Report No. 32253-IN INDIA Urban Finance and Governance Review Volume I (Executive Summary and Main Report) December 2004 Energy and Infrastructure Unit South Asia Region Document of the World Bank CURRENCY EQUIVALENTS Currency unit: Indian Rupee (RS) US$1 = Rs. 44.66 GOVERNMENT FISCAL YEAR April 1-March 31 ABBREVIATIONS AND ACRONYMS BATF Bangalore Agenda Task Force IDSMT Integrated Development of Small and BDA Bangalore Metropolitan Region Medium Towns Development Authority KAS Karnataka Administrative Service BMC Brihan Mumbai Corporation KHB Kamataka Housing Board ("Mumbai") KMA Kamnataka Municipalities Act BMP Bangalore City Corporation KMAS Karnataka Municipal Administrative BWSSB Bangalore Water Supply and Service Sewerage Board KMCA Karnataka Municipal Corporation Act CAA Constitution Amendment Act KTCP Karnataka Town and Country CAS Country Assistance Strategy Planning Act CC City Corporations (Karnataka) KUIDFC Karnataka Urban Infrastructure CCF City Challenge Fund Development Finance Corporation CMC City Municipal Councils (Karnataka) KUWSDB Karnataka Urban Water Supply and CMWSSB Chennai Metropolitan Water Supply Drainage Board & Sewerage Board LG Local Government DA Development Authorities MDF Municipal Development Fund DFID Department for International NGO Non-Governmental Organization Development PWD Public Works Department DMA Directorate of Municipal SCB Slum Clearance Board Administration SFC StateFinanceCommissions EFC Eleventh Finance Commission SWM Solid Waste Management ESW Economic Sector Work TA Technical Assistance GDP Gross Domestic Product TMC Town Municipal Councils GOI Government of India (Karnataka) GOK Government of Kamataka TNUDF Tamil Nadu Urban Development GOM Government of Maharashtra Fund GOTN Government of Tamil Nadu TP Town Panchayats HDFC Housing Development Finance UDD Urban Development Department Corporation Ltd. ULB Urban Local Body HUDCO Housing and Urban Development UDPA Urban Development and Poverty Corporation Alleviation (Ministry) IAS Indian Administrative Service URIF Urban Reform Incentive Fund IBRD International Bank for Reconstruction USAID United States Agency for and Development International Development IDFC Infrastructure Development Finance UWSS Urban Water Supply and Sanitation Company The World Bank Vice President: Praful Patel Country Director: Michael Carter Sector Director Vicent Gouarne Sector Manager: Sonia Hammam Task Team Leader: Dana Weist Foreword ,port was prepared by a team led by Dana Weist, and comprised of Roy Bahl (property tax reform), ,mik Lall (urban economic analysis), Lars Sondergaard (economic and fiscal analysis), K. Mukundan, Ajit Karnik (Maharashtra analysis), Abhay Pethe (Maharashtra analysis), Christine Wong, and Kirida Bhaopichitr. The team is grateful to many counterparts in the States of Kamataka, Maharashtra, and Tamil Nadu for their assistance. Helpful comments were provided by Sonia Hammam, Patricia Annez, Stephen Howes and Soraya Goga, as well as by the peer reviewers, William Dillinger, Robert Ebel, and George Peterson. India Urban Finance and Governance Review World Bank, December 2004 Executive Summary ............................................................................i Intergovernmental Fiscal Relations ........................................................................... ii Governance ........................................................................... iv Options for Reform ............................................................................v 1. Introduction ............................................................................1 2. Analytical Framework ............................................................................2 3. Intergovernmental Fiscal Relations ............................................................................6 Central Roles in Urban Affairs ............................................................................6 State Government Roles in Urban Affairs ............................................................................9 Expenditure Assignment and Levels ............................................................................9 Revenue Assignment and Yield ............................................................................ 13 Intergovernmental Grants ........................................................................... 20 4. Borrowing ........................................................................... 23 5. Governance Aspects ........................................................................... 26 Accountability ........................................................................... 26 Capacity ........................................................................... 32 6. Conclusions and Options for Reform ........................................................................... 33 Enhancing Fiscal Sustainability ........................................................................... 34 Strengthening Institutional Arrangements ........................................................................... 40 7. References ............................................................................ 44 List of Tables Table 1: Comparison of ULB Population and Fiscal Indicators, Per Capita for Selected Years (in Rs) ......4 Table 2: Comparisons of Local Per Capita Expenditure and Revenue, Selected Countries (in US$) .........6 Table 3: 10th Plan Outlay for Central Schemes and Centrally Sponsored Schemes for Urban Development, Urban Employment and Poverty Alleviation (Rs. In billions) ............. .........................8 Table 4: Categories of Expenditures per Capita, Selected States (in Rupees) . .11........................... 1 Table 5:Determinants of Per Capita Expenditures, Selected Indian Corporations . . 12 India: Urban Governance and Finance Review EXECUTIVE SUMMARY India has been in the process of decentralization to urban local bodies (ULBs) since the early 1990s, when the Government of India passed the 74' Constitutional Amendment Act (CAA), which directed states to devolve some of their responsibilities to ULBs. In the decade since passage of these amendments, however, decentralization has been limited, with slow and uneven progress across states. While the CAAs envisioned decentralization of functions, finances and functionaries to enable ULBs to function as "institutions of self government," in reality, fiscal and administrative decentralization have lagged behind political decentralization. Limited progress has also been seen in decentralization to rural local governments (panchayat raj institutions), which was enshrined in the 73rd CAA. ULBs play a critical role in India's development. While its I billion-plus population is predominantly rural, over 300 million people live in urban areas. One-third of this population lives in 35 urban agglomerations or cities with populations exceeding 1 million. The share of GDP generated in urban areas has increased over the past 10 years, and urban poverty levels have declined. Cities are responsible for delivering various public services, yet severe infrastructure shortages in water supply and sanitation, roads, transportation, housing and waste management, and inefficient management have resulted in poor quality services with limited coverage. These inadequate services and worsening environmental conditions disproportionately affect the poor. It is an appropriate time to review the financing and governance of ULBs. The 12t Finance Commission has been constituted, and the possibility of revising the 74h Amendment to strengthen the legal foundation for ULB self governance is being discussed. Awareness of the importance of cities and the need to improve their performance is growing, as evidenced by the priority given to them in the 1W0h Plan. Equally important, ULBs are introducing reforms to improve their financing and governance, and much can be learned about applying these innovations more broadly. Considerable work has documented finance and governance challenges in Indian states and panchayats; this Review complements this work by focusing on ULBs. The Review documents the finance and governance aspects of ULBs in three states (Karnataka, Maharashtra and Tamil Nadu); analyzes why institutional arrangements have promoted or impeded expected outcomes; and assesses the implications of these innovations/impediments on financing and delivering ULB services. These three states were selected because of their large urban populations, different intergovernmental and institutional arrangements, and ongoing urban reforms. The audience for this Review includes Bank staff members working on India, as well as counterparts and researchers interested in comparative analysis of ULBs. From an international perspective, Indian ULBs play a relatively limited fiscal role. In part, this reflects different income levels, but also the fact that Indian ULB roles are generally confined to the provision of infrastructural services such as water and sewerage, sanitation, solid waste removal and street lighting. Unlike urban governments in other countries, they do not generally have significant responsibility for health and education. However, their spending as a share of GDP is considerably less than other countries in the early stages of decentralization where local governments are primarily responsible for providing infrastructure services. Cross-state comparisons of ULB fiscal performance are difficult for two reasons. First, what constitutes an "urban local body" varies significantly across states, and even within a given state. Second, the quality of fiscal data for ULBs is generally poor. Despite these difficulties, comparative analysis shows that ULB total expenditures and total revenues have grown substantially, and on a per capita basis between fiscal years 1996/97 and 2000/01, with average annual spending growth exceeding revenue growth. This trend may indicate an underlying structural imbalance between the costs or demands on ULBs to provide services and their ability to finance those services. Given the population pressures in i Executive Surmnary and Main Report these highly urbanized states and fiscal strains at the state level, this gap is likely to widen over time unless ULBs contain the growth in spending or enhance their local revenue mobilization. INTERGOVERNMENTAL FISCAL RELATIONS As a federation, India has important intergovernmental roles for the Central Government, states and ULBs. The Constitution demarcates functions and finances between the Central Government and state governments. The CAA clearly recognizes that ULBs are "subjects" of the states, hence state-local relations are relatively more important than central-local relations. The Central Government provides funding for ULBs through grants, various central schemes and the recent establishment of incentive funds. However, an evaluation of centrally-sponsored urban schemes has found that they are often underutilized, poorly targeted and limited in achieving their objectives. State Acts define ULB powers, including revenue sources, and autonomy in taxation and expenditure priorities. In addition, state governments determine the number and type of ULB employees, as well as their compensation; oversee ULB budget preparation; issue guidelines and standards for service provision; and issue clearances/concurrence for various financial activities (e.g., procurement) and regulatory issues. States provide considerably more financing than the Central Government, although their flows to ULBs depend in part on their own fiscal health. Large and growing state fiscal deficits have negatively affected state-local fiscal relations. In the three states, delays and reductions in the transfers paid to ULBs were commonly observed. Weak linkages among revenues, expenditures, transfers and borrowing characterize the intergovernmental fiscal relations of the ULBs in the three states. Liabilities and arrears in the repayment of local debt and utilities are endemic in many areas. The CAA defines an "illustrative" list of 18 municipal responsibilities and functions. Expenditure assignment is relatively similar across the three states, except for the devolution of some education and health services to ULBs in Maharashtra. Maharashtra ULBs (especially corporations) provide more core services -- water, sanitation, street lights and roads -- than other states, whereas Karnataka's ULB spending on core services is comparatively low. In all three states, a key ULB function is supplying water, and yet in some instances, state capital spending dominates that sector. Spending differences across ULBs within a state do not appear to be related to poverty levels, as measured by the share of the population in slums. Most State Finance Commisions have not assigned new revenue bases to local authorities, and ULBs have few, autonomous sources of revenue. Effort in collecting existing ULB revenues is low. Nationwide, the central government dominates revenues, and ULBs account for less than 3 percent of total revenues of all tiers of government. In all three states, corporations are generally less dependent on state transfers than other ULBs. Per capita current revenues are considerably higher in Maharashtra corporations, due to their reliance on the octroi. ULBs in Tamil Nadu, on average, collect more current revenues per capita, and are more self-reliant than ULBs in Maharashtra or Tamil Nadu. Most ULBs have a poor record in recovering user charges, especially for water supply, which generally do not cover operating costs, let alone the costs of needed capital investment. Typical problems include poor administration and enforcement of charges and fees; weak information systems for proper billing and collection; political unwillingness to impose full charges; and a culture of non-payment, related in part to the poor quality of services that are provided. These challenges can be overcome, however, as has been shown by the experience of the Bangalore Water Supply and Sewerage Board, which has effectively used information technology to improve its billing and collection systems, and strong enforcement to enhance it's recovery of water charges. Without improved cost recovery, the ii India: Urban Governance and Finance Review quality of utility services will not improve, and opportunities for engaging the private sector in these utilities will be limited. The property tax is a key revenue source, although, at present, it is relatively underused, and has limited buoyancy relative to the overall growth in economic activity. Municipalities have adequate legal powers to collect and recover property taxes tax due. Yet, across Indian municipal corporations, the average collection ratio of property taxes to demand is one-half of annual demand, which is well below collections of state sales taxes, the central income tax and central excise taxes. In many cases, the property tax is used to pay for statewide services via a cess that is surcharged to the basic property tax. These cesses lessen the link between taxes paid and (local) benefits received, and are often not collected nor remitted in full to the State. Recent property tax reforms in Bangalore, among other ULBs, show promising results in increasing the yield of the property tax. In furthering these reforms, a common starting point is to improve property tax administration, and to link financing with service delivery. Focusing on the fundamentals of updating property tax rolls, computerizing billing and collection systems, and strengthening enforcement, are important first steps that can yield significant results. Such reforms can increase the yield of property tax revenues - in some cases quite substantially - but these actions will not increase the overall buoyancy of the property tax system. India faces a major structural problem with its property tax systems, resulting from the failure to resolve conflicts between assessing the true market value of property with rent control ordinances, and other limitations such as the FSI. Moreover, government officials have generally been unwilling to issue new valuation rolls, in some cases for many years. Much of the recent property tax reform in India has entailed stop-gap measures to overcome these problemns, rather than engaging in comprehensive reform. Meanwhile, the growth in property tax revenues has remained anemic. Unless these structural issues in properly valuing property are resolved, improved administration will do little to make the property tax a viable revenue source for local governments, and the gap between local expenditures and revenues is likely to grow over time. State transfers are important resources for most ULBs, however, they achieve few of their intended objectives. Due to state-level fiscal problems, and arrears in local payments that are intercepted from the payment of transfers, payments are generally not timely, stable nor predictable. Nor do these transfers offset fiscal disparities across ULBs. In Maharashtra, state transfers appear to provide a disincentive to local revenue mobilization, with ULBs substituting transfers for their own revenues. Given the high dependence of ULBs on transfers and schemes, their lack of predictability and stability constrains rational budgeting processes at local levels. Most ULBs in Karnataka and Tamil Nadu depend heavily on grants and loans from state and central agencies (many of which are intercepted or adjusted) to finance their infrastructure investments. In Maharashtra, corporations borrow extensively to finance their infrastructure, whereas municipal councils have less outstanding debt. Tamil Nadu is unique among the three states in having established a financial intermediary - the Tamil Nadu Urban Development Fund (TNUDF) - to finance infrastructure through responsible local borrowing, and strengthen local capacity. These grants and loans contribute only a small portion of overall infrastructure financing needs, however. With the exception of the TNUDF, few incentives nor enforcement mechanisms ensure a hard budget constraint for ULB borrowing: ULB arrears in repaying debt and state bailouts are common. Despite prevalent local borrowing, most ULBs fail to meet the basic requirements for creditworthiness: (i) stable, predictable and adequate revenues to support borrowing for capital investment; (ii) managerial iii Executive Summary and Main Report and financial capacity to use debt responsibly, as well as strategic planning for investment; and (iii) a track record of timely repayment of principal and interest. GOVERNANCE State agencies in Kamataka, Maharashtra and Tamil Nadu retain critical roles in planning, financing, and sometimes managing infrastructure and services. These state boards and authorities have weak accountability and limited coordination vis. ULBs. Authority and financing are not congruent with ULB responsibilities, leading to instances where ULBs bear the financial responsibility for decisions in which they had little or no say. This separation of the decision maker (often a state entity), from the financier, the service deliverer, and the ultimate beneficiary, results in the provision of infrastructure and services that do not match local preferences and needs, and are often not repaid nor maintained. ULBs also have limited scope in managing their assets. Land is over-regulated and urban land markets are highly distorted (e.g., land ceiling act, rent control, regulations on conversion of use, high stamp duties). Land use regulations, titling and permitting are generally opaque and ineffective. ULBs face numerous layers of oversight and regulation, with many decisions on procurement and staffing (e.g., recruitment, hiring, staffing levels, pay scales) taken by state officials. ULBs ability to respond to local demands is circumscribed by their inability to hire or fire qualified staff. Central government decisions on public employee wages made through the Pay Commission also have significant impacts on local wage payments. Low limits for municipal approval authority (technical and administrative sanction) and layers of bureaucratic reporting limit ULB autonomy, and discourage them from improving their performance or pursuing innovations. The appointment of senior municipal officials by the Indian Administrative Service or State Administrative Service brings capacity and professionalism to ULBs. However, this structure also directs the accountability of these officials upward rather than toward local politicians or citizens. Local politicians have very limited ability to hold appointed municipal officials accountable for their performance, which in turns limits citizens' ability to hold local politicians accountable. The frequent rotation of these officers further limits accountability to local decision makers and the autonomy of local decision-making. In Kamataka and Maharshatra, financial accountability and transparency of ULBs are generally weak. Municipal budgets are rarely prepared in a uniform format and budget data are often inaccurate. Cash accounting systems are common and are ill-suited for accurately representing the financial position of ULBs; local audits are often performed with delays. Karnataka is in the process of introducing double-entry accounting systems to address these shortcomings. In comparison, Tamil Nadu's ULBs have higher financial accountability and transparency. Detailed guidelines are used in preparing and presenting their ULB budgets, each of which must show a consolidated budget surplus. Tamil Nadu is relatively unique among Indian states in having introduced double entry, accrual accounting in all ULBs in 2000-01; the TNUDF played a pivotal role in introducing and supporting this reform. There seems to be little systematic monitoring and evaluation of the performance of ULBs nor of the projects developed by various boards and authorities in Kamataka and Maharashtra. A noteworthy exception is Bangalore Corporation's negotiation of a performance contract with the Government of Karnataka, which links resource flows with clearly defined performance criteria.' Tamil Nadu is more proactive in monitoring ULB performance, has developed a series of ULB performance indicators, and 'To date, the effectiveness of this contract in promoting performance has not been fully assessed. iv India: Urban Governance and Finance Review improving the quality of ULB fiscal and performance data is a high priority. The establishment of Freedom of Information Acts in Karnataka, Maharashtra and Tamil Nadu, and the reform of procurement processes in Karnataka and Tamil Nadu could also create an enabling environment for enhancing ULB accountability and transparency. Promising developments in promoting accountability to citizens include the use of citizen report cards, citizen charters and enhanced citizen participation in decision making and monitoring (i.e., Bangalore's Agenda Task Force, the Public Affairs Center's Report Card for Urban Services, Mumbai First, etc.) The Bangalore Citizen Report Card is a model for bringing the power of transparency to bear on improved service delivery. OPTIONS FOR REFORM The challenge of urban finance and governance in India is a classic example of incomplete decentralization. Responsibility doesn't match financing, accountability and capacity are weak, and local autonomy and state control are not balanced in ways that create incentives to improve performance. Indeed, these problems are not unique to ULBs in India, but are shared by panchayat raj institutions as well.2 The challenges for rural decentralization are surprisingly similar: limited expenditure discretion; overlapping responsibilities; weak revenue effort and low collection rates; complicated, non-transparent and non-equalizing transfer systems; outdated budgeting and accounting systems; and inadequate information and monitoring systems. As a result of this incomplete decentralization, lackluster ULB performance impedes service delivery, discourages investment, and stifles the economic potential of cities. Improving cities' access to financing is necessary, but not sufficient to overcome these challenges. Nor is it possible to fix the intergovernmental system all at once. A systematic approach is needed over the medium to long-term to improve ULB performance, including state and local actions to enhance fiscal sustainability, and strengthen institutional arrangements to promote accountability and performance. Specific state and local actions needed for this approach are summarized in the table below. Reforms in the overall fiscal system - especially fixing the problem of significant and growing state deficits -- are prerequisites for unleashing the potential of Indian cities. 2 See the World Bank, India Fiscal Decentralization to Rural Governments (2003). v India: Urban Governance and Finance Review Potential State and Local Reform Options Short-term Medium-term Long-term State Actions Fiscal Improve reporting and monitoring of Ensure regular and predictable payment of state De-link local salaries from national Sustainability outstanding ULB arrears (consider TN transfers for recurrent spending wage rates experience) Adhere to equalization formula in allocating Rationalize cesses vis. adjustments Introduce accrual accounting in transfers based on need and capacity in transfers rather than as a part of Maharashtra, and support full Show progress on URIF requirements: phase- local property taxes implementation in Karnataka out rent control laws; repeal Urban Land Assess whether rate setting Identify options for strengthening state Ceiling and Regulation Act; rationalize stamp autonomy is needed, or other fiscal sustainability duties, and reduce rates in a phased way revenue bases should be available to Identify alternatives to intercepting arrears, ULBs including linking sanctions to outstanding Based on successful pilot arrears (i.e., turn off electricity when bills experience, apply sanctions rather aren't paid), and pilot in selected ULBs than intercepts to outstanding arrears Promote responsible local borrowing - stop bailouts, lend to creditworthy ULBs, consider Work out local debt overhang establishing a financial intermediary in (consider TN experience) Maharashtra and Kamataka Institutional Review approval thresholds and raise Reduce frequency of transfer of local officials Allow local officials to hire and fire Arrangements those that unduly restrict local decision Clarify and reduce the role of state boards and their own employees, and set local making authorities in local investment decisions; align wage rates Develop indicators and enhance state decisionmaking for selecting and financing Allow locally elected representatives capacity to monitor local performance investments with maintaining those investments to appoint ULB executives Provide technical assistance to ULBs to Develop systematic approaches/vehicles for Allocate a substantial portion of enhance their fiscal reporting strengthening local capacity (consider TN ULB resources based on the experience with TNUDF) achievement of performance criteria Assess impact of Bangalore's Performance Contract with GoK, and identify other options for incentivizing improved performance India: Urban Governance and Finance Review Local Actions Fiscal Working with local stakeholders, identify Improve property tax administration (i.e., Improve average collection ratio of Sustainability areas for improved service delivery or registration, billing and collection) and deepen property taxes to demand reduced cost coverage of property tax reforms Update and maintain valuation rolls Streamline processes and procedures to Improve cost recovery of local utilities through on a regular basis to reflect reign in costs improved administration and stronger economic growth Link taxes and charges to services enforcement provided In Maharashtra, identify suitable alternatives to Consider public-private partnerships Improve the accuracy and timeliness of the octroi for municipal corporations in providing utility services fiscal reporting Enhance taxpayer services and reduce the transaction costs of paying taxes and fees Develop managerial and financial capacity to manage local borrowing In Kamataka and Maharashtra, improve the timeliness and completeness of local audits Institutional Develop indicators and engage local Encourage competition among local agencies in Strengthen accountability of local Arrangements stakeholders in monitoring local providing services politicians to citizens performance (consider Bangalore Agenda Develop local capacity for investment selection Task Force and Mumbai First experience) and construction management Include performance indicators in local Right size low-skilled employees budget Develop systematic approaches to enhance capacity and skill mix of local employees vii India: Urban Governance and Finance Review 1. INTRODUCTION Cities play a critical role in India's development. While its 1 billion-plus population is predominantly rural, over 300 million people live in urban areas. One-third of this population lives in 35 urban agglomerations or cities with populations exceeding 1 million. The share of GDP generated in urban areas has increased over the past 10 year, and urban poverty levels have declined.3 Cities are responsible for delivering various public services, yet severe infrastructure shortages in water supply and sanitation, roads, transportation, housing and waste management, and inefficient management have resulted in poor quality services with limited coverage.4 These inadequate services and worsening environmental conditions disproportionately affect the poor. Between 1950 and 2000, India's urban population increased from 62 million to 288 million -- an average annual increase of about 3 percent (UN 2000). Even though urban growth rates are expected to drop by about half a percentage point over the next three decades, the total urban population is likely to reach more than 600 million by 2030 (Figure 1). Figure 1: Urban population in India, 1950-20305 That growth trend means that most cities will more 7030 _ than double in size. Already strained to provide 600 services and quality of life to existing urban residents, s 5(30 cities will face tremendous challenges in expanding existing infrastructure and avoiding deterioration of living standards due to congestion, pollution, and lack of basic services. A doubling of population over 30 -o 200l _11 years means that by 2030 there will be a second noo 100 - Mumbai, a second Calcutta, and a second Bangalore o 1 _ - that must be fed; supplied with water, sanitation, and electricity, given public and private transportation options; and whose garbage must be disposed of Year (Deichmann et. al, 2003). Indian cities' responsibility for delivering local services is enshrined in the 74h Constitutional Amendment Act (CAA). Decentralization has been underway since the early 1990s, when the Government of India passed the 73rd and 74"' CAAs, which directed states to devolve some of their responsibilities to panchayats and urban local bodies (ULBs). In the decade since passage of these amendments, however, decentralization has been limited, with slow and uneven progress across states. While the CAAs envisioned decentralization of functions, finances and functionaries to enable panchayats and ULBs to function as "institutions of self government," in reality, fiscal and administrative decentralization have lagged behind political decentralization.6 It is an appropriate time to review the financing and governance of ULBs. The 12th Finance Commission has been constituted, and the possibility of revising the 74h Amendment to strengthen the legal foundation for ULB self governance is being discussed. Awareness of the importance of cities and 3 Urban poverty declined to 32.4 percent in 1993-94, from 38.2 percent in 1987-88; see the World Bank, India Urban Sector Strategy (June 2001.) The Strategy notes that urban poverty is less well analyzed than rural pverty. 4 The 1996 Rakesh Moban Infrastructure Report estimated that annual investments of about US$5 billion (1.5 percent of GDP) are needed to make up for past underinvestments and future demands on urban areas. 5 Source: United Nations World Urbanization Prospects, 1999 Revision. 6 For a discussion of ULB progress in decentralization, see National Commission to Review the Working of the Constitution, A Consultation Paper on Decentralization and Municipalities (September 2001), and for rural decentralization, see The World Bank, Overview of Rural Decentralization in India (September 2000). 1 Executive Summary and Main Report the need to improve their performance is growing, as evidenced by the priority given to them in the 1Oth Plan. The intergovernmental environment shows encouraging trends toward improved fiscal performance, as evidenced by the passage of fiscal responsibility legislation (with clearly defined fiscal targets) in Kamataka and Tamil Nadu, among three other Indian states (World Bank, 2004). Equally important, ULBs are introducing reforms to improve their service delivery and strengthen their local revenue mobilization, and much can be learned about applying these innovations more broadly.' Considerable work has documented finance and governance challenges in Indian states and panchayats; this Review complements this work by focusing on ULBs.8 The objectives of this report are to: (i) describe the finance and governance aspects of ULBs in three states; (ii) analyze why institutional arrangements or (dis)incentives have promoted or impeded expected outcomes, and (iii) assesses the implications of these innovations/impediments on financing and delivering ULB services. The audience for this Review includes Bank staff members working on India, as well as counterparts and researchers interested in comparative analysis of ULBs. 2. ANALYTICAL FRAMEWORK Given the diversity in the intergovernmental and institutional arrangements across Indian states, generalizing about all ULBs is very difficult. This report analyzes ULBs in three states - Karnataka, Maharashtra and Tamil Nadu -- in detail, and attempts to draw general lessons from these case studies. The proposed states have high urban populations, face different intergovernmental and institutional arrangements, are actively pursuing urban reforms, and there is extensive Bank and other experience that can be analyzed. Figure 2 shows projected urbanization rates in Karnataka, Maharashtra and Tamil Nadu. Of these three states, Tamil Nadu is the most highly urbanized (43.9 percent), Maharashtra has similar levels of urbanization (42.4 percent) and Karnataka is the least urbanized (34.0 percent). 7See, for example, Vaidya and Vaidya, "Management Innovations for Municipal Resource Mobilization in India," FIRE Draft Note 30, July 2002. s See for example, Stabilization and Fiscal Empowerment: The Twin Challenges Facing India's States (World Bank, 2004); India Urban Sector Strategy, India Country Assistance Strategy and Update (World Bank 2001, 2002); India . Evaluating Bank Assistancefor Pubic Sector Management - a Country Assistance Evaluation (OED, 2002)., Overview of Rural Decentralization in India (SASRD, 2000), Civil Service Rationalization In India: A Survey Of Issues And Options (World Bank Policy Paper, 2002), among others. 2 India: Urban Governance and Finance Review Figure 2: Urbanization Rates in Karnataka, Maharashtra, and Tamil Nadu r ff , ! . " P r '' w ":1 ' -' X i.1 7~~~~~ -4. NFK From finance and governance perspectives, the distinguishing characteristics of the three case study states are summarized below. Karnataka: Bangalore has successfully reformed its property tax, defined performance benchmarks through a Memorandum of Understanding with the GOK, and promoted community participation in decision making (Bangalore Agenda Task Force); GOK is introducing financial management reformns (double-entry accounting) statewide; has established a Freedom of Information Act and Procurement Act; and is building the capacity of Department of Municipal Administration to monitor ULB performnance and support ULB reforms. Ma ha rashtra: some education and health functions are municipal responsibilities; corporations have autonomous revenues (octroi); private sector construction and management contracts have been established (i.e., Pune water supply and sewerage); state incentive grants are provided for urban water supply projects. Tamil Nadu: SFC recomnmendations largely implemented; accounting reforns implemented statewide (i.e., introduction of double-entry accounting), improved auditing and computerization of financial accounts; private sector participation in selected city services (e.g. sewerage, street lighting), establishment of Tamil Nadu Urban Development Fund, a financial intermediary for ULB infrastructure finance. A sample of ULBs in each state were visited to collect data and conduct in-depth interviews. Population and summary fiscal indicators across these three states are presented below in Table 1. 3 Executive Summary and Main Report Table 1: Comparison of lB Population and Fiscal Indicators, Per Capita for Selected Years (in Rupees) Average Annual Growth Avg. Per Capita Total Per Capita Total Population* Revenues Expenditures Rev. Exp. Karnataka (Year of Data) 96/97 98/99 00/01 96/97 98/99 00/01 1996/97- 2000/01 Corporations (6) 428,832 455 1,009 784 550 997 944 14.6% 14.5% City Municipal Councils (40) 59,906 270 414 644 207 415 583 24.3% 29.5% Town Munic. Councils (91) 29,924 193 329 371 182 284 330 17.7% 16.0% Maharashtra (Year of Data) 95/96 97/98 99/00 95/96 97/98 99/00 1995/96- 1999/00 Brihanmumbai (BMC) (1) 9,910,000 1,548 1,974 2,763 2,073 2,967 3,811 15.6% 16.4% Corporations, excl. BMC (15) 673,929 1,059 1,431 1,786 1,101 1,576 2,310 14.0% 20.4% Munic. Councils -'A' (18) 205,218 715 875 1,033 832 1,149 1,590 11.0% 17.6% 60,326 Munic. Councils- 'B' (48) 578 729 912 667 905 1,123 11.7% 13.9% Munic. Councils- 'C' (163) 21,230 455 597 778 553 661 870 14.3% 12.0% Tamil Nadu (Year of Data) 95/96 97/98 99/00 95/96 97/98 99/00 1995/96- 1999/00 Corporations (6) 1,186,667 429 699 1,037 418 484 1,193 18.3% 24.1% Special Municipalities (13) 179,370 348 506 716 592 375 606 18.7% 17.9% Selection Municipalities (28) 109,351 236 412 948 448 311 871 19.9% 21.8% Grade I Municipalities (36) 63,871 207 368 681 334 288 629 21.7% 21.2% Grade II Municipalities (25) 37,132 156 312 772 269 253 696 23.1% 23.0% Note: Number of ULBs listed in parentheses; data calculated by authors. A few general trends are worth noting at the outset. First, what constitutes an "urban local body" varies significantly across states, and even within a given state.9 Tamil Nadu, the most urbanized state in India, classifies municipalities by average income rather than population. As a result, population size varies largely within each ULB class. For example, the largest Grade I municipality in Tamil Nadu, Aruppukottai, has a population of 93,820 while the smallest, Sathyamangalam, has a population of 34,000. Corporations in Karnataka, the least urbanized state in the sample, generally have smaller populations (average of 428,832) than those in Maharashtra (average of 673,929, excluding Mumbai) or Tamil Nadu (average of 1,186,667). Mumbai Corporation (also known as Brihan Mumbai Corporation, or BMC) in Maharahstra could dominate the entire sample with its 10 million-plus-population.10 At the same time, Maharashtra also has Municipal Council "C" class ULBs with an average population of 21,230. In general, the performance and institutional arrangements for corporations - the largest ULBs - are quite different from other ULBs in scope, authority and magnitude. In contrast, many smaller ULBs, such as Town Panchayats and Town Municipal Corporations in Kamataka, are too small to manage water 9 State Acts define the roles and responsibilities of types of ULBs; see annexes in Volume II for more information. '° Using 2001 data, its population is close to 12 million. As a result, BMC is reported separately in the analysis. 4 India: Urban Governance and Finance Review supply systems or solid waste disposal systems, and these services could be provided by some larger unit to tap economies of scale and scope. Second, the quality of fiscal data for ULBs is generally poor. Except for Tamil Nadu, reported data for ULBs are not audited, nor are they necessarily consistent across years nor across states. For example, current (i.e., revenue) expenditures in Maharashtra and Karnataka generally cannot be distinguished from capital expenditures." Except for the State Finance Commissions which are convened quinquenially, detailed data are not collected regularly, nor do states monitor ULB finances carefully. Because of these inaccuracies, the analyses presented below may have high standard errors, hence in most cases, general trends are discussed, with selected empirical analyses reported. Third, ULB total expenditures and total revenues have grown substantially, and on a per capita basis between 1996/97 and 2000/01 (see Table 1).12 Average annual growth rates in per capita expenditures for many ULBs exceed their per capita revenue growth. This gap is particularly pronounced for corporations in Maharashtra and Tamil Nadu. This trend could indicate that corporations are financing capital expenditures via borrowing or other sources of financing (which are not reflected in total revenues), the impact of the 5b Pay Commission, or perhaps an underlying structural imbalance between the demands on ULBs to provide services and their ability to finance those services. Relative to the other states analyzed, Maharastra ULBs have higher spending and revenue collections, in part because of their broader expenditure responsibilities and sources of revenue. Fourth, Indian ULBs play a relatively limited fiscal role in comparison to selected, other countries.'3 International fiscal comparisons are fraught with problems of non-comparability across data sources and institutions, and such comparisons should be interpreted with great caution. Nonetheless, a comparison of per capita total expenditures and revenues in corporations in the case study states with the average per capita local expenditure and revenue in Poland, Russia, Brazil, Mexico or South Africa (see Table 2) shows that Indian ULBs spend and collect considerably less expenditure and revenues per capita than all of the comparison countries except for Mexico, which is roughly similar to Maharashtra.'4 Corporations in Maharashtra spent an average of $58 per capita, well below Russia's average local expenditure of $349 or Poland's average local expenditure of $358. These expenditure disparities reflect in part the income differentials, but also the fact that Indian ULBs do not have significant responsibility for health and education services, which are an important local responsibility in the countries listed in the table. Similar disparities hold in per capita revenue collections as well. 1 l The Maharashtra Urban Infrastructure Fund report (Kirloskar Consultants, 1998) noted a major issue in municipal accounting is that "there is a general mix up of current and capital works, especially in municipalities." 12 Total expenditures include current and capital expenditures, whereas total revenues include own source revenues and state transfers, but not sources offinancing, such as borrowing, drawing down cash balances or other means of "below the line" financing. 13 While these countries have higher per capita incomes than India, they were chosen because subnational fiscal data were available, and local governments play an important role in these countries. 14 The fiscal indicators for the comparison countries are averaged across all local governments within a country, but exclude provincial or state government fiscal data. In contrast, the Indian data reflect only spending and revenue collection by ULBs. The GFS data do not distinguish municipalities from other local governments. 5 Executive Summary and Main Report 3. INTERGOVERNMENTAL FISCAL RELATIONS As discussed below, the three states have weak linkages among the components of their intergovernmental systems and soft budget constraints.'5 Indeed, liabilities and arrears are endemic in many areas. Table 2: Comparisons of Local Per Capita Expenditure CENTRAL ROLES IN URBAN AFFAIRS and Revenue, Selected Countries (in US$) India is a federation, with important Country/Local Governments Year Exp. Rev. intergovernmental roles for the central (or India - Kamataka ULBs 2000/01 $24 $20 "Union") government, states and ULBs. The India - Maharashtra ULBs* 1999/00 58 45 Constitution demarcates functions and India - Tamnil Nadu ULBs 1999/00 30 26 finances between the Central Government and state governments, and separate legislative, Poland LGs 1998 358 355 executive and judicial branches are constituted ussian Federation LGs 1995 349 347 at each level. The CAA clearly recognizes Brazil LGs 1994 149 140 that ULBs are "subjects" of the states, hence Mexico LGs 1997 51 50 state-local relations are relatively more South Africa LGs 1998 125 125 important than central-local relations. Nonetheless, the Central Government plays a *ULBs except for BMC significant role in funding ULBs through Source: IMF, Government Finance Statistics, various years; grants, various central schemes and the recent calculations by authors establishment of incentive funds. At the national level, the Ministry of Urban Development and Poverty Alleviation (UDPA), the Planning Commission, and the Finance Commission play key roles in urban finance and governance. The role of the UDPA is to provide guidelines for ULB reforms under the 74th CAA, including user fees and taxes, as well as to promote knowledge sharing among state governments. In addition to the Finance and Planning Commissions, central ministries provide assistance to states to implement central schemes.'6 The Constitution mandates that a (National) Finance Commission be constituted every five years to review central and state finances, and recommend the devolution of taxes and grants for the subsequent five years. Following the 74th CAA, the 10th Finance Commission (1996-2001) recommended that a grant of Rs. 10 billion be allocated to states based on the 1971 ratio of inter-state slum to urban population. The 11th Finance Commission (2001-2006) established a comprehensive framework for allocating Rs. 20 billion in grants for ULBs. These grants are meant to supplement the funds that would normally be transferred from state governments to ULBs in 2000-2005, as well as the amounts allocated to ULBs as SFC recommendations.'7 15 With a hard budget constraint; subnational governments are unable to transfer their liabilities (e.g., undelivered services, wage or contractor arrears, unpaid debt or debt service, under-used revenues, negotiated transfers, etc.) to higher levels of government. In other words, higher-level governments do not bail out local governments. See Jonathan Rodden and Jennie Litvack (editors), Decentralization and the Challenge of Hard Budget Constraints (MIT Press, 2003). 16 The States act as implementing agencies for central sector schemes, which are funded entirely by the Central Govemment. In contrast, centrally-sponsored schemes are shared-cost programs between the Central Government and the states, which must contribute to these schemes with various matching ratios, depending on the particular project. See Govinda Rao, Dynamics of Indian Federalism, draft manuscript, 2002. 17 For more information, see Om Prakash Mathur, Approach to State-Municipal Fiscal Relations: Options and Perspectives, National Institute of Public Finance and Policy, March 2001 6 India: Urban Governance and Finance Review ULBs receive grants from 30 central schemes and centrally-funded schemes for urban development and poverty alleviation, which are funded by the GOI, and in some cases, also by states on a shared basis. As shown above in Table 3, in aggregate, these schemes are substantially larger than the Finance Commission grants, and accounted for about Rs. 300 billion in the 10'h Plan. Outside of the National Capital Region, and other regionally targeted schemes, the largest schemes are for HUDCO, VAMBAY, the Integrated Development of Small and Medium Towns (IDSMT), the Mega City Programme, and the Accelerated Urban Water Supply Programme. Generally speaking, the Central Government is responsible for defining the policy framework for these schemes; financing is shared among the central, state and ULB governments; ULBs are responsible for implementing the schemes; and states are responsible for monitoring performance. Evaluation of centrally-sponsored urban schemes has found that they are often underutilized, poorly targeted and limited in achieving their objectives. For example, often less than 75 percent of the support allocated for centrally sponsored programs for low- income housing, the Mega Cities Programme and other, selected programs has been utilized, and existing GOI Slum Improvement Programs tend to be inefficient and poorly targeted.'8 In addition to the grants and schemes mentioned above, a key, Central Government initiative to promote urban reforms is the Urban Reform Incentive Fund (URIF). The URIF, announced by the Union Finance Minister in his 2002-03 Budget Speech, proposes grant funds as a means to encourage state and municipal reforms, to promote an enabling environment for more self-reliant ULBs, and to strengthen ULBs' resource mobilization and accountability. The URIF is not yet operational, but is expected to have an outlay of Rs. 500 crores per year. The key reform measures for the initial phase of the URIF are: * repeal of Urban Land Ceiling and Regulation Act (state) * rationalization of stamp duties (phased reduction in state rates) * reform of Rent Control Laws (state) * introduction of computerized registration processes (ULB) * reform of property taxes and improvement in collection efficiency (state/ULB) * improvement in cost recovery of user charges (state/ULB) * introduction of double-entry accounting systems (state/ULB) The City Challenge Fund (CCF) is also under discussion as another instrument to encourage urban reforms, especially in assisting ULBs in becoming more credit-worthy by partially financing the cost of developing and implementing reform programs. The CCF is expected to be awarded on a competitive basis based on eligibility and award criteria, on-site assessment, disbursement conditionalities, and on-going monitoring. The use of incentive funds to promote urban reforms is part of a broader trend toward incentivizing improved fiscal performance in India. In addition to the two urban incentive funds noted above, the Central Government also provides a Fiscal Reforms Facility to reward states for undertaking fiscal reforms. While the impact of these initiatives on improved performance has not been fully evaluated, they are unique in introducing a performance dimension to fiscal reforms. 18 See Real Estate Reforms. Bringing India's Cities into the Economic Liberalization Program (World Bank, 2003), which provides figures on utilization rates for various programs, as discussed in UDPA The Annual Report 2001-2002 7 Executive Summary and Main Report Table 3: 10th Plan Outlay for Central Schemes and Centrally Sponsored Schemes for Urban Development, Urban Employment and Poverty Alleviation (Rs. In billions) Centrally Central Sponsored Schemes Schemes National Capital Region Planning Board 41.22 Urban MappingO 0.20 Research in Urban and Regional Planning 0.15 Urban Transport (Delhi Metro Rail Corporation and other) 20.16 Equity to HUDCO for Urban Infrastructure 1.00 Computerisation 0.16 Training in Public Health Engineering 0.10 Equity to HUDCO for Water Supply 1.00 General Pool Residential Accommodation 4.00 CPWD Training Institute and North Eastern Zone 0.15 General Pool Office Accommodation 1.50 Modernization of CPWD/Computerisation 0.30 National Urban Information System 0.20 Pool Finance Development Facility 4.00 City Challenge Fund 5.00 North East - Lump Sum Provision 7.00 Integrated Development of Small and Medium Towns 13.05 Mega City 10.50 Low Cost Sanitation 2.00 Extension of Accelerated Water Supply Programme to Small 9.00 Towns Solid Waste Management 0.99 Total, Dept of Urban Development 76.94 44.74 Total Central and Centrally Sponsored Schemes 121.68 Building Material Technology Promotion Council 0.20 Development of Urban Indicators Programme 0.01 HUDCO 145.01 National Cooperative Housing Federation 0.01 Lump Sum Provision for NE Region 4.05 VAMBAY 20.43 SJSRY 5.41 Night Shelter Scheme 0.319 Infrastructure Facilities in the Displaced Persons Urban 0.08 Colonies in West Bengal Total Dept of Urban Employment & Poverty Alleviation 149.28 26.23 Total Central and Centrally Sponsored Schemes 175.51 Source: GOI, Planning Commission, 10th Five Year Plan (2002-2007) 8 India: Urban Governance and Finance Review STATE GOVERNMENT ROLES IN URBAN AFFAIRS Since ULBs are state subjects, state governments play a significantly greater role in regulating and financing ULBs than the Central Government. ULB operations are governed by the provisions in their State Acts, which define the powers of local bodies, including sources of revenue, autonomy in taxation, and autonomy in determining expenditure priorities. For example, state governments determine the number and type of ULB employees, as well as their compensation (through the issuance of wage guidelines); oversee ULB budget preparation; issue guidelines and standards for service provision; and issue clearances/concurrence for various financial activities (e.g., procurement) and regulatory issues. Paralleling the Ministry of Urban Development and Poverty Alleviation, each state has a similar body - e.g., the Department of Municipal Administration - that is responsible for overseeing and regulating ULBs. States play a substantial role in financing capital investment. In 2000/01, they accounted for 57 percent of India's total government capital expenditure (World Bank, 2004). They also provide considerably more financing to ULBs than the Central Government, although their flows to ULBs depend in large part on their own fiscal health. For example, in Karnataka in 2001-02, over 95 percent of the most significant transfers were provided by the State: State Plan Schemes (Rs.685.8 crore); State Finance Commission grants (Rs.590 crore); versus the Central Schemes for urban development and urban water supply (Rs.57.6 crore). The persistence of large and growing fiscal deficits in the States - amounting, on average, to over 4 percent of state GDP (World Bank 2004; Rao, 2002) - has negatively affected ULB financing; delays and reductions in the transfers paid to ULBs are commonly observed in the case states. Given the high reliance of ULBs on state transfers, their fiscal problems cascade onto ULBs, resulting in local fiscal arrears, or further deterioration in local service levels. EXPENDITURE ASSIGNMENT AND LEVELS The 74th CAA defines an "illustrative" list of 18 municipal responsibilities and functions (see Box 1), leaving individual states and SFCs to determine municipal functions, assign expenditures and revenues, and define the financial relations between the state and ULBs.'9 Expenditure assignment is relatively similar across the three states, except for the devolution of some education and health services to ULBs. In Maharashtra, ULBs have obligatory functions for public hospitals and dispensaries, vaccination, epidemic control and prevention of dangerous diseases, medical relief, family planning and welfare etc., as well as primary schools. ULB's in Tamil Nadu have some responsibility for public health and conservancy. ULB's have minor education responsibilities in Tamil Nadu (mostly maintaining school buildings), and education and health responsibilities in Karnataka (formation and maintenance of schools and hospitals, except for selected responsibilities in Bangalore). '9 Revenue and expenditure assignments in the lists are concurrent with the States' responsibilities. Actual devolution of specific revenue sources and expenditure functions to ULBs depends on the willingness of the State government to devolve functions and powers. 9 Executive Summary and Main Report Box 1: Urban Local Body Functions as Defined in the 74th CAA Urban (town) planning Slum improvement and upgrading Regulation of land use and construction of Urban poverty alleviation buildings Planning for economic and social development Provision of urban amenities and facilities (e.g., parks, gardens, playgrounds) Roads and bridges Promotion of cultural, educational and aesthetic aspects Water supply Burials et al. Public health, conservancy and solid waste Cattle pounds, prevention of cruelty to animals management Fire services Vital statistics Urban forestry, environmental protection, Public amenities (street lighting, bus stops, public ecological promotion conveniences) Safeguarding disadvantaged groups Regulation of slaughterhouses and tanneries Compared to urban counterparts in other countries, ULBs play relatively small roles and their functions are narrowly circumscribed and confined generally to the provision of infrastructural services such as water and sewerage, sanitation, solid waste removal and street lighting. This limited role is reflected in the small share of budgetary expenditures accounted for by ULBs, only 3 percent of consolidated government expenditures, or 0.8 percent of GDP in FY 1997-1998 (Rao, 2000). This share is low even in comparison to countries with few social service responsibilities, like Morocco (3.3 percent of GDP in 1998) or Greece (2.1 percent of GDP in 1998).2o Social services that typically absorb major portions of expenditures in municipalities in other countries (and which are core concerns of local legislators), such as primary and secondary education, health, and public safety, largely remain state functions. Major infrastructural services such as electric power, gas pipelines, and public transport also are not typically municipal functions. ULB Spending Levels and Composition Total expenditures (as shown in Table 1) can present a distorted picture because of the "lumpiness" of capital spending.2' Table 4 compares the current (or "revenue") expenditures per capita of ULBs, spending on "core services" and "basic public goods," and total water supply expenditures. Core services are defined uniformly across states, and include water, sanitation, street lights and roads. By this measure, ULBs in Maharashtra (especially corporations) provide more core services than other states. Karnataka's ULB spending on core services is particularly low, given that their values in Table 4 include capital spending. Because of differences in ULB functional responsibilities (notably for health and education) and the way that fiscal data are reported, basic public service expenditures vary more across these states.22 Here, again, corporations in all three states spend more per capita on basic public services, ranging from a high of Rs. 1,489 in BMC and Rs. 1,101 in other Maharashtra corporations, to an average of Rs. 743 in Tamil Nadu corporations, and an average of Rs. 408 in Karnataka corporations. ULBs in Maharashtra and Tamil Nadu generally spend more per capita on basic, public services than those in Karnataka. This is not surprising given their broader responsibility (i.e. for health and primary education in Maharashtra, and for health in Tamil Nadu). 20 See, for example, Morocco Municipal Management and Decentralization (World Bank, 2001). 2'Because fiscal data in Maharahstra were not separated into current and capital expenditures, we defined "current expenditures" as total expenditures minus expenditures on "roads", "sanitation and solid waste" and "other expenditures." 22Basic public services include core services, as well as public health, conservancy and cleaning in Karnataka; education, public health and conservancy, and fire brigade in Maharashtra; and education, public health and conservancy in Tamil Nadu. As noted above, education functions in Kamataka and Tamil Nadu aremainly school maintenance. 10 India: Urban Governance and Finance Review In all three states, a key Table 4: Categories of Expenditures per Capita, Selected States (in Rupees) function for ULBs is Per Capita Expenditures supplying water, and yet in some cases, state Basic Total capital spending Current* Core Public Water dominates the sector. Services Services Supply During the 10th Plan Karnataka (2000-01) period, the Government Corporations 944 249 408 47 of Kavrataka spent an City Municipal Councils' 583 211 401 70 capita annually (or Rs. Town Municipal Councils 330 86 133 21 215 per capita in Maharashtra (1999-00) Bangalore) on water Brihanrumbai (BMC) 3,811 1,360 1,489 384 supply, whereas the 2,310 966 1,101 456 average ULB spent Rs. 47 Corporations, excl. BMC 2,310 496 1,101 145 per capita and Bangalore Municipal Councils -'A' 1,590 419 580 154 spent an average of Rs. 46 Municipal Councils - 'B' 1,123 370 463 197 per capita. The situation MuniipalCounils 'B is somewhat more Municipal Councils - 'C' 870 240 317 88 balanced in Tamil Nadu, Tanil Nadu (1999-00) where the average ULB Corporations 804 307 743 76** Rs. 80 per capita on Special Municipalities 788 261 589 266 water, whereas the Selection Municipalities 571 230 478 130 Government of Tamil Grade I Municipalities 458 158 425 100 Nadu spent Rs. 71 per Grade H Municipalities 396 131 365 26 capita. Only in Maharashtra is *Total expenditures are reported for Karnataka since the breakdown between most of the spending on revenue ("current") expenditures and capital expenditures is not available for 2000- water supply undertaken 2001. by ULBs. The **Does not include Chennai, which has a separate water authority, Chennai Government of Metropolitan Water Supply and Sanitation Board (CMWSS) Maharashtra spent an average of Rs. 86 per capita (or Rs. 120 per capita in Mumbai), while the average ULB spent Rs. 325, and BMC spent Rs. 384 on water supply. About 20-40 percent of ULB spending is accounted for by salary payments, which are outside of the control of ULBs. In Karnataka, wage and salary payments account for about one-third of total expenditures of corporations, 29 percent of CMC total expenditures and 42 percent of TMC total expenditures. In Maharashtra, BMC spent the highest share of total expenditures on administrative costs (49 percent), whereas other corporations spent an average of 18 percent on administrative costs, and municipal councils spent between 25-32 percent on administrative costs. The average Grade II Municipality in Tamil Nadu spent 42 percent of current expenditures on establishment expenditures while the average Corporation spent 24 percent. A general trend across the three case study states is that smaller ULBs spend a larger share of their budgets on salaries. State governments, on average, spend 30 percent of their budgets on salaries, and state employee salaries are significantly higher (over 100 percent) than private-sector salaries (World Bank, 2004). ULBs have no say in determining their employees' 11 Executive Summary and Main Report pay, which is determined by the state.23 States also Table 5:Deterninants of Per Capita control the number of local employees, and the Expenditures, Selected Indian Corporations distribution of workers among categories. Interviews Ln(Exp per capita) with ULB officials revealed perceptions of over-staffing, Octroi (1= Yes: 0 = No) 0.787 (0.307)* and many officials supported state-defined limits on Ln (Transfers per capita) 0.471 (0.094)** staffing. Ln (Slums per capita) -1.643 (1.674) Ln (Mfg Emp per capita) -3.29 (3.439) Explaining the Variation in Current Constant 4.375 (0.511)** Expenditures Observations 51 -squared 0.45 Regression analyses were performed to explain + significant at 10%; * significant at 5%; ** the variation in current expenditures per capita within significant at 1%; standard errors in parentheses; each of the three states.24 It was expected that ULBs with none of the variables exhibited any significant more resources (as measured by their economic base and signs of collinearity. transfers from the state), and greater "need" (as measured Source: Calculated by authors by population size and share of the population living in slums) would be associated with higher spending per capita.25 EXPEND_PC = f(economic base, transfers, size, poverty) In Tamil Nadu, these variables explained differences in per capita spending as expected, with transfers per capita having the greatest explanatory power. Every 10 percent difference in per capita transfers from the GoTN is associated with a 2.1 percent difference in per capita expenditures. Per capita expenditures do not appear to be targeted to the poor, as the share of the population living in slum areas is not statistically different from zero. The model explains only 23 percent of the variation in current expenditure per capita. Only in Maharashtra was a positive and significant relationship found between expenditure per capita, the economic base and the population of the ULB. Every 10 percent difference in population is associated with a 2.1 percent difference in expenditures per capita, and every 10 percent difference in the economic base is associated with a 3.6 percent difference in local spending. The slum population, and grants per capita were not statistically significant. Holding constant the effect of types of ULBs, 40 percent of the total variation in current expenditure could be explained. In Kamataka, differences in current expenditures per capita were related to per capita grants and size, with every 10 percent difference in grants per capita associated with a 2.7 percent difference in expenditures per capita. About 42 percent of the variation in current expenditure per capita could be explained.26 Since the economic base was significantly related to per capita expenditures only in Maharashtra, further analysis was done to test whether this relationship reflected reliance on the octroi, a tax whose base reflects economic activity. The finances of 14 municipal corporations were analyzed over time to 23 The National Pay Commission affects state salaries, which in tum affect local salary levels. 24 These estimates should be interpreted with caution, due to the poor quality of fiscal data and the potential inconsistencies arriving from this simple estimation technique. Ideally, a simultaneous equation model should be estimated to capture the simultaneous determination of expenditure levels and local revenues, based on the independent variables listed above. 25 The economic base was measured by manufacturing employment or the number of industrial properties per capita in the ULB, as income and other measures of economic activity are unavailable at the ULB level. 26 The slum population wasn't available at local levels in Kamtaka, so a simpler specification was estimated. 12 India: Urban Governance and Finance Review determine which factors accounted for variations in current expenditures per capita.27 Not surprisingly, per capita current expenditures varied considerably across these corporations. The average, current expenditure per capita was Rs. 827; ranging from a low in Pune (Rs. 7.6) to a high in Udaipur (Rs. 5,593). In Table 5, the regression coefficients show that, holding other factors constant, per capita expenditures are about 79 percent higher in municipal corporations with octroi. While this figure seems high, it implies that corporations that collect octroi spent about Rs. 2,200 more per capita. It is generally recognized that the octroi is an economically inefficient revenue source, that introduces many distortions in economic activity, and its use has been phased out in many Indian states. Nonetheless, one possible explanation for this finding is that the autonomy and regular (often daily) cash flow of the octroi enable additional spending. Further, every 10 percent difference in state transfers per capita is associated with a 4.7 percent difference in current expenditures per capita. The other two variables - share of population in slums, and per capita manufacturing employment -- do not appear to have any statistically significant effect on per capita expenditures. The model explained about 45 percent of the variation in per capita corporation expenditures. In sum, ULBs in Maharashtra spend relatively more on public services than those in other states, in part due to their broader responsibilities. Maharashtra corporations spend significantly more, in part because of their octroi revenues. In contrast, Karnataka ULBs spend relatively little on public services, although their responsibilities are defined more narrowly than in Tamil Nadu or Maharashtra. Local salaries account for 20-40 percent of ULB spending in the case states, although in most cases these shares are similar to state shares spent on salaries. Even for some services that are clearly a local responsibility (e.g., water supply), state capital spending is significant in some states. Spending differences across ULBs within a state do not appear to be related to the poor, as measured by the share of the population in slums. REVENUE ASSIGNMENT AND YIELD Table 6: Comparison of the Distribution of ULB Current Revenues Current Own Property The 74th CAA required that Revenue Source Tax/Own SFCs be established to consider revenue Per Rev. / Source assignments to rural and urban local Capita Current Revenue bodies. However, most SFCs have not (Rs) Rev. (%) (% assigned new revenue bases to local Karnataka 2000-01 authorities, and ULBs have few, Corporations 362 61% 55% autonomous sources of revenue. City Municipal Councils 80 37% 25% Nationwide, ULBs account for less than Town Municipal Councils 68 53% 17% 3 percent of total revenues of all tiers of Maharashtra 1999-00 government, with the central Brihanmumbai (BMC) 2,619 95% 29% government accounting for 62 percent Corporations, excl. BMC 1,628 90% 19% and states for 35 percent (Rao, 2000). Municipal Councils - 'A' 232 22% 49% Municipal Councils - 'B' 263 28% 48% Table 6 compares current Municipal Councils - 'C' 195 24% 41% revenues for ULBs among the three Tamil Nadu 1999-00 Corporations 547 57% 56% states. A few trends stand out: first, Special Municipalities 464 65% 59% per capita current revenues are Selection Municipalities 346 62% 58% considerably higher in Maharashtra Grade I Municipalities 274 57% 49% Grade II Municipalities 214 54% 48% 27 The corporations analyzed included: Ahmedabad, Surat, Belgaum, Bangalore, Indore, Jabalpur, Nagpur, Pimpri Chinchwad, Pune, Mumbai, Shillong, Jalandhar, Udaipur, Coimbatore. Data were obtained from obtained from various issues of the Statistical Abstract of India. 13 Executive Summary and Main Report corporations (Rs. 2,619 for BMC and Rs. 1,628 for other corporations) than other corporations or ULBs, due to their reliance on the octroi. Second, ULBs in Tamil Nadu, on average, collect more current revenues per capita, and are more self-reliant than ULBs in other states. Third, corporations generally rely less on grants than other ULBs. In Maharashtra, corporations generate about 90 percent of their own-source revenues, whereas corporations in Kamataka and Tamil Nadu generate less than two-thirds of their revenues from their own sources. In contrast, municipal councils in Maharashtra are particularly dependent on state grants, relying on them for as much as three-quarters of their current revenue. (As discussed below, many of these grants are octroi replacement grants). ULB Taxing Powers in Karnataka In Karnataka, the property tax is the most important source of own revenue, accounting on average for 53 percent of own revenues. (Octroi was abolished in Kamataka in 1976.) Property taxes are relatively more important for corporations (62 percent of own revenues) than for smaller ULBs (ranging from 30 percent of own revenues in City Municipal Councils to 28 percent of own revenues in Town Municipal Councils). Other, own revenues include user charges and fees (e.g., water charges, market and license), rents, advertising taxes (especially in large corporations like Bangalore), and miscellaneous receipts. Water charges are the main source of user charges, and in many cases, water rates are set by a ULB (the council must pass a resolution to establish water tariffs) but the receipts are remitted to the Water Board to pay for the capital costs of the water system. On average, water charges levied by ULBs in Karnataka are estimated to cover only 20-30 of operation and maintenance costs, leaving a significant gap in operating costs and no resources for investment in new infrastructure. In contrast, the Bangalore Water Supply and Sewage Board (BWSSB) which provides water supply and sewerage services for the Bangalore urban area, not only builds infrastructure, but also operates and manages the system and sets tariff rates and collect tariffs. The BWSSB has devised innovative ways to collect tariffs enhanced by information and technology systems and is able to achieve cost recovery of up to nearly 100 percent.28 It is the only water board in India that meters each of its 370,000 connections. It has strong collection enforcement; 106 percent of billing is collected. Bills can be paid by cash, bank account, credit card, and online in the spring of 2003. Water service is disconnected within two months of non-payment, and a Recovery Officer on deputation from the Revenue Department auctions moveable and immoveable property (i.e., cars, motorcycles) if payment isn't received by the third notice. Cesses are included in own revenues, even though these are collected on behalf of the GoK (although not always remitted to the State.) ULBs are responsible for collecting the water cess and remitting those funds to Kamataka Urban Water Supply and Drainage Board (KUWSDB) to repay their loans. However, ULBs often do not remit the cesses collected to the KUWSDB, resulting in the GoK often bailing out ULBs. Asset sales are very low, accounting on average for less than 1 percent of own revenues. Loans are relatively small in terms of financing, and are concentrated within corporations. Under Section 94 of 281ts current average cost of production is R 15.2 per 1,000 liters; it is recovering R 14 per 1,000 liters, or a cost recovery rate of 92 percent. 14 India: Urban Governance and Finance Review the Municipal Act, ULBs are required to obtain sanction from the State Government to revise tax rates, and to levy taxes at rates below the specified maximum. ULB Taxing Powers in Maharashtra The taxing powers of ULBs in Maharashtra include: (i) octroi or cess in lieu of octroi (only for municipal corporations), (ii) property tax; (iii) vehicle tax, tax on boats or animals, (iv) sanitary tax on private latrines cleaned by municipal agency; (v) drainage tax, (vi) water tax, and (viii) educational tax. In practice, ULBs have limited autonomy regarding defining tac rates or tax bases. Non-tax revenues are limited to parking fees, permit fees, service fees and user charges, rent from buildings and commercial complexes, development fees for granting permission to construct buildings on vacant plot, and other fees and charges. As noted earlier, the octroi is the most significant source of revenue for Maharashtra corporations, accounting for half of own-source revenue in BMC and even more for other corporations. It is indisputably an economically inefficient tax, that has been repealed for other ULBs in Maharashtra, and in other states. Despite its inefficiencies, the octroi has the advantages of generating revenues that grow in line with economic activity, and that offer a source of daily cash flow. Until other buoyant, certain and regular revenues are available for Maharashtra corporations, it is not recommended that the octroi be eliminated. Other ULBs in Maharashtra without access to the octroi are highly dependent on state transfers, and it appears that these transfers have discouraged local revenue mobilization. A regression of ULB own-source revenues per capita against population, the economic base, the share of population living in slums, and per capita state transfers, estimated in logarithmic form, found that all variables were statistically significant and 63 percent of the variation in own-source revenues per capita was explained. Larger ULBs - measured in terms of population and economic base - generated more own source revenues. ULBs with a higher share of the population living in poverty collected less own- source revenues. The strongest impact on own-source revenues was found to from per capita transfers - every 10 percent difference in grants per capita was associated with 5.3 percent decrease in own-source revenues.29 The property tax is the second most important own-revenue source, accounting on average for about 20 percent of corporations' revenues and almost 30 percent of BMC's revenues. Municipal councils rely more heavily on property taxes (in part because they may not levy the octroi), which generate between 4149 percent of own revenues. Water charges are also significant for ULBs in Maharashtra, and represent about 10 percent of revenues; because of low metering rates, water charges are generally linked to the rateable value of property. Karnik and Pethe (2001) document BMC's performance in collecting water charges (see Box 2 below). Over the period 1995-96 to 1998-99, the average demand created exceeded revenue expenditure by one-third. However, since recovery from demand was only 68 percent, only 59 percent of total expenditures (revenue plus capital) on water supply were recovered over this period. Except for Mumbai and Pune, other ULBs in Maharashtra typically operate water supply systems at a loss, incurring substantial deficits between operation and maintenance expenditures for water supply systems and income received. 29 See the Annexes in Volume 11 for more details. Similar regressions for Kamtaka and Tamil Nadu had poor explanatory power, and transfers were not statistically significant in explaining the variation in own-source revenues per capita. 15 Executive Summary and Main Report Box 2: Indicators of Water Supply Efficiency, Selected Corporations in Maharashtra Navi Mumbai ("New" Mumbai), Pimpri Chinchwad and Thane generate more water than their prescribed norms, with Mumbai generating the most water. Metering of water supply is uniformly poor in the corporations; Navi Mumbai has the highest ratio of metered to total connections (10 percent), which is low by intemational standards. Without meters, water charges are collected with the property tax on the basis of rough estimates of water consumption. Despite the fact that Navi Mumbai corporation has the highest per capita expenditure on water as well as the highest per capita revenues collected from water charges, it also exhibits the highest deficit (Rs. 506.) Navi Mumbai Pimpri Thane Per capita water availability (liters per day) 255.7 188.8 221.9 Norms adopted for water supply (liters per capita per day) 200 180 180 Number of users per connection 11.9 16.0 20.0 Proportion of metered connections to total (%) 10.5 2.8 4.8 Proportion of total revenue to total expenditure (%) 40.7 16.4 54.7 No. of municipal employees per MLD of water supplied 0.3 2.1 1.6 Number of connections per employee 981.4 158.1 141.1 Total expenditure on water supply per capita (Rs.) 852.3 332.0 402.3 Revenue collected from water charges per capita (Rs.) 346.6 59.3 220.1 (Total expenditure - revenue) per capita (Rs.) 505.7 272.6 182.2 Source: Kamik and Pethe, UNDP Report, Part C.: Assessment of Revenue and Expenditure Patterns in Local Bodies of Maharashtra ULB Taxing Powers in Tamil Nadu ULBs in Tamil Nadu have two types of own, current revenues: (i) taxes, non-tax revenues and fees that they collect (i.e., property tax, professional tax, rental income, water fees); and (ii) taxes and fees collected by other agencies (assigned or "shared" revenues). The property tax is the most important tax, accounting on average for 56 percent of own-source revenues and 34 percent of current revenues. Property tax revenues in Tamil Nadu generally do not keep pace with the growth in current expenditures. They include a general levy, as well as a series of specific levies (which are equivalent to cesses.) Property tax rates in Tamil Nadu are not subject to a ceiling. The profession tax accounted for 6-8 percent of own source revenue, and it has grown in line with current expenditures, and could generate additional revenues with proper administration and enforcement. ULBs in Tamil Nadu may also levy an advertisement tax, which is used mostly by large corporations (e.g., Chennai). Three years ago, and as part of the Entertainment Tax, ULBs were granted the right to collect the cable TV tax. This is a potentially promising source of revenue for ULBs, and cable TV revenues in Chinnurli are reported to be nearly as lucrative as property tax revenues (which, it should be noted, are fairly low). Assigned revenues mostly consist of the entertainment tax and the surcharge on stamp duty (or "duty on transfer of property"). The Chennai Metropolitan Water Supply & Sewerage Board (CMWSSB) provides water supply and sewerage services in the Chennai Corporation area. The Board collects water and sewerage taxes (levied as part of the property tax, up to 35 percent of the tax rate) and water charges. Twenty percent of the house service connections are metered, the remaining 80 percent are unmetered connections. Maintenance of water supply facilities is poor, and water supply is often erratic. CMWSSB has arrears exceeding Rs. 150 crores from consumers. The Tamil Nadu Urban Local Bodies Act allows for revision 16 India: Urban Governance and Finance Review of water charges once every three years. Deficits are common in financing water supply and sewerage in Tamil Nadu ULBs. The benefits of fiscal decentralization can be realized only if local governments are accountable and able to respond to the needs of taxpayers. An effective way to provide budgetary accountability and responsiveness is to grant local governments a meaningful degree of tax autonomy, especially in setting local tax rates. Through local tax autonomy, taxpayers become more aware of the costs of services and local officials' actions are subject to closer scrutiny by taxpayers. There are few instances (except for selected taxes in Tamil Nadu), where ULBs are allowed to adjust local tax rates and charges to changing economic circumstances and to respond to constituents' demands for more services (or lower taxes). However, the political unwillingness to levy local taxes to a significant degree is a greater impediment to accountability than the limited autonomy in defining tax rates. Equally disturbing is the poor performance of most ULBs in recovering user charges noted above. Such poor performance can be attributed to many factors: poor administration and enforcement of charges and fees; weak information systems for proper billing and collection; political unwillingness to impose full charges; and a culture of non-payment, related in part to the poor quality of services that are provided. Without improved cost recovery, the quality of services will not improve, and the possibility of engaging the private sector in providing these services is very limited. Property Taxes30 In most Indian states, property taxes are the mainstay of municipal finance. Table 6 showed that except for corporations in Maharashtra and municipal councils in Karnataka, most ULBs rely on property taxes for about half of their own-source revenues, although this share varies by state and individual ULB. Despite their importance, property taxes are not, at present, used to their full potential. An international comparison shows that India uses the property tax far less than other developing countries.3' This limited reliance on property taxes reflects multiple factors: the failure to decide on and implement a property tax structure that is revenue productive and equitable; weak administration (e.g., out-of-date valuation, poor registration) and enforcement; numerous distortions in land and property markets that limit the potential tax base; an unwillingness to assess property at its current market value; and corruption. India seems to be in a state of flux about how it should tax property. Bahl and Linn (1992) identify three basic forms of property taxation used around the world: (i) annual or rental value (ARV) of the property, (ii) capital value (CV) of the land and improvements, and/or (iii) site value of the land. Most Indian cites use ARV as the base of the property tax. Although a few cities have adopted a CV base for the property tax (including those in Kamataka, for example), implementation of the capital value system has not begun. The property tax base in Indian ULBs is more appropriately described as being the rental value of property.32 Various ULBs have adopted a unit value system (also referred to as an area-based system, or form of mass-appraisal), which in essence is a combination of the ARV and CV systems. In this system, values per unit of land (usually per square foot) are estimated and the tax base is the product of this unit 30 See the accompanying report, Urban Property Taxes in Selected Indian States, for a more detailed discussion of property taxes in Bangalore and Mumbai. 31 Roy Bahl "The Property Tax in Developing Countries: Where are We in 2002." Andrew Young School of Policy Studies, Georgia State University, 2003. Some would dispute this comparison on grounds that the IMF Government Finance Statistics, on which these comparisons are based, under-reports property tax data for India, and National Institute of Urban Affairs, Reforming the Property Tax (January 2001). 32 Technically, the base is the amount for which a property could be let by a willing landlord to a willing renter in a market free of encumbrances. 17 Executive Sumnmary and Main Report value and land area, plus the value of the structure. The latter is determined in an analogous way: a basic value per square foot is determined, weighted by construction quality and multiplied by area. This method has been used as a way to get around the limitations imposed on the property tax base by rent controls, or perhaps as a transition to a capital value system. The 11th Finance Commission recommended increased 'property and land based taxes' to enhance local resource mobilization. Various reforms have been introduced, including moving from an ARV to CV system (or variant thereof); improvements in assessment and collection processes; simplification of the registration and payment process; introduction of information technology; etc. that have substantially increased revenue collections in selected ULBs. In principle, self assessment has several benefits, including reducing costs of administration and reporting, as well as reduced opportunities for informal agreements between home owners and appraisers. Prior to the introduction of the self assessment scheme in Bangalore, for example, property values had not been reassessed for 30 years. When it introduced the self assessment scheme, property tax collections increased by 40 crore that year, with nearly 90 percent of this increase due to revised property values (see Box 3). Analysis of recently compiled household data for Bangalore and Pune indicate that the potential revenue yields from broader land and property market reforms could be substantial, and may not impose undue burdens on low-income families.33 For example, in Bangalore, moving from a unit value system to a capital value system that reflects reasonable market prices could significantly increase residential property taxes, estimated to be 26 percent. While this analysis is preliminary, it clearly indicates that a gradual move towards reducing housing market distortions and introducing valuation based on market values of housing units (or vacant land) could build buoyancy into the property tax base and enhance collections. For such a system to work, it will also be essential to publicly disclose updated information on market values in various parts of the urban area(s), as well as to reduce the high stamp taxes that are imposed in some states and that discourage accurate registration of property tax values.34 For example, in Kamataka, the capital value of land is periodically notified in all towns and cities in order to levy the Stamp Duty. The notified capital value forms the basis for imposition of the Stamp Duty. Note that this capital value is a notional value, and not the market driven value. Distortions in land markets caused by rent control provisions, FSI and other distortions, the prevalence of black market property transactions, and disincentives for property registration such as the former 12.5 percent stamp duty (which was very high in comparison to other states) impede the measurement of true market values. 3 The Commissioner of Stamps in Kamataka estimated that the guidance values calculated by his Department were about 60 percent accurate. Recent analysis by the World Bank, based on detailed survey data from 33See the accompanying report, Urban Property Taxes in Selected Indian States for more detail on the estimation methodology, which makes the simplying assumption that property tax liabilities are not shifted backward to property (capital) owners. 34 Until recently, Karnataka's stamp tax rate was 12.5 percent, which could present strong incentives to underdeclare property values when registering them, so as to mimimize the amount of stamp duty paid. 35 While rent control statutes have been formally repealed in Karnataka, their effects are estimated to last another 7 years. 18 -ASS I India: Urban Governance and Finance Review Box 3: Property Tax Reform in Bangalore Historically, Bangalore's property tax was anemic in terms of revenue yield. Various factors explain this poor performance: the cap placed on assessed value by rent control ordinances, the failure to reassess properties for nearly 30 years; weak enforcement, and lagging enumeration of properties on the property tax roll. Since April 2000, the Bangalore City Council has enacted four major changes in the property tax: (i) a unit value method of assessment, (ii) a new method of payment, (iii) legislation for a capital value system, and (iv) a new valuation roll. Kamataka is the first Indian state to approve legislation for a capital value system, whereby the property tax would be levied on the capital value of land and buildings. The land portion of the base is the estimated market value of land based on a willing buyer - willing seller definition. In 2001, BMP introduced an area-based unit value system and a new valuation roll. These reforms surmounted the problem of rent control constraints and a badly outdated list of values. For assessment purposes, the metropolitan area was divided into six land value zones, with each "zone" assigned a rental value per square foot. Buildings are classified according to five residential and 11 non-residential categories (based on construction type, age of structure, and current use), and each "type" is assigned a value per square foot. When designing assessment rates, BMP limited the overall increase in tax liability that a residential taxpayer would face, such that the liability under the self-assessment scheme was no more than 2.5 times the liability under the previous system. While a 250 percent increase may seem very high, in practice, taxpayers' liabilities were very low to begin with. Each taxpayer who chooses to voluntarily self-assess their tax is charged with identifying the location of their property, and declaring the classification of their structure. Based on the schedule of values provided by the taxing authority, they may then calculate their tax liability as part of the "Self Assessment Scheme." The tax rate, set by BMP, is 20 percent for residential properties and 25 percent for non residential properties. If a taxpayer does not choose to "self-assess," BMP will assess the value of the property. For those who do comply, a 5 percent tax rebate is given. The new system also changes payment procedures. Taxpayers may now pay directly at a bank, in designated payment boxes, at the electric utility, and soon on the internet. There is no longer any need for contact with the taxing authority, and taxpayers have many more options for easy payment of their property tax bills. In reforming its property tax, BMP placed heavy emphasis on transparency and enhancing taxpayer awareness. BMP, along with the Bangalore Agenda Task Force, held many events and fairs to inform taxpayers about the changes involved in the self-assessment system, produced numerous booklets and published a new khata, developed a public relations campaign, and offered assistance to taxpayers in filling in the new forms. It also directed considerable attention to informing the Council of the changes entailed in the reform, and made a concerted effort to link visibly improved service levels (i.e., planted greenways, improved pathways) with the change in the tax. This reform has been successful in the short run. It addressed the problems of a badly outdated valuation roll, and taxpayer resistance to updating this roll. It was accompanied by better enforcement and it produced a significant revenue increase. Many cities around the world are unable to overcome these hurdles. Four issues must be addressed for this system to generate growing and sustained revenue: the long run elasticity of the "self-assessment" system, transparency, the administrative costs of maintaining this system, and horizontal equity. Bangalore residents, showed that the market Table 7. Percentage of Property Tax Revenue value of residential properties is probably 18 percent Collected vs. Demand, Major Indian Cities higher than respondents stated property values based on 1999-00 1999-00 annual rents.36 Mumbai 55.6 Mirzapur 31.0 Calcuta 55.0 Ahmedabad 12.5 As important as improving the assessment of Bangalore 63.8 Chennai 63.1 ULB property values is actually collecting and Hyderbad 748 JPaitpnar 5686. remitting the tax. In India, as in other developing Ludhiana 70.0 countries, property tax collections suffer from a weak Source: DA. Ravindra and Vasanth Rao, Property taxpayer culture of not paying taxes fully and Reform in India (UNDP Study, Draft June 2002), voluntarily. The property tax register in each ULB records infornation on property tax liabilities, including outstanding liabilities, as measured by the annual demand, collection and balance. Table 7 shows that a small portion of property tax liabilities are being 36 See for example, World Bank Development Economics Group, Bangalore Urban Household Survey, 2001. 19 Executive Summary and Main Report collected in Bangalore, Chennai or Mumbai. As Ravindra and Rao (2002) note, municipalities have adequate powers (under various Municipalities Acts) to collect and recover property taxes tax due. In fact, these provisions are similar to the tax regimes of state and central govemments. Yet, the average municipal corporation collection ratio of property taxes to demand across India has hovered for decades around half of the demand each year. In contrast, collections of state sales tax, central income tax and central excise are generally around 95-98 percent of demand each year.37 In summary, ULBs have few, autonomous sources of revenue, except for corporations in Maharashtra, which rely extensively on octroi revenues. The property tax is a key revenue source, although, at present, it is relatively underused, and has limited buoyancy relative to the overall growth in economic activity. Recent property tax reforms in Bangalore, among other ULBs, show promising results in increasing the yield of the property tax, although they do not address the fundamental weaknesses in valuation. Charges are also important potential sources of revenue, especially for water supply, however, cost recovery is perpetually low, and generally does not cover costs of operation, let alone the costs of needed capital investment. INTERGOVERNMENTAL GRANTS Grants are an important part of any intergovernmental system. Like intermediate-tier governments in many countries around the world, Indian states generally have access to more buoyant revenue sources (e.g., the VAT, personal and business income, etc.) while local governments are expected to play a key role in delivering selected services. Intergovernmental grants provide a way of offsetting the vertical disparities that come from different expenditure responsibilities and access to different revenue sources. Another justification for state-local grants is to equalize disparities in local income (i.e., horizontal disparities) to ensure a more uniform level of local services across the state. States may also provide transfers to encourage local governments to provide services that benefit the state at large (i.e., have positive externalities), such as water supply, health or education. Finally, the state may want to influence how local governments act, and use transfers as an incentive to change behavior. State grants to local governments are determined by the SFCs, which define the share of states' revenues to be distributed to local governments, and how they are allocated to ULBs. As noted above, ULBs also receive pass-through grants from the Central Finance Commission, as well as many central sector and centrally sponsored schemes. For example, Karnataka has over 50 state schemes for urban development and water supply, and its ULBs derive an average of 40 percent of their revenues from grants and other transfers from the State (see Table 5). Given the importance of octroi collections, corporations in Maharashtra receive relatively little in grants from the State; without octroi revenues, municipal councils are much more reliant (about 75 percent) on grants from the State of Maharashtra. Tamil Nadu ULBs are the most self reliant of the sample states, generating almost two-thirds of current revenues from their own sources. In Karnataka in 2001-02, the most significant grants in descending magnitude of funding were: State Plan Schemes (R685.8 crore); State Finance Commission grants (R590 crore); and Central Schemes for urban development and urban water supply (R57.6 crore). While the State Plan Schemes are largest in magnitude, most of these schemes flow to autonomous bodies rather than ULBs. In fact, ULBs have limited say in how most of these funds are used. And while SFC grants are "untied," they are closely linked to salary payments, which means that ULBs have limited discretion in using these funds. 37 See DA. Ravindra and Vasanth Rao, Property Reform in India (UNDP Study, Draft June 2002), chapter 4, Billing and Collecting the Tax. 20 India: Urban Governance and Finance Review Karnataka's SFC defined the distributable pool for grants to local governments as the "Non-Loan Gross Own Revenue Receipts" (NLGORR) of the State Government.38 These receipts include: gross yields from all taxes, duties and fees levied and collected by the State Govemment, as well as interest receipts. The SFC recommended that ULBs receive 5.4 percent of this pool, that SFC grants be allocated .39 according to a formula with five elements:. SFC grant = 0.33Population + 0.33Area + 0.1 lllliteracy + 0.1 lRoad Density + 0.1 lPersons/Hospital Bed The actual allocation of SFC grants to ULBs does not adhere to this formula because of at least three "moderations" to the formula: (i) Bangalore was "held harmless" in its funding (rather than losing a substantial portion of its pre-existing transfers); (ii) ULB government salaries were protected from any change caused by the formula, and (iii) arrears in payment to state utilities (e.g., KUWS&DB) or loan repayments (e.g., HUDCO, KUIDC) are intercepted by the Urban Development Department. These intercepts are significant in magnitude, and they exacerbate the non-transparency of the allocations. During September 2002, 20 ULBs had arrears averaging 20 percent of their SFC grant that were intercepted. Monitoring and accounting for these arrears is very difficult and an Accounting Commission has been established to sort through the arrears. Complaints abound that Karnataka's transfer system is plagued by shortfalls and delays in payment as a result of state fiscal pressures. The actual allocation of 10th Finance Commission Award transfers in Karnataka, for example, bears almost no relation to the formula defined by its SFC. The transparency and predictability of ULB transfers is further diminished by the intercept of transfers to cover arrears in utility payments and other loan payments. Many of the 30 or so grants in Maharahstra (e.g., octroi, profession tax, road etc.) compensate ULBs for local taxing powers that were repealed. They are often distributed on an ad hoc basis. A Chief Officer from one C-Class Council noted that its grants are typically paid at 12 pm on 30 March (i e., at the last moment of the fiscal year). If a particular ULB does not present its bills to the Treasury, then allocations lapse. Karthik and Pethe (UJNDP, 2002) show that, despite the predictable allocations from the Central Government, the devolution of TFC grants to ULBs in Maharashtra was "haphazard," rather than timely and predictable (see Box 4). Surprisingly, TFC releases were made directly from the State to the ULBs without the involvement of intermediate institutions; this may have strained the absorptive capacity of ULBs.40 3 Report ofthe State Finance Commission Relating to Urban Local Bodies, Govemment of Kaamataka, January 1996. 39 ULBs were allocated 15 percent of the 36 percent of the NLGORR, which is 5.4 percent of the total pool. 40Karthick and Pethe (UNDP report) note that most funds were released in the last month of each financial year. While delays could have been due to pending information from the ULBs about utilization of earlier funds, matching contributions (from the ULBs) etc., they may also reflect a lack of state credibility in releasing the funds in a timely and predictable way. 2 1 Executive Summary and Main Report Box 4: 10th Finance Commission Awards in Maharashtra The Tenth Finance Commission (TFC) recommended transferring Rs. 10,000 million in four yearly grant installments to ULBs, based on the ratio of slum population to the total urban population. By this criterion, Maharashtra received the highest allocation, Rs. 1,329.5 million, i.e., 13.3 percent,' and, starting in 1996/97, it was to receive its annual allocation (Rs. 332.4 million) from the Central Government. The TFC's recommendations were accepted by the Central Government, as long as the funds were not withheld by the State for any reasons, and that the TFC funds would be matched by either the State or by the ULB. In case of inability of any ULB to provide for matching contribution, the State Government was expected to provide for the balance of funds. In 1996/97, the first installment was transferred from the Central Government to the State of Maharashtra. The State failed to raise matching grants, and did not transfer the grants to ULBs. Because of the accumulated funds with the State government for previous years and its reluctance to provide matching grants to ULBs, the Central government did not release the committed grant of Rs. 332.4 million for the year 1999-2000, the last year of Tenth Finance Commission (TFC). Source: Karthik and Pethe (2002). By 1999/00, current grants accounted for 20 percent of current revenue in Tamil Nadu ULBs.4' These grants are devolved according to a formula that first divided the distributable pool between a Rural and an Urban Fund. The Urban Fund is then divided among corporations, municipalities and town panchayats according to population, needs and resource potential. Finally, three Urban Funds are distributed among ULBs of similar classification according to an "interse" allocation that takes into account population, SC/ST population as a share of the slum population, per capita own income, and asset maintenance. Municipalities with salary and pension expenditures below 49 percent of total revenue are eligible for an additional share of the funds.42 Fifteen percent of these grant funds are set aside for equalization and incentive purposes; incentives are designed to reward better performance in collecting taxes, repaying debt service, and promptly implementing schemes. In reality, however, many of the equalization fund allocations are based on discretionary criteria for capital projects rather than equalizing fiscal disparities. Like Kamataka, deductions at source are made from these grants for outstanding debts of ULBs to other institutions or agencies. In Tamil Nadu during the first two years of the SFC award period (1997-98 and 1998-99), funds were released on a more or less timely basis. In subsequent years, releases have been much more erratic - often occurring in the final months of the fiscal year - with deleterious effects on local financial planning, and in some cases affecting the ability of smaller ULBs to pay their employees' salaries. It was reported that in 2001-02, the GoTN only paid one of the four quarterly installments for the SFC grants. Further, the allocation and distribution of shared taxes (akin to state grants) are often contested by ULBs. In sum, state grants are important resources for most ULBs., however, they achieve few of their intended objectives Payment of these grants is generally not timely, stable nor predictable, due to state- level fiscal problems, arrears in local payments that are intercepted from the payment of transfers, and other complications. Hence the objective of offsetting the vertical fiscal gap is not being achieved. Despite the inclusion of "need" or equalization measures in their grant formulae, the considerable uncertainty and deduction of arrears offset the potential equalization of fiscal disparities across ULBs in the sample states. Regression analysis of a ULB's economic base and grants received per capita in 41 In addition to current grants, municipalities and corporations also receive capital grants earmarked towards exper water supply, roads and buildings, storm water drains, sreet lighting, solid waste management, education, others, o Comnmission Grants." 42 See Chapter XIII: "Devolution Mechanism" in Second SFC report for more details. 22 India: Urban Governance and Finance Review Karnataka found no statistical relationship.43 ULB grants in Maharashtra are so numerous and complicated that assessing the equalization impact of these grants is very difficult. The potential equalization impact of the Equalization Fund in Tamil Nadu is offset by the discretionary allocation of these funds for capital projects. Hence the objective of using intergovernmental transfers to equalize fiscal disparities is not being met in the case study states. Relative to the other states, ULBs in Tamil Nadu are more self reliant in generating revenues. The relatively high reliance on state grants in Kamataka and Maharashtra, which often are not paid on time nor in full, misses the opportunity to align local costs (i.e., taxes and fees) with local benefits. Given the high degree of dependence of ULBs on transfers and schemes, they should be predictable (preferably formula-based) and stable in order to enable and encourage rational budgeting at the local level. 4. BORROWING Most ULBs in Kamataka and Tamil Nadu depend heavily on grants and loans from state and central agencies (many of which are intercepted or adjusted) to finance their investments in service expansion and capital improvements. In Maharashtra, corporations rely extensively on borrowing to finance their infrastructure, whereas municipal councils have less outstanding debt. The India Infrastructure Report 2002 shows that these sources can meet only a small portion of ULBs' capital financing needs. For example, the GoK estimates that 184 of its towns do not have adequate water supply, and that an investment of Rs. 764.41 crores is needed to meet drinking water requirements. Investments in improved sanitation facilities are needed in 162 towns, at an estimated cost of Rs. 2713 crores.44 Note that the investments needed in these two sectors vastly exceed the total revenues of ULBs in 2001 (Rs. 1010 crore.) Tamil Nadu's 2nd Finance Commission Report estimated that investment of Rs.2679 crores is needed to reach infrastructure norms for municipalities over the period 2002-07. However, given the relatively weak performance of municipalities, the investment target was reduced by almost three-quarters to Rs.650 crores. Similarly for corporations in Tamil Nadu over 2002-07, investment of Rs.2255 crores (excluding investment needed for CMWSSB) is required to achieve the specified norms. When adjusted for the fiscal performance of corporations, investment of Rs.975 crores is needed. Tamil Nadu is unique among the three states in having established a financial intermediary - the Tamil Nadil Urban Development Fund (TNUDF) - to fund urban infrastructure and build local capacity in accessing debt markets (see Box 5). TNUJDF has been relatively successful in approving over 180 urban projects (Rs. 675.02 crores) for storm-water drains, solid-waste-management schemes, roads, and revenue-generating commercial complexes, wholesale markets and bus stands, and in ensuring timely repayment of its debt. Through its technical assistance fund, the TNUDF has also provided an institutional mechanism for strengthening ULB capacity. With the exception of the TNUDF, few incentives nor enforcement mechanisms ensure a hard budget constraint for ULB borrowing. The Karnataka Urban Water Supply Development Board has, in some instances, provided a second water supply facility to ULBs who have not repaid outstanding loans, nor imposed the minimum tariffs (Rs. 45 per connection per month) specified under Government Order 96. ULB defaults on loan repayments are not confined to water supply projects, and create significant liabilities for state agencies. The Government of Tamil Nadu currently guarantees ULB debt, and there is a long history of default or deferral of ULB debt there. 43 No relationship was found between the number of industrial properties per capita in the ULB (a proxy for the economic base) and the amount of grants received per capita, both when considering the sample of 124 ULBs and within sub-samples. 44 Government of Karnataka, Report of the Second State Finance Commission (December 2002), p. 58-59. 23 Executive Summary and Main Report Box 5: Tamil Nadu Urban Development Fund The Tamil Nadu Urban Development Fund was established in 1996 to develop urban infrastructure within the State. Since 1988, the GoTN has been implementing the Tamil Nadu Urban Development Project (TNUDP) financed by the World Bank, and which established a Municipal Urban Development Fund (MUDF). The MUDF was converted to the TNUDF in 1996. TNUDF's objectives are: * To fund urban infrastructure projects that improve the living standards of the urban population * Facilitate private sector participation in infrastructure through joint venture and public-private partnership * Operate a complementary window, the Grant Fund, to assist in addressing the problems of the urban poor * Improve the financial management of ULBs enabling them to access debt finance from markets According to the Annual Report of the TNUDF (2001-2002), 179 urban projects have been approved at a project cost of Rs. 675.02 crores. Typical projects include storm water drains, solid waste management schemes, roads, and revenue-generating commercial complexes, wholesale markets and bus stands. TNUJDF has a strong repayment rate ("zero non-performing loans") in large part due to mandatory establishment of escrow accounts in advance, and payment of arrears out of monthly Grant Fund distributions. The Grant Fund provides financing for many ULB technical assistance activities, including accounting and financial management, computerization, and the development of performance indicators. TNUDF is looking for opportunities to out-source the design, supervision and maintenance of local projects. Source: TNUDF Annual Report, 2001-2002; interviews with TNUDF officials. Fiscal Balance Table 8 presents information on the Table 8: Share of ULBs with Current Account and current account and overall deficits in ULBs Overall Deficits across the three states. It shows that on average, ULBs ULBs commonly incur current account deficits ULBs with with (which indicates a soft budget constraint), but Current Overall more surprisingly, that current and capital Account Deficit revenues are often insufficient to finance total Deficit expenditures. Note that the figures for Kamataka Karnataka (1998-99) are likely to be misleading, since the fiscal data Corporations 33° 33% reported in Kamataka do not clearly distinguish City Municipal Councils' 32% 42% between current and capital expenditures. Town Municipal Councils 24% 29% Maharashtra (1999-00) Despite the requirement for a balanced Brihamnumbai (BMC) 0% 100% budget in Maharashtra, Table shows that a Corporations, excl. BMC 43% 93% significant share of ULBs also incurred current Municipal Councils- 'A' 67% 78% account deficits, and that many municipal Municipal Councils - 'B' 71% 90% corporations and most municipal councils Municipal Councils - 'C' 59% 71% incurred overall deficits between 1995/96 and Tanil Nadu (1999-00) 1999/00. Detailed fiscal data to understand the Corporations 17% 33% flows behind these deficit figures were only Special Municipalities 15% 85% available in Tamil Nadu. In 1999/00, current Selection Municipalities 39% 79% expenditures exceeded current revenues in more Grade I Municipalities 42% 78% than 40 percent of Tamil Nadu's municipalities. Grade II Municipalities 56% 72% While the fiscal data are somewhat limited in truly understanding how these deficits were financed, it appears that new loans accounted for only 20 percent of total deficit financing. Instead, municipalities appear to have financed their deficits by: (i) running down reserves, as shown through dwindling "opening balances," (ii) building up substantial "non-debt liabilities," mainly pension arrears 24 India: Urban Governance and Finance Review arising from the Pay Commission's 1996 report, retention of funds due to the Provident Fund, Electricity Board, or TWAD Board; arrears on salaries; or library cesses collected but not transferred; or (iii) postponing principal and interest payments on water supply loans due to GoTN or the Life Insurance Company. At the end of 1999/00, municipalities incurred Rs. 3.5 billion in debt, of which Rs. 1.2 billion reflected overdue interest and principal payments on outstanding debt, and Rs. 1.3 billion were incurred in arrears to other agencies or entities. Tamil Nadu's corporations appear to be financing their deficits entirely by issuing new debt.45 Because of the frequent deferral of debt service payment on water sector loans, in 1998, the Government of Tamil Nadu rescheduled outstanding loans drawn by various local bodies since 1945. ULBs were offered a uniform interest rate (13.5 percent), and given 20 years to repay the loan in half- yearly installments. If the ULB failed to repay the loan, the GoTN recovered the money by deducting that amount from the SFC devolutions. According to the SSFC report, this policy has resulted in a sharp increase in debt service payments, which in most cases have more than tripled since 1997/98. Analysis of Borrowing Capacity This section estimates the borrowing capacity of ULBs. These estimates are only rough approximations, and given the poor quality of the data, likely overstate the real borrowing capacity of ULBs. Borrowing capacity was estimated in the following way. First, only those ULBs that have a revenue surplus after meeting current revenue expenditures (including debt service) are considered to have any capacity to borrow, due to their potential ability to service debt out of such surplus. To estimate borrowing capacity, the current surpluses calculated for the most recent year for which detailed data were available were assumed to continue over 15 years, and that only half of these surpluses would be available for servicing new debt obligations. These "surplus cashflows" were then discounted at an assumed rate of 12 percent per annum to arrive at a net present value (NPV); the total amount an ULB -ould hypothetically borrow today and repay over 15 years. By this methodology, borrowing capacity is concentrated in larger ULBs (especially corporations) and not distributed uniformly across ULBs. In Karnataka, about 16 percent of ULBs borrowed funds in 1998/99. Based on the large share of ULBs with debt service expenses, about half of CMCs and almost one-third of TMCs have borrowed in the past. Based on 1998/99 data, almost all Karnataka ULBs could borrow additional funds. In fact, all Corporations and more than 90 percent of smaller ULBs could borrow additional funds. However, once thresholds for borrowing are set, 68 ULBs could borrow $100,000, and only 11 ULBs could borrow more than $1 million. Based on available fiscal data, approximately half of the 244 ULBs in Maharashtra have borrowed since 1995/96. Analysis of ULB borrowing capacity in Maharashtra is greatly complicated by the fact that data are not clearly separated into current and capital expenditures. By the methodology described above, surprisingly, BMC would be unable to borrow, whereas about one-third of ULBs have some borrowing capacity, including more than half of municipal corporations, and about one-third of municipal councils. Of the 90 ULBs potentially capable of borrowing additional funds, 67 could borrow $100,000 and 26 could borrow $1 million. These estimates must be interpreted with great caution; the Maharashtra Urban Infrastructure Fund report noted that the "absence of information on liabilities 45Unfortunately, the SSFC data do not contain yearly debt stocks, nor do they report non-debt liabilities for corporations. Hence it is difficult to verify that this is the case. 25 Executive Summary and Main Report constrains assessment of creditworthiness of the local body, as the current [overall] surpluses are mandatory surpluses.'46 In Tamil Nadu, only 59 percent of municipalities had current account surpluses, and therefore could potentially borrow additional funds. These 60 municipalities could potentially borrow an additional Rs. 2 billion ($41 million) and five corporations could borrow an additional Rs. 5.6 billion ($118 million). Only 50 municipalities and six corporations were capable of borrowing $100,000 or more, whereas only 10 municipalities and three corporations met the higher threshold of $1 million in potential borrowing capacity. Are ULBs Creditworthy? Recent, market-based initiatives (or structured financing) proposed in India and implemented elsewhere rely on narrow revenue streams that are pledged (often with "lock box" guarantees) to repay debt, rather than being based on the adequacy of the overall budget (or general obligation) of the ULB to repay its debt. While many ULBs have some cash flow that could be leveraged for borrowing, flows stemming from individual transactions are insufficient to meet the financing gap for ULBs. Nor do these flows in themselves ensure creditworthiness. Creditworthiness requires: (i) stable, predictable and adequate revenues to support borrowing for capital investment; (ii) managerial and financial capacity to use debt responsibly, including good budgeting practices, the ability to operate and maintain capital investments; as well as strategic planning for investment; and (iii) a track record of timely repayment of principal and interest. By these criteria, ULBs in Tamil Nadu, at the margin, are somewhat more creditworthy: they generate more of their own revenues, which are more predictable; they have higher revenues per capita (excluding octroi in Maharashtra corporations), they have stronger systems of budgeting and financial management (discussed in more detail below); and they have good repayment rates for the TNUDF. In contrast, the significant debt arrears, and unstable and unpredictable revenues of most ULBs would disqualify them from being creditworthy. 5. GOVERNANCE ASPECTS A recent review of public sector reform experience over the past decade defines governance "as the manner in which the State exercises and acquires the authority to provide and manage public goods and services."4 As such, governance includes two main components, namely, the accountability of the State (or government) in acquiring and exercising its authority and its capacity in exercising that authority. ACCOUNTABILITY Accountability requires that the government is responsible to its citizens for delivering public goods and services, and its stewards (politicians/policymakers, bureaucrats, councilors etc.) are prepared to explain and face the consequences for their actions. As noted in the WDR 2004, accountability has three dimensions, which are linked together in a chain: (i) political accountability - the accountability of politicians to citizens as well as businesses and other organized interests, (ii) internal accountability - the accountability of govemment agencies to politicians in their role as policymakers, and within this the accountability of lower levels of government to higher levels of government, and (iii) social accountability - the accountability of "front line" units (i.e., service delivery units) to their respective clients. 46 As reported in Volume II, in our data set, a large number of ULBs have overall deficits. 47 Improving Public Sector Governance: The Grand Challenge? (draft, February 2004) 26 ____ __ I India: Urban Governance and Finance Review In addition to its central role in promoting governance, it is generally acknowledged that decentralization will not achieve its potential allocational efficiencies unless accountability mechanisms are in place.48 Political representation, community voice and participation, and transparency can be instruments to promote accountability. In addition, clarity in who is responsible for what, and mechanisms for monitoring performance are critical for promoting accountability in a decentralized context. Political Accountability The 74h CAA defines a number of elements to promote local, political representation: * State Election Cormmissions supervise and assure that regular and fair municipal elections are conducted * One-third of seats in elected bodies (i.e., councils) are reserved for women and 25 percent for members of scheduled castes * Local bodies cannot be superceded indefinitely; new elections must be held within six months * Ward committees are established in municipalities with populations of 3 lakh or more Metropolitan Planning Committees (MPCs) and District Planning Committees are established to promote consultative planning and to prepare and consolidate development plans. Despite these Constitutional guidelines, the actual experience varies across states and it is unclear how effectively local political institutions represent the views of the electorate, or ensure accountability. Maharashtra and Tamil Nadu provide for direct elections of mayors, who serve for a five-year term, while Karnataka elects mayors indirectly, and they serve only a one-year term. Short tenure in office impedes political accountability, as the elected official could be out of office before fully meeting or understanding the needs of his/her constituents. Mahiarashtra, Karnataka, and Tamil Nadu have held two rounds of elections to date. In Maharashtra, the State Election Commission is considered to be reasonably strong, and most ULBs have held second, and in some cases, third local elections since the 74th CAA. Local election processes were subject to litigation regarding the division and reservation of wards shortly after the CAA was passed, but they have grown and matured in the interim. Recently, however, the State Government has dismissed elected local bodies, eg. Dhule MC-A, Bhivandi MC-A, Nagpur MC.49 The Council is the legislative and decision making body of the ULB. Council members are comprised of elected representatives from each ward in the UILB and nominated persons. Councils in Karnataka are elected every 5 years. In Mysore Corporation, for example, there are 65 wards and 76 council members; each ward has 8-9,000 voters. One-third of council seats are reserved for women, and 25 percent for under-privileged groups; these seats are filled on a rotational basis. This implies that a representative from a ward who is not female nor from an under-privileged group cannot serve as a council member for more than 2 consecutive terms.50 Analysis of the extent of inclusion and exclusion within panchayat raj institutions in Rajastan and Madhya Pradesh has shown limited villager influence on gram panchayat decision making, weak 48 See, for example, Jennie Litvack, Junaid Ahmad, and Richard Bird, Rethinking Decentralization (The World Bank, 1999). 49 UNDP 2002, p. 50 Council members of CCs are called Corporators while those of other ULBs are called Councilors. 27 Executive Summary and Main Report accountability to the electorate at large, and exclusion of disadvantaged social groups.5' Similarly weak accountability from politicians toward urban citizens is likely. Internal Accountability In practice, ULBs operate with relatively little autonomy in decision making, and with few incentives to improve their performance. Their functional responsibilities are compromised by a proliferation of State Boards and Authorities (e.g., Housing Boards, Water and Sanitation Boards, Development/Special Authorities, etc.) with overlapping scope and substantial autonomy, but weak accountability and limited coordination vis. ULBs. State agencies in Karnataka, Maharashtra and Tamil Nadu retain critical roles in planning, financing, and sometimes managing infrastructure and services. In Tamil Nadu, for example, the Chennai Municipal Development Authority is responsible for designing, funding and implementing local civil projects that ULBs are then expected to operate, maintain and repay. Similar Development Authorities exist in Kamataka and Maharashtra. In supplying water in all three states, design, planning and investment decisions remain with the Water Supply and Sanitation Boards, state agencies. Municipal responsibilities are confined to the operation and maintenance of the systems, setting water rates (often below state guidelines) and collecting the charges to be remitted to the Water Boards for servicing their capital costs. The lack of consultation or coordination between state and local authorities appears to have resulted in instances in Kamataka where ULBs have added extensions to the water supply system within six months of its completion, thereby reducing pressure and diminishing the performance of the whole system. The Tamil Nadu Water Supply and Drainage Board is acknowledged in the report of the Second State Finance Commission to have operated as a "monopoly," without consultation with ULBs, and in many cases has passed on unreasonable cost and time overruns to ULBs.52 International experience shows that the participation of the community in water supply projects are more likely to be operated and maintained properly and sustained over time (Gross, van Wijk and Mukherjee, 2001). This participation is crowded out by the heavy state involvement in the water supply sector. It would be more efficient if the level of government providing water services - in this case, ULBs - were responsible for investing in infrastructure, and maintaining it over time. Such an arrangement would promote better community participation, which could strengthen accountability, and potentially provide incentives to pay attention to the financial sustainability and operation and maintenance implications of new investments. The WDR 2004 "accountability chain" is weak in many Indian ULBs for many reasons. The lack of clarity in "who is responsible for what," or concurrent expenditure assignment prevents the average citizen from understanding which body is responsible for which service, and thereby holding an entity accountable for its performance or the quality of public services provided. While ULB responsibilities and functions are defined, and in some cases, concurrently, authority and financing are not congruent with these assignments, leading to instances where ULBs bear the financial responsibility for decisions in which they had little or no say. Joint responsibility across various entities, often without coordination, separates the decision maker (often a State entity), from the financier, the service deliverer, and the ultimate beneficiary, which often breaks the accountability chain, resulting in the provision of infrastructure and services that do not match local preferences and needs. Without accountability to ULBs or the state in these instances, it's no surprise, then, that state governments end up bailing out many of these activities. 51 Alsop et al., Inclusion and Local Elected Governments; the Panchayat Raj System in India (World Bank, May 2001). 52 Report of the Second State Finance Commission, Annexure V-16. 28 India: Urban Governance and Finance Review A further impediment to knowing who is responsible for what is the common practice of allowing ULBs to incur arrears to state utilities, and then intercepting state transfers to offset these arrears. If ULBs were responsible for paying utility bills directly (with concomitant sanctions or loss of power for nonpayment of bills), the consequences of their behavior would be more evident to citizens. Eliminating such intercepts might force ULBs to "become responsible for their city being in the dark." The historical prevalence of utility arrears underscores the complexity of establishing accountability in these relationships, and the need to phase out the arrears out only after having identified and piloted successful alternatives. ULB decision-making authority is limited even for devolved functions, since decentralization is often incomplete. ULBs face numerous layers of oversight and regulation, with many decisions on procurement and staffing (e.g., recruitment, hiring, pay scales) taken by state officials. The extremely low limits for municipal approval authority (technical and administrative sanction) and layers of bureaucratic reporting make it difficult for ULBs to know what their responsibilities are, and discourage them from improving their performance or pursuing innovations. In Karnataka until recently, district- level approval was required for expenditures exceeding 5 lakh, and permits for buildings taller than 2 stories had to issued by the State. The appointment of senior municipal officials by the Indian Administrative Service or State Administrative Service brings capacity and professionalism to ULBs. However, the frequent rotation of these officers limits internal accountability and the degree of autonomous decision-making at the municipal level. At present, Commissioners in most corporations are appointed by the IAS or State Administrative Service. Over the past two years, Chennai Corporation, for example, has had five Commissioners. In Mysore Corporation (the second largest corporation in Karnataka), 29 commissioners have served since 1977 - implying an average tenure of less than one year. Service in the IAS or SAS also directs accountability upward to the state or central governments rather than to local citizens, further impeding internal accountability. ULBs also have limited scope in managing their assets. Land is over-regulated and urban land markets are highly distorted (e.g., land ceiling act, rent control, regulations on conversion of use, high stamp duties).54 Land use regulations, titling and permitting are generally opaque and ineffective. ULB infrastructure is often planned and implemented by state agencies, creating disincentives for adequate operation and maintenance or repayment of debt. In Karnataka and Maharshatra, financial accountability and transparency of ULBs are generally weak. Municipal budgets are not typically prepared in a uniform format and budget data are often inaccurate. Budgets are typically estimated incrementally based on previous year's values, rather than based on an objective analysis of the cost of delivering a service, or outcomes to be achieved, or the liability accrued from long-term investments. An encouraging development is Bangalore's introduction, with the assistance of the Bangalore Agenda Task Force (BATF), of a performance budget, and movements such as JANAGRAHA to promote participatory budgeting and disclosure of municipal performance. Budget controls are generally lacking - Bangalore only recently introduced a variance system. Nor are state and ULB budgets linked, despite the fact that state transfers account for a substantial portion of ULB funding. 53 Historically, ULBs in Karnataka faced very low thresholds for procurement and hiring approvals, with considerable involvement of the DMA and UDD. A typical approval hierarchy would include the Assistant Commissioner, Deputy Commissioner, Divisional Commissioner, Director of Municipal Administration, and then Urban Development Department. Such low limits compromised the autonomy of ULBs and often delayed service provision considerably. In the past, any approval above R25,000 came to the DMA, and it sometimes took 8 months to clear tenders. 54 See India Urban Sector Note (2002) for more details of land market disincentives. 29 Executive Summary and Main Report In contrast, Sections 70 and 71 of the TNULB Act provide detailed guidelines regarding the preparation and presentation of ULB budgets in Tamil Nadu. ULB budgets (except for Chennai Corporation) are submitted to the Department of Municipal Administration and Water Supply (DMA) for approval. Each budget must show a consolidated budget surplus, with at least 1 lakh surplus in the Revenue Budget. Tamil Nadu's DMA periodically monitors ULB budgets, and it requests revised budgets from each of the ULBs in November of each year. Cash accounting systems are common in Karnataka and Maharashtra and are ill-suited for accurately representing the financial position of ULBs, and local audits are often performed with delays.55 The Maharashtra Urban Infrastructure Fund report notes that in municipal accounting "there is a general mix up of current and capital works, especially in municipalities" (Kirloskar Consultants, 1998.) In Karnataka, Bangalore and the Tumkur City Municipal Council are beginning to move to fund-based accounting (i.e., modified accrual accounting), implemented with support from BATF and technical assistance from the ADB. Tamil Nadu is relatively unique among Indian states in having introduced double entry, accrual accounting in all ULBs in 2000-01. Accrual accounting was piloted in 10 municipalities and two corporations beginning in 1997-98, prior to being rolled out across ULBs. Before accrual accounting was introduced, ULBs had reasonable control procedures in place; they had fund-based accounting systems (e.g., General Fund, Water Fund, Education Fund, and Lighting Fund), with requirements for regular reporting of fund balances and statements of liability. Once computerized accounts are fully implemented (expected in 2004), the DMA and Department of Town and Country Planning expect to be able to monitor ULB fiscal performance in "real time." ULBs typically update their trial balance each month or quarter. With full computerization, the DMA target is for all ULBs to calculate their trial balances within 3 days of the close of each month. About 60 percent of ULBs are reported to meet that target presently. Accurate ULB data are limited, and there seems to be little systematic monitoring and evaluation of the performance of ULBs nor of the projects developed by various boards and authorities in Karnataka and Maharashtra. For example, Maharahstra's Annual Administration report, which is supposed to summarize ULB performance, has not been prepared for many years. As noted above, the multiple roles that various institutions play also confounds performance monitoring. While State Finance Commissions (SFCs) monitor fiscal flows, they do not monitor ULB service delivery.56 In partnership with the Bangalore Agenda Task Force, Bangalore has negotiated a performance agreement with GoK for increased grant funding in return for improved performance in defined areas. Tamil Nadu is more proactive in monitoring ULB performance, and improving the quality of ULB fiscal and performance data is a high priority for the DMA. Annually, the DMA issues a performance budget for ULBs, which includes basic information, major finance items, details of works undertaken and various performance measures. These performance measures reflect outputs: kilometers of roads, numbers of street lights, vehicles used, quantity of garbage collected per day, number of water supply and sewerage facilities, number of compost yards, etc. However, the budget does not really report performance information, such as the quality, duration nor satisfaction of services delivered. While internal accountability mechanisms are weak in many ULBs, there are some promising developments. Reforms underway include the introduction of double-entry accounting systems in Tamil 55 See, for example, NCRCL, Karnataka State Financial Accountability and Management Study (Draft, July 2002). 56 SFCs are formally responsible for: reviewing the finances of municipalities; estimating their future financial requirements; design a package of taxes, duties, tolls and fess that can be assigned to municipalities or shared between the state and municipalities, as well as state grants-in-aid; and suggest measures for strengthening municipal finances. See Om Prakash Mathur, Approach to State-Municipal Fiscal Relations: Options and Perspectives (National Institute of Public Finance and Policy, March 2001). 30 India: Urban Governance and Finance Review Nadu and recently introduced in Karnataka, introduction of information technology, and the performance contract negotiated between Bangalore Corporation and the Govemment of Karnataka.s7 State reforms in Karnataka, Maharashtra and Tamil Nadu establishing Freedom of Information Acts and reforms of procurement processes in Karnataka and Tamil Nadu should also create an enabling environment for enhancing ULB accountability and transparency. The scope for administrative reforms to enhance internal accountability is considerable. In recent reforms, the Karnataka UDD has delegated more powers to its ULBs. For example, financial authorization limits have been increased in line with general delegation of powers from 1-5 lakhs for ULB's and 5-20 lakhs for Deputy Commissioners. Further, recommendations of the Committee on Administrative Reforms have led to a significant devolution of powers to department heads, deputy commissioners and ULBs, as well as a simplification of business processes. Social Accountability ULBs do not have an independent tax domain and, except for the octroi in selected states, they are largely devoid of autonomous revenue sources. As noted above, local tax autonomy encourages taxpayers to become more aware of the costs of services, and brings local officials' actions subject to closer scrutiny by taxpayers. Local revenue mobilization is relatively weak in the ULBs examined, which means that this potential scrutiny is likely to be weak. Promising developments in promoting social accountability include the use of citizen report cards, citizen charters and enhanced citizen participation in decision making and monitoring (Bangalore's Agenda Task Force, Public Affairs Center's Report Card for Urban Services, Mumbai First, etc.). The Bangalore Citizen Report Card is a model for bringing the power of transparency to bear on improved service delivery (see Box 6). Tamil Nadu's Department of Municipal Administration (DMA) has been working with ULBs to establish local service centers to facilitate electronic interactions with citizens. For example, Ambattur Municipality has established four service centers where citizens can pay their taxes, apply for licenses and permits, and record and check basic information (addresses, etc.) These service centers offer the possibility of vertically integrating government services - operating as a "one-stop" shop for paying and registering for local and state government services. These centers are linked by computer networks to state and DMA offices, and information is shared daily. It is hoped that they could some day be used to manage daily cash flows. For example, rather than collecting assigned revenues through the Collection Department and then transferring the shared portion back to ULBs, those revenues could be remitted on a daily basis through the service center to the ULB where they were paid. This potential reform could help in enhancing the daily cash flow of ULBs. In addition, Tamil Nadu's DMA envisions developing score cards that would monitor ULB performance according to the performance indicators that it has developed, benchmarked to some norms. It is hoped that by providing more information to citizens through these score cards, a "revolution" would ensue with citizens demanding improvements in ULB performance, and ULBs competing against each other. The DMA is also assisting ULBs in developing City Corporate Plans (CCP), which provide a snapshot of ULB financing, systems and capacity to citizens and businesses. The TNUJDP is piloting the development of CCP in 50 ULBs, and the GoTN is committed to expanding CCPs to all ULBs. 57 The impact of the performance agreement has not been assessed. However, it is one of the first instances where performance benchmarks were clearly defined and linked to financial incentives. 31 Executive Summary and Main Report CAPACITY Capacity refers to the govemment's ability to do things, through supporting processes/systems (e.g. budgeting system), appropriately skilled personnel, and adequate financial resources. As noted above, the appointment of senior ULB officials from the IAS or state administrative service builds ULB capacity, but compromised local accountability. The Kamataka First State Finance Commission Report notes that while ULBs have existed for almost a century, their administrative set-up and information system are still primitive; this is likely true for ULBs in other states as well.58 Many aspects of the organizational structure and administration of ULBs severely limit the decision-making autonomy of municipal govemments, undermine their accountability and incentives, and impede their ability to deliver services effectively. Key staffing challenges for ULBs in Kamataka and Maharashtra are the significant number of vacancies, poor qualifications of many staff, the declining pool of experienced staff due to looming retirements over the next five years and lack of direct recruitment, and significant mismatch of skills, especially in technical areas. In Kamataka, ULB staff are disproportionately concentrated in the lowest skill grade -- Group D staff account for three-quarters of ULB staff. "Special groups" (SC and ST) account for almost 50 percent of the total staff hired in Group D. Because there has been no recruitment of Grade 2 Chief Officers since 1982 in Karntataka, and GoK Executive Order passed in 1992, allowed municipal employees to be promoted to the KMAS cadre based on their years of service in the ULB without meeting any basic qualifications. As a result, 109 ULBs are headed by officers (promoted municipal employees) who have insufficient qualifications to carry out their tasks. To overcome these weaknesses in capacity, large corporations in Kamataka such as Bangalore has focused effort on improving the quality of staff by having a dedicated Human Resource Director and training courses for staff. In addition, initiatives have been made to recognize exceptional performances of officials through non-monetary rewards such as publishing recognition in Corporation's newsletter. District Collectors play less of an oversight role in Tamil Nadu than in other states, concentrating instead on facilitating local officials. Tamil Nadu uses a different model to supervise ULBs, and more importantly, to build local capacity. The DMA appoints Regional Directors to support 12 to 17 municipalities. These Directors are typically former Chief Local Officers who have been promoted to facilitate municipal functions; and they report to the Commissionerate, Municipal Administration, and serve more in an advisory than regulatory mode. While District Collectors are technically responsible for inspecting ULBs, in practice, they have little day-to-day responsibility. Tamil Nadu has also used its Tamil Nadu Urban Development Fund as a mechanism to build ULB capacity, especially in building accounting and financial management capacity in order to access private capital markets, but also including computerization, and the development of performance indicators. An advantage of this approach to capacity building is that ULBs learn these new skills in parallel with using those skills to prepare and oversee infrastructure projects. 58 See Government of Karnataka, First State Finance Commission Report, Chapter XI. 32 India: Urban Governance and Finance Review Box 6: Bangalore Citizen Report Card Satisfaction Across Report Cards 120 - 100 73 80j 2 '199 60 01~~f - El999 '" 40 ,F-31 3 02003 20r BMP BESCOM BESSB BSNL POLICE BDA BMTC Agencies A main finding of the 1994 report card was the pervasiveness of small scale corruption among the city's poor - both in the form of bribes and extortion. The results from the second report card were unusual in that they showed improvements in the quality of service delivery, but a rise in the levels of corruption among the majority of public agencies. One plausible explanation, identified by Samuel Paul, is that citizens were accustomed to accepting corrupt practices and needed to be informed about their rights and the potential benefits of collective action. The third report card, implemented in 2003, reveals both improvements in service delivery and a dramatic decline in levels of corruption among public service agencies. ' Much of the improvement in service delivery and corruption is attributed to persistent public exposure to this problem through media and open debate; growth of civil society groups and the media; and the responsiveness of providers to address these issues. The Bangalore case is one example of how citizens were able to eventually change the culture of corruption once they set their minds to utilizing the infornmation and access available to them.' Source: Samuel Paul, Holding the State to Account :Citizen Monitoring in Action (Books for Change, Bangalore, India, 2002), and presentation by Samuel Paul on Citizen Report Cards: A Comparative Perspective, presented at the World Bank headquarters, 20 November 2003. 6. CONCLUSIONS AND OPTIONS FOR REFORM The challenge of urban finance and governance in India is a classic example of incomplete decentralization. Responsibility doesn't match financing; accountability and capacity are weak, and local autonomy and state control are not balanced in ways that create incentives to improve performance. Indeed, these problems are not unique to ULBs in India, but are shared by panchayat raj institutions as well.59 The challenges for rural decentralization are surprisingly similar: limited expenditure discretion; overlapping responsibilities; weak revenue effort and low collection rates; complicated, non-transparent and non-equalizing transfer systems; outdated budgeting and accounting systems; and inadequate information and monitoring systems. As a result of this incomplete decentralization, lackluster ULB performance impedes service delivery, discourages investment, and stifles the economic potential of cities. Improving cities' access to financing is necessary, but not sufficient to overcome these challenges. Nor is it possible to fix the intergovernmental system all at once. A systematic approach is needed over the medium to long-term to improve ULB performance, including state and local actions to enhance fiscal sustainability, and strengthen institutional arrangements to promote accountability and performance. Such an approach 59 See the World Bank, India Fiscal Decentralization to Rural Governments (2003). 33 Executive Summary and Main Report could lay the foundation for transforming Indian cities into dynamic engines of growth, that could improve overall economic growth and development. Notably, other reforms in the overall fiscal system - especially fixing the problem of significant and growing state deficits -- are needed to unleash the potential of Indian cities. Table 8 below lays out a series of state and local actions over the short-, medium- and long term to enhance fiscal sustainability and strengthen institutional arrangements. ENHANCING FISCAL SUSTAINABILITY Enhancing fiscal sustainability would require various actions, including: reigning in the growth in local expenditures; enhancing local revenue mobilization by expanding and deepening property tax reforms, improving cost recovery of charges and fees, and linking the payment of taxes and fees with the delivery of services; building State credibility to ensure timely, and predictable payment of intergovernmental transfers, and enhancing the equalizing effect of such transfers; and stopping state bailouts and instead promoting responsible local borrowing. Restrain Growth in Spending For fiscal sustainability, the rapid growth in ULB spending must be brought more in line with the growth in ULB revenues. Local wages account for 20-40 percent of ULB expenditures, and it is commonly believed that many ULBs are overstaffed. While ULBs have little control over their wage rates or employee mix, opportunities for gradually outsourcing staff -- especially lower-grade employees or selected administrative functions - abound. Approximately 10 percent of larger ULBs in Kamataka (e.g., Bangalore) are outsourcing many of the functions originally performed by Grade D officials such as park cleaning and garbage collection. Bangalore's approach to outsourcing generally specifies the number of posts (i.e., number of sweepers) to be outsourced rather than an outsourcing of complete tasks (i.e., office cleaning.) Unbundling of local services horizontally and vertically is a necessary first step to improve service efficiencies and competition. ULBs have begun unbundling services in solid waste management, and in introducing private sector participation in the water sector. In concert with the Bangalore Agenda Task Force, and feedback from the Citizen Report Card, Bangalore's departments identified spending programs that could be managed more carefully, and these changes were reflected both in its performance budget and the satisfaction ratings of citizens. 34 India: Urban Governance and Finance Review Enhance Local Revenue Mobilization In addition to the potential property tax reforms discussed below, key ULB actions in enhancing resource mobilization include improving their collection of existing revenue sources, including enhanced cost recovery for utility charges and fees. Improved information systems to enhance the registration, billing and collection of taxes and charges, and stronger enforcement of collections have been shown to generate substantial improvements in ULB revenues. Improving the quality of local services provided may instill a stronger culture of payment of local taxes. Bangalore found that it could overcome some of the culture of non-payment of property taxes by improving its services (i.e., cleaner streets, nicer parks and gardens) before seeking higher property tax payments. At present, local utilities with poor cost recovery are often bailed out by state subsidies that are not transparent, and which are repaid by intercepting SFC transfers. These subsidies discourage ULBs from raising their tariffs or improving their cost recovery. Eliminating such subsidies immediately would create additional contingent liabilities in the utility sector, hence states would need to choose carefully which subsidies would be reduced gradually over time. Improving the information and reporting of utilities would be an important starting point for estimating real costs and identifying opportunities for improved cost recovery. Expand Property Tax Reforms Property taxation in India, has long been and continues to be a vexing issue. The basic problems are to overcome its relative underutilization in Indian cities, to get government officials and the public to move to higher levels of property taxation, and to a base that better reflects the market value of property. Given the successful movement in some India ULBs to reform their property tax systems, this could be an opportune time for significant property tax reform. An encouraging development is the property tax reforms underway and being discussed in various ULBs, which have shown that, with appropriate changes, revenues can be enhanced significantly. It falls to the state govemments to lead any reform effort. But where would India start? Ordinarily, one would point to improving property tax administration, and to the need for focusing on financing service delivery. Dillinger (World Bank, 1998) notes that focusing on fundamentals, that is, updating property tax rolls, computerizing billing and collection systems, and strengthening enforcement, are important first steps that can yield significant results. However, India has a major problem with the structure of its property tax systems, and much of this can be traced back to the failure to resolve the conflict between assessing the true market value of rents, and rent control ordinances. In addition, government officials have generally been unwilling to issue new valuation rolls, in some cases for many years. The recent history of property tax reform in India has been one of stopgap measures to overcome these problems, rather than engaging in comprehensive reform. Meanwhile, the growth in property tax revenues has remained anemic. Unless these structural issues are resolved, improved administration will do little to make the property tax a viable revenue source for local govemments. Property tax reform in India can make progress in three potential areas: structural reforms to improve the revenue yield, buoyancy or equity; administrative improvements, and changes that build taxpayer confidence. 35 Executive Summary and Main Report Increase Revenue Yield If the property tax is to be a mainstay of local government finance in India, then its revenue yield must be increased. This can be done in many ways. The most obvious is a revaluation of property to market levels, either through maintaining a rental value system, or by moving to a capital value system. Bangalore's experience showed that moving from the previous rental value-based assessment to an area-based system increased revenues by around 62 percent. Adoption of market rental values would increase revenue yields in Pune by an estimated 55 percent, and in Bangalore by another 34 percent. For those places whose valuations are still tied to controlled rents, the potential increases are even greater. Second, legal tax rates could be increased. At present, the inclination is to keep tax rates at nominal levels. For example, if Mumbai were to move to a capital value system, the planned rate is 0.02 percent, which would not generate sufficient revenues to make the property tax a major source of financing local government services. To move to a productive level of property taxation, the nominal property tax rates must be increased significantly. Effective property tax rates in industrialized countries are roughly 1 percent of full market value, and in many developing countries these rates approach one half that level. Indian states should consider setting a target level of effective rates based on needs for financing public expenditures. Bangalore is proposing a reasonable starting nominal rate, between 0.3-0.6 percent. Third, states could begin to investigate their exemption policies. In many cases the exemptions are standard and given by all countries, e.g., charities, properties used for religious purposes, foreign embassies, etc. In other cases, exemptions may have been given for exceptional purposes or that stretch the meaning of "charities" or "religious properties." Increased revenues could be gained from removing the preferential treatment of certain types of properties, e.g., residential properties, owner-occupied properties, or vacant land. Finally, state government cesses as surcharges on the local property tax might be removed and their financing shifted to state government sources. This would free up room for increases in the effective rate of local property taxes. Increase Buoyancy Another needed structural reform is increasing the buoyancy of the property tax, that is, having property tax revenues grow automatically to match the growth in local expenditure needs. This requires putting in place a property tax base that will grow as property values grow, and one where the growth can be captured by existing administrative machinery. The switch to an area- based valuation system in several Indian cities increased revenues significantly on a one-time basis, but does not necessarily produce a buoyant system. Buoyancy can be built into the property tax system in two ways. One is to regularly revalue as is required by law. This would imply a large increase in the base, and therefore in tax liability, every fifth year. The other would be to develop a method of indexing the tax base, and then doing a "reconciliation" every fifth year. The fifth year shock would be less severe under this method, but the drawback of this method is that some properties would be inappropriately valued in the intervening years. Enhance Equity Structural change could also improve the horizontal equity of the property tax, whereby equals are treated equally under the tax, and the tax does not interfere with market decisions in inappropriate ways. Among the structural changes that might be considered are to remove the 36 India: Urban Governance and Finance Review commonly-used preferential rates of tax that are levied against residential vs. non residential properties, and against owner-occupied vs. rented properties. A better route is to allow the level of assessed value of the property to be the sole guide in determining the taxation of a property. Rent control is a major problem, and similar properties with and without rent controls might face tax burden differences by as much as a factor of four. Improve Administration There is much room to improve and professionalize the property tax administration, including identification of the tax base and the tax payers, valuation, record keeping, and collections. Some of this can be done even before structural changes are decided upon, but some of the administrative improvements clearly will need to await some decisions about the choice of a property tax base. Identification of properties. For the small sample examined here, and from other reviews, it is clear that all taxable properties have not been identified. Particularly on the urban fringe, many properties are not brought promptly onto the tax roll, and considerable revenue is lost. In Bangalore, for example, evidence suggests that as much as 70 percent of newly developed properties do not enter the tax rolls during the first six years after development. In many ULBs, tax maps are out of date. Some records are not computerized, and there is too little provision for cross checking with other records (e.g., utility bills) to determine property characteristics. Valuation. Proper and up-to-date valuation is the main issue of property tax administration (and policy) in most developing countries. If the tax base does not reflect current market value, the tax cannot be productive, its revenues cannot grow, and it will not be fair in its burden distribution. The first step the Indian states must take is to decide on the tax base and the revaluation period. Valuation itself is a difficult administrative task. If an area-based system or a capital value system is chosen, then basic data must come from the Stamp Duty Office. Given the scant evidence from the case studies analyzed in Annex II, stamp estimates may understate market value by about 20 percent. With stamp and transfer taxes as high as 15 percent of transaction value in some states, there is a great incentive to understate this value. A thorough examination of the efficacy of using stamp data as the basic unit for valuing property is a high priority task. If an area-based system is adopted, as is used now in some of the larger ULBs, then a method of updating the guidance values on a regular basis is necessary. This will require not only reliable values from the Stamp Office, and from the state Ministries of Construction, but also a set of procedures for updating these values. It also will require trained staff, capable of valuing real property, and perhaps a central valuation unit in each state should be considered. There is much to be done to implement a system of this kind. Most local governments do not have a cadre of trained assessors to evaluate property values and update them regularly. A capital value system is even more difficult, because valuation of individual units will be required. In either case, a method for updating any new construction or major renovations, sub divisions, etc. must be put in place. A capital value system will be difficult and costly to implement. Its introduction will require much careful planning, and will take time. The costs and the complexities of introducing a capital value system should not be underestimated. A rental value system could also work, if market rather than controlled rents were to become the tax base and if a tax roll of current market rents could be maintained. But this raises the issue of assessing market rents. With rent controls being repealed in several States, it is likely 37 Executive Summary and Main Report that, over time, the new equilibrium of rental prices will reflect market values. It may take many years before the rental market functions efficiently. Even absent rent controls, land use and zoning ordinances in most Indian cities likely impede the functioning of the housing market. Collections. The collection rate in most ULBs is weak. This is important because an increase in the effective rate of property tax is probably needed. If enforcement is a problem at the current low level of rates, it will be even more of a problem at higher tax rates. Available statistics suggest that a collection rate over 50 percent might be considered comparatively good in ULBs. In most ULBs, a 20 percent increase in the present collection rate should be attainable with stronger enforcement efforts. Collection performance probably varies across cities, and in some cases, enforcement possibilities are weak. The ultimate sanction, confiscation of property, is not a politically attractive alternative in any country. Short of this, however, many steps might be considered: (i) provision to collect from tenants under some circumstances would address some of the problems of absentee landlords, but can be an expensive proposition; (ii) better collection procedures and improved recordkeeping could help increase collection rates; (iii) requiring that taxes be paid in full during the period of a legal challenge could improve compliance. Build Taxpayer Confidence Taxpayers' lack of confidence in the property tax may effect compliance. First, it is necessary to link property tax payments and service delivery. If taxpayers do not see that their taxes buy better public services, they are less willing to pay taxes. In many cases, the tax is used to pay statewide services via a cess that is surcharged to the basic property tax. Eliminating cesses (with an offsetting reduction in SFC grants), which lessen the link between taxes paid and benefits received, and are often not collected nor remitted in full to the State, may also be called for. Second, taxpayers may feel that the tax is unfair in that it burdens other, equally situated owners or occupiers (or businesses) less than it does them. The Indian property tax is a case study in horizontal inequity because of the large number of exonerations and preferential treatments offered. A property tax where only property value mattered in determining tax liability would be more acceptable. Third, taxpayers object to high compliance costs. For a tax that yields so little, the property tax seems to impose high compliance costs on taxpayers. Harassment by collectors has been a particular thorn. Bangalore's "self-assessment system," allows property taxpayers to bypass any direct dealings with assessors, and contributed significantly to its increased tax yield. One of the hallmarks of Bangalore's success was that the property tax system was streamlined, made more understandable to citizens, and easier to comply with. Bangalore also launched an extensive public awareness campaign about the property tax reform, which engaged the Council and citizens in support of the reform. Finally, states must track the performance of ULB property taxes, and stand ready to provide technical assistance, or even legal changes in the tax structure, where necessary. This implies a research function, with detailed property data, and capacity for analytical work.. Understanding the composition of the property tax base, the shares of taxes paid by different groups, and how they might be affected by a potential reform is an important part of designing and sustaining a successful property tax reform. 38 India: Urban Governance and Finance Review Additional state actions to enhance local property tax collections could include reducing stamp duties and offsetting the lost state revenue through commensurate reductions in the SFC grant. A further incentive could be provided to those ULBs that reach reasonable levels in collection ratios of taxes to demand, by granting them greater local autonomy (perhaps in setting tax rates) or access to more elastic bases, with regular cash flow. Urban governments in other countries often have access through local "piggyback" taxes to more buoyant tax bases (i.e., motor vehicle charges, business privilege/licensing taxes). For example, South Africa has assigned a business payroll and turnover tax to subnational governments, and its metropolitan governments raise over 90 percent of all revenues from their own sources. Such sharing is unlikely in the current environment of state fiscal stress for such buoyant bases to be shared by states.60 Revise Intergovernmental Transfer System As noted above, central and state schemes to ULBs have high transaction costs, and funds that are approved are not always drawn down due to weak local capacity. The overall number of urban schemes could be reduced, and those that remain could be restructured to improve their targeting/equalization impact. A critical reform would be to improve the certainty in the timing and amount of state transfers to ULBs. Brazil's revenue sharing system supports state governments in poorer regions and municipal governments throughout the country through a series of formula-based transfers that are made accurately and promptly. Nor do transfers meet their equalization objectives, Intercepting transfers to repay arrears in utility payments or borrowing diminishes the transparency and accountability of the intergovernmental fiscal system, and greatly complicates the administrative costs of the transfer system. Such intercepts should also be phased out, but only after an alternative system of effectiv_ sanctions has been developed and successfully piloted. Promote Responsible Local Borrowing A critical first step in promoting responsible local borrowing is to clarify who owes what. The current system of intercepting arrears in debt service from state transfers leads to considerable confusion at both state and local levels. ULBs often believe that they have repaid their outstanding debt through reductions in their transfers, when the actual amount of debt outstanding may be unrelated to the amounts intercepted from transfers. Improved information and reporting, as well as a phased elimination of intercepting arrears would assist in clarifying who owes what. A second critical step is to align the responsibility for financing and providing infrastructure services, so that the entity responsible for providing the service decides what investment is needed, and is responsible for financing and repaying any borrowed funds. Managing the existing stock of ULB debt is necessary to establish a self-sustaining local credit market. Much of the existing ULB debt was incurred in a dysfunctional system where accountabilities were not clear and there were few incentives to repay debt. This stock of debt provides a significant claim against the existing resources of many ULBs, and introduces liabilities throughout the intergovernmental fiscal system. Tamil Nadu introduced a debt restructuring program that was accompanied by more transparency and stronger enforcement of 60 Indeed, the first State Finance Commission in Tamil Nadu recommended levying a local tax on vehicles not under the purview of Motor Vehicles Act (e.g., mopeds, tractors etc.) but GoTN rejected that recommendation. 39 Executive Summary and Main Report repayment. Similar approaches might be applied in other states to manage the ULB debt overhang. Tamil Nadu is also relatively unique in establishing a financial intermediary - the TNUDF - as a mechanism to establish a local credit market. The TNUDF has succeeded in building local capacity to manage local borrowing, and has also ensured a strong repayment rate. Other states could consider this model as well. Given the generally weak creditworthiness of ULBs and their limited experience with local capital markets, it's unlikely that most ULBs could go directly to the local credit market for financing, although that should be an intermediate goal. Local ULBs would also need to strengthen their own creditworthiness, for example by improving cost recovery, strengthening infrastructure charges and tariffs, managing their expenditures more carefully, improving financial reporting, and building capacity to manage the construction and maintenance of local infrastructure. The significant unanswered fiscal questions in the above sections underscore the importance of collecting and analyzing annual fiscal data on a more systematic basis. For example, available data show growing ULB deficits in Tamil Nadu, but it is impossible to verify how these deficits have been financed. More importantly, without data on debt and non-debt liabilities collected for the Second SFC, it would be impossible to assess the financial stress that ULBs face now and could face in the future. ULB debt service payments have grown rapidly, and, only with detailed information (past and current) on the stock of debt, non-debt and arrears, can future cash flows be projected, and a ULB's ability to borrow accurately assessed. A regulatory framework would need to be established that enforces a hard budget constraint and eliminates the cycle of state bailouts. Such a framework might require the establishment of bankruptcy legislation, creditors remedies and workout procedures for borrower defaults, and possible methods of interceding in ULB management if necessary. South Africa has developed a relatively robust regulatory framework for local borrowing and financial emergencies that might provide insights for Indian states.6' Having experienced three major, state debt crises during the 1990's, Brazil's federal government enacted a series of controls on subnational borrowing, including the Law of Fiscal Responsibility. Finally, for a hard budget constraint to work, individual states must be committed to letting ULBs fail rather than bailing them out. STRENGTHENING INSTITUTIONAL ARRANGEMENTS Clarify Expenditure Authority Clarity in expenditure assignment is needed to avoid fiscal imbalances, assure accountability, and monitor performance. A critical need is for state action to redefine the role of Boards and Authorities toward regulation rather than financing and provision of infrastructure. Devolving responsibility for financing and providing infrastructure to ULBs is both consistent with their roles as defined by the CAA, and more likely to result in sustained local investments. As states lessen their involvement in delivering local services and making related investments, corollary reforms must be undertaken to redefine the role of boards and authorities toward supervision and facilitation, including building capacity in auditing, monitoring and benchmarking ULB performance, regulation of utilities, and building local capacity in delivering services. Administrative streamlining of approvals is also necessary to clarify what ULBs are 61 See for example,-www.treasury.gov.za. 40 India: Urban Governance and Finance Review authorized to do, and to empower them to take more responsibility in managing their own performance. Improve Information and Performance Monitoring Improving transparency and monitoring throughout the intergovernmental system, as well as among ULB politicians, the executive, and citizens is necessary to strengthen accountability and improve performance. Improved information at state, local and national levels, would provide a better sense of what needs to be managed or what areas that are most promising for reform, and how resources are actually flowing among levels and across ULBs. Reforms in budgeting, financial management, accounting and auditing are needed to provide the necessary information on a timely and accurate basis. The accountability chains between citizens and politicians, politicians and bureaucrats, and bureaucrats and citizens must also be strengthened. Granting more autonomy to ULBs in local employment decisions (i.e., deciding which staff are hired, fired and promoted, and how much they are paid) would strengthen the accountability chain between bureaucrats and local politicians, and potentially improve service delivery. Shortening the accountability chain by engaging local citizens in bureaucratic decision making (i.e., budget planning, resource allocation, etc.) as shown in the case of Bangalore . Tamil Nadu has taken the most systematic approach to generating local fiscal information and monitoring local performance, though GoTN recognizes that much more needs to be done in strengthening the links between information for citizens and options for holding local officials accountable. As noted above, information flowing through the intergovernmental fiscal system is often limited, which results in poor decisions and impedes the monitoring of ULB performance. Information flows from ULBs to the state and local citizens, flows among ULBs, and among state agencies could be strengthened. Key ULB actions include enhancing financial disclosure; establishing fund-based accounting systems and conducting timely, external audits; establishing performance targets in budget formats; implementing participatory budgeting; and establishing citizen charters. Actions at the state level include strengthening the databases, indicators and analytical capabilities of the Department of Municipal Administration to monitor local performance, and enhancing information flows among departments, especially the Stamp Department and ULB property tax offices. Publishing information is not enough to enhance local performance. Information must be used to enhance performance. For example, information campaigns on ULB service levels or financing might be used to promote "yardstick competition" among ULBs. As noted above, engaging local stakeholders in monitoring ULB performance yielded substantial improvements in services in Bangalore. Another mechanism for enhanced performance is the Memorandum of Understanding between the GoK and Bangalore that defines specific performance targets and offers financial incentives for improved performance. Given the plethora of reforms needed, a key question is how to initiate and sustain ULB reforms. Karnataka and Maharashtra might consider Tamil Nadu's "wholesale" approach to ULB reform. This approach has three important components: (i) a systemic view of reform (i.e., revising the legal framework, requiring all ULBs to adopt reforms in budgeting, financial management and accounting, etc.), (ii) incentivizing individual ULBs by linking reforms with financing from a financial intermediary (TNUJDF), and (iii) using the TNUTDF to build capacity across ULBs in a systematic way. 41 India: Urban Governance and Finance Review Table 8. Potential State and Local Reform Options Short-term Medium-term Long-term State Actions Fiscal Improve reporting and monitoring of Ensure regular and predictable payment of state De-link local salaries from national Sustainability outstanding ULB arrears (consider TN transfers for recurrent spending wage rates experience) Adhere to equalization formula in allocating Rationalize cesses vis. adjustments Introduce accrual accounting in transfers based on need and capacity in transfers rather than as a part of Maharashtra, and support full Show progress on URIF requirements: phase- local property taxes implementation in Kamataka out rent control laws; repeal Urban Land Assess whether rate setting Identify options for strengthening state Ceiling and Regulation Act; rationalize stamp autonomy is needed, or other fiscal sustainability duties, and reduce rates in a phased way revenue bases should be available to Identify alternatives to intercepting arrears, ULBs including linking sanctions to outstanding Based on successful pilot arrears (i.e., turn off electricity when bills experience, apply sanctions rather aren't paid), and pilot in selected ULBs than intercepts to outstanding arrears Promote responsible local borrowing - stop bailouts, lend to creditworthy ULBs, consider Work out local debt overhang , establishing a financial intermediary in (consider TN experience) Maharashtra and Karnataka Institutional Review approval thresholds and raise Reduce frequency of transfer of local officials Allow local officials to hire and fire Arrangements those that unduly restrict local decision Clarify and reduce the role of state boards and their own employees, and set local making authorities in local investment decisions; align wage rates Develop indicators and enhance state decisionmaking for selecting and financing Allow locally elected representatives capacity to monitor local performance investments with maintaining those investments to appoint ULB executives Provide technical assistance to ULBs to Develop systematic approaches/vehicles for Allocate a substantial portion of enhance their fiscal reporting strengthening local capacity (consider TN ULB resources based on the experience with TNUDF) achievement of performance criteria Assess impact of Bangalore's Performance Contract with GoK, and identify other options for incentivizing improved performance 42 I India: Urban Govemance and Finance Review Local Actions Fiscal Working with local stakeholders, identify Improve property tax administration (i.e., Improve average collection ratio of Sustainability areas for improved service delivery or registration, billing and collection) and deepen property taxes to demand reduced cost coverage of property tax reforms Update and maintain valuation rolls Streamline processes and procedures to Improve cost recovery of local utilities through on a regular basis to reflect reign in costs improved administration and stronger economic growth Link taxes and charges to services enforcement provided In Maharashtra, identify suitable alternatives to Consider public-private partnerships Improve the accuracy and timeliness of the octroi for municipal corporations in providing utility services fiscal reporting Enhance taxpayer services and reduce the transaction costs of paying taxes and fees Develop managerial and financial capacity to manage local borrowing In Kamataka and Maharashtra, improve the timeliness and completeness of local audits Institutional Develop indicators and engage local Encourage competition among local agencies in Strengthen accountability of local Arrangements stakeholders in monitoring local providing services politicians to citizens perfornance (consider Bangalore Agenda Develop local capacity for investment selection Task Force and Mumbai First experience) and construction management Include performance indicators in local Right size low-skilled employees budget Develop systematic approaches to enhance capacity and skill mix of local employees 43 Executive Summary and Main Report 7. REFERENCES Alsop, Ruth et al. 2001. Inclusion and Local Elected Governments: The Panchayat Raj System in India (World Bank). Bagchi, Amresh. "Reforming the Property Tax: Need for a New Direction," November 1997. Bahl, Roy. "The Property Tax in Developing Countries: Where are We in 2002?" Andrew Young School of Policy Studies, Georgia State University, 2003. Bahl, Roy and Paul Smoke. 2003. Restructuring Local Government Finance in Developing Countries; Lessons From South Africa. (Edward Elgar Pub. Co.). Bahl, Roy and Johannes Linn. 1992. Urban Finance in Developing Countries (World Bank). Bangalore Mahanagar Palike, Property Tax Self-Assessment Scheme, 2000. Bangalore Mahanagar Palike, Rationalization of Property Tax in Brihanmumbai.(undated) Campos, J. E. "Improving Public Sector Governance: The Grand Challenge." 2002. (World Bank mimeo, 2004). Deichmann, U., S.V. Lall, A. Suri and P. Rajoria (2003). "Information-Based Instruments for Improved Urban Management." World Bank Policy Research Working Paper 3002. Government of India, Ministry of Urban Development and Poverty Alleviation, The Annual Report 2001-2002. Government of India. Statistical Abstract, various years. Government of India. Planning Commission. 10th Five Year Plan (2002-2007). Government of Karnataka, 2002. Report of the Second State Finance Commission. Government of Karnataka, Report of the State Finance Commission Relating to Urban Local Bodies, January 1996. Government of Tamil Nadu, 2001. Report and Recommendations of Second State Finance Commission of Tamil Nadu. International Monetary Fund. Government Finance Statistics, various years. Karnik Ajit and Pethe Abhay. Assessment of Revenue and Expenditure Patterns in Local Bodies of Maharashtra (IJNDP). Kirloskar Consultants (1998). The Maharashtra Urban Infrastructure Fund Report. Litvack, Jennie et al. Rethinking Decentralization, 1999 (World Bank). Mathur, Om Prakash. Approach to State-Municipal Fiscal Relations. Options and Perspectives (National Institute of Public Finance and Policy, March 2001). Mehta, Meera. Meeting the Finance Challenge for Water Supply and Sanitation ([he World Bank, Water and Sanitation Program.). National Commission to Review the Working of the Constitution. (September 2001) "A Consultation Paper on Decentralization and Municipalities." NCRCL. Public Financial Accountability in Urban Local Bodies in Karnataka, July 2002. Rao, Govinda. (2002) Dynamics of Indian Federalism, draft manuscript. Ravindra, D.A. and Vasanth Rao, Property Reform in India (UNDP Study, Draft June 2002). 44 India: Urban Governance and Finance Review Rodden, Jonathan and Jennie Litvack (editors), Decentralization and the Challenge of Hard Budget Constraints (MIT Press, 2003). Vaidya and Vaidya., "Management Innovations for Municipal Resource Mobilization in India," FIRE Draft Note 30, July 2002. Tamil Nadu Urban Development Fund. TNUDF Annual Report, 2001-2002. United Nations. United Nations World Urbanization Prospects, 1999 Revision. World Bank (2004). Stabilization and Fiscal Empowerment: The Twin Challenges Facing India's States. World Bank (2003). Real Estate Reforms: Bringing India 's Cities into the Economic Liberalization Program. World Bank (2003). World Development Report: Making Services Work for Poor People. World Bank, (2001) India Urban Sector Strategy. World Bank (2002). Karnataka Urban Sector Technical Note. World Bank (2002). India : Evaluating Bank Assistance for Pubic Sector Management - a Country Assistance Evaluation (OED). World Bank (2002). Civil Service Rationalization In India: A Survey Of Issues And Options (World Bank Policy Paper). World Bank (2002). Brazil: Issues in Fiscal Federalism. World Bank (2001). Development Economics Group, Bangalore Urban Household Survey. World Bank. (2001, 2002). India Country Assistance Strategy and Update. World Bank. (2001) Morocco Municipal Management and Decentralization. World Bank (2000). Overview of Rural Decentralization in India (SASRD). World Bank (2003). India: Fiscal Decentralization to Rural Governments. 45 I g S g S g > I