Where is the monkey now?
NOVEMBER 11, 2020
By Vijay Kumar
IN the 25th GST Council Meeting held on 18 January 2018, Mr. Nandan Nilekani told the Council, "the biggest risk of having a mechanism in which the system would do the matching was like taking the monkey on one's back." He was talking about matching of invoices for GST. He also told the Council that:
The core principle of any indirect tax model should be that input tax credit would be provided only on "matched" invoices i.e. legitimate invoices where the supplier had admitted tax liability by uploading the invoices on the portal. This would mean either denial or automatic reversal of credit on unmatched invoices.
- This principle was even more important in GST regime because settlement of IGST became a lot more complex and harder to audit where transactions would have to be settled possibly at invoice level.
- Without matching of invoices, benefit of other related initiatives like e-Way bill system would be diluted.
- Those models were doomed to fail which increased the burden on the taxpayer to correct mismatches or which relied on tax official's intervention to reduce the mismatches.
- Any solution that permitted, in the first place, higher level of mismatch would also fail as it would not permit automatic reversal.
- It was not desirable to entrust the responsibility of invoice matching to the Government.
- A high rate of mismatch of 30% to 40% would provide sufficient cover to fraudsters to easily split the fraudulent claims knowing fully well that detection would be hard.
- It was important to remember that every business - large or small, automated, or manual - routinely compared supplier's invoices with the purchase books and that it was a necessary step before releasing payment.
- Most taxpayers had very few invoices and 93% of the taxpayers have less than 50 sales invoices that needed to be uploaded and 91% of the taxpayers have less than 50 purchase invoices that needed to be accepted.
- GSTR-2 created a separate process and matching and return filing were duplicated, which made it difficult for the taxpayers.
- In order to remove the burden of GSTR-2, it was desirable to align this process with the natural cycle of verification and payment. Comparing supplier's invoices with purchase books all over again for tax claim purpose was a burden.
- By modelling "invoice upload" and "acceptance" as tax "returns" (GSTR-1 and GSTR-2), the model created a perception that there were three returns per month. The structure of forms was so complex that it required a tax professional's help. The concepts like tax on advance payments, its utilisation to offset liability, separate reporting of different types of invoices made GSTR-1 and GSTR-2 more like a tax return form than a statement. Reporting of invoices at rate-level instead of line-item level created more work for the supplier.
- A successful model would be one which aligns with the natural business cycle of verification and payment of supplier invoices.
- The invoices should also be made available to the buyer for acceptance. The key difference from GSTR-1 would be that it would simply mean invoice "upload" and not "filing" of tax return.
- Invoice format and data granularity should match with the actual invoice submitted by supplier for payment namely, invoice item level right from day one and that it should not be rolled up at tax rate or commodity level. Upload should happen on a continuous basis, which would imply that verification and acceptance coincided with the actual business transaction. Invoices uploaded after the 10th of the month would automatically be included in the next return.
- Market forces would evolve a model where invoice would be paid for only after upload on the GST system. Buyer should accept supplier's invoices on the GST system, which would automatically determine the input tax credit.
- The key contrasts from GSTR-2 and pure system matching model was that it was simply an invoice "acceptance" and not "filing" of return and that acceptance could happen on continuous basis, not waiting for all GSTR-1 to be filed. Invoices, once accepted, would be locked and could not be modified by the supplier, thus bringing finality to the transaction. The system should provide robust tools to facilitate smooth acceptance including for offline matching of supplier invoices with purchase books, auto-acceptance capabilities and improved support to GSPs/ASPs for tighter integration with accounting packages.
- Eliminate the concept of "Provisional Credit". However, buyers could "notify" supplier through the system to upload any missed invoice but could not upload or modify it themselves. In this model, there would be no mismatch in the traditional sense and hence there would be no question of any reversal.
- Reversal of input tax credit due to non-payment of tax by the supplier was widely perceived to be unfair to the buyer and the criteria for legitimate invoice should be redefined as one where supplier has admitted Liability by uploading onto the portal and make provisions to recover dues from the supplier rather than penalising the buyer. The GST system would offer multiple channels for uploading and acceptance of invoices and filing of returns as 91% of taxpayers had fewer than 50 invoices in a month i.e. hardly 2-3 invoices per day; small taxpayers with no automated accounting systems could view and accept pending invoices directly on the portal. Small/medium taxpayers with some level of automation could use excel based offline tool to download, compare and accept pending invoices. Large taxpayers with fully automated accounting systems would do reconciliation and acceptance directly in their systems and upload results directly through APIs.
- The gradual transition would involve, in transition phase 'A', to continue with self-declaration of GSTR-3B for payment of taxes and replacement of GSTR-1 with invoice uploading. In transition phase 'B', to continue with self-declared GSTR-3B for payment of taxes; to enable invoice acceptance feature, which accepts input tax credit; introduce system generated GSTR-3 as a read only declaration; have a GSTR-3B versus GSTR-3 comparison report. Under this, the input tax credit compared would include missed invoices. At the end stage, GSTR-3B would be discontinued and would be replaced by GSTR-3; to continue with invoice uploading/acceptance features and enable filing of system generated GSTR-3 as a return including payment capability.
- This system would work well as it was incentive aligned where, if the supplier did not report invoices on time, he would not get paid and the buyer who did not accept invoices in time, would not get input tax credit.
What did the Council do with all these suggestions? It was decided that the model proposed by Shri Nandan Nilekani shall be examined by the Group of Ministers on IT Challenges in GST Implementation, headed by the Hon'ble Deputy Chief Minister of Bihar, in consultation with the members of the Committee on Return Filing and Shri Nandan Nilekani. The issue could then be decided by the Council.
In the 26th GST Council meeting on 10 March, 2018, the Chairperson enquired from the Commissioner (Central Excise), CBEC, as to why revenue administration was apprehensive of the model proposed by Shri Nandan Nilekani. The Commissioner (Central Excise), CBEC, responded that there were several reasons for apprehension.
1. The first and the main reason of apprehension was that there would be no control regarding default in payment of tax on which input tax credit was taken.
2. Secondly, the proposed concept of blockage of uploading of invoices by supplier was an unchartered legal territory. The Central Excise law tried to introduce this provision and the courts quashed it.
3. Thirdly, blocking of invoice upload by a supplier could hurt multiple buyers. If a supplier made supplies to a buyer A in the month of April and defaulted in paying tax for his May return, and made supplies to a buyer B in the month of May, the input tax credit to B would also be blocked due to default in tax payment in respect of supplies made to buyer A.
4. The fourth issue related to limitation in the number of tax officers who could take up audit and scrutiny. He stated that in one year, the Central tax administration was able to do audit of about 40,000 units, return scrutiny of about 30,000 units and anti-evasion cases of about 10,000 units. He stated that the maximum intervention possible by the Central administration would be about 1,00,000 cases. States could possibly make about 2,00,000 interventions in a year as they have similar number of assessing officers but double the number of support staff. This implied that the Central and the State administrations put together could intervene in only about 3,00,000 cases in a year against a taxpayer base of close to one crore. It was in this context that a mechanism of self-policing was very important.
5. The fifth issue was in respect of IGST settlement. If there was a default, then the Centre would lose the entire amount already transferred to the State. He suggested that instead of the Centre losing the entire amount, 50% of such losses should also be borne by the States.
6. The sixth issue was that the system of locking of invoices, as proposed in the model of Shri Nandan Nilekani, could involve compliance load.
The Chairperson stated that Shri Nandan Nilekani also had a meeting with the members of the GoM as well as the officers of the Centre and the States. He observed that broadly, the political executive found the model of Shri Nandan Nilekani to be simpler but the tax bureaucracy considered it to be a risk to revenue.
In the meanwhile, several monkeys have got on our backs and are still there.
Midnight Summons:
The recent judgement of the Telangana High Court - 2020-TIOL-1898-HC-TELANGANA-GST brought to light some very disturbing trends, as can be seen from the following.
At 00:00 hrs on 12.12.2019, the 4th respondent (An SIO in DGGI) issued summons dt.12.12.2019 summoning the 2nd petitioner to appear in person before him at 00:30 hrs. in the office of the 1st petitioner.
The DGGI told the Court that he was available at the spot i.e. Corporate Office of petitioner No.1 and so he was served summons in his office after midnight, in the early hours of 12.12.2019. And there is no bar in making enquiries under Section 70 of GST Act, 2017 in the night..
They also told the Court that the officers posted at the Headquarters of the DGGI at New Delhi have all India jurisdiction while the officers posted at DGGI, Hyderabad Zonal Unit have only local jurisdiction. [As per Notification No. 14/2017 - Central Tax, dated 1st July, 2017, the officers of DGGI from Intelligence Officer to Principal Director General have jurisdiction throughout India and there is no specific All India Jurisdiction for the officers posted at Headquarters.]
The High Court observed,
Though the respondents seek to suggest that such evidence procured by the petitioners ought to be disbelieved by us because Sunshine Hospital is a 'private hospital' and not a Government Hospital, we do not agree with such contention because there is no presumption in law that Doctors in private hospitals do not speak the truth and only Government doctors speak the truth. An injured person is likely to go the nearest available hospital for treatment instead of searching for a Government hospital at that juncture.
The Summons bears a date 12.12.2019 and asks the 2nd petitioner to appear before 4th respondent at 00:30 hrs on 12.12.2019. This prima-facie indicates that it was issued after midnight on the intervening night of 11.12.2019 and 12.12.2019 asking the 2nd petitioner to appear at the ungodly hour of 00:30 hrs on that day.
In view of the admitted fact that the search operations were continued well past midnight and summons were issued to 2nd petitioner to appear at 00:30 hrs on 12.12.2019, we do not accept the plea of the respondents that they did not act contrary to established procedure, that the search proceedings were carried out under proper and applicable law and procedure, and no harm or damage were made to any human/person or property and no sentiments were hurt.
The High Court issued directions:
a. the respondents shall not use any acts of violence or torture against petitioner nos.2 to 4 or their employees in furtherance of enquiry proceedings initiated against the 1st petitioner;
b. the enquiry in the above proceedings against the 1st petitioner shall not be handled by the 5th respondent (a deputy director in the DGGI) and he shall not participate in such enquiry, and it shall be transferred to another official to be designated by the 2nd respondent;
c. any interrogation of petitioner nos.2 to 4 or their employees shall be between 10:30 a.m. and 05:00 p.m. on week days in the visible range of an Advocate appointed by them, who shall not be in hearing range;
d. the respondents shall adhere to the provisions of the CGST Act, 2017 in conducting search, investigation or enquiry in relation to the alleged tax evasion by the petitioners.
Until next week
