CAG and their emphasis on Invoice matching
The CAG’s report on Goods and Service Tax for the year ended March 2018, (Report number 11 of 2019) is exceptional in many ways. This is the first ever critical review by a top Government agency of the GST tax system. CAG reports generally focus on certain aspects of a department’s working, but this report takes a synoptic view of the entire tax system and provides insights in to it. This is also a rare CAG report in that it begins on a very apologetic tone when they state meekly in the preface that their work should not be seen by the stakeholders as a fault-finding exercise. The report, however, is fairly unsparing in criticising the Department for its various failures.
The report has identified that the GST tax administrative system stands on the premises of ‘one hundred per cent invoice-matching and availment of input tax credit, as well as settlement of IGST on the basis of invoice-matching’. The innovation in VAT or the GST tax system is the method by which a tax payer is allowed to reduce his gross tax liability by the amount of tax he has already paid through his vendors on the inputs used by him in the production process. Invoice matching system means that the invoices of inputs should match with the entries of input credit the supplier has taken. The entries relate to all aspects of assessment as noted in the invoice; the description and code number of the goods, their value, the tax paid besides all reference numbers. If the entries in the input invoice match with particulars in the input users ledgers, the system would treat the invoices as validated and further assessment scrutiny would not be needed.
The concern with system based invoice matching as a method of verification in the design of GST system is understandable because for the first time input credit in GST travels across Tax Jurisdictions of different Tax Administrations as a registrant in one State takes credit of tax paid in another state. The invoice matching was planned as a fulcrum on which the GSTN system will turn but even after two and half years of its birth, it is nowhere in sight.
Returns and reporting system created assuming the operation of invoice matching system
The basic electronic reporting system in GST named GSTR1, GSTR 2 and GSTR 3 were created on the assumption that the network will enable cent percent invoice matching (invoice verification). The CAG report explains the reporting system succinctly as below:-
“Para 1.6.2 (d) … It was originally envisaged that suppliers would file invoice-wise details of outward supplies made by them during the month through GSTR-1. The details of outward supplies so furnished by the supplier in GSTR-1 were be made available electronically to the registered recipients through form GSTR-2A. Similarly, details of supplies relating to composition taxpayers, Input Service Distributors and Non-Resident taxpayers as well as Tax Deducted at Source (TDS) by Government departments / agencies and E-commerce operators also were to be made available electronically to the recipients. Thereafter, based on details available in form GSTR-2A, the taxpayer was supposed to furnish form GSTR-2 after including details of other inward supplies.
GSTR-3 is a monthly return with the details of sales and purchases during the month along with the amount of GST liability. Most of GSTR-3 was supposed to be auto-generated from GSTR-1 and GSTR-2 while the taxpayer had to include the details of discharge of liability of tax, interest, penalty, refund claimed from electronic cash ledger and debit entries in electronic cash/credit ledger while filing GSTR-3.”
The system as visualised depends on a high level of automaticity and looks at a situation in future where tax officers may not need to perform routine assessment work. Invoices of a supplier captured in the system would generate the report of supply which can be used by the recipient to claim credit of tax. The third report is made by integrating the reports of supplies received and supplies made. The reliability of this system depends on whether the invoices can be validated/ matched. This system never took off; the GSTR 2 and GSTR 3 have been kept in abeyance ever since. Other reports like GSTR 4,5,6 etc are based on same principles but are meant for different class of tax payers.
As the GSTR 2 2A and 3 have come a cropper, the GST Council in July 2018 decided to roll out revised return mechanism by 1 January 2019. The new date for launch of the new reports is October 2019. It is not yet known what kind of reporting format the GSTN is going to introduce in October and whether invoice matching will be a part of it, but from what have been the public statements of the Government, the new reports are meant to ease operations for tax payers. In the context of whatever has been happening, it is fair to assume that invoice matching is not on the cards. The CAG Report, nonetheless, roots for the invoice matching system and they state, in the preface of the report, as follows:-
“This report identifies invoice-matching as the critical requirement that would yield the full benefits of this major tax reform. It would protect the tax revenues of both the Centre and the States, it would lead to proper settlement of IGST and would minimise, if not eliminate, the tax official assessee interface. In fact, even “assessment” in the sense understood in the manual system may no longer be necessary (returns themselves can be generated by a system that matches invoices); and cases of evasion etc., can be traced by applying analytical tools and AI to the massive data that crores of invoices generate.”
In absence of an effective system of control, the menace of fake invoices has reached an epidemic proportion. The Finance Minister Nirmala Sitharaman in a written reply in the Lok Sabha stated that GST officers have booked 535 cases of fake invoices involving a total fraudulent claim of ₹2,565 crore of input tax credit (ITC) and arrested 40 persons so far in the current financial year 2019-20. In 2018-19, 1,620 cases of fake invoices were registered involving fraudulent ITC claim of ₹11,251 crore under the GST. As many as 154 persons were arrested. If these are the figures of detection then the actual extent of evasion through fake invoices must be an astounding amount. One may wonder if the system would not collapse if corrective systems are not developed fast enough.
Quite apart from the revenue leakage, the system of settlement of IGST has collapsed. States are not able to get transfer of amounts, lying in the Consolidated Fund of India, which have been used to pay SGST by recipients of the goods or services. The data in GSTN system is not dependable to enable easy IGST settlement. The CAG report mentions that a significant amount of Rs 8.19 lakh crore of ITC of IGST was claimed by the taxpayers in their returns during 1 July 2017 to 8 August 2018 against total CGST, SGST and IGST of Rs` 11.93 lakh crore collected during 1 July 2017 to 31 July 2018. Out of Rs. 8.19 lakh crore as stated above, taxpayers of Andhra Pradesh alone claimed IGST-ITC for Rs. 6.49 lakh crore which was considered as highly unlikely. This was brought to the notice of GSTN by audit on 21 August 2018. GSTN replied that the taxpayer had reversed the credit as it was wrongly filed by him in GST 3B.The report puts this as an example of how an unrealistic and erroneous claim of ITC of IGST by one taxpayer, representing 79 per cent of total ITC claim by all taxpayers for a month, was allowed by the system, exposing the vulnerability of the system to fraudulent ITC claims.
Is invoice matching system possible and can it be an effective anti-dote to revenue drain?
Making a software is highly technical, but given the extreme difficulty that was faced by trade and government officials in working the GSTN, the question of technical viability of invoice matching system cannot be pushed away from public debate. Globally the bodies which are in the forefront of IT development are e-commerce companies of USA like Google, Amazon and Salesforce (leaving aside scientific establishments which are clearly outside the scope of this discussion). Government agencies have not reached their levels of sophistication or expenditure in development anywhere. Secondly, the sheer volume of the operation of uploading data on a common portal by millions of GST registrants needs finance for IT infrastructure which will be unprecedented. Added to it is the problem of data quality in reconciliation and auto generation of reports. In a database made of data fed by millions of assessees, to achieve a general data quality of a level that enables the government system to use private party data for running a fully automatic system, is an extremely challenging task at the least. Thirdly getting such disparate stake holders as revenue officials, professional accountants, tax payers and general administrators to come together in the making of a powerful glitch-free IT system needs leadership qualities, which may not be easy to find. Fourthly, fraudsters, who are immediate gainers by absence of an invoice matching system, are not insignificant people who live in shanties; the real people behind the frauds remain faceless and enjoy clout in the society. The difficulties in making an invoice matched GST system is Herculean.
The larger question is whether the system is necessary and whether it will answer the issues of evasion which the Tax Administrators face. The system is non existent but the supposition behind making a system independent of administrators appears unreal. The invoice system is perhaps is not possible in near future.
Design of a reasonable GST system
The Government began the journey of GST with the belief that invoice matching was foundation of the system that would remove tax payer – tax officer interface. This belief is founded on an unrealistic supposition that tax administrators do not play a central role in tax governance and that their function can be replaced by computing machines programmed to control errant behaviour of non-compliant tax payers. The principle on which GST system has been founded flies against reality and the past experience of the working the GST administration proves this. It is officers who have unearthed huge revenue losses through forging of fake invoices and not the GST system. The design of a tax system must recognise that the administrative needs of an efficient tax officer converges with the interest of an honest tax payer and the system must provide officers with the software tools to identify and quickly stop errant behaviour.
The start of GST system itself appears to have been false. They set out to reach the sky and fell down with the ladder. The problem is perhaps in the perception. Man often mistakes the trees for the woods but equally he may mistake the woods for the right tree. Tax is money paid by a person to the state and the best method for checking tax receipts is to cross verify the particulars of payments mentioned in documents, like the invoice or e-way bill, with the bank data base where tax deposits are collected. Tax officers must be able to run a programme which reconciles entries of payment of taxes with invoices uploaded in the system and this alone will help them identify fake invoices or invoices in which for any reason the tax paid on input credit is not the same as tax credit taken on the invoice. A simple system for cross checking tax payment will help root out fake invoices which may or may not have been filed in monthly returns. The system tool must also be made available to registered tax payers or else how would they get to know whether their supplier has deposited the tax amount on the supply with the treasury. A tax registrant faces serious consequences if he takes input credit and it is later found that the tax was not paid on the input invoice. The design for an effective control is far simpler than the complex invoice matching system with which GSTN has been experimenting, and the alternate design would help build a strong relation between tax officers and the tax system.