Finmin does well to scrutinise GST records to catch input tax fraud; use other databases like MCA, income tax, etc.
On the face of things, the department of revenue hasn’t had too much success with its attempt to catch GST evaders; while total collections are around Rs 95,000 crore a month, till November last year, a total recovery of just Rs 1,047 crore was made from people who were floating bogus firms and putting out fake invoices to claim input tax credits. In fact, many of the chief ministers who were complaining about the low GST collections to finance minister Nirmala Sitharaman raised the issue of fake input tax credits, and the need for the Centre to do more about catching the thieves.
It now appears the government has started using data analytics to identify cases of fraud and, according to a press release it put out, a total of 931 cases of fake claims were identified in this manner; all 27,000 claims filed by taxpayers on account of the inverted duty structure—the refunds add up to Rs 28,000 crore—are now to be analysed using analytics. In the case of a Uttarakhand firm, investigators found that fraudsters had created over Rs 600 crore of fake credit. Similarly, Surat investigations showed two export firms using false invoices to avail GST refunds worth RsRs 679 crore via 19 different entities were registered in the names of daily wagers and casual workers. Furthermore, an exporter with ‘star’ status and business worth Rs 50 crore of exports of readymade garments had taken a refund of Rs 3.9 crore while the total GST payment in cash was merely Rs 1,650. The department is already sending notices by matching invoices with TDS returns. Making it mandatory for all current accounts to be linked to a GST number is also an important step in catching evasion.
It is not clear what databases were used to catch the fraudsters, but, logically speaking, the taxman can look at income tax returns—after all, if the firm is claiming a certain credit, it has to have a commensurate turnover as well, and to have this turnover, it needs to be hiring a certain number of people, consuming a certain minimum amount of electricity, paying rent, etc; in short, there are several databases, like the Ministry of Corporate Affairs (MCA) and even the EPFO/ESI etc, and, now that a PAN-Aadhaar linkage has become compulsory in many of them, catching fraud is much easier. Indeed, direct tax department has an Operation Insight, which combines various databases like credit card payments, foreign visits, etc; ideally, the direct and indirect tax departments should share information and access to databases.
While using data analytics—and perhaps AI and machine learning, over time—to catch fraud is a good idea, the GST Council also needs to build in enough checks to keep theft to the minimum. Less tax slabs, for instance, would reduce theft, as would keeping the rates as close to one another as possible; the issue of inverted duty structures, too, will get resolved to a large extent. It would also fix many complications. Right now, for instance, a school filing GST returns does not pay anything on fees, but has to pay 5% on uniform and teaching aids, and 28% for instruments, apparatus and models; apart from this being a nightmare for those filing returns, it opens up the possibility of fraud by misclassification as well. Using blockchain technology, over time, is also a good idea since this involves verification by several people in a value chain for each transaction. Nor is there any reason why refunds should be allowed to be claimed in cash since the trail disappears in many cases. Allowing so many exemptions and, for instance, levying a lower GST without input tax credit is, it is important to keep in mind, simply an invitation to tax fraud.