Thursday, April 08, 2021

Measuring institutional capacity in property tax systems: A case study of ten cities in India

by Diya Uday.

Property tax is ubiquitous with municipal finance. It provides local governments with the means to execute development strategies. In theory, property tax is an ideal candidate for supporting fiscal strategies in decentralised economies because the tax base is immobile making base identification and enforcement relatively easy (Kelly 2013). There are indications, however, that in India, we have not succeeded in doing property taxation well.

A national-level indicator of the performance of property taxes is the percentage of revenue generated from property taxes to the national GDP. Studies indicate that the proportion of revenues from property tax to GDP in India is low when compared with other countries. At the state-level, where property tax is a major source of revenue, there is evidence of revenue shortfalls, indicating the need for reforms. The policy responses for increasing revenues from property taxation, include increasing tax rates, revising taxation criteria and suggesting floor tax rates. But will these interventions be successful in improving the performance of the property tax system in cities?

A key factor in determining the success or failure of any policy intervention is institutional capacity. Policy interventions such as increases in property tax rates assume that ULBs are operating at optimal levels of institutional capacity and therefore increases in tax rates, property values or even improvements in tech infrastructure will optimise revenues from property taxes. In particular, these policies are founded on two main assumptions:

  • that ULBs have adequate human resources and the technical capacity to assess and demand taxes correctly;
  • having assessed taxes correctly, ULBs have the enforcement capacity to collect the entire tax demanded.

To achieve revenue optimisation from property tax it is important to first get tax administration right. Without this, it is unlikely that local governments will be able to capture the full extent of the property tax potential even with tax rate increases or technological interventions. This raises the important question: what is the current capacity of ULBs in property taxation?

In the literature we see indications of deficiencies in the institutional capacity of ULBs in performing some major tax functions like tax collections (World Bank 2004; Mathur et. al 2009; Bandyopadhyay 2014). While these studies give us valuable insights, this literature is not recent. Institutional capacity may have improved over time given the recent concentration of schemes to improve local governance such as the Smart Cities Mission and the Jawaharlal Nehru National Urban Renewal Mission (JNNURM). There is a need for new studies that will give us insights into the current state of institutional capacity.

In this article, we therefore measure the institutional capacity of some property tax functions in a sample set of cities in India. Our aim in doing so is two fold:

  • to gain insights on the current level of institutional capacity in some property tax functions in a sample set of cities.
  • in doing so we attempt to demonstrate that policy interventions must not presume the existence of adequate institutional capacity.

Our findings contribute to the existing literature on the state of property tax administration in India. In addition to this, we question the current approach to measuring administrative functions in the property tax system. We suggest an alternative approach for a more accurate diagnosis of the problems in administration.

A case study of ULB capacity in ten cities

We undertake two levels of analysis. We first examine the institutional capacity of ULBs in property tax collections in a sample set of cities. We then analyse the human resource allocation in the property tax departments in some ULBs. We use our findings to gain insights on institutional capacity in ULBs in a set of sample cities.

Sample selection: Our selection of the cities was driven by the location of the city and the availability of data. Our final selection includes a list of metropolitan and tier-2 cities located across ten different states in India. The selected cities are Chennai, Pune, Indore, Vishakapatnam, Shivamoga, Varanasi, Surat, Warangal, Kota and Bilaspur.

1. Measuring collection capacity

Methodology: We measure collections by calculating the Tax Collection Ratio (TCR), a commonly used method for measuring tax collections. Applying this method, we calculate the TCR as the difference between the tax demand made and the actual tax collected across each of the five years for which the data was available in each of the sample cities (2013-2018). We then calculate the TCR as a percentage value. We use this percentage value as a proxy to demonstrate the level of administrative capacity of a given city by taking 100 per cent as the benchmark. For instance, if the TCR percentage of a given city is 90 per cent, we interpret this to mean that the city has 90 per cent institutional capacity. Such a city has a higher level of institutional capacity when compared with a city in which the TCR percentage is 80 per cent, indicating a higher deficit in tax collections.

Table 1 sets out (i) the average property tax collected in ten cities across five years and (ii) the minimum and the maximum property tax collection across years in the period of study.

Table 1: City-wise average property tax collections (2013-2018)
CityStateAverage TCR (%)Minimum tax collection (as a % of tax demanded in that year)Maximum collection (as a % of tax demanded in that year)
ChennaiTamil Nadu9074 (2013-14)106.60 (2017-18)
PuneMaharashtra96.3687 (2017-18)109.77 (2015-16)
IndoreMadhya Pradesh80.2572.16 (2014-15)106.48 (2016-17)
VishakapatnamAndhra Pradesh114.2924.42 (2017-18)265.52 (2015-16)
ShivamogaKarnataka98.3697.86 (2014-15)99.30 (2016-17)
VaranasiUttar Pradesh9692 (2013-14)98.97 (2017-18)
SuratGujarat84.6576.65 (2015-16)84.21 (2014-15)
WarangalTelangana79.8775.12 (2013-14)82.75 (2016-17)
KotaRajasthan58.8637.04 (2013-14)96.14 (2016-17)
BilaspurChattisgarh90.275.52 (2017-18)122.95 (2013-14)

Source: Author's calculations from Smart Cities Mission data

Findings: We find that no city in the sample has achieved 100 per cent TCR. Only one city i.e. Shivamoga has close to 100 per cent of tax collections. There is a deficit in property tax collection across all the cities in the sample (distance from 100 per cent collection of tax demanded). We do find, however, that half the ULBs in the samples have achieved the goal of 90 per cent efficiency as set by the JNNURM. We also find that there are variations in property tax collection across cities. While in some cities the collections are below sixty per cent (Kota), others have a much higher percentage of collection (Shivamoga and Pune).

We also see a variation in the TCR within the same city. For instance, Kota has a maximum TCR of 96.14 per cent in one year (2016-17) but a low TCR of 37.04 per cent in another (2013-14). Similarly, Vishakpatnam has an over collection of 265.52 per cent in the year 2015-16 but under collection of 24.42 per cent in 2017-18. Even in cities like Pune or Chennai, which have a high average TCR across five years (column 3), the minimum TCR (column 4) and maximum TCR (column 5) vary. In half of the cities in the sample, we also see tax collection exceeding the maximum tax demand in a single year (column 4) for Chennai, Pune, Indore, Vishakapatnam and Bilaspur.

2. Examining human resource allocation

Our second level of analysis examines the human resource capacity in the property tax departments in a set of sample cities. The human resources could affect the TCR in two ways: First, the technical capacity of the human resources to apply the rules correctly. For instance, the ability to correctly identify taxable properties, ascertain property values, apply the assessment formula to a given assessee and determine amounts due. Second, the number of personnel in the department could potentially affect the level of accuracy in tax functions. For instance, an inadequate number of resources could increase inaccuracies. In this analysis, we focus on the second aspect of human resources, the number the personnel to examine whether a higher number officers alone leads to a better TCR.

Methodology: We collected data on the number of officers in the property tax department in the sample cities for which this data was readily available. The cities for which this data was readily available were Chennai, Pune, Vishakapatnam, Shivamoga, Varanasi, Warangal and Bilaspur.

Given the paucity of data on the number of taxable properties in the city, we device an indicator to estimate the number of taxable properties in the city using proxies. For this, we first collect Census 2011 data on the number of households living in permanent structures within the municipal area. We then calculated the number of officers per 10,000 households. We also collect data on the total area (sq. km) of the city and compare this to the administrative strength.

Table 2 sets out the administrative strength of the property tax department, the number of households living in permanent structures within the municipal area, the estimated officer to households ratio (per 10,000 households) and the city area in the sample cities for which this data was available.

Table 2: Comparing city-wise human resource allocation and TCR
CityStateAverage TCR (%)Adminis-trative strength (no. of officers)No. of households in permanent structuresAllocation of officers (per 10,000 households)City area (sq. km)
ChennaiTamil Nadu9027610,40,94831,189
PuneMaharashtra96.36416,83,26717,256.46
VishakapatnamAndhra Pradesh114.29564,25,40916,501
ShivamogaKarnataka98.3612558,826218,477.84
VaranasiUttar Pradesh963211,64,014191,535
WarangalTelangana79.87681,41,7505406
BilaspurChattisgarh90.28057,292146,377

Source: City municipal websites and Census 2011

Findings: We find that some cities with higher a TCR, also have a higher officer to households ratio. For instance, Shivamoga has the highest TCR and the highest level of administrative strength. However, we see that cities with a low TCR, do not have the lowest administrative strength. For instance, Warangal is has the lowest TCR in the sample, but not the lowest officer to households ratio.

We observe that cities with similar TCR scores do not have similar personnel to households ratios. For instance, the officer to household ratios for similar TCR cities such as Varanasi and Pune or Chennai and Bilaspur are false, demonstrating a variation in human resource allocation even across cities with the same TCR levels. Further, cities with a larger area also do not always have a higher allocation of officers. For instance, Varanasi has a smaller area than Vishakapatnam, but a higher number of officers. Chennai has a smaller area than Shivamoga, but a higher number of officers than Shivamogga. We find not consistent pattern in the manner in which human resource allocation is done across cities.

Limitations: (i) We use the number of households living in permanent structures within the municipal limit as a proxy for the number of taxable properties in a city. This does not take into account the commercial property coverage of a city. (ii) Another proxy for the number of properties in a city is the area of a city, however, a larger city may be less dense and have fewer properties than a smaller and more dense city which may have a larger number of properties (iii) The estimates are only as accurate as the data available on government websites.

Learnings for property tax reforms

The findings from our case study offer insights for property tax policy reforms in ULBs:

Presumption of adequate capacity: Our study finds deficiencies in institutional capacity in tax collections across ULBs. From a reforms perspective, even if tax rates are increased, unless the present institutional capacity is improved, revenues from property taxes might continue to be affected. Further, while our study examines the institutional capacity in one tax function - tax collections, it is likely that there are deficiencies even across other functions. This may affect the outcomes from the current set of policy interventions which focus on increasing revenues by changing the design of the tax system rather than fixing the problems in the administration.

Effect of variation across ULBs: Our findings demonstrate a variation in the capacity of ULBs to carry out property taxation. We are therefore likely to see varying levels of success even for the same set of reforms across ULBs because of the different levels of institutional capacity.

Inconsistencies within ULBs: We not only see a variation in the TCR across ULBs, we also see variation in the TCR within the same ULB across different years. This is demonstrated by the variation in the minimum and maximum collection ratios of cities in our sample. This means that even cities with an overall higher average capacity might have low or high collections in a given year. For instance, the minimum TCR in Vishakapatnam is 24.42 per cent across five years and the maximum is 265.52 per cent. Similarly, the minimum TCR in Kota across five years is 37.04 and the maximum is 96.14 indicating a wide variation in the tax collections even by the same authority. While it is unclear why this is the case, this indicates some inconsistencies in capacity levels.

Management of human resources: Our findings indicate that the institutional capacity in property tax systems is not only a function of administrative capacity in terms of the number of personnel. For instance, while we see that Shivamoga has the highest officer to households ratio and the highest TCR, Pune had a lower officer to households ratio but has the second highest TCR. Similarly, despite having a similar TCR, Chennai and Bilaspur have very different human resource allocations. Therefore, increasing the strength of the administration alone may not yield better outcomes in the assessment and collection of property taxes. Instead, improving the technical capabilities of the administration or effective utilisation of the existing human resource capacity by ULBs might yield results. For instance, Bahl et. al 2013, suggest that tax authorities in developing countries are unable to capture economies of scale.

A new approach to measurement

In the course of this study, we found that the existing approach to the measurement of tax functions in the literature has two main problems. First, studies examine tax collections as an isolated administrative function and not as a product of the preceding tax functions. Second, because of this, these studies tacitly assume that the administrative processes that precede tax collections, such as the tax assessment and all the processes that make up tax assessment are accurately done. This in turn affects the diagnosis of the problems in administration.

We posit instead, that the property tax system comprises of a series of interconnected administrative processes that determine the overall outcome of revenue generation from property tax. Each process determines the success of the next. Errors in administering one process will have repercussions for the accuracy and success of the processes and functions that follow. For instance, tax collection is not just a product of the enforcement function of the ULBs. It is also a function of accurately assessing taxes due. Similarly, the accuracy of the tax assessment function is determined by (i) the maintenance of a database of all taxable properties in the city (ii) regular updation of this database, (iii) correct valuation of the properties in the database, (iv) correct application of the tax formula for these valued properties and (v) determining permitted exemptions. Table 3 set outs an indicative list of the functions that work to together form a chain of administrative processes which ultimately determine tax collections.

Table 3: Indicative list of processes involved in tax assessment and collection
FunctionProcesses
A. Accurate tax assessment i. Maintaining a property records database of all taxable properties
ii. Updating the property records database
iii. Correct valuation of properties in the database
iv. Correct application of the tax formula
v. Correct determination of exemptions and concessions
B. Accurate tax collectioni. Making a correct tax demand (= Ai+Aii+Aii+Aiv+Av)
ii. Enforcement to collect tax demanded

When we break down administrative functions into smaller processes and view each function as being linked to the next, the result of measuring of any one administrative function will provide us with insights on the accuracy of not just the function being measured but also the previous functions in the chain of administration. For instance, the TCR of a ULB is an indication of the institutional capacity of not only tax collection but also of assessing tax correctly and getting the processes associated with the functions of assessment and then collection right. In this view, a TCR of 90 per cent potentially indicates not only a failure by the ULB to recover 10 per cent of the tax demanded but also potential inaccuracies in assessment for 10 per cent of the tax demanded, leading to appeals and pending cases on account of which payment might not have been done by assesses.

Our learnings from the case study, therefore, are not indicative of capacity issues just in tax collection, but could also be on account of inaccurate tax assessments. This analysis, in line with reports on poor tax assessments in ULBs.

This approach has two advantages over the traditional approach. It breaks down and highlights all the processes involved in property tax administration. In doing so, it allows us to more accurately diagnose the specific function at which the process fails.

Conclusion

We carried out this case study to demonstrate the importance of institutional capacity in the property tax system of ULBs. We have two main findings which are as follows:

First, we demonstrate that the problems in institutional capacity exist across a majority of our sample cities. This signals that there are potential capacity problems in many if not all cities across India. It is unclear therefore whether the present set of interventions to increase property tax revenues will yield optimum outcomes. Our findings demonstrate that it is important to precede policy interventions with the measurement of institutional capacity in the property tax system. We cannot presume the existence of adequate institutional capacity. This is in line with the literature that suggests that infrastructure and institutions are the foundation for achieving effective policy outcomes (Kelkar and Shah 2019, Pritchett et al 2012, Subramaniam and Felman 2021).

Second, deficits in the TCR are not just signals for improving capacity in tax collections and enforcement but also in tax assessment and all allied administrative processes. It is therefore difficult to diagnose which part of the property tax administration requires reform. A failure at any one point of the system has repercussions for the remaining functions. We, therefore, need a comprehensive framework for measuring institutional capacity at the level of each process of the property tax system, some of which are illustrated in Table 3.

Our study also demonstrates that while most cities have some way to go, some cities have achieved higher levels of TCR than others, indicating that they have perhaps learnt to do assessments and collections better than others. We also see that some cities appear to have achieved better utilisation of administrative strength than others. There are perhaps lessons in tax assessment and collection in these cities that other ULBs in India can learn from. A case study of the good practices in collection and assessment in these cities might offer insights for better property tax administration in other cities in India.

References

Arvind Subramaniam and Josh Felman, The Economy and Budget: Diagnosis and Suggestions, January 2021.

Matt Andrews, Lant Pritchett, Michael Woolcock, Looking Like a State: Techniques of Persistent Failure in State Capability for Implementation, CID Working Paper No. 239 June 2012.

O. P Mathur, Debdulal Thakur and Nilesh Rajyadhyaksha, Urban Property Tax Potential in India, National Institute of Public Finance and Policy, 2009.

Roy W. Bahl, Johannes F. Linn and Deborah L. Wetzel, Governing and Financing Metropolitan Areas in the Developing World, Lincoln Institute of Land Policy, Pages 1-30, 2013.

Simanti Bandyopadhyay, Municipal Finance in India: Some Critical Issues, ICPP Working Papers 14-21. May 2014.

Roy Kelly, Making the Property Tax Work, ICEPP Working Papers. 42, 2013.

Vijay Kelkar, Ajay Shah, In Service of the Republic: The Art and Science of Economic Policy, 2019.

World Bank, India: Urban Property Taxes in Selected States, 2004.

Diya Uday is a senior researcher at the Finance Research Group, Mumbai. The author would like to thank Ajay Shah, Susan Thomas and the anonymous referee for their valuable insights, comments and guidance for this work.

2 comments:

  1. Is there some problem in Surat numbers in Table 1?

    ReplyDelete
    Replies
    1. Yes. Thank you for pointing out the discrepancy. This has been fixed.

      Delete

Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticism is perfectly okay; uncivilised language is not. We delete any comment which is spam, has personal attacks against anyone, or uses foul language. We delete any comment which does not contribute to the intellectual discussion about the blog article in question.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.