Financial Viability of Small and Mid-sized Cities in India: A New Approach

Abstract

We develop a novel framework for performance evaluation of governments, identifying three strands embedded in a public finance cycle: (a) expenditures are getting translated into services, (b) service delivery bringing about revenues, and (c) revenues resulting in expenditures. An intuitive construct is conceptualised, and a slack-based Data Envelopment Analysis Model is mapped to each of these strands. Stylised facts derived from the possible consequences by comparing the input/output radial efficiencies and the off-radial slacks thereafter, offer a nuanced analysis of financial viability for 910 small and mid-sized urban local governments (ULGs) in southern India. We find that, irrespective of size, services fail to generate sufficient revenues in ULGs, primarily due to poor revenue collections. Most of the ULGs could have incurred existing levels of expenditures with lesser revenues, indicating suboptimal resource utilisation. Smaller ULGs suffer from supply bottlenecks in services, which are maximum in roads, followed by water supply. The mid-sized ULGs face leakages of expenditures in general, while the smaller ULGs, in establishment expenditures in particular. Policy recommendations suggest revising water tariffs, parking fees, incurring expenditures on the maintenance of roads, regular monitoring and audits, and effectively devolving and utilising tax and non-tax revenue handles.

JEL: C14, H41, H70, H71, H72, R51

Keywords

Public finance nonparametric estimation off-radial efficiency financial viability

I. Introduction

Empirical literature on performance assessment of any level of government primarily centres around how effectively the government translates its spending into the delivery of public services. This emanates from the rationale of optimal provision of services by the governments discussed in public finance theories when preferences of the population are not revealed (Musgrave, 1959; Samuelson, 1956). However, the other two strands on how service delivery helps generate revenues and how revenues influence expenditures for governments are hardly explored in the empirical literature. As a result, the assessment of government performance remains somewhat incomplete. Can we offer a framework using all three strands and develop a new methodology for assessing the performance of governments? Through this framework: can we identify service inadequacy and expenditure leakages in government? Can we assess the problem of poor revenue collection and the inadequacy of revenue handles? Can this framework help us estimate poor revenue channelisation and underutilised revenue handles?

We begin by discussing the public finance theories within which each of these strands is rooted. An analysis of these three distinct strands together would build up a thorough and nuanced empirical evaluation of the performances of any level of government, and is necessary to get an accurate understanding of the financial viability associated with the process.

In the first strand, expenditures can be treated as inputs for providing services. This emanates from the role of the government in establishing physical, environmental, technological, financial and social infrastructure in the economy (Stiglitz, 2005). It is imperative for governments to intervene in the provision of public goods because individuals generally do not reveal their preferences (Musgrave, 1959; Samuelson, 1956). Moreover, it is the role of the government to ensure that citizens can be provided with the minimum acceptable standards of living.

It is primarily the governments at the local level that are entrusted with the responsibility for providing basic services like water supply, sewerage, street lighting, solid waste management and roads to the population in India. The local governments have better information on the tastes and preferences of their population and therefore can spend in accordance with the demand for services (Oates, 2008; Slack, 2011). There has been empirical evidence that fiscal decentralisation influences the composition of expenditure on infrastructure, as found in the case of Italian regional administrations over the period 1996–2008 (Grisorio & Prota, 2015). Delegating such responsibilities to higher levels of government could cause a non-optimal supply of services. Therefore, local governments should incur expenditures to ensure optimal service delivery to the local population. In fact, according to the cameralist school of economic thought, expenditures by the government are considered as investments for generating future streams of revenue (Backhaus & Wagner, 1987). This can happen only when the government spends on the provision of public goods, which could generate revenues. Thus, service provision would require expenditures being incurred by the governments. The expenditure by the government would be rendered inefficient if it does not translate into service provision.

The second strand, where services are crucial inputs for generating revenues, emerges from the government’s function of collecting revenues through tax and non-tax sources (Buchanan, 1949). Olson (1969) calls for ‘fiscal equivalence’, which means that those who receive the benefits of collective goods must pay for those benefits. Public goods like water supply, roads, public parks, healthcare or education fall under such category. Therefore, service provision should ideally lead to revenue generation for the governments. Taxes, which are a major source of revenue for local governments, are higher in communities that spend on public goods (Oates, 1969). People are always willing to pay more to live in a community that provides high-quality services (Bahl & Bird, 2008; Bird, 2010; Oates, 1969). Utility maximising consumers would move to the jurisdiction which provides them with the greatest benefits (services) over costs (taxes paid) (Tiebout, 1956). In fact, regional competitiveness may depend on infrastructural services (Kitson et al., 2004).

The third strand is rooted in the government’s critical role of redistribution of income and wealth (Musgrave & Musgrave, 1989). The role of the government is not limited to generating and collecting revenues. The revenues collected must be spent on infrastructure, services and other social welfare goals. So, we can identify revenues as inputs to generate spending. The expenditures will depend on the number and nature of revenue generation handled by any level of government, which would maximise social welfare (Buchanan, 1949).

The literature on fiscal federalism assigns the redistribution of incomes through taxation and expenditures, as a major role to the local governments (Bird, 1993). Own revenues play a significant role. Residents hold local officials accountable if local public services are financed through locally imposed taxes and user charges (Olson, 1969). This makes raising revenues through taxes and user charges important for spending on collective goods or services. Therefore, higher expenditures would require higher revenue generation through taxes and user charges. Subnational revenues should be enough for subnational units to cover expenditures (Bird, 1993). If revenues generated are not utilised for expenditure on services, this could indicate a lack of willingness to spend or the existence of unutilised public money.

The aspects of financial viability explored above can be used to assess the performance of governments and some existing lacunae, empirically. Using these, we develop a conceptual framework and, thereafter, an empirical methodology for the performance evaluations of governments. While this framework and the methodology can be applied to any tier of government, the present article analyses the performances of urban local governments (ULGs) in India, as the local government performances are rarely evaluated in India due to complexities in operations, with intergovernmental dynamics and the paucity of data.

The article is structured as follows: a conceptual framework analysing the performances of the ULGs is described in Section II. Section III discusses the empirical literature and gaps. Section IV presents the data, variables and the methodology used in the article. Section V offers a set of stylised facts by listing down the possible consequences of applying our models and their implications on the financial viability of cities, theoretically. Section VI discusses the results of the article. The article ends with conclusions and policy recommendations in Section VII.

II. Conceptual Framework

We interpret the efficiencies of the governments from three different angles in an input-output framework as indicators of performance. These efficiencies are relative, so there might exist a possibility of further improvement for the best performers. In the presence of resource constraints, tracking these improvements is extremely important, as we cannot leave aside any possibility of strengthening financial viability. The present framework, for the first time, provides a methodology for a comprehensive analysis of performance evaluation incorporating all possible sources of improvement.

The first strand, which considers expenditures as inputs and services as outputs, indicates how effectively the expenditures incurred by the governments are translated into the provision of services. The possibilities of inefficiencies through misappropriation and/or leakages in expenditures can be traced through this strand, from the input side. We can trace the possibilities of inefficiencies through service inadequacy from the output side.

The second strand considers services as the input and revenues as the outputs. The rationale is that the generation of own revenue sources of the governments (tax and non-tax) depends on the services provided. The possibilities of inefficiencies through an inadequacy of revenue handles and/or inappropriate or unrevised rates for taxes (and/or user charges) can be traced from the input side of this strand. We can trace the possibilities of inefficiencies through poor revenue collections from the output side.

The third strand considers revenues as inputs and expenditures as outputs. The revenue generated by the governments should be spent on service delivery. It is important to see how effectively the revenues generated are converted into expenditures. The possibilities of inefficiencies through ineffective revenue channelisation can be traced from the input side of this strand. We can trace the possibilities of inefficiencies through underutilised revenues from the output side.

The three strands and their interdependence can be depicted with the help of the following diagram (Figure 1).

The framework identifies the sources of gaps and leakages in the performance of governments from the perspective of financial viability. The potential for further improvements in the best-performing governments can also be identified. We apply the methodology to the Indian ULGs, facing complex problems like lack of autonomy, poor infrastructure, lack of human resources and mismanagement in operations.

III. Empirical Literature and Gaps

We look at the existing method and the relevant model specifications used in the literature to measure the performance of governments. Empirical literature analysing the performances of local governments using frontiers, estimate their levels of efficiency through a non-parametric (Afonso & Fernandes, 2008; Afonso & Venâncio, 2019; Balaguer-Coll et al., 2010; Benito et al., 2010, 2014; Bosch et al., 2012; de Borger & Kerstens, 1996; de Borger et al., 1994; Kalb, 2014; Narbón-Perpiñá et al., 2019; Salazar-Adams, 2024; Sampaio de Sousa et al., 2005; Seifert & Nieswand, 2014; Sorensen, 2014; Storto, 2013; Worthington & Dollery, 2000) or parametric approach of stochastic frontier analysis (Agasisti & Porcelli, 2023; Bucci et al., 2023; Ibrahim & Salleh, 2006; Kalb, 2014; Kalb et al., 2011; Šťastná & Gregor, 2015). The studies using a nonparametric approach mostly consider expenditures as inputs and service delivery as outputs, in a two-stage analysis. In the first stage, they estimate the technical efficiency scores using Data Envelopment Analysis (DEA), and in the second stage, they explain them using exogenous factors like population, level of education of the population, personal income of the population, stage of decentralisation, geographical distance from the city centre, level of commercial activity, rate of urbanisation, importance of tourism, and unemployment rate to name a few.

There exist certain gaps in the literature. First, all the studies that have attempted efficiency analysis have assessed the performance of local governments through their efficiencies in service provision. This has been explained as the ‘first strand’ of efficiency of governments in our article. The other two strands have not been cultivated to estimate the performance of local governments empirically, which can prove to be extremely important. Second, the nonparametric estimation models are only concentrated on analyses of radial efficiencies but do not consider off-radial efficiencies through slacks, leaving an important aspect of further improvement in performance.

We have a few studies which assess the performance of Indian ULGs by looking at their fiscal health (Bandyopadhyay, 2015; Bandyopadhyay & Rao, 2009; Bandyopadhyay & Sharma, 2020; Sridhar, 2007). Most of these studies only compare the revenue capacity with the expenditure needs of the local governments. Many ULGs have not been provided the autonomy to perform all the functions and finances delegated under the 74th Amendment Act (a list of functions and finances devolved to Indian ULGs under the 74th Amendment Act of the Indian Constitution is provided in Table A1, Appendix A). Most of the local governments face budgetary rigidities due to high dependence on grants or transfers and/or ineffective devolution of certain functions.

Another major gap found in the literature is that the studies tend to bypass the problems of smaller cities. Even the urban growth policy in general is skewed towards big city regions and fails to address the economic needs of smaller cities (Beel et al., 2020). Quite often than not, the small and mid-sized cities, which are huge in number, support a bigger chunk of the population and yet remain neglected. In India, although the proportion of small and mid-sized cities is way higher than that of Municipal Corporations in the total number of local governments, it is the latter that are more densely populated (Table 1).

Table 1.

Proportions of Different Types of Urban Local Governments (ULGs) and Their Population Distribution in India.

Local Governments	Number	Proportion in Total Number of ULGs (in %)	Population (in Million)	Area (sq. km)	Population Density (Number per sq. km)
Municipal corporations	209	5	172.21	22,919.13	7,514
Municipality	2,171	51	112.02	35,126.41	3,189
Town panchayats	1,879	44	32.53	19,356.10	1,680

Source: Ahluwalia et al. (2019). Authors’ calculation.

If these small and mid-sized cities grow in terms of their infrastructure provision and finances, they can attract a population from bigger cities, which are already saturated. Also, an increase in property tax rates of big cities in prime areas may result in urban sprawl, as inhabitants from the core areas would move to the surrounding smaller cities (Ermini & Santolini, 2017). This would ensure a more balanced growth of cities of different sizes, which would lead to sustainability. It makes it more imperative to study if smaller cities are performing well enough to absorb the population of bigger cities in the wake of higher taxes. Smaller cities also warrant a separate study because these cities might have operations on a smaller scale and are more vulnerable.

IV. The Present Study: Data, Variables and Models

The present study considers Tamil Nadu and Karnataka, which are amongst the better-performing states in India. A careful analysis of the Finance Commission reports (11th, 12th, 13th and 14th) shows that both Tamil Nadu and Karnataka demonstrate better performances across various fiscal indicators when compared to other Indian states and all states’ combined average levels. The indicators on revenue mobilisation, fiscal prudence, debt management, infrastructure development and fiscal self-sufficiency show a consistent performance of these two states vis-à-vis other Indian states (Government of India [GOI], 2000, 2004, 2009, 2015; RBI, 2024). On the other hand, some economic and development indicators (like state per capita income, unemployment rate, fertility rate and poverty) also reveal a superior performance of these two states (RBI, 2024) not only in the study period but also in recent times, which speaks about the consistency of the superior performances in these two states. These findings justify the choice of these two states as benchmarks for this analysis.

The distribution of ULGs across municipalities (mid-sized) and town panchayats (small) in the two states is provided in Table 2.¹ For the present study, we do not include municipal corporations in the analysis because these are big in size and have different finance sources and functions. We focus on the other local governments, which are smaller in size, larger in number, and face serious administrative and operational challenges.

Table 2.

Categories of Urban Local Governments (ULGs) Across Karnataka and Tamil Nadu Considered for the Study.

State/ULG Type	Municipality	Town Panchayat
Karnataka	133	69
Tamil Nadu	149	559
Total	282	628

Source: Government of Tamil Nadu (2016) and Government of Karnataka (2018). Authors’ computations.

Based on the conceptual framework discussed above, we consider three models using DEA. Each model has two versions: ‘a’ denotes the ‘output version’, and ‘b’ denotes the ‘input version’ of the same model. The variables considered for expenditures, service delivery and revenues are provided in Table 3:

Models 1a and 1b: Expenditure as input and service delivery as output.

Models 2a and 2b: Service delivery as input and revenue as output.

Models 3a and 3b: Revenue as input and expenditure as output.

Table 3.

Variables Used for Efficiency Estimation.

Expenditure²	Services	Revenue³
Operation and maintenance cost Labour cost Establishment expenditure	Total annual water supplied in a ULG (in liters) Total road length in a ULG (in km) Total number of streetlights in a ULG Solid waste transported after collection (tonnes per day) for Karnataka Storm water drains (in km) for Tamil Nadu Length of roads cleaned daily in a ULG (in km)	Tax revenue Non-tax revenue

Source: Authors’ selection of variables.

Note: ULB: Urban local body.

The data on expenditures, services and revenues used for the two states are for the year 2013–2014, which are sourced from the unit level surveys done by the Fourth State Finance Commission of Karnataka (constituted in 2015) and Fifth State Finance Commission of Tamil Nadu (constituted in 2014).⁴ While we recognise that we are constrained by the non-availability of data for the current year, we have attempted to utilise the most recent data available to knit three different strands of public finance theories for estimating the performance of governments. Since the thrust of the article is to offer a methodology for assessing the performance of the government, any other constraint related to the availability of data may be considered a minor limitation.

We use slack-based DEA (Banker et al., 1984) to assess the performance of ULGs. We not only analyse the radial efficiency scores but also attempt to visualise the untapped potential of ULGs through off-radial inefficiencies. For the first time, we attempt a comprehensive analysis of ULGs having slacks in this article. For this, we consider the proportion of ULGs recording slacks in a particular input or output.

Let x be the vector of inputs and y the vector of outputs. Let x^j = (x₁ _j , x₂ _j ,…, x_nj) be the input vector and y^j = (y₁ _j , y₂ _j ,…, y_nj) be the observed output vector of firm j (j = 1,2,3,…,N).

The slack-based input-oriented measure of technical efficiency of a firm t, producing output y^t from the input bundle x^t, under Variable Returns to Scale, is derived from the solution to the following problem.

\min θ^{~} = θ - Є (\sum_{k = 1}^{m} s_{k}^{+} + \sum_{i = 1}^{n} s_{i}^{-})

s . t . \sum_{j = 1}^{N} \begin{matrix} λ_{j} x^{j} + s_{i}^{-} = θ x^{t}; \end{matrix}

\sum_{j = 1}^{N} \begin{matrix} λ_{j} y^{j} - s_{k}^{+} = y^{t}; \end{matrix}

\sum_{j = 1}^{N} λ_{j} = 1;

λ_{j} \geq 0 (\forall j = 1, 2, \dots, N)

Here $Є$ is an infinitely small positive number.

Let $(θ^{*}, λ_{1,}^{*} λ_{2,}^{*} . . \dots ., λ_{N}^{*})$ be the optimal solution. Define $(x^{t} = θ^{*} x^{t})$ . Then the input-oriented technical efficiency of the firm t is given by ${T E}_{I}^{t} (x^{t}, y^{t}) = θ^{*} .$ If an input slack is positive in an input-oriented model, it means that there exists a further possibility of reducing the individual inputs even after the ULG has reached the input efficiency frontier.

The slack-based output-oriented measure of technical efficiency of firm t producing output y^t from the input bundle x^t is obtained from the solution to the following:

{\max \emptyset}^{~} = Φ + Є (\sum_{k = 1}^{m} s_{k}^{+} + \sum_{i = 1}^{n} s_{i}^{-})

s . t . \sum_{j = 1}^{N} \begin{matrix} µ_{j} x^{j} + s_{i}^{-} = x^{t}; \end{matrix}

\sum_{j = 1}^{N} µ_{j} y^{j} - s_{k}^{+} = \emptyset y^{t};

\sum_{j = 1}^{N} µ_{j} = 1;

μ_{j} \geq 0 (\forall j = 1, 2, \dots, N)

Here $Є$ is an infinitely small positive number.

Define Φ* y^t = y^t*. Then the output-oriented technical efficiency of firm t is given by

{T E}_{O}^{t} = 1 / Φ *

$s_{k}^{+} a n d s_{i}^{-}$ indicate the output and input slacks, respectively, at the optimal solution. If an output slack is positive in an output-oriented model, it means that it is possible to expand a particular output by the amount of slack even after the ULG has reached the output efficiency frontier.

V. Stylised Facts

We see that a mapping of each of the strands to a DEA model is straightforward. We offer a set of stylised facts below, spelling out the possible consequences of interpreting the radial output and input efficiency scores, comparing the output and input versions of DEA models for each strand. We also interpret the off-radial efficiencies, considering the proportion of ULGs recording slacks in outputs or inputs.

Models 1a and 1b: Expenditure as Input and Service Delivery as Output

The radial estimation of input inefficiency would mean that the ULG could have provided the same levels of services by incurring lower expenditures. We interpret this as a misappropriation or leakage of expenditures.

Radial output inefficiency would mean that there is a possibility of delivering higher levels of services with the same levels of expenditures. We interpret this as service inadequacy.

If the output-oriented radial inefficiency is greater than the input-oriented radial inefficiency, we can say that the extent of service inadequacy is greater than the extent of misappropriation of expenditures or leakages of expenditures.

The off-radial estimation of positive input slacks would mean that there is a scope for further reducing misappropriation/leakages of expenditures with the same levels of service delivery once a ULG has reached the efficiency frontier. Similarly, a positive output slack would mean that there is a possibility of providing higher levels of services with the same levels of expenditure after a ULG has attained 100% efficiency.

Models 2a and 2b: Service Delivery as Input and Revenue as Output

In this strand, radial input inefficiency would mean that the local body could have generated the same levels of revenues with lesser vectors of services. This could mean that some of the services are not charged, which could be a cause of an inadequacy of revenue handles and/or inappropriate or unrevised tax rates and user charges.

On the other hand, output inefficiency would mean that, given the same levels of services, the local body could have generated higher levels of revenues. If the service delivery is not translating into revenue generation, this shows a possibility of poor revenue collection. It could also be possible that the quality of service delivery is so poor that residents are not willing to use them and, hence, pay for those services. In addition to this, there could be administrative delays in collecting revenues, the revenue handles might not exist or might not have been devolved to the local governments.

If we find that radial inefficiency in the output-oriented approach is higher than that in the input-oriented approach, we could say that the extent of poor revenue collection is greater than the extent of the existence of missing revenue handles/uncharged services.

The off-radial estimation of positive input slacks would mean that there is a scope for further reducing the uncharged services at the same levels of revenue generation once a ULG has reached the efficiency frontier. Similarly, a positive output slack would mean that there is a possibility of improving revenue collections further, with the same levels of services, once a ULG is on the output efficiency frontier.

Models 3a and 3b: Revenue as Input and Expenditure as Output

Hereby, radial input-inefficiency, we mean that the local governments could have generated the same levels of expenditures with even lesser revenues. We interpret this as ineffective revenue channelisation. The radial output inefficiency would mean that the local governments could have incurred more expenditure with similar levels of revenues. This signals underutilised revenues by the local governments.

If the radial inefficiency in the output-oriented model is greater than that in the input-oriented one, we can say that the extent of ineffective spending is greater than the extent of ineffective revenue utilisation.

A positive input slack would mean that there is a scope for reducing ineffective revenue channelisation further, given the same levels of expenditures, once a ULG has reached the input efficiency frontier. A positive output slack would mean that there is a possibility of further incurring effective spending, given the same levels of revenues, after a ULG has attained 100% efficiency.

VI. Results and Analysis

Radial Inefficiency

Radial inefficiency measures give an idea about how much, on average, the outputs can be expanded (in case of an output-oriented model)/inputs can be contracted (in case of an input-oriented model) to reach the output/input frontier. Each model is run for each of the two states, for each size class, separately (Table A2, Appendix A).

It is interesting to find that output inefficiencies are higher in town panchayats in both Models 1a and 2a; however, input inefficiencies are higher in municipalities in both Models 1b and 2b. For Models 3a and 3b, output and input inefficiencies are higher in the municipalities. The differences in the median values of output and input efficiencies are highest between Models 3a and 3b and lowest between Models 2a and 2b. The municipalities are worse off in four out of six models of performance analysis.

We also compare the median output inefficiency and input inefficiency of local governments in Karnataka and Tamil Nadu (Table 4). We highlight the results where the difference between input inefficiency and output inefficiency scores is statistically significant according to the KS-test and the t-test (for descriptive statistics of output efficiency and input efficiency scores of all ULGs, refer to Table A2, Appendix A).

Table 4.

Output and Input Efficiency Scores in Urban Local Governments (ULGs) of Karnataka and Tamil Nadu: Some Comparisons.

ULGs	Models	Strands	Output Efficiency Score (Median)	Input Efficiency Score (Median)	D-statistics of KS-test (p Value)	t-statistics with Equal Variance (p Value)	Comparison of Inefficiencies¹	Interpretation
Karnataka municipalities	Model 2a and Model 2b	Service delivery is input and revenue is output	0.21	0.40	.692 (.000)	6.818 (.000)	Output inefficiency > Input inefficiency	Extent of poor revenue collection > Extent of existence of uncharged services
Tamil Nadu municipalities	Model 1a and Model 1b	Expenditure is input and service delivery is output	0.68	0.61	.208 (.003)	−2.057 (.041)	Input inefficiency > Output inefficiency	Extent of misappropriation/leakages of expenditures > Extent of service inadequacy
	Model 2a and Model 2b	Service delivery is input and revenue is output	0.53	0.7	.329 (.000)	−4.022 (.000)	Output inefficiency > Input inefficiency	Extent of poor revenue collection > Extent of existence of uncharged services
	Model 3a and Model 3b	Revenue is input and expenditure is output	0.54	0.41	.282 (.000)	4.119 (.000)	Input inefficiency > Output inefficiency	Extent of ineffective revenue channelisation > Extent of revenue underutilisation
Tamil Nadu town panchayat	Model 1a and Model 1b	Expenditure is input and service delivery is output	0.5	0.55	.161 (.000)	3.582 (.000)	Output inefficiency > Input inefficiency	Extent of misappropriation/leakages of expenditures < Extent of service inadequacy
	Model 2a and Model 2b	Service delivery is input and revenue is output	0.33	0.57	.361 (.000)	−9.794 (.000)	Output inefficiency > Input inefficiency	Extent of poor revenue collection > Extent of existence of uncharged services
	Model 3a and Model 3b	Revenue is input and expenditure is output	0.66	0.5	.556 (.000)	10.898 (.000)	Input inefficiency > Output inefficiency	Extent of ineffective revenue channelisation > Extent of underutilisation of revenues

We analyse results in two ways. First, we look at the similarities that exist across size classes. Then we highlight the differences across size classes.

We find that, regardless of the size of the cities, the extent of poor revenue collection is higher than the extent of the existence of uncharged services. This means that even when a service is chargeable, revenues are not generated/collected. This could be because of the poor administrative capacity of the local governments to enforce the taxes and user charges and/or to collect the taxes and user charges from the residents.

A major source of own revenue for local governments is property tax. There is an exportation of property tax from residential to non-residential properties, which leads to poor revenue generation. In addition, the local governments have poor information on the property tax base, wide exemptions and the existence of vacant properties (Rao, 2013). ULGs in Karnataka have been identified with underassessment/non-assessment of property tax, low collection efficiency, and non-compliance, resulting in lower own resources (Karnataka State Finance Commission, 2018). Rajah (2022) highlights the problem of poor allocation of resources to ULGs by the State Government of Tamil Nadu

We find that in both municipalities and town panchayats, the collection efficiency, which is defined as the ratio of collection of revenues to estimated demand raised for revenues by the government in a financial year, is abysmally low (Table 5). The smaller cities in both states perform worse than mid-sized cities.

Table 5.

Collection Efficiency of Urban Local Governments (ULGs) in Karnataka and Tamil Nadu.

Collection Efficiency	Municipality (Middle-sized City)	Town Panchayat (Small City)
Karnataka	65	57
Tamil Nadu	42.4	14.7

We also find that the extent of ineffective channelisation of revenues is greater than the extent of ineffective spending from revenues across all city sizes. This highlights the problem of underutilisation of revenues in all local governments. It could be due to a lack of will and planning in spending those resources.

The analysis shows that in the smaller cities, the extent of service inadequacy is greater than that of the leakage of expenditures. The results are strikingly opposite for mid-sized cities, where misappropriation/leakage of expenditures is higher than the extent of service inadequacy. This is to be expected, that small-sized cities have inadequate administrative capacity to operate as effectively as a mid-sized city. This could cause a greater inadequacy of service delivery in smaller cities as compared to mid-sized cities.

Additionally, we conduct a correlation analysis to understand the association between input-efficiency and output-efficiency scores in a ULG for each strand, across all city sizes (Table 6). Irrespective of the city sizes and strands, we find that there exists a high and statistically significant positive correlation between input-efficiency and output-efficiency scores. This implies that as the extent of leakage of expenditures goes up in a ULG, the level of service inadequacy also goes up and vice versa. Also, the higher the extent of inadequacy or underutilisation of revenue handles, the poorer the revenue collection and vice versa. Furthermore, ineffective revenue channelisation in a ULG is accompanied by underutilised revenues and vice versa.

Table 6.

Correlation Coefficients Between Output Efficiency and Input Efficiency Scores Across Different Strands in Urban Local Governments (ULGs) in Karnataka and Tamil Nadu (Level of Significance 5%).

Models	Municipalities			Town Panchayats
Models	Model 1a	Model 2a	Model 3a	Model 1a	Model 2a	Model 3a
State: Karnataka
Model 1b	0.9335*			0.9382*
Model 2b		0.7522*			0.8880*
Model 3b			0.7921*			0.6277*
State: Tamil Nadu
Model 1b	0.9036*			0.6364*
Model 2b		0.8573*			0.7451*
Model 3b			0.8067*			0.8328*

Note: * indicates 5% level of significance.

Off-radial Inefficiency

Radial efficiency is defined in terms of the difference between the actual position of the decision-making unit and its ideal position, which radially hits the frontier. There can be further possibilities of improvements by saving resources for a radially efficient unit through off-radial movements, indicated by the presence of slack in the inputs or outputs concerned.

We look at the proportion of ULGs having slacks in output- and input-oriented models. We want to know what proportion of ULGs may have further scope for contracting the inputs or expanding the outputs. Our analysis of off-radial inefficiency is attempted according to the different components of expenditures, services and revenues used in the models, in all the size classes of cities. The detailed results are given in Table 7.

Table 7.

Off-radial Efficiency Analysis on the Basis of Slacks.

Model Orientation	Model Specification	Outputs	Municipality	Town Panchayat
Model Orientation	Model Specification	Outputs	Proportion of ULGs Having Output Slacks (%)	Proportion of ULGs Having Output Slacks (%)
Output-oriented DEA	Model 1a: Expenditure as input and service delivery as output	Water supplied	18.09	41.56
		Road length	13.12	64.65
		Number of streetlights	30.14	12.74
		Solid waste transported	26.24	44.59
		Length of roads cleaned	48.58	19.75
	Model 2a: Service delivery as input and revenue as output	Tax revenue	6.38	45.54
	Model 2a: Service delivery as input and revenue as output	Non-tax revenue	13.12	30.10
	Model 3a: Revenue as input and expenditure as output	Operations and maintenance expenditure	8.16	7.17
		Labour cost	48.23	8.92
		Establishment expenditure	42.55	20.86
Model Orientation	Model Specification	Inputs	Municipality	Town Panchayats
Model Orientation	Model Specification	Inputs	Proportion of ULGs Having Input Slacks (%)	Proportion of ULGs Having Input Slacks (%)
Input-oriented DEA	Model 1b: Expenditure as input and service delivery as output	Operations and maintenance expenditure	3.90	40.45
		Labour cost	32.62	17.52
		Establishment expenditure	27.30	29.62
	Model 2b: Service delivery as input and revenue as output	Water supplied	34.75	70.22
		Road length	23.40	66.40
		Number of streetlights	28.01	21.66
		Solid waste transported	18.44	26.11
		Length of roads cleaned	50.35	62.74
	Model 3b: Revenue as input and expenditure as output	Tax revenue	7.45	8.76
	Model 3b: Revenue as input and expenditure as output	Non-tax revenue	9.93	6.53

Note: DEA: Data envelopment analysis; ULGs: Urban local governments.

It is interesting to note that there exist slacks in all the components of the three categories of variables used, that is, services, expenditures and revenues, in both size classes. It is very clear that the nature of challenges is different across size classes. As far as expansions of services are concerned, keeping expenditures unchanged, the highest unutilised potential for improvement in mid-sized cities is recorded in conservancy services, whereas that in smaller cities is recorded in the provision of roads. While mid-sized cities record the highest unutilised potential for improvements in the expansion of non-tax sources, given the levels of services, for smaller cities, it is the tax component. As far as the highest unutilised potential for improvement in mid-sized cities for expansion of expenditures, given the existing sources of revenues, is concerned, it is the labour cost component, whereas that in smaller cities is the establishment expenditures.

As far as the contractions of expenditures are concerned, keeping service levels unchanged, the highest unutilised potential for improvement in mid-sized cities is recorded in labour cost, whereas that in smaller cities is recorded in operations and maintenance expenditures. While mid-sized cities record the highest unutilised potential for improvements by contraction of conservancy services, given the existing levels of revenues, for smaller cities, it is water supply. As far as the highest unutilised potential for improvement in mid-sized cities through contraction of revenues, given the existing levels of expenditures, is concerned, it is the non-tax component, whereas that in smaller cities is the tax component.

VII. Conclusions

Small and mid-sized cities in India have been growing rapidly, but most of these cities are not equipped to provide basic services and, thus, a good quality of living to their populace. The 74th Amendment Act in the Constitution of India aimed at giving the ULGs autonomy to ensure effective provision of services. However, despite three decades of proposed reforms, most of the cities do not have effective autonomy with respect to their functions and finances.

This article, for the first time, proposes a novel framework to offer a holistic and nuanced analysis to assess the performance of governments by defining three distinct strands of performance rooted in the public finance literature. This framework can be replicated for assessing the performance, and thereby the financial viability, of any level of government. In this article, we assess the performance of small and mid-sized cities in Southern India, empirically, using this approach.

The results suggest that the consequences we face are partly due to an ineffective decentralisation process and administrative failures to put checks and balances in place. What we find is interesting, which may also lead to important policy recommendations relating to the financial viability of these classes of cities. This, in turn, would impact the fiscal sustainability of the governments.

First, when we consider expenditures incurred by the ULGs in determining the service delivery, we see that small and mid-sized cities are faced with an inadequacy of services. There can be two reasons: either the levels of expenditure incurred are less than those required, or there is misappropriation of expenditures. The supply bottlenecks in smaller cities, particularly with respect to water supply and roads, need to be addressed. The deterioration of roads should be checked by the maintenance of roads at regular intervals. So far as the water supply is concerned, groundwater depletion can be arrested by stopping the wastage of water. Revising the water tariffs in the states, concomitant with the demand for water, can help reduce wastage. The revision of water tariffs should be accompanied by the effective collection of tariffs by the ULGs.

The problem of service inadequacy can also be addressed by taking appropriate measures to eliminate misappropriation of expenditures and ensuring better utilisation of each revenue handle by tapping the corresponding revenue base. Often, service inadequacy is a result of the misappropriation of expenditures, which in turn is a result of a lack of administrative machinery to ensure checks and balances in the system. Also, even lower levels of expenditures, if utilised properly, can bring in higher service levels. So is the case with revenues; if utilised better, even lower revenues can result in expenditures which are higher than the existing levels. So, the starting point is proper invigilation of the existing revenue channelisation and expenditure utilisation. Administrative reforms like performance evaluation of local governments through systematic monitoring and effective audits are required to check misappropriations and leakages in expenditures in both small and mid-sized cities. The smaller cities are crippled with leakages in establishment expenditures, which also warrant a timely audit of budgets. In mid-sized cities, the local governments must ensure that there is adequate recurring expenditure for the proper upkeep of services.

Second, services can help generate revenues in the ULGs. There can be two reasons. The inadequacy of services might fail to attract the desired levels of revenues because the revenues which are generated are not collected for administrative lacunae, or the revenue handles are not exercised in the city. We find that the ULGs in India, in general, are characterised by poor revenue collections, which are an outcome of low administrative efficiency, and it requires an effective devolution of revenue handles to the ULGs. To improve revenue generation, some of the revenue streams can be devolved to the ULGs, such as tax on vacant land, electricity cess, fire tax, drainage tax, tolls on roads and bridges, tax on vehicles, development charge, sanitary cess and tax on parking spaces in any non-residential building. Ensuring effective devolution of finances can help to rediscover, implement and utilise the missing revenue handles.

Third, the revenues might not get converted to expenditures because of three reasons. It might be the case that the revenues generated are too little to be spent; the productive routes in which revenues can be spent are too little; or revenues are being spent but are not a part of the accounting process. The nature of revenue sources is important to consider for the utilisation of revenues. The problem of ensuring expenditures on services from the revenues generated can be resolved if the ULGs bring in more non-tax revenue handles under the purview, because the non-tax revenues can bring in more accountability in expenditures. Also, a periodic audit of accounts should be done to ensure timely and effective utilisation of revenues for expenditures on services.

While the High Powered Expert Committee Report on Indian Urban Infrastructure and Services (GOI, 2011) has suggested that the levy of user charges should be a major reform agenda for augmenting the revenues from the non-tax sources, subsequent Central Finance Commission Reports have recommended that user charges should be designed properly, such that local bodies can at least recover the operation and maintenance costs of services. This has been reiterated by many State Finance Commissions in India, which have further recommended that the rates of user charges should be revised periodically, and that people should be encouraged to pay user charges by educating them on the merits of such charges. Parking fees, one-time street charges, and fees from mobile towers are some of the heads for which rates need to be revised. Among the new components that are proposed are the conservancy charges, congestion charges and cable operator charges. These would be instrumental in bringing about sustainable and inclusive urban infrastructure services, and thereby an inclusive growth and development in urban India.

Footnotes

Acknowledgement

The authors would like to thank the anonymous reviewers and the editor/s for their valuable suggestions and comments. This has helped improve the quality and clarity of this manuscript. An earlier version of the paper was presented in North American Productivity Workshop held during June 8-12, 2020, hosted by Miami Herbert Business School, Florida, USA. It benefited from the comments by the experts and participants in the conference.

Declaration of Conflict of Interests

The authors declared no potential conflicts of interest regarding the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

Notes

Appendix A

References

Afonso

, & Fernandes

(2008). Assessing and explaining the relative efficiency of local government. Journal of Socio-Economics , 37(5), 1946–1979. https://doi.org/10.1016/j.socec.2007.03.007

Afonso

, & Venâncio

(2019). Local territorial reform and regional spending efficiency. Local Government Studies , 46(6), 888–910. https://doi.org/10.1080/03003930.2019.1690995

Agasisti

, & Porcelli

(2023). Local governments’ efficiency and its heterogeneity—Empirical evidence from a stochastic frontier analysis of Italian municipalities 2010–2018. Applied Economics , 55(25), 2902–2927. https://doi.org/10.1080/00036846.2022.2107988

Ahluwalia

I. J.

, Mohanty

P. K.

, Mathur

, Khare

, Roy

, & Mangla

(2019). Finances of municipal corporations in metropolitan cities of India . New Delhi: Indian Council for Research on International Economic Relations (ICRIER). https://icrier.org/publications/finances-of-municipal-corporations-in-metropolitan-cities-of-india/

Alegre

J. G.

(2010). Decentralization and the composition of public expenditure in Spain. Regional Studies , 44(8), 1067–1083. https://doi.org/10.1080/00343400903365151

Backhaus

, & Wagner

R. E.

(1987). The cameralists: A public choice perspective. Public Choice , 53(1), 3–20. https://doi.org/10.1007/BF00115651

Bahl

R. W.

, & Bird

R. M.

(2008). Subnational taxes in developing countries: The way forward. Public Budgeting & Finance , 28(4), 1–25.

Balaguer-Coll

, Prior

, & Tortosa-Ausina

(2010). Decentralisation and efficiency of local government. Annals of Regional Science , 45, 571–601.

Bandyopadhyay

(2015). Local government finance: Challenges in revenue-raising at the Municipal Corporation of Delhi. Commonwealth Journal of Local Governance , 16/17, 59–82.

10.

Bandyopadhyay

, & Rao

(2009). Fiscal health of selected Indian cities (World Bank Policy Research Working Paper No. 4863). World Bank.

11.

Bandyopadhyay

, & Sharma

(2020). Assessing the performance of Kolkata Municipal Corporation: Implications for service delivery (International Center for Public Policy Working Paper 20-02). Georgia State University, Andrew Young School of Policy Studies.

12.

Banker

R. D.

, Charnes

, & Cooper

W. W.

(1984). Some models for estimating technical and scale inefficiencies in data envelopment analysis. Management Science , 30(9), 1078–1092.

13.

Beel

, Jones

, & Plows

(2020). Urban growth strategies in rural regions: Building The North Wales Growth Deal. Regional Studies , 54(5), 719–731. https://doi.org/10.1080/00343404.2019.1669783

14.

Benito

, Bastida

, & García

J. A.

(2010). Explaining differences in efficiency: An application to Spanish municipalities. Applied Economics , 42(4), 515–528. https://doi.org/10.1080/00036840701675560

15.

Benito

, Solana

, & Moreno

(2014). Explaining efficiency in municipal services providers. Journal of Productivity Analysis , 42, 225–239.

16.

Bird

R. M.

(1993). Threading the fiscal labyrinth: Some issues in fiscal decentralization. National Tax Journal , 46(2), 207–227.

17.

Bird

R. M.

(2010). Subnational taxation in developing countries: A review of literature (Policy Research Working Paper 5450). World Bank.

18.

Bosch

, Espasa

, & Mora

(2012). Citizen control and the efficiency of local public services. Environment and Planning C: Politics and Space , 30(2), 248–266. https://doi.org/10.1068/c1153r

19.

Bucci

, Ferrara

, & Resce

(2023). Local government spending efficiency and fiscal decentralization: Evidence from Italian municipalities. Applied Economics , 47(7), 680–699. https://doi.org/10.1080/00036846.2023.2169241

20.

Buchanan

J. M.

(1949). The pure theory of government finance: A suggested approach. Journal of Political Economy , 57(6), 496–505. https://doi.org/10.1086/256880

21.

de Borger

, & Kerstens

(1996). Radial and non-radial measures of technical efficiency: An empirical illustration for Belgian local governments using FDH reference technology. Journal of Productivity Analysis , 7(1), 41–62.

22.

de Borger

, Kerstens

, Moesen

, & Vanneste

(1994). Explaining differences in productive efficiency: An application to Belgian municipalities. Public Choice , 80(3/4), 339–358.

23.

Ermini

, & Santolini

(2017). Urban sprawl and property tax of a city’s core and suburbs: Evidence from Italy. Regional Studies , 51(9), 1374–1386. https://doi.org/10.1080/00343404.2016.1190448

24.

Government of India (GOI). (1993). The constitution (Seventy-fourth Amendment act) .

25.

GOI. (2000). The report of the 11th Finance Commission (2000–2005) . Ministry of Finance.

26.

GOI. (2004). The report of the 12th Finance Commission (2005–2010) . Ministry of Finance.

27.

GOI. (2009). The report of the 13th Finance Commission (2010–2015) . Ministry of Finance.

28.

GOI. (2011). The high powered expert committee for estimating the investment requirements for urban infrastructure services . Ministry of Urban Development.

29.

GOI. (2015). The report of the 14th Finance Commission (2015-2020) . Ministry of Finance.

30.

Government of Karnataka. (2018). Report of the Fourth Finance Commission, Karnataka . State Finance Commission.

31.

Government of Tamil Nadu. (2016). Report and recommendations of Fifth Finance Commission, Tamil Nadu . State Finance Commission.

32.

Grisorio

M. J.

, & Prota

(2015). The impact of fiscal decentralization on the composition of public expenditure: Panel data evidence from Italy. Regional Studies , 49(12), 1941–1956. https://doi.org/10.1080/00343404.2013.859665

33.

Ibrahim

F. W.

, & Salleh

M. F. M.

(2006). Stochastic frontier estimation: An application to local governments in Malaysia. Malaysian Journal of Economic Studies , 33(1–2), 85–95.

34.

Kalb

(2014). What determines local governments’ cost efficiency? The case of road maintenance. Regional Studies , 48(9), 1483–1498. https://doi.org/10.1080/00343404.2012.731044

35.

Kalb

, Geys

, & Heinemann

(2011). Value for money? German local government efficiency in a comparative perspective. Applied Economics , 44(2), 201–218. https://doi.org/10.1080/00036846.2010.502110

36.

Karnataka State Finance Commission. (2018). Report of the Fourth State Finance Commission, Karnataka . Government of Karnataka.

37.

Kitson

, Martin

, & Tyler

(2004). Regional competitiveness: An elusive yet key concept? Regional Studies , 38(9), 991–999. https://doi.org/10.1080/0034340042000320816

38.

Musgrave

R. A.

(1959). The theory of public finance . McGraw-Hill.

39.

Musgrave

R. A.

, & Musgrave

(1989). Public finance in theory and practice . McGraw-Hill.

40.

Narbón-Perpiñá

, Balaguer-Coll

, & Tortosa-Ausina

(2019). Evaluating local government performance in times of crisis. Local Government Studies , 45(1), 64–100. https://doi.org/10.1080/03003930.2018.1507908

41.

Oates

W. E.

(1969). The effects of property taxes and local public spending on property values: An empirical study of tax capitalization and the Tiebout hypothesis. Journal of Political Economy , 77(6), 957–971.

42.

Office of the Registrar General & Census Commissioner, India. (2011). Census of India 2011: Provisional population totals, urban agglomerations and cities . Ministry of Home Affairs, Government of India.

43.

Oates

W. E.

(2008). On the evolution of fiscal federalism: Theory and institutions. National Tax Journal , 61(2), 313–334.

44.

Olson

(1969). The principle of ‘fiscal equivalence’: The division of responsibilities among different levels of government. American Economic Review , 59(2), 479–487.

45.

Rajah

N. L.

(2022 March 10). Tamil Nadu: How urban local bodies can work better. The Times of India .

46.

Rao

M. G.

(2013). Property tax system in India: Problems and prospects of reform (Working Paper No. 2013-114). National Institute of Public Finance and Policy.

47.

Reserve Bank of India (RBI). (2022). Report on municipal finances .

48.

RBI. (2024). Handbook of statistics on Indian states 2023–2024 .

49.

Sacchi

, & Salotti

(2017). The influence of decentralized taxes and intergovernmental grants on local spending volatility. Regional Studies , 51(4), 507–522. https://doi.org/10.1080/00343404.2015.1111512

50.

Salazar-Adams

(2024). The influence of organisational and staff characteristics on local governments’ efficiency: The case of Mexican municipalities. Local Government Studies , 50(2), 287–306. https://doi.org/10.1080/03003930.2023.2198218

51.

Sampaio de Sousa

, Cribari-Neto

, & Stosic

B. D.

(2005). Explaining DEA technical efficiency scores in an outlier corrected environment: The case of public services in Brazilian municipalities. Brazilian Review of Econometrics , 25(2), 287–313.

52.

Samuelson

(1956). Diagrammatic exposition of a theory of public expenditure. The Review of Economics and Statistics , 37(4), 350–356.

53.

Seifert

, & Nieswand

(2014). What drives intermediate local governments’ spending efficiency: The case of French départements. Local Government Studies , 40(5), 766–790. https://doi.org/10.1080/03003930.2013.812962

54.

Slack

(2011). Financing large cities and metropolitan areas (IMFG Papers on Municipal Finance and Governance No. 3). Institute on Municipal Finance and Governance.

55.

Sorensen

R. J.

(2014). Political competition, party polarization and government performance. Public Choice , 161, 427–450. https://doi.org/10.1007/s11127-014-0168-0

56.

Sridhar

(2007). Reforming urban service delivery in developing countries: Evidence from a case study in India. Economic & Political Weekly , 42(33), 3403–3413.

57.

Šťastná

, & Gregor

(2015). Public sector efficiency in transition and beyond: Evidence from Czech local governments. Applied Economics , 47(7), 680–699. https://doi.org/10.1080/00036846.2014.978077

58.

Stiglitz

J. E.

(2005). The role of government in economic development. In A. Bagchi (Ed.), Readings in public finance (pp. 156–175). Oxford University Press.

59.

Storto

(2013). Evaluating technical efficiency of Italian major municipalities: A data envelopment analysis model. Procedia - Social and Behavioural Sciences , 81, 346–350.

60.

Tiebout

C. M.

(1956). A pure theory of local expenditures. Journal of Political Economy , 64(5), 416–424. https://doi.org/10.1086/257839

61.

Worthington

A. C.

, & Dollery

B. E.

(2000). Measuring efficiency in local governments’ planning and regulatory function. Public Productivity and Management Review , 23(4), 469–485.