Demonetisation and the Delusion of GDP Growth

The Modi-led National Democratic Alliance decided to demonetise 500 and 1,000 currency notes on 8 November 2016 thereby destroying the value of 86% of the currency in circulation in one stroke. This was unprecedented in every sense. No other country had ever experimented with such a policy measure to curb the size of the black economy. No other country had ever imposed on itself such a massive economic shock endogenously.

Whilst this move (combined with other measures aimed at crunching the size of the black economy) will yield benefits in the long term, we highlight the fact that the move undeniably affected the gross domestic product (GDP) growth rate adversely in the short term. Even as the negative impact appears obvious, the need to explain and quantify the same is necessitated by the fact that the country’s apex statistical body has declared that the GDP growth in the third quarter of 2016–17 (3QFY17) was recorded at 7.1% year-on-year (YoY) in real terms, which is just 10 basis points (or 0.10) slower than the GDP growth rate of 7.2% YoY recorded in the first half of 2016–17 (1HFY17).

The subsequent part of this article is divided into four distinct segments. The first segment highlights why it is almost impossible that India’s GDP growth was unaffected by demonetisation given India’s large informal sector and its high dependency on cash transactions. The second segment attempts to query the official GDP statistics by showing how either the formal GDP growth data or the bank credit data provided by the Reserve Bank of India (RBI) can be true. In the third section we show using high frequency indicators provided by industry bodies that economic momentum slowed decisively in 3QFY17, and then recovered in the fourth quarter but not to the same levels seen pre-demonetisation. Finally, we conclude with our view regarding India’s GDP growth prospects in 2017–18. We highlight that even as cash comes back into the system, we expect the informal sector to deliver a lower GDP growth in 2017–18 than it would have been the case if the government had not aggressively enforced an increase in tax compliance as it has been pursuing over the second half of 2016–17 (2HFY17).

Adverse Impact

Why is it hard to imagine that India’s GDP growth was unaffected by demonetisation in 3QFY17? Unlike more developed economies, India’s informal sector is large and more importantly it is labour intensive. The informal sector accounts for approximately 40% of India’s GDP and employs close to 75% of the Indian labour force according to the National Sample Survey Office (NSSO). In absolute terms, this means that the informal economy generates GDP roughly worth $950 billion (bn) and provides employment to 398 million (mn) of India’s total labour force of 530 mn (historical growth rates are applied to NSSO’s 2011–12 estimates to arrive at workforce number for 2017–18).

Given that the informal sector relies on cash heavily (from making payments to employees, acquiring raw material from suppliers to collecting revenue from customers) and given that this sector accounts for a large chunk of the Indian economy, prima facie it appears unlikely that a move aimed at destroying the value of 86% of the cash in the system had no impact on the level of economic activity of this segment of the Indian economy. Besides, it is worth noting that India is a heavily cash-reliant country where more than 80% of all transactions take place in cash according to RBI data.

Now the plot thickens in the Indian context, because despite the size of the informal sector being material, there is a serious inadequacy of data in capturing the health of this sector on a contemporaneous basis. The main source of pan-India informal sector data is provided by the NSSO. To complicate matters: (i) this data is published with a lag of a year; and (ii) this data is captured at a frequency ranging from two to five years. For instance, this data is available for 1999–2000, 2004–05, 2009–10 and 2011–12. Therefore, to independently gauge the short-to-medium term impact of demonetisation on the informal sector, we relied on a combination of qualitative and quantitative tools.

We first conducted a survey spanning 88 small- and medium-sized enterprises (SMEs) across India. The responses were collected over the post-demonetisation period spanning 22 November to 2 December 2016. Even as the sample size is very small, the extreme nature of the findings suggests that the SME sector was undeniably under duress in the weeks post demonetisation. The survey yielded four key takeaways, namely: (i) SMEs in India remain highly dependent on cash with more than 50% of the costs and receivables being transacted in cash; (ii) sales growth for SMEs was likely to come under meaningful pressure in the short term with almost 50% of SME promoters believing that their sales would collapse by 40%–80%; (iii) 59% of the SMEs expected competition from the organised sector players to pick up in the medium term; and (iv) non-performing loans in the SME segment seemed likely to increase for lenders as many SMEs said they will not be able to collect dues from their customers and could see a prolonged adverse impact on their sales. To be specific, 32% of surveyed SMEs felt that they will not be able collect more than 50% of their dues from their customers.

Apart from the survey, we travelled extensively over the course of December 2016 and January 2017 into the interiors of India focusing mainly on the SMEs in the informal sector. The first and most important takeaway from these travels was that businesses in the unorganised/informal sector that sold products mainly in cash and also covered their costs in cash were affected most severely. For instance, Panipat in Haryana is the textile hub of North India. It is a 310 bn industry with 60 bn worth of goods being exported. It employs around 3,50,000 workers. Whilst our interviews suggested that the export-focused units were largely unaffected, the domestic component of the industry saw business activity fall by 40%–80% as this component of the business was more reliant on cash. As a result, almost half of the 3,50,000 workers employed in the region had been temporarily laid off as demand collapsed in the domestic market and there was no cash to pay wages.

Similarly, we visited Tiruppur in Tamil Nadu which is a textile hub of South India, though much larger than Panipat. The textile industry in Tiruppur is roughly 500–600 bn in size and employs around a million workers. Here too, the export-focused units and the units operating in the organised segment did not experience major disruption due to demonetisation. However, the domestic segment, of which 40%–50% of the market is unorganised was facing problems that were similar to those in Panipat. The units were running only three days a week (compared to seven days before demonetisation) owing to the lack of demand. A bank official with a local private-sector bank in Tiruppur made the point that prior to 8 November 2016 their branch used to disburse 40 mn weekly to their clients as they needed to pay their labourers. This amount had dropped to 6 mn due to the shortage of cash.

Despite the overwhelming evidence suggesting that India was dealing with an unprecedented economic shock and the fact that the informal sector was crippled by the shortage of currency post-demonetisation, the Central Statistics Office’s (CSO) quarterly GDP estimates for 3QFY17 failed to capture the same. This, to be clear, is likely to be an error of omission and stems from two technical reasons. First, the quarterly estimates published by the CSO by definition are a result of an extrapolation exercise based on partial data from the organised sector. This is likely to be problematic as 3QFY17 marked an inflexion point and such a point cannot be captured by the formal estimates if they rely on extrapolation. Second, the quarterly numbers published by the CSO estimate growth in the informal economy using formal economy-related data. This can be problematic especially because the informal sector of India is where the GDP growth slowdown is likely to be concentrated. In fact, the Economic Survey for 2016–17 admitted that the CSO’s GDP statistics will underestimate the impact of demonetisation noting that “the national income accounts estimate informal activity on the basis of formal sector indicators, which have not suffered to the same extent. But the costs have nonetheless been real and significant.” However, even the authors of this fine document, for an inexplicable reason, refused to admit that the adverse impact on GDP growth was meaningful. In their assessment, the demonetisation move would only shave off 30–50 bps from India’s steady state GDP growth rate.

CSO and RBI Data Mismatch

Does the GDP growth data and the monetary data provided by the RBI add up? As per the CSO, the GDP growth in 3QFY17 was recorded at 7.1% in real terms and at 10.6% in nominal terms. Separately, the RBI reported that the bank credit growth in 3QFY17 collapsed to a multi-year low of 6% YoY. Reading these two data points about the same economy makes it difficult to imagine how a bank-funded economy like India (where about 60% of corporate funding comes from the banking sector) managed to grow at the pace that it did when bank credit growth was languishing at a multi-year low.

In fact, history suggests that under the new GDP series, the delta between nominal GDP growth and nominal bank credit growth reached an all-time high in 3QFY17 (Figure 1). In other words, never before has the GDP growth data been so out of sync with the bank credit data provided by the RBI. Even as the CSO does offer technical reasons to explain this anomaly, for an investor looking to know where economic momentum in the country is headed, it is clear that the CSO data is underestimating the GDP growth slowdown which materialised in India in the second half of 2016–17. It is worth noting that even after adding non-bank borrowing to the bank credit data, credit growth trends remain weak (Figure 2, p 19).

Besides the bank credit growth angle, another way of triangulating the validity of the GDP growth data is through the “velocity” route. In a very simplified sense, the nominal GDP can be thought of as the product of “currency in circulation” (CIC) and the velocity (“v”) with which this money circulates. The CIC is a variable that the RBI/central government controls whilst the v tends to be determined by the levels of business confidence in an economy as well as the underlying structure of the economy. Consequently, v tends to be extremely stable and is known to be very sticky. For instance, under the new GDP series, the implied v in India (that is, derived by dividing the CSO’s estimate of quarterly nominal GDP by the CIC) has been range-bound between 2.0x and 2.3x. However, if the nominal GDP data provided by the CSO for 3QFY17 is correct, then that would suggest that India’s v miraculously shot up to 4x times in this quarter (Figure 3). Again, it appears highly unlikely that at a time when business confidence plummeted to historic lows is when the velocity of CIC shot up to a multi-year high.

Finally, it is worth noting that as at the end of 3QFY17, the CIC stood at 9.4 trillion or 6% of GDP in relative terms. This is exactly half of the CIC that was maintained in the economy before 8 November 2016 at 12% of GDP. Even if we were to make the generous assumption that v remained unscathed by the record economic uncertainty that businesses experienced in the second half of 2016–17, it is difficult to imagine how GDP growth too remained unscathed even as the remonetisation process has been far from complete (Figure 4). For instance, the latest data from the RBI suggests that CIC stood at 13.6 trillion or 9% of GDP as of 7 April 2017.

The True Story?

What do high frequency macro indicators tell us? Even as the official GDP data continues to be far removed from the economic realities on the ground, high frequency indicators (HFIs) continue to tell a different story.

We look at 13 such HFIs (Table 1) in order to get a sense of the real health of the economy in an environment where the official GDP data in India is losing credibility.

As is evident from the table, 3QFY17 saw gauges relating to consumption in the economy such as two-wheeler sales, passenger vehicle sales, and retail credit growth nosedive. For the same period, the CSO reported that the private final consumption expenditure growth in India grew at 10% YoY in real terms. It is worth noting that consumption expenditure on transport equipment accounts for 14% of the total spends. Thus, it is difficult to imagine consumption growth rising to a multi-year high even if the monsoons were good at a time when this chunk of consumption expenditure was under visible pressure.

As regards the fourth quarter of 2016–17, on expected lines, gauges related to consumption have started recovering from the lows of 3QFY17, but are nowhere near the levels seen in the second quarter of 2016–17 (Table 1).

A Full Recovery?

Will GDP growth in FY18 recover fully to the levels that were expected before demonetisation? As the economy recovers from the shock of demonetisation and as remonetisation takes place, a recovery from the painful lows seen in 3QFY17 is but obvious. However, what is critical to note is that the informal economy will shrink rapidly as the government’s crackdown on black money will make it difficult for businesses to operate in the informal sector without paying taxes.

Even as the effects of demonetisation are fading as the pace of remonetisation improves, the government has launched a host of regulations which will affect the functioning of the informal sector in the short term. On 8 February 2017, Parliament passed the Wages Act mandating business units to pay workers through the banking system and not in cash. The act seeks to enable employers to pay wages to workers through cheque or directly crediting salaries into their accounts. The new legislation will help keep in check the exploitation of workers engaged in the unorganised sectors and they will get exact salaries in their bank accounts. If businesses continue to pay the wages in cash then the law mandates that they will not be able to offset the wages as costs. This will increase the wage costs for the firms operating in the informal sector as the businesses in this sector kept their wages cost low when they paid in cash by avoiding a host of other benefits which should accrue to a worker by law. Moreover, increased tax compliance and the advent of the Goods and Services Tax regime could make a host of informal sector businesses unviable. The government has also empowered the junior-most officials in the income tax department to conduct raids.

Finally, it is worth noting that under Finance Act, 2017, cash transactions above 3 lakh will not be permitted: (i) to a single person in one day; (ii) for a single transaction (irrespective of number of payments); and (iii) for any transactions relating to a single event. Amendments to the Finance Bill, 2017 propose to lower this limit from 3 lakh to 2 lakh. Therefore, even as the informal sector will be in a much better state in the fourth quarter of 2016–17 as compared to 3QFY17, it seems highly unlikely that this segment of the economy will find itself undertaking the same activity levels that it would have if all these measures had not been launched over the course of the second half of 2016–17. In other words, even as the ephemeral impacts associated with demonetisation abate, we expect the informal sector to deliver a lower GDP growth in 2017–18 than it would have been the case if the government had not aggressively enforced an increase in tax compliance.

Whilst analysts and economists can differ on the extent of the impact the demonetisation had on GDP growth in the second half of 2016–17 and on 2017–18, the fact that it was material and the fact that it was adverse is undeniable.