The Income Tax Department's Time Series Data

Income Tax India Official recently released Time Series data of direct tax collections for the period 2000 - 2022. This was a positive move since the last available time series data was till F.Y 2018-19. What is more helpful is the indirect taxes data that was included in this series, saving time for data analysis of total tax collections of Government of India. This is an attempt at making simple inferences from the freshly available data.

Comparision of Direct and Indirect Taxes Growth

1. From FY 2000 to FY 2006 - the first six years of the century and probably even before that, Indirect Taxes were the bigger contributors to total tax revenue. Direct Taxes were just about one-third of the total. The situation flipped in FY 2007 and has remained that way since. In FY 2009, direct tax collections reached a peak, as a % of total tax revenue, contributing to 60% of the total collections.

What is remarkable is the growth in direct taxes during the five year period between FY 2015 and FY 2019. This was a period of relentless reforms which included #Demonetization, Income Disclosure Scheme, the revised #Benami Act, Operation Clean Money and the crackdown against #shell companies.

The impact of the pandemic is stark and unmistakable. What is more stark is the recovery which coincided with the second spate of reforms which included the e-Verification Scheme, CPC - 2.0 and substantial new investments in the Department's infrastructure.

2. A comparison of growth rates of Direct and Indirect Taxes does not show a regular pattern or any correlation. However, two shock events can be compared from a tax perspective,

Shock event 1 - 2008 financial crisis. Shock event 2 - CoVID Pandemic.

The 2008 event impacted both direct and indirect tax collections severely. Look at the drop in collections in 2008-09. Even though India was always touted to have been reasonably insulated from the global economy during that period, tax collections took a clear hit.

Compare this to the impact of the pandemic on collections. In comparison to Direct Tax Collections which saw a drastic drop, Indirect Tax Collections looked secure. This is a crude reflection of the relatively inelastic nature of consumption taxes - Pandemic or no pandemic, rice, wheat, eggs and pulses will be consumed and GST collections will come in. However, the same was true in 2008 as well. Why didn't we see a drastic fall in indirect tax collections in 2020? The only answer I can think of is the government's Direct Benefit Transfer which touched 6.3 Trillion Rupees in FY 2020-21, the worst hit year. These transfers ensured consumption and consequently, GST Collections. The miracle of JAM( JanDhan account, Aadhaar and Mobile) which made this possible is stuff that deserves its own article.

Another FY that stands out is 2015-16. Growth rate of Direct and Indirect Tax Collections moved in opposite directions. While nothing spectacularly wrong took place in the Indirect Tax ecosystem, the Direct Tax administration unfurled several reforms during the period which included the Income Disclosure Scheme - a reform with immediate impact on overall collections.

The Tax - GDP Ratio

The Tax-GDP Ratio of India as recorded in the OECD - OCDE 's Annual Tax Administration Survey requires to be revised upward. As per the OECD, for the calculation of Tax - GDP Ratios, all kinds of taxes and levies (including Social Security Contributions) which are mandatorily collected by the Government of that country are to be included. While the OECD-TAS 2023 shows Tax-GDP Ratio of 8%, the figure used by the United Nations is much higher, at 18.3%. An attempt was made to arrive closer to the UN figure for FY 2022-23.

While the time series data provides direct and indirect tax collections, the following payments need to be included to arrive at the real Tax-GDP Ratio (EPF and ESIC figures collected from individual Annual Reports, other figures approximated using data available for individual Metros)

a. Total receipts of the EPFO under its three flagship schemes namely the Provident Fund Scheme, the Pension Scheme and the Deposit Linked Insurance Scheme - Amounting to approximately Rs. 2,80,000 Cr.

b. Total receipts of the ESIC - Approx. Rs. 15,300 Cr.

c. Professional Tax, Stamp duty, Motor Vehicles Tax and Property Tax - Approx. Rs. 2,70,000 Cr.

These amounts would lead to a significant uptick of approximately 2.5%. This is shown in the total Tax GDP Ratio figure for FY 2022-23 in the graph above. That would take the Tax GDP Ratio to 13.5%, pushing India above the United States in the OECD ranking list. However, this still appears to be on the lower side compared to the UN figure of 18.5%. The UNU-WIDER - United Nations University World Institute for Development Economics Research comments document does not provide much detail on the break-up of the figure and this needs further analysis.

3. State - wise Growth

The time series data provided state-wise direct tax collections during the period FY 2017-18 to FY 2022-23. This data was quite interesting. First, quite naturally, it was interesting to see the states that saw the highest growth in direct tax contributions over these 6 years,

While the names on the left side of the graph did dampen my spirit, it was those on the right that were true revelations. 3 of the top 5 states are from the North-East. It was heartening to see the investments in the North East over the last decade bear fruit. Meghalaya doubled it's contribution to the country's direct tax purse from Rs. 800 Crores in 2017 to Rs. 1600 Crores in 2022. Telangana is an outlier from becoming a new state in FY 2013-14 to crossing 400% growth over the last 6 years. Karnataka and Gujarat are states that continue to shine with high growth rates notwithstanding their significant bases. Andhra Pradesh and Kerala lag behind their geographical cousins, Karnataka, Telangana and Tamil Nadu. However, where Kerala and Andhra Pradesh lose out on contributions to the exchequer, they make up through remittances to the country's forex reserves.

It is when we look at percentage contribution of each state to the nation's direct tax kitty, the big names comes to light. An undeniably skewed pattern comes out where 75% of the total collection comes from the top 6 states, 3 from the south, 2 from the west and the Capital. How to make sense of this? Now if you look at the break-up of total tax revenue, it is almost 50% a piece between corporates and individuals. So about Rs. 8 Lakh Crores comes from just 10 lakh companies while the remaining 8 Lakh Crores comes from about 7 Crore individuals. So the tax-density if you may call it, of companies is very high and this is reflected in the states where companies are registered.

So, wouldn't it be logical to correlate number of companies registered in each state with tax collections?

What we get is a highly correlated (at least visually) set of variables. You can see how direct tax collection is impacted by the number of companies registered in the state. A pan-India company with headquarters in Kerala might be registered for Income Tax purposes in Mumbai purely for operational purposes. Take banks for example, several banks which were founded in Karnataka, moved their registered addresses to Mumbai in the recent past. The business ecosystem of Mumbai is attractive for large businesses. Manpower, funding, networks are all accessible.

Now, while both the variables are correlated, it can be seen that the average direct tax paid per company varies widely from state to state. From the data, it appears, an average company in UP, West Bengal, Telangana and Bihar pays less tax than a company in Maharashtra, Andhra Pradesh or Karnataka. Does this imply a state's preference or non-preference towards tax planning? This can be confirmed only if we had a state-wise break-up of Corporate Income Tax collected from ITR-6 and ITR-7. Since the data is a sum of corporate and non-corporate taxes, meaningful inferences would be difficult to draw from this data.

But what about the remaining 8 lakh Crores? I presume, there will be natural correlation between the number of individual ITRs filed in each state and the tax collections of that state.

Visibly, the correlation is clearly less stronger than the one with company registrations.

The strength of these visible correlations was tested through the Ordinary Least Squares Method using Python.

The number of companies and the number of ITRs filed in each state have a statistically significant correlation to the total direct taxes collected in the state - indicated by a positive coefficient greater than 1 and a p-value < 0.05.

Overall, the data is a gold-mine for those interested in numbers and taxes to dive into. There are three more tables, Cost of Collection, Pre and Post Assessment Collections and Number of persons filing ITRs. These are yet to be analyzed.

Do comment if you have something to add about the inferences above.

Image Credit - #Copilot