Published by Westland Publications Private Limited in 2019 Ist Floor A Block, East Wing, Plot No.

40, SP Infocity, Dr. MGR Salai, Perungudi, Kandanchavadi, Chennai 600096

Westland and the Westland logo are the trademarks of Westland Publications Private Limited, or its
affiliates.

Copyright © Surjit S. Bhalla, 2019

ISBN: 9789388689120

The views and opinions expressed in this work are the author’s own and the facts are as reported by
them, and the publisher is in no way liable for the same.

All rights reserved

No part of this book may be reproduced, or stored in a retrieval system, or transmitted in any form or
by any means, electronic, mechanical, photocopying, recording, or otherwise, without express written
permission of the publisher.
 For all true liberals
Remember, to be objective is to be cruel—to self
Acknowledgements

Hobbies, like passion, can be painful. When hobbies are passion, it can be doubly painful. As I was
to find out when I embarked on this adventure of writing a book about a favourite interest of mine—
the analysis of elections. As anybody who has been courageous enough to forecast an election knows
—the best of your intentions, the hardest of your work, the deepest of your insights, and the best
counter advice of your friends and family will often not be enough to prevent you from deep
embarrassment—a wrong forecast. ‘Shame on you! How could you get it so wrong, when it is
intuitively obvious to the most casual observer,’ and so on.
It has happened to me not once, but twice before. Once in 1991 and the second in 2016—Rajiv
Gandhi and Hillary Clinton were my favourites. They both lost, unexpectedly. Given this history, I got
wise. I would write a book on the electoral economy of India from 1952 to 2019—enough for
everybody to disagree with in every chapter, and as a consequence, generate some agreement, and
some ‘rights’ versus a forecast that may possibly go wrong.
That is the origin of this book. A desire to combine economic and political events in the context
of the electoral history of India. Gives one an opportunity to vent on several issues of interest to
many, and with a casual insertion of the one that most people are most interested in—the forecast.
An ambitious project requires support, and in that regard I have been luckier than most. I am
thankful, indeed indebted, to Family and Friends—now that’s a TV serial worthy of consideration.
The family has been supportive, and indulgent, in allowing me freedom, and space, to pursue this
passion. Ravinder, Simran and Sahil—the last few months, I have been both absent and absent-
minded. The extended debates on the political economy of India have been more helpful than you can
imagine. Discussions with Suman Bery have helped to sharpen my thinking, and analysis, of the
political, economic, and social transformation project that India is today.
Good friendly interactions with Arvind Virmani and Lord Meghnad Desai have helped sharpen
my understanding of Indian politics. Ankur Choudhary, Sanjay Dutt, Varun Khandelwal, and Abhinav
Motheram were ever helpful with discussions, and comments on drafts. Abhinav also helped out
considerably with the research, and Rohini Sanyal was very helpful in the initial stages of this
project. I am also thankful to Pradeep Chibber and Rahul Verma for their useful comments.
My editor Sudha Sadhanand has been more than kind, and supportive of this study, and I
apologise for the delays and optimistic predictions of finishing. Thank you also for making this book
read better, which should help it sell more.
By instinct I am a contrarian—and how that is consistent with a DNA++ disposition, I honestly
do not know. But if being both contrarian and optimistic helps make correct forecasts, I welcome the
combination.
 Contents

Introduction
1. What is Election 2019 All About?
2. Democracy Begins
3. The Importance of Nehru
4. Does the Economy Matter?
5. The First Reformers
6. Sonia & the RSS: Myth or Reality?
7. Opinion Polls, Voter Attitudes, and Lies
8. The Birth of the Fake & Faulty Twins—News & Analysis
9. Mandir, Cows & Guns
10. Battle of the Elites
11. Modi and Inclusive Growth
12. The Empire Strikes Back?
13. A Second Innings for Narendra Modi?
14. Democracy Helps Usher in a New India
Select Bibliography
Yeh jo public hai, sab jaanti hai
Yeh jo public hai…
(The people know everything
Nothing is hidden from the public…)
– Roti, 1974
Introduction

It is show, sorry election, time. A few weeks hence, starting 11 April, the biggest tamasha in the
world is going to happen—the Great Indian Election. By virtue of its population, everything about
India is great, which is true—except of course China. But China is not a democracy—hence, we will
stick to the greatest tamasha unfolding in India.
This book is about the expectations around the winners and losers of this ‘experiment’, and
sixteen other experiments that have happened since the first election in 1952. But on a serious note, I
do less than justice to India, and to democracy in India, by calling it either a tamasha or a circus. This
choice of word was deliberate. I wanted to highlight the change that has already occurred in India
over the last seventy years.
Not so long ago, the elections were a grand festive affair with billboards and spectacles galore.
As India has become more developed (less poor), the opportunity cost of time (the economist’s
description of forced leisure time) has gone up i.e., we have less time for these events as more of us
are otherwise occupied.
At the same time, even as time-value has gone up, turnouts in Indian elections have also
increased. In Lok Sabha elections, the average turnout—defined as the number of voters as a ratio of
the electorate i.e., those eligible to vote—has also increased, from a low of 53.5 per cent in 1975, to
a high of 66.4 per cent in the last 2014 general election. For state elections—referred to as assembly
elections—the turnout reached a high of 76.3 per cent in the 2018 elections held in six states. This is
just one of the many indicators on offer to suggest that democracy is thriving, and well, amongst the
Indian populace.
Not so long ago, the philosopher-economist and Nobel Prize winner, Prof Amartya Sen wrote a
book called The Argumentative Indian. No book of his has ever rung truer. While this book does not
discuss the causes behind what seems like a quintessential trait amongst we the people, let me just
state that this is a characteristic that has been pointed out to me by every foreigner I have met. And no
Indian disagrees. We are like that only. In the first chapter, I shall discuss the origins of Indian
democracy suggesting that India became one, and stayed one, because of two necessary conditions—
being colonised by the British, and being ethnically, and culturally, very diverse. If you add
argumentation to the list, as one should, it really cements the case against those who believed, and
continue to believe, that democracy is not the preferred habitat of the Indian.
I perfectly understand that the ‘worth’ of my book might be decided by the accuracy of my
forecast for the 2019 election. But that would be a huge disservice to the author! This book is mostly
about the evolutionary process of Indian democracy, about where it was at inception in 1950 when the
Constitution mandated a democratic form of government, about where it is today, and everything that
has happened in-between.
Let us recount where we were at the time of independence. One of the poorest countries in the
world (the rank in US dollar terms was eighth from the bottom, and twenty-third from the bottom in
1996 PPP1 dollars), we have ‘come a long way, young woman!’ No matter what the criteria, India is a
vastly transformed economy, with the added advantage that the mindset of the people took a
significant rightward shift in 1991 and since 2014, there has been another decisive change in their
thought-process. India was among the poorest of the poor, with over eighty-five per cent of the
population defined as absolute poor by the World Bank ‘dollar a day’ poverty line. We were therefore
a very, very poor country in 1950.
Over the next thirty years, matters did not much improve. Poverty was estimated to have
declined marginally to seventy plus per cent in 1971 when Indira Gandhi launched her garibi hatao
(remove poverty) campaign. Twelve years later, poverty in India was still a very large sixty per cent.
Cast vs. Caste
It is therefore not very surprising that much of the discussion about who votes and why (one of the
major concerns of this book) centered around caste. Many of you must have heard the refrain—people
in India ‘caste’ their vote. But gradually, the genuine explanation about the reasons for voting began
to change. Education levels and per capita incomes began to improve. From an illiterate population
with the average educational attainment of the above fifteen year olds at only one year in 1950, we
progressed rapidly, especially in the last fifteen years, to an average educational attainment close to
eight years in 2018. What about poverty? The latest 2017–2018 consumption data is likely to show
absolute poverty to be close to the mid-single digits range.2 This is according to India’s official
poverty line which is almost exactly equal to the authoritative World Bank official poverty line of
1.90 PPP dollars in 2011 prices. In 2011, the poverty line was equal to a consumption level of 893
rupees per capita per month, or thirty rupees per day. In present prices, the poverty line is close to
1,300 rupees per person per month.3
What this translates into is the fact that from being a poor country in 1950 and 1983, India has
rapidly transformed into a middle-income economy (albeit lower middle income). This should mean
that the aspirations of the people, and their voting behaviour, has undergone a large structural change.
Stated differently, but pointedly, caste may no longer be an important factor in Indian elections, and it
is therefore high time that the media, politicians, and non-empiric intellectuals, recognised this simple
reality.
One might ask—what does this have to do with the analysis of elections? Everything. For
example, even today, many ‘enlightened’ ones (including good friends of mine) will enter into the
minutiae of caste differences by explaining the caste pattern of voting in municipal, assembly, and Lok
Sabha elections. It is an absolute fact (there is no other kind) that individuals of different castes vote
differently from election to election. It is also a fact that the same caste person will vote differently in
different elections. So how does elaborating on how a person from the Kurmi or Jat caste shall vote
in a given election help in forecasting her vote? It doesn’t, and that is the point.
Let us look at it differently. The caste composition has not changed for the last seventy years. Yet
castes have booted governments out, and in. In post-independent India, the citizen has not only
endured the Emergency, recessions, assassinations, poverty, and booms, but also many incarnations of
caste politics—Congress governments, V.P. Singh’s regime, governments run by the Bharatiya Janata
Party or BJP, and sundry coalitions of all of the above. These have been the varied outcomes of her
vote; what part did caste play? If I am missing something, please correct me. But prove first that caste
is still a factor, and the reasons thereof, in this globalised, middle class, Internet age.
And Then Came Fake News
Indians are a hugely modest lot, and this does not go with being argumentative. In contrast, the
Americans are moderately less argumentative—maybe the ordering on argumentation depends on
democracy and the size of the electorate. Just look at how the Americans have appropriated the
invention of fake news to themselves, as they did after the 2016 Trump–Clinton election. It was the
first national election in the world where fake news played a big part in the final outcome.
Furthermore, the Brexit referendum also involved fake news and this happened a few months earlier
in June 2016.
But I am more concerned with invention at home. And the invention originated in India after the
2014 election, also referred to as the ‘Modi election’. The argumentative Indian is shy about her
inventions. Even a cursory perusal of Indian elections over the last four years will convince any
sceptic that the ‘trophy’ belongs to India, its politicians, and journalists.
But what role could fake news possibly play after the election? A lot—and we discuss this
question in some detail in Chapter Eight (see p. 111). The 2014 invention most likely had to do with
the surprisingly large win by the Narendra Modi-led BJP—and the very large loss for the Congress.
How can the Congress party revive itself after winning as few as forty-four seats? Its previous low
was 114 seats in 1999, and it formed the government in 2004, albeit winning just thirty-one seats
more than the 1999 election. In 2009, it managed to gain an additional sixty-one seats to reach 206,
making the fall from grace in 2009–2014 a cumulative total of 162 seats.
The important point is that the revival of the Congress, and it ruling the Centre, is not an
impossibility. While it may require fresh, new strategies, the memory of a Congress government for
long periods of time is still recalled in people’s minds.
At the time of the 2014 election, the Congress had ruled India for all but twelve plus years of its
sixty-two years of having contested elections—the first election was in January 1952. The Congress
as the main Opposition lasted just three years—1977–1980; just one-and-a-half years from 1989–
1991; and eight years from 1996–2004.
The memory of Rajiv Gandhi’s illustrious grandfather, India’s first prime minister, Jawaharlal
Nehru’s wins provides the requisite moral heft for a Congress revival. Nehru won three elections in
succession, with a seat share of 74.4 per cent (1952); 75.1 per cent (1957); and 73.1 per cent (1962).
In 1984, a year when he lost his mother to assassins on 31 October, Rajiv Gandhi won 76.4 per cent
of the seats.
Indeed, pleasant memories of fabulous wins, but consider what happened in 2014. The country
was faced with a new paradigm—no dynasts, worse a chaiwalla (as the Congress derisively called
Narendra Modi several times starting with the 2014 campaign; a chaiwalla being a tea seller) won
282 seats in his first foray as a prime ministerial candidate. Worse, his fabulous win restricted the
Congress party to not even gain ‘respectability’ as the major Opposition party. (This honour goes to a
party obtaining the second largest number of seats, as long as this number is more than ten per cent, or
fifty-five seats—the Congress won only forty-four seats.)
After ruling India uninterruptedly for decades, and establishing a hegemony, it was obvious that
the ignominy of a terrible loss was unbearable for the established Congress elite to handle, and they
correctly surmised that they had to fight by non-traditional means to at least get to the magic mark to
be the only large and credible Opposition party, if not substantially beyond. One of the most important
aspects of that non-traditional means was none other than the deployment of fake news, a selection
that endeavours to create a favourable political atmosphere for the (weak) challenger. Also, it is not
something that is of much use to the incumbent, particularly an incumbent with as large a majority win
as Modi; and the first majority win in an Indian election in thirty long years.
Let us however not forget that the largest victory in Lok Sabha was recorded by Rajiv Gandhi in
December 1984: he obtained 415 seats. The election was held soon after his mother and prime
minister of India, Indira Gandhi, was assassinated by her Sikh bodyguards on 31 October 1984. Such
an overwhelming victory was never witnessed before, or ever since.
The fake news strategy is to turn conventional wisdom upside down—convincing voters of an
alternate reality, by creating a smokescreen for them to believe that what they see is not what they get.
In this unending cycle, as it were, the basis of wisdom needs to change—and it can change if
‘alternate’ facts to the prevailing reality are provided. If successful, this strategy can work wonders,
as it did in the US in 2016.
Here Come The Cows
As we will discuss later (in a whole chapter at a minimum), both ex-ante and ex-post, this was a
brilliant strategy on part of the Congress or the United Progressive Alliance or UPA. The plan was to
make people believe that the Modi government was no different from the UPA government—at best it
was UPA with a cow, an ‘eloquent’ phrase coined by a former member of the BJP and one of India’s
leading intellectuals, Arun Shourie. To what extent is this strategy likely to succeed? We will know in
May 2019.
There is of course no doubt that the reference to the cow is the unfortunate tendency amongst
large elements in the BJP (and as it turns out, large members of the Hindu community) and its
affiliates to think that the animal is holy, a mother, a holy mother—and hence needs to be protected
like an endangered species. This categorisation of the BJP is a bit unfair since (detailed in Chapter
Ten, see 147) the real fault lies with the Indian Constitution which provides for protection of the cow,
and extreme undue protection was provided to the cow in October 2005 by the Supreme Court, just
one and a half years after the UPA won a narrow victory in 2004.
Meanwhile, with the explosion of social media, the instrument and practice of fake news gained
both respectability as a tactic, and prominence as an electoral weapon for the challenger. President
Donald Trump’s Electoral College victory in 2016 gave fake news a credibility which was hitherto
never attained. In 2016, it was rumoured that the Congress party had hired Cambridge Analytica to
manage its 2019 electoral campaign—yes, the very same firm that had allegedly helped Trump spring
a huge surprise victory. It is conjectured that his victory was in part based on the use of fake news.
It is worth remembering that the father of fake news or analysis was none other than Joseph
Goebbels who reasoned that if you incessantly repeat a lie, it will become a ‘reality’. With social
media, it is much easier to reach people, and convince people, of lies. However, even as the reality
remains the same—to change people’s perception is the mantra. This is what has been happening over
the last few years in India, and has now reached a crescendo because national elections are no more
long drawn-out processes, but are concluded in barely thirty days.
The Role of Ideology
It would be a grievous error if I do not state the role of ideology in discussions about politics, and
economics, and practically all of social science research. It is all pervasive in journalism and is at
the background of nearly every piece of substantive social science research, whether it be from the
pen of Nobel laureates or the computers of mere mortals. Which is as it should be, because it is
ideology which provides the major impetus, the adrenalin, for all good research. However, if
ideology is present in research, the impact of fake news is enhanced.
Let us hope this is not the ‘reality’ that one has to live with henceforth. What separates good
social science research, and better journalism, is the separation of ideology (and opinion) from the
use of empirical evidence. In my various opinion columns (in leading Indian newspapers and the
count is well above 1,500), I have attempted to use empirical evidence whenever possible (indeed,
my critics accuse me of presenting evidence even when it’s not required).
It is easy to practice ideology—we all do, and should not refrain from admitting this fact.
Ideology is a very important part of most individuals, especially those in the social sciences. If you
delve into politics, either as a participant or as a practitioner, then, almost by definition, for you
ideology is important. The same applies to political science (or the study of politics), as also
economics (the making of policy which affects lives).
Recognising this reality still leaves us with a conundrum. How does one arrest the invasion of
ideology (a private matter or view) into discussions on policy (a public point of view)? If the
practice of ideology in public discourse has to be shunned, then the only recourse is to use empirical
evidence to substantiate one’s point of view. What separates analysis from ideology is evidence—and
we will use plenty of that in the course of this book.
The first chapter in the book is devoted to a discussion about what is the 2019 election all
about? India is at a cross-roads (it always is!), but there is a good chance that this fork is far more
important than the many it may have faced so far. After all, clues to the present are often contained in
the past, if not actually determined by the past. Hence, it is imperative that we take a peek into history
on the origins of our democracy, and the paths we have taken to arrive at 2019.
This book is about elections—hence a discussion on how they unfolded since the first one was
held in 1952 is in order. Chapter Two (see p. 16) delves into the adoption of democracy as the
preferred form of government, and the pros and cons of this selection. At the time of independence,
not many gave India any chance of succeeding as a democracy. It was poor, illiterate, and had a
minuscule middle class. But it succeeded—and how! What made this happen? In retrospect, and there
is considerable empirical evidence to support this result, India was destined to be a democracy.
There were two components which ensured this much desirable outcome. First, the fact that India was
colonised by the British, and the second, that India is diverse.
It is found that the path of development, and the hurdles to Indian success, have their genesis in
the Constitution. That is as it is supposed to be; but unfortunately, the Constitution has not served India
as well as it could or should have. It is telling that the Indian Constitution has been amended at a
faster pace than most people change their toothbrush.
There were other problems associated with the birth of a new order, and the inherent diversity
of India. The Constitution forced the country to adopt an extremely inefficient, and unjust system to
redress historical wrongs, and also skirted genuine attempts at providing equal opportunities. We
adopted a reservation system of ‘quotas’ for those affected by centuries of discrimination, rather than
a more just system of affirmative action. Alas, even seventy years after independence, India is
affected by this major historical wrong.
Chapter Three (see p. 34) in this book focuses on how important Jawaharlal Nehru was for the
development of a newly-independent India. While India gradually began to experience the fruits of
democracy, and the establishment of a liberal political order, economic policies were anything but
liberal. Therefore, even as India enjoyed plenty of political freedom, there was precious little
economic freedom.
On another note, somewhat relevant, there is very little evidence at a global level that
democracy hurts growth. For every dictatorship like China that has succeeded, there are ten others
that have not. During its low growth phase parallel to India, 1950–1980, China was a dictatorship
with little economic freedom and a shoddy economic growth, just as it was in India. The scourge of
poverty ensured that economic growth did not matter much for voters, and the Congress merrily ruled
for forty-plus years, except for three years of Emergency (no political freedom i.e., a classic
dictatorship).
The fourth chapter salutes the arrival of the new economic order—the importance of the
economy, rather than caste or other determinants for victory in elections. The new kid on the block
(for India) was the old adage made famous by President Bill Clinton—It is the economy, stupid.
Economic reforms began to be introduced post 1990, and along with it came a shift in voting
behaviour—caste began to matter less and less with economy more in focus. It is now nearly three
decades since the heady days of bold economic reforms witnessed in the early 1990s (covered in
Chapter Five, see p. 64).
For the voters, economic reforms and their continuance began to form an important basis for
voting preferences. Unfortunately for politicians, their dream run of not performing, and playing to the
galleries, but nevertheless winning elections, began to undergo a transformation, and non-
performance became a significant issue. It seemed as if the days of entitlement were finally over.
Chapter Five is about the first set of economic reformers in India, individuals who helped
cement the importance of economy amongst voters—P.V. Narasimha Rao and Dr Manmohan Singh of
the Congress, and Atal Bihari Vajpayee of the BJP. These were men who accelerated the much needed
economic reforms, and stepped on several toes. That is what reforms do—possibly this stepping on
toes was a greater factor, than the good economic performance, that led to the re-election defeat of
both Narasimha Rao and Vajpayee. Dr Manmohan Singh won the election in 2009 because of faster
growth (not economic reforms because there weren’t any during 2004–2009), and lost in 2014
because of low economic performance (and economic de-reform).
Chapter Six (see p. 81) is concerned with two extremely important political factors in
contemporary Indian politics—Sonia Gandhi and the Rashtriya Swayamsevak Sangh (RSS), and
attempts to sift myth from reality.
The next chapter discusses an aspect which has come to be recognised as one of the necessities
in a democracy, and something that people like me have indulged in for the last thirty odd years, but
whose value-add to the system is not apparent. I am talking about the analysis of polling behaviour,
and the art of forecasting elections. It is not in the least obvious that opinion polls affect electoral
outcomes, but there is enough of a connection felt by none other than the Election Commission of India
that bans the publication of any opinion poll days before the first vote is cast.
But there is a problem, and one caused by human nature—people lie, not only to their close
ones, but also to pollsters. Chapter Seven (see p. 94) recounts a very early attempt (1988–1989) at
developing a lying index of voters’ intentions, an index that proved absolutely accurate at forecasting
the 1989 defeat of Rajiv Gandhi. In many ways, this was the first attempt at deciphering the genuine
voting intentions of voters, in contrast to who they may have said they would vote for. As an aside,
Facebook and the Trump election campaign was well anticipated by my analysis of the 1989 Rajiv
Gandhi loss.4 But before one gets carried away, a statutory warning about the past (and future)
accuracy of this method. The lying index was also massively correct at forecasting the 1991 election
results—except for the two important and populous states of Uttar Pradesh, and Bihar. To find out the
why and how, I urge you to read the chapter.
Chapter Eight is about the nature of the 2019 electoral campaign to be adopted by the Opposition
—the discovery, and use of fake news. Let’s not forget, all political parties have fake news in their
arsenal—what is revealing is that fake analysis is mostly a feature used by the Opposition; the
incumbent stands to gain little by the deployment of fake news.
Both Rao and Vajpayee lost their re-election bids after stepping on many toes and primarily
because they brought about several bold and major economic changes. What does this fact imply
about the re-election chances of the one who has stepped on several hundred toes and brought about
maximum economic reforms in the shortest period of time—Narendra Modi? This is what the rest of
the book is all about, starting with Chapter Ten (see p. 147), which documents a necessary transition
about a maturing economy in a vibrant democracy. It is a Battle of the Elites—between the old and the
new, between the entrenched and emerging, between going with consensus, and stepping on toes—
between the bottom seventy per cent and the top thirty per cent. It is not a pretty battle for those
accustomed to privileges, and influence. Amidst all this, the person who has brought about this
transformation is facing re-election—Narendra Modi.
Chapter Nine (see p. 131) highlights what I consider to be the biggest mistake of Narendra
Modi’s policies in the last five odd years: his failure to correct the viral nature of cow politics.
However if any blame needs to be apportioned, then it should be at the doorstep of the Supreme Court
of India which had pronounced the necessity of enforcing a complete ban on cow slaughter. As is well
known, the cow is a major part of the economy when alive (milk) and dead (meat and leather), and its
appropriation by large elements of the BJP and its subsidiaries has sounded a death knell for the
economy, and led to murderous assaults on Muslims, the only community that allows a cow to be
‘touched’. In twenty-first century India, we don’t need to be reminded that the worship of a cow is a
private affair. The merging of a private affair with politics is a combustion few societies can afford.
There have been more economic reforms instituted in the last five years than in the previous
twenty or more years (which is in Indian history, given that serious economic reforms only started in
1991). Chapter Eleven (see p. 160) documents the socio-economic performance under Narendra
Modi’s tenure as prime minister.
Chapter Twelve (see p. 181) focuses on the myriad expectations for the forthcoming 2019
elections. The conventional wisdom at the time of writing this book (end-February 2019) is as
follows: it’s no longer a given that Modi will be the next prime minister; a hung parliament,
reminiscent of the 2014 election, is the popular expectation; the Congress is expected to more than
double its 2014 tally and reach three figures; and finally, the BJP’s tally of 282 seats is expected to be
sixty to a hundred seats lower i.e., in the range 180–220 seats. This chapter also provides a detailed
calculation (forecast) of the chances of an anti-Modi alliance led by the Congress and comprising a
motley group—the Samajwadi Party (SP), the BSP or Bahujan Samaj Party, the Trinamool Congress,
and the TDP— all singularly dedicated to the removal of Narendra Modi as prime minister.
Will Modi and BJP emerge with a respectable showing of 240 plus seats in 2019? A loss of only
forty odd seats, despite structural transformation, will be an enormous vote of confidence in Modi’s
favour and the reform process that he initiated. This is the major thrust of Chapter Thirteen (see p.
199).
Even as the last chapter, fourteen, will be a conclusion, let me reiterate that the success of this
book should not be entirely dependent on the accuracy of forecasting. That is my hope, but not
necessarily my expectation. Let the games begin.
As we go to press, there is important news that could affect Election 2019. On 14 February, a
terrorist attack killed forty people belonging to the Central Reserve Police Force (CRPF). In a
significant departure from past retaliations, India opted to go into Pakistan’s territory and attacked
terrorist camps. A month hence, India will go to the polls. How will the Pulwama massacre, and
India’s retaliation, in Balakot, affect voting behaviour? A difficult question to answer, given the extent
of uncertainty over what the terrorists will do next, as well as the uncertainty regarding responses of
government on both sides of the border. Our best guess is that the impact on the election is unlikely to
affect the seats of the ruling party by more than ten per cent of whatever estimate was present prior to
Pulwama.

1 Purchasing Power Parity—the now universally adopted measure of a country’s GDP in a common
currency—PPP dollars.
2 Preliminary results of this NSSO survey are likely to be made available in June 2018.
3 See Surjit S. Bhalla, Imagine There’s no Country for an extended discussion on the different
poverty lines and the relationship between the lines and absolute poverty.
4 For those wondering, educated white women very likely lied to pollsters about their preference for
Hillary Clinton.
Rotey huey aate hain sab, hansta hua jo jayega
Woh muqaddar ka Sikandar, janeman, kehlayega…
(We arrive crying into this world (at birth), but he who leaves this world laughing, he
shall be known as Sikandar [king] of his own destiny…)
– Muqqadar Ka Sikandar, 1978
What is Election 2019
All About?

No politician loves a challenge—but everyone loves a competitive challenge. Which is why, even as
an ardent life-long cricket fan, I will sometimes root for the opposition half-way through a match, just
so that adrenalin can flow (but in the end, victory should be India’s).
Elections-Predictions
Predicting Indian elections has been relatively straightforward, perhaps even easy until now. Opinion
polls originated in India in 1980 with Dr Ashok Lahiri and Dr Prannoy Roy partnering with ace
British pollster, David Butler, to produce a very accurate forecast for India Todaymagazine. Legend
has it that the forecast was within inches of the final Congress tally of 353 seats, and since this was
the very first opinion poll, the publishers developed cold feet and went with a number which was
some distance away (but still close) to 353. I believe the ‘rumour number’; understandably, the
publisher was afraid of getting the very first poll wrong.
It may be recalled that just three years earlier, Indira Gandhi had been unceremoniously dumped
by the public. The Janata Party coalition had come unstuck owing to internecine squabbles and other
issues, and Indira Gandhi had come out all guns blazing to get back into the saddle. However, there
was no way she could have repeated her 1971 victory riding on the iconic, garibi hatao slogan.
Therefore, the conservative and safe prediction of a much lower victory forecast of around 300 seats.
Something similar may be happening with opinion poll forecasts for the 2019 election—play it
safe by forecasting a hung election—there will always be time to make a new, latest, better forecast.
I recall how thirty-five years ago, the election forecast for 1984 was far easier. There was a
sympathy wave for Rajiv Gandhi in the aftermath of his mother’s assassination. Five years later, the
1989 election also started as a sure-shot victory for him. The initial polls indicated the Congress
party winning about a 100 seats less than the 1984 tally of 415 seats, but still comfortably above the
273 required for an absolute majority. Later, the opinion polls hovered around the 195–215 seat tally,
close to the eventual 197 mark. It was in that year that one could say with much confidence that
opinion polls had arrived in India, a decade after its inception.1 After that landmark election, it was
not until 2014 that any party would receive seats above the majority margin of 273.
For the 1991 election, forecasts were jeopardised by the tragic assassination of Rajiv Gandhi on
21 May in Sriperumbudur (Tamil Nadu), just a day after the first day of polling. The Congress
obtained its second lowest tally ever—234 seats versus the 154 seats it had got in the ‘Emergency
election’ of 1977. Be that as it may, but P.V. Narasimha Rao managed to cobble together a respectable
coalition and instituted bold economic reforms. Most importantly, this was the first elected Congress
government which was not ruled by a member of the Nehru-Gandhi family. Sonia Gandhi, widow of
Rajiv Gandhi, chose to stay away from the action and refrained from any involvement until she was
‘drafted’ to lead the party in 1998.
Coalition governments became the norm till 2014 and suddenly, from a lonely occupation in
1980, Ashok Lahiri and Prannoy Roy were joined by hordes of experts. There was more money to be
made and had for elections, as proven by the satta bazaar (illegal betting).

The 2014 election was indeed a turning point, but it needn’t have been. The pre-poll numbers did
indicate that the Congress was in trouble in terms of matching its 2009 seat share of 206 seats, but a
straightforward majority win was a long shot. Almost everyone forecast a hung parliament, and many
pollsters had the UPA in pole position. I remember being extra nervous on the morning of counting
day; no one likes being proven wrong, especially in front of a national audience. Moments before the
counting began on 14 May 2014, one of my co-panelists on India Today channel and I had a bet on
which party would form the government. I bet on the BJP, having published my forecast of 300 odd
seats for the party and its allies just a few days earlier in The Indian Express.
Finally, every pollster was surprised, and some of us not so pleasantly, by the 282 seats that the
BJP won on its own and primarily because of Narendra Modi. The NDA alliance exceeded all
estimates and won in 336 constituencies.
Five years later, the forecasts at the time of writing this book are the same as in January 2014. It
is a coalition fight between UPA-III and NDA-III, as it were. In my view, this is a dream forecast
election for a psephologist. It is a difficult election to forecast; but once over, it will define, and
dominate polling, politics, and economics. To get this one right is a dream—to get this one somewhat
wrong will be fine. To get it totally wrong—enough to hang up one’s boots? Possibly. I will speculate
(estimate) on what is likely to happen in the May 2019 elections and provide different scenarios. The
reason: this is too important an election to go with just one estimate, especially four months before
election time.
But first, a mandatory caveat. I learnt very early on, from none other than Ray Fair, the father of
economic voting models, the treasured secrets of successful forecasting. Just two secrets—forecast
often, and always remind people when you are right. If I am wrong, well, life goes on, but now with
an additional barrage of kind trolls (I realise that is an oxymoron, but emphasis is needed). If right,
well, we will just have to wait till the next election to go wrong.
The caveats out of the way, let us get down to the business of analysis, and forecasting. This
election is a fight between the Mahagathbandhan and Modi. The Opposition parties, led by the
Congress, want to stop, from their vantage point, ‘dictator Modi’. Towards this end, they have
gathered together, enemy vs. enemy, united to achieve a common goal.
As usual, the people will decide. But for various reasons, detailed elsewhere in this book, but
primarily because of the near unparalleled success of Narendra Modi as a politician, the polity is
polarised. But one may ask: isn’t polarisation almost always the case in any democratic election,
anywhere in the world? It is, so what is so polarising about the 2019 election? Nothing special,
except for the first time ever, the well-ensconced ‘establishment’, the old elite, finds itself on the
wrong foot. Prior to the 2014 election, this group was confident that Modi would not win on his own,
and even if he were to become prime minister, it would only be for a short duration. We all remember
how V.P. Singh lasted for a mere eleven months post the 1989 election. Many expected a similar fate
for Narendra Modi at the head of a minority coalition.
However, it was as if history awaited Narendra Damodardas Modi, the chaiwalla. It may be
recalled that during the 2007 Gujarat election, the Congress had disparagingly described him as a
maut ka saudagar, a messenger of death. That is how derisive the old elite was prior to 2014.
Coming back to our basic question, in the final analysis, the 2019 election will be broadly
decided by four issues, and how the polity responds to them.
Modi vs. Mahagathbandhan
It is Modi vs. the Rest, and the latter realise that they better hang together (or they will hang
separately, state by state). Hence, the talk for the last year, and especially after the 2018 state
elections, towards the formation of a Mahagathbandhan, a Very Big Alliance. As we shall discuss in
the next chapter, one of the prime attributes of India is its diversity, second only to its ability to argue,
and disagree. The reason coalition governments have succeeded in India, after the near mandatory
initial monopoly of a founding party (also known as the first mover advantage), is because we come
armed with different points of view. The reason the Index of Opposition Unity concept was invented
in India by Lahiri–Roy was because the political tradition, perfected by the Congress, was to divide
and rule. And, of course, the Congress learnt it from those who wrote the original book on the
creation of empires.
Is 2019 ‘the’ Presidential Election?
From the BJP’s point of view, the election will be converted into a presidential election. While India
is a parliamentary democracy, it is also true that most elections around the world have willy-nilly
become presidential elections. Therefore, the BJP doesn’t need to create an ‘aura’ of a presidential
election—the reality is out there, and in my view, has been forever.
And that is the most basic flaw in the 2019 Opposition model. They don’t have a leader; phrased
equivalently, they have many wannabe leaders. While the old establishment will hem and haw about
how India is diverse, comprises many states, is a parliamentary democracy and not a presidential
system, the fact is that most electoral contests in India have been conducted presidential style.
Therefore, the Opposition has to decide on who to choose as a leader, which they haven’t at the
time of writing this book. Let me be upfront here: not that there aren’t any—the problem is that there
are far too many. Will it be the dynast, Rahul Gandhi, fighting for what he thinks is his due place in
history? Will it be the original street rebel, Mamata Banerjee, chief minister of the large (forty-two
out of 543 seats) state of West Bengal? She was a member of the Congress party once, but later built
her career by taking them on. Indeed, she was inducted as Minister without Portfolio by Atal Bihari
Vajpayee in 2001, and she joined the UPA government in 2009 as Minister for Railways, before
proceeding to trump the Communists, and the Congress, in subsequent state elections.
The other contender is the former chief minister of Uttar Pradesh, a former member of
Vajpayee’s Cabinet, and leader of the Bahujan Samaj Party, Mayawati. As is well known, UP is a big
state, and one which sends eighty members to parliament. You win the state, you rule. But Mayawati
has had to team up with her sworn enemy, Akhilesh Yadav of the Samajwadi Party. Truth be told, it is
not Akhilesh who is Mayawati’s sworn enemy, but his father, Mulayam Singh Yadav. Members of the
senior Yadav’s team imprisoned (and assaulted) Mayawati back in 1996, three years after her party
then led by Kanshi Ram, and Mulayam had teamed up to fight the 1993 state election.
The problem with the Mahagathbandhan is that there is only one possible leader, Rahul Gandhi,
but the regional leaders, for reasons unknown (or known), are not able to unite behind him.
Objectively, the party which obtains the largest number of seats should head the coalition, and elect
the prime minister; no matter what the calculation, or who the calculator, the Congress party is near
certain to be the single largest Opposition party. In Indian politics, there are no certain certainties,
and Chapter Twelve (see p. 181) evaluates and forecasts the number of seats that the Congress is
likely to obtain in 2019.
This is the dilemma facing the Opposition—to unite, or not unite, behind Rahul Gandhi.
Let us now turn to a question which has been raised ad nauseam. Does Rahul have what it takes
to be prime minister of India? His presence in active politics is more than a decade old, and I have
always thought of him as the reluctant ‘prince charming’. Indeed, I had once described the Congress’
dilemma as follows—the leader wants to be a rock star, but his mother wants him to be a dentist.
But that was the Rahul Gandhi of yore. Today, he is combative, and seems comfortable doing his
job. Which brings us to the basic premise: is the Presidential election becoming more even? One gets
a taste of it in Chapter Twelve, in which I deep dive to figure if an even contest has indeed become
accepted as reality for 2019.
Professionals vs. Amateurs
For the last year or so, the political agenda in the country has been dictated by the Congress party.
Whether the discussion revolved around the perceived lack of jobs, or the welfare of the poor,
especially poor farmers, it is the Opposition’s voice that has been the loudest. A major reason for this
anomalous development is that the election campaign for the last few years has been a contest
between professionals and the amateurs. It is a well-known secret that several senior bureaucrats,
journalists, and industrialists have at one time or another supped at the Congress’ table. The grand old
party ruled India for most of the last seventy years, and we all know that experience does count in
politics. In the opposite corner are the amateurs, people who gained their status by winning the old-
fashioned way—they earned their victory. The BJP, for instance, won because of the promise of
development, and people’s trust in Modi as someone who could bring about change in the system. But
then, they lacked experience in handling the highly influential English media; whatever handling they
did, it was not very effective.
This failure to manage the elite, let alone win them over, was acknowledged by Prime Minister
Modi himself. In an interview to ANI on the first day of 2019, when asked about his one failure, Modi
stated:
But one thing, I am not surprised, I could neither make the Lutyen’s world part of me or me a
part of them. I did not want them to be a part of me as my background is different. I am a
representative of the non-elite world. I could not win them over, still trying how to win over
such forces.
Therefore, it is safe to surmise that management of the old elite, which effectively controls the media,
especially the English media, goes to the Opposition.
However, mismanagement of the elite was not the only mistake that the Modi government made,
or its lone failure. Several mistakes are detailed elsewhere in this book; in particular, and as
mentioned in the Introduction to the book, the government erred in not forcefully stopping the
gaurakshaor cow protection movement (see p. 131, Chapter Nine for details). This must easily rank
as the biggest failure of the Modi government.
The Bottom 70 per cent vs. the Top 30 per cent
You want to call it a class struggle, you can. But it isn’t. It is about the nature of development under
Narendra Modi. It is correct to state that the last five years have seen much greater improvement in
the lives of the ‘have-nots’. The traditional definition of the have-nots was the absolute poor; my
definition of those who have really benefitted from the Modi reforms are the bottom seventy per cent.
It is equally true that the top thirty per cent have relatively lost out. By relatively I mean that the rate
of welfare growth for the bottom seventy per cent has been greater than ever experienced before; and
that while the top thirty per cent have increased their incomes, it is lower than what they experienced
in UPA-I and UPA-II, and definitely lower than their expectations.
The data on this 70–30 divide is quite compelling. Let us examine what has happened to
personal income tax payments, primarily because of demonetisation. Tax compliance has increased
manifold with the number of income tax payers increasing from a level of forty million in fiscal year
2013–2014 to seventy plus million (projected) in 2018–2019. More important than the number of new
tax payers in the system is the increase in tax compliance. One direct measure of tax compliance is to
look at the percentage gain in tax revenue and compare it to the percentage gain in nominal incomes.
In the steady state, this ratio is close to unity i.e., if incomes go up by ten per cent, and tax rates stay
the same, then tax revenue should increase by ten per cent. But what happened in both 2016–2017 and
2017–2018 is that personal income tax revenues increased at twice the rate of nominal incomes.
Somebody was being ‘forced’ to pay more tax, and it wasn’t the salaried tax payer whose tax was
deducted at source. Salaried tax payers, most of whom reside in the 70 to 80 per cent bracket of
individual incomes, cannot avoid tax because it is deducted at source.2 Then, who is paying more tax
per unit of income? The shopkeepers, doctors, the lawyers, the accountants, and the self-employed:
same people who were avoiding tax before.
I have an easy litmus test for identifying the person who is now against Modi (but voted for him
in 2014). He is paying more tax. He is against demonetisation. The reason: because his net black
income has been reduced by demonetisation, which did hurt the old rich elite the most.
The bottom seventy per cent has seen large increases in income, and seen that across-the-board.
Inflation is in low single digits (around three per cent) and low inflation primarily helps the bottom
seventy per cent. One of Prime Minister Manmohan Singh’s major beliefs was that India should strive
for low inflation, because it primarily helps the poor. He was right.
On the social side, the bottom seventy per cent has witnessed large improvements. Especially
the women, which is good news. Indeed, targeting the woman voter appears to have been a conscious
strategy of the BJP, and amongst Modi’s first initiatives was the announcement of Beti Bachao, Beti
Padhao (Save the daughter, educate the daughter) policy. In the recent political history of India, no
prime minister had ever dared to speak publicly about the despicable practice of killing a girl-child
before birth. Prior to the advent of technology in the mid-1980s, Indians by and large controlled the
sex-ratio (number of females per male) by deliberate neglect of the girl-child; hence, she would die
early and the parents could indulge in their sex preference attitudes by attempting to have a male-
child. Post the mid-1980s, amniocentesis (the sex identification of a child before birth) allowed
parents to indulge in this criminal practice, by aborting a female foetus.
The second part of the policy—Beti Padhao—refers to the recognition that a girl-child needs to
obtain education as a result of which her standing in society is enhanced, and there is little or perhaps
no reason left to either neglect her, or to abort her before her birth. It is rather surprising that the so-
called Left-oriented ‘make everything a right’ regime of Sonia Gandhi (and her National Advisory
Council or NAC) did precious little in terms of policy initiatives to help the girl-child.
The second big gender equality initiative of the Modi government was the Swachh Bharat or
Clean India campaign. Again, the genteel old India circles never bothered to look at the simple reality
of the bottom seventy per cent. Open defecation is a disease-habit, and one which primarily affects
the poor of India, and particularly the women. About fifty per cent of India was indulging in this
practice, and while seasoned economists lamented the fact that stunting and malnutrition was higher in
India than in much poorer sub-Saharan Africa, they ignored this most important cause.
In the annals of Indian policy measures, Swachh Bharat is easily the most ambitious, the most
necessary, and the most successful social reform movement. Whether open defecation is now eighty
per cent or ninety per cent (and the old elite is happily providing statistics that India is not as open
defecation free as claimed by the government), the fact remains that this is the fastest improvement
ever witnessed in India on an issue which involves the betterment of crores of people. We must all
learn to give credit where due—particularly, the old elite must learn.
However, it is not as if all is well for the bottom seventy per cent. There is farmer distress,
caused not by too little produce to sell, but too much produce. Food prices are low, and farmers have
not experienced large income gains. This needs new policy responses, and responses do not include
increasing the Minimum Support Price or MSP for farmers. Transformation of agriculture will not
occur by doing more of a bad policy. India needs a targeted basic income policy for the bottom third
of the population. In the 2019–2020 Interim Budget, announced on 1 February 2019, the Modi
government announced what maybe the beginning of a large scale targeted income policy for the
bottom half, or bottom third, of the population.3
Fake News: Big Time
One of the truly big innovations in the 2019 election campaign is the use of fake news. Given the
reality mentioned earlier about the economic progress and especially concerning the lives of the
bottom seventy per cent, what is a challenger to do? Say that it isn’t so—which is precisely fake
news. Hence, no matter what the initiative, no matter what the policy, the Opposition’s argument is
that it isn’t so. GDP growth data—cooked; data on the number of toilets—invented; increase in tax
revenues, same complaint. When cash came back to the levels before demonetisation, fake news
peddlers said: ‘Look, we told you so—demonetisation was a failure.’
The fake news campaign is important enough—for politics, elections, and intellectual honesty –
to devote one entire chapter in this book, and more.
What Does It All Add Up To?
Back to the original question. What is election 2019 all about, and who will win? It is about the old
and new elite; about fake news and the underlying reality; about the transformation that Modi and his
government has affected. Change is never easy, especially rapid change. Moreover, slow change often
results in no change, because vested interests have enough time to rally their forces.
There is no doubt that a fresh new development that India witnessed in recent history was the
chaiwalla taking charge. If Modi wins, and I think that as of now it is not an unlikely prospect, it will
be because of the support of the bottom seventy per cent of the population, for whom he has done
more than any other political leader. As discussed above, Modi has won some of the initial ‘rounds’,
but also lost some.
Chapters Twelve and Thirteen are devoted to a forecast of seats that the Congress and BJP can
expect in 2019. There is uncertainty—hence there is more than one forecast. But we do not
completely hedge—there is a forecast we consider as firm and as final as circumstances permit. As
we said in the Introduction, Pulwama can add (or subtract) up to ten per cent seats to our final
forecast. If there are any major changes, I shall communicate those via my website, ssbhalla.org.

1 But it was not an easy convergence (between opinion polls and reality); see Chapter Seven for a
detailed discussion on opinion poll forecasts, and what may have been the first use of identifying true
intentions through the development of a lying index.
2 An individual is only required to file taxes if she has income above Rs. 2.5 lacs. About 70 per cent
of the worker population has incomes below Rs. 2.5 lacs—hence, the nomenclature of the 70–30
divide.
3 See my joint paper with Karan Bhasin—‘Targeted Basic Income Policy’ for a detailed discussion
on the subject.
Na tum hamein jano, na hum tumhein janein
Magar lagta hai kuch aisa, mera humdum mil gaya…
(You don’t know me, neither do I
But it seems like I have found my soul-mate; my lover…)
– Baat Ek Raat Ki, 1962
Democracy Begins

As I mentioned in the Introduction, this book is essentially about democracy, and development, and
elections are an integral part of democracy. A legitimate question asked by many scholars since the
start of Indian democracy in 1952, when India held its first general election, was: how did India end
up being a democratic State? No one gave India a chance to succeed as a democracy; that it did may
be more than just luck, or prescient foresight on the part of its founding fathers.
In her book, On How India Became Democratic, Dr Ornit Shani states that, ‘theorists of
democracy have conventionally seen the establishment of India’s democracy as a product of
elite(emphasis mine) decision-making and institutional design. In this view, popular democracy, and
the constitution, were endowed from above (emphasis mine) by discerning nationalist leaders and
intellectuals.’1
Note the words emphasised in the quote—the elite and democratisation as an institution imposed
from above. Along the same lines is the assertion by many that it was because of liberalism and the
liberal (and this is in the nineteenth century sense of a liberal) nature of India’s leaders, especially its
first Prime Minister Jawaharlal Nehru, that India embarked upon its possibly premature experiment
with democracy. Add to this the widely held popular view at the time of independence, and even now,
that India’s adoption of democracy was a natural progression of the British Raj. The same institutions
stayed, only the headmasters changed from white to brown.
The above are some of the reasons why people expected India to be a democracy. There were
equally other compelling reasons to reckon that even if India adopted a democratic framework, it
would not remain one for long. There were two reasons put forth for this postulation—first, India was
an illiterate, developing country and hence not ready for democratic governance; second, that it was
too diverse, and multi-ethnic to survive as a democratic nation. In addition, at the time of
independence, there was no known example of a poor country being democratic.
I want to present an alternate explanation for why India became democratic, something I first
explored in my manuscript on the middle class published in 2007, Second Among Equals—The
Middle Class Kingdoms of China and India,2 and later in my book, The New Wealth of
Nations(2017). The explanation combines the reasons for India being a democracy, and of not staying
one, to offer a view that incorporates both.
Colonisation and Democracy
Now, let us get back to the basic question—why is India a democracy and why has it remained so?
Firstly, India became a democracy because of its ‘heritage’. This inheritance consisted of two
aspects: it was not only a British colony, but also ethnically and culturally diverse. The former
guaranteed the presence of democratic institutions and the latter hinted that there were very few
alternatives.
Democracy is not only the only form of government that guarantees minority rights, but it also
guarantees important roles for different ethnic and cultural groups.
While casual evidence also supports this conclusion, the above-mentioned two results are based
on empirical research. Upon gaining independence, almost all British colonies adopted a democratic
system. Whereas there are no examples of a German, Portuguese, or a French colony ever becoming
democratic. The independence wars in Algeria and Vietnam are recent reminders that non-British
colonisers were most uncomfortable with democracy; in direct contrast, the British were aligned to it.
The second part of the democracy ‘heritage’ was India’s diversity. The first being colonised by
the British. There is considerable empirical support for the British heritage (and diversity)
hypothesis.3
Relating a popular democracy index to diversity and colonisation, one obtains the following
result:4
Democracy = -2.58 + 1.31*ethnic + 2.14*UK colony
For 1960, the predicted probability for India being a democracy was seventy-three per cent, when the
highest probability for democracy for South Africa was at seventy-six per cent. Indeed, India has the
fourth highest predicted probability. The results suggest that Indian democracy is not such a great
surprise, and especially, that the Indian democracy experiment, believed by many to be sui generis,
was just not so.
Furthermore, India may have succeeded as a democracy because it was the onlypolitical system
compatible with a heterogeneous population. Most analysts focused on India’s poverty and illiteracy
in 1950, not expecting India to be democratic, and not fully appreciating that only democracy can
keep everyone the least unhappy. A democratic process gives, at least in theory, every group and each
individual a chance to participate in decision-making. A small chance, one might say, but an infinitely
higher chance than if the system was non-democratic, either as a monarchy or Communist
dictatorship, and all possibilities in between. It is important to appreciate the existence of these small
probabilities. The fact that they exist is the glue for solidifying expectations and for perpetuating
democracy.
The logic of Indian democracy can therefore be summarised as follows: inheritance of British
institutions meant a strong, positive, and an initial proclivity towards democracy. The vote
empowerment of different social, cultural and religious groups meant that each group, especially the
small ones, had a strong stake in democracy. A correlate of this empowerment was the desire among
all groups for a united India; for only in a united India would each non-majority group have a stake.
Logically therefore, democracy was most likely the preferred choice among most sections of society.
For instance, the Indian subcontinent has several democracies, each with its own pattern. The
two most ethnically diverse economies in the region are, India and Sri Lanka. Both have stayed the
course (albeit with brief forays into dictatorships). The two most ethnically homogeneous economies
in the region are, Pakistan and Bangladesh; both have not stayed consistently democratic. Add
Myanmar (ethnically homogeneous) and Nepal to the mix and the democracy-diversity story is
complete.
Does Democracy Hurt Growth?
Democracy was, and is, politically correct. But is it economically correct? If elections are broadly
about economic performance, then has democracy served the people well? A popular myth in India is
that democracy hurts economic growth. This is often cited by experts as the reason why India has
failed to equal China’s growth after matching it for 480 years between 1500 and 1980. Consistent
with this interpretation is the observation that East Asian economies, the ones which have done better
than India, were all dictatorships during their period of out-performance.
The facts do not support this apology (for India’s under-performance). The fault is not
democracy, it is us, and the politicians we elect, and the economic policies the elected
representatives pursue. While we shall examine economic performance during different election
periods, what we need to establish at present is whether democracy has helped, or hindered,
economic development in different countries.
Table 2.1 documents the growth experience for developing countries for two periods, 1950 to
1979, and 1979 to present. Prior to 1980, it shows that non-democracies grew faster; post 1980, the
opposite is true. Further, average democratic growth is almost two percentage points (ppt) higher than
average non-democratic growth—3.8 per cent per annum compared to only two for the non-
democrats. The broad averages confirm that there is little substance to the hypothesis that democracy
is a hindrance to economic growth.

Table 2.1: Does democracy hurt growth?

Pre Globalisation Post Globalisation

(1950–1979) (1979–2012)
Democracy Democracy
Region No Yes Average No Yes Average
(in % per year) (in % per year)
East Asia 3.2 2.9 3.2 3.6 3.7 3.6
Russia & EE 4.2 2.7 4 1 0.4 0.8
Latin America 3.1 2.9 3.1 1.2 1.7 1.3
MENA 2.6 1.5 2.6 2.3 1.9 2.3
South Asia 2.4 1.3 1.4 2.2 4.2 4
Sub-Saharan Africa 0.8 2 1.1 0.9 1.3 1
Average 2.9 1.4 2.2 2 3.8 2.8

Source: World Bank, World Development Indicators, IMF, World Economic Outlook and Polity IV.
Notes: 1) MENA contains the countries of the Middle East and North Africa. EE stands for the countries of Eastern Europe.
2) Democracy as measured in Polity IV data. The value ranges from 0 to 10; a country is defined as democratic if it obtains a
value greater than 5.
3) Cells contain weighted averages of per capita income growth.
4) The data above does not include China.
Democracy: The Constitution
Democracy and elections are not perfect solutions to the organisation of a society—but they are still
better than all other alternatives. Democracy also requires a Constitution—rules about organisation,
and custom, and behaviour. Several centuries ago, rules of behaviour were either prescribed by kings
or religious leaders, and we all know how that ended up—by yielding to democracy.
It is therefore our Constitution that authorises our freedom, and our elections. In 2019, close to a
billion people will be eligible to vote. In many ways, and despite several amendments, the
Constitution is indeed a proud achievement. But I want to strike a slightly discordant note.
Our Constitution, laudable as it is, emphasises the rights of the State rather than the rights of
individuals. That is the major reason it is so bulky (the longest written Constitution) and so confusing
—for the citizen, the politicians, and even perhaps the Supreme Court.
Has the Constitution served India well? The ledger, according to many, is heavily positive; the
answer, an overwhelming yes. But we need to ask—what is it in the Constitution that makes it
susceptible to frequent change? (124 primary amendments to date—and a pace of replacement faster
than the disposable toothbrush.)
Is the problem the Constitution, or the interpretation of the Constitution? Many a time, Supreme
Court orders seem to have a particular fondness for opaque decisions—for example, what is the
constitutional origin of the need to play the national anthem in movie theatres, and before every show,
no matter how A-rated the movie might be or how objectionable its contents? If it is not in the
Constitution, the parliament needs to pass a law. No law (at least to my knowledge) has been passed
by the parliament on the playing of the anthem in movie theatres. Maybe it is a tradition; all traditions
are not enshrined as law, nor should they be.
Indian Constitution—Longest and Most Amended
As mentioned earlier, at the time of independence, literacy levels in India were frighteningly low. But
the writers of our Constitution were learned, scholarly, and were desirous of creating a fair and just
society. The Constitution writers also recognised the limits to social engineering. Hence, they
included a section on Directive Principles—theories that did not have the weight of law, but were
suggestions for the direction that the future laws should take. I am not a lawyer, but why include in the
Constitution thoughts about what the government should do in the future. If so important, why not then
(in 1950 itself)? One of the modern, progressive ‘thoughts’ in the Directive Principles—ban on cow
slaughter.
The somewhat in-built, and extreme prejudices in the Indian caste system made it mandatory for
social and political leaders to provide special consideration in the Constitution for those at the very
bottom of the Indian caste system, the Scheduled Castes (SC) and Scheduled Tribes (ST). It is
important to note that in the Constituent Assembly debates, the idea of using an economic criteria to
help the disadvantaged was explicitly discussed—and rejected by two most important thinkers and
writers of the Constitution—Jawaharlal Nehru and Dr B.R. Ambedkar.
The disadvantaged were to be identified by caste; the policy remedy: quotas for education and
jobs. But there were other methods—affirmative action rather than reservations (quota)—for
providing for the downtrodden. The building of schools, colleges, health centres, provision of
scholarships, transfer of ‘basic income’ are just some of the measures the Indian State could have
taken to right historical wrongs.
Before proceeding with the analysis of reservation policy, an important, but not much
acknowledged fact needs to be highlighted. There is no explicit policy, no explicit statement, about
job or education reservations in the Constitution. The only explicit reservation policy contained in the
Constitution is for seats in the legislature i.e., reserved constituencies for SCs and STs. Even this
reservation was for ten years—through amendments, it has been renewed every ten years, and the next
amendment is due in 2020.
Reservation Policy on Education and Jobs
The time has come to examine the logic, legality and practice of education and job reservations for all
castes and classes. The reason, amongst others, is that there is a lot of confusion, some of it deliberate
on the part of the interest groups, about what the Constitution does, or does not say, about
reservations. There is also considerable doubt about the difference between a policy of
‘reservations’and a policy of ‘affirmative action’; additionally, much confusion about eligibility.
Economic Criteria for Reservations
In the last session of parliament (December 2018) and barely a few months away from the general
elections, the Modi government introduced a major new initiative to help the bottom seventy per cent
—reservations for the economically backward. More precisely, a ten per cent quota in education and
jobs for Indians not presently covered by any law pertaining to reservations; a quota for the
Economically Weaker Sections (EWS) of society.
It is safe to surmise that this is the first law in India that explicitly benefits the economically
underprivileged, and does so neither on a caste nor on the basis of religion.
This is in a series of ‘bombs’ (amendments) on reservations in the last seventy years.
Amendments that have had political and electoral consequences. The first reservation bomb exploded
with the Constitution, when after considerable debate in the Constituent Assembly, it eliminated the
use of an economic criteria for special favourable discrimination on the part of the State. A good
beginning was aborted before we even had a Constitution to guide us. This was possible (aborting the
much-needed economic criteria) due to the active support of both Jawaharlal Nehru and Dr B.R.
Ambedkar. Incidentally, the EWS recommendation brings back to policy the original spirit of the
Constitution.
The second problematic rule was within the Constitution itself when it talked about education
and job reservations. Initially, the Constitution introduced it for only the SCs and STs; the politicians
thought about it some more and extended the benefits to OBCs (Other Backward Classes). In practice,
the OBCs are mostly Hindus, even though more than fifty per cent of the Muslim population also
belong to this category.
There is a lot of confusion regarding OBCs, and some of it may have been deliberate. The 1931
Census measured caste proportions, and Hindus who were not SC, STs, or upper caste, were
classified as OBC or Other Backward Caste. The Mandal Commission uses OBC as Other Backward
Class. So while a Muslim can be an OBClass, she cannot be OBCaste, since having a caste means
that you are a Hindu. This confusion is being tested in our courts. Query: are OBClass Muslims
eligible for OBCaste benefits a la Mandal.
It was V.P. Singh who, during his brief tenure as prime minister between 1989 and 1990,
introduced the (Mandal) OBC amendment. The proposal was to increase the quota system for those
who were poor, but not SCs or STs. In 1990, the law extending benefits to the OBCs was passed, but
it did not take long for the electorate to recognise that free benefits could be had just from the change
of one’s caste to the category of ‘Other Backward Class’. In the end, as I view it, there is no pride
lost in ‘converting’, and a lot of freebies received, especially admission to prestigious government
institutes of learning, and then secure plum jobs in government (both state and Central). Therefore,
much like grade-inflation, India had caste-inflation.
Let me document this inflation. The percentage of population that was classified as OBCaste in
1999–2000 was thirty-six per cent. Let us examine how accurate was this National Sample Survey
Office (NSSO) estimate? There are two reasons why this estimate is about right—first, this
approximation of OBCaste in 1999–2000 is very close to the thirty-one per cent estimate obtained
during the 1931 census. The difference between thirty-one per cent (in 1931) and thirty-six per cent
(in 2000) can easily be explained by Partition and the migration of OBCaste Hindus from across the
border.
However, the NSSO survey five years later found that the OBCaste numbers in the national
population had increased to forty-one per cent. In a span of five years, 1999–2000 to 2004–2005, the
OBCaste population grew at an annual rate of 4.4 per cent.
The point to ponder over is whether an OBCaste population growth rate of 4.4 per cent a year is
plausible? It is not, because even for the Hutterites (who do not practice any birth control and have an
average fertility rate of nine children per woman), the population growth rate has never exceeded four
per cent a year. It is biologically not possible, not even for the most fertile, and hence it is obviously,
caste-inflation. As a consequence, in order to obtain the irrational government-mandated manna, the
non-OBCs simply changed their caste. Gradually, more and more castes, economically deserving or
not, joined the over-arching OBC category.
The Minority Reach of the Ten per cent Quota
The EWS ten per cent quota will help the Muslim minority catch up in education. It requires no
special insights, but due to the overall lack of good primary and secondary education for the non-
elite, the dominantly poor (SC, ST and Muslim) students had little access to good education.
However, because of the special reservation quotas, the SCs and STs expanded their education base
(and were also able to get government jobs). That Muslims fell backward because of the non-
availability of reservations as shown clearly by the data (Table 2.2).
In 1983, an average SC/ST youth had 2.3 years of educational attainment, and an average
Muslim youth (between the ages eight and twenty-four) 2.9 years; in 2011–2012, an average SC/ST
youth had 6.4 years of education, and a Muslim youth had the least educational attainment in the
country at 6.2 years.
It has been argued erroneously that Muslims ‘culturally’ do not have as great a demand for
education as the Hindus, and therefore their rate of progress has been lower. But perusal of data for
other Muslim countries suggests that this is not the case. Every (predominantly) Muslim country has a
higher rate of improvement than India, and three Muslim countries (Indonesia, Malaysia and
Bangladesh) have had exceptional rates of improvement, with Bangladesh being the real outlier
(according to the international data compiled by economists Barro and Lee). In Pakistan, there has
been an eighty-four per cent increase in the average level of education, almost twice the rate of
increase of an average Indian Muslim.
The net result is that, all things considered, today there is only one community deserving of any
expansion in affirmative action in education—the Muslims. There are no data to substantiate that the
relatively privileged OBCastes should get any additional benefits; indeed, the data suggest that the
Indian government should at least make an all-out effort to bring the Muslims up to par with the
SC/STs before the passage of any legislation for any other community. Further, as the Sachar
Committee Report (2005–2006) had documented, Muslims in government jobs are about a third
lower than their population levels.

Table 2.2: India: Youth educational attainment, 1983–2011/12

Social category Average years of schooling

1983 1993/94 1999/00 2004/05 2007/08 2009/10 2011/12
Dis-privileged 2.5 3.4 3.9 5.4 5.5 6 6.4
– SC 2.5 3.4 4 5.5 5.7 6.1 6.6
– ST 2 3 3.6 4.9 5.3 5.8 6.1
– SCST 2.3 3.3 3.9 5.3 5.6 6 6.4
– Muslims 2.9 3.7 4 5.4 5.4 5.9 6.2
Privileged 4.3 5.2 5.5 6.9 6.8 7.2 7.8
 All groups 3.6 4.5 4.8 6.3 6.3 6.7 7.1

Source: NSSO data, various years, author’s computations

Notes: 1) Youth is defined as those between 8 and 24 years.
2) Dis-privileged defined as SC, ST, plus Muslims.

What deserves emphasis is that the ten per cent quota for the Economically Weaker Sections of the
population (which is around thirty per cent of the total population) is the first reservation system
initiated for those who are currently not covered by government support. Therefore, it will be the first
time that poor non-OBC and non-SC/ST individuals will get a chance, and given that Muslims are the
poorest (economically weakest), they should obtain preference in the ten per cent quota amongst the
EWS.
As is now well known, considering it has been discussed threadbare at several TV debates, the
eligibility criterion for this class has been fixed at eight lakh rupees per family. To many, this seems a
bit high, and it is objectively quite high. Also, much has been made of the income surveys conducted
by the National Council of Applied Economic Research or NCAER (the Indian Human Development
Survey or IHDS), and wage surveys conducted by NSSO, that the number of eligible households with
eight lakhs as income is a whopping ninety-nine per cent. But these surveys are able to capture only
25-30 per cent of personal incomes. In reality, the eight lakh limit makes only twenty per cent of
households ineligible.
But where did the Modi government obtain the guidance for fixing this limit? Very likely from
limits which were fixed for the OBCs by different governments since 1993. The OBC creamy layer in
1993 was defined at one lakh rupees per family. A decade later in 2013, this limit was raised to six
lakhs (there were interim and intermittent increases as follows: in 2004, 2.5 lakhs; in 2008, 4.5
lakhs). Note that all these increases were instituted by the Congress party. During this period, the total
increase in limit was 700 per cent; and the increase in Consumer Price Index or CPI was 413 per
cent. In 2017, the NDA government increased the limit to eight lakhs, an increase of thirty-three per
cent with a CPI increase of twenty-one per cent.
That the old elite may have discriminated against minorities is indicated by their lack of
response to the exorbitant OBC creamy layer income expansion undertaken by the Congress
government in 1993 and the UPA government in 2013. Now that the NDA government has instituted
the change, critics are crying foul. Is it because some ‘poor’ Muslim, Sikh and Christians will now
get the benefit that was almost the exclusive preserve of the OBCaste Hindus before?
It needs to be reiterated that one of the major innovations in the EWS initiative is that now
Muslims and other religious minorities (mostly Christian) will also benefit from the social policies
i.e., quotas and reservations. An additional benefit, which isn’t emphasised by most, is that the doors
are open for a real debate, a genuine reconsideration of the skewed social welfare policies of the last
seventy years.
From the perspective of the impending 2019 elections, the EWS quota policy is likely to turn out
positive for the BJP. There is an added twist—by including a clutch of religious minorities, Modi
might have done something for the equitable future of the country.
However as I said at the beginning of this chapter, reservation policy should have been in the
first place based entirely on an economic criteria—as per the original design of those who wrote the
Constitution. An economic-based criteria would have been consistent with the liberal ethics of both
Dr Ambedkar and Pandit Nehru. It is very likely that the two learned men realised that they were
introducing an economic and moral wrong—in the first flush of independence, which had come after a
long drawn-out struggle, they may have omitted to foresee the intended and unintended consequences
seventy years hence. But that in no way means that no subsequent government, however far removed
in years and ideology, should not take a serious re-look at this important aspect of social and
economic policy.

1 Cambridge University Press, 2017.

2 Submitted to the Peterson Institute for International Economics, Washington, D.C., May 2017. This
manuscript was the first document to define a global measure of the middle class, and to present
estimates of the middle class in over 180 countries of the world for the period 1850–2030.
3 See Surjit S. Bhalla, India: Democracy, Growth and Development 1951—2014, paper prepared
for the Democracy Consensus, 2014, p. 8, available at democracy.cde.org.za
4 Statistical details: Probit regression, democracy index is coded as Yes No for 76 countries in 1960;
pseudo-R2 = 0.4205, and being a UK colony is most statistically significant (t-statistic of 4.8); the
(Fearon) diversity index is significant at the 10 % level of confidence (t-statistic of 1.5).
Chhodo kal ki batein, kal ki baat purani
Naye daur mein likhenge mil kar nayi kahani
Hum Hindustani…
(Forget the past, it’s old and best forgotten
Let us write a new chapter in a new era
We, the Indians…)
– Hum Hindustani, 1960
The Importance of Nehru

Citizen Raj was ushered in under the leadership of Jawaharlal Nehru in 1952. Owing to his
indisputable stature, what Nehru emphasised became policy, and law. However, our concern in this
chapter is with the important role that he played in setting the tone and content of India’s economic
policy. As must be obvious by now, this book considers economic policy and its outcomes as the most
important feature of a democracy. Political rights of citizens are sacrosanct; but no human being can
live by rights alone.
Except for two years when Lal Bahadur Shastri was prime minister, and the three years of Janata
Party rule, the first forty-two years of independent India were ruled by members of the Nehru-Gandhi
family. These were formative years for a young nation, and it is important to understand the economic
legacy of Nehru. The reason: what the dynasty thinks, and does, affects the lives of Congress
supporters, and millions of other Indians. That is a given; and it is important for us to understand the
dynamics of this ‘entitlement’. For that, we need to start at the very beginning.
The Beginning
Jawaharlal Nehru became India’s prime minister on 15 August 1947; in October 1949, India had a
written Constitution; and three years later, the first general election was held in 1952. Nehru was
prime minister for seventeen years and won three successive elections, each with more than forty-five
per cent vote share and close to seventy-five per cent seats.
These initial Nehru (and Congress) wins should not be over-interpreted. Notwithstanding his
untimely death in May 1964, the Congress party continued to rule till the loss in the sixth general
election held in 1977.
History suggests that this was slightly in excess of what was happening around the world. The
Federalist Party in the US (George Washington’s party) ruled from 1789 till their loss to Jefferson’s
Democrat-Republican (D-R) Party in 1800 (the fourth election). The D-R party, in turn, ruled for
seven consecutive terms. Closer to home, elections in Sri Lanka alternated between the Bandaranaike
family (husband, wife, and daughter) and the United National Party or UNP. Pakistan was also ruled
by the Bhuttos, father, daughter, son-in-law, and now grandson—but they were nowhere as successful
as the Nehru-Gandhis (father, daughter, youngest grandson, elder grandson, grand daughter-in-law,
and now great grand-son). While the PRI (Party of the Mexican Revolution; now called the
Institutional Revolution Party) ruled Mexico for decades, it was not so when Mexico had national
elections; the first general election in Mexico was won by PRI’s Ernesto Zedillo in 1994, the second
in 2000, by Vicente Fox, who was an independent candidate.
The Nehru-Gandhi dynasty is exceptional and its influence over a large population is therefore
without any parallel in contemporary political history. Which is why it is important to understand the
mindset, the thinking, and the ideology of the patriarch, Jawaharlal Nehru.
Nehru’s Economic Vision—Was He a Socialist?
A constant debate amongst Indian economists is of about the significance of the role Nehru played as
‘Founding Father’ and ‘First Citizen’ after Mahatma Gandhi. That Nehru was a tall leader in the realm
of human liberties is beyond much doubt. That he was an influential international statesman is widely
accepted. However, it is also true that fifteen years into his prime ministership, his foreign policy
began to be questioned after the loss of the 1962 war with China.
At the time of India’s independence, the world was emerging out of the double whammy of
Depression and the Second World War. Meanwhile Nehru, who was in the midst of India’s freedom
struggle, was also influenced by the goings-on in the erstwhile USSR and China. Not only had Russia
ostensibly defied Western imperialism and grown rapidly, it had also adopted economic
totalitarianism as its weapon of success. On the other hand, China had led a ‘poor-people’s’
revolution. The ideologues, led by the liberal Indian Prime Minister Jawaharlal Nehru, were so
impressed by the economic trajectory of the Soviet Union that they saw no contradiction between the
simultaneous practice of heightened political freedom and deep economic controls at home. Nehru
was convinced that only the State could efficiently ‘enforce’ economic development, and in this
context, there really was very little difference between the temples of destruction in Russia, China, or
India.
There was such an adverse atmosphere for any support for economic liberties (loosely, the exact
opposite of what socialism allows) that the elite were quick to pronounce it as heartless and
unpatriotic.It may come as asurprise to many, but the removal of economic controls is still frowned
upon by many members of the Indian policy-making elite, both old and new. Yet another surprise in
this regard is that apart from the Congress and the Communists, even the Swadeshi Jagran Manch or
SJM (an affiliate of the RSS, which is considered to be a ‘guiding force’ for the BJP) opposes reform
of India’s antique and retarded labour laws.
Socialist is not a four-letter word, so the attempt with the sub-heading is not to be provocative;
rather, it’s a genuine query. The conventional wisdom in India is that the real socialist was Indira
Gandhi—after all, it was she who had inserted the ‘socialist’ word in the preamble to the
Constitution. On the other hand, Nehru was considered a liberal primarily because of his political
views. However, the reality is that Nehru endorsed socialist philosophy as early as 1931.
According to Granville Austin:1
the content of the [Congress] party’s socialism became clear in its 1931 Karachi Resolution.
Among other things, it said that ‘key industries and services, mineral resources, railways,
waterways [and] shipping’ were to be government controlled, and the government was to
safeguard the interests of ‘industrial workers’ and women and children…..The Congress
Socialist Party—formed in 1934—of which Nehru was a supportive non-member supported
a policy of ‘elimination of princes and landlords and all other classes of exploiters without
compensation’ (emphasis mine) and ‘redistribution of land to peasants.’
Implementing Socialism
As early as 1948, the Congress party adopted the Industrial Policy Resolution, a policy document that
was to become a formal part of Indian planning, as annexure to the Approach paper to the Second
Five Year Plan in 1956. It was at this juncture that heavy State interventionist policies found their
way into the discussions of the Indian Constitution, and the Constitution itself. As previously noted,
the Constitution is almost all about the rights of the State; most constitutions (and definitely the good
ones) concentrate largely on the rights of individuals.
In 1950, the Constitution of India (heavily influenced by Nehru) came into being, and it
contained an important section titled the Directive Principles. These principles did not have the force
of law—for example, the State could not be sued if the promise of universal primary education was
not met (which is incidentally one of the Directive Principles). But the State was directed to adopt
policies which would enhance their intervention into these Principles.
Some Pointers of Socialism
The controversial bank nationalisation initiated by Indira Gandhi was contained in the Industrial
Disputes Act of 1947, which included a list of industries which could be declared public in the
interests of the State or for the overall development of the nation. Second on the list was ‘banking’,
third was cement, fifth was cotton textiles, sixth was foodstuffs, seventh was iron and steel. The fact
that even foodstuffs and textiles were contained in a list of activities to be nationalised made a
mockery of the belief that the public sector was ‘forced’ to step in because of the private sector’s
recalcitrance. Whether it was the operation of hotels, or making bread, and later computers, and even
condoms, government ownership and production was considered necessary, and vital.
It all began as an innocent claim by the original planners that the country needed investment, and
the private sector was just not available, let alone willing to do the job. No surprise that it all ended
as socialist, backward India with not much to show during the first forty-three years of independence.
As mentioned earlier, although it is believed that it was not Nehru’s legacy, but Indira Gandhi’s
ideology, which led to the insertion of the phrase, ‘socialist state’ into the Indian Constitution, there is
enough evidence to show that the sequence of events or the thought process leading India to be a
socialist State, in word and deed, was inherited from Nehru and the Congress party that he dominated.
Amongst the many recent ifs being discussed in India is what would have happened if Sardar
Vallabhbhai Patel had not passed away in 1950. The reason: he was considerably less of a socialist
than Jawaharlal Nehru and a believer in market forces. In his book, The Man Who Saved India,
Hindol Sengupta starts Chapter Seven with the following quote from Sardar Patel: ‘The so-called
slogan of Socialist to march forward is hollow talk.’
The chapter and the book documents in great detail of how Patel’s ideas about the economy were
orthogonal to that of Nehru’s.
The Second Five-Year Plan and Beyond
State intervention was undoubtedly the theme of the Second Plan. Both in ideology, and deed,
Jawaharlal Nehru was more than an armchair socialist as evidenced in the following statement made
by him in Hindustan Standard, dated 17 May 1958:
Socialism to some people means two things: Distribution which means cutting off the pockets
of the people who have too much money and nationalization. Both these are desirable
(emphasis mine) objectives.
Two years after the publication of the Second Plan, China launched its interventionist campaign
called the Great Leap Forward in 1958. As documented in my middle class manuscript (see p. 18 for
a complete reference), the two countries have followed each other at several critical historical
junctures—for instance, China formulated the one-child policy in 1976, close on the heels of the ill-
fated sterilisation drive initiated by Indira Gandhi and executed by her son, Sanjay Gandhi during the
Emergency in 1975.
The penchant for control which was displayed in the economic domain, extended to culture as
well. In the late 1960s, India banned Louis Malle’s brilliant nine-hour documentary called Phantom
India. A few months earlier, India had obliged Russia by changing the title of the film, From Russia
with Love, to From 007 with Love.
Therefore, the lack of any consideration for the concept of economic freedom has been systemic
in modern India, and credit, or blame, lies squarely with the leadership, specifically the Congress
party and the Nehru-Gandhi dynasty. So pervasive was this dominance that several Indian
intellectuals experienced economic freedom only in the mid-1990s, and only after a non-Nehru-
Gandhi Congress leader had changed course in the early 1990s. For instance, one of the world’s
leading philosophers and champions of liberty, Nobel Prize winner, Dr Amartya Sen, was to
recognise economic freedom somewhat belatedly in his 1997 book, Development as Freedom.In this
book, Sen gives credit to both political andeconomic freedom as pathways to development; for very
long, he had primarily stressed on the former.
Nehru Was a Socialist Because of the Times
An alternative ‘explanation’ for Nehru’s socialism is that the intellectual environment had made him
turn against his liberal instincts. Hence, it wasn’t a ‘mistake’. If that be the premise, then we must also
accept the fact that our leaders were not leaders, but followers. But even this argument is only
partially valid. There were prominent dissenters to the socialist order—for example, B. R. Shenoy
and Milton Friedman. And the intellectual giant that Nehru was, he did not pay much heed to any of
the prescient observations on the consequences of socialism laid out by the future Nobel Prize
winner, Friedrich von Hayek in his 1945 treatise, The Road to Serfdom.
Be that as it may, in so far as Shenoy was concerned, he seems to have waged a lonely battle in
the 1950s and 1960s, and was later joined by noted economists, Jagdish Bhagwati and Padma Desai
in the late 1960s. Dr Shenoy emerged as a lone dissenter to the path-breaking socialist Second Five
Year Plan (1956–1961), by consistently opposing the extreme socialist and/or communist leanings of
Nehru and his daughter, Indira Gandhi. He had the foresight to predict that economic freedom (the
opposite of socialism) was not on the list of freedoms recommended by India’s populist leaders.
Shenoy’s objection was to the fact that for Indian politicians, economic freedom meant discrimination
in favour of the rich, and hence immoral. For Indian socialist leaders, including Nehru, until the poor
became rich, one could not, or rather should not, even conceive of freedom for the rich. The reason:
economic authoritarianism was an integral part of Nehru’s vision, and his legacy.
Indian politicians and intellectuals who swore ‘in the name of the poor’ required the State to
play a decisive role in order to ensure wealth for all and growth for the poor. The fact that the
ostensible benefits of this intervention never reached the poor was a point for future historians to
ponder—and something we consider in Chapter Five (see p. 64).
As a result of this thought process, Indian planners in the mid-Fifties proceeded to assume
draconian controls over economic activity, but within an essentially free political system. Shenoy
objected at several points against the State’s monopoly capitalism. There is this ‘not an’ apocryphal
story about a senior policy maker who when asked by me about the reasoning behind some highly
unreasonable financial market controls in the mid-1990s, said, ‘If you ask for logic, we will not help
you!’
The man who was the main architect of Indian planning and controls was none other than P.C.
Mahalanobis, a brilliant statistician who had Prime Minister Nehru’s support. In this context, some
scholars still maintain that Nehru was essentially a liberal and a reformer, and if India went astray
under his watch, then it was due to his advisers, in whom he placed his trust. There is an interesting
comment about planning and Mahalanobis by Milton Friedman, one of the several advisers to India on
economic policy.
In a memo on his advisory visit to India, Friedman wrote about Indian planning as follows:
Mahalanobis began as a mathematician and is a very able one. Able mathematicians are
usually recognized for their ability at a relatively early age. Realizing their own ability as
they do and working in a field of absolutes, tends, in my opinion, to make them dangerous
when they apply themselves to economic planning. They produce specific and detailed plans
in which they have confidence, without perhaps realizing that economic planning is not the
absolute science that mathematics is.
Higher Growth Under Nehru or the British?
A complete assessment of Jawaharlal Nehru’s contribution to economic development and alleviation
of poverty is to examine his record in the context of his environment i.e., how well did he do versus
what should have been the outcome. One set of expectations is to assess whether the economy was
better under Nehru or the British. Second, how did India compare with other poor, emerging
economies?
Critics of the Nehru model should acknowledge that India’s growth rate accelerated markedly in
post-independent India. As a free country after 200 odd years of foreign rule, we obviously did better
than we did under the Raj. It is also true that this phenomenon was enjoyed by practically all the
developing countries in the world. Table 3.1 shows the regional, and growth rates for selected
countries in the period 1890–1950, and 1950–1980. We did just as well as our neighbours, although
Myanmar topped the list. It is also a fact that almost all poor East Asian economies do systematically
better than India.

Table 3.1: India’s post-independence boom: In perspective

Country Growth Before and After

1890-1950 1950-1970
(average growth per year (in %)
South Asia
India 0.1 2.1
Sri Lanka 0.3 1.1
Pakistan 0.2 1.7
Myanmar -0.7 2.4
Nepal 0.1 1.2
East Asia
China -0.3 2.7
Indonesia 0.4 1.7
Korea 0.2 4.7
Malaysia 1.1 1.6
Taiwan 0.5 5.3
Philippines 0.3 2.9
 Thailand 0.1 2.8
Vietnam 0.1 0.6

Source: World Development Indicators, World Bank (2016); World Economic Outlook, IMF (2017)
Nehru: In Retrospect
In summary, Jawaharlal Nehru made two major economic mistakes. First, he failed to expand primary
and secondary education. Instead, he chose to build somewhat elitist models of wisdom (the IITs). No
one denies the importance of higher institutions of learning. Equally, all should appreciate that for a
poor and illiterate economy, a broad-based expansion of education was of utmost importance. This
would have been both equitable, and productive. Instead, as mentioned in the previous chapter, Indian
policy makers tried to compensate for their extremely short-sighted policy on education with a bad
short and long-sighted policy of reservations.
Then there were two major economic mistakes which were initiated by Nehru’s daughter, Indira
Gandhi, and since then followed by every Indian prime minister, and as result of which the poor are
still paying a heavy price. The first was the nationalisation of banks which allowed banks to be
brought under political and bureaucratic control, rather than the discipline of the market. This led to
inefficiencies, losses, and the accumulation and burden of large Non Performing Assets or NPAs. The
system was hollowed from within with rampant corruption.
The second major policy error was a classic in-the-name-of-the-poor policy—the Public
Distribution System or PDS which was initiated by her and expanded significantly by Sonia Gandhi
in 2013 via the national Food Security Act (FSA). One manifestation of that initial policy and the
FSA was large-scale corruption with half of the food meant for public distribution not reaching the
public.
There were other policy blunders which heavily influenced Indian economic policy for at least
forty-odd years. Such blatantly erroneous policies were the result of an ideological outlook, a
philosophy, which emphasised State intervention over private initiative—also known as socialism.

1 Granville Austin, Working a Democratic Constitution: A History of the Indian Experience, New
Delhi, Oxford University Press, 2003.
Mudh mudh ke na dekh mudh mudh ke
Mudh mudh ke na dekh mudh mudh ke…
(Don’t turn around as yet and look
Don’t turn around as yet…)
– Shri 420, 1955
Does the Economy Matter?

Is the Indian political landscape changing? Has our collective concern moved away from caste,
temples, religion to economic matters—housing, jobs, nutrition, and welfare? In the first quarter of
2019, the government reclaimed the deed to the disputed land around Babri masjid in Ayodhya, so the
Ram temple issue is not going away anywhere. But what has also dominated the airwaves is farmer
distress, high real interest rates, a slowing economy, and low inflation. Not to forget NPAs of banks
or the value of the rupee.
Just a few of the economic questions we shall discuss for each period since the first election
held in 1952. Data for several important economic variables are reported in Tables 4.1a and 4.1b—
per capita GDP growth, inflation, real wage growth, real interest rates, rainfall (percentage deviation
from normal) etc. The attempt is to assess broad trends, and aggregate relationship between the
economy and votes. We look at all the sixteen elections held to date, and speculate (just a bit) on the
forthcoming seventeenth general election. One approach to reading the table is to look back over the
period in question and see if the economy actually corresponded with electoral change. For how long,
implicitly at least, was vikasor developmentan important voting parameter? Narendra Modi fought on
the slogan of development in 2014, and persists, despite provocations, to emphasise Sabka Saath,
Sabka Vikas, translated as Collective Efforts Inclusive Growth. Was this a radical departure from
previous regimes?
It is not that non-economic issues are not important. They are and are discussed elsewhere in the
book. A purist economist would however say that political and social change is best revealed via
economic data.
The Importance of Delta (Change)
Economists use the term delta to denote change (from the first letter of a Greek word, meaning
change). For most parameters, it is the study of changes, rather than levels, that is of greatest interest.
For example, the study of how poor you are in a particular year may not be as important for voting
behaviour as the change in your income (since the last election). This chapter will focus on the
changes, which quite frankly, can be problematic. Note that the delta sign in Greek is denoted by the
sign of a triangle and that is a sign ‘as hazard forward’ in a road sign. Change can mean travel into
something different, somewhere unknown, somewhere hazardous.
Elections 1-4: 1952 to 1967
The first four general elections in India can be considered as ‘grace period’ elections. The Congress
party won repeatedly, but with a very low growth record. The average GDP growth during the
twenty-five-year period was just 3.5 per cent a year. The well-known economist and teacher, Raj
Krishna had mischievously and correctly dubbed it as the ‘Hindu rate of growth’. The per capita GDP
growth, the one that really matters for welfare, grew at a paltry 1.5 per cent a year. Droughts were a
big story for two successive years, 1965–1966 and 1966–1967, with the percentage deviation from
long-term mean recording -19.1 per cent, the largest rainfall deficit in the history of independent
India.
There was little improvement (read change) in people’s lives, yet the Congress was the
preferred party. Therefore, economy did not matter in the least. What did matter was political
freedom—people were grateful that the nation was finally free. In a newly-independent nation,
economic freedom and economic performance were eclipsed and perhaps rightly so by the collective
relief and euphoria of the people, and the Congress was viewed as the party which had fought the
most for freedom.
Election 5: 1971
The 1971 election can be characterised as the first election when the economy really mattered. Indira
Gandhi fought the election with the evocative slogan, garibi hatao—remove poverty, a sure sign that
the Indian voter was moving away from tradition, and most importantly, caste. She won with 43.7 per
cent of the vote, a significant three ppt improvement over the 40.8 per cent of the vote she gained in
1967. In terms of seats, Indira Gandhi won 352 seats out of the 441 she contested. Her strike rate of
43.7 per cent was only marginally lower than what her father had achieved in the first three elections.
Conclusion: Indira Gandhi had arrived and on her own.
Election 6: 1977, The ‘Emergency’ Election
The sixth Lok Sabha election was due in March 1976. But on 12 June 1975, Justice Jagmohanlal
Sinha of the Allahabad High Court stunned the nation by delivering a verdict against Indira Gandhi
which barred her from holding any high office for six years. Even as she had obtained a stay order,
she in turn stunned democratic India by imposing an Emergency the next day, 25 June 1975. The
Supreme Court overturned her guilty verdict a few months later on 7 November 1975.

Table 4.1a: Economic performance during federal governments, 1952–2018

Time Period Population Growth (CAGR) Performance

(average)

GDP GDP per Capita CPI GDP /CPI

(in mil) (Y-o-Y, in %) (growth ratio)
1952–56 396 4.5 2.7 0.1
1957–61 434 3.3 1.3 3.3 0.39
1962–66 480 2.3 0.2 9.4 0.02
1967–70 527 5.4 3.1 1.6 1.94
1971–76 591 2.7 0.5 6.9 0.07
1977–79 655 2.4 0.4 4.4 0.09
1980–84 716 5.3 3.2 6 0.53
1985–88 788 5.4 3.2 7.6 0.42
1989–90 839 5.6 3.5 7.3 0.48
1991–95 901 5.1 3.1 9.6 0.32
1996–97 964 5.9 4 7.9 0.51
1998 992 6.5 4.6 12.5 0.37
1999–03 1050 5.7 4 3.8 1.05
2004–08 1130 8.1 6.6 5.9 1.12
2009–13 1220 6.9 5.3 9.7 0.55
2014–18 1300 7.1 5.8 4.4 1.32
Source: CSO, National Accounts; RBI, Handbook of Indian Statistics; author computations.
Notes: The difference between GDP and GDP per capita growth rates is the rate of growth of population.

Let’s assume there was no court case, no conviction, and no Emergency. What would the ‘economic
voter’ have done in 1976? Those were early days of globalisation (actually one would date its advent
around 1980), but the world did witness a quadrupling of the oil price in October 1973. CPI inflation
for the four years post the garibi hatao promise averaged fifteen per cent; GDP growth averaged 1.6
per cent per annum; and per capita growth averaged -0.7 per cent. In all likelihood, poverty rates
would have also increased significantly, as one can speculate in the absence of data for it. Those
were still ‘early’ days for the Nehru-Gandhi dynasty, but what the data does suggest is that Indira
Gandhi would have been tested even without the Emergency in 1976. It is therefore an open question
that given the euphoria over India’s win in the 1971 war against Pakistan and the formation of
Bangladesh, whether a hard landing awaited Indira Gandhi in 1976.
The post Emergency election was held in March 1977, and the hastily cobbled Janata Party
coalition won a decisive mandate of 345 seats with a vote share of 51.9 per cent. If the Janata
coalition can be considered as one cohesive unit, then this would rank as the highest ever election
victory in Indian history, more than three ppt higher than Rajiv Gandhi’s landslide victory seven years
later in 1984.
Between 1971–1976, per capita GDP growth averaged only 0.5 per cent per annum, and
inflation averaged, 6.9 per cent. In the next three years of Janata rule, inflation fell, as did per capita
income growth. The economic slowdown, and political infighting, led to the fall of the Janata Party in
1980.

Table 4.1b: Economic performance during federal governments, 1952–2018

Time GDP Rainfall Real

Period Agriculture

Wages Bank Repo 10 year

rate Rate yield
CAGR (in %) (% from CAGR % % %
normal) (%)
1952–56 3.6 3.2
1957–61 2.1 8.2
1962–66 -0.8 -4.4 -4.8
1967–70 6.7 -0.8 3.2
1971–76 0.8 -4 -0.9
1977–79 -0.6 -4.4 4.4
 1980–84 5.5 -1.5 2.4 3.4 2.1
1985–88 3.2 -4.3 -0.7 2 3.3

1989–90 2.6 2.6 -0.6 2.4 3.8

1991–95 2.3 0.3 -7.7 1.7 2.4
1996–97 3.4 -2.7 6.7 2.9 4.7
1998 5.8 -3.5 -3 -4.2 -1.1
1999–03 2.1 -10.2 0.6 3.1 3.7 4.8
2004–08 2.8 -3 -0.1 0 0.8 1.3
2009–13 4.9 -4.6 6.9 -2.9 -3.4 -2.1
2014–18 2.3 -7.6 0.6 3.4 2.2 2.7

Source: CSO, National Accounts; RBI, Handbook of Indian Statistics; author’s computations.
Notes: Real wages refer to wages of a ploughman; all real variables are computed as the difference in nominal minus the rate of
CPI inflation.
Elections 7&8: 1980 & 1984, The Family Returns
Given no glue to hold together a clutch of satraps, the Janata Party coalition collapsed within two
years of its grand formation and the Indian voter brought back Mrs Gandhi with near identical vote
share (and seats) as in 1971 which she had won on the back of the garibihatao slogan.
It is interesting to note that both the 1980 and 1984 wins were under exceptional circumstances
—the former because of the ignominious collapse of the Janata Party, and the latter because of Indira
Gandhi’s assassination.
The 1984 outsized victory was obviously because of a sympathy wave for Indira Gandhi and
unfortunately also because of the anti-Sikh riots which followed immediately thereafter. It may be
recalled how Rajiv Gandhi had fanned anti-Sikh sentiment by providing the following ‘explanation’
or rather ‘excuse’ for the pogrom:
We must remember Indiraji. We must remember why her assassination happened...When
Indira’s assassination happened, there were riots in the country. We know that the hearts of
the Indian people were full of anger and that for a few days people felt India was shaking.
When a big tree falls, the earth shakes.
Election 9: 1989, The Beginning of the End of Nehru-Gandhi
Dynasty?
In 1989, Rajiv Gandhi lost his re-election bid against his own party member-turned-Finance
Minister-turned-rebel, V. P. Singh. The Congress’ vote share fell to a respectable 39.5 per cent, just
marginally lower than the forty-three per cent share of 1971 and 1980. However, the seats fell
dramatically as well—to almost half the 1971 and 1980 levels. This was the first big sign of a major
change in voters’ behaviour and attitudes.
Throughout the 1980s, per capita GDP growth averaged a healthy 3.2 to 3.5 per cent, and CPI
inflation had fallen back to its long run equilibrium value of around 7.5 per cent. Ideally, these two
important economic indicators shouldn’t have led to Rajiv Gandhi’s defeat in 1989. In retrospect,
there are two possible explanations for this: first, some fallback was expected from the outsized 1984
victory. Second, growth was good, and recorded a significant acceleration; inflation was contained
and close to the international median for developing economies.
But most importantly, the out-sized loss had to do with the voters’ impatience with the Congress
party, and in their willingness to try out something new. And not only did they find an anti-corruption
messiah in V.P. Singh, but the sixty-four crore Bofors gun deal scandal provided the ideal ignition for
an anti-corruption victory.
Finally, the Indian economic voter had arrived just when the education levels were higher, the
country was less poor, and the influential middle class had increased to a double-digit level. As
mentioned earlier, the first forty years after independence did not matter, but there was a palpable
change four decades after.
Election 10: 1991, Beginning of the Coalition Era
Bill Clinton won the 1992 election with the slogan, ‘It is the economy, stupid.’ In India however,
economy had begun to affect politics a year earlier. By the time of the 1991 election, GDP growth had
slowed to a crawl—the three-year GDP growth path in 1989–1990 was 6.1 per cent; and 5.3 and 1.4
per cent respectively in 1991–1992. Oil prices had doubled in 1990 because of the Kuwait war, and
globalisation and its full impact was felt by all economies across the globe.
Much like the 1984 election, the 1991 election was also fought under the cloud of an
assassination. Rajiv Gandhi was killed a day after polling began while on a campaign trail to Tamil
Nadu, in Sriperumbudur. The Congress party soldiered on despite a great tragedy and won the
election led by P.V. Narasimha Rao, with 232 seats.
Meanwhile, after his spectacular victory in 1989, V.P. Singh lost the 1991 election in the
aftermath of Rajiv Gandhi’s assassination. It needs to be mentioned here that the ‘messiah-politican’
Vishwanath Pratap Singh attempted to establish an alternative for the monolithic Congress and was at
the helm of a coalition called the Janata Dal (under the National Front umbrella), a combination of
many parties, including the BJP. The Janata Dal also went the Janata Party way—lasting just a year;
Chandrashekhar, the ‘young turk’ was prime minister for a mere seven months before Rajiv Gandhi
pulled the carpet from under his feet and sent him home.
After 1984, even as the call for building the Ram temple in Ayodhya was getting louder, and on
another note, Punjab was faced with a secessionist movement, religion had segued into politics. And
so had the scourge of Indian electoral politics: caste. Starting with the 1991 election, Mandal politics
meant that every Indian now knew her caste, because there were non-food handouts (education, jobs)
associated with knowing your caste. This is when Indian democracy came to be best known as being
based on ‘cast(e)ing your vote’.
Be that as it may, V.P. Singh failed to harvest any political gains despite his virulent
championship of caste politics.
After assuming office as India’s ninth prime minister, Narasimha Rao, along with his Finance
minister, Dr Manmohan Singh initiated wide ranging economic reforms in the areas of industry and
trade, which resulted in the per capita GDP growth accelerating from 3.3 per cent average in 1992–
1993, to a 4.7 per cent average for 1994 and 1995 (see Chapter Five, p. 64). Hence at the time of the
1996 election (April–May), Rao had the highest two-year GDP growth average ever recorded in
India. The 1988–1989 two-year average was marginally higher, but that was as a follow through to
the drought-induced low agricultural growth in two successive years, 1986–1987 and 1988–1989
(rainfall deviation from normal of -9 and -12 per cent respectively).
Ideally, the high growth should have signalled a re-election victory for Rao in 1996, if one were
to go by Bill Clinton’s mantra. But there were discordant signs. The growth in real unskilled
agricultural wages was -7.7 per cent, the lowest ever recorded for an election period. In addition,
CPI inflation had accelerated to 9.6 per cent, nearly the highest recorded average ever.
Dr Manmohan Singh had once famously said how inflation was politically harmful and it
wouldn’t be wrong to surmise that it was perhaps drawn from his experience during the 1996
election. History repeated itself in 2014 when inflation reared its head yet again, and this time
around, Manmohan Singh was the incumbent prime minister.
Elections 11-14: 1996–2004, Maturing of Non-Congress Coalitions
It would take nearly twenty years before India would witness a majority government at the Centre.
Within a space of four years, three elections were held in 1996, 1998 and 1999 and each one was a
coalition government with the United Front forming the government in 1996, and the BJP in 1998 and
1999.
Atal Bihari Vajpayee became the prime minister at the head of a minority government in 1996
and left after a short thirteen-day stint. The next to take centre stage in the coalition drama was H.D.
Deve Gowda, the Karnataka strongman at the head of the United Front government.
However in the midst of hastily cobbled coalitions and frequent adjustments for political gains,
the BJP had succeeded in registering its presence as a major player; it may be recalled that this was
the same party which had won a meagre two seats when the Congress had surged forward in 1984.
The vote shares of the two major parties in these four elections (1996, 1998, 1999, and 2004)
tells an interesting story—an average of only 4.3 ppt separated the two polls of coalitions. And the
Congress had the higher vote share of 27.4 per cent. Thus, the real kingmakers were the alliance
partner(s).
The Vajpayee administration took forward the economic reforms initiated by Narasimha Rao and
Manmohan Singh, and gradually, trade, monetary and exchange rate policy began their ascent to the
centre of economic policy.
It was at this juncture that India made a word fashionable—disinvestment, i.e., partial
privatisation, which meant that the share of government in State-run enterprises began to be reduced.
During this period, 1999–2003, per capita GDP growth averaged four per cent which was the
highest for any prior five-year period. (The previous highest five-year average was 3.2 per cent
witnessed during 1985–1988.) The inflation record was truly spectacular—an average of only 3.8 per
cent, a steep decline of nearly five ppt from the levels prevailing in the previous four decades.
Also shown in Table 4 is the ratio of GDP growth to inflation rate. In the 1970s, the American
economist Arthur Okun had developed a misery index for the US economy which comprised the total
of unemployment and inflation rates. Unfortunately, a parallel index for India is not possible because
of lack of data on unemployment. What can however be estimated is the ratio of GDP and inflation
rates—the higher this ratio, the better the performance—hence we label it as the GDP performance
index. (Finance Minister Arun Jaitley had recently used this index to demonstrate how well India has
grown post 2014.) For the 1999–2004 period, the growth/inflation hit its highest level for any
election period since 1967–1970, which had recorded an artificial high of 1.94 because of a huge
increase in the numerator (GDP growth after two successive drought years in the mid-Sixties) and a
large decline in the denominator (inflation decline post the inflation hike due to bad rainfall).
These were reasons enough for the BJP to feel confident that it had a very good chance of
winning the 2004 election. As it turned out later, much like what Narasimha Rao had faced, an
impressive economic record was not good enough for winning elections.
Let us therefore examine what led to the BJP’s defeat in 2004. It is more than apparent that two
major economic factors may have played a crucial part—high real interest rates and bad rainfall.
With inflation down steeply, nominal policy rates did not keep pace, and the worst rainfall of ten per
cent less than normal was witnessed in election history. We shall return to this subject in the next
chapter in some detail.
Elections 15 & 16, 2009 & 2014: Outlier Elections?
The 2009 election was a cakewalk for the UPA alliance—the Congress won 206 seats on its own
with 28.6 per cent of vote share. The BJP’s vote share during this election was ten ppt lower. Five
years later, the BJP won 282 seats with a thirty-one per cent vote share, while the Congress recorded
its worst ever vote share, which was below twenty per cent, and managed only forty-four seats.
Going into 2019 elections, analysts have to decide which of these two elections was an outlier.
Economic determinants of voting are consistent with both the 2009 and 2014 results. For
election 2009, GDP per capita growth (average between 2004 and 2008) at 6.6 per cent per annum
was the highest ever recorded in India, and inflation was at a reasonable average of 5.9 per cent. The
1.12 growth/inflation ratio was also higher than before, and marginally exceeded the 1.05 ratio
recorded in 1999–2004. The real interest rates also favoured the incumbent UPA—an average of just
zero per cent; rainfall at three per cent below the long-term average was also normal; agricultural
growth was low, just 2.8 per cent per annum, and growth in real agricultural wages was zero.
In the 2009–2013 period (as input into the 2014 election), the average GDP growth declined to
only 5.3 per cent per annum. The real shock for the economy (and the Congress) was an acceleration
in the inflation rate to 9.7 per cent per annum—the highest for any election period in history, even
when compared to the OPEC-induced domestic and worldwide inflation of the 1970s. On the plus
side, rainfall was close to normal, and real agriculture wages grew at 6.9 per cent per annum, the
highest ever. But none of this could help the Congress’ electoral chances.
High inflation was one of the three most important economic factors explaining the 2014
Congress debacle. The other two—corruption to an unprecedented level, along with per capita GDP
growth falling to 5.3 per cent from the 6.6 per cent average that had lent the Congress the
unexpectedly large seat gain in 2009.
Election 17: May 2019
Let us dive straight away into the economic record for the last five years. The per capita average
GDP growth is fifty basis points higher at 5.8 per cent, and the average inflation is more than 500
basis points lower. The GDP/inflation ratio at 1.32 is almost 2.5 points higher than the low 0.55 level
recorded during 2009–2013. However, what is glaring is that the real bank rate at 3.4 per cent is the
highest level since the early 1980s. Growth in real agricultural wages at 0.6 per cent per annum is
low, but considerably higher than was obtained in the five years preceding the Congress’ victory in
2009.
The government, through various redistribution schemes (transfers for housing, toilets, provision
of gas cylinders for the poor etc.) has supplemented the incomes of the poor. These in turn have raised
the expectations of those who have not yet got the transfers (in particular, for the construction of
toilets and houses) and therefore there is a new uncertainty—the arrival of the revolution of rising
expectations. Therefore, what one needs to ask is whether those who have not obtained transfers be
inclined to vote more for the Opposition or for the BJP?
There are of course other determinants of elections—most importantly, sentiment and in a multi-
party eco-system, the nature of electoral alliances, but more about this later in the book.
Behti hawa sa tha woh
Udti patang sa tha woh
Kahan gaya usey dhoondo
(He was like the flowing breeze
He was like a kite soaring in the sky
Where did he disappear
Let’s find him…)
– 3 Idiots, 2009
The First Reformers

The election of P.V. Narasimha Rao as prime minister was a breakthrough for the Indian polity, and a
departure from the socialist norms under which Indian economy had laboured for the previous fifty
years. A point which needs to be reiterated: the reason it was an important first was because Rao
wasn’t a Nehru-Gandhi. The last time this had happened was post Jawaharlal Nehru’s death on 27
May 1964, after which Lal Bahadur Shastri became prime minister (I am not taking into account the
short stints of Gulzarilal Nanda in this context). While on a tour to Tashkent in 1966 in the former
USSR, Shastri passed away after suffering a heart attack, which was one year before the next
scheduled election.
Post Rajiv Gandhi’s assassination in May 1991, the ‘natural successor’, Sonia Gandhi abstained
from taking on any political role, and Narasimha Rao was elected the leader of the party, and won the
mandate in the same year with 232 seats (of the required 273 for a majority). Both Dr Sanjaya Baru
(media advisor to Dr Manmohan Singh) and author, Vinay Sitapati1, argue in their respective books
that Rao’s personality, ideology, and inclination was heavily in the socialist-communist mould. He
proceeded to select Manmohan Singh to be his Finance minister, the first and only time (unfortunately)
that a rank political outsider, a technocrat-bureaucrat, was chosen for the second most important post.
In time, Rao-Singh became the most famous doubles pair in Indian political-economic history, so
much so that their names have become synonymous with economic reform the world over. As it
should be, for in quick-time, they set about dismantling the all-obtrusive license raj. A regime where
firms were required to take a license (read bribe) from the government for most investment projects
that they wanted to undertake. India began to rely more on the (dreaded) market; and State-sponsored
socialism became less of an influence. In the main, the target of reforms was liberalisation of the
industrial economy, the area where blind State intervention had been the highest.2
Licensing was no longer required, except for eighteen industries of critical importance,
presumably for the security of the country! Foreign trade, and foreign investment, was opened up.
Import tariffs were reduced, and export growth was emphasised. Industrialists no longer had to beg
and bribe government officials, at least not to the same extent as before. Reforms led to an
acceleration in long-run GDP growth from around 5.3 per cent in the 1980s to 6.3 per cent in the
1990s.3 From closed minds and a closed economy, the move toward an open, competitive economy
had begun.
Therefore, the Rao-Singh regime deserves major credit for the beginning of the change in the
mindset. A mindset change is a long-term project, and India is still undergoing transformation in this
regard. For instance, the one sector, which also happens to be the most important, which was
neglected by the Rao-Singh reforms was agriculture—it continued to be planned, controlled, and
corrupt. This lack of agriculture reform since then has gone on, although the first steps towards
dismantling a corrupt, and wrong, agricultural policy began to be undertaken post 2014.
 In retrospect, even as the reforms initiated by the Rao-Singh duo was a watershed moment in
India’s economic history, it was nevertheless questioned not only by several planners, and the
socialists, the old elite, Sonia Gandhi (who was carrying forward her mother-in-law’s legacy), the
Communist parties of India, but also the affiliate of the BJP, Swadeshi Jagran Manch. What is most
revealing is that this anti-reform alliance continues to this day.
Economic Reforms—How Rao Lost His Re-election Battle
Are economic reforms the kiss of political death? This is a question Narendra Modi faces in 2019.
History suggests that according to the economic criteria, he maybe in for a tough fight. Not only
Narasimha Rao and Dr Singh, but also the late Atal Bihari Vajpayee—the three reformers, lost their
bids for re-election. I have always claimed (as I continue to do throughout this book) that economy
matters, and matters a lot, politically. So, am I reading history wrong?
I don’t think so. Reasons: Rao lost because he was caught in the vortex of the beginning of
another seismic change in India’s political economy—the structural decline of the Congress party. It
also did not help that the traditional dynasty loyalists within the Congress may have actively worked
towards his defeat. Some evidence for this hypothesis is obtained from the treatment meted out to him
after his death on 23 December 2004, six months after the beginning of the Sonia-led UPA rule—P.V.
Narasimha Rao’s body was denied permission to be cremated in Delhi. This inexplicable, and nasty,
treatment meted out to Narasimha Rao has been well documented by both Sanjaya Baru and Vinay
Sitapati in their respective books.
The electoral story behind Rao’s defeat in 1996 was as follows. He came in with 232 seats in
1991; he left with 140 seats, a decline of ninety-two seats; seventy seats were lost in just three big
states accounting for 127 in the 543 seat parliament. In Madhya Pradesh, the Congress party’s tally
declined from twenty-seven wins in 1991 to a meagre eight in 1996; the largest INC win since then
has been twelve in 2009.
In yet another state, Maharashtra, the Congress’ score declined from thirty-eight wins in 1991 to
only fifteen in 1996; the largest INC win there was thirty-three seats in 1998. Since then, the INC has
managed to win only ten, thirteen, seventeen, and two seats in successive elections. (Maharashtra
contributes forty-eight seats; the second highest individual state contribution to 543.)
In Tamil Nadu, the Congress’ seat count declined from twenty-eight wins in 1991 to zero in
1996; the largest INC win since then was ten seats in 2004, and eight, and zero again in subsequent
elections.
The Rao-Singh economic reforms helped all the states, and particularly the industrial states of
Maharashtra and Tamil Nadu. But this is where they lost the most heavily. One has to thus score the
1996 Rao loss to structural factors associated with the Congress, factors which would manifest
themselves in spades, except for the 2009 election (which Manmohan Singh won for the Congress).
The 1996 election was therefore the second crack in Congress’ electoral misfortunes, the first
having occurred just seven years earlier in 1989. Rajiv Gandhi, with 197 seats, brought about a 150
seat decline from the norm of 350, the average expected level of Congress’ seats in the preceding four
decades. With regard to his 1984 achievement, the fall was an astounding 218 seats. Even in the
subsequent 2009 peak, INC gained just nine seats more than what Rajiv Gandhi had managed in 1989.
The Defeat of Manmohan Singh, circa 1999
I had the privilege of being closely associated with the (doomed) Manmohan Singh’s Lok Sabha
electoral campaign in 1999. He ran from South Delhi, obtained the highest vote share there for the
Congress since 1984 (46.3 per cent) and still lost the election to BJP’s Vijay Kumar Malhotra by six
ppt. There was no doubt that Manmohan Singh was widely respected, had a successful tenure as
Finance minister, and was considered near unique among politicians as being personally, and
scrupulously honest. The first technocrat to be a high-level Cabinet minister, and the darling of the
middle classes; honest and knowledgeable and efficient—a dream candidate for the steadily
expanding middle class of South Delhi (and India). Yet the unthinkable happened—he lost. Why?
One explanation could be that the Congress leadership did not want him to win. The dynasty,
actually all dynasties, want to protect their fiefdoms, and the Congress, to this day, has been playing
that game. Simply put, Manmohan Singh was the first (and as it turns out, the last) credible threat to
the dynasty.
On 2 September 1999, I proceeded for a meeting with the ‘brain trust’ of Dr Singh, which was
headed by my close friend, Isher Judge Ahluwalia, one of India’s leading economists and wife of
Manmohan Singh’s trusted aide during 1991–1996, Montek Singh Ahluwalia. Upon arrival at the
meeting, I was met with glum faces—prior to this meeting, we were confident that Manmohan Singh
would win, for all the reasons mentioned earlier. But today was different—on arrival, I was handed a
small hand-written note on what Dr Singh was made to say at an election rally. It was then we knew
that Singh would lose, despite ours, and his best efforts.
As reported in rediff.com, dated 2 September 1999, the note said:
Former finance minister Manmohan Singh, Congress candidate for the Lok Sabha from South
Delhi, today accused the Rashtriya Swayamsevak Sangh, ideological parent of the Bharatiya
Janata Party, of being involved in the 1984 anti-Sikh carnage in the capital.
Addressing reporters at the Press Club of India, Dr Singh said the 1984 riots were ‘a black
spot and the saddest event’. But the Congress as an organisation had no role in it, he
claimed. ‘It should not have happened.’ He dismissed as ‘BJP propaganda’ reports that Sikh
voters had asked him during campaigning how they could vote for a party that had ‘blood on
its hands’ (emphasis mine).
And two days later, rediff.com reported yet again as follows:
According to Govindacharya, ‘Evidence implicating Congress leaders in the riots, or rather
the mayhem that followed the assassination of Prime Minister Indira Gandhi, is well
documented. What more proof does Manmohan Singh require?’
In the Congress camp, activists campaigning for Dr Singh were also upset. ‘What was the
need to bring this issue up again?’ said one. ‘After last year’s assembly election it seemed as
if the Sikhs were finally coming back to the Congress. This could upset them.’
The issue takes on an increased sensitivity as Dr Singh, himself a Sikh, had refused to contest
the election from Delhi unless those leaders tainted by the 1984 riots, in which more than
2,000 Sikhs were killed, were denied tickets by the Congress.
His original statement had been made at the Press Club of India in response to a question
asking him if he felt comfortable contesting the election as the candidate of the very party that
was responsible for those riots.’
Manmohan Singh: 2004–2008
When Manmohan Singh first accepted to be prime minister in 2004, implicit in his contract was the
understanding with Sonia Gandhi that he would be in charge of running the government, and she
would run the Congress party. This was the much-vaunted, CEO-Chairman model, explicitly and
implicitly enunciated by both. Manmohan Singh had held every conceivable job in the economic
establishment in India, and from all accounts, had been given a free hand by P.V. Narasimha Rao to
run the economy, and implement economic reforms, as he saw fit. It was therefore natural for Dr
Singh to believe in Sonia’s assurances that he would be master of his own successes and failures.
The inherent weakness in the CEO-Chairman model of governance was exposed by none other
than a leading Congress politician and ex-chief minister of Madhya Pradesh, Digvijaya Singh. In an
interview to rediff.com in 2008, he had stated: ‘Personally, I feel this model hasn’t worked very well.
Because, I personally feel there should not be two power centres and I think whoever is the PM must
have the authority to function.’
In the 2009 election, the Congress chalked up an impressive electoral victory. From a paltry 145
seats in 2004 (and 114, 141 and 114 in the previous three elections), its seat count advanced by sixty-
one seats, the largest jump since 1971 when it jumped by sixty-nine seats, and ignoring the outlier
year of 1980 (199 seat increase). The 2009 increase was also just one seat short of the sixty-two seat
increase achieved by Rajiv Gandhi in 1984.
In retrospect, who should be credited with the Congress’ 2009 victory, the CEO or the
Chairwoman? There are divergent views—the differences arise with regard to one’s views on the
performance of the economy, and considerations about the relevance, and importance, of Sonia
Gandhi’s leadership. The economic facts are however well known: GDP growth between 2004 and
2008 expanded at 8.1 per cent, and per capita income at 6.6 per cent per annum; agricultural growth
was an average 2.8 per cent, and real interest rates averaged 0.8 per cent; even the rainfall was good
and the world economy had its best five years ever; growth in agricultural wages (ploughman) was
zero (actually -0.1 per cent per year). With such a record, the economy had to be an important factor.
Were economic reforms responsible for this spurt in GDP growth in UPA-I? Not according to
Montek Ahluwalia, who was then Chairman of Planning Commission. In February 2009, months
before the election, I was the host of an NDTV interview show called Tough Talk. My first guest,
very appropriately, was Montek, a close friend, and one of the best policy economists that India has
produced. I asked him to document any economic reform that the UPA had engineered to make the
growth spurt possible. His short and candid answer: ‘We abided by the Hippocratic oath—we did no
harm.’
The Congress won the 2004 election and did not believe that Atal Bihari Vajpayee, and the
NDA, had gifted it an economy in the take-off stage. It believed it won in 2004 because the ruling
NDA was selling India a false reality—it was not an improving middle class economy as the
campaign slogan of ‘India Shining’ suggested. In Congress’ ‘reality’, India was still a poor country
and in need of old-style socialist reforms. The Congress gloated, not just in its victory, but in feeling
vindicated that India was just as poor as it had thought.
The 2009 election came along. India had grown at more than eight per cent an annum for five
years, and large parts of the Congress still believed that the poor had voted them in—not realising
that it was the middle class that was increasingly important, and dominant, and one that had indeed
voted them in. The middle class does not much care for populist platitudes; it worries about how its
taxes are spent. It is concerned about corruption, and about both crony socialism and crony
capitalism. This was to become apparent in 2014.
The vote share of the Congress increased from 26.5 per cent in 2004 to 28.5 per cent in 2009—a
not very significant increase. The highest recorded seat per vote percentage (a concept discussed at
length in Chapter Twelve, see p. 181) in India was BJP’s 9.1 in 2014; the highest recorded by
Congress was Rajiv Gandhi’s 8.6 in 1984. No matter how one would have estimated it, the best yield
would have advanced the Congress’ seats by an extra twenty; yet, it gained more than three times this
‘maximum’ estimate. The only conclusion is that it was economic performance that won Congress one
of its best victories, ever.
Congress’ Expectations for 2019
It is not a coincidence that news reports, and discussions, around Congress’ expected come-back
‘victory’ points to the exception of 2009 being repeated. There are other similarities in this
expectation. All the three members of the Gandhi dynasty—Sonia, Rahul and Priyanka—were fully
involved in the 2009 election. Identification of whether the 2009 election should correctly be
attributed to the economy (Manmohan Singh) or non-economic factors (caste, dynasty) will be
substantially resolved by the results of the 2019 election. We all wait expectantly.
UPA II—2009–2014
But what about the largest election loss of the Congress in 2014—economy or politics? Again, the
vote goes to economy, but what remains to be determined is whether the loss had more to do with Dr
Singh’s economic policies, or Sonia’s politics (and economics)?
The second term of the Sonia Gandhi-Manmohan Singh government, 2009 to 2014, was marked
by a severe decline in economic conditions, and a marked increase in high level corruption. Starting
in 2010, few months passed by without some reference to corruption in the highest places.
Prime Minister Manmohan Singh was not considered personally corrupt; unfortunately for the
Congress party, it was not given the benefit of doubt in 2014. If high-level corruption was the only
major flaw, the Congress might well have got re-elected in 2014. But the five years, 2009–2014 were
also the highest inflation years in India; GDP growth had also steadily declined from around nine per
cent per annum to around five per cent.
During Vajpayee’s reign (1999–2003), CPI inflation in India averaged 3.8 per cent per annum, a
radical decline from the steady 7-9 per cent inflation encountered during the previous fifty years.4
This advantage was squandered away by the Congress, and especially so during UPA-II when
inflation averaged 9.7 per cent between 2009–2013, near double the rate in UPA-I (5.9 per cent), and
near triple the rate during NDA-II (1999–2003, an average of 3.8 per cent). GDP growth had also
steadily declined from around 8.1 per cent per annum to around 5.5 per cent in the final two years of
UPA-II.
There is substantial evidence that the economic decisions made by the UPA couldn’t possibly
have been made by an economist, let alone an economist of as much repute as Manmohan Singh. As
an economist qua economist, he recognised, more than fifty years ago, the role that exchange rates
played in generating exports, and did so when the fashion in India, and most of the world, was one of
export pessimism. As an economic policymaker, he was the Finance minister who helped initiate, and
implement, the beginning of a change in the socialist-controlled economy mindset. He has consistently
been vocal about his opposition to inflation, arguing repeatedly that it was both wrong economics and
wrong electoral politics (the poor and the middle class gets hurt most by high inflation).
A halving of GDP growth, a doubling of inflation rates, a twenty per cent depreciation of the
rupee, and record current account deficits in 2013 reflected the deep abyss the Indian economy was in
as the Congress entered the 2014 election. In the Opposition camp, Narendra Modi was the prime
ministerial candidate, a chief minister with a stellar performance record. A historical mismatch
maybe?
The 2014 election was, in equal parts, reflective of the candidacy of economic reformer,
Narendra Modi (see Chapter Eleven, p. 160 for details) and of Sonia Gandhi and the Congress
possessing an outdated ‘in-the-name-of-the-poor’ mindset. This is the same attitude that proposed an
employment guarantee programme in the biggest growth boom in India’s history; the same mindset that
introduced retrospective tax legislation at a time of India’s greatest need for foreign capital; the same
mindset that tried to win over farmers by generating the highest inflation in India’s history. Also a
mentality that ultimately exploited the loyalty of its most transparent loyalist.
Vajpayee: 1999–2004
Atal Bihari Vajpayee became the first NDA prime minister in 1998, and subsequently in 1999 when a
successive election was held after Sonia Gandhi won a vote of no-confidence stating that she had the
requisite 273 votes required for victory. In the 1999 election, the Congress won only 114 seats to
BJP’s 182.
The Vajpayee administration was witness to a continuation of economic reforms—the trend of
trade reforms accelerated, and monetary and exchange rate policy began their ascent to the centre of
economic policy. And India contributed a new word to the English language—disinvestment, i.e.,
partial privatisation. The share of government in State enterprises began to be reduced.
The Vajpayee era could easily be described as the beginning of the golden age of Indian
economy. Economic reforms continued (started in 1991) and these were in large part responsible for
the acceleration in GDP growth between 2004–2009. His government moved towards a realistic
exchange rate and a realistic real interest rate policy. The deep overvaluation of India’s exchange rate
was reduced5, and real policy rates increased over the preceding election period, 1996–1998, by
over 250 basis points. At the same time, real rates on government securities declined by 450 basis
points between 1999 and 2003. This paved the way for the economic boom witnessed during UPA-I
from 2004 onwards; after averaging around twenty-five per cent of GDP for the previous five years,
the share of investment in GDP increased by five ppt in fiscal year 2004–2005. As is well known, the
fiscal year starts in April, and no one in the UPA government claimed that they engineered any policy
to push the investment rate upwards.
The Vajpayee government was also the first to be fiscally responsible—the central fiscal deficit
declined from 6.4 per cent in 1997–1998 to 4.7 per cent in 2003–2004. However, the largest
economic success of the Vajpayee government was in reducing inflation—between 1992 and 1998, it
averaged 9.1 per cent; in the five years starting 1999, inflation averaged 3.8 per cent.
What went wrong in the economy during these five years? India experienced the worst five year
rainfall in Indian history post 1947—10.2 per cent lower than normal. The average for the previous
forty-nine years—just one per cent below normal. Growth in real agricultural wages averaged a
minus 0.8 per cent per annum. As part of financial sector reforms, the real yield on ten-year
government securities averaged 4.9 per cent per annum; this rate had averaged 2.7 per cent over the
previous decade. In summary, bad weather slowed down agricultural and GDP growth, and growth in
agricultural wages; financial sector reforms kept the investment rate constant, and real borrowing
rates high. The India Shining campaign was initiated in 2003–2004 when GDP growth had galloped
to eight per cent per annum after having averaged only 4.5 per cent during the previous three years
(accompanied by the first year of good weather).
The economy was very likely a factor in Vajpayee’s re-election defeat; electoral politics was
also partly responsible. The BJP lost forty-four seats between 1999 (182 seats) and 2004 (138 seats).
In the coalition battle in 2004, the BJP was near-tied with the Congress (145 seats), but the latter had
more allies. The BJP lost eighteen seats in Bihar (five vs. twenty-three in 1999), and nineteen in Uttar
Pradesh (ten vs. twenty-nine seats). In other words, just two states made up for almost the entire
BJP’s loss in 2004—and defeat in the coalition game.
The Congress gained twenty-four seats in Andhra Pradesh, and eight in Tamil Nadu. The
Dravida Munnetra Kazhagam or DMK changed its alliance from BJP to the Congress in 2004; this led
to a swing in favour of the INC plus of twenty-four seats in Tamil Nadu alone (sixteen won by the
new ally, DMK and eight more by the Congress).
Thus, alliance math, the wrath of god (bad weather), and a disciplined macro-economy were
responsible for the reformer, Atal Bihari Vajpayee’s loss in 2004. In some ways, economic
performance under Modi (low agricultural growth; low growth in agricultural wages; and very high
real interest rates) suggests a close correspondence with the 2004 election. In addition, GDP growth,
while reasonable at around 6.5-7 per cent over the last five years is below expectations. How these
factors play out is considered in Chapters Eleven (see p. 160) through Thirteen (see p. 199). Until
then, like we say, wait till the last vote is cast.

1 Sanjaya Baru, 1991: How P.V. Narasimha Rao Made History, 2016; and Vinay Satpati, Half Lion
—How P.V. Narasimha Rao Transformed India, 2016.
2 Except for agriculture, where heavy State intervention continued. If only Rao-Singh had begun
reforms in this and social sectors like health and education.
3 For the 1980s, the average of 5.3 % excludes two growth outlier years 1982 (2.9 %) and 1988
(10.2 %). For the 1990s, the outlier reform year of 1991 is excluded (1.4 %).
4 The onion-induced CPI inflation of 12.3 per cent occurred in 1998.
5 A large part of the credit for the exchange rate policy should go to the career bureaucrat and ex-RBI
Governor, Dr Bimal Jalan. He came into office at the time of the East Asian currency crisis, and left
with the rupee, inflation, and balance of payments in control.
Ajeeb dastaan hai yeh
Kahan shuru kahan khatam
Yeh manjilen hain kaun si
Na woh samajh sakey na hum….
(Strange are these stories
For they have no beginning, nor an end
Strange are these destinations
Alien to both him and me….)
– Dil Apna Aur Preet Parai, 1960
Sonia & the RSS: Myth or Reality?

This chapter examines two myths which aren’t strictly speaking, myths—one is a personality and the
other, a socio-cultural-political organisation. Supporters of the personality think that it was she who
saved the Congress party, and gave it an expanse which was hitherto unknown. Amongst other things,
her detractors think that her influence on politics is vastly exaggerated, and her influence on economy
policy, less than benign.
Supporters of the Rashtriya Swayamsevak Sangh or RSS think the organisation is the glue that
both holds and holds up Hindu society—and that it is the hand that rocks the BJP’s cradle and has
helped it to dominate Indian politics from 2014 onwards. In the run up to the 2019 elections, the same
group also thinks that if the BJP was to win again, it will be in no small measure due to their efforts,
and support. There is of course an opposite view to this and which is similar to what Sonia’s critics
say about her: an exaggerated sense of influence.
Most of us decode the meaning of reality as: what you see is what you get. And I am sure we are
all in agreement that myths are harder to pin down. Actually, so is reality as was vividly brought
home to the world by Kurosawa in his film, Rashomon. The genius of that film, one of the ten best
ever in any genre or language, was in the way it communicated so brilliantly that while we think we
recognise truth, the truth of reality is often more complex, because each person brings her own
subjectivity to the experience.
However, compared to truth or reality, myth is relatively easier to define—nevertheless, the
dictionary gives at least two definitions. First, a myth is ‘a traditional story, especially one
concerning the early history of a people or explaining a natural or social phenomenon, and typically
involving supernatural beings or events.’ And second, a myth is ‘a widely held but false belief or
idea’. My definition would be closer to the second.
My analysis follows the traditional pattern—documentation of evidence; after all, I am
examining political myths, not religious myths. The latter gets confounded by blind faith, although the
same maybe the case with political myths.
My task becomes easier because politicians have the simplest of goals—to win elections. Let us
examine the winnability of Sonia Gandhi, and the RSS’ impact on BJP’s victory. Apart from politics,
it is also mandatory that we analyse their policies which may or may not result in electoral wins.
As I was writing this chapter, it was announced that Priyanka Gandhi had finally joined active
politics and was given the charge of reviving the Congress party in Uttar Pradesh. The story, and the
evidence, presented below bears an eerie resemblance to the aura created by the party about
Priyanka’s mother decades ago.
At the outset, it is important to understand the circumstances under which Sonia Gandhi was
‘drafted’ by the Congress party. After the 140 seats that the Congress won in 1996 (to BJP’s 161),
Prime Minister P.V. Narasimha Rao resigned as leader of the party, and the Congress chose Sitaram
Kesri, a decades-long loyalist and an OBC to take his place. There are many stories about why Kesri
finally lost out, but one that did the rounds was that he lacked charisma, and apparently mishandled
outside support that the Congress gave (and withdrew and reinstated) to the United Front government
during 1996–1998. Meanwhile, the Congress went back to worrying over its disastrous performance
in the elections—the decline from 232 seats to 140 seats was a major blow, and correctly seen as one
which, if unstopped, would push the party into oblivion.
Since independence, and surprisingly although not mandated, the Congress party has never
elected a leader from outside the family bastion (if a family member was available for election). I
have little hesitation in saying that there has been more than a sense of entitlement of the family in the
generation of this unusual practise. It is also true that the Congress ‘attempted’ to go outside the family
after Lal Bahadur Shastri’s sudden death in 1966 at the age of sixty-six. Amongst several non-family
aspirants to the post (also shorts stints by the relatively unknown, Gulzarilal Nanda), Indira Gandhi’s
attempt to upset the strong patriarchal order of Indian society (and the Congress) by seeking to lead
the party was greeted with derision. The well-known socialist leader, Dr Ram Manohar Lohia called
her a goongi gudiya or dumb doll to highlight her lack of experience and unbridled ambition.
Politics and dynastic rule notwithstanding, Mrs Gandhi’s elevation was the first signal to the
Western world that Indian women could equal men in every aspect, and particularly in the rough-and-
tumble of politics.
Let’s now look at the circumstances which led to the stepping up of Sonia Gandhi. First, there is
no denying that given the culture of the Congress, they had little choice but her. It may be recalled that
Indira Gandhi didn’t need to be drafted in 1966; she had openly challenged the ‘Syndicate’ (old or at
least older men in the Congress party) and emerged victorious. The world watched as Indira proved
her prowess and influence—she fought for respect, and she gained it. In contrast, Sonia was primarily
viewed as the saviour and wasn’t called upon to prove her worth. It is up to history and analysts to
assess whether her influence on Congress’ performance was or is real, or mythical.
The moot point—what does her political record look like? She formally assumed power in
April 1998, but a year earlier (31 March 1997 to be precise), the Congress party under her leadership
(or the formal head, Sitaram Kesri) had withdrawn support to the United Front government. The
Congress obtained the same seats as in 1996 (actually one more), but three ppt less votes than what it
had done under P.V. Narasimha Rao in the 1996 election. The following year, after a full eighteen
months in power, the Congress obtained the lowest seats until then, 114, but kept its vote share equal
to the 1996 level of twenty-eight per cent.
The same story continued in 2004; the Sonia-led Congress won 145 seats and obtained 26.5 per
cent vote share. In the three elections spread over six years, the Congress’ vote share declined by a
few ppt from twenty-nine per cent, and the number of seats stayed constant around 140. By any
objective criteria, one can conclude that Sonia Gandhi had at best a minimal effect on the political
fortunes of the oldest party.
In 2009, the vote share remained just a notch below the 1996 level, but the Congress won 206
seats. In the previous chapter, I mention how most of the credit for the political recovery (albeit
partial) should be directed towards Dr Manmohan Singh. Thus, the bottom line is quite
straightforward—there is nothing in record to suggest that Sonia was a successful politician—except
(and this is important), for a very important fact that she is the glue that kept the flock together.
After the surprise victory in the 2009 election, Sonia and her party set a target of 273 seats.
(While we are keeping score, my record in seat calling for the four ‘Sonia Gandhi’ elections—1998,
1999, 2004 and 2009 was surprisingly good. The record would get even better in 2014.) This hubris,
the belief that only the Congress party has the wherewithal to rule India, has been the Achilles heel of
the party for the last three decades.
The Sonia Gandhi Effect on Policy
The guiding policy in India’s attempt to remove poverty has been to ‘directly’ involve the State. In
contemporary times, the most convincing advocate of direct poverty alleviation has been Prof
Amartya Sen. His contention being that the macro growth process is inherently too slow to help
poverty alleviation. An integral part of this view is that developing economies like India are naturally
prone to greater inequality, and hence, policy has to be extremely pro-active towards pro-poor
growth.
Sonia Gandhi’s economic policies are consistent with Sen’s worldview. She has been at the
forefront of portraying an image that much like her mother-in-law, she is also the champion of India’s
poor. Her belief being that in a predominantly poor country, the poor need to be ‘bought’. Note that
there are two dimensions to the concern for the poor. First, every society, and all political and
economic leaders must enact policies efficiently and fairly to eliminate poverty. It is a well
established fact that whenever poverty is eliminated, according to a given absolute poverty line, the
mandate moves on to improvement and opportunities for the relatively poor. These relative poor exist
in every society (even in Sweden) and at all times (they shall even in 2100).
The second dimension is one of political calculation. Every political leader has to strategise
who or what her core constituency is. For instance, in the 2012 US election, Mitt Romney of the
Republican Party decided that the bottom forty-seven per cent would not vote for him, so why bother.
Sonia Gandhi perhaps believes that India has existed in a time warp for the last forty years and being
pro-poor will guarantee a 273-seat majority.
In the context therefore, Sonia Gandhi strongly advocated two policies—introducing a job
guarantee programme of a 100 days for every rural poor family (the Mahatma Gandhi National Rural
Employment Guarantee Scheme also known as MGNREGA) and the Food Security Act guaranteeing
five kilograms of cereals per month for seventy-five per cent of the population. The latter was an
extension of the (corrupt) Public Distribution System which has been in operation since the 1970s.
For any poor country, food distribution scheme is a very important government policy for the
alleviation of poverty. What the PDS requires is an elaborate government machinery to first procure
grain and rice from the farmers (not to rely on individual agents or the market to procure food since
the market is an ‘evil’ institution), then another government machinery to provide this procured food
to government ‘fair’ price shops, from whom the poor people buy food at a discounted price. Those
buying food grains from such shops have to be in possession of a ration card to make them eligible to
receive food at subsidised rates. I am aware that many readers may well know about this, but then I
owe it to those who may not. Therefore, the entire system seems like an elaborate and meaningful
institutional structure, at least on the surface.
The Employment Guarantee Scheme programme was introduced via an Act of parliament in
2006–2007 and by 2008–2009, it covered the entire country. What comes out clearly in the language
and intent of the Act is that the jobs programme is an income supplement scheme for the poorest of
the poor i.e., those who are in desperate need of incomes and are willing to do unskilled manual work
will be able to obtain work for a maximum of a 100 days. However, job programmes are not new in
India; the first such guarantee scheme was started in Maharashtra in 1973, and therefore unlike what it
was touted to be, there was considerable experience and expertise available with the government to
implement such a programme.
As always, the issue has perennially been about the efficacy of these in-kind welfare
programmes. That such schemes were likely to lead to large scale corruption first became known in
1985, when the newly elected Prime Minister Rajiv Gandhi, after a day-long meeting with several
young bureaucrats proclaimed that several Indian institutions had failed miserably in reaching the
poor. He laid particular emphasis on estimates given to him by field officers that only about fifteen
per cent of every rupee spent for the poor actually reached the poor.
Several studies have documented that Rajiv Gandhi’s estimate was indeed close to the truth,
even twenty-five years after he made the statement. According to the 2011–2012 National Sample
Survey Office or NSSO surveys, about fifteen per cent of PDS money was actually transferred to the
poor, and fourteen per cent of MGNREGA money.
The broad conclusion on the Sonia Gandhi effect on votes, to put it shortly and simply—very
little. Her effect on policies meant for the poor—no improvement in the corruption over the estimate
obtained by Rajiv Gandhi in 1985. But money spent by the government, as a fraction of the GDP, a lot
more and therefore, greater corruption.

There is no question that the RSS is the largest cadre-based organisation in the world. It was set up
with a particular intent and in so far as its core principles are concerned, it has delivered on its
promises. Our focus in this chapter is however to examine the limited question of whether the RSS
was able to exert its influence over and in favour of Narendra Modi in 2014. The near universal
assumption, and conclusion, is that it did; indeed, many argue that without the help of the RSS, Modi
would not have been able to muster a majority for the BJP.
One test of the importance accorded to the RSS is via membership records. Unfortunately, such
records are not publicly available, but we can apply a test which comes close –whether the RSS has
any coattail effect on the fortunes of the BJP. According to the web dictionary, ‘The coattail effect is
the phenomenon wherein a popular politician in higher office attracts votes for candidates of the same
party in lower offices. This usually occurs because voters who greatly support the popular politician
turn out to cast ballots in greater numbers, and voters generally support the same political party
across the board.’ So, does the RSS have coattails? As the above definition suggests, the only manner
in which this question can be answered is via post-poll data.
I have studied the data on the 2014 Lok Sabha election and shall focus on the most
straightforward question put to voters—who did you vote for in 2014? Several other questions were
also asked, but we will examine the answers to this simple question. Each respondent (voter) was
asked about his/her caste, sex, age, and socio-economic background, before asking which party the
individual had voted for in 2014.
The survey results are reported in Table 6.1. Data are reported in three columns for several
important background indicators—and the average voting preference for each category. The data
should be read as follows: the number in the first column is the voting preference of the residents of
Uttar Pradesh (UP) and Maharashtra (Mah). These should reflect the base case i.e., under the
assumption that these states have the maximum number of RSS workers and RSS-aided votes, the BJP
vote in the rest of the country should be lower than the vote obtained in the RSS states. Take the Hindu
UC (Upper Caste) vote. In UP+Mah, 70.1 per cent of the voters said that they voted for the BJP. In the
rest of the country, the average voting percentage was 70.4 per cent. No difference.
 Table 6.1: If RSS not there, higher BJP vote share?

Variable Probability of Voting for NDA

In UP & MH1 Excluding UP & MH Difference

(in %)
Sex
Female 48.7 53.1 4.3
Male 52.5 55.1 2.6
Urbanicity
Rural 51.9 51.5 -0.4
Urban 49.8 55.5 5.7
SC+ST2

No 52.6 53.5 0.8

Yes 41.3 56 14.6
Age Group
18-30 51.6 55 3.4
31-49 49.4 54.4 4.9
>50 51.4 52.2 0.8
Income
Poor 49.3 59.5 10.2
Middle 50.6 53 2.3
Rich 56.3 60.8 4.5
Hindu Upper Caste
No 43.6 49.7 6.1
Yes 70.1 70.4 0.3

Source: Lok Dataset, Post 2014 election survey, September 2014.

Notes: 1) UP & MH refer to the states Uttar Pradesh and Maharashtra respectively.
These two states are conventionally considered as the strongholds for the RSS (Rashtriya Swayamsevak Sangh).
2) SC and ST refer to Scheduled Castes and Scheduled Tribes respectively.
3) Categories highlighted in bold are found to be statistically significant at 95% confidence interval.

Now let us look at the SC/ST vote. In UP+Mah, only forty-one per cent of them voted for the
BJP. Elsewhere (not UP+Mah) there is less RSS support than in UP+Mah; hence, the SC/ST
propensity to vote for the BJP will be less in states other than UP+Mah. We find the opposite, and by
a large amount. There is an extra fifteen per cent of SC/ST vote going the BJP way. This effect is very
large and in the wrong direction. Stated differently, the RSS effect on the BJP vote is negative—when
the RSS is present, the BJP vote goes down, ceteris paribus.
Any RSS Effect on Economic Policies?
The answer is similar to that found in the case of Sonia Gandhi. Let us start at the very beginning to
establish a premise. The Swadeshi Jagran Manch or SJM is the economic arm of the RSS, and holds
near identical views on economic matters as the Communists or the Congress party. A short summary:
anti-Americanism (opposition to the nuclear deal); belief in archaic and outdated labour laws; high
MSP prices for rich farmers as a means of raising farm incomes for all farmers; belief in high income
tax rates; a strong role for the public sector in production, etc. The list is endless.
I have presented indirect statistical (and heuristic) tests for whether Sonia Gandhi has vote-
getting power for the Congress, and whether the RSS obtains more votes for the BJP. In both cases,
we do not find support for the conventional wisdom. I agree: more research and empirical evidence
is needed on this important topic. I hope my unconventional findings will spur more investigation and
debate, for these two myths or reality are here to stay.
Tadbeer sey bigdi hui taqdeer bana le
Apne pe bharosa hai toh yeh daanv laga le….
(Do not deliberate on planning your future
Trust in yourself and take the plunge to emerge victorious….)
– Baazi, 1951
Opinion Polls, Voter Attitudes, and Lies

Perception affects opinion, and opinion results in votes. And votes help win elections. That is the
reality, though one might like to think otherwise. Closely linked to opinion is the recent viral
phenomenon called fake news, a subject covered in the next chapter.
How does the public know about the prevailing opinion? Via opinion polls. But are opinion
polls accurate? If they were, one needn’t wait till election day to find out who won. The 2014
election in India had forecast a hung parliament, and we all know that Narendra Modi won the
election with an absolute majority. In the 2016 US election, only three opinion polls had predicted a
victory for Donald Trump, and all were wrong—they had Trump with a higher vote share; in terms of
votes, Trump lost the election by close to two ppt. So, one may ask and rightly so: what is going on?
What is going on is that Artificial Intelligence (AI) robots do not vote, humans do. People lie—
that is human nature. The purpose of analysis, and inference, therefore, is to separate truth from the
lies in the information contained in opinion polls.
The Facebook-Trump episode1 highlighted how AI and Big Data can be hugely successful in the
identification of true intentions. But I want to highlight an earlier, largely successful attempt, to do the
same some thirty years ago. That story appears a bit later in the chapter—first, how did I get here?
My first indulgence in now a life-long hobby or interest, in interpreting poll data, and
forecasting election results, started in Princeton in 1971 when I was pursuing a graduate course in
Econometrics (a sophisticated term for applied statistics). The assignment at hand from my teacher,
Orley Ashenfelter (who was so good that I asked him to be my Ph.D. thesis adviser—and he agreed!)
was to write a paper on any subject of my choice.
I didn’t have to look far. I just went across the street and spoke to another young professor who
was beginning to make his mark in economic (and political) circles. Those were very early days in
the use of what now goes by the title ‘big data’. The reason: we had very slow computers in 1971,
and big data was defined in terms of the speed of the computer. Prof Ray C. Fair had developed a
model to forecast US election outcomes based on economic variables (it was published in 1972). He
was kind enough to give me access to his data on the US economy (quarterly data on unemployment
and GDP growth) and vote shares in the past US elections. The data were like a toy to a kid—and that
kindled my obsession with the analysis and forecast of elections.
In 1987, I received a call, which was later followed by a visit from a gentleman who would go
on to become one of India’s leading lawyers—Raian Karanjiwala. I was told that V.P. Singh wanted
to be prime minister and if I could help ascertain if he stood any chances. I accepted the challenge,
and not only helped in designing the questionnaire, but also interpreted the results of the trial opinion
poll.
My reward—the 1989 election was in full swing and Vir Sanghvi, then editor of the now defunct
Sunday magazine, was looking for a pollster to analyse the election. I got to meet Ranjit Chib, one of
India’s original pollsters, and learnt about opinion polls. More learning from both Dr Ashok Lahiri
and Dr Prannoy Roy, the pioneers of psephology in India, who were not only my colleagues from the
Delhi School of Economics in the early 1980s, co-founders of the Policy Group, but also very good
friends. With help from all of the above, how could I not do justice to forecasting?
I soon discovered that my tryst with opinion polls was proving to be a trial by fire—Rajiv
Gandhi, winner of the largest electoral victory in India was facing a challenge from V.P. Singh, a
former member of the Congress party who was Finance minister in Rajiv’s Cabinet.
The on-ground buzz at the time: nobody I knew expected Rajiv to repeat his sweeping mandate
of 1984 (49.4 per cent of the vote and 415 out of 545 seats); the conventional wisdom was that he
would win 300 seats. Opinion polls held a month before the election had also made similar
predictions; a private poll for the Congress party also put the number at around 295 seats, and The
Hindu newspaper’s forecast stated that the Congress would manage 267 seats, just short of the magic
number of 273 seats.
Meanwhile, the Sunday magazine’s poll suggested that if the raw data were to be taken at face
value, then the Congress would win by a landslide with a total vote share of fifty-five per cent and
obtain 325 seats.
I met Vir Sanghvi for detailed discussions on the survey. In the questionnaire, I had sought
opinions on various issues; via correlations, I developed a lying index. The lying index was nothing
more than a tabulation of inconsistencies. The net resolution of these inconsistencies led to my final
forecast for Sunday—Rajiv Gandhi and the Congress party will lose the election with a final tally of
215 seats. In the actual event, the Congress obtained 197 seats. My polling career was launched—
first time lucky, but not forever! It is important to understand that the game of making errors is never
over—especially in the sport called psephology, as I was to discover in the 1991 Lok Sabha election.
But first—some fundamentals of election polling.
Polling and Forecasts—The Basics
At one level, psephology seems like the easiest job in the world. Just go out and ask some people
what they think. Forget what they think. Just ask them who they will vote for, and ask who they voted
for the last time. Tabulation of these responses should give you the right answer.
But obviously, it often does not. If it were so easy, then the oft-repeated question to a pollster
would not be—‘So, what happened? Where did you go wrong?’ There are several possibilities of
‘errors’ between prediction and reality. First and foremost, the preference for a particular candidate
can change. Indeed, this is the most common explanation given by pollsters when their forecast goes
awry. Second, a non-scientific sample may have been chosen; not representative of the underlying
population in terms of its composition i.e., age distribution, propensity to vote, sex, income,
occupation etc. Third, there might have been errors in the choice of constituencies chosen for an
aggregate state or national level forecast.
The biggest hurdle however still lies ahead. Most often, public interest is directed at a single
outcome—who will win. In a two-party system, as in the United States, counting of opinions can
closely approximate the votes each candidate (party) will obtain, and hence correctly predict the
winner. But even in the US, there is the problem faced by the Electoral College—votes are not
counted equally, or who you vote for may not be who you get. This is possibly the major reason for
going wrong—the uneasy, and non-linear, relationship between the counting of votes, and the counting
of seats.
However, a two-party contest only involves identifying which party can get more than fifty per
cent of the vote. In a three-party contest, a candidate can theoretically win with a voting percentage in
the wide range of 33.4 per cent and more. Viewed differently, a landslide victory (defined as a two-
thirds share of seats) in a two-party game results with a vote share of around 52-55 per cent. In a
competitive three-party contest, the landslide vote share needs to be as little as 35-40 per cent; and in
a four-party contest, the landslide share need be only around 30-35 per cent. This highly non-linear
relationship between votes and seats is therefore the quicksand. The voting percentage has to be
exceedingly accurate in order for the seat prediction to be broadly correct. Consequently, pollsters
devote extra efforts towards identifying homogeneous ‘vote’ constituencies, and then isolating the
relationship between these constituencies and a state’s voting pattern. Swing constituencies are also
identified i.e., those which most frequently pick a winner. In addition, pollsters try to ascertain the
change in the vote—which party did they vote for the last time, and which party now. It is easy to see,
and understand, how errors in forecasting can occur if a wrong ‘representative’ constituency is
chosen.
Often, erroneous predictions have resulted from both opinion and exit polls. The
excuses/explanations for wrong forecasts most often converge on the first, and the most convenient
explanation is—the voter changed her mind. However, it should be emphasised that this excuse
cannot apply to exit polls. Such polls, by definition, are ex-post, and need to be little more than
simple counting exercises i.e., counts of yes vs. no in a scientifically selected sample. Since the art or
science of sample, and constituency selection is well developed, there are no obvious explanations
for exit polls going drastically wrong. But they do.
Identification of a consistent explanation for varied divergences between forecast and reality is
not a simple exercise, and I have mentioned several sources of errors. But at this stage of the election,
the counting isn’t over yet. There is an additional, and large, source of error identified earlier—the
poll interviewees may lie to pollsters.
People lie. Not all the time, and not even most of the time. Possibly, they lie very rarely. But
when they do, it can cause serious errors. And given the high price of going wrong, it pays pollsters
to heed, and calculate, a ‘lying index’. The back-to-back surprises of Brexit and Trump has brought a
new urgency to assess the magnitude of this lying. Not all people lie, but some do, and perhaps in a
systematic way—but who is lying (apart from the answer to the existential question of why they are
lying), is an important part of the puzzle.
If there is lying, the obvious question is: can it be measured, and measured successfully? The
success of developing a lying index involves several dimensions, but the measure should pass two in
particular: it should broadly reflect lying when there is lying, and not reflect lying when there is no
lying. In other words, there should be no ‘bias’ in the estimate of lying.
The Trump victory revealed extensive use of voter attitudes by Big Data analysts like Michal
Kosinski (as revealed in The Data that Turned the Whole World Upside Down.2 It is worth quoting
the article in some detail:
Anyone who has not spent the last five years living on another planet will be familiar with the
term Big Data. Big Data means, in essence, that everything we do, both on and offline, leaves
digital traces. Every purchase we make with our cards, every search we type into Google,
every movement we make when our mobile phone is in our pocket, every ‘like’ is stored.
Especially every ‘like’. For a long time, it was not entirely clear what use this data could
have—except, perhaps, that we might find ads for high blood pressure remedies just after
we’ve Googled ‘reduce blood pressure’.
On November 9, it became clear that maybe much more is possible. The company behind
Trump’s online campaign—the same company that had worked for Leave.EU in the very early
stages of its ‘Brexit’ campaign—was a Big Data company: Cambridge Analytica.
This quote illustrates the substantive point that the accuracy of opinion polls can be enhanced (as well
as the productivity of targeted advertisements) by knowledge of the opinions and/or attitudes of
voters. However, a respondent’s view is not necessarily related to the direct question of: who you
were going to vote for, or what product you are buying. An early experiment of voter attitudes, and
psyche—and/or lying—occurred in India in the 1989 Lok Sabha election.
Measurement of Lying: India 1989
In October 1989, the then prime minister Rajiv Gandhi called elections a few months ahead of
schedule. Several opinion polls at the time predicted that he would win easily, albeit with a
substantially reduced majority.
However, a deep dive into the Sunday poll data showed an interesting anomaly—while Rajiv
Gandhi’s predicted vote count was high, so were his ‘negatives’—a significant proportion of voters
disapproved of Rajiv Gandhi’s handling of the economy; his handling of communal divisiveness; his
position on issues relating to regional autonomy vis-a-vis three states (Punjab, Kashmir, and Assam);
and his involvement in corruption. These data were therefore highly inconsistent with the voting
pattern which suggested a victory for him.
The reason the above negatives were identified was because ‘lying’ was anticipated in the
design of the questionnaire. Most election opinion polls in the late 1980s rarely asked respondents for
their views on the economy, or their views on social policies.
Therefore, for the first time ever, in addition to the first question: ‘Who will you vote for if the
election were held today?’, the respondents were also asked about their views on political and social
issues. As a rule, some assumptions have to be made regarding the interpretation of views shared by
respondents. In particular, one important assumption is the following: respondents may have a
motivation for lying on the important personal question of who they intend to vote for; but on
political, social, and economic issues, the respondents are more at ease and do not mind stating their
views. It is this difference that allows one to extract, statistically, the percentage of respondents who
lie when asked about their voting preference.
As the above-mentioned premise indicates, one’s voting preference may be a function of the
candidate’s views on various issues, and the voters’ socio-economic-cultural background. For
different people, different factors are paramount. It nevertheless is likely that there may be (is) a
well-defined statistical relationship between one’s assessment of the candidate’s views on various
issues and one’s eventual voting preference. And it is this well-defined structure that helps identify
the ‘outliers’ i.e., those who do not fit the established pattern.
About fifteen per cent of voters lied in the 1989 Sunday magazine poll. As a result, the adjusted
vote for Rajiv Gandhi became forty-one per cent rather than the lying inclusive figure of fifty-five per
cent. The predicted seats were 215, and not 325. Interestingly, in the state elections which followed
in 1990, the lying index was close to zero, and in no state was the lying proportion more than ten per
cent; for most states, the proportion was less than five per cent. We tasted success yet again as the
prediction of seats was accurate. But there was something else which was even more interesting.
Lying Continues—The 1991 Election
After accurately predicting national and state elections in 1989 and 1990, came 1991 and yet another
national election. According to the exit polls, the Congress alliance was set to get only 200 seats,
while the BJP around 160 seats out of the 524 being contested. The final figures were: Congress, 232
seats, and the BJP, 120 seats i.e., the exit poll was considerably off the mark for both the parties. As
mentioned earlier, this election was majorly affected by the assassination of Rajiv Gandhi. While the
exit poll was way off the mark, one opinion poll (conducted before the assassination) had predicted
that the Congress would obtain 240 seats, while another predicted the Congress winning 302 seats.
This poll was the one I had jointly conducted for the Sundaymagazine.3
Even twenty-eight years later, my friends have not forgiven me for the error. I was off by sixty
seats, but had correctly predicted the winner. The error being the same as the exit polls. I kept
thinking: a mere sixty votes made my friends question my credibility. But thankfully, my prediction for
the BJP was close, and most accurate—I had said, 110; it got 120.
The 1991 election was unusual in that for the first time, three strong parties were pitted against
each other—the middle-of-the-road Congress party led by Rajiv Gandhi, the Right-wing (at that time,
nota liberal Right-wing party in terms of economic policies) BJP led by L.K. Advani, and the Left-of-
Centre Janata Dal led by former prime minister, V.P. Singh. The lying index, and seat forecasts, were
readied for each of the three parties in fourteen states i.e., forty-two election forecasts.
While the election results in some states and/or constituencies may have been affected by the
assassination of Rajiv Gandhi, it is unlikely that the overall result was affected in any significant
manner. The lying index was deemed to be significant for only the BJP i.e., it predicted that its
percentage was higher in opinion polls than it was likely to be in reality. At the time, most election
forecasts, as well as the India Today exit poll, had said that the BJP would win in the neighbourhood
of 155 seats. The lying index adjusted forecast of Sunday was 110 seats for the BJP, which finally
turned out to be the most accurate—the BJP won 120 seats. The exit poll forecast therefore was off by
close to thirty per cent in terms of seats.
The Sunday forecast for the states revealed that it was exceedingly accurate in thirty-eight of the
forty-two constituencies. (Table 7.1 maps the results of the 1991 election; the tabulation is by zones.
Note the accuracy of the opinion poll for all regions and states of the country, except two large states
—UP and Bihar. We got it wrong in Uttar Pradesh where we gave too many seats to the Congress and
too few to the Janata Dal, as also in Bihar.)
I remember the post-mortem of the Sundayforecast, and its large pro-Congress error for two
states. We inferred that three major factors may have been responsible for the anomaly. First, the two
states showed the largest proportion of lying in 1989—fifteen per cent in Uttar Pradesh and sixteen
per cent in Bihar. Further, even after adjustment for lying, these states had the largest over-prediction
of seats for the Congress. Thus, there seemed to be a consistent pro-Congress bias in the sampling
(and lying!) in UP and Bihar. Second, these two were the only states in the country facing a three-way
contest in 1991. And in such instances, the lying index may have become intractable i.e., instances of
‘I intend to vote for A; I hate B; but state that I will vote for C’ are very difficult to identify in a
consistent fashion.
The third error was to not take adequate account of the new emerging reality—the beginning of a
structural decline of the Congress party and the emergence of coalition politics. There was an entirely
new shift in the two large states which my opinion poll got wrong. For instance, Mulayam Singh
Yadav and Lalu Prasad Yadav had ejected the Congress in UP and Bihar, respectively. For these two
states, I had forecast that the Congress would win sixty-six seats—it won six, and gave my friends an
opportunity to call me out!
However, the good news was the accuracy with which we could predict the BJP’s seat share;
the Congress error was reflected in the National Front error, the coalition to which both the Yadavs
belonged.

Table 7.1: Seat forecast errors, 1991 Indian General Election

Congress BJP National Front

Predicted Actual Predicted Actual Predicted Actual
North Zone1 43 41 15 19 0 0

– UP & Bihar 66 6' 46 56 25 68

South Zone 108 89 5 5 17 17
East Zone 29 27 1 0 44 43
West Zone 56 56 43 37 2 0
Total 302 232 110 120 88 130

Source: Data obtained from SUNDAY opinion poll, May 1991.

Notes: 1) North Zone excluding Uttar Pradesh(UP) and Bihar; Assam not included in forecast.
2) Actual total is for all states, big and small; forecast only for big states.

My overall forecast for the Congress party was 302 seats; it obtained 232, an error of seventy seats.
Barring the two large states, and I am aware that they are extremely significant for the overall
political health of the nation, I was right for the rest of the states. And it was not as if there were large
errors cancelling out. The error in South Zone—nineteen out of 108 seats; East zone—two out of
twenty-nine; West zone—fifty-six forecast for Congress, and the same obtained.
The 2019 Election—What are the Polls Saying?
At the time of writing this chapter, opinion polls are unanimous in their forecast that the 2019 election
will result in a hung parliament. Yet, ‘attitude surveys’ are sounding positive about Narendra Modi—
but with a large drop in 2018 to a level in excess of fifty per cent. Simply put: the electorate is
polarised about the personality of the prime minister. Is this emerging as the second 1991 election in
terms of opinion polls, and forecasting?
Table 7.2 summarises what the opinion polls are saying. In August of 2018, and before the state
polls in November, the NDA was sitting comfortably with a majority. But there has been a decline
from August 2018 to now; it has been uneven, but from peak to trough the decline is about fifty seats
for the NDA. This is mirrored in the increase for the UPA, with others staying broadly constant at 140
seats.
Part of the decomposition between NDA, UPA and others is affected by the party alliances
assumed by any pollster. What is more meaningful is to look at the seats to be obtained individually
by the two major parties, BJP and INC. The August 2018 India Today poll had the BJP obtaining 194
seats and the INC ninety-six (NDA, 281 and UPA, 122 seats). The January 2019 poll by the same
organisation saw the BJP seat share improving marginally to 202 seats, and the INC staying constant
at ninety-seven seats. The latest opinion poll, Times Now-VMR, has the NDA at 252 seats, but it does
not report a tally separately for the BJP or INC. The fifteen seat increase for the NDA is mostly in the
BJP camp, so we will place the ‘reading’ of this poll for BJP at 217 seats. Going into the election, the
base opinion poll forecast for BJP is 209 seats (average of India Today and VMR) and the base result
for INC is ninety-seven seats.

Table 7.2: Opinion polls for Indian General Election, 2019

Date published Polling agency Projected Seats

NDA UPA Others

2019
Jan-19 Times Now–VMR 252 147 144
Jan-19 ABP News–Cvoter 233 167 143
Jan-19 India Today–Karvy 237 166 140
Jan-19 VDP Associates 225 167 150
2018
Dec-18 India Today 257 146 140
Dec-18 ABP News–C Voter 247 171 125
Dec-18 India TV–CNX 281 124 138
 Nov-18 ABP News–C Voter 261 119 163
Oct-18 ABP News–CSDS 276 112 155
Aug-18 India Today–Karvy 281 122 140

Source: As reported in the media.

Opinion Surveys
There are two sources of information on opinions about the performance and popularity of Prime
Minister Modi. If individual data were available, this would go into the construction of the lying
index for 2019. The first source is the Pew Research Center, one of the most respected survey
organisations on global and individual country attitudes. In their 2017 report4, the following statistic
was cited: seventy-eight per cent of the public had a favourable view of Modi in 2013, and that this
had risen to eighty-eight per cent in 2017. In their latest global survey on economic conditions5, only
fifty-six per cent said that their economic condition was satisfactory, while eighty-three per cent
thought so in 2017. This is a large decline, indeed the largest in 2018 across the world. On an
absolute basis, India was well above the median answer of forty-five per cent for twenty-eight large
global economies (China was not part of the project).
In the first edition of Firstpost, now a weekly magazine, survey results on ‘trust’ for India were
reported. The survey was conducted by Ipsos, another world attitudes pollster. In January 2019, it
undertook a large scale attitude on trust survey. On leadership qualities, Modi scores higher than any
of his political opponents, and scores higher than Rahul Gandhi by a two-to-one margin—52.8 per
cent for Modi, 26.9 per cent for Rahul Gandhi. On the issue of trust, Modi is trusted by 74.4 per cent,
with the principal Opposition party, the Congress scoring 53.3 per cent.
Thus, what one can say going into the 2019 election, and proceeding towards our discussion of
vote forecasts in Chapters Twelve and Thirteen (see pp. 181 and 199) is the following: there has been
a decline in Modi’s popularity over the last year, and that the 2019 election is no longer a sure bet for
either him or the BJP. The Indian polity is also possibly more polarised than ever before, and hence
all forecasts are suspect, especially those which do not explicitly account for lying. I mentioned at the
beginning of this chapter that this is a dream election for a pollster, especially if she gets it right!

1 It is believed that President Trump’s election campaign employed the use of ‘intelligence’ from
Facebook posts to target individuals with fake news so that they would not vote for the Democratic
candidate, Hillary Clinton.
2 https://motherboard.vice.com/en_us/article/how-our-likes-helped-trump-win
3 My co-author in the prediction—my good and very smart friend, Sanjiv Kumar, a student of mine at
the Delhi School of Economics, and my boss in the Treasury Department of the World Bank.
4 Pew Research Center, Brue Stokes et. al., ‘Three years in, Modi remains very popular’, 15
November 2017.
5 The question asked was, ‘the current economic situation in our country is good.’
Kal ki daulat, aaj ki khushiyan
Unki mehfil, apni galiyan
Asli kya hai, naqli kya hai
Poocho dil se mere…
(The money they earned yesterday, gives them pleasure today
Their great gatherings; our lives on the streets
What is real, what is fake
Ask my heart for the truth…)
– Asli-Naqli, 1962
The Birth of the Fake & Faulty Twins—News & Analysis

One of the biggest transformations in the world of the last decade is the explosion of instant
information and dissemination. This phenomenon is met by the rise of Internet usage, and use of smart
phones. It is now an accepted norm that the current US President only communicates via Twitter, and
in India, the Finance Minister writes blog posts. A decade ago, this would have not only been
considered as breaking protocol, but also an impossibility.
Everyone is in the same game. Winning friends and influencing people has been converted into
convincing people to vote for you (or buy your not-so-different product). It would be rather foolish to
not accept that we live in the age of information, an era where marketing is king. At the same time, the
opportunity cost of time has increased, and individuals have a lot more things to do. The time taken
for ingesting information however is a luxury, and making quick decisions has become mandatory.
These in turn are based on all the information available, and the computer in our bodies sends out
information to the brain—‘You should vote for X.’
So what is it looking for a voter who shall decide who rules New Delhi in 2019? If we were to
do a rough and quick profile of such a person, then here is a person who experienced sixty odd years
of dominant rule by the Congress party, and nearly a decade (in two five year halves) of rule by the
BJP. This voter is also considerably more educated, as more than twenty-five per cent of those in the
eighteen plus years age-group are now going to college. What also sets them apart is that they are not
only consuming a lot of information, but are able to translate that experience into intelligent decision
making.
However, despite all the information on hand, and the ability to take clear-cut decisions, what
she faces is the eternal conundrum: who should I vote for? For the Congress party, which she
traditionally may have done till 2014, or gravitate towards the BJP? That is the question.
The answer obviously depends on factors which shall influence her final decision. And here, the
news is good for informed decision making. According to international surveys (the Pew International
Center, for instance), Indians have a much larger faith in democracy than their ‘equivalent’
counterparts in the rest of the world. Recent research has also shown that India is more urbanised than
conventional data will have us believe—the official urbanisation rate is nearly twice the official
Indian definition of thirty-one per cent urbanisation in 2011–2012 and nearly thirty-four per cent in
2017.1
It is also the case that the rate of turnout has steadily increased in India and reached close to
seventy-five per cent in the 2018 state elections.
Yet another point which impacts the voter turnout number is that Indian laws make it very
difficult for migrants to vote—make it impossible, in fact. The migrant has to return home to vote.
More than one-third of the Indian resident is a migrant, and about ten per cent of these are from
outside the state which goes to elections. If no migrant returns home to vote, a seventy-six per cent
voting rate (as observed in recent elections) is a near perfect 100 per cent intention to vote.
Obviously, many migrants return home to vote—the only point I wish to make is that a very large
proportion of Indians actually vote.
This fact is not lost on our politicians. The received wisdom is, and supported by data, that the
urban voter is more likely to vote for the BJP than vote for the Opposition. The received wisdom also
is that the macro-economy has performed better under NDA regimes than UPA rule—and especially in
case of inflation. Both in 1998 and 2014, the NDA regime had to battle with the overload of bank
NPAs bequeathed to them by the Congress-United Front coalition, and the UPA respectively; and these
NPAs were, in both cases, in the double-digit range.
Let’s make no mistake, the voter understands the economic performance of both regimes in
minutest detail. Somewhat surprisingly, despite an impressive economic performance, the voter did
vote against the NDA in 2004. The Opposition hopes that she will do the same in 2019.
This is where fake news, and its twin, faulty analysis, comes in. They may be siblings, but there
is an important distinction between the two. Fake news is used to move sentiment against the
incumbent or the person in ‘lead’ position. For instance, the US election, via Cambridge Analytica,
seems to have been genuinely affected by the wild accusations against Hillary Clinton, which were
absorbed by an ever-willing electorate eager to fall for the narrative.
The former MP from Odisha, Jay Panda, articulated the role of fake news in post 2014 state
elections as follows:
In any event, it is surely no coincidence that, real or exaggerated, this narrative of rising
intolerance has been peaking around elections. Just as in the earlier phase of reported church
attacks bunched around the Delhi state election, so too now the crucial Bihar election is
undoubtedly a catalyst. The bigger question is, for whom? For no one side or party has a
monopoly on such tactics.2
There has now begun a clampdown, and self-regulation, by organisations in both the US and India. In
what seems to be a concerted effort, companies such as Google and WhatsApp etc., are ‘controlling’
the spread of fake news. But what this also suggests is that fake analysis will spread, and indeed has
already been used extensively in India.
A few pointers first. What is fake analysis, and how does one distinguish it from faulty analysis,
and how does one distinguish it from non-faulty, non-fake analysis? It is difficult, but not impossible,
and sometimes it is easy to identify.
Don’t Confuse Me with Facts, My Ideology is Made Up
The difference between fake and faulty is in motivation. If I knowingly commit a mistake in the
analysis, then it is fake; if I make an honest error, it is faulty. An extreme and exaggerated way to
recognise fake analysis is to understand the distinction between fools and knaves. Fools make honest
errors; knaves do fake analysis. All of us at one time or another have been ‘fools’.
A majority of fake news and/or fake analysis centres around the economy—and somewhat not
co-incidentally, fake news involves the conclusion that the economy under Narendra Modi is doing no
better than it was under UPA-I and UPA-II. The Opposition feels that if the public is made to believe
in this simple ‘fact’, then the political objective of winning the next election is more than half
achieved.
We will look at several, approximately a dozen, economic facts, and the associated fake
analysis. I mentioned in the Introduction to this book that India invented the political use of fake
news/analysis. Now I present the proof.
Evidence on Fake News/Analysis

1. GDP data are manipulated

It all started in January 2015, barely seven months into Modi’s prime ministership. The new GDP
series was finalised and published by the Central Statistical Organization (CSO) in end January 2015.
One point needs to be clarified and only because we tend to smell a conspiracy almost everywhere—
the entire staff, bureaucrats, and the economic experts involved in the construction of the series were
appointed by the previous UPA government. The chairman of the National Statistical Commission,
which approved the new series, Dr Pronab Sen, was a UPA appointee. The new 2011–2012 base
GDP series had been vetted by international organisations and conformed to the system of national
accounts laid down by the UN. Staff members of the World Bank and the IMF were actively involved
in the construction and revision of the GDP series. It is not even clear how many of these international
experts were of Indian origin, let alone sympathetic to the new government.
The new series showed that GDP growth during the last three years of UPA-II (fiscal years 2012
through 2014) were lower than previously estimated. The debate that followed was unique to India,
and the world, because GDP changes to a new series are routine, and had occurred at least six times
in India and never before with any discussion, let alone debate. Count them—GDP base year changes
occurred in 1960–1961, 1970–1971, 1980–1981, 1993–1994, 1999–2000, 2004–2005, and now
2011–2012. I challenge anyone, and everyone, to show me one instance of any discussion on base
year change in India.
Yet what transpired on release of these data was extraordinary. For the first time, official
government statistics began to be questioned. It is well known that GDP, and other government
statistics, are problematical around the world. They go through revisions, and more often than not,
lead to surprises and raised eyebrows. But malpractice has never been the issue. Except in 2015—
and every data release thereafter. Why did it happen in India, and only post 2014? If you have a better
explanation than politics, and political ideology, I am ready to listen.

2. GDP back series understates true GDP growth in UPA years

Three years later, in mid-July 2018, a draft report of a government appointed committee headed by a
leading economist (and a very good friend), Dr Sudipto Mundle, was leaked to the press. This draft
report controversially (and on a preliminary draft basis) concluded that prior to 2011–2012, GDP
growth under UPA-I and UPA-II was higher than that indicated by the old 2004–2005 series. The
same experts who had criticised the new GDP series as faulty (and politically motivated) now had no
hesitation in accepting a faulty interpretation of the back series constructed with the new series.
Subsequent to the Mundle draft report, the CSO released its own official estimate of GDP
growth between 2004–2005 and 2011–2012. This new 2011–2012 base back series lowered GDP
growth relative to the 2004–2005 series. I will now get into some technical details (documentation of
evidence) which strongly indicates that the lowering of GDP growth for 2004–2005 to 2011–2012
was entirely to be expected, and primarily because of the surprise low employment growth between
2004–2005 and 2011–2012. This low employment growth only became known in 2013 and was
added as an input into the revisions. Let me explain.
 There is one large sector of the economy—Wholesale and Retail Trade (WRT)—whose GDP
estimation was directly dependent on NSSO employment data. This sector, in the old 2004–2005
series, accounted for sixteen per cent of GDP; in the new 2011–2012 series, it accounted for only
eleven per cent. What was the reason for this large decline in the share of this sector?
Let us look at the traditional method of estimating the contribution of the WRT sector to GDP. On
the release of the NSSO employment data (approximately every five years), the CSO looked at
employment gains of this sector and arrived at an estimate of the sector GDP. In 1999–2000, there
were 34.4 million people working in WRT, which increased to 41.7 million in 2004–2005, yielding
an annual growth rate in employment of 3.9 per cent per annum. For the 2004–2005 to 2011–2012
period, the growth in WRT employment collapsed to only 0.2 per cent per annum. This is why the
share of this sector was reduced to eleven from sixteen per cent.
A crude back of the envelope calculation shows that GDP growth between 2004–2005 and
2011–2012 was over-estimated by 0.6 per cent per annum by the earlier 2004–2005 series.3 There
are many other ifs and buts in estimation of GDP data; these debates are normal and healthy. As the
old saying goes, if you have two economists, there are three conclusions. But what is a first in India,
and it has happened post the 2014 election, is that constructive debate has now yielded ground to
ideological disagreements; worse, laced with dirty politics.

3. Investment decline under NDA

Another fake news favourite is that investment (ratio of gross fixed capital formation to GDP) rates
declined under NDA and at the time of writing this chapter, are even below the UPA levels inherited
by Modi. In 2013–2014, the year preceding UPA’s defeat, investment was 31.3 per cent—the latest
data for 2017–2018 shows that this share has declined to 28.6 per cent—a 0.7 ppt decline for each
year of the Modi government. Hence, it does appear that the investment rate has declined.
One hidden (or squirreled away) fact from the discussion is the reality that the BJP inherited a
broken state banking system (also known as very large NPAs) and that this may have led to the
decline in investment shares. Nevertheless, a close perusal of the data seems to suggest that contrary
to fake news, investments have done reasonably well under the Modi dispensation. Furthermore,
much like any other economic variable, investment should be measured in real terms i.e., after
adjustment for inflation. If that is done, then the investment share in GDP (2011–2012 prices) was
31.1 per cent in 2013–2014, and, after reaching a low of 30.7 per cent in 2015–2016, is now at 31.4
per cent—an increase of thirty basis points over four years rather than a decline of seventy basis
points per year. A very different picture is painted by real investment rather than the flawed nominal
share of investments—yet it is nominal investment share that is being used to show that GDP growth
cannot be seven per cent plus as stated by the CSO, but is likely to be lower.

4. Demonetisation—an economic disaster

A constant theme running through the 2019 election narrative of the Opposition is how the ‘draconian’
demonetisation policy of November 2016 brought unprecedented pain for the ordinary citizen, and the
voter should (will) punish the BJP for it. Over the next few years, I am sure that several Ph.D. theses
examining the political and economic impact of demonetisation will be written. For the moment,
however, I just want to concentrate on the fake or faulty analysis on this subject.
 I remember 8 November 2016, the day demonetisation was announced. I was on my way to the
studios of India Today for a discussion with fellow political and election junkie, Rajdeep Sardesai
on the Clinton-Trump election. I had written a few articles on how Clinton would easily win the
election. I obviously got it wrong, but before you give up reading my forecasts, consider this—only
three opinion polls had Trump winning, and all of them had him winning with a higher vote share.
That, as it turned out, didn’t happen. As Bill Clinton might say, It’s the Electoral College, stupid! In
reality, all opinion polls were wrong, as was my forecast. My analysis showed that women,
especially college-educated white women, would vote in larger proportions for Clinton. That was not
to be, so my second major poll forecast went awry—the first boo-boo was the 1991 Indian election
as discussed in Chapter Seven (see p. 94).
Meanwhile, my studio visit was cancelled and I sat at home answering Rajdeep’s question—‘Dr
Bhalla, what do you think of Mr Modi’s demonetisation announcement?’ I had had fifteen minutes to
think it over. In 2002, I had written a background note for a Ministry of Finance Task Force on direct
taxes (one of the first ever) headed by one of India’s leading reform economists, Dr Vijay Kelkar,
which examined the pattern of income tax evasion in India. The magnitude—only fifteen per cent of
tax revenue owed to the government was received by the government.4 This low percentage, I had
added, was due to a combination of two realities—many individuals who should have paid income
tax did not; and many individuals paid less than what they should have actually paid. In 2006, the
Internal Revenue Service (IRS) began publishing the magnitude of tax evasion in the US on its
website. The findings: about eighty-two per cent of income tax due was paid.
As is well known, there is tax deduction at source in India for salaried workers, and hence they
cannot avoid paying taxes. The individuals who did not pay the due taxes in India were the white-
collared professionals (surgeons, doctors, lawyers, accountants, stock brokers, etc.) and of course,
shopkeepers and owners of small businesses. The large business tycoons have also avoided paying
their due share of tax, but that is a different tax (corporate income tax) and not considered here.
Over the years, I consistently updated the tax compliance ratio and found that it had remained
stuck at around 15-20 per cent. In January 2017, I updated the analysis along with Arvind Virmani, a
close friend, an ex-World Bank colleague, and a diligent and dedicated economic reformer, and one
who has the huge advantage of having worked with the Indian government for over twenty years. This
update, which was undertaken two months after demonetisation, showed that income tax compliance
in 2015–2016 was still extraordinarily low at around twenty-five per cent.5
Back to what caused this digression in the first place. When Rajdeep asked me about
demonetisation, I blurted out that it was a good policy move because it would help raise the tax
compliance ratio in India. Black money is tax evaded money—hence, a decline in tax evasion is a
decline in black money.
In recent years, there has been a fundamental shift in income tax compliance in India. In three
years, direct income tax compliance has increased to at least the mid-thirties, a ten ppt gain. The
number of tax payers has doubled in the last five years, from thirty-eight million in 2013–2014 to
about seventy-six million in 2018–2019. It is very difficult to not acknowledge the role that
demonetisation has played in bringing about this fundamental revolution in tax payment, and tax
compliance.

5. Demonetisation failed—all cash is back

Popular news is all about the failure of demonetisation. The fake news evidence—cash is back. At
the time of demonetisation, cash with the public was around eighteen trillion rupees, of which 15.3
trillion was demonetised (all five hundred and thousand rupee notes). As recently as February 2019,
the cash in the system was also eighteen trillion rupees, and the argument now made is that this means
that demonetisation came a cropper. The proof that this is fake news/analysis is embedded in this
statistic. There was no single argument to the effect that cash will not return to the pre-demonetisation
level. Or even exceed it.
The demand for cash is part of a transactions demand for money, a subject that has been studied
for centuries. The issue is of excess cash in an economy. The proof that cash to GDP levels in India is
substantially lower than before demonetisation is given by the simplest of all statistics: cash is now
back to demonetisation levels, even a few per cent above it, but nominal GDP is about 15 to 20 per
cent higher. QED.
Yet those opposed to the present government, aided and abetted by financial experts who should
know better, continue to peddle the lie (fake news) that demonetisation failed because all the
confiscated cash, and more, is back.

6. Government is overstating GST revenue targets

One of the most significant economic reforms in India was the introduction of the Goods and Services
Tax (GST) in July 2017. This is a national tax, and required passage in both houses of parliament, and
was supported by all political parties. Like everything else in India, although its execution has been
less than optimal but in no way should the performance of GST enter into the fake news category.
Whatever be the theory –it has worked, not worked as well as we expected, or any combination in-
between—it must be a subject of debate, discussion, and evaluation, and in ordinary times, should not
even be a subject of faulty analysis, let alone a fake one.
We return to the recurring theme yet again, and that being that the 2014 loss has induced a spirit
of fakeness in the Opposition. At least that is what the data shows. Observe the following—no less an
expert as former Finance Minister P. Chidambaram wrote the following about GST revenue estimates
for 2018–2019,6 ‘Based on revised estimates, it appears that …GST (revenue) will grow at a
whopping 67 per cent… Notwithstanding the doctor’s concerns, a very ill person has left an
aggressive will.’
The GST revenue was budgeted to increase from 444.6 trillion rupees in 2017–2018 to 743.9
trillion in 2018–2019. That is a sixty-seven per cent gain and Chidambaram is right in his math, but
definitely not in his logic. He omits to clarify that the 2017–2018 GST revenue was for eightmonths
(July 2017 through February 2018) at 55.6 trillion a month. The Budget 2018–2019 target is for
twelve months at sixty-two trillion a month. Thus, the GST Budget revenue increase was targeted at
10.7 per cent on the correct, non-fake, month to month basis.
Job Needs & Job Creation
Let’s now tread into real choppy waters, and focus on some facts on jobs. Prior to 2009–2010, there
was only the NSSO employment survey every five years. Since 2011–2012, there have been two
official surveys on employment—one a comprehensive NSSO survey in 2017–2018, but now called
the Periodic Labour Force Survey or PLFS, which will henceforth be conducted annually. Second, a
national labour force survey, which has been undertaken every year since 2010–2011—it is
conducted by the Labour ministry and is called the Employment-Unemployment Survey (EUS). The
EUS surveys were also national, but not as ‘exhaustive’ as the NSSO survey and are therefore being
discontinued.
While this might change by the time you read this book, the reality is that the PLFS data and
report has not yet been released officially. Instead, what we saw was leaked news report on some
selected contents of the report. The leak occurred simply because of the reluctance on the part of the
government, and that for me is good non-fake news—the media is alive and well and kicking and
screaming and correcting the government when it is manifestly wrong.
What has this arcane detail got to do with fake news? Everything. Job creation, or lack thereof,
is big news in this election year. In this chapter, I have discussed how esoteric concerns with ‘base
year’ is big fake news—we will now examine how ambiguous and spotty employment data has added
grist to the fake news mill.

7. Where are the jobs?

The Opposition has consistently claimed that Narendra Modi and the BJP promised to create ten
million jobs a year in their 2014 election campaign. The Congress’ prime ministerial candidate,
Rahul Gandhi, has upped the ante (fake news) by claiming that Modi had promised twenty million
jobs a year!
The Indian economy added a 100 million jobs over a space of thirty years (1993–1994,
employment of 330 million) and 2011–2012 (429 million). Between 2009–2010 and 2011–2012, one
million jobs were lost.7 Is it even plausible that the BJP promised to create even ten million jobs a
year, a promise that would increase jobs at nearly twice the fastest rate ever experienced (between
1983 and 1993–1994, sixty-two million jobs were created over a space of 10.5 years). A simple
lesson about spreading a lie is that it should be credible.
How much of fake news is Rahul Gandhi’s job claim promise? In the BJP Election Manifesto
2014, there is the following statement about jobs: ‘The country has been dragged through 10 years of
Jobless Growth by the Congress-led UPA Government.’ At a campaign rally in Agra in August 2013,
Narendra Modi did talk about the lack of job generation in the UPA years, but there is no reference to
the promise of creating ten million jobs per year anywhere in the media. Even after being challenged
to provide evidence, the Congress and/or its spokespersons have refused to comply.
What we do know is that precious few jobs were created between 2004 and 2011 (ten million
over seven years). So why did the Congress party’s seats jump between 2004 and 2009 (from less
than 150 to more than 200)—because they had created so few jobs? The average rate of job growth
was 2.4 per cent per annum in the NDA years spanning 1999–2004, and the BJP lost the election. The
average annual job growth between 2004 and 2009 was only 0.5 per cent per annum, and the
Congress had its best showing since 1991.
 The link between job growth and election victories is, at best, a puzzle; at worst, it is an
unknown. However, in the era of fake news, job creation, or lack thereof, appears to be solid
campaign material for the Opposition.

8. Demonetisation led to a loss of jobs

A private data company, Centre for Monitoring Indian Economy or CMIE, has been producing
employment-unemployment reports thrice a year—January-April, May through August, and September
to December. The survey started in January 2016 and in mid-2017, the CMIE announced that 1.5
million jobs were lost in the first four months of 2017. The BJP had just won the UP election by an
overwhelming margin in April 2017, and the CMIE report provided headline ‘evidence’ that
demonetisation had not worked.
I am afraid that both the headline and the analysis were misleading. What the author had done
was to compare the change in employment between September-December 2016 and January-March
2017. This comparison was definitely not kosher i.e., it did not account for seasonality in
employment. Fortunately, the CMIE data (posted on their website) allows one to compare like with
like i.e., employment (and unemployment etc.) in the eight months before demonetisation, and the first
eight months post demonetisation from January-August 2017.
The results obtained were in complete contrast to the interpretation posted by CMIE.
Employment had actually increased at a healthy rate in the eight months post demonetisation,
compared to the eight months before demonetisation. For the population aged between 15-24,
employment increased by seven million, and for the adult age group 25-64, a healthy 12.7 million, or
at a rate of 3.7 per cent, the highest over the last thirty-five years (previous years, NSSO data).
Contrast this with the one million plus jobs lost due to demonetisation announced by CMIE. Faulty
analysis—maybe?

9. India needs twelve million jobs a year—really?

For economists and politicians, an important concern is the equation between jobs needed and jobs
created. The run-rate (job need rate) is the number of jobs created per year that will keep the
economy at a given ‘neutral’ rate of unemployment. In a 2017 report8, the World Bank estimated that
India needed close to eight million jobs a year. On the other hand, Dr Raghuram Rajan, distinguished
economist and former Governor of the RBI, in a speech at Berkeley on 10 December 2018, stated that
India needs to ‘employ 12 million people coming into the labour force every year.’
For the relevant age-group >=15 years, population per se is expanding at a rate of fifteen million
a year. The average all-India fertility rate is close to replacement level at 2.1 births per woman. The
15-24 age group will expand by only 0.4 million per annum over the next five years. Over the last six
years, the population above fourteen years expanded at fifteen million a year. Enrolment in schools
and colleges expanded at six million a year. Hence, population net of education expanded at only
nine million a year. The labour force participation of this group, at a high fifty per cent rate, yields
only 4.5 million jobs as needed per year. With increased labour force participation expected in the
next few years, the expected figure is 5.1 million jobs a year—not eight million, and definitely not
twelve million.
Non-Economic Fake News
As in the US, fake news is often of the non-economic kind and therefore subject to multiple
interpretations. In this chapter, we have only considered economic fake news, the kind that can easily
be exposed. In Chapter Eleven (see p. 160), we shall examine news pertaining to attacks on civilians
(e.g., communal riots) to establish how much of the ‘news’ on communal equations, and fear, is real
or fake.

1 See Reuben Abraham on urbanisation, 2018.

2 Jay Panda, ‘Perceptions Matter’, timesofindia.indiatimes.com, 28 October 2015.
3 This is just a crude calculation and there were other methodological changes in GDP estimation
(most importantly, in the measurement of growth in manufacturing). I only want to highlight the effect
of WRT employment on GDP growth. Previously, we had believed that 16 % (WRT) was growing at
3.9 % per annum, or contributing 0.62 % to GDP. Now, we know that WRT growth was only 0.2 %
per annum, and its weight in GDP was reduced to 10 %, yielding a .02 % contribution. The difference
between the two is 0.6 % or 60 basis points of lower GDP growth than previously believed. Note that
productivity growth is assumed to be the same in the two calculations.
4 See Table 1.3. p. 10, in Report of the Task Force on Direct Taxes,Ministry of Finance, December
2002.
5 See Surjit S. Bhalla & Arvind Virmani, ‘Towards an Income Tax Revolution’, The Indian Express,
14 January 2017.
6 ‘Across The Aisle: Good doctor, bad patient’, The Indian Express, 4 February 2018.
7 See Bhalla-Das, 2018.
8 Andres et. al., (2017), ‘Precarious Drop: Reassessing patterns of female labor force participation
in India’.
Tu Hindu banega na Musalman banega
Insaan ki aulad hai insaan banega…
(You shall grow up to be neither Hindu nor Muslim;
You shall grow up to be human; just as you were at the
time of your birth…)
– Dhool Ka Phool, 1959
Mandir, Cows & Guns

If it is election season, can religion and cows be far behind? Or communal politics? The largest
minority in India are the Muslims, constituting approximately 14.2 per cent of the population. With
this population size of 186 million, India is the second largest Muslim country (Indonesia being the
first with 230 million).
India has two large vote banks—the Scheduled Caste and Scheduled Tribes, and Muslims. There
is considerable evidence to show that the SC/STs are not an ‘assured’ vote bank for any party. It is
nowhere apparent that they vote as a bloc; it varies from constituency to state.
That has not been the case with the Muslim population—they have consistently notvoted for the
BJP (around 8-10 per cent in all the elections since 1998, including 2014)1, and were a consistent,
and reliable supporters of the Congress party till the late 1980s. Since then, they have shifted their
allegiance to different regional parties.
In close elections, and especially where Muslims form a sizeable chunk of the population
(around 20-40 per cent), the effect of a bloc Muslim vote can be important. Since independence,
Muslim appeasement has been a recurring theme for most political parties, and if that in itself is not
overtly opportunistic, it is also bad politics. The latest proof of this tendency, especially within the
Congress party: the recommendation by Rahul Gandhi to withdraw the Triple Talaq bill when it
comes up for vote in the Rajya Sabha in 2019. Any similarities between this non-liberal view and the
appeasement politics of Rajiv Gandhi in 1986 is not co-incidental—everything is fair in elections,
stupid.
As is well known and something which is obviously unfortunate, there are several fault lines in
the Hindu-Muslim divide, but the two most divisive components in recent history are—the Ram
temple in Ayodhya, and cows. That these should be central issues in the fastest growing economy in
the world, is both counterproductive and anti-development. Both have grave political and electoral
implications.
Temple Politics
The disputed Babri masjid or mosque in Ayodhya stood at a location (till it was razed to the ground
on 6 December 1992) which traditional Hindus have always believed to be the birthplace of Lord
Ram. That in essence is the kernel of the mandir-masjid dispute which has gone on for seventy long
years.
According to several devout Hindus, Lord Ram, who is an avatar of Lord Vishnu and hence one
amongst the trinity including Shiva and Brahma, needs to be accorded His rightful place in the land of
His birth. They therefore want to rebuild the temple, while the Muslims want to retain ownership of
the mosque.
After the Somnath to Ayodhya rath yatra undertaken by L.K. Advani in 1990, the temple issue
had whipped up communal passions to such an extent that the nation was stunned into silence after the
demolition of the disputed mosque. It is therefore true that the BJP had raked up the temple issue to
obtain political gains. But there is more to this case.
It was none other than Rajiv Gandhi who had made the first blunder and created a schism
between the two communities. The 1985 Shah Bano judgement was evidence of his effort to appease
the conservative Muslims and when he followed it up by opening the locks of the Babri masjid, he
played into the hands of fundamentalist Hindus as well.
Since then, the battle-lines are clearly drawn between the two communities—the focus being on
restoring dignity to their respective gods.
The Ramjanmabhoomi-Babri masjid dispute has not only led to unprecedented communal strife,
but the involvement of the Supreme Court at regular intervals. At election time, the issue gains
prominence, although it is heartening to note that the Vishwa Hindu Parishad or VHP, an affiliate of
the RSS, announced that they won’t agitate on the temple issue until after the elections.
My personal view has always been that in the interests of unity, harmony, peace, and all things
good, the Muslim clergy, or whoever is in charge, should consider yielding to the Hindu demand for a
Ram temple. It is rather unfortunate that very few (any?) of the liberal Muslim intellectuals and artists
have come out openly in favour of this proposition and only because it is for the larger good of the
nation, which belongs to all, but foremost deserves peace and development.
I concede that religion is important around the world, and particularly in India. I grew up in a
Sikh-Hindu household, and in a very religious home at that. Both my parents were observant, but
never once did they impose their religious beliefs on the children or on each other. On the few
occasions we went to a temple, it was to a gurudwara. But during all festivals at home, particularly
Diwali, we prayed to Lord Ram. My parents were emphatic that religion was about a personal belief.
It is therefore said, and not incorrectly, that Hinduism is a way of life.
Be that as it may, while the Muslims won’t give up on the Babri masjid, Hindu politicians want
to keep the issue on a boil to make political gains. Just imagine the aftermath of the counter-factual.
What if both communities yield to the other’s demands and adopt a conciliatory stand. I am sure that
the political narrative shall do well to move on from what has come to be a millstone around our
necks and focus on other urgent issues leading to better governance.
Cow Politics
The other big divide in Hindu-Muslim relations is politics in-the-name-of-the-cow. What needs to be
recognised by the large Hindu majority is that the ban on cow slaughter is in equal parts, if not more,
responsible for divisiveness than the mandir. It is also an issue whose history is misunderstood
across party lines.
In January 2019, Kamal Nath, the newly installed Congress Chief Minister of Madhya Pradesh,
invoked the National Security Act to arrest people suspected of cow slaughter. You read it right—it is
the draconian Security act that is used to arrest people suspected of cow slaughter. I am sure it is
largely believed that only the RSS (and the BJP) think in these terms, but if even the so-called
‘liberal’ Congress party is thinking, and acting similarly, what hope is there for the poor cow and the
politics around it? Kamal Nath also hinted that he might implement a cow tax to build cow shelters. It
is little wonder that his proclamations sound eerily similar to what Yogi Adityanath, the chief minister
of the largest state in India, Uttar Pradesh has been insisting on. What is then the difference between
the liberal secular Congress Hindu and the conservative BJP Hindu when it comes to the deeply
flawed policy on cows—none. More on this below, but first a discussion on how we got to this sorry
state.

The Constitution on cow slaughter ban

Surprisingly, it starts with the Constitution, where, in the Directive Principles, it is stated that,
The State shall endeavour to organise agriculture and animal husbandry on modern and
scientific lines and shall, in particular, take steps for preserving and improving the breeds,
and prohibiting the slaughter (emphasis mine)of cows and calves and other milch and
draught cattle.
The Directive Principles are not law, but are suggestions for ‘law’. The Constitution writers perhaps
recognised they were treading on thin ‘constitutional’ ice when they suggested a ban on cows and
buffaloes, but not goats and pigs or any other animal. Goats give milk, so why are we allowed to eat
mutton? A major drawback of the Indian constitution is that it wants India to be a Hindu state, but
shies away from saying so. But our Supreme Court judges are not shy from interpreting the
Constitution in a ‘Hinduised’ manner.
It became obvious that the cow is more than equal when in a 1959 judgement, the Supreme Court
allowed the killing of cows only after they reached the age of sixteen. The reasoning behind the
judgement being that cows of that particular age could be killed to make way for younger cows. A
little more than four decades later, a 2005 Supreme Court judgement (rarest of rare cases, a seven-
member bench) reversed the 1959 judgement and said that cows and bulls could not be slaughtered,
period, and gave the following grounds for doing so:
It cannot be accepted that bulls and bullocks become useless after the age of 16. This is
because till the end of their lives they yield excreta in the form of urine and dung which are
both extremely useful for production of biogas and manure. An old bullock gives 5 tonnes of
dung and 343 pounds of urine in a year which can help in the manufacture of 20 cartloads
of composed manure (emphasis mine).
This judgement must also rank amongst the rarest of rare—condemning millions of poor farmers
because of the benefit of twenty cartloads of manure per cow! Note how the court invokes only
economics, and bad economics at that, rather than attribute any religious beliefs to its survival or
otherwise.
Cow Slaughter Ban
The rigorous implementation of cow slaughter ban has its origins in a 2016 demand by the RSS
sarsanghchalak, Mohan Bhagwat (at an event to mark the birth anniversary of Lord Mahavir in New
Delhi on 9 April 2017). This was less than a month after the UP assembly election, which the BJP had
won with an overwhelming majority. In his speech, the sarsanghchalak unleashed what sounded
potentially damaging for the Modi-led BJP at the Centre. Even as I shall leave it to historians to
determine its true import, the fact of the matter was that Bhagwat had come out openly for a complete
ban on cow slaughter. ‘We have been striving to protect the cow. ‘Sampoorna Bharatvarsh mein
gauvansh hatya band ho, ye hamari ichcha hai’(We wish the killing of cows and its progeny be
banned in the entire country). He recommended a change in state laws (the cow may have been of
great concern to the Supreme Court, but it is a state subject). ‘Wherever in the state governments we
have dedicated swayamsevaks, those state governments have brought in laws. But our desire is that
the law should be for the entire country.’
The implementation of the hard-line cow policy, as advocated by the RSS chief, was accelerated
with the BJP appointing Yogi Adityanath as chief minister of Uttar Pradesh. While there was no intent,
deliberate or otherwise, the consequence of this policy witnessed an upsurge in communal tensions,
and innocent Muslims were targeted by extreme Hindu elements acting in the name of the cow. What
is apparently not realised by our cow-saving politicians and social engineers is that the cow slaughter
ban itself increases poverty and social tensions and therefore encourages illegality relating to cows
i.e., both the Hindu seller and the Muslim buyer are desperate to get on with their lives—the violence
is most likely due to the economic strain imposed on the poor via the necessity for feeding an
unproductive cow.
Cow slaughter is a major source of income for both the Hindu farmers and Muslim butchers. For
the poor, the killing of the cow is a necessity, not a spiritual urgency. Holy as it might be, a cow needs
food to survive. When she is producing milk, she is a major asset to everybody; when she stops, the
cost of feeding a cow is prohibitive (about 40-60 rupees a day). Even at the lower end, it comes to
around 15,000 rupees a year. That is close to the poverty line for humans. How are the poor then
supposed to finance the upkeep of her holiness? Cow economics 101 says that there are about fifteen
million buffaloes and cows beyond the stage of milk production. That adds up to a national
expenditure of 22,000 crores a year, co-incidentally the exact same amount as the government has
promised in the 2019–2020 Budget as basic income support for poor farmers.
The much talked about farmer distress is both due to over-production, and the misguided cow-
slaughter policy, and one endorsed and sanctioned by such respected institutions like the Supreme
Court, the RSS, BJP, and the Congress. The farmer is unable to feed the cow; the cow eats plastic,
and dies. Or enters schools and other places for shelter, refuge, and food. Or, the Hindu farmer
illegally sells the cow to a Muslim butcher, as he used to do before. In the recent past, this has taken
on the proportion of a political tragedy, when a Muslim butcher gets lynched by a Hindu mob. The
telling example of how the poor Hindu goes scot free, and not the poor Muslim, is proven by the
following incident.
In April 2017, six innocent individuals were transporting cows and calves bought at a cattle fair
in Alwar, Rajasthan; one of them was a Hindu named Arjun. The group had papers documenting that
they had legally bought the cattle. The Hindu was let go by the vigilantes acting in the name of the
cow; the other five, all Muslim, were severely beaten, and one of them subsequently died of the
lynching by the Hindu mob.
The ill-thought out cow policy therefore has had grave consequences. Muslims have been
lynched, and the violent gau-rakshaks (cow vigilantes) have gone unpunished. This has
understandably led to a fear psychosis among both the Muslims, other minorities and Hindus. Not a
good advertisement for any government, or any political party, let alone one accused by the media of
having a pronounced anti-Muslim bias; or one facing a tough re-election battle. And additionally, one
made tougher by the fact that the ten per cent Muslims who had voted for the BJP in 2014 will be
reduced to very low figures; in addition, genuine Hindu liberals and non-cow Hindu fundamentalists
may have been put-off by this cow irrationality.
RSS & Modi—Not on the Same Cow Page?
The RSS is a firm advocate of banning cow slaughter; that we know for certain. At a minimum, we
know that Mohan Bhagwat, the RSS supremo, is ardently for it. Although the RSS was not set up to be
a political organisation, it transformed itself into one and today, Bhagwat is also a politician, in
charge of a major socio-cultural-political organisation. Seldom are politicians not in competition, and
some of them often opposed to politicians within their own party.
The question regarding cow slaughter ban is the following—is Prime Minister Narendra Modi
on the same page as the leader of the RSS? Some divergence is noted in the statements of Mohan
Bhagwat and Modi, suggesting that they don’t see eye to eye on this issue. As discussed earlier, there
are large differences in the economic approaches of the RSS and BJP, as was made quite evident
during the late Atal Bihari Vajpayee’s tenure, and now also with Modi at the helm.
In mid-October 2015, Narendra Modi commented, some would say belatedly, on the lynching of
Mohammad Akhlaq who was beaten to death by a mob which suspected him of having stored and
consumed beef. Several members of the BJP were among the suspects in the case. Modi stated:
I have said it earlier too. Hindus should decide whether to fight Muslims or poverty. Muslims
have to decide whether to fight Hindus or poverty. Both need to fight poverty together. The
country has to stay united, only communal harmony and brotherhood will take the nation
forward. People should ignore controversial statements made by politicians, as they are
doing so for political gains.
At a meeting on 6 August 2016, Prime Minister Modi went several steps further—four Dalits had
been flogged by cow vigilantes in Una, Gujarat. Modi called them,
anti-social elements hiding behind the mask of gau-rakshaks… I get so angry at those who
are into the gau-rakshakbusiness. A gau-bhakt (cow devotee) is different, gau seva(cow
protection) is different. I have seen that some people are into crimes all night and wear the
garb of gau-rakshaksin the day. 70-80% will be those who indulge in anti-social activities
and try to hide their sins by pretending to be gau-rakshaks. If they are true protectors, they
should realise that most cows die because of plastic, not slaughter.
Less than two months later, Mohan Bhagwat pointedly disagreed with Modi. He described gau-
rakshaks as good people:
Countless good people are working for cow protection….These good people are working
within the ambit of law and Constitution….The administration should see to it that those
creating trouble should not be compared with cow protectors. There should be a distinction
between the two.
What makes the RSS’ stand somewhat untenable, and inconsistent, is because it is at great variance
with the views of two major scholars who are till date revered for their erudition and spiritual
worldview by the cadre—Swami Vivekananda and Vinayak Damodar Sarvarkar. In addition, there
isn’t much support for cow worship in ancient Hindu texts (the Ramayana and Mahabharata included).
During a speech in Madurai, Tamil Nadu, at the World Congress on Religions, dated 11-27
September 1893, Swami Vivekananda said:
 There are two sorts of truth we find in our Shastras (contained in the Vedas and Smritis). The
Smritis speak generally of local circumstances, of duties arising from different environments,
and they change in the course of time….(emphasis mine) There was a time in this very India
when, without eating beef, no Brahmin could remain a Brahmin; you read in the Vedas how,
when a Sannyasin, a king, or a great man came into a house, the best bullock was killed; how
in time it was found that as we were an agricultural race, killing the best bulls meant
annihilation of the race. Therefore the practice was stopped, and a voice was raised against
the killing of cows.
He then added:
As time rolls on, more and more of the Smritis will go, sages will come, and they will change
and direct society into better channels, into duties and into paths which accord with the
necessity of the age, and without which it is impossible that society can live. Thus we have to
guide our course...; and I hope that every one of us here will have breadth enough, and at the
same time faith enough, to understand what that means, which I suppose is the inclusion of
everything, and not the exclusion (emphasis mine).
As mentioned earlier, even Sarvarkar was dismissive of cow worship and said:
A substance is edible to the extent that it is beneficial to man. Attributing religious qualities
to it gives it a godly status. Such a superstitious mindset destroys the nation’s intellect….Let
the movement for cow protection be based and popularized on clear-cut and experimental
economic and scientific principles. Then alone shall we achieve genuine cow protection
like the Americans (emphasis mine).
Note the specific reference to a people known for their beef-eating.
Sarvarkar then proceeded to add:
The cow should not be the emblem of the Hindu nation. The cow is but a milch symbol of the
Hindu nation. By no means should it be considered its emblem. The symbol of Hindutva is
not the cow but the man-lion (reference to Narsimha, considered the fourth incarnation of
Lord Vishnu….; emphasis and clarification mine). Whilst considering the cow to be divine
and worshipping her, the entire Hindu nation became docile like the cow. It started eating
grass. If we are to now found our nation on the basis of an animal, let that animal be the
lion….We need to worship such a Narsimha. That and not the cow’s hooves, is the mark of
Hindutva(emphasis mine).
The terrible murder of Akhlaq happened before the Bihar election in 2015 which the BJP lost; the
Alwar murder happened post the UP election in 2017, a state where the BJP scored a spectacular
victory. This statistic is telling and the only inference one can draw in the context of electoral politics
is that much like his spectacular victory in the 2014 general elections, Narendra Modi was yet again
acknowledged as the undisputed leader of the BJP post the UP assembly elections. There is little
rational argument as to how he benefits from any of this abominable Muslim hunting or the cow
slaughter ban? He doesn’t—but he can be hurt by all the ill-will Muslim hunting generates among the
large middle class that will elect the next leader of India.
Mohan Bhagwat’s demand that there be a national law against cow slaughter is suggestive of
two interpretations—first, that it is against the best interests of India, the BJP, and Prime Minister
Modi, and not necessarily in that order. Second, that the RSS seems to be out of synch with the
modern Indian reality. There is however no gainsaying the fact that the RSS and its supremo has the
firm support of several BJP members, as also some from the Congress party in his mission to enforce
a ban on cow slaughter.
However, there is hope for those caught in the cow war—several Indians, Hindus, Muslims,
poor farmers and also the elite, are opposed to both the conceptualisation of the cow policy and its
implementation. Note that until the 2005 Supreme Court decision, the law was not so extreme and the
citizens had learnt to look the other way when faced with any transgressions of a loose law. The
court’s decision defied rationality and common sense, as was obvious in the decision itself—
recourse to cow dung and cow urine as justification for why banning cow slaughter was profitable.
How odd is the Hindu obsession with the cow? Not very, if you consider what happens in the
most advanced economy of the world—the United States of America. It is useful to compare the
‘rationality’ of a holy cow with another constitutionally mandated ‘rationality’—the right to bear
arms.
The second amendment to the US constitution states: ‘A well regulated militia, being necessary
to the security of a free state, the right of the people to keep and bear arms, shall not be infringed.’
The US constitution was written at a time when the founding fathers were fighting a war against the
colonisers. A legitimate question asked by many (including various US Presidents) is whether the
circumstances, and the country, and the world has changed in the last 250 years. Do the Americans
really need guns to protect themselves today against an invading army? Isn’t a nuclear bomb enough
of a weapon? Does the gun lobby realise that more Americans have been killed by guns in the US
since 1968 than in all wars that the country fought?
Therefore, in a way, guns are America’s sacred cows and it appears that they are not dealing
with the problem any better than we are. Both countries victimise innocent citizens in the name of an
abstract indefensible belief. Hopefully, the politics of tomorrow will be different.

1 See Sanjay Kumar, ‘Who did India’s Muslims vote for in general election’, BBC News, 30 May
2014.
Sab kuch seekha hamne, na seekhi hoshiyari
Sach hai duniya walon, ke hum hain anari…
(I learnt everything, except to be clever
It is true that I am naïve…)
– Anari, 1959
Battle of the Elites

A very important structural change is presently happening in India—what can easily be termed as a
historic battle. The changing of the elite, a rare occurrence in any society, even as the old elite, quite
naturally, is loath to give up on its privileges.
This changing of the guard is relevant because it has been brought about by the candidacy of
Narendra Modi in 2014, and his incumbency in 2019. No matter how you define the elite, the
conclusion remains that Modi is the new elite. He belongs to the exclusive club because he is the
prime minister, and with reasonably impressive popularity ratings. The other and more significant
factor is that he is radically different from all his predecessors, even from within the BJP, and
obviously from the Congress. And this is the last but not the least ingredient for his stupendous
success—Narendra Modi represents the beliefs of the single largest component of the new elite—the
middle class.
As Election 2019 approaches, be aware that beneath all the analyses and opinions, this
momentous battle of the elites is behind every utterance, every forecast. This may sound rather
simplistic, but then one also needs to recognise that regardless of political affiliation, the old elite has
broadly the same political and economic philosophy, characterised by Western-style social liberalism
and Fabian economic socialism. However, over time there have been changes in all viewpoints—for
instance, people are not as openly feudal as they were before. Nevertheless, it is critically important
for analysts and politicians to appreciate the tension between the old and the new when they read
about political developments in English-language reports: it is my firm opinion that the liberal press
shares the cosmology of the old elite and is consequently structurally biased against Narendra Modi
and the new order that he represents. The tension runs very deep, and very likely will affect the 2019
election, as it did in 2014.
What are the principal causes of this change in the structure of the elite? Actually, in my view,
there is only one cause—the rise of the middle class. As pointed out by me in two books (published
in 2007 and 2017 respectively), throughout world history of the last 400 years, reforms (read
changes) have been inspired and championed by the middle class.
The New Elite—the Middle Class
Who is the middle class? Definitions differ, but the poverty line in advanced economies maybe the
best starting point.1 Conventionally, we think of the world as having three classes—the poor, the
middle class and the rich. By definition, the line defining the poor is also the line defining the
beginning of the middle class. The line defining the rich is the top of the middle class and the
beginning of the rich. In present prices, the middle class begins at approximately one lakh rupees per
person per year.
In 1950, the middle class was a minuscule one per cent of the Indian population, but today it has
grown to be close to fifty per cent. There is little doubt that this class is currently the quintessential
elite—it is new, it is educated, and it has a modern outlook. And yes, unlike the old, it is heavily
disposed towards merit.
The interests of the middle class are different than all other classes, above and below. Those
below are concerned the most with making ends meet; they are poor, or near (absolute) poor. The rich
are most concerned with preserving their relative status, whether it be feudal, or monopoly
capitalistic. Thus, it is dominantly the middle class that is interested in economic, political and social
change. As we well know, reform enhances growth, which accelerates both the absolute and relative
value of the middle class.
Therefore, expansion of the middle class is crucial to the overall development of a nation, and
particularly the political process. When this change occurs, the established old elite has to make a
choice—either be co-opted and allow itself to be ‘ruled’ by the newly educated, technocratic, and
empowered by global values middle class, or allow itself to be overrun by history.
Electorally speaking, the old elite in India chose to fight the middle class. It accidentally won in
2004 (it may be recalled that the old elite, the Sonia Gandhi-led UPA, obtained the same 140 odd
seats as in 1996 when P.V. Narasimha Rao lost the general election) and continued with its noblesse
oblige policies which was fifty years old.
Old and New Elite—Identifiers
The first pointed old and new elite battle took place within the UPA when Dr Manmohan Singh was
prime minister, and his recommendations for policy and personnel got over-ruled by the
representative of the old elite, Sonia Gandhi. Indeed, as Manmohan Singh’s then media adviser,
Sanjaya Baru controversially asserts in The Accidental Prime Minister: The Making and Unmaking
of Manmohan Singh, that Sonia Gandhi was actively involved in the decisions made by the prime
minister.
But Dr Singh did win one important battle for the new elite—in 2005, he championed the
nuclear deal with the US. It may be recalled that it was heavily opposed by most of Congress’
coalition partners—especially the Communists—but to his credit, Manmohan Singh was able to
achieve the near impossible. Being anti-American was a sine qua non of the old elite, while the
signing of the nuclear deal with the US (and being anti anti-American) was the biggest sign of a
different Indian order, and one increasingly inhabited by the new elite.
A good heuristic model, or benchmark, or representation of the old and new elite is provided by
two of India’s most distinguished economists—Nobel prize winner Prof Amartya Sen personifies the
old elite with his recommendations against the market and belief in the enlightened efficacy of the
bureaucrat, and the system; whereas, Jagdish Bhagwati represents the upstart believer in individual
and market freedom. I had earlier offered a simple yardstick of identifying people opposed to
demonetisation—if you paid a larger fraction of your income today than before (and larger not
because of tax rates but because of greater compliance), you would be against demonetisation.
Analogously, one identifier of the old and new elite is whether you think Sen or Bhagwati is a better
economist!
It is not a co-incidence that one of the principal advisers to the Congress party, and in particular
to Sonia Gandhi’s alternate power and policy centre, the National Advisory Commission or NAC,
was Amartya Sen. And a prominent economic adviser to Narendra Modi was Bhagwati. It may also
not be a co-incidence that Sen is a Bengali and Bhagwati a Gujarati!
Modi—The Challenger
The dominance of the old elite could only be ended with the arrival of a card-carrying member of the
new elite. Narendra Modi, the son of a roadside tea seller, could be many more things as his
detractors claim, but he is the polar opposite of a Fabian socialist. His aspirations were (are)
quintessentially middle class, and it is no co-incidence that policy documents during the last five
years have constantly mentioned the aspirations and demands of the new educated middle class. He
came to power on the back of a successful transformation in the state he had ruled for a decade. In
Gujarat, GDP growth accelerated from 5.4 per cent per annum in the decade prior to 2002, to 9.7 per
cent per annum in the decade post 2002 (the comparable numbers for all-India growth rates were 5.6
and 7.6 per cent respectively).
Narendra Modi instinctively challenges the old way of doing things—he gnaws at the comfort
zone of the old elite at every given opportunity and that is also a reason for the staunch opposition to
his re-election. The tussle is between the old, established way of doing business, and the new
approach, whereby family origins and influence carry considerably less weight than ever before.
It is prudent to acknowledge that technology and the middle class are taking over most of the
influential policy roles of the old elite. For instance, anti-poor policy is best handled via direct
transfer of incomes; tax payments are enhanced with proper identification (Aadhaar, the twelve-digit
unique identity number) and the GST.
The prime minister has hit the old elite where it is most vulnerable—money and influence. He
engineered a crackdown on corruption; the discretionary powers of bureaucrats, which was taken as
an added perk to influence decisions were reduced; and post demonetisation, there was a distinct
increase in tax compliance. When it hurts the pocketbook, it is only natural to complain, which the old
elite has done relentlessly since 2014 by asking repeatedly: ‘What real change has Modi effected?’
They constantly remind themselves and their constituencies how demonetisation was not only
draconian, but a sheer waste of an effort; the NDA, nothing more than the UPA tethered to a cow, and
so on and so forth.
In addition to the economic policies, which are a recent addition, Narendra Modi is unpopular
with the influential old liberal elite for two main reasons. First, he is viewed as a champion of
Hindutva — i.e., that India is predominantly a Hindu nation, which directly contradicts with universal
secular liberalism ethos of the old elite. Note that throughout, Modi has constantly stated that he
believes in vikas for everybody; also, that he has never called India a Hindu nation. The old elite,
however, has never stressed on the nuances.
Second, and perhaps more importantly, the old elite hasn’t forgiven him for the role he allegedly
played (although he was acquitted by the courts) in the 2002 Gujarat riots, which proves, in their
eyes, Modi’s anti-Muslim bias. Let us examine the two objections below.
Hindutva
The major identifying characteristic of the new political order, and the new elite, is its adherence to
the philosophy of Hindutva. But what is Hindutva, if viewed from their prism? Simply, that India is
dominantly a Hindu nation and because of the simple statistic that eighty per cent of the population is
Hindu, down from a 100 per cent two thousand years ago.
The old elite are well within their rights to ask: first, the Indian Constitution does not say so in
so many words; second, even if this statistic were to be taken seriously, this certainly is not a
proportion to make any Hindu insecure. But the perceived reality amongst certain sections is that the
Hindu is insecure in her own land, and this is what the BJP has capitalised on for many years.
But why does the Hindu feel insecure? A consistent view of the old elite, and the Congress
leadership, has been that India needed to shed religion as part of modernisation, and that this was
necessary for the multi-ethnic nation to stay intact. Like their Western counterparts (an important
aspect of the old elite is its attainment of high levels of Western education), the emphasis was more
on Western customs and attitudes, and even to the extent of neglect and possibly contempt for the
‘Hindu’, or Indian, way.2
Additionally, an important objective of the Congress party was the achievement of a ‘secular’
State, in which there was a complete separation of church and State—according to most, if not all, a
laudable goal. Additionally, the rise of fundamentalist Islam over the last few decades has underlined
the appropriateness of the secular goal. And at the same time made increasingly unpopular Muslim
vote bank politics as practiced by the Congress.
To this day, the old elite has refused to accept Narendra Modi as an individual worthy of being
prime minister. The antagonism is at two levels—Modi was not educated in an English-medium
school (English being the definite lingua franca of the old Indian elite) and hence not ‘worthy’ of
aspiring for, let alone achieving, the high office of a prime minister. This is classic class bias and the
old elite is quite unsubtle about its contempt for Modi (the old elite continues to refer to him as
chaiwalla at various platforms).
Godhra: Gujarat 2002
On 27 February 2002, Gujarat witnessed its worse episode of Hindu-Muslim riots following the
burning of a stationary train compartment of Sabarmati Express in Godhra, in which fifty-eight Hindu
pilgrims were attacked and killed by a Muslim mob. The crime which took place in the wee hours of
the morning and over what was perpetrated to be a minor scuffle, soon snowballed into a virulent
communal conflagration resulting in the deaths of 790 Muslims and 254 Hindus. Narendra Modi was
at the time the chief minister of Gujarat.
Subsequently, several investigations were carried out and also a 541-page-long report by a
Supreme Court-appointed Special Investigation Team (SIT) in May 2012, which rejected any
culpability on the part of the sitting chief minister.
Let us be clear that even though Modi was cleared of any direct involvement in the 2002 riots by
Indian courts, the fact remains that he was the chief minister of Gujarat and hence was, by definition,
accountable for whatever happened under his watch.
Delhi: 1984
Now consider an equally (if not more) horrific tragedy some eighteen years earlier—the 1984
pogrom against the Sikhs, which was in retaliation against the assassination of the then Prime
Minister Indira Gandhi by her Sikh bodyguards. It was more in the nature of a pogrom in which Sikhs
were selectively targeted and killed. According to all available accounts, lumpens had a free hand to
wreak havoc and roamed around the streets of Delhi looting, raping and killing hapless Sikhs for
nearly three full days.
It was only as late as the third day of the carnage when a curfew was announced and the Army
was summoned to stop the carnage, which was one of the worst in the history of post-independent
India. I am aware that comparisons are odious, but even so, the curfew remained unenforced until
the fourth day. According to several reports, approximately 3,000 Sikhs were reportedly killed in
Delhi alone, with an all-India figure hovering around 8,000 deaths.
If Narendra Modi was responsible for Godhra because the riots happened under his watch,
objective logic should also conclude that Rajiv Gandhi was responsible for the pogrom against Sikhs
—it happened under his watch and he issued a statement seemingly condoning the killing of hundreds
of Sikhs: ‘When a big tree falls, the earth shakes.’
Although the last of the inquiry commissions on the 1984 killings cleared Prime Minister Rajiv
Gandhi of any wrongdoing in its February 2005 report, it pointedly held Congress leaders, Jagdish
Tytler and Sajjan Kumar culpable. What was most astonishing was that both were handed tickets by
the Congress party to contest the 2004 parliamentary elections with the former even serving in the
UPA-I Cabinet. He however remains the subject of a criminal inquiry by the Central Bureau of
Investigation (CBI) till today; while thirty-four years after the incident, Sajjan Kumar was convicted
for his role in the riots in December 2018.
While delivering the judgement in 2009, Additional Sessions Judge, Surinder S. Rath had said:
Though we boast of being the world’s largest democracy and Delhi being its national capital,
the sheer mention of the incidents of 1984 anti-Sikh riots in general and the role played by
Delhi Police and state machinery in particular makes our heads hang in shame in the eyes of
the world polity.
It is almost as if he was expressing the anguish of a nation that the old elite reiterates is both secular
and democratic. However, even as they have moved on from that terrible tragedy, their focus has been
unwavering on the tragedy of the Godhra riots. In the 2007 Gujarat state election, the Congress party
ran a campaign with a slogan equating Modi with a merchant of death (maut ka saudagar).However,
he not only won that year but yet again in 2012; but the old elite refused to acknowledge his win.
Later, he was barred from entry to Western nations, and as late as 2013, after being elected the
leader of the BJP for the impending 2014 election, Modi was ‘uninvited’ from a conference at the
University of Pennsylvania due to his alleged role in the 2002 Gujarat riots.
I have described very briefly facts pertaining to both tragedies and what clearly emerges is a
strong bias, and inconsistency, with which the old elite views the Sikh riots and the role of the
Congress party, and the alleged role of Narendra Modi and his party members in the 2002 riots.
Therefore, amongst large sections, the old liberal elite commands little respect, and little credibility.
There is no doubt that both 1984 and 2002 are major blots on India’s democratic record, and
need to be condemned in no uncertain terms by both the old and new elite. But that is obviously of
little consequence for the old elite who seem to evaluate Modi only for his alleged role in the
sectarian violence of 2002. But like I had quoted the opening lines from a popular Hindi film song in
the Introduction to this book, Yeh jo public hai, yeh sab janti hai…, this double standards of liberals,
including the press, did not go unnoticed in the 2014 campaign, and continues to play a role in the
unpopularity of the Congress party, and the declining popularity of the old liberal elite.
India rightly prides herself as a democracy and as a society where justice matters. Arguably, if
the judicial commissions for the 1984 pogrom had been seen as impartial, and punishment to its
perpetrators had been quickly meted out rather than some of the accused being rewarded with Cabinet
posts, the 1992–1993 Mumbai riots and Gujarat 2002 may have been avoided. The lesson from the
1984 riots for me is that you could not only literally get away with murders, but also be rewarded for
it. This has contributed to the creation of perverse incentives.
What impact is the battle of elites going to have on Election 2019? The old elite continues to
oppose Modi, and fancy the Congress. That is its right—but democracy and politics dictates that one
should have respect for the electorate and the changes in society. One of the main reasons that the old
elite representative, the Congress party, is in the doldrums is because it has failed to notice, and
respect, that India has changed from the illiterate and feudal order prevalent at the time the Nehru
dynasty assumed control in India. From a poor nation, India today is an aspiring middle class country.
Communication technology ensures that news travels, and travels instantly, and that old-style politics
does not work anymore. Inconsistencies (like on Godhra and Delhi) are noted by the public, as is fake
news and faulty analysis (see Chapter Eight, p. 111).
It is not only my firm belief, but of scores of others, that in contrast to other political leaders
before him, Narendra Modi represents the new aspirational elite, one that is anti-feudal, anti-political
dynasty, and more responsive to the needs and demands of a ‘new world’ order. And more in sync
with, and execution of, what both the old and new elite desire, inclusive growth—a subject
considered in the next chapter.

1 See The New Wealth of Nations, and Second Among Equals—The Middle Class Kingdoms of
India and China, for details on the definitions of the middle class.
2 See the discussion in Chapter Three (p. 34) about beliefs of Nehru, an old elite.
Kuch toh log kahenge, logon ka kaam hai kehna
(Some people will say what they have to; it’s their job to say what they will….)
– Amar Prem, 1972
Modi and Inclusive Growth

The 2014 election was about change in India—and obviously change for the betterment of all citizens.
Almost five years have passed, and in a few months, the country will post its verdict. The people’s
choice will most likely be determined by changes in the political, economic, and social environment.
This chapter discusses how all the different factors affecting the economy, and elections, have
unfolded so far.
Expectations and Performance
On 4 December 2018, the editorial in the prestigious London-based newspaper, Financial Times,
concluded, ‘Modinomics has yet to deliver for many in India.’ There are two major critiques of
Narendra Modi’s handling of the economy, and hence the delivery of achhe din(good days, a promise,
albeit unofficial and more like a slogan, made by the prime minister and his party in 2014). First, that
demonetisation not only had zero positive effects, but huge negative effects, especially for the poor
and the emerging middle class—broadly, Modi’s support base. We examine the veracity of this
opinion/conclusion below, and especially on whether economic growth under Modi has been pro-or
anti-poor.
The second assessment, not related to the first, is that Modi violated all expectations of
economic reform, and therefore failed in delivering achhe din, because he did not privatise the
economy. Let us not concentrate much on privatisation, because not only is it ‘not’ an economic
reform, it is irrelevant to any assessment of the Indian economy at this juncture—as it was in 2014;
and irrelevant for analysing voting behaviour.
Corruption
Two issues dominated the 2014 election—corruption and development. Somewhat surprisingly,
opposition to Modi has taken the form of not what actually happened—i.e., ‘Did you experience
better conditions than earlier?’ But more like a reminder, ‘But Modi promised!’ In other words, one
should vote Modi out because the incumbent has not delivered fully on all his promises. Note that this
is an unreasonable and impossible expectation, and can never be fully met.
No one expected corruption to be eliminated in five years; given that it has not been eliminated,
anywhere, for the last 5,000 years, its complete defeat surely is not what people expected. Yet you
find reasonable men saying that they are not voting for Modi because he promised an end to
corruption: ‘Look, I just paid a bribe; I know corruption is still there, and this is not as it should be.’
Let us accept that there is corruption, but the moot point is that the people are most likely to
judge Narendra Modi on whether corruption has significantly reduced. Limited evidence in the form
of surveys do suggest that corruption has reduced; while anecdotal evidence supports the premise.
Inclusive Growth
Every government, at least since the reforms initiated in 1991, has spoken about ‘inclusive growth’. It
wasn’t called the same back then—perhaps ‘growth with a human face’, or ‘pro-poor growth’, or
some variant thereof. The origin of the phrase ‘inclusive growth’ can definitely be attributed to the
UPA-I period, when, post the continuing market-oriented reforms, successive governments were
criticised for not doing enough for the poor, or indeed making their condition worse than ever before.
How often have you heard: the rich are getting richer and the poor are getting poorer? Very often, and
you still hear it today, even from respected academics. To put it bluntly and succinctly, there is no
bigger fake news than the assertion that economic growth, at least in India, has made the poor poorer.
We await the results of the 2017–2018 NSSO consumer expenditure survey—the source for
determining the level of poverty in India. Unfortunately, the last such survey was undertaken way back
in 2011–2012; analysts like myself and Homi Kharas of the Brookings Institution have concluded that,
when released, the NSSO 2017–2018 data is very likely to show that absolute poverty in India will
be considerably reduced from the fourteen per cent level of 2011–2012; and that the new government
should (will) raise the poverty line to a level encompassing at least twenty-five per cent of the
population.1
Economic Performance: 2014–2018
For some unspecified reason, the weather gods have not been kind to NDA governments. As noted in
Chapter Five (see p. 64), in the five years 1999–2003 (NDA-I), average rainfall was more than ten
per cent below normal—the worst rainfall in any consecutive five-year period. No sooner had NDA-
II entered office with a promise of doubling agricultural income over seven years, it was greeted with
two consecutive drought years, 2014–2015 and 2015–2016.
Undaunted, the Modi government persisted with economic reforms, including an attempt to arrest
the spread of black money i.e., demonetisation. That happened in the first good agricultural year,
2016–2017. In the fourth year, 2017–2018, the government introduced the mother of all tax reforms,
the Goods and Services Tax or GST. In May 2016, a few months before demonetisation, the
government had introduced yet another major economic reform—Insolvency and Bankruptcy Code
(IBC).
These are undoubtedly the three big reforms and easily match the 1991 reforms in both scope
and effect. In addition, very large-scale social sector reforms were undertaken by the Modi
government. Initial attempts at reforming the rather tough agricultural sector were also made. Adding
all of this up, it is a fair proposition to state that more economic (and social sector) reforms have
been introduced in these five years than in the previous twenty-five years (1991–2014).
However, some economic reforms are disruptive in the short run e.g., demonetisation and GST.
They increase uncertainty and dampen economic activity. It is an unfortunate reality that the RBI/MPC
was not helpful in counter-acting the negative short-term consequences of economic reforms.2 When
the Monetary Policy Committee or MPC tightened in the face of huge disinflationary effects of
reforms, it went against the macro-economic hypotheses about policy and growth. Normally,
loosening follows expectations of below normal GDP growth; in India, all the reform policies
pointed to higher uncertainty, lower inflation, and lower GDP growth. Yet, the MPC acted perversely
by raising policy rates to Indian, and world record levels. This has definitely slowed the pace of
arrival of achhe din. How much this will affect the vote count, we will only know in May 2019 (when
results become known). It may be recalled that several policies to hasten the pace of achhe din for the
bottom seventy per cent of the population were undertaken in the last few years, and are documented
elsewhere in the book.
As chronicled in Chapter Five (see p. 64), two major economic reformers, P.V. Narasimha Rao
and Atal Bihari Vajpayee lost their re-election bids in 1996 and 2004, respectively. They were the
original reform-oriented prime ministers. The question remains whether Narendra Modi will escape
the ‘reform jinx’ in 2019 or not.
What Did the Economic Reforms Deliver?
Growth has been lower than expected, but there have been several positives. For example, direct tax
collections have shown the highest buoyancy i.e., the biggest (tax revenue) bang for each buck (unit of
GDP growth). This is because tax compliance has considerably increased.
Aided by tax buoyancy, and spending restraint, the fiscal deficit is more in control than ever
before; its best days lie ahead. In addition, inflation is in control with the five-year CPI average close
to four per cent (if not below).This has allowed the macro growth performance indicator (GDP
growth/CPI inflation) to record its best performance in the last fifty years.
Counting some of the major macro-economic achievements: GDP growth, second best at 7.1 per
cent, with 2004–2008 at 8.1 per cent (old GDP series) the best ever. Inflation, second best ever, with
NDA-I marginally lower and the best. Tax to GDP ratio, best ever. Direct tax buoyancy, ditto. Tax
compliance, likewise. Fiscal deficit control—the best ever.
Jobs
But what about jobs? According to leaked reports from the national employment survey conducted by
the NSSO in 2017–2018, the unemployment rate topped six per cent—the highest rate in forty-five
years. (In 2011–2012, the unemployment rate was two per cent.) Such a large increase in the
unemployment rate is indicative of a very low growth rate in job creation, very likely even negative.
In 2013–14 (the year preceding the Modi election and therefore the inherited economy), a job survey
conducted by the Labour Bureau, Ministry of Labour and Employment, showed that the unemployment
rate was 3.4 per cent. Still, the rate jump from 3.4 to six per cent is large.
The Modi government, owing to a knee-jerk reaction, has till date not released the report or the
underlying raw data. This is a huge mistake and it will be desirable if the government cuts its losses
and releases the data and the report.
One peculiar feature of the leaked NSSO report, and a feature which hasn’t received much
attention, is the wage growth for employees in the formal sector (the top thirty per cent), and the wage
growth for employees in the daily wages sector (approximately the bottom seventy per cent of the
work force). We are reporting these data without malice or prejudice; indeed, we think the reported
wage growth data to be the end result of not fake news, but very faulty survey and/or faulty analysis.
NSSO data for 2017–2018 (one with unemployment rate at a forty-five-year high) shows that
real wage growth for the bottom seventy per cent of casual workers rose at an unprecedented Indian
(and possibly worldwide) rate of 7.6 per cent per annum for six years (between 2011–2012 and
2016–2017, nominal wage growth at 12.7 per cent; inflation at 5.1 per cent per annum). The same
survey shows that nominal casual worker wage growth for rural females was at 17.6 per cent per
annum compared to only 12.9 per cent for rural males.
In contrast, for salaried workers (the top thirty per cent), real wage growth was not only
considerably lower, but a negative 0.8 per cent per annum. Never has this economic paradox been
observed for such a long period of time. For 120 million upper class workers, a negative wage
growth, and for 230 million workers, a positive and large wage growth—can this be correct?
The big conclusion from the NSSO 2017–2018 ‘leaked’ report—the highest unemployment rate,
in addition to the most inclusive wage growth ever documented, and the most gender-correct wage
growth ever observed. Neither the ‘good news wage’ nor the ‘bad news unemployment data’ is likely
to be correct. Some believe these data represent a ‘true’ picture of the Indian economy, and that the
data will have a bearing on the May election. But that can only be ascertained once the data are
finally released.
Revolution of Rising Expectations
A consistent policy adopted by the Modi government has been to provide meaningful assets for the
poor—assets that the upper class (the top thirty per cent) can take for granted, but are life-changers
for the bottom seventy per cent. Modi government’s major schemes have also had women as the prime
beneficiaries. The list—liquefied petroleum gas cylinders so the women don’t cough, or die, from
smoke inhalation; Beti Bachao, Beti Padhao (Save the daughter, educate the daughter); financial
inclusion in the form of bank accounts in the recipient’s (primarily a woman) name for welfare
transfers from the centre; toilets for the poor (read women); and housing for the poor.
That the government was particularly focusing on the women for its benefit programmes was
visually brought home to me when I watched a public relations video that the government had
prepared to showcase its achievements in end May 2018. A very large part of the visuals only had
women—a girl child, a young girl, young women, and also an old woman—as central characters.
Table 11.1 presents a simple tabulation on how inclusive growth, and women-intensive, the last
five years have been:

Table 11.1 Inclusive growth 2014–2018

Fiscal Year
Inclusive Growth Indicator 2013–14 2017–18
Ease of Doing Business (World Bank rank) 134 77
Tax Collection (lakh crore) 6.4 10
Personal Income tax (filers, in crores) 3.8 6.4
Bank Accounts (% of population) 53 80
Bank Accounts (Gender gap; %) 20 4
Jan Dhan Accounts opened* (in crore) 12.5 33.7

Villages Electrified* (per cent) 96.7 100

Solar Power (in Giga Watts) 2.6 12.3

LPG Coverage (cooking gas cylinders); % of households 30 89.5
Rural Road Construction (km/day) 69 134
Rural Toilets Built under SBS (in mil) 4.8 21.3
DBT—Direct Benefit Transfer (in Rs. 000 crore) 7.4 102.1
DBT Beneficiaries (in mil) 10.8 63.3
Sources: Ministry of Electronics and Information Technology; pmjdy.gov.in, uidai.gov.in; Ministry of Drinking Water and
Sanitation. Syed Zafar Islam, Indian Economy, 2019.
Note: SBS refers to Swachh Bharat Scheme. * The first year refers to 2015–16.

The people behind these abstract growth and performance statistics (incidentally, all show
exceptional growth rates) came alive during my visit to Guna, Madhya Pradesh in the fourth week of
November 2018. I was part of a fifteen-member Limousine Liberal ‘team’ which has over the last
twenty years, religiously been visiting states going to elections. As is our custom, and our wont, we
travelled to villages to obtain a sense of things on the ground (the hawa).3 I must concede that this
visit was my most heart-warming experience as a (development) economist.
In a dominantly Schedule Caste village near Guna, I was confronted by six angry young women,
who were not only vocal, but extremely self-assertive (in a very positive way). These women were
demanding their rights.
As documented in Table 11.1, the BJP government has an extensive toilet building, and a house
building programme around the country. India has always had subsidies in the name of the poor, but
for the first time, I was able to witness these subsidies actually reaching the poor. As it turned out,
some of the women had received their toilet building grants; some had not. Those who hadn’t
received their grants for the free toilets, and/or grants for free pucca housing, complained loudly.
These complaints echoed the discussion I recently had with an Uber driver (Jitan) in New Delhi.
He had told me that one of the biggest changes that the Modi government had brought about was that
people, especially poor people, were now aware what rights they had i.e., what government benefits
they were eligible for. This knowledge was creating expectations, and unfulfilled demands were
responsible for a revolution of rising expectations.
It is hard to tell which way this revolution will go in terms of voting in Election 2019. Will the
poor and emerging middle class be grateful that the government is now providing tangible, life-
improving, benefits? Or will they feel upset that they have to wait their turn? For that we would have
to wait until May 2019.
Modi’s Mistakes & the Fear of Criticism
It is normal in a democracy that the government be criticised for making a mistake. The Modi
government has received more than its fair share of criticism. That is as it should be. It also is the
case that some believe that there isn’t enough criticism—and the lack of criticism is a sign of fear on
the part of the electorate, perhaps of retribution. How valid is this fear? And how logical the belief
that this is structurally different than the ‘fear’ felt under previous governments? We examine the
veracity of these fears below.
I know recounting personal experience is not proof, but doing so might help shed light on this
important belief, or concern, or fear. I have documented how the economic performance has been
considerably better under Modi than what it was earlier. But the performance has not been devoid of
major mistakes.
Some of these economic (and non-economic) mistakes were made while I was a part-time
member of the Prime Minister’s Economic Advisory Council (PMEAC), from September 2017 to
November 2018. All through my stint, my public and private evaluation of the government was
unaffected by my advisory position. I can categorically, and emphatically, state that the government
never attempted to stymie my criticisms in any manner. If an adviser to the government was free to
criticise publicly, why on earth would the government bother to restrain others?
Policy Mistakes
The 2014 Budget was so bad that it could have as well been produced by the UPA-II government. For
example, apart from not initiating any reforms, the Budget kept intact the retrospective tax legislation,
the one policy candidate Modi was very critical of.
Another mistake, mentioned earlier (and mistake in my eyes) was the creation of the MPC as the
sole arbiter of monetary policy. Full disclosure: many consider MPC to be a significant economic
reform. However, its actions have been the most inexplicable. Highest real rates in the world (latest
estimate is around four per cent) and more than three times the rate advocated by the ex-Governor of
RBI, Urjit Patel-led MPC during its first meeting in October 2016. On that occasion, the MPC had
emphatically said that it was targeting a real rate of 1.25 per cent.
Interestingly, the most ardent supporters of MPC’s decisions are also those who are openly
against the Modi administration. I mentioned earlier on in this chapter that there is a belief, and fear,
about disagreeing with the Modi dispensation. Curiously, very little objection was raised to the Urjit
Patel imposition of a ban on MPC members talking to the public either before, or after, the decisions
were made. No central bank, no MPC, has this restriction; they all have a window when the members
can freely discuss, and express their opinions. (If exceptions exist, they are likely to be very few. Let
us put it this way, important economies do not have this limitation.)
In the 2018 Budget, the Finance ministry re-introduced the long-term capital gains tax. This
decision of reversing a fourteen-year-old policy was not based on any economic logic. False morality
perhaps, but not good economics.
And then came yet another ill-advised economic decision in Budget 2018—the raising of the
Minimum Support Price for farmers to fifty per cent above cost. This was seen as helping farmer’s
distress. As @twitter would say, LOL—Laugh Out Loud.
Another mistake—the targeting of students who disagree with the government’s political and/or
economic viewpoint. Smart governance generally means you don’t target students (or the media). It
always backfires, and is hard to defend. The Modi government is guilty of having violated the ‘cease-
fire agreement’ with students on several occasions. The entire attention on student protesters at the
Jawaharlal Nehru University or JNU was not only uncalled for, it was politically naive. It is hoped
that the new administration will recognise and respect this simple home truth.
Riots and Communal Tensions
As discussed in Chapter Nine (see p. 131), there is increased tension between the two largest
communities—the Hindus and Muslims, and many claim that relations between them have never been
worse. This is a subjective view, and it behooves all of us to incorporate this view, and absorb it
critically. One can conclude based on the simple maxim—where there is smoke, there is fire. At the
same time, it is the case that around the world, tensions have heightened between communities, within
societies. The role of instant communication and opinions as exemplified on WhatsApp and Twitter
should not be brushed aside. Both discourse and opinions should be based on concrete aggregate
evidence—and on the pace of change, which I think is asking for a bit much in the times that we live
in.
Let us look at whatever data are available, across time and countries, on societal tensions. The
battle of elites is also a reality, and a constant campaign run by the old elites is that the situation has
never been so bad, and every riot, every lynching, every murder, every crime against women is a first.
I exaggerate, but I hope you will agree that a detached reading of the data on the sensitive topic of
communal tensions will be helpful.
In so far as the cow slaughter issue is concerned, widely believed to be the most important
reason for the increased tension between the two communities, there is so much anxiety that leading
intellectuals claim that they are living in an unprecedented fear economy, and especially one for the
Muslims. One issue of Seminar, a thoughtful journal on contemporary political and social themes,
was titled Cow, and said amongst several other things:
UNEXPECTED resources of energy, organisation and dedication have revealed themselves
in the fury over the cow agitation during the last year. This has its positive aspects: if such
fervour could be let loose all over the country in lighting drought, illiteracy, backwardness,
India could assume a different look. The fact that the cow has acquired a position of such
symbolic consequence, denotes the seriousness of the problem. No political party, no
candidate seeking votes, dared open his mouth against the agitation whatever his personal
persuasion. This only proves that the problem needs to be looked at soberly so that national
energy should not be wasted.
It all began with Article 48 of our Constitution in the chapter on the Directive Principles of
State policy. ‘Organisation of agriculture and animal husbandry.’Since then, a rash of
legislation appeared in the different States. (There a re-instances of peasants, breaking the
arms and legs of useless cattle which they cannot support, in order to sell them for slaughter.)
There is now a total ban on killing cows in Gujarat, Madhya Pradesh, Uttar Pradesh and
Orissa. Here, cattle other than the cow can be slaughtered if they have passed the milch stage,
or when they are of no use for breeding, or work, or if they are incapacitated.
This article resonates with what has happened over the recent cow slaughter Muslim bashing years. It
is informative that the article appeared in the first issue of the journal in May 1967; the founder-editor
of the journal, Romesh Thapar, was a well-known Leftist journalist and a card carrying member of the
CPI (M). As also pointed out in Chapter Nine (p. 131), the cow is neither a BJP nor a Modi problem
—it is much larger and requires the intervention of Hindu society.
We shall now present all the available data to test the proposition that communal incidents
and/or deaths have increased over the last five years. The evidence is presented so that the reader can
make up her mind regarding the increase (or decrease) in civilian and/or communal violence in India.
Communal Riots—Such a Long & Sad History
Independent India has seen frequent occurrences of communal riots. One of the first major communal
riots was the infamous, ‘Bombay riots of August 1893’ in which almost a hundred people were killed
and 800 injured.
The 1893–1945 period was a particularly violent period for the erstwhile Bombay Presidency,
the region broadly covering today’s Maharashtra and Gujarat4. Some of the reasons which sparked the
killings between the two communities are worth noting—discord over cow slaughter; disputes over
the playing of music near religious places; the building of prayer halls, mosques, and temples—not
very different from the debates we are witnessing even today.
As the official statistics on historical records of communal riots are sparse, we have to perforce
rely on newspaper reports and piecemeal evidence combining various sources. Ashutosh Varshney
and Steven Wilkinson provide the most comprehensive, ‘Dataset on Hindu-Muslim Violence in India,
1980–1995’, as reported in The Times of India, Mumbai edition. (This dataset has been extended by
other researchers and now covers the period 1950–2010.) The key finding in the data is that while
Hindu-Muslim violence had risen in the decade and a half before 1995, it had declined significantly
during the period 1995–2010, except for the upsurge of violence concentrated in the state of Gujarat
in 2002.

Table 11.2 Conflicts in India (1989–2017)

Overall Per Year

Events Fatalities Events Fatalities

1989-1991 667 10302 222 3434

1992-1996 1220 10338 244 2068
1997-1999 1169 5610 390 1870
2000-2004 4224 12617 845 2523
2005-2009 4008 9110 802 1822
2010-2014 2357 4663 471 933
2015-2017 1144 2251 381 750

Source: Sundberg, Ralph, and Erik Melander, 2013, ‘Introducing the UCDP Georeferenced Event Dataset’, Journal of Peace
Research, vol.50, no.4, 523-532.
Note: GDELT event database. Years are ordered according to Lok Sabha election years.

We shall also examine alternative international resources for data on communal events. One such
resource is Global Data on Events, Location and Tone or GDELT’s event database which records
over 300 categories of physical activities around the world, from riots and protests to peace appeals
and diplomatic exchanges dating back to January 1979. The limitation however is that the events in
GDLET’s dataset are neither manually coded nor cleaned and rely entirely on parsing keywords from
newspaper reports. It is therefore possible that the events are under-reported; however, it is still
useful to examine the trends of conflict over a period of time.

Table 11.3 Riots and violence against civilians in India (2016–18)

2016 2017 2018

Number of
All Events 12742 12907 17092
Riots and Violence 11955 12076 15810
against civilians
Per Capita (per million)
All Events 9.8 9.8 12.8
Riots and Violence 9.2 9.2 11.9
against civilians

Source: Armed Conflict Location & Event Data Project (ACLED); available at www.acleddata.com

The civilian and Hindu-Muslim disputes in the GDELT database for India from 1989–2017 are
presented for each election period. The data does not indicate that there was any marked increase in
events and/or fatalities after Modi was elected. If anything, the data suggests a marked decline.
Another cross-country data resource to examine conflict trends across the world is the Armed
Conflict Location & Event Data Project or ACLED database. This is more comprehensive in mapping
each conflict event to its location, source, and other details. Further, each event is manually reviewed
as opposed to the GDELT database (which is reviewed by a computer code). However, there is a
limitation to the ACLED database—it only covers the most recent period starting from 2016 for India.

Table 11.4: State Elections and violence—are they correlated?

All Events Riots

2016 2017 2018 2016 2017 2018
Tripura 64 124 220 64 123 215
Meghalaya 44 53 93 24 45 80
Nagaland 23 32 68 14 29 58
 Karnataka 568 490 536 566 487 534
Chhattisgarh 98 102 179 33 30 63
Madhya Pradesh 99 182 175 98 180 165
Mizoram 5 8 28 5 8 28
Rajasthan 239 293 237 239 289 229
Telangana 106 174 218 104 169 211

Source: Armed Conflict Location & Event Data Project (ACLED); available at www.acleddata.com
Note: All the states listed had their assembly election in 2018.

As is evident, for the years 2016 and 2017, incidents of violence are near constant. However, as
shown in Table 11.3, there is a marked increase in 2018. A statewide documentation (Table 11.3)
reveals that this increase was concentrated in a few election-bound states in the North-East and
Chhattisgarh. The other election-bound states of Karnataka, Madhya Pradesh, and Rajasthan do not
show any increase in violence.
It is for economic historians to decide whether the economic inclusive growth record of
Narendra Modi’s tenure has been one of the best India has ever witnessed—if not the best. Our
purpose has been to present all the data available, whatever the issue, so that the public (and of
course the voter) can make informed judgements. Regardless of any informed opinion, the people will
anyway decide one way or the other.
The next two chapters contain electoral forecasts which are statistically independent of
whether the economy, or inclusive growth, or caste considerations, will determine the outcome of
Election 2019. Will this inclusive growth record go un-noticed by the voting public? A hint of what
awaits the reader—electoral forecasts are not entirely inconsistent with the hinted economic forecast.
We all keenly await the result.

1 Three studies reach the same conclusion that when the NSSO data are released, they will show
absolute poverty in India in the mid-single digits according to the official Tendulkar poverty line.
Surjit Bhalla, ‘Financing Basic Income for the Bottom 50 %’, The Indian Express, 7 January 2017;
Homi Kharas, Kristofer Hamel, and Martin Hofer, ‘The start of a new poverty narrative’, Future
Development, 19 June 2018 and Surjit Bhalla, ‘India Middle Income Now—Raise the Poverty Line’,
The Indian Express, 30 June 2018.
2 The MPC came into being in October 2016 and has by law the mandate to set monetary policy rates
in India. It is a six-member committee and three of the staff members have to be from the RBI, with
the Governor having an additional vote in case of a tie. The committee meets six times a year and, to
date, it has always followed the views of the Governor.
3 In his book, The Road to Democracy, Ruchir Sharma provides a summary of his insights from the
Limousine Liberal trips.
4 Meena Menon, ‘Chronicle of Communal Riots in Bombay Presidency (1893–1945)’, Economic and
Political Weekly, 2010, pp. 63-72.
Jayen toh jayen kahan, samjhega kaun yahan
Dard bhare dil ki zubaan…
(Where must one go, for no one understands
The grief which engulfs the heart…)
– Taxi Driver, 1954
The Empire Strikes Back?

The first chapter of this book posed a question: ‘What is Election 2019 all about?’ and inferred that it
was going to be a straight fight between Narendra Modi and the Mahagathbandhan. It also highlighted
the overall crux of several opinion polls—unanimous in predicting a hung election i.e., no alliance
would win a majority of 273 seats. If that be the case, then obviously no individual party, either the
BJP, or Congress, would win. Second, in so far as individual party seats is concerned, the consensus
had converged on 200 seats for the BJP and a 100 for the Congress party.
It is said that a week is a long time in politics, and as a natural corollary, also in forecasting. In
January 2019, opinion polls were sanguine, and unanimous, about a hung parliament. A month or so
later, while no new opinion polls have happened, the mood has gradually shifted in favour of the BJP.
This shift is based on fissures that have appeared within the Grand Alliance, suggesting that it may not
succeed as originally expected.
My expectation is that the forecast contained in this chapter will not materially change between
now and counting day. A hung parliament is unlikely—I differ from the consensus in this important
regard.
This chapter is primarily about the old dominant national party of India—the Congress. The next
chapter will document the rise and rise of the ‘new kid in town’, the BJP. As the line in old Westerns
went, ‘This town a’int big enough for both of us.’
2018 Assembly Elections—a Mixed Blessing for the Congress
It all started with the assembly election results in mid-December 2018—initially expected to win
only in Rajasthan, the Congress won all three states including Madhya Pradesh and Chhattisgarh,
which it did rather convincingly. In Madhya Pradesh, it won more seats, but marginally less votes.
The unexpected success changed the CW (Conventional Wisdom) forecast for 2019 from done (for
the BJP) to hung. This also changed expectations, within the yet to be formed alliance. Prior to the
wins, the Mahagathbandhan was unsure about choosing a leader, considering Rahul Gandhi had
somehow failed to convince the regional satraps of his credentials. But the victories compelled
another look, and overnight, he became the ‘unofficial’ leader of the Opposition, which was anyway
given an official stamp within his own party the day he had taken over as vice president. There was a
renewed sense of hope—media circles were abuzz with how the Congress could now easily win 140
odd seats—after all, this was the number in 2004 when it had formed the government. So why not in
2019?
This is precisely where the problems began for the alliance. As a rule, sage political advice is
to out-perform expectations; was 140 seats considerably above normal expectations? Electoral math
shows that unless this is the most unusual election in India’s history, the Congress winning 140 seats
has a very low probability—much lower than even Rahul Gandhi becoming the next prime minister.
The Alliance Math
There are strong assumptions about voting behaviour embedded in alliance math. An alliance works
through seat-sharing, and seat-sharing according to the counter-factual i.e., what did the parties
comprising an alliance achieve if (or when) they fought individually. This performance forms a
legitimate basis for the new alliance seat management.
The first indication of the problematic electoral math facing the alliance came from the
announcement that the two major parties in UP—BSP and SP—had agreed to a seat arrangement
withoutthe Congress. Out of eighty seats, seventy-five were to be shared equally between the two
parties, while two seats were left for the Congress, and three for the Rashtriya Janata Dal or RJD.
An unbiased calculation based on performance yields the same decision. In 2014, the number
one and number two positions for the respective parties in UP were as follows: Congress (2, 6), SP
(5, 31) and BSP (0, 34). Objectively speaking, the Congress should contest around seven seats, while
the BSP-SP combine should contest the remainder. This is almost what the BSP-SP had decided—
thirty-eight seats for the BSP, thirty-seven for SP, three for the Rashtriya Lok Dal or RLD, and two
seats for the Congress.
As was expected, the Congress did not view the decision objectively. Instead, it retaliated by
inducting the reluctant but charming dynast, Priyanka Gandhi, sister of Rahul, into the battle. As a
result, a strong two-way fight became a two-and-a-quarter way battle.
As is well known, alliances work best if each party has a large individual ‘glue’ factor, and this
may also have been a factor for the SP-BSP combine to exclude the Congress. Historical record
suggests that the strongest glue is with the BSP and the weakest with the INC. In Lok Sabha elections,
the Congress’ vote in UP, since 1996, has averaged at eleven per cent and in a wide range, six to
eighteen per cent; both the SP and BSP have averaged at twenty-three per cent in a narrow eight-point
range—twenty to twenty-eight per cent. Regardless of what the calculation spin is, a SP-BSP alliance
sans the Congress is what the most objective calculation yields.
This ‘humiliation’, albeit objective, was a deal-breaker for the Congress. It had envisioned a
path to 140 seats (back to the ‘norm’ since 1996) and must have realised that it would fail to come
anywhere closer to that number without a significant presence in either UP or Bihar. Hence, its
decision to manly accept the SP-BSP challenge and contest all the eighty seats in UP. Though there
might be agreements between the three parties, the basic math is not going to change.
How will the battle for UP be resolved? Before answering the question, a brief foray into the
origins and outcomes of past electoral alliances.
Electoral Alliances—Origins
An alliance is a collection of regional parties surrounding their ‘core’. The BJP contested its first
election in 1984, and in its second foray in 1989, it obtained a national vote share of 11.5 per cent
and eighty-six seats in parliament, almost twice of what the Congress had obtained (forty-four) in
2014. BJP’s rise and Congress’ fall is best exemplified by this simple statistic—the former in its
second election gained twice as many seats as the latter achieved in its sixteenth election and 130
years of existence. The conclusion: the BJP is on the rise, while the Congress is a declining national
party, and the fall is on a rather slippery slope. More than an alliance, a large vote swing in its favour
is what is required to bolster the Congress’ fortunes.
Prior to 1989, the only other national party of any consequence was the Communist Party of
India (Marxist), but it didn’t have a pan-Indian presence—a force to reckon with, it had an interrupted
and impressive run in West Bengal and was part of the alliance in Kerala. Beginning the late 1980s,
and particularly post the 1984 election, the hegemony of the Congress began to be challenged. In
retrospect, the true outlier in Indian elections was 1984 when the Congress won an out of the ball
park victory. But soon thereafter, a former member of the Congress party and once India’s Finance
Minister, V.P. Singh began to challenge Rajiv Gandhi, and the floodgates opened.
The preferred choice for a wannabe political leader has been to either form or join a regional
party, or at least since the mid-1980s to join the BJP. There was no incentive for thinking individuals
to join the Congress, because the dynastic principle made it impossible for anyone to be a top dog,
other than a Nehru-Gandhi. In all probability, this simple fact may have been responsible for the steep
decline in the Congress’ fortunes in a span of twenty-five years—from 415 seats in 1984 to one-tenth
of that figure in 2014.
In time, regional satraps established their hegemony, and became aware that aligning with the
Congress would prove a death knell for their respective parties; what they however acquiesced to
was engage in a ‘friendly’ pre-poll competition and also collaborate after the polls. In this scenario,
the voter was spoilt for choice—all political parties began drawing on the Congress’ voter. As a
consequence (and you don’t need to be a mathematician to calculate), the Congress’ vote share
declined structurally. Over the next twenty-five years, its vote share went down by twenty ppt and the
BJP’s vote share saw an increase in the same proportion. Sample this: the Congress’ vote share in
1989 and 2014 was 39.5 per cent and 19.1 per cent, respectively; whereas the BJP’s vote share was
11.4 and thirty-one per cent. Both differ by exactly (within a small decimal point) 20.5 ppt.
The Politics of Coalitions
The arithmetic of adding up coalition votes requires that the difference between the alliance and
BJP’s vote (in previous elections) be large and somewhere in the neighbourhood of 5-10 per cent.
Note that as a general rule, the national incumbent party enjoys a jump in the state vote in the Lok
Sabha elections. This jump averages 2 to 4 ppt.
The inherent contradiction in an anti-BJP alliance is that the existence and popularity of regional
parties is dependent on competing with the Congress. The Congress has negative coattails in the eyes
of many—an alliance with it would mean less joint votes than otherwise. I present several examples
of electoral math and alliance failures below to ascertain if they place the Mahagathbandhan in a
comfortable position in 2019.
UP Alliances
SP-INC alliance: In the 2014 Lok Sabha election, the two parties fought separately and garnered
29.7 per cent of the vote; in the 2017 state election, a much-publicised SP-INC partnership
yielded nearly two per cent less votes i.e., twenty-eight per cent.
SP-BSP alliance: In the 1993 UP state election, the two parties fought together and obtained
twenty-nine per cent of the vote; three years later, they fought separately and obtained forty-one
per cent of the vote.
UP results: Alliances have led to less votes, than if coalition members fought separately. The
BJP’s vote share in UP was 42.3 per cent in 2014. The BSP and SP together obtained 41.8 per
cent. Going into 2019, the BJP is marginally ahead; the mildest of swings towards the BJP
(around three per cent) would mean that it would not win less than fifty seats.
West Bengal
The firebrand Chief Minister of West Bengal, Mamata Banerjee, obtained thirty-nine per cent of the
vote in 2014, and 44.9 per cent in the 2016 assembly elections. In 2019, she does not benefit from
aligning either with the CPM or INC—she is credited with having vanquished both in state and
general elections. The gap between her party and the INC plus votes is about ten ppt. The Congress is
faced with a dilemma—if it really wants the Mahagathbandhan to win, then it has to be a very junior
partner in both West Bengal and UP. This the Congress party is not willing to do, which in turn should
help the BJP increase its vote share and seats in West Bengal in 2019.
The Success of the Mahagathbandhan—Bihar in 2015
While there are several notable examples of the Mahagathbandhan’s failure, the one outstanding
success was in the 2015 Bihar assembly elections. There are obviously several reasons for what
many in the media called a ‘surprise’ victory, but if one were to cite a single reason for the outcome,
then it was their joint ex-ante vote share which was about sixteen ppt higher than that of the BJP.
In the 2014 Lok Sabha elections, the BJP won twenty-two seats in Bihar and the three-party
conglomeration comprising the JDU, RJD, and INC (which were fighting separately) won eight seats.
The vote shares were as follows: BJP, 29.4 per cent; the alliance, 44.3 per cent. A year later in the
2016 assembly election, the BJP’s vote share declined by five ppt, and the alliance’s came down by
2.5 ppt. The vote share comparison was as follows: BJP, 24.4 per cent; the alliance, 41.8 per cent. As
it turned out, it was obviously a no contest. The alliance worked because of the big gap in vote shares
—fifty-three seats for the BJP, and 178 for the combined alliance.
There is one clear inference from all the data across states, time, and the type of election. Pre-
poll alliances tend to work if there is a large ex-ante difference in vote shares between the alliance
and the Opposition. However, the problem with allthe alliances in the offing (Bihar, Maharashtra,
UP) is that the joint vote share of the alliance is less than what the BJP and its partners obtained in
2014. It’s been proven time and again that in a two-party contest, small differences in vote shares get
magnified. This means that for Narendra Modi to be unseated from a single-majority status will not be
easy, and for the NDA to be unseated as the single-largest coalition seems very, very, difficult.
What seems most likely, and this is bereft of any ideology or emotion, is that the 2019 election
could be a referendum on the survival of the Congress as a major national party. It will centre as
much around Modi’s popularity as the spectre of a Congress-muktBharat or an India without the
Congress party.
The Vote for Congress—What to Expect in 2019
Let us now examine state by state of what can be expected of the Congress in 2019 by using several
methods to yield a forecast.
The average of all the positive changes in the vote shares of the Congress party, 1957-2014, is
four ppt. All the negative changes average 5.5 ppt. The highest swing, 8.2 ppt in 1980. The largest
decline, 9.3 ppt in 2014, almost identically equal to the 1977 debacle. There is a reasonable
likelihood that the 2014 vote share was exceptionally low—hence, a maximum 7.5 per cent swing
assumption is possible. As is normal for such calculations, the swing is applied uniformly to all the
constituencies.
There are three suppositions that we obtain and all of them yield different estimates. The final
and fourth estimate is based on my judgement call on the swing model and other models
(assumptions).
Congress’ Vote History: Revival Most Unlikely
Chart 12.1 reports both the vote share and seats obtained for all the last sixteen elections. What we
observe is that the Congress held on to a steady vote share of around forty-five per cent till the mid-
1970s. It lost about one ppt of vote per year between 1980 and 1996, and stabilised at around twenty-
eight per cent till (and including) the 2009 election. Since then, there has been a large jump
downwards of around nine ppt as witnessed in 2014.
The state elections subsequent to the last Lok Sabha election have also not been encouraging for
the Congress. The weighted vote share in state elections post 2014 averaged 21.9 per cent compared
to 23.1 per cent between 2009 and 2014. The post-2014 number includes the Congress’ wins in the
Hindi heartland in November 2018. The BJP, on the other hand, increased its vote share by a robust
eight ppt; from 19.7 per cent between 2009–2014 to 27.6 per cent between 2014–2018.
Yield of Votes—Vote Share Per Seat
If the Congress were to move from nineteen per cent of the national vote to twenty-three per cent,
there would be a swing in its favour of four ppt. One method of estimation (forecast) involves
approximating the yield per vote i.e., what are the number of seats a party can expect if its vote share
increases (or decreases) by one per cent?
This is where the Congress story, in terms of numbers, becomes somewhat alarming.
Chart 12.2 documents the historical yield for the INC. In most elections, the Congress obtained
approximately seven seats for each one per cent vote share. Although there are fluctuations, there is
also a broad consistency—i.e., until 1989 when the Congress yield was at nearly the same level as in
the 1977 ‘Emergency election’.

In hindsight, I would say that this was a ‘surprise’ result; most of us believed that the 1977 election
result was a genuine Emergency-induced aberration, yet the polling data suggests that the ‘Bofors’
election in 1989, was just as bad. There is however a short recovery in the INC yields, before the
resumption of the downward trend. If the 2009 outlier is ignored, then the 2014 yield matches the
prediction of the ‘trend line’ from 1991 onwards.
What kind of recovery in seat yield can be expected in 2019? The first marker in this recovery is
four, which is the strike rate obtained in 1999. If this were to happen, the recovery rate, co-
incidentally, would be almost the same as obtained by Indira Gandhi in 1980 (the move then was from
4.4 in 1977 to 7.6 in 1980). If both these events were to happen (the Congress’ vote share increase of
five per cent to 24.3 andthe strike rate increase from 2.3 to four), then one obtains a forecast of
ninety-seven seats for the party.
This ninety-seven seat forecast is based on the assumption that the people of India are as frustrated
with the Modi regime in 2019 as they were with the Congress in 2014. If this assumption is
unrealistic, and I believe it is, then the Congress will have to be satisfied with a number considerably
below ninety-seven.
A realistic possibility is a three ppt increase in the vote share to twenty-two per cent anda
recovery of strike rate to three; this yields a figure close to sixty-nine seats for the Congress. All of us
can of course play with these numbers (that is why the historical record has been presented) and come
up with our own guesstimates.
I want to present yet another method for estimating the number of seats a party can obtain—this
is based on margins of victory in the previous election. This is a time and pollster-honoured method,
and one that I also lay great store by. Opinion polls have two mandatory questions in each poll—who
are you going to vote for, and who did you vote for the last time. If answers to these questions are
obtained, the circle of information is complete—one can statistically estimate the election outcome
with reasonable accuracy, provided the sample is adequate, and the polling method sound.
What Do the 2014 Margins Say About 2019?
Table 12.1 presents a complete picture of how the INC fared in 2014. Out of 543 seats, the Congress
won just forty-four and lost 420. The frequency distribution of the wins and losses are also presented
in the table, which is according to the margin of victory (or loss)—the winning margin is the
difference between the voting percentage of the party (Congress) and the party which is in second
position in a given constituency. For calculating losses, the margin is the difference between the vote
share of the losing party and the vote share of the winning party.
If a positive swing is assumed, then it means that the seats won in the previous election are
retained. The analyst can now choose to forecast from a menu of assumptions about margins. If the
assumption is made that there is a positive swing of five per cent, then the Congress can be expected
to win seventy-eight seats—forty-four that it won in 2014, plus the thirty-four that it lost by five ppt or
less. This is Scenario 1 (referred to as Sc1).
Two other assumptions about the magnitude of the swing yield the upper bound (swing of 7.5 per
cent) and lower bound (a ‘mixed’ swing of five per cent i.e., it wins all the seats that it lost with a
margin of five per cent, but loses half the number of seats it won with a margin of less than five per
cent).
Scenario 2 incorporates the assumption of a very large 7.5 per cent increase in vote share—as
mentioned earlier, such a large swing would bring the Congress back to its glory of the last twenty
years or so. But even this large assumption gets the Congress back to only ninety-seven seats.
Interestingly, this is equal to, or slightly less than, the consensus of the opinion polls.
The third estimate is essentially the Sc2 estimate (ninety-seven seats) minus half the low margin
(<5 per cent) winning seats that the Congress might lose. By definition, we relax the assumption of a
uniform swing with this supposition. Even this loss is expected to be half of the nineteen seats won by
the INC with a less than five per cent margin. This yields an estimate of sixty-nine seats for the
Congress—a twenty-five seat jump from the 2014 result.

Table 12.1 How did the INC fare in 2014 General Elections

Seats Contested Seats Winning Margin Losing Margin

(%) (%)
Won Lost <10 >=10 <10 >=10
Big States 502 39 387 26 13 62 325
Andhra Pradesh 42 2 39 1 1 4 35
Assam 14 3 10 3 0 3 7
Bihar 40 2 10 1 1 3 7
Chattisgarh 11 1 10 1 0 3 7
Gujarat 26 0 25 0 0 2 23
 Haryana 10 1 9 0 1 2 7
Jharkhand 14 0 9 0 0 2 7
Karnataka 28 9 19 7 2 10 9
Kerala 20 8 7 7 1 7 0
Madhya Pradesh 29 2 27 0 2 5 22
Maharashtra 48 2 24 2 0 2 22
Odisha 21 0 21 0 0 5 16
Punjab 13 3 10 2 1 7 3
Rajasthan 25 0 25 0 0 3 22
Tamil Nadu 39 0 39 0 0 0 39
Uttar Pradesh 80 2 65 0 2 2 63
West Bengal 42 4 38 2 2 2 36
Small States 41 5 33 4 1 9 24
Total 543 44 420 30 14 71 349

Source: ECI; author’s database on Indian Elections.

Note: Andhra Pradesh includes Telangana.

Table 12.2 Forecasts for 2019 based on 2014 margin distributions – Congress

Seats Forecasts for 2019 - Margins

Total Won Lost Sc1 Sc2 Sc3 ScF
Seats
Big States 502 39 387 69 85 61 54
Andhra Pradesh 25 2 23 5 6 5 2
Assam 14 3 10 4 5 3 2
Bihar 40 2 10 3 4 3 1
Chattisgarh 11 1 10 4 4 4 6
Gujarat 26 0 25 0 1 0 3
Haryana 10 1 9 1 2 1 1
 Jharkhand 14 0 9 1 2 1 1

Karnataka 28 9 19 13 16 11 10
Kerala 20 8 7 13 14 10 6
Madhya Pradesh 29 2 27 4 6 4 4
Maharashtra 48 2 24 2 3 2 2
Odisha 21 0 21 2 4 2 0
Punjab 13 3 10 8 9 8 4
Rajasthan 25 0 25 1 2 1 5
Tamil Nadu 39 0 39 0 0 0 0
Telangana 17 0 16 0 0 0 1
Uttar Pradesh 80 2 65 2 3 2 4
West Bengal 42 4 38 6 6 6 2
Small States 41 5 33 9 12 8 3
Total 543 44 420 78 97 69 57

Source: ECI; author’s database on Indian Elections.

Notes: Sc1—INC wins all seats which it had lost with margins less than 5 per cent.
Sc2—Sc1 plus INC wins half the seats it had lost with margins in the 5-10 per cent range i.e. margin of victory is 7.5 %
Sc3—Sc1 minus INC loses half the seats it had won with margins <= 5 per cent.
ScF—Our most likely scenario incorporating all factors (swings, margins, judgement).

The final column labelled ScF, is the forecast for the Congress on its own in 2019. This includes the
information contained in Scenarios 1 through 3 plus our judgement of the likely outcome. The
previous three estimates were mechanical in nature i.e., they conformed to the principle of Untouched
By Human Hands or UBHH (these forecasts do not incorporate the changing alliances and seat
arrangements between different parties). In the aggregate, one obtains only fifty-seven seats for the
Congress, just a thirteen seat jump from the low forty-four seats that it got in 2014. The good news is
that Karnataka is likely to be the only state with a double-digit outcome in 2019—in 2014, in no state
did the Congress reach this milestone.
Finally, the Congress’ seat forecast points towards two facts—history and math are not in its
favour. If one is proven wrong, and the Congress reaches the triple-digit mark forecast by opinion
polls, it will be on a strong path to recovery. If one is proven right, then the Congress will very likely
have to re-invent itself, and seriously stop being feudal and dynastic. Election 2019 is show time for
the Congress party.
Woh Sikandar hee doston, kehlata hai
Haari baazi ko jeetna jisey aata hai…
(The one who can snatch victory out of a losing battle is a king like, Alexander… )
– Jo Jeeta Wohi Sikandar, 1992
A Second Innings for Narendra Modi?

Amongst several others, there are two perceptible points about the political trajectory of Narendra
Modi. He became chief minister without ever having been an MLA. 1 Second, he became prime
minister without being a Member of Parliament. One could also add a third ‘firsts’ to this list—he
also won his first state and national election by huge margins. Yet another act also makes it more
interesting—he won three consecutive elections in Gujarat (2002, 2007, and 2012). Will history
repeat itself in 2019, and if so, will he get a hat trick in 2024?
But then, 2024 is some distance away, and for 2019, the numbers indicate that there is not much
of a chance of him not being the next prime minister. Indeed, history may just as well repeat itself,
both for the Congress and the BJP.
We looked at the Congress side of the ledger in the previous chapter; the nature of alliance math
being an important consideration. This chapter is about the expected seats for the BJP in 2019.
As far as alliances go, the BJP lost Chandrababu Naidu (TDP) as an ally in Andhra Pradesh, but
got Nitish Kumar (JDU) on its side in Bihar. The traditional allies—the Shiromani Akali Dal or SAD
in Punjab, and the Shiv Sena in Maharashtra remain, while the All India United Democratic Front or
AIUDF in Assam is uncertain at the time of writing this chapter. Potentially, the big alliance gain for
the BJP may perhaps be in Tamil Nadu, where it has entered into an alliance with the All India Anna
Dravida Munnetra Kazhagam or AIADMK, and other smaller regional parties.
However, our central concern in this chapter is about the number of seats that the BJP might
obtain on its own in 2019. Towards that issue we now turn. But first, some preliminaries about voting
and opinions.
Negatives Matter More
In my career as a pollster, I have been struck by how the negatives matter more than the positives.
That great observer of human nature, William Shakespeare, correctly concluded several centuries
ago:
The evil that men do lives after them;
The good is oft interred with their bones.
The hope of the Opposition hinges on this very fact: the people will definitely act on the negatives.
The young and upcoming pollster, Pradip Bhandari informed me that during his several road trips to
different parts of India in January 2019, he did not detect any hatred for Modi. But, on the other hand,
some sections of the English news media provide ‘evidence’ on a daily basis that hate (or strong
dislike) for Modi is widespread, and much beyond the confines of the elite Lutyens’ quarters.
The overall lack of negative views about Narendra Modi makes it difficult to believe that his re-
election may be problematic. I am aware that truth shall elude us until counting day, but Modi’s
popularity ratings should not be worrying him or his party. Of course, these ratings can be wrong—
and/or the opinion polls can as well be.
Opinion Polls on A Taxi Ride
This may come as a surprise for many, but taxi rides work as an important input for assessing
political and social trends. We now live in times when taking a taxi or cab is not a poor, but a better
substitute, for driving one’s own car and there is no one I know who won’t strike up a conversation
with the cabbie on a number of issues, and someone like me, on politics for sure.
Much like the boatmen of Banaras, the taxi drivers often have an inside view of the mood of the
nation. I know that many equate a taxi driver conversation with lazy journalism or analyses, but
tomorrow if somebody were to finance my poll, it would only include auto and taxi drivers (at least
about eighty per cent of the sample).
My informal poll in January 2019 revealed an overwhelming support for Modi in Delhi, about
eighty per cent of taxi drivers gave a thumbs up for the prime minister. It may be recalled that the
strike rate for the Aam Admi Party or AAP among auto drivers in 2015 was also eighty per cent.
Interestingly, a majority of taxi drivers did not mention the name of a particular party, but
invariably identified the leader or star attraction of a party—Modi, Rahul, Mayawati, or Akhilesh.
This proved that we had, for a large part, moved towards a Presidential style of elections—today,
leaders are far more important than political parties.
I concluded that there were several reasons for the taxi drivers’ ratification of Narendra Modi.
Amongst the first was that he is incorruptible and in equal measure, that he wants to help India
progress. Another surprising view (at least for me) was that Modi had helped India’s image gain
credence abroad. I thought international relations mattered only for the educated elite, but taxi drivers
around the world, and in India, are now definitely more aware than they were ever before. Then came
the conclusion—look, Modi must be given a second chance only because he fully deserves it; five
years is too short a time period for correcting the past mistakes.
Has he done anything that has offended them, something he should not have done?
While demonetisation seems to have been forgotten by most in this category (excluding the urban
elite), two negatives came to the forefront. The first being the cow policy; I have yet to meet a person,
rich or poor, elite or non-elite, who can find any rationale for protecting cows or for the inane policy
of banning cow slaughter. Now that the Congress in Madhya Pradesh has announced a policy
endorsing, and replicating, the outlandish cow shelter policy of the RSS, it must be the case that I met
all the wrong people.
If a driver said, and many often did, that this is a scam-free government, I would ask: what about
the Rafale deal? There was an acknowledgement that it was possibly a scam—at least all taxi drivers
agreed to having had heard about it. But it was also brushed aside by many as follows: it is just one
corrupt deal; there were many before.
Let me state that I have been conducting these taxi driver interviews for more than a decade.
There is an interesting parallel here apropos Dr Manmohan Singh (and the Congress), and the BJP
and Narendra Modi. For a long time, and even now, no one ever thought that Dr Singh was personally
corrupt, which is what is said about Narendra Modi as well. The difference however, and a near
universal feeling, is that it was impossible for Manmohan Singh to have not known about the large
scale corruption in the party; and he tolerated it. Two aspects are however different today from
yesterday—Modi is in control, unlike Manmohan; and Modi is openly scornful of corruption, and
people believe that he acts against corruption.
I have faithfully recounted what I recorded—you might say that this closely corresponds to my
own views. This is true. But then the cabbies’ views corresponded with mine even when I was an
ardent supporter of Manmohan Singh, right through the 1990s and until 2010, when I changed tack for
much the same reasons as the taxi drivers.
The Modi Effect on BJP’s Vote Share
Back to the Modi effect—how do we measure it? The last time the BJP performed well (in terms of
vote share) was in the 1998 and 1999 elections with 25.6 and 23.8 per cent of the national vote,
respectively. The last time the Congress performed impressively was in 1996 and 1999—their vote
share equal to 28.8 and 28.3 per cent, respectively. Again in 2009, the vote share was close to these
levels—28.6 per cent.
It is contended by many that the BJP’s vote share of thirty-one per cent in 2014 was an
inaccurate reflection of the ‘inherent’ vote for the BJP. In other words, it was an outlier, was bumped
up temporarily, and that it would revert to its ‘original’ magnitude in 2019.
The BJP’s vote share in 2014 of thirty-one per cent represented a 12.2 ppt increase over 2009.
What might have been the BJP’s ‘normal’ vote share in 2014? If one were to exclude 2014, then the
best election years for the party were 1998 and 1999—the BJP obtained 182 seats and an average
vote share of 24.7 per cent (25.6 in 1998 and 23.8 per cent in 1999, respectively).
Given these facts, one can estimate the Modi effect on the BJP vote in 2014 as follows: we shall
compute, at an individual state level, the difference in BJP vote shares between 2014 and the late
1990s. The vote share in 2014 is so far the best BJP has ever done; the vote share in the late Nineties
is the best that the BJP achieved previously; the difference is 5.4 per cent of the national vote, which
is one approximation of the Modi effect. In 2014, the Congress vote share declined by 9.3 per cent
(19.3 per cent in 2014; 28.6 per cent in 2009). Some of this large decline likely went to the BJP, and
some went to regional parties. If some of the Congress’ secular decline went to the BJP, then the net
Modi effect would be somewhat smaller than 5.4 ppt.
Charts 13.1 and 13.2 document the performance of the BJP since their second election in 1989.
Several noteworthy facts are embedded in this—first, that in the three elections in the mid-1990s, the
BJP averaged around 175 seats. During these years, the yield of the BJP vote was around 7.5 i.e.,
each one per cent vote brought the BJP around 7.5 seats (the same level as in 1989). This increased to
9.1 in 2014. One estimate of the extra seats that the BJP may have obtained in 2014 due to the Modi
effect is obtained by assuming that the seat yield of the BJP remained at 7.5—hence, with thirty-one
per cent of the vote, the BJP should have secured 232 seats; hence, by obtaining 282 seats, the Modi
seat effect in 2014 was fifty seats.
The opinion polls suggest only a one ppt decline for BJP’s vote in 2019; thus, my first estimate
for the BJP in 2019 is 225 seats—under the assumption that there is no Modi effect in 2019 and that
there is a one ppt decline in the vote share as per the opinion polls. It is interesting to note that this
estimate is near the upper end of most opinion poll forecasts (also note that neither forecast factors in
the unfortunate and uncertain effects of the Pulwama attack in the Valley in February 2019).
My second estimate is that the Modi effect persists in 2019 in terms of yield i.e., a vote share of thirty
per cent and a seat yield of 9.1—this indicates a razor-thin majority estimate of 274 seats for the BJP.

Let us now turn to our forecast of seats based on the margin of victories and losses in 2014 (as was
done for the Congress in Chapter Twelve, p. 181). It may be recalled that for the Congress, we had
assumed that there would be a positive swing in its favour. For the BJP however, we shall assume a
negative swing—a positive swing will automatically mean a seat share greater than the 282 reached
in 2014.
 My aim is to be conservative, or to go for lower bound estimates. The evidence from the state
elections post 2014 however suggests that there will be a positive vote swing for the BJP. If one goes
by assembly elections, the swing for the BJP for years 2014 through 2018 were as follows: 9.5 per
cent; 6.3; 5.6; 16.2; and 1.5 per cent, respectively i.e., 9.5 per cent in the state elections held in 2014
and 1.5 per cent in the state elections held in 2018 (including obviously the elections in Karnataka,
Chhattisgarh etc.). It is interesting to note that the 2014 state elections, held at the same time as the
Lok Sabha elections, closely mirror the vote swing for the Lok Sabha.

Table 13.1 : How did the BJP fare in 2014 General Elections

Seats Contested Seats Winning Margin (%) Losing Margin (%)

Won Lost <10 >=10 <10 >=10
Big States 502 256 135 64 192 30 105
Andhra Pradesh 42 3 22 2 0 12 10
Assam 14 7 6 2 5 4 2
Bihar 40 22 8 10 12 3 5
Chattisgarh 11 10 1 3 7 1 0
Gujarat 26 26 0 2 24 0 0
Haryana 10 7 1 1 6 0 1
Jharkhand 14 12 2 5 7 2 0
Karnataka 28 17 11 8 9 6 5
Kerala 20 0 18 0 0 1 17
Madhya Pradesh 29 27 2 5 22 0 2
Maharashtra 48 23 1 1 22 1 0
Odisha 21 1 20 1 0 4 16
Punjab 13 2 1 1 1 0 1
Rajasthan 25 25 0 5 20 0 0
Tamil Nadu 39 1 8 0 1 2 6
Uttar Pradesh 80 71 7 17 54 2 5
West Bengal 42 2 40 1 1 2 38
 Small States 41 26 11 8 18 1 10
Total 543 282 146 72 210 31 115

Source: ECI; author’s database on Indian Elections.

Even if the 2014 state elections are ignored, the average swing for the BJP in 2015–2018 (weighted
by size of the state electorate) was a large 7.5 per cent; while for the Congress it was -0.7 per cent. In
2019, even a two per cent extra vote for the BJP will increase its seat share to 297 (an average
historical yield of 7.5); just a one per cent swing will get them to 290. In other words, it is not
difficult to imagine a scenario whereby the BJP gets close to the second magic figure of 300.
What if the BJP gets less votes than it did in 2014, as predicted by several pollsters. In other
words, what if there is a negative vote swing? The methods employed by me to compute the expected
seat loss are a mirror image of the calculations done to estimate the seat gain for the positive vote
swing for Congress. Scenario 1 (Sc1 in Table 13.2) assumes that the BJP loses all the small margin
(<5 per cent) constituency victories in 2014 i.e., an against swing of 5 per cent. This yields twenty-
nine losses for the BJP (Table 13.1); hence, Sc1 computes the number as 282 minus twenty-nine, or
253 seats.
The second scenario (Sc2) is where the average margin is a decline of 7.5 ppt. It may be
recalled that the largest loss for the BJP ever was the 3.4 ppt decline in 2009—from 22.2 per cent in
2004 to 18.8 per cent in 2009. What we are assuming in Sc2 is more than twice this decline. If this
were to happen, it would be a wave election, somewhat like the 1977, 1980 or 1984 elections.
However, the likelihood of this happening, going by past precedents, is very small. But even if the
wave election were to happen, the BJP will, by this calculation or method, be down to 232 seats.
Scenario 3 parallels Sc3 for the Congress—except this time, the estimate is that the BJP loses
all the seats it won with <5 per cent margin (i.e., 253 seats) plus it wins half the seats it lost with a <5
per cent margin (twenty-four seats). This yields a Sc3 estimate of 265 seats.
Regardless of the method chosen, the lowest estimate is 232 and the highest: 265. One reason for
these ‘high’ estimates for the BJP is that it won the 2014 election with very high margins. In other
words, the Opposition has significant hurdles to overcome to claw back some of the lost votes.
One example of testing the impediments is to analyse the recent state elections in two states that
the Congress won in 2018—Rajasthan and Madhya Pradesh. It is a reasonable argument that
December 2018 was close to the nadir of the Modi government, at least in terms of popularity.
However, it needs to be mentioned here that these elections were before the highly popular
interim Budget presented by the Modi government on 1 February 2019. One parameter of the
popularity of the Budget was the way in which the satta bazaar estimate for the BJP moved up to 225
seats from 200. Hence, a very reasonable assumption is that the BJP vote was at its bottom in
December 2018—and at this stage, the BJP was winning thirteen of the twenty-five parliamentary
seats in Rajasthan, and seventeen of the twenty-nine seats in Madhya Pradesh. This is not an estimate,
but the actual summation of assembly votes according to parliamentary constituencies.
In addition, there is an assembly to Lok Sabha vote effect for the government at the Centre, in
this case the BJP. Historically, this effect has averaged around 2 to 4 per cent for the incumbent. In a
two-party close fight, a three per cent extra vote can have a large outsized effect on the seats. (For
instance, the vote and seat swings in Kerala on the basis of small differences in vote shares.)
 Forecasts for 2019 based on 2014 margin distributions – BJP

Total Seats Scenarios for 2019

Seats
Won Lost Sc1 Sc2 Sc3 ScF
Big States 502 256 135 230 211 241 244
Andhra Pradesh 25 2 23 2 1 2 3
Assam 14 7 6 6 6 7 9
Bihar 40 22 8 17 15 19 19
Chattisgarh 11 10 1 7 7 8 5
Gujarat 26 26 0 26 25 26 24
Haryana 10 7 1 7 7 7 7
Jharkhand 14 12 2 10 9 11 8
Karnataka 28 17 11 14 12 15 12
Kerala 20 0 18 0 0 0 2
Madhya Pradesh 29 27 2 25 24 26 25
Maharashtra 48 23 1 23 23 23 23
Odisha 21 1 20 0 0 1 6
Punjab 13 2 1 1 1 2 4
Rajasthan 25 25 0 23 22 24 20
Tamil Nadu 39 1 8 1 1 1 3
Telangana 17 1 16 1 1 1 2
Uttar Pradesh 80 71 7 65 60 68 62
West Bengal 42 2 40 2 2 2 10
Small States 41 26 11 23 21 24 30
Total 543 282 146 253 232 265 274

Source: ECI; author’s database on Indian Elections.

Notes: Sc1—BJP loses all seats which it had won with margins less than 5 per cent.
Sc2—Sc1 plus BJP loses half the seats it had won with margins in the 5-10 per cent range.
Sc3—BJP loses all seats with margins less than 5 per cent plus it wins half the seats it had won with margin 5-10 per cent.
ScF—Our most likely scenario incorporating all factors (swings, margins, judgement).
A ‘Preferred’ Estimate of BJP Seats
Just as in the previous chapter, the final column shows my ‘mixed’, but final estimate for the seats that
the BJP is likely to obtain in 2019—ScF. This estimate is based on multiple models, multiple
considerations, and multiple factors. Some of the factors considered are: swings in assembly
elections, nature of alliances, considerations of the most recent assembly elections, historical trends
in votes and yields etc. All these determinants lead to my razor-thin majority estimate for the BJP on
its own—274 seats.
This estimate is before the terrible Pulwama tragedy. A ten per cent increase in seats for the BJP
will get it past 300; a ten per cent negative impact will get them close to 250 seats.
The estimates for Uttar Pradesh deserve special mention. All three scenarios, Sc1 to Sc3, yield
estimates upwards of sixty seats for the BJP in 2019. This is understandable, given the BJP had large
victory margins, and that was because the BSP and SP had fought separately in 2014. However,
despite the alliance, I shall say this here: the BJP is likely to win around sixty-two seats in UP in
2019. Let me explain how. To put it simply, alliances in general, and particularly in UP, have not
worked. In 2014, the BJP’s vote share was 42.3 per cent; the BSP and SP together obtained 41.8 per
cent, and the Congress, 7.5 per cent. As noted in Table 13.1, the BJP victory margin exceeded fifteen
per cent in forty-one seats. In the current scenario, with the Congress sitting out of the alliance, it is
closer to (but not quite) a three-way fight. We also have to factor in the 2017 assembly election
results when the Congress and SP were allies. Their joint vote share was twenty-eight per cent, down
1.7 ppt from the 29.7 per cent vote when they had fought separately in the 2014 Lok Sabha election.
There is also the question of how many SP votes move towards the BJP or the Congress. All things
considered, an estimate of sixty-two seats for the BJP in UP does not look like an extreme bet.
2019: Seat Conclusion
In this entire exercise, maybe we circled the whole world and came back to where we started—
practically the same electoral result in 2019 as in 2014. The BJP loses a few seats and the Congress
gains a few. If one is proven right, then there will be consequences for political leaders, political
parties, and most importantly, India. Narendra Modi’s stature as one of the world’s most successful
politicians will have been established. If our figures are right, then it might also well mean the
beginning of the end of the Nehru-Gandhi dynasty.

1 I am basing this conclusion on a conversation with Lord Meghnad Desai, a person known for his
sound knowledge of history, economics, and politics.
Hum logon ko samajh sako toh samjho dilbar jaani
Jitna bhi tum samjhoge utni hogi hairani
Apni chattri tum ko de den kabhi jo barse pani
Kabhi naye packet mein beechen tumko cheez purani
Phir bhi dil hai Hindustani…
(You may try understanding us
But the more you do, the less you will
Even as we can lend you an umbrella in rain
We can also sell you old stuff in a new packet
But all said and done, our hearts are Indian….)
– Phir Bhi Dil Hai Hindustani, 2000
Democracy Helps Usher in a New India

Our study of India reaches the one big conclusion, also reached by many students of history, politics
and of course, economics—India is one big, noisy, messy democracy.
As with all democracies, over time and space, the feeling of some is: Armageddon is here—or
as we say in Hindi, ‘Ab marr gaye,’ (Now we are dead). But we deal with it, life goes on, India
marches forward. The rest is secondary. When India goes to the polls, this is what we celebrate—
change amid diversity.
Despite several doomsday predictions, I believe and always have that our country has many
assets. It is an ancient civilisation, and for reasons for historians to ponder over, it is a beautifully
diverse one. Over centuries, people migrated to India, and settled here. At one time, I am sure, India
had more gods and faiths than people.
India is also a country in transition, and the change here is not just a fall out, but in addition to
the change that the world is witnessing. This is therefore, double change, which sometimes, not
incorrectly, gets translated into double trouble. The transformation across the world is peaking at an
all-time high, which is magnified by the all-embracing effects of climate change. The domestic
transformation is the handing over of the baton from the old elite to the new. From an economy
dominated by an influential socialist elite to an educated, technocratic, middle class elite.
The hegemony of the new elite is a reality that the world has come to acknowledge. However,
the manner in which it has occurred elsewhere is different from the way it has unfolded at home. It
has primarily to do with the pace of change—in most advanced economies, it is glacial and seamless.
In India however, it has been rapid, as is bound to occur in any economy transiting from a very poor
economy (circa 1980) to a lower middle-income economy (circa 2018). A country which shifted from
276 US dollar per capita in 1980 to 2016 dollars per capita in 2018 showed an annual growth of 5.2
per cent per annum for thirty-eight years. For aggregate GDP, the growth has been seven per cent per
annum in dollar terms for thirty-eight years; in (real) rupee terms: eight per cent per annum for thirty-
eight years. My plea, in this and other writings over the years, is that we (especially the old, rarefied
elite) should recognise this change, welcome it, and most importantly, wholly embrace it.
This economic transformation has brought about a political transformation, and this is what is
most troubling to the old, established order. The new elite is not as sophisticated, and definitely not
as professional as the old. As I have elucidated in Chapter One (see p. 1) of this book, the new elite
is made up of amateurs who are just about in the process of assuming the mantles of control—the old
elite has been at it for the better part of sixty-five years. All we are saying, give the new a chance. In
any case, not that they are going to wait for permission.
An intermediate outlier in Indian electoral history was 1977. At that time, most analysts had felt
that the Janata Party was the beginning of an era; a force which had challenged the hegemony of the
Congress party. As it turned out rather swiftly, the Janata Party was a failed experiment. Power
struggles made worse by internecine skirmishes dashed all hopes of annihilating the monolithic
Congress party led by Indira Gandhi.
Four decades and more later, 2019 seems to be the second structural change new era marker in
Indian history; the first being 1996. As discussed in Chapter Five (see p. 64), the defeat of a non
Nehru-Gandhi, P.V. Narasimha Rao, was the change marker for the Congress party. This was the
writing on the wall—with, or without, the dynasty, the Congress days were ending. For the next
eighteen years, India went through a transitory phase, and Dr Manmohan Singh helped the Congress
attain some respectability in 2009. However, even that and more was squandered away with the
worst instance of macro-management in the history of Indian economy over any given election period.
I can only blame bad economics for this debacle, and leave it to my historian friends to decipher how
the Indian National Congress landed itself in such a royal mess despite the much-touted bench-
strength of Sonia Gandhi.
In all probability, and if one can say in a classic case of history repeating itself, Election 2019 is
looking like a repeat of 2014. I have forecast a gain of thirteen seats for the Congress, which is above
the low 2014 reality; for the BJP, it is 274 seats—a single majority on its own, albeit with a loss of
eight from the 2014 level of 282 seats. If I am proven correct, then no can stop Narendra Modi from
consolidating, and ushering in, a new India. And that is the real BIG story of this election.
One final point—economics and politics aside, the old elite thinks that Modi is the cause of all
their problems. Wrong, and wrong at the core. He is the resultof the transformation, not the cause. If
Modi did not exist, India would have had to invent him.
On a personal note, let me say that it is a great time to be seventy years old and to witness the
transformation. India is exciting, troublesome, entertaining, and full of hope for a brighter, more
shared, future. This is India—as the slogan of Air India had said so evocatively in a tourism ad some
forty years ago—‘Come to India. It is not just another country, it is another world.’
Select Bibliography

Reuben Abraham and Pritika Hingorani, ‘India is already a land of cities, not villages’,
Bloomberg.com, 19 January 2019; https://www.bloomberg.com/opinion/articles.
Austin, Granville. Working a Democratic Constitution: A History of the Indian Experience, New
Delhi, Oxford University Press, 2003.
Barro, Robert J. and Jong-Wha Lee, Education Matters: Global Schooling Gains from the 19th to
the 21st century, New York, Oxford University Press, 2015.
Baru, Sanjaya. 1991: How P.V. Narasimha Rao Made History, New Delhi, Aleph Book Company,
2016.
Baru, Sanjaya and Meghnad Desai. The Bombay Plan: Blueprint for Economic Resurgence, New
Delhi, Rupa Publications India, 2018.
Baru, Sanjaya. The Accidental Prime Minister: The Making and Unmaking of Manmohan Singh,
New Delhi, Penguin Random House, 2015.
Bhalla, Surjit S. ‘Freedom and Economic Growth: A Virtuous Cycle?’ In Democracy’s Victory and
Crisis, ed. Axel Hadenius, Cambridge, UK, Cambridge University Press, 1997.
Bhalla, Surjit S., Imagine There’s No Country: Poverty, Inequality and Growth in the Era of
Globalization,Washington, Institute of International Economics, 2002.
Bhalla, Surjit S., Second Among Equals: The Middle Class Kingdoms of India and China, Peterson
Institute of International Economics, Washington; photocopy May 2007. [Available
at]:https://ssbhalla.files.wordpress.com/2015/08/second_among_equals_-_omega_august-25-
2015.pdf
Bhalla, Surjit S., The New Wealth of Nations, New Delhi, Simon & Schuster India, 2017.
Bhalla, Surjit S. and Karan Bhasin, ‘Towards a Targeted Basic Income Policy for India’, 2019.
Available at ssbhalla.org.
Bhalla, Surjit S. and Tirthatanmoy Das, ‘Population, Education and Employment in India: 1983–
2018’, a background report for PM Narendra Modi’s Economic Advisory Council, November 2018.
Available at ssbhalla.org.
Brass, Paul. ‘The body as symbol in the production of Hindu-Muslim violence’, Religion, Violence
and Political Mobilisation in South Asia, 2005, pp.46-68.
Chhibber, Pradeep K., and Rahul Verma, Ideology and Identity: The C hanging Party Systems of
India, New Delhi, Oxford University Press, 2018.
Corbridge, Stuart, Nikhila Kalra, and Kayoko Tatsumi, ‘The search for order: understanding Hindu-
Muslim violence in post-partition India’, Pacific Affairs 85, no. 2, 2012, pp. 287-311.
Desai, Meghnad, The Raisina Model: Indian Democracy at 70, Penguin Random House, 2017.
Dhattiwala, Raheel, and Michael Biggs, ‘The political logic of ethnic violence: The anti-Muslim
pogrom in Gujarat, 2002’, Politics & Society 40, no. 4, 2012, pp. 483-516.
Friedman, Milton, ‘Mahalanobis Plan’, Unpublished Memo, 1955. Available on
http://www.indiapolicy.org/debate/Notes/fried_opinion.html
Hayek, Friedrich A von, The Road to Serfdom,University of Chicago Press, 1945.
Islam, Syed Zafar, ‘Indian Economy: Prospect and Retrospect’, 2019.
Jaffrelot, Christophe, India since 1950: Society, Politics, Economy and Culture, Foundation
Books/Yatra Books, 2012.
Jensenius, Francesca R., Social Justice Through Inclusion: TheConsequences of Electoral Quotas
in India, Oxford University Press, 2017.
Kapur, Devesh, and Milan Vaishnav, eds., Costs of Democracy: Political Finance in India, Oxford
University Press, 2018.
Kumar, Sanjay, Changing Electoral Politics in Delhi: From Caste to Class, Sage Publications India,
2013.
Koninsky, Michael, ‘The Data that turned the world upside down’, 2017. Available at
https://motherboard.vice.com/en_us/article/how-our-likes-helped-trump-win
Mitra, Anirban, and Debraj Ray, ‘Implications of an economic theory of conflict: Hindu-Muslim
violence in India’, Journal of Political Economy 122, no. 4, 2014, pp. 719-765.
Nellis, Gareth, Michael Weaver, and Steven Rosenzweig, ‘Do parties matter for ethnic violence?
Evidence from India’, Quarterly Journal of Political Science 11, no. 4, 2016, pp. 249-277.
Ninan, T. N., Turn of the Tortoise: The Challenge and Promise of India’s Future, Oxford University
Press, 2017.
Rajeshwari, B., Communal Riots in India: A Chronology, 1947–2003, Institute of Peace and Conflict
Studies, 2004.
Sen, Amartya, Development as Freedom, Oxford University Press, 1999.
Sen, Amartya, The Argumentative Indian: Writings on Indian History, Culture and
Identity,Macmillan, 2005.
Sengupta, Hindol, The Man Who Saved India, New Delhi, Penguin Viking, 2018.
Shani, Ornit. How India Became Democratic: Citizenship And The Making Of The Universal
Franchise. Cambridge University Press, 2017.
Shenoy, B.R., Economic Policy Resolution of AICC at Bangalore and Indian Economic and Social
Progress, Economic Research Center, New Delhi, 1969.Sitapati, Vinay, Half-Lion: How P.V.
Narasimha Rao Transformed India, Penguin Random House, 2015. Vaishnav, Milan, When Crime
Pays: Money and Muscle in Indian politics, Yale University Press, 2017.
Varshney, Ashutosh, Battles Half Won: India’s Improbable Democracy, Penguin Random House,
2014.
Varshney, Ashutosh and Joshua R. Gubler, ‘Does the state promote communal violence for electoral
reasons?’ India Review 11, no. 3, 2012, pp. 191-199.
Wilkinson, Steven I, ‘Putting Gujarat in perspective’, Economic and Political Weekly, 2002, pp.
1579–1583.
Wilkinson, Steven I, Votes and Violence: Electoral C ompetition and Ethnic Riots in India,
Cambridge University Press, 2006.
Wilkinson, Steven I, ‘Communal riots in India’, Economic and Political Weekly 40, no. 44/45, 2005,
pp. 4768–4770.