These amendments appear to be driven by three forces. First, for operational convenience. Second, for plugging loopholes in the GST base. Third, for accelerating devolution of receipts to states, though on a provisional basis. The basic structure of the GST is not being touched. The purpose of this article is to highlight the opportunity to further amend the GST laws to simplify the IGST structure, lighten the compliance burden on dealers, and encourage exports.
The chart details collections of CGST, SGST, IGST and the aggregate GST collections (which includes compensation cess, which is not shown in the graph) over the past eight months.
Two questions arise here. First, why is there a consistent difference between CGST and SGST collections over this eight-month period? Second, why have IGST collections significantly increased and CGST/SGST collections fallen correspondingly over the last three months?
First, it should be noted that CGST and SGST are applied at the same rate on the same base. For example, the GST rate of 18% comprises a CGST of 9% and SGST of 9% levied under their respective acts, but collected in aggregate through one return. When the same rate of tax is levied on the same base by both the Central government and the state governments, the collections from these taxes should be the same. As the chart shows, this is not the case. The collections are different. The difference was about ₹ 13,150 crore in January. While this difference has narrowed to ₹ 6,000 crore in June, the gap remains. One explanation is that it arises from a differential set off of input tax credit on CGST and SGST due to divergent rates of Central excise and state value added tax (VAT) applied on goods which were held in stock as on 1 July 2017, when GST was implemented. Since transitional stocks have different levels of Central excise and state VAT embedded in them, the input tax credits set off for CGST and SGST are different and therefore the collections of CGST and SGST are different. This is a valid explanation, but only in so far as it relates to transitional stock. More than a year has elapsed since the GST was implemented. Most companies turn over their inventory at least twice a year if not more. Using this conservative guideline, all transitional stock should have been sold by December 2017. Once fresh inventory is acquired under the GST framework, there should be no difference between CGST and SGST collections after December 2017. Another explanation is that the difference arises from the carry over to the subsequent year of input tax credit on capital goods purchased, which could be different for central excise and VAT. If so, the difference should steadily decrease from July 2017. While the variance has decreased, a significant difference persists and is unexplained.
Second, IGST has more than doubled over the last three months. How has this striking increase taken place? The e-way Bill was introduced in April. Could this be correlated with this significant rise ? However, there has been an equally significant fall in the CGST and SGST revenue over these three months . Why have CGST and SGST revenue fallen during this period? Is this fall linked to the rise in IGST revenue? Further, the present unreconciled balance in the IGST account is reportedly about ₹ 1.5 trillion. The present unreconciled balanced in the Compensation Fund is about ₹ 38,000 crore with a cumulative cess collection so far of about ₹ 86,000 crore against which about ₹ 48,000 crore has been released to states so far. A state government has accused the Central government of unreasonably crediting the unreconciled balances in these two funds to the Consolidated Fund of India, unwarrantedly bolstering the Union’s fiscal position, to the detriment of the fiscal position of the states. Apparently, in response, one of the proposals under consideration seeks to amend the Compensation Act “to distribute the amount remaining unutilized in the Fund amongst the Centre and the States in the manner provided". This proposal seems to recognize that the compensation being collected is beyond the requirement of the state governments. Two points arise for consideration. First, the priority should be providing relief to the consumer by lowering the rates of compensation cess (which is anyway supposed to be temporary) rather than going against the grain of the GST and enriching the states by redistributing excess balances in the Compensation Fund. Second, the GST Council in its recent meeting reduced the tax rates for a number of items in the 28%, 18%, 12% and 5% slabs. This would involve a loss of revenue, which may result in increased compensation claims over the next few months. For both these reasons, is the proposal to distribute unutilized Compensation Fund balances to the Centre and states appropriate?
The answers can be determined only after goods and services tax network (GSTN) publishes intrastate and interstate-wise details of tax and cess collected and adjusted, sales and purchase data, tax rate and commodity wise. Like census data is released for research purposes, individual firm data could be aggregated to address privacy concerns and released to the public. GSTN carries a wealth of trade transactional data and any policy change in the GST design should be based on analysis of its data. It is not clear that the present amendment proposals have benefitted from such an analysis.
One step that can be taken immediately by the Central government to address the large outstanding and unreconciled balances in the IGST account is to simplify the IGST. GST is the sum of CGST and SGST. In interstate transactions, SGST cannot be levied since states cannot levy taxes across their borders. To prevent arbitrage and promote tax efficiency, the dealer must face the same tax rate for interstate and intrastate transactions. The IGST levied is essentially a “wash" tax, applied as an equivalent to SGST in interstate transactions. Thus, ideally IGST should comprise only the SGST and nothing else. For intrastate transactions, the dealer should pay the sum of CGST and SGST. For interstate transactions, the dealer should pay the sum of CGST and IGST. CGST revenue will remain distinct and allocable to the government of India. SGST revenue will remain distinct and separable and allocable to the individual state governments. IGST will be allocated to the states where the downstream sales have occurred on the basis of adjustments claimed by the relevant dealers.
The present IGST framework is unnecessarily complex, leading to convolutions in reconciliation. The IGST at present is the sum of SGST and CGST, not only SGST as proposed above. This results in the anomalous requirement that the CGST Act which has been legislated by the Union be made applicable only to intrastate transactions and not to interstate transactions. This is a jarring approach, restricting the application of a tax levied by the Union to intrastate transactions, when the Union has the powers to levy tax on all transactions both within and across states. This modus has made the IGST framework lumbering and ponderous while contributing to unreconciled balances in the IGST Fund.
If the IGST were made the equivalent of the SGST only and CGST had been applicable to both interstate and intrastate transactions, then the unreconciled balance in the IGST account today would have been at half its present level, i.e. about ₹ 75,000 crore instead of the present ₹ 1.5trillion. The government could have legitimately recorded a rise in CGST revenue to the extent of half the IGST presently collected (representing the CGST portion of the present IGST). States would have been reassured that the entire balance in the IGST account (excluding B2C transactions) is earmarked for allocation to states based on claims made by dealers in the downstream sales states. A countervailing duty (CVD) on imports to the extent of CGST would need to be applied, but otherwise this arrangement will simplify the levy, collection and allocation of IGST revenue.
A further simplification will be to amend the IGST Act to truly zero rate exports. Internationally, in countries which have adopted the GST, exports are zero rated. This means a zero rate of tax is levied on exports, enabling a refund of input tax credit to the exporters. Under the present IGST law, the dealer is required to pay IGST on his exports (net of input tax credit claims) and then seek refund of the IGST paid by him. Admittedly, a letter of undertaking mechanism is in place in lieu of IGST payment, but many are unable to avail it. The levy of IGST on exports increases an exporter’s compliance burden, enlarges his working capital requirements and expands his turnaround time. This also inflates IGST revenue to the extent of IGST levied on exports (which needs to be refunded) and presents an incorrect picture of not only the IGST balance available for adjustment to states, but also the gross GST revenue. If the IGST Act is amended to truly zero rate exports, these problems will be resolved.
In sum, as part of its process of amending the GST Acts, the GST Council should simplify the structure of the IGST and truly zero rate exports by making appropriate amendments to the IGST and the CGST Acts. It should also consider reducing the compensation cess if the Compensation Fund continues to record surpluses after the recent rate decreases. In addition, the GST Council could consider publishing aggregates of GSTN data commodity and rate-wise as well as interstate and intrastate-wise so that researchers can examine questions related to GST policy.
V. Bhaskar and Vijay Kelkar are, respectively, former special chief secretary (finance), government of Andhra Pradesh, and vice-president of the Pune International Centre.
Comments are welcome at email@example.com