUC large number of participants are at present either The role oftransaction cost andpolicy. Participation driven into informal rental transactions or are in land rental markets is not costless. Those interested completely rationed out of the rental market; (ii) even in participating need to acquire information on market modest changes in the policy environment for land conditions while actual participation requires the rental will have far-reaching impacts on the livelihoods screening of possible applications, negotiation, and of a large number of households; and (iii) allowing enforcement of payments. The existence of such those who rent land informally to be officially transaction costs will drive a wedge between the recognized is likely to have a significant and positive amount of resources expended by renters and that impact on their livelihood. A more detailed analysis received by landlords. This will expand the range of of rental markets is thus clearly warranted. producers who remain in autarky, thereby reducing the number of households able to gain access to land EFFICIENCY AND EQUITY IMPACT through rental as well as the amount ofland transacted OF LAND RENTAL: CONCEPTUAL through such markets. Anything that reduces these ISSUES AND INTERNATIONAL transaction costs will thus increase rental market participation, total output and productivity, and EVIDENCE social welfare. Policies, such as the restrictions on Endowments and ability as basic determinants ofmarket rental imposed in India, also increase the transaction participation. To understand the rationale for land costs ofland market participation, for example, they rental market functioning, let households be endowed force market participants to spend resources on with fixed amounts oflabour and land but different circumventing them or entering into informal levels of agricultural ability. As it is costly to supervise agreements. They would, therefore, have similarly hired labour in agricultural production (Binswanger negative impacts on social welfare. et al. 1995), it will be more efficient for households Urban evidence worldwide demonstrates significant to allocate their labour between self-employment in policy impacts. A number of studies have analysed farming on their own land and off-farm employment the impact of rent ceilings and other forms of policy at the going wage rather than cultivation using wage restrictions that increase transaction costs of land labour. It can then be shown that it will be those rental in urban contexts. In these settings, rent control with higher levels of ability and lower per capita served as a textbook example for policies that can endowments of land who will tend to rent in land effectively transfer resources in the short term but which they will use more efficiently than those from will be associated with inefficiencies in the medium whom they are renting. This implies that rental markets to long run (Arnott 2003). The key reason is that, by will transfer land to 'land-poor but efficient' producers fixing rents below their equilibrium level, rent controls and that will increase overall productivity in the reduce the supply of new housing or the maintenance economy (Deininger andJin 2005). Also, any increases of existing units by landlords who face an artificially in the wage for off-farm employment will increase reduced price. Rental restrictions would indeed transfer the supply of land to the rental market and thus the resources from landlords to sitting tenants when they amount of land transacted in them. This will result are imposed but also make access to rental property in a reduction of the rental rate and, in a risk-free more difficult for those who were not renting when environment, will make everybody better-off. the controls were imposed (Basu and Emerson 2000). LAND LEASE MARK With a constant or decreasing number of beneficiaries legislation, landlords will have an incentive to and an increasing number of new entrants who need artificially subdivide land and subsequently leave it to access land through now distorted markets, the idle. All this would imply that the impact of rental social cost of keeping land rental restrictions in place restrictions in rural areas will go far beyond the price is expected to increase over time. In practical terms, effects on which the attention of the urban literature this has led policymakers in many urban areas of the has been focused. world to realize that rent controls do not benefit the Land rental restrictions in India. Most Indian states poor and that a more flexible approach is needed restrict the legality of land leasing. This includes a (Arnott 1995). complete prohibition ofleasing for all except disabled However, rental restrictions could have larger impact landowners in the Telangana parts ofAndhra Pradesh in rural areas. Although empirical evidence from rural and in Bihar, Karnataka, Madhya Pradesh, Orissa, areas is more limited, a number of reasons would lead Uttar Pradesh, Himachal Pradesh, Kerala, and Jammu one to expect a larger and more significant negative and Kashmir. Other states (Gujarat, Haryana, impact of rental restrictions on the functioning of such Maharashtra, Punjab, Orissa, Rajasthan, West Bengal, markets than on those in urban settings. First, as owners and the Andhra part ofAndhra Pradesh) permit leasing of urban housing stock have fewer opportunities to but limit the amount of rent that can be charged or revert to self-cultivation (or cultivation with wage make leases heritable and non-transferable with some labour) than rural landowners, the supply of housing states (Assam, Gujarat, Haryana, Maharashtra, and to urban markets will be less elastic, and thus the Punjab) establishing a right for the renter to purchase negative supply response less pronounced than in the the property after some time. Even where leasing is case of rural land. Second, to the extent that rural rents allowed, the rents that can be received by the landlord are defined in kind-often as a share of output- are limited and tenants' rights are inheritable but contract terms in rural areas will be less flexible than contingent on personal cultivation, that is, they cannot in urban ones, limiting the scope for circumventing be legally transferred to others through sub-lease rental restrictions by adjusting rental rates. Third, arrangements by the original tenants. The rent ceiling the rights given to tenants in rural areas are often is defined either in terms of multiples of the land non-transferable and heritable but incomplete (that revenue or as a share of output, such as one-sixth of is, still requiring them to pay rent to the landlord), output in Gujarat (Kutch area), Maharashtra, and reducing both parties' incentive for making land- Rajasthan (in case the landlord does not provide any related investments. Limits on sub-leasing could have inputs), a fifth or fourth of output in Andhra Pradesh, a pronounced impact if, with generational change, Bihar, Rajasthan (if the landlord provides inputs), the original tenants are no longer able to farm the Orissa, Tamil Nadu, and West Bengal (unless the land they received in the most efficient way while landlord provides capital, manure, and seed in which their offspring may have taken up non-agricultural case the ceiling is half of the produce), and a third of occupations. Finally, for rural (but not urban) land, the crop in Punjab and Haryana. disincentive effects of wage-labour-based cultivation We use three measures to represent rental imply that land use or ownership arrangements will restrictions. The first is the number of tenancy and affect productive efficiency (Binswanger et al. 1995). ceiling laws enacted since independence, something Also, if tenancy laws are combined with land ceiling that also allows us to maintain comparability with the existing literature (Besley and Burgess 2000). At TABLE 7.2: Household characteristics by rental the same time, it is well known that counting laws in market participation in 1999 the absence of a measure for their content is at best a Rent-in Autarkic Rent-out very imperfect proxy for the constraints actually faced Basic Characteristics on the ground. And even if it were, the ability to Landless (o/o) 37.34 26.29 0.00 implement such laws has often lagged significantly Land endowment (ha) 1.27 2.02 2.87 Household size 6.91 6.04 5.54 behind the legislative zeal (which may itself be a Head's age 47.41 48.98 51.65 substitute for actual implementation). To deal with Head with at least 49.50 48.51 61.53 this, we complement the number of laws with the primary (o/o) share of households which actually benefited from Wealth and asset tenancy and ceiling legislation in each state, the endowments variable that was used earlier. In addition to providing Consumption exp. 1346 1549 2213 p.c. (Rs) a measure for the eagerness of a given state to Value of all assets (Rs) 33,839 465,68 62,466 implement land rental legislation, doing so is also Financial and 19.23 22.69 27.160 justified by the fact that in all states the ability to off-farm (o/o) transfer land that was received through either of these Farming and 21.67 20.91 13.26 means is highly restricted. 3 livestock (o/o) House and cons. 59.10 56.41 59.58 durables (o/o) EMPIRICAL EVIDENCE ON Number of observations 308 6366 802 EFFICIENCY AND EQUITY IMPACT Source: Own computation from 1982 and 1999 ARIS/REDS surveys. OF LAND RENTAL IN INDIA Land rental is importantfor the poor. Table 7.2 presents nature of their land market participation supports descriptive evidence on household variables the notion that rental provided opportunities for poor disaggregated by their type of land rental groups to access productive resources and thereby participation, that is, whether they rent in, rent out, improve their well-being. The value of all assets owned or remain in autarky. Comparing per capita land by those renting in land is at Rs 33,839, more than endowments across the three groups (0.20, 0.36, and 25 per cent below the average, compared to levels of 0.64 ha for renters, autarkic households, and asset ownership close to the mean for autarkic landlords, respectively) suggests that, as expected, households and about 33 per cent higher than the mean land rental provided an opportunity for land-scarce for those renting out. Clearly, it is the asset-poor who and labour-abundant households to gain access to benefit from the market-mediated land access. land. Note that, of those who rented in, 37 per cent Land rental allows pursuit ofoffform opportunities. were landless. Comparing levels of consumption and The data also illustrate that households who rent out asset ownership for households which differ in the land are endowed with higher levels of human capital (61 per cent have a head who had at least primary 3 As none of the Indian states permit sub-leasing of lands to education, compared to less than 50 per cent for which tenants had received permanent rights and most states also impose restrictions on transfers ofland received in the course autarkic and renting households) and are much more of implementing ceiling legislation, this is an indicator of direct likely to engage in salaried employment (30 per cent restrictions on the operation of land rental markets. as compared to 11 per cent for those renting in and LAND LEASE MARK 16 per cent for autarkic households). This supports includes structural factors affecting overall rental the notion that, as non-farm opportunities increase, market participation and for cut points marking the those with little advantage in agriculture will tend transition between renting out and autarky as well to move out, thereby making land available to as autarky and renting in, respectively, are reported provide human and physical capital assets to those in Table 7.3. 4 The regressions support the descriptive with lower levels of ability, and also give them non- evidence of rental markets opening up an important farm opportunities, with a possibility of gaining access avenue for land access to the poor. Landless households to (additional) land and improving their livelihood. are estimated to be significantly more likely to gain This is consistent with the notion that renting will access to land through the rental market. Similarly, provide opportunities to accumulate experience higher land endowments are estimated to increase and capital and can thus constitute an 'agricultural households' propensity to supply land to the rental ladder' that will allow landless people to accumulate market while endowments with family labour, land (Spillman 1919) and contribute to greater especially of members in the 14-60-year age group, diversification of income sources and occupational increase the propensity to rent in land. This supportS mobility in rural areas (Alston and Ferrie 2005). the notion that, by transferring land to those with Ability to rent in land increases returns to labour. higher endowments of family labour, markets will Estimation of a production function (not reported) improve the scope for gainful employment oflabour allows us to compute the marginal product oflabour in rural areas. Fears about free operation of rental from agricultural cultivation and compare it to the markets leading to 'reverse tenancy' that would push wage rate for casual agricultural labour. Doing this marginal farmers from the land, and give way to suggests that both males and females obtain a value concentration of operational land holdings are thus marginal product of about Rs 150 per day engaged unlikely to be justified. in agricultural self-cultivation. We are unable to reject Wealth barriers to rental market participation have the hypothesis of equality of such returns among males disappeared. Contrary to widespread belief which and females, implying that females are as productive assumes that markets favour the wealthy, the lack of as males. Noting that the casual wage rate in agriculture significance of households' total asset ownership as well as non-agriculture is less than Rs 50 per day, suggests that, consistent with descriptive statistics, it appears that land rental provides a very attractive land rental markets are not biased against the poor opportunity to improve household well-being even although the composition of a household's asset after subtracting the rental payment. portfolio does matter. 4 Right-hand side variables include the following: (i) a set of Econometric evidence household characteristics including total asset and land Rental markets allow landless and land-poor households endowments, age, education, household composition, and to gain land access. To assess the impact of restrictions dummies for caste as well as landlessness; (ii) a measure of agricultural ability that is derived from a stochastic frontier on determinants of land rental participation, we production function to make inferences on the impacr ofland estimate an ordered probit model on pooled data markets on productivity of land use; (iii) mean village income from 1982 and 1999 where renting out, autarky, and to assess the response of rental markets to higher non-agricultural renting in are coded as 1, 2, and 3, respectively. activity; and (iv) the three measures for rental restrictions as Estimated parameters for the main equation that discussed above. A-LAND POLICIES FOR GROWTH AND POVERTY REDUCT TABLE 7.3: Determinants of participation in land rental markets Policy variable considered No. of tenants recognized Tenancy laws Main equation Cultivation ability 0.248*** 0.247*** (2.63) (2.61) Landless dummy 0.623*** 0.746*** 0.762*** (18.09) (15.16) (15.24) Land endowment (ac) -0.012*** -0.029*** -0.029*** (4.63) (8.56) (8.35) Members below 14 years 0.054*** 0.054*** 0.058*** (6.22) (5.65) (5.96) Members aged 14-60 years 0.063*** 0.056*** 0.055*** (7.97) (6.42) (6.29) Head's age 0.021*** 0.026*** 0.026*** (3.44) (3.56) (3.59) Head's age squared/100 -0.025*** -0.028*** -0.029*** (4.34) (4.14) (4.24) Head has primary or above -0.148*** -0.117*** -0.135*** (4.59) (3.23) (3.67) Mean village income (log) -0.090*** -0.1 06*** -0.078*** (3.42) (3.56) (2.64) Total assets (log) 0.010 0.009 0.007 (0.59) (0.44) (0.32) Off-farm share in total assets -1.194*** -1.617*** -1.685*** (5.43) (4.63) (4.65) Lower bound (rent out to autarky) Policy variable -12.300*** -4.294* -0.121 *** (6.50) (1.90) (5.38) ST/SC dummy -0.200*** -0.138** -0.136** (3.85) (2.16) (2.10) OBCdummy -0.1 05** -0.040 -0.030 (2.49) (0.81) (0.60) 1999 dummy 0.527*** 1.426*** 1.359*** (8.73) (13.09) (12.51) Upper bound (autarky to rent in) Policy variable 12.697*** 8.816*** -0.023 (4.18) (2.64) (1.00) ST/SC dummy 0.166** 0.101 0.110 (2.52) (1.44) ( 1.57) OBCdummy 0.148** 0.166*** 0.170*** (2.42) (2.64) (2.72) 1999 dummy -0.239*** 0.143* 0.097 (3.41) (1.75) (1.19) Observations 11331 8820 8711 Log likelihood -4564.94 -3513.55 -3436.48 Notes: Robust z-statistics in parentheses; * significant at I 0 per cent; ** significant at 5 per cent; *** significant at 1 per cent; Constants and regional dummies included throughout but not reported. LAND LEASE MAR Land rental increases productivity of land use. To but also supply. 6 This is in line with evidence suggesting identifY the impact of rental markets on productivity, that landlords will be less ready to rent out their land we estimate a stochastic production frontier (not if regulations either imply that part or all of their reported) that provides a measure of agricultural ability property rights to land that is rented out may be lost for each household in the sample. This measure is or limit their ability to freely negotiate the amount then included in the regression and in all the three of rent to be paid. This result obtains irrespective of specifications, the coefficient on our measure of the variable chosen to empirically represent policy. agricultural ability is highly significant, suggesting Comparing coefficients suggests that the number of that it is households with higher levels of ability who laws passed had the least impact. This was followed gain access to additional land through the rental by implementation of ceiling and tenancy legislation, market. This implies that rental markets can be and is consistent with the notion that ceilings expected to increase the overall productivity ofland pose less of a threat than tenancy laws which, in use in the economy. 5 contrast to the former which are applicable only to Land rentalfocilitates off-form development. We also large owners, apply to all market participants find that higher levels of education, measured by the irrespective of the size of their holding. This is also completion of at least primary education by the head, true because enforcement of the latter is less politically increase the propensity to supply land to the rental controversial and administratively complex than that market due to higher opportunity cost oflabour for of ceilings. We also note that, even after adjusting for more educated individuals. Mean village income the factors discussed earlier, SCs, STs, as well as other increases the tendency to rent out, implying that as BCs are less likely to rent out land than others. This the level of income increases with overall development, finding may relate to lower levels of social capital and households will be more likely to move out of thus either less opportunity to find partners in rental agriculture and supply their land to the rental markets or to protect property rights to rented out market, thereby allowing those with higher levels of land. The highly significant 1999 dummy illustrates agricultural ability to increase their holdings and that supply of land to rental markets has increased income levels, similar to what is happening in China. significantly over time, beyond the expansion due The highly significant coefficient on ability also to higher village income. This result is robust to the supports our hypothesis of rental markets transferring choice of policy variable. land from less to more efficient producers, pointing Rental restrictions make it more difficult to rent in to the potential for land transfers through rental to land. Turning to the (upper) bound between autarky significantly enhance efficiency. and renting-in, the fact that coefficient estimates for Rental restrictions reduced supply ofland to the rental all policy variables are positive suggests that these market. We find that policy measures not only reduced restrictions depressed demand, making it more demand (as indicated by the upper bound equation) difficult to obtain land through rental markets. Across specifications, the level of significance is much higher 5 Results with ability are reported separately because the fact that ability is defined only for households or dynasties who 6 To interpret the coefficients in the lower bound equation engaged in agricultural production in either of the two periods marking the transition from renting out to autarky, note that a reduces the sample by about 2500 observations. The fact that positive coefficient implies expansion of the rent-out regime this does not significantly change our results provides an additional and better functioning of rental markets while a negative robustness check. coefficient suggests the opposite. DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIOI\J···· for variables relating to the intensity of enforcement; TABLE 7.4: Impact of restrictions on productivity in fact the number oflaws, while of the expected sign, and equity is not significantly different from zero at conventional Lower bound (rent out to autarky) levels, suggesting that such legislation affects market Share receiving tenancy rights -26.524*** outcomes only if measures are taken to implement (3.09) Tenancy rights*ability 32.764*** it. The magnitude of the coefficients which can be (2.60) interpreted as a rough measure of the difference in ST/SC dummy -0.134** impact between tenancy and ceiling legislation is (2.09) considerable. Also, backward and scheduled castes OBCdummy -0.042 or tribes are more likely to remain in autarky. Over (0.85) 1999 dummy 1.427*** time the size of the autarky area has decreased, that (13.07) is, land rental markets have become more active, partly Upper bound (autarky to rent in) offsetting the negative impact of rental regulation. Share receiving tenancy rights -0.109 While this is an encouraging sign, one should note (0.01) that, given the magnitude of the coefficients, almost Tenancy rights*landlessness 35.528*** a century will be required for the time trend to fully (2.82) Tenancy rights*ability 7.689 offset the impact of tenancy legislation. This, together (0.51) with evidence that circumventing such legislation is ST/SC dummy 0.092 normally easier for the rich than the poor (Yugandhar (1.31) 1996; Thangaraj 2004) would imply that expecting OBCdummy 0.170*** the passage of time alone to eliminate the negative (2.71) effect of tenancy regulation is unlikely to be a realistic 1999 dummy 0.144* (1.75) policy option. Observations 8820 Log likelihood -3502.61 Equity and efficiency effects of Note: Main equation omitted for space reasons. rental market participation Rental restrictions reduce equity. The discussion provides provide some interesting insights. The upper bound strong support for the hypothesis that land rental equation suggests that a major impact of tenancy restrictions constrained the ability of the poor and regulation on the demand side is to prevent access landless to acquire land that would help them to to land by landless and more efficient producers; in improve their income and social status. There may, fact, after accounting for both these effects, the however, still be an underlying trade-off between coefficient on the policy variable by itself is no longer equity and efficiency (Besley et al. 2004). To make positive but turns negative and highly significant. inferences on this, we interact the policy coefficient One explanation consistent with this is that sitting with a dummy for landlessness or the level of tenants who already own land but are not necessarily productive efficiency. Key results from doing so (with the most efficient producers, especially when they the coefficients for the main equation omitted) are become old and their children do not want to continue reported in Table 7.4. Coefficients are broadly in farming (as was seen in the Philippines), benefit consistent with those reported earlier and in addition, from tenancy regulation at the cost of landless but LAND LEASE MARKET more productive producers. The latter are constrained THE GENDER DIMENSION OF by the transaction costs imposed through tenancy LAND ACCESS THROUGH RENTAL restrictions and are unable to effectively express their The gender gap in wage rates is large and significant. As demand in the market. already discussed, a key reason for the attractiveness Rental restrictions also reduce economic efficiency. ofland rental markets is that cultivation ofland (even Concerning the impact of tenancy restrictions on if only on rented land), can provide higher returns efficiency, the upper bound equation suggests that, in to labour than what is offered in local markets for addition to rationing out efficient producers on the wage labour, in addition to reducing the risk of being demand side, and contrary to what is intended, tenancy unemployed. If, as the literature suggests, rural wage regulations encourage more efficient producers to labour markets are characterized by high levels of supply their land to the rental market, possibly because gender discrimination, that is, wage differences that they are in a better position to sidestep the transaction are not related to differences in the productivity of costs imposed by this intervention. As tenancy men and women in agricultural production, the restrictions make it more difficult for them to gain ability to access land through rental should be access to additional land that would allow them to particularly attractive for women. Table 7.5 compares make best use of their skills, producers with high wages for unskilled agricultural labour between men levels of ability are estimated to supply land to the and women in key Indian states. It suggests that wage rental market. This suggests that tenancy restrictions rates for men are about one-third higher than those can make the economy worse off and that the gains associated with their removal should be more than sufficient to compensate possible losers. The coefficient TABLE 7.5: Gender gaps in agricultural wage rates for the policy variable in interaction with ability Wage rate (Rs/day) suggests that, in states where tenancy restrictions are Male Female prevalent, producers with low levels of ability are Andhra Pradesh 45.36 33.94 significantly less likely to rent out their land. This Assam 48.14 49.06 would imply that, where such restrictions exist, Bihar 40.05 33.42 possible efficiency gains from low-ability producers Gujarat 44.46 40.32 moving out of agriculture and making their land Haryana 74.14 59.00 Karnataka 43.22 31.92 available to those with higher levels of ability are less Kerala 101.36 60.93 likely to materialize. This is consistent with descriptive Maharashtra 43.26 28.37 evidence from much smaller samples in individual Madhya Pradesh 32.30 27.42 states. 7 Orissa 41.00 30.79 7 Punjab 68.53 59.13 In a small survey in Karnataka, 94 per cent of respondents Rajasthan 54.50 55.51 answering definitively stated that existing tenancy restrictions Tamil Nadu 54.24 29.33 harm the landless; 91 per cent of respondents answering definitively stated that the existing tenancy restrictions harm Uttar Pradesh 47.51 44.57 landowners; and 38 per cent of respondents answering West Bengal 41.75 32.02 definitively reported that at least one farmer in their village keeps All India 46.35 33.70 land fallow rather than renting it out because renting may lead No of observations 2,322 1,699 to the loss of such land (Hanstad eta!. 2006). Source: Own computation from NCAER (1999), REDS Survey. .lA-LAND POLICIES FOR GROWTH AND POVERTY REDUC for women, a difference that is statistically highly one-third lower than those received by males significant in all cases. (column 1). This gender gap persists even if other factors are Economic development alone will not narrow the accounted for. To make the case that the observed gender gap. To assess whether such large differences differences in wage rates are indeed a sign of discrimi- are just a temporary phenomenon that will disappear nation, rather than due to pre-existing differences in automatically with economic development, we interact ability or other factors, it is necessary to estimate a the male dummy with the mean level of income in wage equation that corrects for selectivity. Table 7 6 the village (column 2). Results suggest that, contrary reports the results of using a Heckmann 2-stage to what one would expect, the difference between estimation to do so. One notes that selectivity is male and female wages tends to increase as the village indeed an issue with the probability of participation grows more affluent. It is also interesting to find that, in agricultural wage labour markets being significantly everything else being equal, wages are lower in villages higher for scheduled castes and tribes, male indi- where land is distributed in a more unequal fashion, viduals, and landless households while-consistent as measured by the Gini coefficient ofland distribution with expectations-wealthier households and those (column 3). with higher levels of education are less likely to Self-employment (including on rented land) increases participate. Even adjusting for these factors, we find opportunities for women. While differences in human that females obtain significantly lower wages-about capital or physical ability between men and women TABLE 7.6: Determinants of daily wages for casual labourers in India Specification (1) (2) (3) Male dummy 0.331*** -0.087 1.848*** (37.23) (0.58) (9.89) Male* village income 0.043*** -0.261 *** (2.79) (9.74) Age (log) 1.863*** 1.865*** -0.057 (9.97) (9.99) (0.38) Age squared (log) -0.263*** -0.263*** -0.104*** (9.83) (9.84) (11.78) Agricultural worker -0.1 04*** -0.1 03*** -0.0 17* (11.82) (11.69) (1.78) SC/ST dummy -0.016* -0.016* -0.028*** (1.67) ( 1.68) (2.85) Other backward -0.029*** -0.028*** 0.022 caste dummy (2.93) (2.87) (1.59) Village average 0.048*** 0.018 0.040*** income (log) (5.46) (1.33) (2.58) Inequality of -0.126** village land (Gini) (2.39) (contd.) LAND LEASE MARK Table 7. 6 (continued) Specification (1) (2) (3) Selection equation (Determinants of participation in wage employment) Age (log) 13.158*** 13.158*** 13.152*** (32.40) (32.40) (32.39) Age squared (log) -1.923*** -1.923*** -1.922*** (33.06) (33.06) (33.04) Household size 0.005* 0.005* 0.005* (1.85) (1.85) (1.73) Male dummy 0.877*** 0.877*** 0.876*** (39.02) (39.02) (38.95) Years of education (log) -0.243*** -0.243*** -0.241 *** (20.89) (20.89) (20.65) SC/ST dummy 0.273*** 0.273*** 0.272*** (10.40) (10.40) (1 0.35) Other backward 0.058** 0.058** 0.060** caste dummy (2.25) (2.25) (2.32) Village average -0.145*** -0.145*** -0.144*** income (log) (6.15) (6.15) (6.09) Village level -0.235 inequality (Gini) (1.63) Dummy for landless 0.525*** 0.525*** 0.527*** households (16.42) (16.42) (16.47) Value of financial -0.1 09*** -0.1 09*** -0.1 09*** assets (log) (15.87) (15.87) (15.88) Value of agricultural -0.046*** -0.046*** -0.046*** assets (log) (11.89) (11.89) (11.88) Value of house and -0.386*** -0.386*** -0.385*** durable goods (log) (27.05) (27.05) (27.01) Value of non-farm -0.101*** -0.101 *** -0.101 *** assets (log) (19.04) (19.04) (19.01) Inverse Mills Ratio 0.049*** 0.049*** 0.048*** (5.41) (5.35) (5.27) No. of observations 29104 29104 29104 Notes: Robust z-statistics in parentheses. *significant at 10%; **significant at 5%; ***significant at I per cent. State dummies included throughout but not reported. could be a reason for observed differences in wage rates, is no difference in the productivity of male and female estimating a production function where male and labour in own agricultural production. This suggests female labour are entered separately (not reported) that the gender gaps estimated here are indeed a sign suggests that this is not the case; in fact we are not able of discrimination but also implies that the relative to reject the hypothesis that the coefficients on male benefits from providing women with access to land, and female labour are equal to each other, that is, there that is, the difference between what they could earn DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCT! from agricultural cultivation on such land-even Plan (Government oflndia 2002). 8 In fact, by dem- after the rental fee has been deducted-and their onstrating the significance and magnitude of these opportunity wage in casual agricultural labour, is impacts quantitatively, our results reinforce the significantly higher than that for men. This is consistent importance of taking action on this issue, especially with anecdotal evidence from Andhra Pradesh where in commercialized areas. a large number of women organized in SHGs found Replace rent ceilings with enabling regulations. land rental to be a very attractive opportunity and often Where tenancy is allowed but is subject to maximum used the group to obtain access to such an opportunity, rent levels and/or a minimum lease term, legislation possibly even paying more to the landlord than was is required to remove maximum rents or minimum allowed under existing rent ceiling legislation. Case length terms, or-in cases where this is not feasible studies suggest that a combination of women having politically-at least replacing these with what is better access to finance, as well as social pressure and reasonable and enforceable, given the local conditions. availability of alternative opportunities for landowners Attempts to impose rules that are not attractive to either has led to a significant increase in women's land rental of the parties will, therefore, encourage short-term market participation, with a consequent positive contracts or drive transactions underground, thereby impact on their ability to generate income. reducing incentives for longer-term investment, depriving tenants of any form of protection, and POLICY IMPLICATIONS making them unable to access bank loans and crop insurance. Instead of trying to enforce such rules, the Freeing land rental couldyield significant benefits. Our government should aim to reduce transaction costs analysis suggests that rental markets can make a and encourage long-term rental contracts that would significant contribution to productivity and equity and ·encourage investment and sustainable management, that the pro-poor nature ofland rental has improved in addition to allowing tenants to get access to the significantly over time as wealth biases that had ear- necessary institutional support. Options to do so lier characterized such markets have disappeared. include dissemination of rental options and Rental markets are more active in places with higher standardized contracts (which can be in writing, levels of non-farm activity, thus making an increas- especially if they are for a longer term). ingly important contribution to diversification of Explore options for making land reform permanent. livelihoods in rural areas. The fact that tenancy and Experience from other countries suggests that freeing ceiling laws are found to reduce rather than increase rental can yield significant benefits but also that land access by landless and more efficient producers political resistance from sitting tenants can make its is a cause for concern. Consistent with our earlier implementation difficult. Where beneficiaries from results on land reform, it suggests that, even in situ- past tenancy reform are not full owners, it would be ations where such laws may have led to significant prudent to explore market-based options that can social gains in the past, maintaining them could develop into an increasingly potent obstacle to pro- 8 ' .. .freedom in leasing of land, both "leasing in" and "leasing ductivity growth and land access by the land-poor, out" will help generate income for both lessee and lessor/ particularly women. This supports the desirability contractor. A legislation needs to be enacted to facilitate the land of eliminating land rental market restrictions that is utilisation by making land transactions easier and facilitating articulated in the government's most recent 5-year leasing and contract farming' (Government oflndia 2002, p. 528). LAND LEASE MARKET help to convert them into owners, for example, giving will have to be taken forward by individual states. This protected tenants an option to 'buy out' landlords, implies that action to liberalize rental markets is based on mutual agreement, with the possibility of likely to be taken first in situations where the potential government or N GOs providing guidance on 'fair' land for rental markets is already high. Further analysis of values and assistance in negotiating such agreements state-level data could help support the case for reform on a village by village basis, and possibly even in and identify specific ways of implementing it. obtaining access to sources of finance for implementing Furthermore, careful and impartial monitoring of such them. A programme with similar characteristics is efforts to assess their impact will be critical to generate already being contemplated in West Bengal and, if the momentum for further reform within a state and implemented, it could provide valuable lessons for to make the case for reform in similar settings. As our other states. results suggest that women could derive particularly CarefUlly evaluate state pilots. As in the case ofland large benefits from such a measure, careful attention administration, any initiatives to liberalize land leasing to potential gender-differentiated impacts is warranted. CHAPTER Land Sales Markets Policymakers have often been concerned that in rural markets are justified or what alternative policy areas where households are not able to fully insure measures could be considered. 1 against shocks due to credit market imperfections, distress sales may have a negative impact on both EMPIRICAL EVIDENCE ON equity and efficiency. In other words, farmers would OPERATION OF LAND SALES be forced to sell off their land, often to usurious MARKETS moneylenders or other unscrupulous persons, at Households who bought land significantly improved their bargain prices that are well below the productive value welfore. Descriptive statistics on household welfare of the land just to ensure their survival in the face of are presented in Table 8.1, separately for those who a shock. As they will not be able to re-acquire the land remained autarkic as well as those who purchased or through purchase once prices return to normal, this sold land. We find that over the 17-year period covered would leave them permanently landless, an outcome by our data, 15 per cent of households bought land that is likely to be undesirable both from equity and and about 8 per cent sold land. We note that on average efficiency points of view. the land sales market seems to have transferred from Indeed, historically, distress sales were a major those with higher endowments ( 1.9 ha on average for factor that led to the accumulation oflarge amounts sellers) to those with less land (1.32 ha for purchasers) of land by the powerful, though often not very and that 13 per cent of households which originated efficient, landlords and moneylenders. However, even from a landless dynasty 2 were able to make the though the extent, incidence, and impact of land sales are among the most hotly debated policy issues 1 One problem in all surveys dealing with land transactions is on land, most of the empirical evidence to support that, because parties who sold land may have dropped out of such arguments is based on case studies or surveys the sample, there may be considerable selectivity which may bias the results. In our case this danger is reduced by the fact of small samples that may not be representative for a that respondents in 1982 were randomly selected and that in larger population. Our data, therefore, provides an 1999, information on all the descendants of a given household opportunity to obtain evidence on the nature of such or dynasty has been collected. transactions in India. This should help to identify 2 We use the dynasty's land endowment as it may well be the wh~ther restrictions on the operation of land sales case that a specific individual or household had not yet received LAND SALES MARK TABLE 8.1: Household characteristics by market to acquire land for inheritance purposes, and the years participation status since the household became independent to control Market participation status for life cycle phenomena. Variables affecting the Purchase Autarkic Sale transition between selling of land and autarky and Dynasty land 1.32 1.46 1.92 autarky and purchase of land, respectively, include endowment (ha) whether the household is an ST or SC, something Landless dynasty 0.13 0.23 0.00 that should make land market participation (in Asset value 1982 16,101 16,440 15,695 28,399 28,822 contravention of government rules) less likely, the Asset value 1999 43,374 Income p.c. (Rs) 1982 1572 1432 1603 level ofinitialland inequality in the village which could Income p.c. (Rs) 1999 4063 2470 2439 increase the propensity to participate in markets if there Expenditure p.c. (Rs) 1982 1310 1202 1355 were an equalization effect or reduce it if credit market Expenditure p.c. (Rs) 1999 1909 1579 1724 considerations predominate and the incidence of No. of observation 892 4581 459 drought shocks which would be expected to increase Source: NCAERARIS/REDS surveys 1982 and 1999. land market transactions. The drought shock variable is defined as the number of times the area experienced transition to ownership of land. Most importantly, a drought shock, that is, had less than 50 per cent of we note that, even though their initial level of income the long-term level of precipitation during at least two and assets was not significantly different from the consecutive growing seasons. To proxy for the presence average, their level of assets and income in 1999, is of insurance mechanisms, we add two variables, more than 50 per cent above the mean while their namely the availability of a bank and whether the level of consumption is about 20 per cent above the employment guarantee scheme (EGS) was available average. Clearly then, those who bought land became in the village in 1982. In both cases, we interact these much better-off whereas land sellers who remained variables with the drought shock variable to assess the in the sample are not significantly worse off than the extent to which results can indeed be explained as a average household in the sample. consequence of consumption smoothing. To explore factors that might underlie these Land sales transactions are not efficiency reducing. phenomena, we estimate an ordered probit regression First, the positive and significant coefficient of the similar to what was applied earlier for the case of dummy for landless dynasties and the negative and land rental. Given that we look at the behaviour of significant dynasty land endowment in all dynasties over a rather long period of time, the right- specifications support the notion that land sales hand side variables included are slightly different. In markets did provide at least some opportunity for particular, some of the variables included in the main the landless and those with low land endowment to equation are the dynasty's land endowment, the total access land that had emerged from the descriptive value of initial assets, the number of unmarried sons statistics. At the same time, the positive and and daughters in 1982 as a proxy for the tendency significant coefficient on the level of initial asset endowments, together with the fact that the land in ownership but instead cultivated parents' land-which coefficient on the landless dummy is much smaller may have been inherited upon the death of the parents-under than what was observed in the land rental equation, another arrangement. Clearly, such a household cannot be suggests that the redistributive potential ofland sales considered landless. markets is more limited than that of rental markets. DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTIO TABLE 8.2: Determinants ofland sales SC/ST dummy 0.198*** 0.195*** market participation (3.47) (3.41) Specification Drought shock -0.130*** -0.132*** (4.55) (2.69) I II EGS in 1982 0.134** 0.196** Agricultural ability 0.171** 0.170** (2.27) (2.01) (2.38) (2.35) EGS 1982 shock -0.028 Unmarried sons in 1982 0.063*** 0.063*** (0.85) (3.33) (3.36) Bank access in 1982 -0.131 *** -0.232** Unmarried daughters in 1982 0.022 0.023 (2.61) (2.57) (1.00) (1.02) Bank in 1982 shock 0.039 Landless dynasty dummy 0.124*** 0.124*** ( 1.26) (2.89) (2.87) Observations 5930 5930 Dynasty land endowment -0.004*** -0.004*** Log-likelihood -3835.67 -3830.86 (3.66) (3.66) Notes: Robust z-statistics in parentheses; * significant at 10 per Value of total assets (log) 0.044** 0.045** cent; ** significant at 5 per cent; *** significant at 1 per cent. (2.34) (2.40) Expend p.c. 1982 (log) -0.029 -0.028 (0.78) (0.77) Still, the positive and significant coefficient of No. of years independent 0.006** 0.006** agricultural ability suggests that, in general, it was (2.06) (2.00) households with higher levels of agricultural ability Lower bound (sell to autarky) which were able to access land through the ability sales Village land Gini 0.911*** 0.912*** market although the minuscule magnitude of the (4.64) (4.61) estimated impact-the most efficient household in Village income growth 1982-99 1.095 1.189* our sample is 4 per cent more likely to purchase land (1.58) (1.69) than the least efficient households in the sample- SC/ST dummy -0.419*** -0.405*** also points towards the limited productivity-enhancing (5.40) (5.21) Drought shock 0.1 05*** 0.234*** role of land sales markets. Households with a larger (3.55) (4.46) number of unmarried sons aged 5 to 25 in 1982 were EGS in 1982 -0.150** -0.001 also more likely to purchase land. (2.14) (0.01) Shocks increase land sales activity. The positive and EGS 1982 shock -0.063* very significant coefficient of the shock variable in the (1.78) lower bound equation suggests that shocks encouraged Bank access in 1982 0.168*** 0.400*** households to sell their land, implying that in India's (2.65) (3.56) Bank in 1982 shock rural areas, insurance against large-scale and co-variate -0.085** (2.40) shocks of the nature captured by this variable remains Upper bound (autarky to purchase) difficult to come by. This interpretation, and the Village land Gini 0.138 0.149 implicit hypothesis that government-supported safety (0.94) (0.95) net programmes such as the EGS provide a substitute Village income growth 1982-99 -1.893*** -1.909*** for such insurance, is supported by the negative and (3.15) (3.15) significant coefficient on the EGS. The negative coefficient on the interaction between EGS and the an ST dummy when introduced separately from each shock variable supports this interpretation, suggesting other (not reported) casts at least some doubt on the that safety net programmes can indeed help to mitigate interpretation of the quantitatively rather large the adverse impact of shocks to involuntary land sales coefficient that can be attributed to the specific policy in India. Even though the coefficient on bank access restriction under consideration, which should after all in the lower bound equation is positive, implying that apply only to STs and not to SCs rather than a villages that had a bank in 1982 were characterized general stigma oflower castes. This is to some extent by higher levels of land market activity than those supported by the positive and also highly significant who did not, the interaction between bank presence coefficient on the SC/ST dummy in the upper bound and drought shocks is negative, suggesting that the equation which suggests that this group also faced EGS is not the only way to provide an insurance significant restrictions in participating on the demand substitute but that access to a sufficiently diversified side of land sales markets. network of bank branches can perform a similar role. Credit market imperfictions are important. The IMPROVING THE FUNCTIONING hypothesis of significant credit market imperfections OF LAND MARKETS is further supported by the negative and significant While it supports the view that there is reason to be coefficient ofbank access in the upper bound equation. concerned about distress sales, the surprising part of This is reinforced by the fact that higher levels ofland our results is that land sales markets are not inherently inequality in the village, which can serve as a proxy bad for the poor but instead offer the most promising for such imperfections, significantly increase land route for the landless to acquire land. This is in line market activity on the sale side but are not estimated with other studies on land sales markets in India. For to have any impact on land purchases. Results from example, a recent study from West Bengal finds that, the transaction cost equations also reveal that even in this state, that has arguably implemented economically more dynamic villages (as proxied by land reforms most ardently, land sales markets have, the mean rate of income growth in the village over since the late 1970s, transferred more land to the the 1982-99 period) reduce the transaction costs poor than land reform implementation (Bardhan and associated with purchasing of land but do not have Mookherjee 2006). Although the factors underlying an appreciable impact on land sales. this phenomenon, for example, the role ofland ceilings, Caste restrictions have some impact. Finally, the remain to be explored, it suggests that land sales coefficient on a dummy indicating whether a markets should not be neglected by policymakers household belongs to SC or ST should provide an and that possible links between rental and sales- indication of the extent to which restrictions on tribal where, for example, tenants in a free rental market will land sales do have an impact. Our results in this be able to acquire skills and, via longer-term contracts, context are mixed. First, the significant and negative eventually be able to acquire land ownership through coefficient in the lower bound equation indeed points purchase--could be an important path out of poverty. to a much lower incidence of land sales among SCs In this respect, a number of issues are of relevance. and STs. At the same time, the fact that we cannot Changes in land use. Changes of land use from reject that hypothesis of equality between an SC and agriculture to non-agriculture, especially but not only at the urban periphery are to be expected in a growing of individuals or groups to transfer their rights is highly economy and should be possible as long as they restricted, land acquisition will have to be resorted do not generate negative externalities. However, a to more often. This is true even in situations where number of states, for example, Karnataka, Maharashtra, no public purpose is involved and negotiation and Tamil Nadu prevent acquisition of agricultural between private parties may be more expedient and land by non-agriculturists and allow changes in land advantageous, than 'land taking' by the public sector. use only upon special permission. In practice, this This is especially relevant given that bureaucrats are precludes industrialists or other non-agriculturalland often ill-equipped to represent local communities and users from negotiating directly with landowners. that the Land Acquisition Act's (LAA) main intention Instead, they have to go through land acquisition by was to make it easy for the state to acquire land, the state which often acquires the land for a fraction rather than protecting the rights oflandowners and of its true value from farmers and then sells it on to providing them with appropriate compensation. While the final users at a huge profit. In the areas where this amending the LM to remedy these defects is an is an issue, landowners will be well aware of market important issue for policy, it is equally critical to conditions and there is no reason to prevent them define more clearly as well as strengthen land rights of from negotiating with private industry directly. This individuals and groups in areas where land acquisition can, in fact, be encouraged by eliminating land is likely, so as to allow more decentralized negotiation. restrictions such as the ones mentioned above, improving land administration in general, and opening PROTECTING LAND OWNERSHIP up land records to the public. AND ACCESS BY THE POOREST Subdivision restrictions. A number of states continue With the exception ofland ownership ceilings, land to maintain subdivision restrictions to prevent sales markets in India operate in a rather unrestricted fragmentation ofholdings below a 'minimum' size that fashion. However, the government maintains strict ranges from 0.5 to 1.25 ha. In rural areas, with little restrictions on sale ofland from tribals to non-tribals or no evidence for increasing returns to scale in to eliminate as far as possible the impacts oflandlessness agriculture, such restrictions are difficult to justifY from which are considered particularly undesirable in the a productivity perspective. Moreover, experience in case of tribals who in many cases rely on land for the India and other countries suggests that they are rarely main part of their livelihood (Government oflndia, effective in preventing subdivision bur instead lead to National Committee on the Development ofBackward such processes taking place in an informal manner- Areas 1981). However, procedures to ensure recovery greatly increasing the potential for land-related conflict of land from tribals who sold in contravention of and, in addition, making it very difficult for the actual regulations are difficult to enforce and have not cultivators (who cannot register the land in their name) prevented large-scale land loss by tribal populations. to gain access to credit. In peri-urban areas, the problem The importance of the issue is illustrated even by is even more acute as it prevents small farmers from official statistics from the land records which are selling their land in small pieces to benefit from high likely to leave out a large number of benami holdings, prices and encourages informal deals and corruption. that is, land in tribal areas that is held by tribals in the Eliminating such restrictions is, therefore, desirable. name of non-tribals. For example, in the scheduled Compulsory land acquisition. In an environment areas of Andhra Pradesh, more than half of the land where land rights are often ill-defined and the ability LAND SALES MARKET·.· is held by non-tribals and census figures point cases currently require the government to resort to towards a continuing decrease in the ratio of cultivators expropriation-with amounts of compensation to agricultural labourers in India's tribal areas. determined in what often seem rather arbitrary Consistent with our failure to find a dear impact of ways-constraints that had originally been intended restriction on land sales by tribals, this suggests that to improve the welfare of tribal populations may well in many cases such restrictions may be largely end up worsening it. There is now a growing current ineffective. The high cost of enforcing such a policy, all over the world to deal with such situations by and the near impossibility of recovering land once it giving rights to the groups which-as long as they has been sold, highlight the need to explore alternative follow transparent and democratic procedures-can options. In fact, other countries' experience with decide about the specific way in which land rights restrictions on land sales suggests that even in cases should be defined (including possible restrictions on where they would be justified, that is, where those land alienation) and allocated internally based on selling land do not receive a realistic value for their their own mechanisms and preferences. Given that land at the point of sale, such restrictions may be existing legislation such as the PESA already defines difficult to implement and, rather than avoiding land an institutional framework at the local level that has sales by the poor, may further depress the returns of many of these properties, it would be worthwhile to those who have no other choice due to imperfections consider that, instead of deciding in a centralized in markets for credit and insurance (Box 8.1). A and rather paternalistic fashion whether or not land number of considerations for a more flexible regime should be alienable, it would be better to gradually might be relevant. devolve this authority to the community level Prevent inefficient land loss by improving markets provided that such a decision is taken at the local for credit and insurance. There is little doubt that level in a transparent way, and with the possibility of prohibition ofland sales is a costly second-or even enlisting the support of pubic servants in its third-best policy and it would be more effective and enforcement. Having a debate on this issue, followed preferable to deal with the root cause rather than the by a dear policy decision would be highly desirable. symptoms. Both international experience and our Deal with the huge backlog ofland alienations in a empirical results suggest that the first line of defence quick and flexible way. Even if policy is changed to against productivity-reducing land sales should be allow for more flexibility and greater community effective implementation of safety nets that would control in transfer of land, a large number of land eliminate the need for individuals to dispose of their transfers from tribals that have already been land to cope with shocks in the first instance. Banks undertaken in contravention of the law will have to and/ or federations of SH Gs which can provide access be dealt with. This is in some way similar to instances to credit in situations of distress would perform the (already discussed) where systematic adjudication same function and should be encouraged. efforts are needed to update the spatial framework Emphasize (group) rights rather than prohibitions. and link it to textual records because of too many While wholesale prohibition of land sales will not irregularities in non-tribal areas. Any efforts to clarify always prevent socially undesirable land alienations, and enhance the land rights enjoyed by tribal people it may have the impact of preventing desirable land will need to be combined with a mechanism to transfers where groups could negotiate with outsiders resolve this backlog in a quick and comprehensive to obtain significant benefits. To the extent that such way. In a similar experience in Mexico (Box 8.2), INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION BOX 8.1: GMNG LAND RIGHTS TO GROUPS AS AN ALTERNATIVE TO PROHIBITING LAND SALES Restrictions on land sales markets can increase the costs associated with certain actions, but if the rewards from circumventing them are high enough, will not eliminate them. For example, owners who have no desire to farm tend to disregard the temporary prohibition of land sales in Nicaragua and circumvent it through long-term rentals with the promise to sell which, because of the associated insecurity, leads to them receiving much lower land prices. The danger of beneficiaries' undervaluing their land could be reduced through other means, and the goal of preventing small landowners from selling out in response to temporary shocks would be better served by ensuring that they have access to output and credit markets and to technical assistance, and by providing safety nets during disasters to avoid distress sales. One case where temporary restrictions on land sales can be justified is in the presence of significant knowledge gaps. This was the case, for example, in many transition countries in Eastern Europe where, after decades of collectivism, new landowners did not really understand the working of land sales markets and were thus in danger of disposing of a hugely valuable asset way below its true value, something that would have led to negative social consequences and re-concentration of land in the hands of speculators. To avoid this, many counties introduced moratoria on land sales, for example, for the first 10 years after land had been privatized, to prevent quick sell-offs at unrealistically low prices. In some transition and developing countries !for example, Albania and Mexico), but also in developed ones !for example, France or Germany) a permanent 'right of first refusal' is in place. This means that before selling land to an outsider, neighbours and other village members lor in some cases other government bodies) must be given the opportunity to acquire the land at the same price for some period, usually 30 or 45 days. This not only helps to allay fears of being bought out by outsiders but also makes the prices paid for land public knowledge and prevents land from being disposed of at fire-sale prices-in which case somebody can offer a higher price. A second reason why communities may restrict land sales is that, in closely knit (indigenous) communities, disposition of land by one individual can generate negative externalities for the whole community which has the obligation to care for him or her after the loss of land !Andolfatto 2002). This implies that giving communities the right to make a conscious choice by the group and the group has clear and transparent mechanisms for changing the land tenure regime, such restrictions are unlikely to be harmful. If traditional social ties loosen or the efficiency loss from the sales restriction becomes too high, groups are likely to allow sales to outsiders in some form. For example, if an outside investor requires land, it may be more efficient for the group to bargain as a whole and to jointly obtain the benefits accruing from such a transfer land possibly invest them in social goods) rather than having individuals do so. For example, in Mexico, communities were recently given the ability to decide-by majority voting-whether they would want to eliminate the restriction on sales to outsiders. Quite surprisingly, in a country that has about 10 times the per capita income of India, only about 12-15 per cent of the communities-largely those living in peri-urban areas where demand for land was very high and where often large chunks of land had already been sold informally-<:hose to freely allow sale of land by individuals. This suggests that they did perceive that the benefits from maintaining land internally enforcing) land sales restrictions for the time being exceeded the costs from doing so, especially in view of the fact that they would always be able to change this policy should the need arise. LAND SALES MARKET BOX 8.2: INGREDIENTS OF A COMPREHENSIVE PROPERTY REFORM- THE EXAMPLE OF MEXICO From the 1917 revolution until 1992, Mexico implemented a large-scale process of land reform, which distributed more than 100 million hectares, or 50 per cent of the arable area, from large farms to the 'social sector' which comprised the so-called ejidos. Despite the physical achievements, the desired benefits did not materialize because of numerous restrictions on tenure security and transferability of land within the ejido. To overcome these, in 1992 the Mexican government adopted a comprehensive set of reforms that focused on modifications of the legal environment and institutional changes. The goal was to not only eliminate restrictions on the functioning of land markets but also empower communities to choose the property rights regime most suitable to their needs (that is, either communal, individual, or mixed), increase security of tenure and investment by issuing certificates of land ownership, and deal with the backlog of land conflicts that had been inherited from the past. To do so, a large-scale systematic programme of land regularization, called 'PROCEDE', was implemented which yielded quite impressive results at least in terms of the area certified. One key lesson from this experience is that, in view of the fact that many (informal) land transactions that were essentially illegal had taken place, an effective way of resolving disputes was a must. To allow this, a system of 42 land courts and one appeals court was created. In 5 years, these courts dealt with nearly half a million conflicts, thereby helping to eliminate a huge backlog of cases that had accumulated from the past. In dealing with conflicts in a systematic manner, these courts were explicitly instructed to seek out of court settlements and special mechanisms were put in place to ensure accessibility by the poor. Since small farmers who had been under the tutelage of local authorities for a long period of time would still find it difficult to ascertain their rights-especially to correct past irregularities-a special institution, the Procuraduria Agraria, was created to provide legal assistance to landholders, represent them in court dealings, perform an ombudsman function, and supervise the implementation of the regularization programme. Evaluation of the programme suggests that, in addition to the economic benefits that were associated with clarification of land rights and receipt of a secure and unambiguous document, beneficiaries often perceived the main impact to be in helping them to gain independence from local political bosses who had long used land as a means for exercising political power. efforts to improve the land administration system tenure security to tribal areas and populations. The had to be accompanied by an involved and rather goal should be to have these in place by the time the complex effort of adjudication-which included Tribal (Forest) Land Bill is passed so as to ensure special incentives for out of court settlements-to that-in contrast to many other progressive and far- resolve the pending disputes and conflicts. Similar reaching laws in India-this piece oflegislation can approaches to systematic conflict resolution will be swiftly implemented and have a real impact on need to be thought about for programmes to extend the ground. CHAPTER Towards an Integrated Land Policy While few Indian policymakers disagree on the system. Experience also illustrates that doing so importance ofland administration and land policies provides an opportunity to significantly enhance tenure for the poor, lack of systematic empirical evidence security, investment incentives, and access to credit, on the status ofland administration in different states while at the same time simplifYing the system to reduce and on the impact of specific policies on intended the transaction cost of securing and transferring beneficiaries has in the past often resulted in a highly, land rights. The transition towards comprehensive and unnecessarily, polarized debate on these issues. management ofland administration, although by no This report departs from the premise that the means automatic, is thus definitely within reach. opportunities provided by new technology and New data used for this study also suggest that the overall economic growth in India warrant a closer pessimism about the poor being able to benefit from look at empirical evidence to inform policymaking land (rental or sales) markets that had traditionally in these critical areas. This chapter first summarizes underwritten government interventions to overcome some of the more surprising findings from the review perceived imperfections in these markets-which may of empirical evidence before drawing out some possible have been justified in the past-no be longer be implications and options for policy. warranted. We find that rental markets work in favour of the poor, that restricting them will reduce KEY CONCLUSIONS FROM productivity and equity, and that unclear land records reduce the scope for land rental. Also, land sales EMPIRICAL ANALYSIS markets emerge as an important channel to provide Modernization of land administration has made land access. While prevention of involuntary land loss greater progress than is often realized by observers, by the poor through distress sales is a desirable goal, including policymakers at various levels. Although our analysis implies that restricting transferability of progress varies across states, which moved at very land is rarely the most appropriate measure to do this different speeds with no single state having developed and it must be complemented by other instruments. an approach addressing all the areas to be covered, One issue that emerges clearly from our analysis putting together the experience of various Indian states is that the synergies between land administration and allows clear identification of the steps to be taken to land policies are stronger than are often assumed. The put in place a more comprehensive land administration lack ofland rights in many marginal areas and outdated TOWARDS AN INTEGRATED LAND POLICY or contradictory records may be more important than policy on this issue. This makes it difficult to guide traditional market imperfections in reducing asset implementation of programmes in this area, to ensure endowments by the poor, the ability to use land in the that results obtained will be sustainable, and to make most effective way, and the potential for land markets states aware of sequencing issues such as preconditions to operate. Similarly, a merely technical approach to for moving from textual to spatial records. As a land administration that neglects critical policy result, criteria for allocation of Central CLR/CMLR dimensions, for example in terms of stamp duties, funds among projects or states are not well articulated may not be sustainable. The recent HSA Amendment and disseminated, implying that funds are often spent illustrates the potential impact of policy measures to in sub-optimal ways and the scope for ex post change the way rights are defined. With well-defined accountability remains limited. To deal with this it is land rights, elimination of restrictions on lease desirable to establish a high level technical committee, markets will have a positive impact and it will be with a long enough tenure to allow to taking a long- possible to put in place policies to help avoid distress term view. This committee could include practitioners sales by the poor and reduce the need to rely on from successful states, the National Informatics Centre ex post intervention to deal with this issue. (NIC) technical experts, academics, and policymakers In view of this, it is desirable for the government from the MoRD as well as the MIT and other to: (i) build on the successes in land administration ministries affected. but integrate the systems to eliminate duplication at A key task of this committee, to be accomplished the operational level; (ii) systematically clarify land in the short term, would be to draw on the existing ownership including resolution of disputes and a basic experience to formulate a policy on land administration spatial reference; (iii) expand beyond settled areas to through a participatory, consultative process. The ensure that the most needy will not be left out; and policy should be clear on the strategic vision and (iv) ensure that operational and policy issues are tackled should contain specific recommendations for the in tandem. This is a task of huge proportions which design of programmes and funding arrangements to existing institutions are not well equipped to handle attain this goal. Once in place, proposals for funding on their own. Experience from computerization by individual states can then be justified and judged of textual records suggests that, if the regulatory based on: (i) their fit into the overall strategy and framework is clear and enforced effectively, PPPs offer compliance with key elements (for example, the switch a major opportunity to scale up successful approaches. from manual to computerized records); (ii) the extent To take advantage of this opportunity, greater focus of to which they generate true public goods; (iii) financial the public sector on regulatory functions and evaluation contribution by beneficiaries or state governments of pilots to determine their potential is required. in line with their means; and (iv) proven capaciry for implementation or appropriate outsourcing OPTIONS TO FULLY REALIZE arrangements. This will also help to clarify the targets THE POTENTIAL OF LAND to be accomplished by specific interventions, which in turn can serve as a basis for allocation of future funds. ADMINISTRATION Beyond this immediate task, the committee would Establish a high level technical committee. Despite be responsible for defining and possibly also supporting considerable successes with computerization of textual pilots, overseeing their systematic evaluation, records, the MoRD does not have a well articulated ensuring that lessons are codified in regulations, and !NDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION. recommending ways to scale up successful experiences. capabilities to combine satellite imagery with existing This would include the tasks of exploring options village maps and other readily available spatial products for survey, recommending appropriate pilots, and to generate a basic cadastral index map would be a lower issuing regulations and manuals for specific situations cost option to provide a comprehensive framework, (from sporadic/systematic subdivision surveys to identifY gaps, and on this basis establish criteria to complete loss of records and unsettled areas), as well address spatial data problems in an affordable manner as investigating whether changes in the institutional and in the near term. This could easily be combined structure to ensure greater consistency and quality with preservation of existing village maps through of surveys (especially in urban areas) and to regulate vectorization and geo-referencing to link to the an emerging private sector would be appropriate. cadastral index maps. It could feed into a more systematic process of Pilot ways ofimproving textual and spatial records monitoring in a participatory framework (for example, for well-defined situations. In line with the ultimate workshops of technical experts, web sites, etc.) to goal of the land administration system, the purpose disseminate principles and good practices in land of piloting should be to establish processes that can administration at the technical level and to increase be scaled up rapidly to improve the overall (that is, awareness of recent developments that would prevent textual and spatial) records and formulate regulations individual states from re-inventing processes from that can help to do so, possibly by subcontracting scratch. to the private sector. To achieve this, pilots should Integrate textual records. Limited integration of be targeted to archetypical situations arising from records with registration records increases the risk of the nature oflndia's land records (that is, unrecorded inconsistencies among different types ofland records subdivisions, inconsistencies across records, decay/ that can give rise to land conflict. To deal with this, loss of maps, change of land use patterns, unsettled it is desirable to refine and roll out systems for back- lands). It would be ideal to set up a technical working office integration of records and registry, including group to steer this process with the goal of producing definitions and standards to ensure inter-operability results, to feed inro a broad debate on this topic, in a at the technical level. This should be combined with 12-18 month time frame. more systematic evaluations of the impact of specific Allow private sector participation in surveying, elements ofland record computerization on household focusing government on a regulatory role. Given the size welfare and decisions by other economic actors (for of the gaps in spatial data and the limitations that example, banks). This would give the issue the policy make it difficult for the public sector to address them attention it deserves, overcome bureaucratic resistance comprehensively, the almost complete prohibition to such a step, and provide guidance as to priority of private participation in survey is surprising and areas for attention. inconsistent with international best practice and Provide a basis for statewide spatial coverage. Large India's own experience in computerizing textual records. amounts of money have been, and continue to be, Efforts to change this should focus on: (i) providing spent on surveying pilots with ill-defined objectives a regulatory framework for the application of a the results of which are rarely subjected to rigorous range of survey methods with defined precision evaluation." On the contrary, pilots often make officials requirements; (ii) strengthening capacity in the lose sight of the forest for the trees. Drawing on India's private and public sectors; and (iii) revamping survey TOWARDS AN INTEGRATED LAND POLICY. processes, for example, shifting from paper-based to if desired, transfer land to investors rather than having electronic ones to reduce cost and make surveys more to go through the government and often receiving affordable. only very limited compensation. Provide options for a wider range of ownership patterns. Although expansion of survey coverage is OPTIONS FOR POLICY REFORM critical to ensure that poor people in marginal areas Reduce stamp duty rates and e>.:plore the scopefor replacing are able to gain secure land rights, in many situations, them with a land tax. There is little doubt that the award of individual title may not be the most high rates of stamp duty currently assessed upon appropriate option; indeed some observers have linked registering land transfers push people into informality such individualization with the break-up of traditional while reducing government revenue. Reducing these community structures and widespread land transfer. rates, which are very high by international standards, To prevent these, it is important to have a menu of is likely to be necessary to ensure the sustainability tenure options, including communal ones, available of any improvements made in land administration. and to allow groups to choose freely and depending To make such a step revenue neutral, it may be useful on their specific needs, with the possibility of making to consider combining it with an increase in the land the transition to individual holdings at a later stage tax for specific groups, possibly to be shared between if desired. states and local governments. While such a decision Complement restrictions on tribal alienation with will not be politically easy, it is likely to have a more flexible mechanisms providing them with property rights. profound impact on India's land administration system While there is little doubt that alienation of lands than a transition towards a title registration system. through distress sales is an extremely undesirable Eliminate restrictions on land markets. All over the outcome that should be avoided, increasing rates of world land rental markets allow rural dwellers to join tribal landlessness suggest that regulations are often the rural non-farm economy in a way that provides not effective in preventing the same. In the short term, those who stay back with access to additional the most promising means to prevent tribal land productive resources. Indian evidence shows that alienation is likely to be effective safety nets, something rental restrictions reduce equity as well as efficiency. that could possibly be combined with mechanisms It is thus desirable to: (i) make leasing legal where it for communities to have a greater say in whether or is currently prohibited and replace rent ceilings with not land should be transferable, for example, a right regulations to facilitate rental markets instead of of first refusal or community consent for sales. constraining them; (ii) allow transferability of land Providing tribals with real property rights, either by land reform beneficiaries at least through lease, and individually or as a group, would in the long term explore options for making the gains from such reform make a more important contribution to their permanent; (iii) drop restrictions on subdivision and productive development and help them to avoid on sale of land to non-agriculturalists which have distress sales. Therefore, the longer-term goal should little economic justification; and (iv) review legislation be to implement systematic programmes that would on compulsory land acquisition and, subject to the recognize tribal land rights and according to accepted prevention of undesirable externalities, allow farmers principles of policy resolve whatever conflicts exist as or their representatives to negotiate directly with and, a result of past alienations in contravention of the law. INDIA-LAND POLICIES FOR GROWTH AND POVERTY REDUCTION Develop state-specific roadmaps to improve landpolicy the land administration structure, it is important to and administration and monitor closely. Contrary to base such reforms on a broad policy dialogue to help the case of land administration where the need for set priorities, sequence implementation, and monitor change is widely acknowledged and a wide array of realization of the desired impact. While certain experiences can be drawn upon to identify at least strategies need to be state specific, there is considerable the first steps on this way, policy issues remain more scope for assistance by the centre and for comparing controversial. To ensure that any reforms are properly experience and learning across different states. sequenced and synchronized with improvements in APPENDIX R Policy Matrix Issue Action Timing 1. Benefits from computerization of textual records Expand computerization, integration, and use of textual are often not realized records to ensure full coverage The MoRD's policy on land administration is not clearly articulated. Task a high level technical committee, including practitioners, Short-term It is thus not effective in providing guidance to implementation academics, and policymakers with formulation of a policy on (ST) of programmes in this area and states are not aware of the land administration in a participatory, consultative process. The preconditions that need to be in place to ensure successful policy should be clear on the long-term strategic vision and implementation of even 'pilot' projects (for example, contain specific recommendations for the design of programmes on surveying). and funding arrangements to attain this goal. Also, as a result, criteria for allocation of central CLRJCMLR Require that proposals for funding by individual states be funds among projects or states are not well articulated and justified on the basis of standard criteria such as: (i) fit into disseminated, implying that funds are often spent in sub-optimal the overall strategy; (ii) generation of true public benefit; ways and the scope for ex post accountability remains limited. (iii) appropriate financial contributions by beneficiaries or other governments; and (iv) proven capacity to undertake the work. Principles and good practice lessons from computerization of Issue guidelines and make best practice more widely available. ST land records are not widely disseminated, especially at the Make funding of CLRJCMLR initiatives contingent on technical level. As a result, decision-makers at the state level are tangible progress towards best practice outcomes. Systematize often unaware of recent developments and lack strategic vision. monitoring of outcomes in a participatory process (for Although digitization of records has made considerable progress example, workshops of technical experts) to create peer pressure in most states, computerized records are not used in all of them, and debate solutions to common problems. suggesting that they may become obsolete, putting into doubt the long-term viability of this exercise. (contd.) Issue Action Timing Limited integration of records with registration records increases Refine and roll out systems for back-office integration of ST risks of inconsistencies among different types of land records that records and registry, including definitions and standards to can give rise to land conflict. ensure inter-operability of systems. Paucity of systematic information on magnitude of benefits Encourage ongoing and potential future evaluations of ST from modernizing different elements of land administration computerizing land records on households' and other (and the synergies between them) makes it difficult to give economic actors' (for example, banks') decisions. the topic of integration between registry and records the priority it deserves. 2. Coverage of spatial data remains spotty and vast mismatches Establish statewide cadastral index maps combining satellite with textual records remain but their extent is not known to imagery with existing data decision-makers There are wide gaps in the spatial data but the magnitude and Create index map by combining satellite imagery with existing ST/ type of such gaps is unknown and, as a consequence, no strategy spatial data. medium- for dealing with them in a systematic way (and according to Use this cadastral index map to identify the magnitude of gaps, term (MT) well-defined processes) exists. outcome from pilots to establish cost-effective ways of dealing Absence of a comprehensive spatial framework makes it with them, and overlay with other layers to establish possible impossible to define a strategy for improvement and measure cost-recovery strategies. progress towards it. Provide output-based Goi funding for such activities, possibly by redirecting resources from surveying pilots and equipment. Existing maps are in precarious condition and in many cases no Take up preservation of existing village maps through ST/MT longer usable. Initiatives to preserve them are not systematic vectorization and geo-referencing to link to the cadastral index and the products of doing so are not integrated in the process maps. Combine with scanning of FMBs where needed. of land administration and management. Develop low-cost mechanisms to reconstitute the spatial basis through desk work (using the redundancy between records) in cases where maps have been lost. Non-compulsory registration leads to inconsistencies between Make registration compulsory, exempting certain types of ST/MT different sets ofland records. High rates of stamp duty lead to transaction from fees. non-registration of land transfers, driving users into informal Explore the scope for reducing stamp duty, possibly in line arrangements. with a shift towards a higher level of land taxation (possibly exempting the smallest farmers) that is shared between state and local governments. (contd.) Issue Action Timing 3. Processes to verifY and update spatial and textual data Pilot and evaluate processes to establish such a link; adapt together are lacking regulations to allow large-scale application of successful ones The public's limited ability to access land records prevents errors Make the computerized records available on the Internet after ST from being detected and remedied and reduces confidence in they have been validated in a participatory manner. the system. Despite significant differences in cost and precision of surveying Implement pilots to compare the suitability and cost- ST/MT methods, no systematic comparison between them has been effectiveness of different surveying methods as a basis for conducted for Indian conditions. formulating recommendations and regulations. Wherever appropriate, change paper-based workflows into digital ones. Even in situations where pilot re-survey has been undertaken it Make availability of a strategy to integrate records and registry MT is often not integrated with land records or registry. a precondition for moving towards activities to improve spatial data. Ensure that whatever spatial data are generated in this way will be integrated into the existing system. High level ofland-related conflict precludes investment and Based on pilot studies, develop regulations to resolve these MT productive land use by 'freezing' many properties, especially in conflicts in an incentive-compatible way, ideally simultaneously peri-urban areas where land values are increasing (and maps are with a comprehensive updating of textual and spatial land highly inaccurate), in tribal areas where land alienation in records. Establish incentives, with sunset provisions, for settling contravention of the law is an issue, and in agricultural areas conflicts out of court in an expeditious manner. where a large number of disputes over ceiling land are still pending. Outdated regulations for surveying often make it impossible to Establish technical working group to: (i) explore options for ST/MT adopt 'modern' survey technology that could significantly reduce survey, recommend appropriate pilots, and issue regulations the cost of doing surveys. and manuals for specific situations (from sporadic/systematic Uniformly high precision standards and adherence to one type subdivision surveys to complete loss of records and unsettled of technology unduly increases cost of survey, making it more areas) and (ii) investigate whether changes in the institutional difficult to cater to a wide range of users and situations. structure-to ensure greater consistency and quality of surveys Institutional responsibility and technical standards for surveys especially in urban areas and to regulate an emerging private of different types are not clearly specified. sector would be appropriate. Prohibition of private sector participation causes long backlogs Provide the basis for developing technical capacity (for private ST/MT for subdivision surveys, reduces the ability to respond to or public sectors) by reviewing appropriateness of schools and demand for land-related services, and-via reduced revamping curricula. ~.. (contd.) Issue Action Timing competition-constrains adoption of new technology and Create regulatory environment (including educational quality of service provision. requirements, licensing, disciplinary measures) for operation of the private surveying sector. Revise existing regulations to facilitate use of a broader range of technology (for example, photogrammetry) and define the uses to which it can be put. Central government support to programmes for modernization Distinguish roll-out from innovation programmes and ST of land administration does not take into account the extent to consider providing funding for the two categories separately which they will generate public goods. and with different levels of matching by states. 4. Land rights of disadvantaged groups are often unclear and Disseminate existing legislation; where needed, clarify not enforceable provisions or formally recognize such rights Although the 2005 amendment of the HSA significantly Conduct systematic dissemination campaigns through ST strengthens women's inheritance rights, limited awareness collaboration with grassroots organizations. restricts women's ability to demand its enforcement. Evaluate the impact of the HSA Amendment on women's welfare and use this to assess the scope for similar legislation to extend to non-Hindu women. Land rights for tribals and others in unsettled and other marginal Provide opportunities for group ownership, together with a MT areas, even if located on revenue lands, are often not recognized. clearly defined process of making the transition from such an Individualization of land ownership may not be the most arrangement to individual land holdings. desirable and cost-effective option to secure tribal rights and Conduct systematic surveys of marginal areas which, for various may instead lead to negative consequences. reasons, have not been surveyed in the past. Strengthen local institutions and implement provisions of PESA at the state level. The fact that large tracts of land do not have updated and Develop and subsequently scale up processes to provide an ST/MT consistent records reduces their value, impedes investment, and integrated and up-to-date land record in an affordable manner. reduces supply of land to the rental market. 5. Restrictions on land markets prevent land access by the End restrictions on land leasing, develop and review obstacles poor, hurt productivity, and have a particularly negative impact to land sales markets on women Restrictions on land leasing limit access to land by the poor, Make land leasing legal in states where it is currently prohibited, ST/MT reduce productivity ofland use, and thus retard development of starting in states and areas with obvious potential for efficiency- rural areas. enhancing land transactions. (contd.) Issue Action Timing Abandon rent ceilings and instead replace them with legislation that will enable better functioning of rental markets. Carefully evaluate state-level pilots to abandon land rental restrictions to make the case for similar action by other states. Land reform (tenancy) legislation no longer transfers land to the Allow transferability of land received through land reform by MT poor but continues to exert a very negative impact on land rental beneficiaries, especially through sub-leasing. Explore options market activity. for giving ownership rights to long-sitting tenants through decentralized mechanisms that do not involve the government bureaucracy. Restrictions on subdivision and land use change in a number of Review and, where possible, eliminate unwarranted restrictions ST/MT states make transfer of land to non-agricultural purposes on land sales. unnecessarily cumbersome. Instead of allowing landowners to Encourage direct negotiations between landowners and potential negotiate directly with potential buyers, expropriation under the buyers in peri-urban areas and provide guidance for these. Land Acquisition Act is invoked even in cases where direct Revise the Land Acquisition Act and clearly circumscribe its negotiation would be preferable. applicability. 6. Prohibition of tribal land alienation is often ineffective Complement such legislation with options to improve access to insurance and decentralized governance Restrictions on transferability of tribal lands are costly and difficult Complement bureaucratic control of preventing land alienation ST/MT to enforce, have limited impact on the ground, and do little to with safety nets, a right of first refusal, community approval strengthen the rights of tribal groups. to sales, and collective bargaining. Improve the land administration system in tribal areas to provide the basis of information needed to enforce any regulations relating to land. Prohibition of tribal land alienation is paternalistic and does Strengthen land rights of tribal populations instead of restricting ST/MT nothing to prevent involuntary land sales and provides few them. Develop ways to implement systematic programmes to incentives to manage land sustainably, improve it, or take actions clearly establish rights and resolve conflicts in tribal areas, to be to ensure its most effective utilization. scaled up once the Tribal Land Rights Bill is passed. Note: ST = short term; MT = medium term. Source: Authors. APPENDIX Figures and Tables FIGURE A2.1: Illustration of a tippan (field measurement book) APPENDIX (i) Eventually included in the!n- Taluk of _ _- ; : : - . - , - - - - - formedaa per G. 0. Ms. 5o Dated _ _ and Later in the Mandai o(~ KukUIIP!t& ------·· formed as per Q, 0. Ma. No 496 Rey (U) nt•u\.. 5·19l5__ iii) The V1Uage Code No. printed above refera to old talult. NQ.461 BARAPALU -..,..1113 K.Ms NO. 466 SUKURU REFERENCE PAllt Pare-nt Village IE I Thotched house D FootPath NO. 485 ~~~ r.n~c DALJMPUT@] Village bol.lndary stet;oo 057{72 ~ '/illage trijunctioo slatm" G Mtncr circuit stoilon I I Village ooun Rs 50,000). Connecti- forest survey. be rolled out to 150 available anywhere in application for EC Rs 1020 crore in vity to revenue planned. 17,244 villages with this year and the state (including and CC, registration 1999-2000. Remit stamp duty at 5289 million survey remainder next year). web-based requests). anywhere, and model Banks. Use of software nos, 35,205 million Capital cost of the deeds from the web. as a dissemination tool subdivisions, and 2407 project repaid to the Indexes are uploaded for information to field million town survey government using via dial-up to central offices. nos. 151 towns, 6 computerization fee server daily. corporations. 558 ofRs 100. SROs in 50 regn districts and 9 zones. West 18 districts, 66 sub- CORD implemented Computerization of Legacy records (50 Duty 6% in Panchayat User charges to be Bengal divisions, 341 blocks, in 4 offices of market value lakh) expected to be areas, 1Oo/o in Kolkata determined based on 3358 GPs, and 40,782 Hooghly district- monitoring register completed in 2006. and Howrah, 8% in proposals received from villages. p31; roll-out expected database planned. Data entry is being rest of state; private vendors. to be completed by done for 1999-2003 registration fee 1.1 %. end 2007 through records. Total revenue in PPP (BOOT). 2004-5 was Rs 942 crore as compared to Rs 430 crore in .. 1999-2000 . TABLE A2.3: Status of modernizing survey in selected states State Last revisional survey Village maps FMB digitization Private surveyors Status of town Pilots/ remarks digitization surveys Andhra On average, the last Village maps in Of 8.57 million sur- Private surveyors Surveys undertaken Experimental re-survey Pradesh survey was done at the Nalgonda and vey numbers, FMBs/ allowed but work to only in parts of the in 17, 25, and 7 villages turn of the century. Karimnagar (< 10% tippans available for be scrutinized by 30 urban centres; even in Guntur, Chittoor, of total) vectorized. 7.72 million (0.852 divisional deputy where they have, and Karimnagar, million missing), 6.7 inspector against a records are not upto respectively (75,000 ac million have been 10% fee but only 60 date. total) using ETS and scanned. Digitization have been trained Problems with RoR GPS by private sector of 93,000 FMBs in thus far and will in peri-urban areas since 2001. Progress Chittoor district start operation soon. are huge. slow (< 50% finished ongomg. No provision for in Guntur). mapping habitation IUS with 36 crore lands. budget. Bihar First survey around No village maps No village maps Not allowed Maps not updated as 1900; 5 districts started digitized digitized. part of mutation revisional survey in process, thus our of date. 1959 and have final hearings underway. 4 started in 1981 and work still ongoing. Gujarat ftJl surveys (in ex- All 24,085 VMs Pilot in 539 villages No private surveyors 2.3 million property 100 individuals trained princely states and scanned; 4050 geo- of 2 districts. allowed (actually cards are available in in use of ETS and GPS. Bombay presidency) referenced and says not prevalent 66 city survey offices from before indepen- 2348 digitized rather than not covering 7 Municipal dence. After that, survey allowed). Corporation, 172 un-surveyed areas in municipalities, and Kutch, Saurashtra, etc. 518 towns. Some to 1975, followed by 100,000 mutations 2500 village survey. annually. Surveys in progress for 0.9 million properues. {contd.) State Last revisional survey Village maps FMB digitization Private surveyors Status of town Pilots/ remarks digitization surveys Himachal Re-survey should be Not undertaken. Not undertaken. No private surveyors No information given. Average time for survey Pradesh undertaken every 10 allowed. of a revenue village is years in urban and 2 years. every 40 years in urban areas but current records in Shimla more than 100 years old. Karnataka There have been no All village maps have Scanning of tip pans Licensed surveyor Last city survey was Pilot in Coorg district regular repetitions of been scanned and done in 8 taluks scheme implemented done in 40 cities about stopped in 1995. ETS settlement surveys; only half have been (survey records) (out for surveys; work to 100 years ago. 222 survey in Maddur taluk sporadic in individual vectorized by of a total of 13 where be reviewed and city surveys planned; (Mandya dist.) started in locations. KSASARC it was taken up). certified at taluk 48 underway but no 2002; completed in Vectorization planned office. 2280 persons date for completion. 2006 in 33 villages. but not started yet. with licence. Phodi (subdivision) Hudbusth backlog is 146,726. Kerala No information given. No scanning of village Pilot project for The state has stopped A total of 3 villages Pilot 'Torrens offices' maps in the state. digitization of FMBs, issuing licences to surveyed by GPS and since April 1995 in (No village maps exist which makes sub- private surveyors. ETS; prior to this all Kottayam and in Kerala-villages are divisions much easier No survey done in chain and crosstab. Eranakulam aimed to divided into blocks in 6 taluks, to be cases of subdivision. make survey mandatory and only these have expanded to 48 other Survey to be done for before registration; there maps) villages. free by government. are also cost issues. Malharashtra Original survey in All 44,000 village Scanning of tippans Survey is not All 22 municipal Land tiding pilots in 1901, last revisional maps have been and pharnis (sub- compulsory for corporations surveyed, Aundh in Pune munici- survey in 1927. In rural scanned, vectorized, division maps) comp- subdivision requests. plus 4500 of 6000 pality, circles Hadolati areas, all surveys older and geo-referenced leted in 2 taluks on a Backlog of 2 million villages with a popu- and Chandola in teshil than 30 years; in urban by MRSAC and are pilot basis. It is now subdivisions where lation more than 2000 Ahmadpur and Mukhed ones 1Oo/o surveyed available to the public. taken up on a routine no survey yet. eligible for city survey. (Latur, Nanded district) during last 10 years basis. Private surveyors 159 out of 220 approved in 2002. Very (contd.) State Last revisional survey Village maps FMB digitization Private surveyors Status of town Pilots/ remarks digitization surveys and 15% within 20 hired and subcontr- municipalities sur- limited progress in years. acted by department. veyed; rest to be developing procedures to completed soon. investigate/adjudicate rights. Madhya Re-survey started in Digitization of maps No information No private surveyors. Survey of 135 out of Re-survey of one-third Pradesh 1976-7; declared in 12 tehsils (3000 provided. 369 urban areas of villages completed in completed in 2000 villages) at cost of Rs (nazoob started in the 1970s with although only 5717 18 million (Rs 6000 1964. Aerial/ground theodolite, less than villages out of 18,119 per village or Rs 2700/ survey for 111 one-third in the 1980s completed and 11 dis- sheet). Discontinued completed. and early 1990s using tricts not covered at all. due to mis-match in Survey of abadi aerial photography and areas. (habitation) lands is theodolite with detailed Maps missing for planned using total ground survey. 1121 villages, 43 stations; cost estimates dealt with. have been prepared. Orissa Average age of existing Digitization of 4600 FMBs scanned Not allowed. No information given. DFID project. settlement survey is 20 cadastral maps done but not digitized; years. Plain table survey in 6 tehsils on pilot discontinued. conducted with chain basis at a cost of takes about 7 months Rs 3000 per map. for one village (250 ac.). About 3800 out of 2.18 lakh maps, i.e. 1.5%, are digitized. Punjab Current settlement Only village maps for Plans for digitization No private surveyors. No information given. Software for digitization period started in 1889 28 out of 13,001 of field maps does not seem to be (i.e. no new settlement villages digitized. (musavis) through standardized. completed since then). PPP but no quality standards given. Rajasthan 64 of 241 tehsils Not done; plans to No information No provision for Cities not covered by Plane table used surveyed less than 20 do so exist. available. private surveyors even survey department universally. 88 of 241 (contd.) State Last revisional survey Village maps FMB digitization Private surveyors Status of town Pilots/remarks digitization surveys years ago, 123 with though scarcity of tehsils have been surveys more than 20 staff is mentioned converted to metric years old, and 54 are as a key problem. system. 8 tehsils flown notified for survey (29 under aerial the first time since photography but independence); cadastral maps have not completion is expected yet prepared. in 2015. Tamil Last full survey in 1917. Village maps in 26 Pilot in 4 taluks No private surveyors Natham survey almost Pilots with ETS/GPS in Nadu UDR (updatation of out of 30 districts digitized 1.22 lakh currently. completed, town Konathyi villages (1.08 registers) drive in 1987 scanned and stored FMBs but too expen- survey done for 5 out sq km in 4 months and to update FMBs and as raster images. sive (Rs 100 per of 6 corporations and 1250 mdays) and 4 efforts to update FMB). NIC is Pilo- 82 of 151 municipal villages in Chennai natham land. Hill ting COLLABLAND towns. (4.22 sq km) in 9 survey undertaken for in Perambalur at months with 2243 858 sq km of non- Rs 10/FMB. mdays). agricultural land. West Initiated in 1972-4, 700 maps in Hooghly 892 mouzas digitized No private surveyors Only Kolkata (100 sq Some aerial and satellite Bengal apparently completed. in 2001 but no plans under pilot project. allowed (has to be km is recognized as survey in Purulia. for follow-up or confirmed by staff). urban). No information Surveys of Salt Lake modernization. Survey is free. given on status of and Kolkata using ETS survey there. have been undertaken as well as cadastral mapping in Purulia by photogrammetry funded by state government. Source: PriceWaterhouseCoopers (2006). References Agarwal, B. (1994), A Field of One's Own: Gender and _ _ (2003), 'Tenancy rent control', Swedish Economic Land Rights in South Asia, South Asian Studies, Policy Review, Vol. 10, No. 1, pp. 89-121. Cambridge; New York and Melbourne: Cambridge Banerjee, A.V., P. J. Gertler, and M. Ghatak (2002a), University Press. 'Empowerment and Efficiency: Tenancy Regorm in Aghion, P. and P. Bolton (1997), 'A Theory of Trickle- West Bengal', journal ofPolitical Economy, Vol. 110, Down Growth and Development', Review ofEconomic No. 2, pp. 239-80. Studies, Vol. 64, No.2, pp. 151-72. _ _ (2002b), 'Empowerment and Efficiency: Tenancy Agrawal, V. K. (2006), 'Andhra Pradesh Pilot Project on Reform in West Bengal', journal ofPolitical Economy, Integrated Land Information System', Paper presented Vol. 110, No. 2, pp. 239-80. at the MoRD-WB Workshop on Land Policies and Banerjee, A. V. and L. Iyer. (2004), 'History, Institutions, Administration in India, New Delhi: 5 and 6 January. and Economic Performance: The Legacy of Colonial Ahuja, M. and A. P. Singh (2006), 'Evaluation of Land Tenure Systems in India', MIT Working Paper Computerization of Land Records in Karnataka', 02-27 Cambridge, MA. Economic and Political Weekly, 7 January, pp. 69-77. Bardhan, P., S. Bowles, and H. Gintis (2000), 'Wealth Aim, J., P. Annez, and A. Modi (2004), 'Stamp duty in Inequality, Wealth Constraints and Economic Indian states: A case for reform', World Bank Policy Performance', in A.B. Atkinson, and F. Bourguignon Research Paper 3413, Washington DC: World Bank. (eds), Handbook of income distribution, Amsterdam: Alston, L.]. and]. P. Ferrie (2005), 'Time on the Ladder: North-Holland. Career Mobility in Agriculture, 1890-1938' NBER Bardhan, P. and D. Mookherjee (2006), 'Land Reform, Working Paper 11231, Cambridge MA: National Decentralized Governance, and Rural Development Bureau of Economic Research. in West Bengal', Paper prepared for the Stanford Andolfatto, D. (2002), 'A Theory ofinalienable Property Centre for International Development Conference on Rights', journal of Political Economy Vol. 110, No. 2, Challenges of Economic Policy Reform in Asia, pp. 382-93. Stanford, CA, 31 May-3 June. Appu, P. S. (1997), Land Reforms in India: A Survey of Basu, K. and P.M. Emerson (2000), 'The Economics of Policy, Legislation and Implementation, New Delhi: Tenancy Rent Control', Economic journal, Vol. 110, Vikas Publishing House. No. 466, pp. 939-62. Arnott, R. (1995), 'Time for revisionism or rent control?', Benjamin, D. (1995), 'Can Unobserved Land Quality journal of Economic Perspectives, Vol. 9, No. 1, pp. Explain the Inverse Productivity Relationship?', Journal 99-120. ofDevelopment Economics, Vol. 46, No. 1, pp. 51-84. REFEREN Berry, R. A. and W R. Cline (1979), Agrarian Structure Carter, M. R. and F. J. Zimmerman (2000), 'The and Productivity in Developing Countries, Baltimore: Dynamic Cost and Persistence of Asset Inequality Johns Hopkins University Press. in an Agrarian Economy', Journal of Development Besley, T. and R. Burgess, (2000), 'Land Reform, Poverty Economics, Vol. 63, No. 2, pp. 265-302. Reduction, and Growth: Evidence from India', Chau, N. H. (1998), 'Land Reforms in the Presence of Quarterly journal ofEconomics, Vol. II5, No.2, pp. Monitoring Costs and International Trade', Review of 389-430. International Economics, Vol. 6, No. 4, pp. 564-79. Besley, T., R. Burgess, and B. Esteve-Volart (2004), Conning, J. and J. A. Robinson (2002), 'Land Reform 'Operationalising pro-poor growth: India case study', and the Political Organization of Agriculture', CEPR London: DFID, BMZ, GTZ, KfW, and the World Working Paper 3204, London. Bank. Deb roy, B. (2000), 'Some issues in law reform in India', Binswanger, H. P., K. Deininger, and G. Feder (1995), in J.J.Dethier (ed.), Governance, Decentralization and 'Power, Distortions, Revolt and Reform in Agricultural Reform in China, India and Russia, Dordrecht: Kluwer. Land Relations', Handbook ofDevelopment Economics, Deininger, K. (2003), Land Policies for Growth and Poverty 3B, pp. 2659-772. Reduction. A World Bank Policy Research Report, Oxford Binswanger, H. P. and M. Elgin (1988), 'What Are the and New York: World Bank and Oxford University Prospects for Land Reform?' in A. Maunder, and A. Press. Valdes (eds), Agriculture and Governments in an Deininger, K., M. Maertens, P. Olinto, and F. Lara Interdependent World, Proceedings of the Twentieth (2002), 'Redistribution, Investments and Human International Conference of Agricultural Economists, Capital Accumulation: The Case of Agrarian Reform Aldershot: Ashgate. in the Philippines', World Bank Discussion Paper, Bird, R. M. (2004), International Handbook of Land and Washington D.C. Property Taxation, Cheltenham, UK and Northampton, Deininger, K., D. Ayalew, and T. Yamano (2006), 'Legal Mass.: Elgar. knowledge and economic development: The case of Bourguignon, F. and T. Verdier (2000), 'Oligarchy, land rights in Uganda', World Band Policy Research Democracy, Inequality and Growth', journal of Working Paper 3868, Washington DC: World Bank. Development Economics, Vol. 62, No.2, pp. 285-3I3. Deininger, K. and S. Jin. (2003), 'Land Sales and Rental Brown, P. K. and M. A. Hepworth (2002), 'A study of Markets in Transition: Evidence from Rural European land tax systems', Working Paper WP02PB I, Vietnam', World Bank Policy Research Working Cambridge, MA: Lincoln Institute of Land Policy. Paper 3013, Washington DC: World Bank. Burgess, R. (200I), 'Land and Welfare: Theory and _ _ (2005), The potential ofland markets in the process Evidence from China', London School of Economics of economic development: Evidence from China', Working Paper. journal ofDevelopment Economics, Vol. 78, No. I, pp. Burns, T. A. (2004), 'Computerization of land records: 241-70. Building on Karnataka's experience', Wollongong: Eswaran, M. and A. Kotwal (1985), 'A Theory of Land Equity International. Contractual Structure in Agriculture', American Cardenas,]. C. (2003), 'Real Wealth and Experimental Economic Review, Vol. 75, pp. 352-67. Cooperation: Experiments in the Field Lab' ,Journal of Fabella, R. V. (2003) 'The Comprehensive Agrarian Development Economics, Vol. 70, No.2, pp. 263-89. Reform Program and Coase Theorem', Philippine Carter, M. R. (1984), 'Identification of the inverse Review of Economics, Vol. 40, No. 1, pp. 109-14. relationship between farm size and productivity: An Feder, G. (1985), 'The Relation between Farm Size and empirical analysis of peasant agricultural production', Farm Productivity: The Role of Family Labor, Oxford Economic Papers, Vol. 36, pp. 131-4 5. Supervision and Credit Constraints', journal of DIA-LAND POLICIES FOR GROWTH AND POVERTY REDUC Development Economics, Vol. 18, Nos 2 and 3, pp. Lieten, G. K. (1996), 'Land Reforms at Centre Stage: 297-313. The Evidence on West Bengal', Development and FIG Commission 7 (2006), 'Cadastral Template, A Change, Vol. 27, No. 1, pp. 111-30. worldwide comparison of cadastral systems', Lobo, A. and S. Balakrishnan, (2002), 'Report card on Melbourne: PCGIAP, University of Melbourne, and service of bhoomi kiosks: An assessment of benefits FIG. by users of the computerized land records system in Calor, 0. and J. Zeira (1993) 'Income Distribution and Karnataka', Working Paper, Bangalore: Public Affairs Macroeconomics', Review ofEconomic Studies, Vol. 60, Centre. No. 1, pp. 35-52. Manthorpe, ]. (2004), 'Comparative Analysis on the Government of India (1966), 'Implementation of Land Cadastral Systems in the European Union. United Reforms: A reveiw by the land reforms implementation Kingdom: National mapping, land valuation, land committee of the National Development Council', registration', London: Permanent Committee on New Delhi: Planning Commission. Cadastre of the EU. _ _ (2002), 'lOth Five Year Plan (2002-2007)', New Mearns, R. (1999), 'Access to Land in Rural India', Policy Delhi: Planning Commision. Research Working Paper 2123, Washingon DC: Government of India, National Committee on the World Bank. Development of Backward Areas ( 1981) 'Report on Moog, R. S. (1997), Whose interests are supreme? Development of Tribal Areas', New Delhi: Planning Organizational politics in the civil courts in India, Commission. Ann Arbor, Michigan: Association for Asian Studies. Hanstad, T., T. Haque, and R. Nielsen (2006), 'Improving Nettle, K (2006), 'Tides vs deeds: International experience Land Access for India's Rural Poor', Paper presented and implications for India', Paper presented at the at the MoRD!World Bank workshop on 'Land Policies MoRD!World Bank workshop on 'Land Policies for for Accelerated Growth and Poverty Reduction in Accelerated Growth and Poverty Reduction in India', India', New Delhi: 5 and 6 January. New Delhi: 5 and 6 January. Haque, T. and A. Singh (1986), Agrarian Reforms and Nugent, ]. B. and J. A. Robinson (2002), 'Are Institutional Changes in India, New Delhi: Concept Endowments Fate?' CEPR Working Paper 3206, Publishing Company. London. Jaamdar, S.M. (2006), 'Integration of phodi process with Panda, P. and B. Agarwal (2005), 'Marital violence, human mutation in bhoomi', Paper presented at the MoRD- development and women's property status in India', WB Workshop on Land Policies and Administration World Development Vol. 33, No. 5, pp. 823-50. in India, New Delhi: 5 and 6 January. PriceWaterhouseCoopers (2006), 'Status assessment of Kaushik, A. (2005), 'Nature of land market land records, registration, and survey in 14 Indian interventions in India', Institute for Sustainable states', New Delhi, (mimeo). Development, Noida. Quisumbing, A. R. and]. A. Maluccio (2003), 'Resources King, R. (1977), Land Reform: A World Survey, London: at Marriage and lntrahousehold Allocation: Evidence G. Bell and Sons. from Bangladesh, Ethiopia, Indonesia, and South Kumar, K., P. R. Choudhury, S. Sarangi, P. Mishra, and Africa', Oxford Bulletin of Economics and Statistics, S. Behera (2006), 'Contested landscapes, dispossessed Vol. 65, No. 3, pp. 283-327. people: Land, forests, and tribals in Orissa', Paper Raju, K., K. Akella, and K. Deininger, (2006), 'New presented at the MoRD/World Bank workshop on opportunities to increase land access in India: The 'Land Policies for Accelerated Growth and Poverty example of Andhra Pradesh', Paper presented at the Reduction in India', New Delhi: 5 and 6 January. MoRD/World Bank workshop on 'Land Policies for REFEREN Accelerated Growth and Poverty Reduction in India', Upadhyay, S. (2006), 'Tribals and their lands: Legal issues New Delhi: 5 and 6 January. and concerns with special reference to Madhya Rawal, V. (2001), 'Agrarian Reform and Land Markets: Pradesh and Andhra Pradesh', Paper presented at the A Study of Land Transactions in Two Villages of West MoRD/World Bank workshop on 'Land Policies for Bengal, 1977-1995 ', Economic Development and Accelerated Growth and Poverty Reduction in India', Cultural Change, Vol. 49, No. 3, pp. 611-29. New Delhi: 5 and 6 January. Rezende, G. C. d. (2005), 'Politicas trabalhista e fundiaria Warriner, D. (1969), Land Reform in Principle and Practice, e seus efeitos adversos sobre o emprego agricoloa, a Oxford: Clarendon Press. etrutura agraria e o desenvolvimento territorial rural no World Bank, (2004), Doing business in 2004: Under- Brasil', Texto para discussao No 1108, Rio de Janeiro: standing Regulation, Washington DC: World Bank, Instituto de Pesquisa Economica Aplicada (IPEA). International Finance Corporation, and Oxford Saxena, N. C. (2005), 'Updating Land Records: Is University Press. Computerisation Sufficient?', Economic and Political _ _ (2005a), Urban Land Markets in India: Analytical Weekly, Vol. 40, No.4, pp. 131--46. Background Papers. Washington DC: Urban and Spillman, W]. (1919), 'The agricultural ladder', American Infrastructure Sector Unit, South Asia Region. Economic Review, Vol. 9, No. 1, pp. 170-9. _ _ (2005b), 'World Development Report 2006: Thangaraj, M. (2004), Land reforms in India, Volume 9, Equity and Development', Washington DC: World Tamil Nadu-An Unfinished Task, New Delhi: Sage Bank and Oxford University Press. Publications. Yugandhar, B. N. (1996), Land reforms in India, Volume Thorner, D. (1976), Agrarian Prospect in India, New 3. Andhra Pradesh--People's pressure and administrative Delhi: Allied Publishers. innovations New Delhi: Sage Publications. Transparency International India (2005), 'India corruption Zimmerman, F.]. and M. R. Carter (2003), 'Asset Smooth- study 2005', New Delhi: Transparency International. ing, Consumption Smoothing and the Reproduction UNECE (1996), 'Land Administration Guidelines, With oflnequality under Risk and Subsistence Constraints', Special References to Countries in Transition', United journal of Development Economics, Vol. 71, No. 2, Nations Economic Commission for Europe, ECE/ pp. 233-60. HBP/96, New York and Geneva. ISBN 019568959 - 3 lllltlllllllllllllllllll 9 780195 689594 Rs 395