## Monday, January 16, 2017

### What does the tax data tell us about the state of the economy?

by Suyash Rai.

One of the most interesting questions in Indian macroeconomics today is: how are we faring since late 2016? In this article, I seek to analyse data on tax revenues and obtain some clues about the performance of the economy.

In a press release published on January 9, the central government reported the following increases in tax collections during April-December 2016, compared to the corresponding period of previous year:

1. Central excise duty: 43 percent.
2. Service tax: 23.9 percent.
3. Customs duty: 4.1 percent.
4. Corporation tax: 4.4 percent.
5. Income tax: 24.6 percent.

These are large values. Holding other things constant, they suggest buoyant economic activity. However, when looking at tax data, we have to look at the extent to which other things are indeed constant. When analysing tax data, in order to read the state of the macroeconomy, we need to adjust for the part of the tax revenues which are on account of 'Additional Revenue Mobilisation' (ARM). Two kinds of ARM are:

1. An increase in tax rate: additional revenues due to higher rate do not indicate robustness of the underlying activity.
2. An administrative measure: additional revenues from one-time administrative measures (eg. a tax amnesty scheme) may not reflect the underlying economic activity.

Let us walk through the major taxes, and see what we can tell, and what we do not know.

### Excise duty

The biggest increase in tax collection has come from excise duty. The collection during April-December 2016 was 43 percent higher than the corresponding period in 2015. Collection grew by 45 percent in April-October, 33.7 percent in November, and 34.8 percent in December, compared to the corresponding periods of previous year.

Excise duty rates during the reference period

Were the rates constant? Between April-December 2015 and April-December 2016, there have been certain changes in excise duties. Basic excise duties on these products were increased in five steps between November 6, 2015 and January 30, 2016. The detailed notifications can be found on the CBEC website. The cumulative impact of these increases is:

• Unbranded petrol: increased from Rs.5.46 to Rs.9.48 per litre
• Branded petrol: increased from Rs.6.64 to Rs.10.66 per litre
• High speed diesel: increased from Rs.4.26 to Rs.11.33 per litre
• Other diesel: increased from Rs.6.62 to Rs.13.69 per litre

Taxation of petroleum products is unfortunately the backbone of India's excise duty collection; we have failed to build a more broad-based indirect tax system. In 2015-16, the total excise duty (including cesses) collected was Rs.2,84,142 crore. From this, about Rs.1,94,061 crore, or 68.3 percent, was from crude oil and petroleum products (including cess on crude oil).

The increases in the specific rates for the four products would thus have an important impact upon the overall excise duty collections. Consider the excise duty on unbranded petrol. Since five rounds of rate increases happened between November 2015 and January 2016, when we compare collections in April-December 2015 versus April-December 2016, we are comparing periods with different applicable rates. During April-December 2015, the basic excise duty per litre on unbranded petrol was Rs.5.46 between April 1 and November 6, Rs.7.06 from November 7 to December 16, and Rs.7.36 from December 17 to December 31. However, throughout Apri-December 2016, the basic duty per litre of unbranded petrol was Rs.9.48. So, while comparing collections during these two periods, we need to deduct the additional revenue due to the higher rates.

Increase in collection during April-October

A press release on December 9, 2016 reported the indirect tax (excise duty, service tax, and customs) collection with and without ARM. During April-October 2016, the growth in indirect tax collection with ARM was 26.6 percent, while without ARM it was 8 percent. So, the total ARM was 18.6 percent of indirect tax collection during the same period of previous year. This is Rs.71,315 crore.

To the best of my knowledge, there has been no ARM in customs. ARM in service tax collections can be estimated by comparing applicable rates during the two periods being considered. During April-October 2015, the applicable rate for April and May was 12.36 percent, and it was 14 percent for June-October. So, the average rate during the period was 13.53 percent. During April-October 2016, the rate was 14.5 percent for April and May (Swachch Bharat Cess of 0.5 percent was introduced in November 2015), and 15 percent for June-October (Krishi Kalyan Cess of 0.5 percent from June 1, 2016). So, the average rate was 14.86 percent. The estimated increase in service tax collection due to ARM was (14.86-13.53)/13.53 percent = 9.8 percent of collection during corresponding period of previous year. This is Rs.11,059 crore. So, the remaining ARM, i.e. Rs.60,256 crore is estimated to have come from excise duty increases. Since the total increase in excise duty collection during April-October 2016 was Rs.66,485 crore, the increase without ARM was Rs.6229 crore, or 4.22 percent higher than excise duty collection in April-October 2015 (Rs.1,47,670 crore).

Increase in collection during November and December

During November 2016, the growth in indirect tax collection with ARM was 23.09 percent, while without ARM it was 8 percent. This suggests that, in November 2016, indirect tax collection from ARM was 15.09 percent of indirect tax collection in November 2015. This is Rs.8259 crore. Customs collections had no ARM. The effective service tax rate during November 2015 can be assumed at 14.25 percent, as half of the month had 14 percent rate (effective from June 1 to November 15, 2015) while the other half had 14.5 percent rate (0.5 percent Swachch Bharat Cess introduced on November 15, 2015). In November 2016, the rate was 15 percent (including Krishi Kalyan Cess of 0.5 percent). So, the ARM in this November's service tax collection is estimated to be (15-14.25)/14.25 percent = 5.26 percent of service tax collection in November 2015 (Rs.14,870 crore). This is about Rs.782 crore. Deducting this, we get Rs.7477 crore of ARM for excise duty in November. The total reported increase in excise duty collection in November 2016 was Rs.7477 crore. So, the increase in excise duty collection in November without ARM is estimated to have been zero. This suggests a deceleration from 4.22 percent growth in the preceding months..

Since these are nominal values, in real terms, collection may have declined in November. Excise duty becomes due at the time goods leave the factory. One did not expect any significant demonetisation impact on excise duty collection for November, as production and factory clearance schedules for the month were not likely to have been significantly affected by a decision taken in second week of the month. We saw this in the auto sector data, where November data showed a decline in sales but an increase in production, while the December data shows a decline in both sales and production. The latest IIP numbers, which show an increase in industrial production in November, also confirm this. So, the estimate of excise duty collection without ARM in November is surprising. Further, since retailers of petroleum products were allowed to accept old notes, the impact on these products should have been lower. So, to the extent increase in excise duty collection is an indicator of underlying economic activity, the news from November may be worse than expected.

It is difficult to make a reasonable estimate of excise duty collection without ARM in December, because, unlike the press release on December 9, the press release on January 9 does not include details about ARM, not even in the aggregate. When government releases December's data about excise duty collected without ARM, we can do this analysis for December.

Increase in excise duty collection (in percent)
April-Oct Nov Dec
With ARM
45 33.7 34.8
Without ARM (estimate)
4.22 0
NA

### Service tax

During April-December, 2016, service tax collections were up 23.9 percent compared to the corresponding period last year. During April-October, the increase was 27 percent, while it decelerated to 15.5 percent for November and 3.67 percent for December. This deceleration is significant, but we need to understand the increase without ARM.

Service tax rates during the reference period

Service tax rate has been increased thrice during the reference period

1. June 1, 2015: rate increased from 12.36 percent to 14 percent
2. November 15, 2015: Swachh Bharat Cess of 0.5 percent took the rate to 14.5 percent.
3. June 1, 2016: Krishi Kalyan Cess of 0.5 percent took the rate to 15 percent

Increase in collection during April-October

In the section on excise duty, I estimated that ARM for service tax during April-October was 9.8 percent of service tax collection during the corresponding period of previous year. The overall increase in service tax collection in April-October was 26.9 percent. So, the increase without ARM would have been about 17.1 percent.

Increase in collection during November and December

In November 2016, collection increased by 15.52 percent compared to November 2015 - from Rs.14,870 crore Rs.17,178 crore. In the previous section, I estimated that about Rs.782 crore of service tax collection in November 2016 may have been on account of ARM. After deducting this ARM from collection in November 2016, the increase without ARM was 10 percent.

The increase in service tax collection in December 2016 (Rs. 22,449 crore) was 3.67 percent higher than that in December 2015 (Rs.21,655 crore). Since, the service tax rate applicable in December 2015 was 14.5 percent, while that in December 2016 was 15 percent, ARM is estimated to have led to 0.5/14.5 percent = 3.45 percent increase in service tax collection in December. So, in this estimate, the increase in service tax collection without ARM was 0.22 percent. If this analysis is correct, it shows a significant deceleration in rate of increase in service tax collection without ARM: from 17.1 percent in April-October, to 10 percent in November, to 0.22 percent in December.

Increase in service tax collection (in percent)
April-Oct Nov Dec
With ARM
26.9 15.52 3.67
Without ARM (estimate)
17.1 10 0.22

### Customs duty

During April-December 2016, customs duty collection has increased by 4.1 percent, compared to the same period in 2015. During the corresponding period in previous year, the growth in collection was 17 percent. For November 2016, collection increased by about 16 percent. Partially, this may have been because this year Diwali was in October, while last year it was in November. Customs collections are mostly done on working days. November 2015 had fewer working days than November 2016. This effect might explain a part of the increase. Still, growth in customs collections in November was significant. This is along expected lines, as import orders usually do not get cancelled with a short notice. However, for the month of December, the customs collection was 7.6 percent lower than the same month in 2015. It is too early to say what this decline means. In the past also, there have been months when customs collections declined even when there was no obvious explanation.

### Corporation tax

Corporation tax collection during April-December was 4.4 percent higher than the corresponding period last year. Last year, during the same period, the growth in corporation tax collection was 11.74 percent. Collection in December 2016 was 4.7 percent lower than that in December 2015. December is one of the months for deposit of advance taxes. This low collection may indicate that firms have revised their profit forecasts downwards. However, it is difficult to draw a conclusion, as there may be other explanations. For instance, disproportionate refunds may have been made in December. We do not have the data to draw a conclusion.

### Income Tax

During April-December, income tax collection was 24.6 percent higher than that in the corresponding period of 2015. However, a key factor here is the income declaration scheme that ended on September 30, 2016, and had mandated payment of tax, surcharge and penalty by November 30, 2016. The expected tax inflow from the scheme was about Rs.30,000 crore. This is a form of additional revenue mobilisation. Hence, unless we know the increase without this ARM, it is difficult to interpret the number. For instance, if Rs. 25,000 crore was collected under the scheme, the increase in income tax collection during April-December would be just over 8.3 percent.

### Conclusion

The analysis presented here suggests that the reading of tax collection numbers as signifiers of robust economic activity may be too optimistic. I have had to estimate some of the numerical values above because the data releases on tax collections have been parsimonious on details. This is especially true of the release on January 9. The consistent inclusion of details about additional revenue mobilisation for different taxes for each month would make it easier to conduct economic analysis using tax data.

Update: MR Madhavan pointed me to another source of additional collection under direct taxes this year. From this year, 75 percent advance taxes have to be paid by December 15, while till last year, this was 60 percent. This may have been a significant source of additional collection in December. So, the fact that corporation tax collections this year have been lower than they were in December 2015 is much more indicative of decline in economic activity.

The author is a researcher at National Institute of Public Finance and Policy. Views expressed here are personal.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.