<link rel="stylesheet" href="//fonts.googleapis.com/css?family=Open+Sans%3A400italic%2C700italic%2C400%2C700">JayEnAar, Author at Aam JanataSkip to content

About JayEnAar

Public health physician, Commentator and blogger on public policy, Statistics and Data science enthusiast, Liberal.

1

Governments all over the world resort to spin to present themselves in the best possible light.

The boundary between gloss and outright lies can often be blurred. Statistical information is the easiest to fudge. The actual numbers may not be fabricated; that would be too obvious. Instead the spin masters use a few standard tricks: they select which numbers to present and which to hide, and they drawing unwarranted conclusions from a headline number. Official statistics present every opportunity to cross the line from disingenuous spin to downright misleading propaganda - or lies, in plain English.

Statistics is like a bikini, as the wit said, what it reveals is interesting; what it conceals can be vital.

Here’s an example of misuse of official statistics from India.  The Finance Minister tweeted a headline number about annual revenues from direct taxes. For the first time in 2017-18 it had crossed the 1 trillion rupee mark.

But he did more than just mention that 10.02 lakh-crore number. He made 2 further statements - one was implied and the other is an explicit and clear conclusion. Both bear further examination.

The statement that the 2017-18 revenue was 18% higher than in the previous year was not there just for completeness. The implication is that it was a substantial rise; it was a creditable, praiseworthy performance of the Indian economy. Is that really true?  

An 18% rise, on the face of it does sound a whopping rise! And it would be creditable if it had bucked the trend. Is that the case?

We do have the figures going back to 2001. They are here, on page 2 of this report from the Govt of India.

The data cover the years 2001-01 to 2016-17. Lets take the 10.02 trillion Rupees number given by the FM in his tweet for 2017-18 and add it to the existing time series data. The following chart shows the growth in direct tax revenues over the years.

Now, set in the context of previous years’ data the 10.02 trillion figure for 2017-18 looks like it is no more than the trend growth.

What about the  18% y-on-y growth? That sounded impressive. Again, taking the entire data series and plotting the y-on-y % aage increase over the previous year, we get this interesting plot.

So, now the 18% looks par for the course; its good but can it factually be described as ‘historical’? The y-on-y growth has returned to levels last seen in 2010-11, and is nowhere near the amazing near 40% growth seen in 2006-7, and the 36% seen the following year.

Now lets turn to the second part of the statement. The FM drew a clear conclusion about what led to the (not so) creditable 18% growth in 2017-18. He ascribed it unequivocally to ‘accountable governance’ under his leader, the Prime Minister. Leaving aside for the moment the question of the extent to which the Govt allows itself to be held to account, if the 2017-18 performance is ascribed to the leadership of the Prime Minister, the question that begs to be asked is this:

What led to the  9% growth in 2014-15, the 6.6% rise in 2015-16,  and the somewhat better 14.5% in 2016-17? You can’t claim credit for a good (if not stellar) performance and deny responsibility for a dismal performance.

References:

  1. Income Tax Department,  Time Series Data, Financial Year 2000-01 to 2016-17. See https://www.incometaxindia.gov.in/Documents/Direct%20Tax%20Data/Time-Series-Data-2016-17.pdf

1

DeMonetisation did not promote the uptake of digital transactions

Driving India towards a less-cash, digital payments economy was one of the aims claimed by the Prime Minister when he invalidated 86% of India's circulating currency. The reasoning was that India was a largely cash-based economy; if circulating cash was reduced, people would rapidly move towards electronic, or digital payment systems for their commercial transactions. That at any rate was the hope.

Did it happen?

Rupa Subramanya thinks it did. She claimed as much in a blog in the Hindustan Times. She accepts that the original aim of taking out black money has not been met, given that almost all of the Specified Bank Notes (SBNs) hav now been returned for exchange or deposit. But she goes on to say that her research shows that the secondary aim of pushing the country towards digital payments and away from a cash based economy has been achieved. To quote from her article:

....several key components of digital payments such as Point of Sale Debit and Credit (PoS) purchases, National Electronic Fund Transfer (NEFT), Immediate Payment Systems (IMPS) and mobile banking, are way above their pre-demonetisation trends

.....

The bottom line of the research conclusively demonstrates that there was a structural break after November 2016 with a permanent increase in digital payments and decrease in the relative importance of cash. Whatever you may think of the original goals and whether they succeeded, it’s clear digitisation is one demonstrable success story of demonetisation.

She expresses the use of digital payments not in absolute terms but as a proportion of total M3 Money. However, though she says she has published her findings,  they are not in peer reviewed journals. Rather the findings are published in research papers for the Observer Research Foundation - a think tank. I haven't seen these papers and therefore cannot comment on the methodology of the research. In any case she does not cite a source, nor does her article present the full results of her analysis. Her data source, though is the same that I have used earlier in a series of tweets and in a twitter Moment. This is the Reserve Bank of India's Database on Indian Economy.

I believe her conclusions are premature, they may even be misleading or wrong, based as they are not on absolute value of payments but on payment volumes as a ratio of M3.

I present here my own much simpler and more intuitive analysis of the RBI data set and draw very different conclusions.

Data and Methods: The RBI dataset consists of monthly transaction amounts for each of several different modes of digital transactions. The data goes back to 2004 and the latest available data is for August 2017, 10 months after DeMonetisation. Rather than look at just at the figures a few months either side of D-Day (DeMonetisation-Day if you are a fan or Disaster-day if you are a critic), I suggest it is best to look at the entire period. Since these are time series data (defined as data collected consistently with a defined periodicity - in this case monthly) the most obvious and simple technique would be to chart the data against time, draw a vertical line at D-day and look for a change in the trend . If D-Day did indeed result in qualitative sustained change in aggregate behaviour the change in trend would be obvious. I used the statistical programming language, R and the charting package ggplot2 to draw and annotate the charts. These statistical programmes are widely used in academia and business.

Results: The results of my analysis are best presented as a series of charts. They speak for themselves.

The key point to get is that by looking at the entire time series for each of the main digital payment modalities, two conclusions leap out immediately.

One, that in the first few months after D-Day there was a spurt the volume of payments made by digital means.

Two, for some payment modalities, they have subsequently fallen back to levels that were seen well before D-day. In particular retail electronic clearing, and plastic card volumes are effectively back on the same trend growth they always were since long before D-Day. Mobile banking transactions in particular were going up steeply in the months before D-Day was even a glint in anyone's eye; they went up even more steeply after D-day - and here's the crucial point, they have more latterly dropped right back. If we superimpose what we know about the re-introduction of new currency notes this looks like a perfect fit. As cash was re-introduced into the system, people began giving up on mobile banking transactions.

Another set of charts looks at the volumes (i.e number) of transactions.

Here there appears to be a small shift upward that, despite some month on month fluctuation appears to be settling down at a level clearly higher than anything seen pre D-Day. In the case of mobile transactions it is small-ish numbers and starting from a very low base; in the case of digital transactions its a step change from about 1.4 billion transactions a month to about 2 billion. But the volumes transacted appear not to have shifted much at all - it is in keeping with long-running trend and it is certainly not a step change.

Undoubtedly, there has been a steep growth in the number of Point of Sale outlets, as shown here:

Starting from a very low base, this is only to be expected given the huge Government push including cash incentives and subsidies for the take up of POS machines. The extend to which these have penetrated much beyond the largest urban centres and the plushest retail outlets is the big question. The last chart above is based on data published by NITIAayog

Conclusions. My analysis leads me to conclude that any effect of DeMonetisation on the use of digital payment systems has been transient, small and short-lived. Some change has occurred (POS terminals for example) but the fact that both retail electronic clearing and card usage is back on what I call 'trend growth' (i.e. on the same trend as obtained before D-day) would suggest that there has not been a structural change that can be confidently ascribed to DeMonetisation.

Post-script discussion. In all the commentary on Digital Payments insufficient attention has been paid to a most fascinating report that was published on Oct 5 by Visa India. Amitabh Kant, CEO of NITIaayog, wrote the foreword to this report. Nobody who read the detailed figures or had taken in the measured  recommendations in this report would have supported a sudden, cataclysmic and disruptive withdrawal of 86% of the currency, certainly not with the intention of promoting a digital payments economy. Among the key findings of the report are:

  • Cash usage costs the economy 1.7% of GDP (Note: borne largely by the State)
  • According to a 2014 World Bank survey, only 0.38 percent of women above 15 years old used the internet to make payments compared with 2.04 percent
    of men; 3.25 percent women had used a debit card versus 5.25 percent of men.
  • The cost of a point-of-sale (POS) terminal in India ranges from INR 8,000 to INR 12,000. The annual operating cost is INR 3,000
    per terminal. Low transaction volumes especially outside of Tier 1 cities, make it unviable for banks to expand their footprint into such segments.
  • RBI and the Govt of India already had a plan to transition to a less-cash economy.
  • If India  invested a total of INR 58,000 crores (USD 8.6 billion) over the next five years through tax
    rebates, it could not only expedite the pace of payment digitisation but also save about INR 70,000 crores (USD 10.4
    billion) in that period through a reduction in the cost of cash with a potential to save 4.7 lakh crores (USD 70 billion)
  • If we invested 60,000 crores and undertook a series of reforms and regulatory changes cash use could come down in 2025 from 1.7% of GDP to 1.3%.
  • In particlar see exhibit 7 of the rport which details the benefits from a sustained programe of policy changes as well as investments to improve the infrastructure for digital transactions. Effectively, a 5 year programme of sustained policy implementation and investment would potentially result in a growth of digital payments for Personal Consumption Expenditure from 4% to a whopping 36%, a drop in cash need from 11% of GDP to 10%.

 

A year on the Prime Minister's Great Idea may have turned out to be a dud

Mao ZeDong’s Great Leap Forward [1]  has to be the most outstanding example of the devastating harm from the unintended consequences of a state policy that aimed to modernize and develop an entire country.  It resulted in the deaths of 45 million Chinese in 4 years [2].

That was possible only because China was a Communist dictatorship and Mao held absolute power over both Party and the people of China. He decided it was a good idea and the Great Leap Forward happened.

In contrast, Mr Modi’s sudden, dramatic and hugely disruptive announcement of a year ago on Nov 8 2016 [3] was a tame affair; only a few score people died[4]. Like the Great Leap Forward, it too was one man’s Great Idea [5]; the aims were similarly laudable even if the goalposts kept changing; unlike Mao ZeDong though, Mr Modi was an elected leader of a Party that had won a decisive mandate.

Black Money was a major problem, declared the Prime Minister, and it called for a dramatic, decisive and bold step. Effective midnight 8th Nov 2016 the 500 and 1000 Rupee notes would be raddi ('worthless pieces of paper' to use the PM's words). New notes would be issued, including inexplicably a 2000R-Rupee note; and people left holding the old notes would be able to exchange them at banks or deposit them for credit to their accounts.

Almost everyone in India was affected and quite a few overseas Indians. Those with real black money (held as cash, you were safe if all your illegal wealth was held as gold or real estate)  found ingenious ways to convert their illegal stash of old notes into bank deposits.

The others, especially the poor, suffered the most. Day-to-day commercial transactions seized up. Shopping for groceries, taxi rides, buying a train ticket, paying the utility bill - all the routine stuff of everyday life became hard. Given that the notes that were declared illegal made up fully 86% of the currency-in-circulation, and that for all but the richest urban citizens India was still predominantly a cash economy, this was hardly surprising. Daily wage labourers lost livelihoods; victims of domestic violence lost the money they were hiding from violent partners; small businesses saw customers turn away; smaller businesses and street traders could not afford to take up the offer of Point of Sale equipment.  The rural sector was worst hit;  agricultural markets collapsed in a state-ordained market failure. [6]

But there was also widespread support for a ‘decisive strike against the rich and the corrupt’; in the days of chaos that followed, support for the Prime Minister hardly wavered. The cause was a noble one and people were prepared to make personal sacrifices for the national good. In time the economy would pick up, more of the informal cash-driven sector would be persuaded, cajoled or dragged into the formal, digital-transaction banked sector, the tax take would rise and India would become a modern rich economy. Trillions of rupees would not be returned to the banks by rich crooks and the ensuing windfall would be put to good use in building up national infrastructure. That, at any rate, was the hope.

It remained a forlorn hope. None of the claimed benefits materialised.

By June 2017, even the Govt's staunchest media supporter, SwarajyaMag.com acknowledged that the move had not lived up to the expectations. [7] press Very little Black money has been unearthed. After much delay a discredited central bank finally came out with the figures that almost all of the notes in circulation have been handed in [8]. There were no major prosecutions for tax evasion or illegal money laundering.

The process of re-monetisation with the new notes gradually picked up and by the 1st anniversary the total currency in circulation was back to 85% of what it used to be. Cashless transaction rose in the early days after Nov 2016 as people were forced to use alternative means of payment but have since fallen back to previous levels as currency became available.[9]

The wider economic damage too has been widely acknowledged. GDP growth fell back to levels last seen in the worst years of the previous  regime.[10] Jobs growth just did not materialise.

The Great Idea of 2016 will continue to be assessed, studied, debated and analysed for a long time, [11] But some questions may never be unanswered for many years to come.

  • What advice and analysis went into the formulation of the policy? Were experts consulted at all?
  • What was the role of economic and finance policy institutions like the Reserve Bank of India and NITI-Aayog? Did they play a role in the formulation of the policy and its implementation or were they relegated to serving as mere apologists for the ill-effects of a decision taken by an autocratic Prime Minister?
  • Why did Cabinet not protest at being ensconed in a room without access to mobile phones as the decision was announced?
  • Was it not the role of Parliament to hold the Government to account?
  • Will there ever be an independent cost-benefit analysis of the decision?

 

References:

  1. See this wikipedia account of the Great Leap Forward.
  2. See: this review of a book on the subject. I acknowledge that I have not read the book in the original.
  3. See this article in the Scroll.In for a review of how the news was covered in the newspapers on Nov 9th 2016.
  4. Deaths attributable directly and solely to the scrapping of notes was always going to be difficult. That a number of deaths occurred in queues is undeniable. Were they caused by the need to stand in queues? That's more difficult. Arguably the distress, the economic harm, the job losses and the lost wages/livelihoods and savings took its toll on ordinary people. The exact number of deaths became a political ding-dong  that diverted attention from the bigger question of the wisdom of the policy.
  5. We'll never know for certain that the final decision to go ahead and DeMonetise the currency was entirely Mr Modi's. There has never been a proper enquiry. All the indirect evidence points to it being either solely or largely his decision and his alone. Much later on it emerged that the RBI Board met on the morning of the 8th Nov and agreed to a Govt proposal but the delay in publishing this resolution leads to the suspicion that it was a hastily put together fig leaf. See: this and this . There's also speculation that a war on cash was one of the suggestions put forward to Mr Modi by an engineer and keen campaigner for tax reform Mr Anil Bokil of the Pune based ArthaKranti Foundation . Its worth noting that these ideas have no traction among mainstream economists.
  6. See the writings of P Sainath on the effects of the noteban on rural economy of India.  https://ruralindiaonline.org/articles/demonetisation
  7. SwarajyaMag.com is an online journal that is openly and avowedly right wing and a keen supporter of the PM's party. In an unexpected op-ed piece on June 14 2017, R Jagannathan the editor declared Demonetisation to be a failure but argued that the critics were right for the wrong reasons. Their criticisms, he argued was led more by animosity towards Mr Modi than by any special economic insight. But even I, as an amateur student of economics, argued in my blog of 16 Nov 2016, a week after the decision to demonetise, that it was a flawed policy that would do nothing to root out black money. I argued that DeMonetisation would cause tremendous hardship and loss to large numbers of people, that it would not deliver its claimed benefits, that there were other better targeted means of combating black money.  It was, I argued neither necessary nor sufficient to make a serious dent in black money. At that time, it is important to note, the stated aim was to eliminate black money. The push to a digital cashless payments system came later on.
  8. The earlier, almost gleeful, expectation was that as much as 3.5 to 5 trillion rupees worth of high denomination notes would not be handed back in and would be a free windfall for the Reserve Bank of India which would see a dramatic drop in its liabilities. This would be a huge bonanza in the form of a one off dividend from RBI to the Govt. This euphoria evaporated when someone pointed out that a decline in liabilities affected the balance sheet but would not lead to a profit and the RBI act  required it to pay a dividend only out of annual profits from banking activities. In the event the actual dividend that RBI paid out to the Govt in 2017 actually fell by almost half compared to the previous year. The losses arose out of scrapping the old notes, printing new ones, and the extra logistics costs of shipping the new notes out to where it was needed.
  9. I published a twitter thread and a moment with analysis of month-by-month time series data right up to August 2017 of the amount of money that flowed through non-cash digital payments systems. These include bank-to bank systems, like real time gross settlements used by businesses, paper-based payments systems (bank drafts and cheques), retail electronic payments, credit and debit card payments, and mobile banking payments. These charts show that any effect of DeMonetisation has been at best short-lived. There has been a growth in the number of point of sale terminals but from a very low base, and a growth in the number of subscribers to mobile phone based payments systems.
  10. The GDP growth slowdown has been widely commented upon. The standard Govt response has wavered between arguing that DeMonetisation was necessary medicine for a a backward economy built on cash-fuelled corruption, and a counterattack that the slowdown is not due to Demonetisation but was in the making long before Nov 2016. As arguments go both are own-goals and ill-serve the Govt's credibility.  
  11. The Harvard Business review paper argues that the 4 lessons to learn are
    • Choose your experts carefully. Mr Modi may have been influenced by a few cranks posing as economic experts with not so much out-of-the-box ideas as off the wall thoughts.
    • Dont ignore basic data. All the evidence was that only 6% of black or illegal wealth was held in cash. Not attacking the sources of corruption - politicians, real estate, and big businessmen meant tha instead of a targeted approach we had an assault on everyone - honest and weak included, in which the rich and corrupt got clean away.
    • Consider human behaviour. People found a way out of the cash crunch both to manage their poor honest lives and to squirrel away whatever illegal cash they held. Digital transaction was already growing as fast as it could given the infrastructure available, so as soon as new cash came into the system any spurt faded away.
    • Beware of digital silver bullets. India came 41st out of 42 countries just ahead of arch-rival Pakistan in the infrastructure needed to support a digital payments eco-system. However 'bold' and 'decisive' an executive ordz er cannot replace patient attention to detail.

Many of the problems and issues in India – witness the political debate and the controversies that are so rife on twitter and other social media – are to do with either the relationship between people and the state, or the fine balance between the competing rights of individuals and communities.

It may be a generalisation but I believe that at its core these disputes arise from a fundamental lack of clarity in our own minds of the status of the people.

Are we citizens or are we subjects?

Until  70 years ago there was no question – we were subjects of one king or imperial power or the other. Even if you go back to before recorded history we were subjects of a King who later came to be regarded as God. Kingdoms large and small were sometimes replaced with empires as conquests superseded earlier victories. 

That changed dramatically with the ‘tryst with destiny’ in 1947 when almost overnight we became, in theory at any rate, citizens of a free sovereign nation. In 1950 we became a Republic when we gave ourselves a We-the-People type Constitution, thus setting in stone, so to speak, the role of the people as free citizens in determining the future.

But has this change of status from subject to citizen been reflected in how the State treats us; in how we treat and how we demand to be treated by the state;  and how we respond to each other and react with the State? Do we feel ourselves to be empowered citizens or do we behave like loyal (or disloyal) subjects? Do those we elect to run our affairs treat us as free citizens to whom they are accountable, or as subjects who can be taken for granted?

Make up your own mind But consider the distinction I make between citizens and subjects.

Citizens have rights and freedoms. They have independence of thought, speech and choice; they have agency. Citizens elect their governments to run the affairs of State – their State. They implicitly consent to laws that their elected representatives frame for the greater good of all citizens. When citizens are deprived of their basic freedoms by the agencies of the State it is only by the fair and impartial rule of law, and only after due process. When the State uses its power unreasonably citizens protest and expect at the very least not to be ostracised for it. 

Subjects, on the other hand, are people who have a duty, explicitly stated or otherwise, whether or not they agree, to serve the interests of the ruler. Such rights as they do enjoy are granted to them  at the will and behest of the ruler, capable of being restricted or removed at will. The ruler frames laws mainly for his benefit and implements these laws arbitrarily and not always with due process. As a consequence subjects can have new obligations imposed on them and rights and freedoms abrogated at the whim of the ruler.

A nation of citizens has a truly representative government with transparency, accountability and the rule of law applying equally to both rulers and citizens. The institutions and agencies (that run the country according to the will of the people as expressed through an elected legislature) do so impartially because firstly, they follow a code that requires them to respect the citizens, and secondly, they are independent of the state.  

It is possible in a country of nominally free citizens under an electoral democracy to end up with power going to a small group or ruling class that, whether benign or otherwise, acts mostly in its own interests with little accountability and little or no need for openness. It applies the law selectively, sparing its friends and cracking down on those who would question it. Indeed, the ruling class in a such a country with (de jure) citizens as  (de facto) subjects uses the law as a tool for oppression rather than for ensuring a free society. Such a power elite needs necessarily to subvert democratic processes, control institutions, and suppress dissent in order to maintain its hold on power.

Apply these criteria to the myriad ways in which the State in India interacts with the people, either liberating them or constraining them, either facilitating personal and collective freedoms or restricting individual liberties and group rights. Think about these criteria in relation to how we as individuals and communities interact with each other. Do we support each others’ freedoms or do we we deliberately or inadvertently reinforce our role as loyal subjects?

As a Twitter correspondent of mine put it, its actually quite simple:

Subjects are granted some rights by a sovereign Ruler.

Citizens are sovereign and grant some (temporary) power to a selected ruler.

I believe there is a long way to travel – before we can say we are no longer subjects but free citizens.

Apologists for Prime Minister Modi’s Demonetisation policy are now claiming it led to a  big rise in direct tax revenues for Financial Year (FY) 2016-17. See, for instance this from Swarajya: 

Does this stack up?

The obvious logical flaw in this line of reasoning is the well-known ‘Post hoc ergo propter hoc’ fallacy. This is a Latin phrase that means, ‘After this, therefore because of  this’. The logical fallacy lies in the assumption that X caused Y because Y happened some time after X. It is possible that Y happened despite X, not necessarily because of X. Indeed if X had not happened then it is at least conceivable that something even better than Y might have resulted; we’ll never know. But without further analysis the conclusion is always suspect that X directly or indirectly caused Y.

How do we judge the claimed rise in 2016-17 tax revenues? Is it something we might have observed anyway? Regardless of what happened on Nov 8? Luckily for us the tax revenues in each year are part of a time-series data set published by the Government. The full report is here:

Table 1.1 of this report shows the tax revenues for each year since 2000-01. The first few rows of this table are shown here.


I have examined the trend in tax revenues by drawing trend charts. The following charts show the trend in direct tax revenues for Personal tax and for Corporate tax.

Tax Collection 2000-01 to 2016-17

Personal Income Tax

Corporation tax

These charts clearly show that the trend over nearly 2 decades has been one of  steady year-on-year growth in direct tax revenues.

This is of course to be expected. India’s GDP has grown over this period, economic activity has expanded, wages have risen and more people are tax payers - albeit the vast majority of people are too poor to be liable for income tax. In 2016-17 the total number of direct tax assessees (including companies, partnerships, trusts, individual, and 6 other categories) stood at 62.7 million in a country of 1,250 million people.

The key point is this: the growth in tax revenues in 2016-17 is just trend growth; it’s part of the pattern that we saw for the last decade. At least that much growth would almost certainly have happened even if the PM had not declared 86% of the currency to be no more than worthless scraps of paper.

Arguably it  might have been higher if demonetisation had not battered the economy as hard as in fact it did; we’ll never know. But to claim that the rise this past year is one of the benefits of demonetisation is to clutch at straws. It is a deliberate misuse of the official statistics; It is a cynical attempt to hoodwink the public into thinking that the sacrifices they willingly made in the months after Nov 2016 was worth it. The least the Government can do is not  insult their intelligence by claiming benefits on a fallacious and specious use of statistics.