Scarcity of resources for urban infrastructure is a universal concern in developing economies. Also, prudent mandate of macroeconomic management has led to a reduction in hand-downs from higher governments. The sub-national governments have had to look at several alternatives with a sense of urgency. In this paper we focus on the possibilities of the sub-national governments accessing, the financial markets in general and the debt market in particular. Our paper focuses on the need to create virtual entities – self-help groups amongst the urban local bodies – that could expand the domain of eligible ULBs. We visualise a scheme for capital market access by ULBs, which would work without the state acting as an intermediary and also without any new institution being set up. We provide the theoretical underpinnings, illustrate and operationalise the idea with the help of data related to ULBs in the state of Maharashtra.

ABHAY PETHE, MALA LALVANI

A

This paper focuses on the possibility of the sub-national governments accessing the financial markets in general and debt market in particular. We focus here on the need to create virtual entities – self-help groups amongst ULBs – that could expand the domain of ULBs eligible to access the credit market. We illustrate and operationalise the proposed scheme with the help of data pertaining to ULBs in Maharashtra. The paper is divided into four sections. Section I gives an overview of the finances of the ULBs in Maharashtra, and draws attention to some known sources of revenue generation. Section II briefly looks at the Indian experience of sub-national borrowing in the capital market. Section III suggests a methodology and provides illustrations for operationalising a new idea that is the main theme of this paper. Finally, in Section IV we conclude.

Revenues of ULBs can be broadly classified as revenues from own sources and those from external sources, such as grants from the state and loans. Again, own sources of revenues can be categorised as tax revenues and non-tax revenues. There are specific provisions in the state laws, regarding taxation powers of the ULBs. Article 243 X of the Constitution, inserted after the 74th Constitutional Amendment (CA) envisages that states should devolve additional taxation powers to ULBs, so as to make them financially competent for discharging the added functional responsibilities, mandated by the succeeding Article 243 W. However, in Maharashtra, there has been no such devolution of taxation powers, which would have been expected since it would have been in consonance with the process of decentralisation. Instead, we have seen that the taxation powers of small ULBs regarding octroi have been withdrawn by the state in March 1999. Hence, the taxation powers of the ULBs are limited to its traditional sphere and have not gone beyond various existing provisions in the state laws.

What is important to note in this context is that even within the list that delimits the taxation powers of ULBs, there are provisions in the state laws that further reduce the flexibility of the ULBs. This has been illustrated in the case of property tax. Karnik et al (2004) point out the municipal corporations in Maharashtra can levy property tax as a percentage of annual ratable value of the property, and ceilings for such percentages are laid down by the state in three different acts. Brihanmumbai Municipal Corporation has no autonomy regarding the components and rate for each component of the tax, while Nagpur has limited autonomy. All other municipal corporations or MCs (governed by BPMC Act) have autonomy regarding the rate of tax in case of components related to water supply and sewerage only. However, there is no freedom to any MC regarding inclusion of any new component or changing the tax base to some other, say, area.

The important point that we wish to make here is that the power of ULBs in Maharashtra appear to be highly restricted with respect to both, the tax and the non-tax sources of revenues, which constitute their own sources. This has forced the ULBs to be dependent on the state for their finances. These constitute external sources of finance, which complement the own sources of ULBs. However, grants from the state are not devolved by objective or fair principles but are ad hoc in nature.

Thus it would appear that revenue growth of the ULBs has been constrained by such inherent structural bottlenecks like limited autonomy regarding taxation, small bandwidth for nontax revenues and the unpredictable nature of funds flowing from the state. The problems faced by ULBs have been worsened by the fact that the 74th CA has further added to the list of services to be performed by local bodies. As already noted, the functional devolution to the ULBs not being matched by supporting financial devolution has lead to too many responsibilities chasing a narrow resource base. The problem is further aggravated by the stipulation in the municipal acts that ULBs must balance their budgets. The plethora of contraints on the ULBs has made assessment of their functioning difficult. It is difficult to establish whether non-performance represents dereliction of duty or inability to perform due to lack of funds. Tabulated below is the share of grants and major sources of own revenue in total income of municipal councils and corporations.

The point that we would like to emphasise is that given the resource crunch faced by the government, the ULBs cannot help but depend on themselves (singly and collectively) for resource mobilisation in the foreseeable future if de facto decentralisation has to occur. Some much discussed alternative sources of revenue that could be explored for improving the finance of ULBs are briefly discussed in the subsection below [see Karnik et al 2002 for greater details].

Improving overall finances of the ULBs in general would involve various steps like: prudent management; spending resources on appropriate items; cutting costs; minimising unproductive expenditure by spending on identified priorities; selection of appropriate low cost technologies; proper maintenance and timely replacement of exhausted infrastructure; private sector participation; and identification of socially essential subsidies and elimination of inessential ones.

ULBs would need to improve on both fronts, viz, improve implementation of existing sources of revenue and tap new sources. A major source of tax revenue for the municipal corporations is property tax. The recovery percentage of property tax, in its present structure, is around 60 to 65 per cent, in the case of corporations and between 40 and 50 per cent in municipal

The state level pooled fund scheme being proposed in India is along the lines of the “Bond Bank” scheme prevalent in the US. In the late 1960s the state level bond bank concept emerged in the US to support the borrowing by smaller municipalities which otherwise find it difficult to tap the capital market. A bond bank is a state sponsored intermediary that borrows from the capital market and then lends to small municipalities either by subscribing to their bonds or by direct lending. The issuers benefit from lower administrative costs that are spread among a large number of borrowers. Pooling of credit also results in a higher overall credit rating for the issue as the credit risk is diversified among a broader range of municipalities. The bond banks are self-sustaining and charge the municipalities a fee for their service. This US version simply does not go far enough since it leads to veritable cherry picking that is not in consonance with the developing status of our country nor with the inclusive form of developmental agenda that we have mandated for ourselves.

The scheme that we would like to suggest would extend the existing concept and yet, which would work without the state acting as an intermediary and also without any new institution being set up. We believe that the existing financial institutions can do the job with enabling legal and regulatory provisions. Our scheme would provide incentives for not only the best of the ULBs to come together and access the capital market, but also demonstrates that strong and relatively weak ULBs too could come together an obtain a credit rating as a “Virtual” Entity. Needless to say that such a coalition would emerge only if it gives to each partner at least as much as it would have got if it had independently accessed the market, i e, that the standard imputation conditions apply. The scheme proposed here shows that such incentives could indeed be built into the system.

Theoretical Underpinnings

In this subsection we elaborate on the theoretical underpinnings of our idea/scheme that we have been discussing. First we consider the space of ULBs which forms the universe of discourse for the purposes on hand. This set will need to be partitioned into three subclasses, viz, one that has ULBs that are “good”, those that are not very good but those that have some redeeming features and finally those ULBs that are – in a sense – beyond repair. In order to rank the ULBs, there would be a set of criteria that provide the filter for such a categorisation. Formally therefore we have the following:

Let Ui (i = 1...N) denote the N urban local bodies. In a specific case these may be municipal corporations or A,B,C municipal councils. In this general framework we make no such distinction. We use ‘L’ different criteria, not necessarily of equal importance in order to gauge the rating/ranking of ULBs. We use a twostage filtering strategy to categorise the entire set of ULBs into three exclusive classes, U(I), U(II) and U(III). As the first stage filter we use Per Capita Revenue Surplus. All ULBs with a revenue surplus gives us the first of these classes, viz, U(I) whom we shall denote as “Cherries”. Within the category U(I) we could use some stricter criteria and obtain a subset which denote as “Super-Cherries”. This is simply an illustrative mechanism of picking the Best of the Best. This is strictly not essential to our argument. All the ULBs recording a per capita revenue deficit define a joint set of {U(II) ∪ U(III)}. To distinguish U(II) from U(III) we make use of second stage filtering strategy. In our exercise we have identified three criteria for this purpose, which we discuss in the next section. We then define a benchmark for each of these criteria and from {U(II) ∪ U(III)} we identify U(II) as being those which pass the test for at least one criteria. The U(II) group is then “Best amongst the Worst” and are termed as “Salvageables”. All the revenue deficit ULBs which fail the test for all three criteria are the U(III) group and are termed as “Duds”. The ULBs in this category need direct intervention and/ or support by the state to strengthen them enough so that “they may be able to walk on their own steam”.

Let us now briefly talk about the different financial instruments (Ij) that could come into play. One could off-hand mention some such as on lending by a FI, muni-bond (specific to ULB or project), general or structured debt obligations or pure vanilla or indeed some derivative. Obviously the choice of any one or more of the instruments depend on the funding agency, as well as the ULBs that come together for the purposes of raising resources and the type of projects that are being contemplated. The general theoretical precept here is that the normal sequence is from junk bonds to loans to investment grade bonds. Obviously, we are not interested in junk bonds that can be issued by the institutions that have a poor financial health. Being concerned here with public bodies, such junk bonds are not advisable and indeed our argument has been that in such cases (especially in the extreme), the state has to intervene in a direct fashion and in a decisive way to help such ULBs to get out of the rut that they find themselves in, in the process redefining the role of the state. The loan exposure is already in existence and has been increasing in importance for all categories of ULBs. However, the loans do not constitute bank exposure to the extent feasible has been argued by us elsewhere [Pethe and Ghodke 2002]. In that paper we have underlined the scope and suggested measures towards remedying this situation. In this paper however, the focus is on the third category, viz, investment grade municipal bonds. In the Indian case, the institutional framework is already in place for IPOs and this has been successfully exploited in several cases. The point is that unless this is extended to general public offering where individual investors are incentivised to bid for and hold this paper, we would not be able to fully exploit the potential of this avenue. It is also common knowledge that one of the necessary conditions for this to happen is the existence of vibrant secondary market in this paper.

The literature on bond issue/market/pricing is common knowledge in literature. Obviously, the credit rating of the bond issuer (ULB for us) is a crucial input in deciding on the premium. The size of the issue is also important and it is pertinent to point out that there are obvious economies of scale to be reaped. The significance of this rather obvious observation is that – with some regulatory changes – it may be possible for a ULB to go in for an issue size in excess of its immediate requirement and use the excess cash profitably for onward lending to ULBs or indeed to finance a project – by way of an extension – of another ULB and reap technological economies. All this of course is contingent on the particularities of the type(s) of projects under consideration.

Broadly, the set of universe of projects {Pk} can be loosely subdivided (say) into P(I), P(II) and P(III). Among these P(I) will be comprising of projects whose purpose is to supply goods that are in the nature of basic amenities. These would normally be made up of goods with overwhelmingly pure public good characteristics. The second class P(II) will have projects that deal with “merit” components and acceptable “marketable” characteristics. Finally, the third class P(III) will be “almost pure” market goods that have conventionally been produced/provided by ULBs. There is of course another way of looking at the classification. This has to do with the underlying production technologies and the efficiencies to be reaped through scale effects. Also, one can categorise the projects depending on whether space (i e, contiguity for implementation) is a consideration. But the latter two categorisations are not immediately relevant. Finally, there is an aside that is nevertheless important. The identification and proposing of bankable projects is a special skill that requires considerable capacity. It is usually the case that such capacity rarely exists with “weaker” ULBs. Whilst this provides a long term agenda for capacity-building efforts, in the shorter run, this is one more reason why the different ULBs (strong and weak) need to come together to make a headway.

contract amongst each other and with the funding/underwriting agency). Third, a thriving secondary market (and not just IPO which already exists) for muni-bonds must come into being.

Our empirical illustration tries to demonstrate the following cases: Case 1: U1, U2 ∈ U(I) come together to form a virtual entity V1 such that NPV(V1) = α. We then have to argue why such a coalition formation is feasible and utility enhancing at the aggregative level as well as at the micro level for all concerned (imputation problem). Case 2: We illustrate the formation of the virtual entity V2 made up of U1 and U2, where U1 ∈ U(I) and U2 ∈ U(II) and consider the same problem as in the earlier case. This case is particularly important in that it represents a cross over possibility, underscoring the development argument for “conceptual extension” made earlier in the paper.

It needs to be argued why such coalitions provide incentives for intra-contract. Various arguments have to be made based on scale and portfolio principles. There are also reasons based on technology and contiguity and arbitrage. It is quite well known

– from elementary finance literature – that through construction of portfolio the risk associated with the portfolio is pegged at a level that is less than the weighted sum of the individual risks of the components that go to make up the portfolio. In standard symbols this can be formally written as: σ (V1) < w1 σ (U1) + w2 σ (U2). The implication, in our case is that the virtual entities will be able to get access to credit/bonds at easier terms than individually. Thus when both the ULBs coming together are strong, there is still an incentive for them to come together. But more pertinently, even when one of them is weak – due to associated high risk – there may be a case for them to come together when the composite risk is acceptable for the purposes of credit disbursement, i e, σ (V1) < α and α < σ (U2), where α is the acceptable level of risk for lending or debt issue. Here of course the imputation problem becomes crucial for creating an incentive for the stronger ULB to join the coalition. In both the cases mentioned above, all the concerned parties in a Pareto improving way may share the resulting “spoils” through bargaining and contract setting. The other aspect has to do with the fact that there are overheads involved in the process of incurring debt, with the obvious implication that going for large loans or bond issue will be rather more cost effective, after all the cost argument of the supply function of loans or bonds is, ceterus paribus, monotonically inversely related to the quantum involved. These and other scale economies provide an important economic rationale for going in for collusion between ULBs. The scale economies can also be rationalised from technological angle in fairly obvious way. The argument of spatial contiguity will depend on the particularities of the specific projects involved. All this of course presumes that there is a possibility (enabled by regulatory/legislative) of reaping arbitrage gains. We now turn to empirical illustrations.

Empirical Illustrations: The Benchmarks

A total of 238 municipalities were considered for the year 2000-01. The district of Jalna was left out on account of lack of availability of relevant data. Since the exercise is purely illustrative any such data problems will certainly not vitiate our results. Also, we chose to keep the city of Mumbai out of the sample considered. Results obtained by including and excluding Mumbai city have been found to be markedly different [Karnik et al 2002]. Inclusion of Mumbai would have certainly influenced our illustration as the benchmark norms would have been significantly different.

PUG = (Expenditure on public goods)/Total Expenditure

In this illustration (Table 3) Navi Mumbai and Thane are both our “Cherries”, i e, they are revenue surplus and they also pass the test for all our three criteria. Clearly, if they come together and pool in their assets the virtual entity, which is a combination of the two will have a superior credit rating and a higher revenue surplus and therefore a higher NPV of the loan raised. A coalition of two such ULBs is obviously beneficial to both on scale as well as – in this case – on contiguity argument. Two points to be noted here, one, that they are now able to access together a quantum of funds that are greater than what each one could have therefore given that they had the same credit rating, the gain to be made is by borrowing large amount resulting in economies of scale from the side of the FI (say bank) or the issue of debt side (arising mainly out of reduction of transactions costs. This means that they can go for large technology projects or use the left-over resources for further on-lending (should the regulations allow) to slightly weaker and smaller ULB and making gains from such a financial trade. Two, noting that the two ULBs are contiguous, it may be possible to go for large spatial projects than would lead to delivery of service that in a broken form would be costlier to provide.

In this paper we have drawn attention to the significance of financial markets’ access by sub-national governments. We have focused on the ULBs in Maharashtra. Contextualising the financial status of the ULBs, within the parametric environment provided by the current economic scenario in India, we have argued that there is an urgent need to address the issue of resource crunch vis-à-vis functioning of ULBs and their mandate. Whilst there are several well-documented strategies in the literature, in this paper we have focused on conceptually extending the concept of “pooled fund” which requires the working out of “intra-group incentives”. We have also illustrated our argument of “possibilities” along with the economic and developmental rationale with the help of computations conducted with real data.

Implicit in the above arguments are several threads. One is that the existing infrastructure fund could be used to facilitate underwriting of the projects that are being undertaken by coalition of ULBs coming together as virtual entities. This should help the ULBs to float debt and with other enabling changes help initiate the process of ushering in healthy secondary market for this kind of paper. The underlying argument also indicates the required move away by the government from taking on the mantle of direct producer or provider of infrastructure facilities. The fund should

Name of Name of ULB Class (as per ULB Type Revenue NPV the District the ULB 2001 Census) Surplus

Thane Navi Mumbai E Cherry 422255 1922931 Amravati Amravati E Salvageable -27897 Virtual entity = (Navi Mumbai + Amravati) 394358 1795889

NPV has been computed for 50 per cent of the revenue surplus at 7 per cent rate of interest for a period of 15 years.

be seen only as seed money to be used for purposes of creating an environment where the ULBs are incentivised to take up project through coalition formation and perhaps capacity-building consultancy. The other implicit argument here is that the banks should learn (and be allowed) to look at the coalitional formation and encourage them by taking ever-increasing exposure as a matter of policy mandate or indeed as serving their profit motive. These institutions can be useful for creation of policy framework as they do bring to the table considerable amount of relevant experience from consortium formation from lending side. Yet another important implication – an aside – of this paper is the underlying a need to redefine (by delimiting and refocusing) the role of the state in this regard. In our scheme of things the state is visualised as a facilitator. Legislative changes would need to be explored to provide an enabling environment for the scheme to be operational. Also, the “Duds” would need to be identified and their infrastructure needs have to be taken care of directly albeit in an accountable way by the state such that it would help them to graduate from being “Duds” to “Salvageables” and then to becoming “Cherries”. Perhaps a pre-infrastructure fund could be created for this purpose.

Thus much more important further work that needs to be done in order to actualise the potential that lies in the domain of the design of feasible policy implementation mechanism. Given the situational possibilities, the financial institutions and ULBs will each need to be incentivised to come forward and create “happy accidents”. The government will have to set the stage by promulgating enabling policy initiatives using the existing infrastructure fund along with provision of a pre-infrastructure fund

– may be with the help of the RBI – that will jump-start the process of economic coalition formation between the ULBs as well as help the “Duds” along. The good news is that the present economic ethos is pushing various institutions – like the financial institutions and ULBs – up a steep learning curve. The need to change is no more a question of one’s taste or volition but indeed of survival.

mr:

Email: mlalvani@economics.mu.ac.in

[We would like to acknowledge the comments by an anonymous referee which has enabled us improve on the paper. Data used in the study is based on data collected for a World Bank consultancy which is gratefully acknowledged.]

Bagchi, S (2001): ‘Financing Capital Investments in Urban Infrastructure: Constraints in Accessing Capital Market by Urban Local Bodies’, Economic and Political Weekly, January 27.

Financial Institutions Reform and Expansion Project-Debt Market Component (FIRE-D): Developing A Municipal Credit Rating System in India, Project Note no 20, November 1999.

Ghodke (2004): Accessing Capital Markets by Urban Local Bodies in India: An Assessment of Municipal Bonds, India Infrastructure Report, 3iNetwork, OUP.

Karnik et al (2002): ‘An Approach to Designing Intergovernmental Transfers: A Comprehensive Study of Urban Local Bodies in Maharashtra’, Study Submitted to UNCHS and UNDP, May.

Karnik, A, Anita Rath and J C Sharma (2004): ‘Reforming the Property Tax System: Simulated Results for Mumbai City’, Economic and Political Weekly, August 21.

Pethe, A M and M Ghodke (2002): ‘Funding Urban Infrastructure: From Government to Markets’, Economic and Political Weekly, Vol XXXVII, No 25, June 22.

Sood, P (2004): ‘Country Case Study: India in Freire M et al (ed), Subnational Capital Markets in Developing Countries from Theory to Practice, WB-OUP.