Study on State Revenue Augmentation in the wake of COVID-19 Pandemic

The onset of the COVID pandemic in 2020 dealt a dual fiscal shock to the Government of the National Capital Territory of Delhi (GNCTD). Running a revenue surplus budget prior to 2020, the government found its expenditure rising sharply as it rallied to meet the rising healthcare needs of millions of its citizens. Simultaneously, the slowdown in economic activity led to a drastic reduction in public revenue.

It was in this context that the Dialogue and Development Commission of Delhi (DDC) was tasked with carrying out a comprehensive study on revenue augmentation possibilities across multiple departments of GNCTD. Given its background and expertise in carrying out such in-depth studies on public finance, DDC brought on board CEGIS to study measures to augment the government's revenue sources and boost economic growth by reforming its tax structure and administration.
 

APPROACH AND PROGRESS 
 

A Memorandum of Understanding (MOU) was signed between the Department of Finance, GNCTD and CEGIS in December 2020 for a period of 5 years, with the express purpose of increasing public revenue. CEGIS has since been embedded with the DDC and the Department of Trade and Taxes (DOTT) and has been actively supporting the government's efforts on revenue augmentation on various fronts.

The approach has focused on the five key levers of public revenue augmentation, namely
 

1. Tax design,
2. Tax administration,
3. Tax avoidance and evasion,
4. Dispute resolution, and
5. Interaction with other taxes.

Leveraging both local administrative expertise as well as international scientific research, DDC and CEGIS have identified the applicability of each of these levers to various tax streams to design potential interventions. Three forms of interventions have thus been identified and affected during this period.

First, rigorous, advanced analytics have been leveraged to identify the scope for modifications, the extent of distortions, and avenues for evasion. Both descriptive and inferential analysis, in addition to advanced machine learning systems, have been brought to bear. These have been applied to various contexts including – Mapping of tax collections to economic activity to identify underperforming sectors; Identification of fake firms and “invoice mills”; Assessing the impact of previous and ongoing design changes on tax collections.

Second, taxpayer-facing interventions have been instituted to improve compliance by easing the compliance burden and costs for taxpayers in the aftermath of a difficult period for businesses and citizens and boosting economic activity in struggling sectors. These include overarching policy changes, largely pertaining to tax design as well as programmatic interventions targeted at specific subgroups within the taxpayer base. Some examples of the former include offering a 20% discount on circle rates in February 2021 which has led to significant growth in the real estate sector, with both transactions and tax collections not only matching but exceeding pre-COVID trends. Permanent changes in tax rates and guidance values are also being assessed by the government. On the latter front, nudge-based interventions to improve first-order filing compliance, waivers and amnesties for past dues, and lottery-based systems to improve citizen engagement and reduce evasion at the last mile, are some of the interventions already implemented or in the process of finalisation and roll out.
 

The third form of intervention is inward looking which seeks to reform tax administration to increase efficiency and reduce costs. Being designed and implemented in collaboration with the DOTT, this has entailed
 

1. Rationalisation of workforce allocation based on changing workload and priorities,
    
2. Creation of a performance management system to track outcomes and create structures for supportive supervision,
    
3. Creating and institutionalising competency-based capacity-building systems drawing from the principles of Mission Karmayogi to improve department capacity, and
    
4. Upgrading and rationalising the ward structures to better reflect the changed economic landscape of the state of Delhi.
    

As of mid-2022, several of these interventions have been piloted and implemented with constant monitoring, learning, and evaluation support. Whereas some of these are long-term changes which are likely to improve outcomes over a period, even at the end of the first year, some impact has been seen in terms of improved revenue collections. The two tax streams in which the implementation has progressed the furthest, namely GST and Stamp Duty, Delhi has seen the highest collections in recent years (discounting the COVID years). The overall objective of the exercise is to bring the tax to GSDP ratio of the state of Delhi to comparable levels as traditionally high-growth states like Maharashtra and Karnataka.

A brief description of some key interventions undertaken through this exercise is provided in the following sections.
 

A. Department Of Trade And Taxes (DOTT)

The Goods and Services Tax regime is yet in its formative years, with policy modifications being affected. One of the key challenges for tax administrations across the country has been to pivot its structures and processes to the new system. DDC and CEGIS have been working with DOTT on strengthening organisational and individual capacity as well as data-driven processes for improving taxpayer compliance.
 

1. Nudge Station
 

One of the key systems that are being worked on is a “Nudge Station'' for faceless taxpayer communications and instituting non-coercive methods to improve compliance. Currently geared towards improving filing compliance for regular (and payment-based) returns, the system can be repurposed and expanded to achieve other objectives.

A Randomised Control Trial was conducted by CEGIS in collaboration with Professors from the University of California and Ashoka University. Several nudge-based interventions (reminder, deterrence or a customised deterrence message based on past filing behaviour) were tested with 2.13 lakh taxpayers with IVR calls from September to November 2021. The findings indicate a significant impact on both non-filing and late billing patterns, as well as revenue generation, of the intervention. Heterogeneity and network analyses also provide directions for future nudges and other departmental actions, thus acting as the first response measure in case of non-compliance. This system is now being institutionalised and scaled up.
 

2. Artificial Intelligence and Machine Learning
 

One of the key opportunities provided by the GST system is the availability of large amounts of detailed taxpayer
data on a regular basis. To effectively utilise this data and tackle evasion, methods are being applied to identify fraudulent firms. Data across different sources are being merged and analysed to recognise patterns of such firms making it easier to track them. The first iteration of the model, containing a set of basic/limited engineered variables, has been run to test the effectiveness and capacity of the hardware. The model has also provided the probability scores for each prediction, which is currently under analysis.
 

The ML model can support targeted physical inspection of suspect firms and can be repurposed to identify not only invoice mills, but also identify frauds at the transactional level. For example, a firm may be conducting legitimate business but may be conducting a small percentage of illegal transactions to reduce tax liability marginally. We can also develop algorithms to identify such transactional-level frauds. Similar models can also be used by other departments for their own purposes, such as controlling expenses on rolling/regular contracts, or for HR performance management, to name a few.
 

3. Competency-based Capacity Building 
 

Through this collaboration, the officials of the Department of Trade and Taxes are being supported through Competency based Capacity Building for technical, behavioural, and organisational aspects along the principles which are being applied in Mission Karmayogi. There is a need to upgrade and upskill the frontline cadre (previously trained in VAT) to deal with theoretical and practical aspects of the new taxation system. These pertain to both hard and soft skills such as assessment, scrutiny, seizures, audit, client(public) engagement, sectoral/industrial expertise, IT skills etc. Along with the update of the skillset, there is a need for officials to stay informed about the latest developments in the field of GST given its dynamic nature.

A competency-based approach to capacity building primarily involves mapping roles, activities, and competencies to positions across the department. This process is followed by the identification of capacity-building interventions that address mapped competencies. To individuals, this approach offers clarity of role and job expectations. The department can also leverage this system for hiring decisions, setting expectations, and identifying training needs. A comprehensive annual training calendar including induction and refresher training will be prepared. In addition, a Knowledge Management Tool is being created. Since GST is a highly technical and dynamic field, officials are expected to stay abreast of the latest developments such as court rulings and amendments in addition to the details of the GST Act and associated processes.
 

4. Bill Bannao, Inaam Pao
 

A statewide lottery programme (Bill Bannao, Inaam Pao) has also been designed through the DDC – CEGIS collaboration to improve last-mile compliance and increase citizen engagement in tax enforcement. The self-enforcement mechanism of any Value Added Tax breaks down at the retail level due to a lack of incentive for the buyer to report their invoices, thus creating space for underreporting turnover. By creating an incentive (through lottery draws and cash prizes) for the consumer to share their invoices, there is a deterrent effect as well as the generation of third-party data.
 

B. Department of Revenue
 

DDC and CEGIS have also engaged with the Revenue Department to provide analytical support to improve tax design and minimise economic distortions. The objective is not only to increase the quantity but the “quality” of revenue, by ensuring that economic activity is not constrained.

One key aspect of this work has been the redesigning of the guidance values for the levy of stamp duty on property transactions. The design for these circle rates for Delhi has not been updated both in terms of slabs and rates for the past 12 years and 8 years respectively. The existing circle rate slab has 8 categories ranging from A to H. Delhi has grown twice in terms of the population since the last revision and the number of categories currently is too few and no longer represents the current property market size and range.

A study report has been submitted to the Revenue Department recommending an intermediate sub-categorisation based on transaction values of properties in neighbourhoods in the past, increasing the current 8 categories to a total of 26. These subcategories would be a precursor to complete re-categorisation from scratch later on for a second phase of reforms which would then mirror the real market values of properties in Delhi.

In addition to the design change, DDC and CEGIS had recommended a temporary rate cut for stamp duty which would help boost sales of properties which had slumped during the pandemic. In lieu of this suggestion, the Revenue Department introduced a 20% rate reduction in circle rates from February 4th, 2021, to June 2022. The intervention was for a total of 17 months which also saw two waves of the pandemic, the delta wave in April 2021 and the Omicron wave in January 2022. Despite the two subsequent waves of the pandemic, the rate cut has helped bump the market back to pre-covid levels both in terms of registrations and revenue. The table below gives a comparison of the registrations and revenue during the intervention to the period prior to the intervention (First covid wave) as well as a normal year (2019).



The temporary rate cut resulted in a ~52% increase in registrations (3,254 per month) and a ~55% increase in revenue (₹104 crores per month) for the 17-month period. Even if compared to a normal year (2019 in this case), the rate cut showed a ~24% increase in registrations (1,853 per month) and a ~17% increase in revenue (₹47 Crores per month).
 

C. Department of Transport
 

The transport sector in Delhi is uniquely impacted by its tax policy. Characterised by largely open borders with surrounding states, static motor vehicle tax rates on static vehicles have lost pace with market changes over the last decade and led to arbitrage opportunities and tax competition with neighbouring states, especially for the registration of higher-end vehicles. This also defeats the purpose of ensuring the higher cost of vehicles with higher fuel consumption and emissions, while adversely impacting state revenues. Further, given the nature of the economy and its centrality as the hub of north India means that out-of-state transport vehicles travel through the city on a regular basis without payment of additional taxes.

In view of these aspects, DDC and CEGIS have proposed that the design of taxes on the purchase of private vehicles be brought on parity with neighbouring states to avoid competition without triggering a race to the bottom. Simultaneously, a Passengers and Goods tax is proposed on out-of-state buses, especially those which do not meet modern emission standards, to recover the social costs associated with such movement.