Index

S.no Topic Page No.

1. Objective 1

2. Exits poll 2

3. Purpose 3

4. Organizations that conduct election exit polling 4

5. Challenges of exit polls 5

6. Incorrect prediction 6

7. Sampling 8

8. Probability 9

9. Probability in election 10

10. 11
Correct prediction

11. Tamil Nadu election 2021 13

12. Sums 17

13. Conclusion 19

1
Objective
To discuss the prediction of the outcome of an election based on exit polls

Introduction
An election is a formal group decision-making process by which a population
chooses an individual or multiple individuals to hold public office.
Elections have been the usual mechanism by which modern representative
democracy has operated since the 17th century. Elections may fill offices in
the legislature, sometimes in the executive and judiciary, and for regional and local
government. This process is also used in many other private
and business organisations, from clubs to voluntary associations and corporations.
The universal use of elections as a tool for selecting representatives in modern
representative democracies is in contrast with the practice in the democratic archetype,
ancient Athens, where the Elections were not used were considered
an oligarchic institution and most political offices were filled using sortition, also
known as allotment, by which officeholders were chosen by lot.
Electoral reform describes the process of introducing fair electoral systems where they
are not in place, or improving the fairness or effectiveness of existing
systems. Psephology is the study of results and other statistics relating to elections
(especially with a view to predicting future results). Election is the fact of electing, or
being elected.

2
Exits poll
Exit polls are very similar to any other polls except for the fact that they are
exercised when people exit from the voting booth on election day to learn about the
candidate they chose. These can be online polls or offline polls, both are conducted in
an attempt to predict the official results. Since the exit polls are conducted only after
people cast their votes, they are more reliable than opinion polls.

Since the answers of exit polls are almost confirmed, they are considered to be
genuine sources to estimate the results of the elections. Exit polls can also impact the
opinions of those who are still to vote since they are post-election polls, but in
comparison to other poll methods such as opinion polls, the switch in opinions is not
as drastic in exit polls.

Exit polls have consistently been reliable sources to gather demographic

information of voters. Since voting is always anonymous, it becomes critical for
electoral candidates to know who participated in the voting process – polls such as
exit polls are the sole source to gather demographic details and other information such
as reasons why the voter voted for a particular candidate.

History
There are different views on who invented the exit poll. Marcel van Dam, Dutch
sociologist and former politician, claims to be the inventor, by being the first to
implement one during the Dutch legislative elections on February 15, 1967. Other
sources say Warren Mitofsky, an American pollster, was the first. For CBS News, he
devised an exit poll in the Kentucky gubernatorial election in November that same
year. Not with standing this, the mention of the first exit polls date back to the 1940s
when such a poll was held in Denver, Colorado

3
Purpose
Exit polls are also used to collect demographic data about voters and to find out
why they voted as they did. Since actual votes are cast anonymously, polling is the
only way of collecting this information.
Exit polls have historically and throughout the world been used as a check against,
and rough indicator of, the degree of election fraud. Some examples of this include
the 2004 Venezuelan recall referendum, and the 2004 Ukrainian presidential election.
They are used to command a mandate as well as to determine whether or not a
particular political campaign was successful or not.
The distribution of votes is not even across different polling stations, and also varies at
different times of day. As a result, a single exit poll may give an imperfect picture of
the national vote. Instead, the exit poll is primarily used to
calculate swing and turnout. Pollsters return to the same polling stations at the same
times at each election, and by comparing the results with previous exit polls they can
calculate how the distribution of votes has changed in that constituency. This swing is
then applied to other similar constituencies, allowing an estimate of how national
voting patterns have changed. The polling locations are chosen to cover the entire
gamut of society and where possible, to include especially critical marginal seats. Data
is presented in one of three ways, either as a table, graph or written interpretation.

Problems
Like all opinion polls, exit polls by nature do include a margin of error. A
famous example of exit poll error occurred in the 1992 UK General Election, when
two exit polls predicted a hung parliament. The actual vote revealed that Conservative
Party Government under John Major held their position, though with a significantly
reduced majority. Investigations into this failure identified a number of causes
including differential response rates (the Shy Tory Factor), the use of inadequate
demographic data and poor choice of sampling points.
Because exit polls require a baseline to compare swing against, they are not reliable
for one-off votes such as the Scottish independence referendum or the UK EU
membership referendum. Because exit polls can't reach people who voted by postal
ballot or another form of absentee voting, they may be biased towards certain
demographics and miss swings that only occur among absentee voters. For example,
in the May round of the 2016 Austrian presidential election, exit polls correctly
pointed to a narrow lead for Norbert Hofer among those who voted at a polling
station. However, the postal votes (which made up about 12% of the total vote) were
slightly but definitively in favour of his rival Alexander Van der Bellen, and
ultimately gave Van der Bellen victory. This could be considered a non-U.S. example
of the phenomenon known as "blue shift" in the U.S.

4
Organizations that conduct election exit
polling
In the United States, the National Election Pool (NEP), consisting
of ABC, AP, CBS, CNN, FOX News, and NBC, conducts a joint election exit poll.
Since 2004 this exit poll has been conducted for the NEP by Edison Media Research.
Edison uses probability-based sampling. In 2020, in-person interviews on Election
Day were conducted at a random sample of 115 polling locations nationwide among
7,774 Election Day voters. The results also include 4,919 telephone interviews with
early and absentee voters.
The release of exit poll data in the US is controlled. In the 2012 election protocols to
quarantine the release of data were put in placE.
In Egypt, the Egyptian center for public opinion research (baseera) conducted in 2014
two exit polls; the constitution referendum exit poll and the presidency elections exit
polls.
In South Korea, KBS, SBS, and MBC conducted a joint exit poll to decide the
president of the country.
In India, the exit polls are conducted by private news broadcasting channels and
newspaper agencies such as The Times Group, CVoter, India Today,etc. India being
the largest democracy in the world has many agencies publishing exit polls which
roughly predict the outcome of the elections.

5
Challenges of exit polls
Exit polls face many challenges especially given the changing nature of
how and when people vote. These challenges include:
 Changes in the way people vote:- The share of all voters who cast their votes
before Election Day using by-mail absentee ballots or in-person early voting
has increased from approximately 16% of voters in the 2000 presidential
election to an estimated 42% of voters in the 2016 presidential election.
Telephone surveys of absentee and early voters are used to supplement the
NEP’s national exit polls as well as state exit polls when there is a high
proportion of absentee and early voters.

 Selective refusal:- Slightly less than half of voters asked to participate in an

exit poll agree to complete the questionnaire and those who refuse sometimes
are different from voters who agree to be interviewed. The exit polls attempt to
account for differential non-response by adjusting the data using visibly
observable voter characteristics such as gender, age and race. However, if
Republican voters decline to fill out an exit poll questionnaire at a higher rate
than Democratic voters, the exit polls can produce a slight Democratic
overstatement. As noted previously, to deal with this the exit poll results are
also weighted to reflect the actual election outcomes.

 Early leaks:- In the past, early and incomplete exit poll results were sometimes
leaked on Election Day. Today, major news organizations have agreed not to
release any exit poll data that could characterize the outcome of a race before
all of the polls in that state have closed. The embargo is necessary because an
exit poll sample is not representative of the entire electorate until the survey is
completed at the end of the day. Different types of voters turn out at different
times of the day. Since 2006, in order to prevent early leaks of exit poll data,
the NEP has limited access to exit poll data to a small group of analysts in a
quarantined location until a few hours before the polls close.

 Election regulations:- Over the years, some state election officials have

attempted to impose restrictions on how close exit poll interviewers are able to
stand to the exit of the polling location. News organizations have brought suit
against more than a half-dozen states that have tried to impose distance
restrictions of 100 feet of more. In every case, courts have thrown out these
restrictions. However, there are some states that impose interviewer distance
restrictions of up to 75 feet and these distance restrictions do adversely affect
response rates.

6
Incorrect prediction
 2004 Lok Sabha elections: Riding on the triple victories in Madhya Pradesh,
Chhattisgarh and Rajasthan state elections, Atal Bihari Vajpayee government
dissolved the Parliament early, hoping for re-election. The ruling BJP went for re-
election with a slogan of 'India Shining', but the slogan apparently boomeranged.
The exit polls then had predicted BJP-led NDA to secure over 240 to 250 seats.
But when the actual results came, the numbers were completely opposite and the
Congress ..

Exit polls NDA UPA OTHERS

Aajtak ORG - 248 105 105
MARG
NDTV- AC Nielsen 230 100-120 100
Star News-C-Voter 263 86-98 86
Zee News 249 117 117
Average of Exit 255 105 105
Polls
Actual Final Result 219 137 137

 2014 Lok Sabha Elections: Though most exit polls had predicted a victory for
BJP-led NDA, but none had predicted a majority to the BJP. Majority to exit polls
were predicted BJP's win but just short of the majority mark of 272. However,
when the actual result came the NDA scored a major victory with tally well above
300 and BJP alone crossing the majority mark. Congress conceded a heavy defeat
and was reduced to just 44 seats.

Exit Polls NDA UPA OTHERS

India Today- 261 110 150
Cicero's

CNN-IBN, CSDS 270 102 159

India TV-C Voter 286 100 153
Actual Results 336 60 147

7
 UP Assembly Elections 2017: Just months after demonetisation, ruling Smajwadi
party and Congress had stitched an alliance to beat the Modi juggernaut in the
most-populous state. All exit poll results had predicted a hung assembly in Uttar
Pradesh, with BJP emerging as the largest party. However, all the exit polls proved
wrong and the BJP surprised everyone by winning over 300 seats, which was way
above what exit polls had predicted. The saffron party got 265 seats more in 2017
..

Exit Polls BJP SP-Congress BSP

India Today-Axis 251 38 28
survey
CVOTER 161 141 87
ABP-CSDS 170 168 67
NEWSX-MRC 185 120 90
Today's Chanakya 285 88 27
Actual Result 325 19 47

 Bihar assembly election 2015: Exit polls predicted a photo-finish for the the
bitterly-contested Bihar assembly elections in 2015. Nitish Kumar was fighting for
the re-election in the state on the development poll-plank.
This election saw highest voter turnout in Bihar assembly polls since 2000, with a
56.8% voter turnout. Most exit polls predicted a mixed picture with no clear
majority to any alliance. However, in actual result the RJD-JDU-Congress scrored
a thumping victo ..

Exit Polls BJP+ JDU+ Others

Today's Chanakya 120 117 6
ABP News Nielsen 155 83 -
Times Now-CVote 130 108 -
India Today-Cicero 111 122 -

Final Result 58 178 -

Sampling
8
 sampling basically means selecting people/objects from a population in order to
test the population for something. For example, we might want to find out how people
are going to vote at the next election. Obviously, we can't ask everyone in the country,
so we ask a sample. 
When considering a particular population, it is usually advisable to choose a sample in
such a way that everyone is represented. This is not easy and requires careful thought
about sample size and composition. Often questionnaires are devised to identify the
required information. These need to be idiot proof, so questions need to cover all
alternatives and give little scope for variation.

Example on sampling
A bus company attempted to estimate the number of people who travel on local
buses in a certain town. They telephoned 100 people in the town one evening and
asked 'Have you travelled by bus in the last week?'
Nineteen people said 'Yes'. The bus company concluded that 19% of the town's
population travel on local buses.
Give 3 criticisms of this method of estimation.
In answering this question, there are no 3 correct answers. As long as what you say is
plausible and sensible, you should get the marks. For example, you might say:

 100 people in a large town is not a large enough proportion of the population to
give a good sample.
 People who travel on local buses once a fortnight may have said no to the
question. They nevertheless travel on local buses.
 On the evening that the sample was carried out, anybody travelling by bus
would be out.

Probability
9
 Probability is the branch of mathematics concerning numerical descriptions
of how likely an event is to occur, or how likely it is that a proposition is true. The
probability of an event is a number between 0 and 1, where, roughly speaking, 0
indicates impossibility of the event and 1 indicates certainty. The higher the
probability of an event, the more likely it is that the event will occur. A simple
example is the tossing of a fair (unbiased) coin. Since the coin is fair, the two
outcomes ("heads" and "tails") are both equally probable; the probability of "heads"
equals the probability of "tails"; and since no other outcomes are possible, the
probability of either "heads" or "tails" is 1/2 (which could also be written as 0.5 or
50%).
These concepts have been given an axiomatic mathematical formalization
in probability theory, which is used widely in areas of study such
as statistics, mathematics, science, finance, gambling, artificial intelligence, machine
learning, computer science, game theory, and philosophy to, for example, draw
inferences about the expected frequency of events. Probability theory is also used to
describe the underlying mechanics and regularities of complex systems

Example
Draw a random card from a pack of cards. What is the probability that the card drawn
is a face card?
Solution:
A standard deck has 52 cards.
Total number of outcomes = n(S) = 52
Let E be the event of drawing a face card.
Number of favourable events = n(E) = 4 x 3 = 12 (considered Jack, Queen and King
only)
Probability, P = Number of Favourable Outcomes/Total Number of Outcomes
P(E) = n(E)/n(S)
= 12/52
= 3/13
P(the card drawn is a face card) = 3/13

Probability in election
10
Multicandidate elections with a single winner suggest several questions about the
manner in which the preferences of a group of individual voters are aggregated into
a single social choice. Obvious examples are the national presidential primaries in
the major political parties. However, non-political exercises such as the ranking of
job applicants or college football teams provide other examples. If an individual’s
preference is viewed as a ranking of the available choices then the literature on
probability models for rankings (see the survey by Critchlow, Fligner and Verducci)
may be used to analyse methods for combining preferences. Several probability
models are used to analyse the results of a five-candidate presidential election of the
American Psychological Association. In addition, simulated data generated by
parametric probability models is used to consider the merits of a variety of voting
systems.

Role of probability and statistics

Probability and statistics hold the key for enabling our students to better
understand, process, and interpret the vast amounts of quantitative data that exist all
around them, and to have a probabilistic sense in situations of uncertainty. To be able
to judge the validity of a data-supported argument presented to them, to discern the
believability of a persuasive advertisement that talks about the results of a survey of
all of the users of a particular product, or to be knowledgeable consumers of the data-
intensive government and electoral statistics that are ever-present, students need the
skills that they can learn in a well-conceived probability and statistics curriculum
strand.

11
Correct prediction
 Tamil Nadu 2016:- However, exit polls in Tamil Nadu turned out to be a big fail, as
almost all of them believed that the late former chief minister Jayalalithaa’s
AIADMK would be thrown out thanks to anti-incumbency. They predicted a win for
the DMK-Congress alliance.
The Axis-My India exit poll saw the DMK-Congress alliance getting between 120
and 140 seats of the 234-member Tamil Nadu Assembly. It saw AIADMK getting 90
to 110 seats, IANS reported at the time. The News Nation TV exit poll gave 114-118
seats to the DMK-Congress alliance and 95-99 seats to the AIADMK.
Instead, come counting day, the AIADMK returned to power with a whopping 136
seats.

12
 Bihar 2015:- The results of the October-November 2015 Assembly election here
were hugely awaited. That’s because it was Bihar and because of the ‘grand alliance’
of Nitish Kumar’s JD(U), Lalu Yadav’s RJD and the Congress, which was to take on
the BJP for a majority in the 243-seat Assembly.
The ABP-Nielsen poll said the ‘grand alliance’ would get 130 seats and the BJP and
allies would get 108. Times Now along with C-Voter gave 122 seats to Nitish’s
‘grand alliance’ and 111 to the BJP and allies.
Come counting day the result were not nearly as close as the exit polls predicted.
Nitish’s grand alliance won a landslide victory getting 178 seats.

Tamil Nadu election 2021

13
 According to all the exit poll predictions opposition DMK led by MK Stalin is
set to win big in the 234 assembly seats in Tamil Nadu. The DMK-led opposition
alliance respectively was projected ahead of their respective rivals by all the exit polls.

 The DMK led alliance in the state is projected to win 16-0170 seats out of the
234 seats in Tamil Nadu

 Republic-CNX Exit Poll predicts that DMK + is all set to return to power in
Tamil Nadu

 The break- up projected by the Republic-CNX Exit Poll is:

DMK+ allies->160-170
AIADMK+ allies->58-68
MNM+ Allies-->00-02
AMMK+ Allies--> 04-06

MK is all set to emerge as the single largest party in the state by bagging137-147 seats
and AIADMK in the second position by 49-59 seats
According to ABP-Cvoter Exit Poll Results 2021 predicted that opposition DMK is
set to return in Tamil Nadu.

 The break- up projected by the ABP-Cvoter Exit Poll is:

DMK+ allies--> 160-172 seats
AIADMK+ allies->58-70
MNM+ Allies-->00
Others-->00-07

 The break- up projected by the IndiaToday-Axis My India is:

DMK+ allies-->175-195
14
AIADMK+ allies-->38-54
MNM+ Allies-->00-02
Others-->01-07
 The break- up projected by the Today's Chanakya is:
DMK+ allies-->164-186
AIADMK+ allies-->46-68
MNM+ Allies-->
Other-->0-

15
 AIADMK

TMCV P-MARQ R-TV CNX Chanakya AMI Poll of poll

DMK

TMCV P-MARQ R-TV CNX Chanakya AMI Poll of poll

AAMK

TMCV P-MARQ R-TV CNX Chanakya AMI Poll of poll

16
17
SUMS
1) These are the observations of various people participating in exit polls. Calculate
mean.

CI fi xi fixi
0-10 45 5 225
10-20 56 15 840
20-30 87 25 2175
30-40 24 35 840
TOTAL 212 4080

MEAN = ∑fixi /∑fi = 4080/212 = 19.24.

2) Create a bar graph based on TN exit polls.

TN ELECTION 2021 OPINION POLL RESULTS

200

180 177

158
160 151

140
SERIES 1
120 SERIES 2
100 SERIES 3

80 76 74

60 49
40

20
0 2 3
0
DMK ADMK AMMK

18
3) In page17 of this document , the graph represents of TN election poll results find
the average votes for all the parties.

DMK = ( 160 + 165 + 164 ) /3

= 163
ADMK = ( 56 + 58 + 46 ) /3
= 53.33
AMMK = ( 4 + 4 + 0 ) /3
= 2.66

4) Taking the US elections on account probability of Joe Biden (A) and probability of
Trump (B) winning are given. Calculate P(B/A). Here the probability of A is true!

PROBABILITY MEN WOMEN

P(A) 45% 57%

P(B) 53% 42%

AVERAGE P(A) = 45+57/2 = 51%

AVERAGE P(B) = 53+42/2 = 47.5%
Using Bayes’ theorem,
P(A/B) = P(B/A) * P(A) / P(B)
= (51/47.5) * 51 / 47.5
= 1.073 * 51 / 47.5
= 54.723 / 47.5
= 1.520.

⸫ probability of A given B is true for 1.520

19
Conclusion
An election exit poll is a survey based on interviews with voters as they
leave (or exit) their balloting locations. To estimate the outcome of an election in a
particular constituency, a sample of its smallest voting units (often referred to as
precincts or polling locations) is drawn and at least one interviewer is assigned to each
sampled location on Election Day. On a pre-determined and systematic basis, the
interviewer approaches people who have already voted in order to obtain an interview.
For example, depending upon the expected number of voters at the sample location,
the interviewer may approach every third or fifth person who leaves the voting
location. The interviewer usually hands the voter a questionnaire on a clipboard and
asks him or her to fill out the questionnaire, fold it up, and deposit it in a survey ballot
box. In most instances the exit poll questionnaire is self-administered to protect the
privacy of the voter and the confidentiality of each individual’s responses.

Bibliography
https://www.hindustantimes.com/lok-sabha-elections/history-hits-and-misses-
controversies-all-about-exit-polls/
https://www.financialexpress.com/india-news/exit-poll-2021-from-tamil-nadu-to-
uttar-pradesh-four-times-when-exit-polls-proved-wrong/2243045/
https://revisionmaths.com/gcse-maths-revision/statistics-handling-data/sampling
https://economictimes.indiatimes.com/topic/exit-poll-predictions
https://www.aapor.org/Education-Resources/Election-Polling-Resources/Explaining-
Exit-Polls.aspx
https://www.newindianexpress.com/nation/2019/may/22/when-exit-polls-went-the-
wrong-way-1979935.html

20