Home | Biodata | Biography | Photo Gallery | Publications | Tributes [Back to Surveys List] Political - Socio - Economic-Surveys

In the Parliamentary elections held in Tamil Nadu in January 1980 the Congress (I)-DMK alliance had a massive victory over the Janata-AIADMK alliance and this result was what one would have expected. There were only three main parties, viz., the All India Anna Dravida Munnetra Kazhagam with a popular support of about one-third and the Dravida Munnetra Kazhagam with a support of about one-fifth of the voters. The day the DMK joined with the Congress (I), it was clear that the DMK-Congress (I) would win the elections.

This was not so obvious to the average voter and people turned to the Astrologers and Wise Men. The newspapers published the different predictions made by astrologers of different schools. Astrological predictions are usually based on the time of birth (lagna), the dates of birth (nakshatra) of the political leaders and the time of elections. Each astrologer follows a suitable astrological text and predicts the victory of his favourite leader. Such predictions are based on a magical model of the world and its affairs. Since the planetary system has a certain regularity of movement it is possible to predict with accuracy the position of the different planets and the direction in which they will be found with reference to the fixed stars. The ability to predict the position of the planets is used to predict the outcome of elections in the Indian continent. The fact that this magical view of the world is quite strong in India was evident during the grand phenomenon of total eclipse of the sun on February 16, 1980 when people took shelter in their dwellings and all transport came to a halt. The central and state governments including the Government of West Bengal declared a public holiday.

With the astrologers getting busy on the one hand, the people also turned to their Wise Men. Newspapers were full of opinions of different people who made sweeping statements about the outcome of the elections. Wise Men based their knowledge on the newspaper reports as well as their traditional models of society. To them the country was made up of different caste groups. They assumed that all people of one caste would vote for only one particular candidate in a given constituency. They went about finding out the different proportions of castes in a given constituency even though there is no official data available. Since most of the Wise Men belonged to the higher castes this gave them an opportunity to affirm their superiority. They would make statements about the Harijans and other poorer sections making people more and more aware of caste-distinctions.

Instead of going to the Astrologers and Wise Men one could go directly to the people and conduct an opinion poll¹ and that is what we did during the second half of December 1979.

There are two problems associated with public opinion surveys. The first is to have the technical know-how to conduct a survey and the second is to be unprejudiced in reporting the results of the survey. Many Wise Men believe that we do not have, in India, the necessary technical know-how and that the agencies that conduct the surveys are dishonest !

Sampling Procedure

To gauge the relative popularity of political parties and leaders it is not necessary to use the method of complete enumeration. A well-designed sampling procedure^{2, 3} would give the results to any desired degree of accuracy. The larger the sample, the smaller the sampling error provided the sample is a well-chosen sample. The term, sampling error is used in a technical sense and its value can be calculated using appropriate formulas. A sample of about 2000 respondents would lead to a sampling error of about one per cent when the estimated proportion is about thirty-five per cent. If the true proportion of supporters of a certain party is about thirty-five per cent then a random sample of 2000 respondents would yield a proportion varying between 34 to 36 per cent in a majority of instances.

The problem of estimating the true proportion of supporters of a certain political party in a large population of voters can also be viewed  from a slightly different point of view and values called confidence intervals  calculated but we shall not follow this procedure here. The crucial problem is to take a sample that will truly represent the population of all voters. If voters are chosen in a haphazard manner then the results will have only anecdotal value and nothing more.

The principle behind the proper method of sampling is that every unit in the population must have the same chance of being chosen in the sample.

We distributed all the units of the sample among all the taluks of Tamil Nadu in proportion to the population of the taluk. If a taluk had less population compared to a second taluk, a smaller number of units was allotted to the first sample compared to the second. In each taluk the number of sampling units was allocated to the rural and the urban strata in proportion to their sizes. In other words, if half the population of a certain district was classified as urban then the number of respondents in the sample allocated to the taluk was divided between rural and the urban sectors equally. If the number of respondents to be interviewed were to be forty in a taluk and half of the taluk were urban then twenty respondents were chosen from the urban areas and another twenty from the rural areas of the taluk.

Once the number of respondents for each taluk was fixed then villages/towns were selected using random number tables such that the larger villages had a greater chance of being chosen in the sample compared to smaller villages. For the sake of convenience we chose five or six respondents from each village selected for the purpose. At the village level the selection of respondents could be made by drawing lots from the electoral rolls. Instead of drawing lots we use in practice random number tables for drawing the sample. Thus we ensure that each respondent chosen in our sample is a true representative of the people.

In addition to the so called sampling error that we have discussed, there are the non-sampling errors. The investigator may be dishonest, the respondent may not speak the truth and the responses may not be entered properly and so on. Many checks and cross-checks are introduced to reduce the non-sampling errors in our study.

We conducted our first state-wide survey in 1971 when contrary to general expectations we found that the DMK party was quite popular in the State. We also conducted studies in January 1976, January 1978,⁴ January 1979⁵ and now in December 1979. Many of the results were stable showing the power of the method of scientific sampling.

Data Processing

Data were collected in printed questionnaires and the answers were coded into numerical form by the student investigators. These were carefully scrutinised by the data-processing team and punched on to computer cards. The punched data were once again checked and fed into a computer along with the computer programme written by members of the department of statistics. The results were obtained in the form of tables in computer sheets.

The Survey

More than a hundred student volunteers interviewed about 1800 respondents selected from 400 villages/towns of Tamil Nadu. Interviews were conducted from December 15 to December 24 and a vast majority of the respondents co-operated with the student volunteers. A group of students and staff worked round the clock and computer printouts were obtained on the morning of December 27, 1979 and preliminary results were made available to the press the same day.

Relative Popularity of Political Leaders

The multipurpose survey had one main question on political leaders. Each respondent was asked to indicate against the name of each leader whether the respondent supported the leader strongly, was against him/her, have a good opinion of him/her, never heard of the leader or had no opinion to offer. The respondent was allowed to strongly support only one leader. If the respondent insisted on strongly supporting more than one leader he was reckoned to have a good opinion of both, and not to have supported any one leader strongly.

Mrs Indira Gandhi had the strong support of thirty-five per cent of the people (Table 1) in December 1979. It will be of interest to compare this figure with the ones obtained from surveys conducted in the earlier years. In January 1978⁴ twenty percent  of the people strongly supported her (Table 2) and the figure rose to twenty-seven per cent in January 1979.⁵ A linear trend fits fairly well to the data showing that her popularity had been rising at a steady rate of seven to eight per cent per year after the Janata Party came to power.

In addition to the thirty-five per cent who strongly supported her in December 1979, forty per cent had a good opinion of her and only sixteen per cent were opposed to her.

The then Chief Minister, Mr. M. G. Ramachandran was strongly supported by thirty-three per cent of the people interviewed. This figure had remained more or less stable over the two previous years. In addition about twenty per cent had a good opinion of him but almost forty per cent were opposed to him. Even though the level of strong support had remained stationary, the level of opposition had gone up over two years since in January 1978 only about one-fifth of the people were opposed to him. Mr. M.  Karunanidhi the leader of the Dravida Munnetra Kazhagam had strong support of about twenty per cent of the people. During the surveys held earlier he had the support of about seventeen per cent of the people. In Table 2 it is seen that in December 1979 more than ninety per cent of the voters strongly supported one leader or the other but in the survey of January 1978 the total percentage of committed voters was only about eighty or so. Either people were more clearly committed to some leader or other in December 1979 compared to the earlier surveys or the investigators were more successful in the recent survey in assessing the true opinion of the respondents.

More than forty per cent opposed M. Karunanidhi but almost thirty per cent of the people interviewed had a good opinion of him. In January 1978, forty per cent were opposed to him and only about twenty per cent had a good opinion of him. In other words the proportion of voters committed to Mr. Karunanidhi had increased only marginally and opposition to him had remained more or less at the same level. The additional good will that he had received in the form of good opinion could be from the strong supporters of Mrs. Indira Gandhi.

Only four per cent of the people interviewed were strong supporters of Mr. Morarji Desai and from Table 2 it is seen that his popularity had been going down steadily over the years in Tamil Nadu. Ten per cent of the people had not even heard of him.

There were hardly any support for Mr. Charan Singh and eighteen per cent of the people interviewed had not even heard of him.

When the political parties form alliances, it is difficult to get a clear picture. A rough picture will emerge when we look at the problem in terms of party support after the announcement of candidates.

Relative Popularity of Political Parties

In Tamil Nadu the Congress (I) Party led an alliance consisting of itself, the DMK and the Indian Union Muslim League. The AIADMK formed an alliance with the Janata, the Communist Party of India and the Communist Party of India (Marxist). Many other parties lent their support to the different fronts even though they had not set up any candidates. At the time of the survey the party candidates for each parliamentary constituency had been announced. A respondent from a given constituency may strongly support a particular party but the party may not field its own candidate. To get a clear understanding of patterns of political support it is more useful to study the support to party leaders than to the party itself. Our survey of January 1978 revealed that between 90 and 95 per cent of strong supporters of a political leader also strongly supported the party. During the time of elections one would expect the percentage to be even higher. We shall be primarily concerned in this study with the support of the leaders than with the political parties. However we shall also show that there was a fairly clear picture about the level of popular support each party could expect in the elections.

The total number of parties that formed the alliances for the parliamentary elections in Tamil Nadu is 18 but in our questionnaires we made provision mainly for four main parties, viz., the Congress (I), the Janata, the AIADMK and the DMK. There was some provision for other parties belonging to the different alliances but they were not given separate identity while processing the data. Both the communist parties together had only about three per cent of committed voters in our earlier surveys and since sample was not large enough to estimate such small proportions they were not listed as separate parties in the questionnaire.

Our pre-election study of the Coimbatore constituency showed that the CPI candidate was no match for the DMK candidate.⁶

To get a broad picture of party position, we also approached the problem from another angle. Taking the Congress (I) party, 512 respondents said that they would definitely vote for its candidate and in contrast to this only 284 said they would definitely not vote for its candidate (Table 3). This indicated a massive victory for the Congress (I). On the other hand, 124 respondents said that they would vote for the Janata Party compared to 208 who said they would definitely not vote for it. This showed that Janata candidates did not have much support in Tamil Nadu. Taking the AIADMK party, 449 respondents said that they would definitely vote for it and in contrast to this 579 respondents said that they would definitely not vote for it. In addition 46 voters said that they were undecided. This indicated that the AIADMK did not have any chance of winning the elections. For the DMK 327 respondents said that they would definitely vote for it, 345 said that they would definitely not vote for it and 65 said that they might or might not vote for it. Unlike in the case of the other three parties the position of DMK was not so clear since the difference between the strong supporters and the opposers was not very much. If the election propaganda gained momentum and the alliance partners faithfully voted for the DMK it could still win seats.

A detailed analysis of relative support of the parties is not attempted here since many voters had not heard the names of all party candidates at the time of our survey.

Another analysis was made on the basis of regional support for the parties and in almost every region the Congress candidates had a clear advantage over the rivals. It was quite easy to conclude that Congress (I)-led-alliance was poised for a massive landslide victory.

The survey revealed the relative popularity of different political leaders, and the potential support for the different parties.

Now we shall take up the question of who were the people who supported the leaders.

Occupation of Respondents

In the state-wide surveys we have not collected any data on caste or income of the respondents. However occupation of the respondent was collected and it is an indicator of socio-economic status.

Our sample was divided into seven groups in terms of the main occupation of the respondent. They are the landless labourer, the agriculturist (including the tenant cultivator), the professional (including school teachers), the trader, the self-employed craftsman, the office worker and the rest (called any other}. The last category includes the housewife if she had no other occupation. We have taken into consideration only the committed supporters who strongly support a particular political leader.

In Table 4 we find that only 28% of the landless labourers surveyed supported Mrs. Indira Gandhi and this figure is lower than the overall support of 35% she enjoys. Even among the self-employed craftsman only 31% strongly supported her. In contrast to this she had more than average support from the office workers and the professionals. There is no doubt that Mrs. Gandhi had the support of all categories of people but some supported her more than the others.

Mr. Morarji Desai had an overall support of 4%. His support among the landless labourers was half this percentage. Among the traders and the self-employed craftsmen he had more than average support.

Mr. M. G. Ramachandran had the strong support of one-third of the people interviewed. Even though he had support among all classes of people his popularity was greatest among the poorest of the poorthe landless labourers. Forty-six per cent of them were his committed supporters. No other leader in Tamil Nadu got so much support from the landless labourers. He got much less support (about 20%) from the professionals and the office workers. Even though the agriculturists were organised to demonstrate against MGR's government, he still enjoyed the support of 33 % of them. One-third of the traders as well as self-employed craftsmen strongly supported him.

The DMK leader Mr. Karunanidhi enjoyed the strong support of about 20% of the people. His influence cut across all classes and he had a little more than average support from the office workers, the agriculturists and the professionals.

All the four political leaders get support from all classes of people but some sections gave greater support than others to their favourite leader.

Place of Residence of Respondents

We shall take up certain indicators of social change such as urbanisation, education and younger age. We shall first deal with the place of residence of committed voters. The values are presented in Table 5.

The places of residence are divided into five convenient categories, viz., the hamlet with a population of less than 2000, the village with a population not exceeding 10,000, the small town with a population between 10,000 and one lakh, the large town with a population exceeding one lakh and finally the city of Madras.

The residents of hamlets are the least informed compared to others. For instance on the average ten per cent of the people interviewed had not heard of Mr. Morarji Desai. Among the residents of hamlets 16% had not heard of Mr. Desai and among the residents of small towns 10% had not heard of him. Twenty-six per cent of the residents of hamlets had not heard of Mr. Charan Singh compared to the overall average of eighteen per cent of the State.

An interesting pattern about the percentage of committed voters in cities reveals that the city averages are fairly close to the overall state averages for the different leaders (Table 5). The results of our earlier survey in January 1978 did not have such a pattern.

The committed voters of Mrs. Indira Gandhi were spread through all the five categories of places of residence. Compared to the overall value of 35% she received more support from the residents of hamlets (42%) and large towns (41%). The level of support from small towns as well as Madras City are about the same as the overall support that she received.

Mr. Morarji Desai received more than average support from small towns and large towns.

Mr. M. G. Ramachandran's committed supporters were fairly evenly spread through all the five categories of places of residences. His greatest support of 37% came from the residents of villages with a population lying between 2,000 and 10,000. The least committed support of about 29% came from the residents of hamlets, small towns and large towns. The residents of Madras city were very close to the average value of 33%.

The committed support to Mr. Karunanidhi was fairly evenly spread through all the five categories of places of residence. The lowest percentage of committed supporters were from large towns with a population exceeding one lakh.

The influence of political leaders was spread over all categories of places of residence.

Education of Respondents

The level of education is divided into four groups, viz., the illiterates, those who had had an education upto the VIII standard, those who had had a high school education (IX-XI standards) and those who had had a college education. The level of illiteracy as published in official statistics is often misleading since the official figures are calculated taking the total population of the state including new-born babes. The earlier practice was to omit those below five years of age. In an earlier study⁷ we found that sixty-nine per cent of men and forty-three per cent of women could write a pre-determined passage in Tamil, working to an average of fifty-six per cent of literacy for adults. However, in the present study we have classified a person as having gone to school even though the person may not be functionally literate. To sort out those who are functionally literate we had another question on reading of newspapers and we shall discuss it later.

Mrs. Indira Gandhi's committed supporters are fairly evenly spread among all the four classes (Table 6). However among those who were interviewed a pattern was discernible, viz., the higher the level of education the greater the support for Mrs. Indira Gandhi. Compared to the average rate of 35.4%, among the illiterates only 32% strongly supported her. Among the college-educated, however, about 38% strongly supported her. As far as Tamil Nadu is concerned, the educated people certainly gave their support to Mrs. Indira Gandhi more than to any other leader.

Mr. Morarji Desai got only about 2% support from the illiterates compared to his overall value of 4.18%. As far as Mr. M. G. Ramachandran is concerned there is a clear pattern that has emerged out of this classification. Among the illiterates 41% are his committed supporters. No other leader has such a strong support from the illiterate group as he does. As the level of education of the voter goes up the proportion of his committed supporters steadily goes down. Among the college educated only 23% strongly support him. We have already pointed out that he has very good support from the landless labourers and it is clear that he is the leader of the poor and downtrodden sections of the people of Tamil Nadu.

The pattern of support for Mr. Karunanidhi is just the opposite. Even though he has his supporters among all categories of people, the higher the level of education of the voter the greater is the support for Mr. Karunanidhi.

Functional Literacy and Reading Habits

Very often those who had attended school for a few years slip back to illiteracy. Furthermore many claim to be literates when they can only sign their name. In our study a person is classified as illiterate if he cannot read. Even among those who are illiterates there are many who maintain a fairly high level of political awareness by regularly participating in newspaper-read-aloud sessions. It is fairly common in many villages where a newspaper is read aloud to a group of persons by one of the local men. In this section we divide the people on the basis of their reading and listening habits. The adult voters are classified into five main categories. The first group consists of persons who read newspapers/magazines in English only. The second group consists of those who read newspapers and magazines in English and Tamil and other languages. This group contains all those who have a good working knowledge of English. The third group consists of those who read papers and magazines in Tamil but not in English. The fourth group consists of those who normally listen to newspapers being read to them and they will be more aware politically than the next group which consists of voters who neither have the newspaper read to them nor read it.

Almost half of the people in our sample belonged to the group which read newspapers and magazines in Tamil only. Less than one per cent of the sample belongs to the group which reads only in English, and for all practical purposes the percentages given for this group in Table 7 may be ignored. Among the illiterates the ratio of those who listen to paper-reading to those who do not, is approximately in the ratio of 3:4.

In our sample of about 1800 respondents only 13 said that they read only English papers and magazines. Out of those 13 respondents 7 strongly supported Mrs. Indira Gandhi. Among these who read English papers along with language papers 41% strongly supported Mrs. Indira Gandhi. In that group no other leader gets so much support like Mrs. Indira Gandhi. This once again confirms the fact that she got more support from the educated than the uneducated. The educated gave her support in spite of the fact that many English dailies were against Mrs. Indira Gandhi's return to power. She received the least support from the group which neither read nor listened to newspapers.

Mr. Morarji Desai also got a little more support from the educated than from the uneducated people.

The pattern of support Mr. M. G. Ramachandran received was quite the opposite. He got tremendous support (43%) from the group which neither read nor listened to papers. He got least support (24%) from the group which read English as well as the language papers.

Mr. Karunanidhi, on the other hand received most support among those who read Tamil papers only and least support from the group of people who neither read papers nor listened in the paper reading sessions.

Age of Respondents

Respondents are classified into five categories on the basis of their age. The youngest group is made up of people who are below 30 years of age and they form a little more than 30% of the sample. The next age group (31-40) forms about one-third of the sample. The next group (41-50) forms about one-fifth of the sample. Those above 50 years of age form about 15 per cent of the sample.

In general the committed supporters of the political leaders are spread over all the age groups and the results are presented in Table 8.

Compared to the average value of 35%, Mrs. Gandhi gets more support (41%) from the 51-60 group. They probably represent those who once belonged to Congress (0) who would remember the struggle for Independence led by the Indian National Congress. She gets the lowest support of 32% from the youngest group (21-30) made up of citizens born after the achievement of Independence. By and large she gets more support from the middle-aged and old groups than from the younger groups.

Mr. M. G. Ramachandran on the other hand is the leader of the youngest group; about 40% of them strongly support him. As the age-level goes up, the level of support goes down for MGR among the people of Tamil Nadu. He has more followers among the young and dynamic sections than any other leader.

Mr. Karunanidhi's support is fairly evenly spread through all the age-groups. He gets a little more committed support from the 31-50 group than in other younger or older groups. In interpreting the data one must bear in mind that the percentages are based on a sample of about 1800 respondents and that they are not population figures. Much significance cannot be attached to small differences such as those found between the overall percentage and the group percentages.

Sex

In a sample of 1794 respondents men formed 65% and women 35% and the relative support the leaders received from these two groups are presented in Table 9. It is fairly common to get unequal representation of men and women in sample surveys conducted in India. One can make the necessary corrections in the final estimates of parameters by giving them equal weights. Alternatively the investigators may be given strict instructions to get equal number of men and women but this would also introduce some bias as it would ignore uncooperative women respondents. Most of our student investigators are young men and they find it difficult to make women respondents answer their questions. Even when women investigators go many female respondents refuse to answer questions. Very often we plan multipurpose surveys including questions on public health and child welfare so that we may get better cooperation from the respondents. In an earlier survey we found that out of all women respondents only 40 % said that they would vote to their favourite party independent of their husbands. This shows that in voting studies, women's opinion cannot be given equal weight with men's opinion. We treat the proportion of men and women found in the sample as a kind of natural weightage for estimating the percentage of strong supporters of a leader. Mrs. Indira Gandhi's support is significantly higher among women (39%) than among men (33%). Our earlier surveys did not show so much difference in the level of support. The data indicate a clear swing among women in favour of Mrs. Indira Gandhi. Among women she is the most popular leader and is decidedly more popular than Mr. M. G. Ramachandran.

In our sample Mr. M. G. Ramachandran got a little more support among women than men but such a small difference cannot be treated to be of much significance. Even in our earlier study of January 1978, the percentage of women supporters was 34.3 which is the same as the present figure (34.65%). In other words MGR had maintained the same level of popularity among women voters.

Mr. Karunanidhi gets less support from women (14%) compared to men (23%) and this pattern is similar to the one that we obtained in January 1978. The main reason why there are fewer supporters for him among women is that he is blamed for having suspended prohibition in Tamil Nadu during his period of office.

Prohibition

Prohibition was one of the few issues in which there was difference between the AIADMK and the DMK parties. Mr. M. G. Ramachandran remained the champion of the cause of prohibition but after the defeat of his party in the parliamentary elections of January 1980 he had second thoughts about his prohibition policy.

Respondents were asked to agree, or disagree with the statement, 'Prohibition must continue in Tamil Nadu'. Eighty-two per cent agreed with the statement and only 17% disagreed with the statement. This proportion of about four-fifths of the people supporting the policy has been a stable feature in our earlier surveys also.

In 1979, Mr. M. G. Ramachandran withdrew some of his earlier orders on the strict enforcement of prohibition and the curbing of drunkenness. Respondents in our sample were asked to agree or disagree with the statement, 'I welcome relaxation of enforcement of prohibition'. Only 24% agreed with the statement and 72% disagreed with the statement showing that there was no popular support for the policy of relaxation.

Two-thirds of the respondents felt that the level of drunkenness among the people had increased and 73% had seen people in a drunken state during the previous week. In spite of the fact that prohibition was not well-enforced a vast majority of the people supported the prohibition policy.

Summary of Findings

The strong support of Mr. M. G. Ramachandran (33%) and Mr. M. Karunanidhi (20%) had remained more or less stable over a period of two years. While Mr. Morarji Desai went down steadily to 4% Mrs. Indira Gandhi's support grew steadily from 20% in January 1978 to 35% in December 1979. The combination of Mrs. Gandhi's and Mr. Karunanidhi's influence led to the massive victory of the Congress (I)-DMK victory in the Parliamentary elections of Tamil Nadu in January 1980.⁸ Mr. Charan Singh had no support in Tamil Nadu and 18% of the respondents had not even heard of him.

Party loyalties of respondents cut through all the different sections of the population in terms of occupation. A high proportion (46%) of landless labourersthe poorest of the poorstrongly supported Mr. M. G. Ramachandran. Mrs. Gandhi, on the other hand had greater support among the professionals and the office workers than the landless labourers or self-employed craftsmen.

Mrs. Gandhi enjoyed greater support among the educated than the uneducated. In contrast to this, Mr. M. G. Ramachandran had greater support among the illiterates than among the educated.

Mr. M. G. Ramachandran had the maximum support (40%) from the youngest group of voters whereas Mrs. Gandhi got the lowest support from the youngest group.

A significant proportion of women (39%) strongly supported Mrs. Gandhi and this level is much higher than the level of support received from men (33%). Mr. M . Karunanidhi got less support from women compared to men. Mr. M. G. Ramachandran received a little more support from women (35%) than men (32%) but much importance cannot be attached to this marginal difference.

Prohibition is one of the main issues in Tamil Nadu politics and a vast majority (82%) of the people support the continuance of prohibition. A good majority (72%) of them were against any relaxation in enforcing prohibition even though the prohibition policy had failed to arrest the incidence of drunkenness among people.

Acknowledgement

More than a hundred student-volunteers of the Statistics Department of the Madras Christian College took part in the survey and were responsible for the coding and punching of data. All the members of staff of the Statistics Department gave their time and support. Miss S. Bala and Miss S. Lalitha organized the students and gave them special training in the collection of data. Mr. R. Chandrasekaran wrote the computer programme in PL/I language. The author wishes to thank them all.

TABLE 1
Relative Popularity of Political Leaders (Percentages) December 1979

Political Leaders	Not heard of	Strongly support	Strongly oppose	Have good opinion of	No opinion	Total
Indira Gandhi	0.06	35.45	16.44	41.03	7.02	100.00
M.G.Ramachandran	0.45	33.00	38.91	19.84	7.80	100.00
M.Karunanidhi	0.17	19.62	41.25	28.87	10.09	100.00
Morarji Desai	9.92	4.18	26.92	29.77	29.21	100.00
Charan Singh	18.28	0.45	31.05	11.32	38.91	100.00

TABLE 2
Trend of relative strength of committed supporters
Political Leaders over a two year period (Percentages)

Month and Year	Indira Gandhi	Morarji Desai	M.G.Ramachandran	M.Karunanidhi	Total of committed voters
January    1978	19.75	12.95	31.17	16.59	80.46
January    1979	26.97	07.87	31.28	16.61	82.73
December 1979	35.45	04.18	33.00	19.62	92.25

TABLE 3
Support for candidates of Political Parties

Party	Definitely vote for	May or may not vote for	Definitely not vote for	No opinion
Congress (I)	512	59	284	112
Janata	129	29	208	48
All-India Anna Dravida Munnetra Kazhagam	449	46	579	81
Dravida Munnetra Kazhagam	327	65	345	60

TABLE 4
Occupation of committed supporters of political leaders (Percentages)

Occupation	Indira Gandhi	Morarji Desai	M. G. Ramachandran	M. Karunanidhi
Landless labourer	28.13	1.95	46.48	17.58
Agriculturist (also tenant)	35.51	3.27	32.86	23.27
Professional	40.22	2.17	19.57	23.91
Trader	33.33	12.50	33.33	18.33
Self-employed craftsman	30.92	7.24	32.24	21.05
Office worker	41.60	4.00	20.80	25.60
Any other (also housewife)	38.28	3.76	32.02	15.21
Overall percentage	35.45	4.18	33.00	19.62

Table 5
Place of residence of committed supporters of Political Leaders (Percentages)

Place of Residence	Indira Gandhi	Morarji Desai	M. G. Ramachandran	M. Karunanidhi
Hamlet (less than 2000)	41.82	3.27	29.21	20.33
Village (2000 - 10,000)	31.52	3.96	37.08	19.90
Small Town (10,000 - 1 lakh)	35.89	5.21	29.32	18.63
Large Town (more than 1 lakh)	40.58	7.25	28.99	14.49
City (Madras)	34.96	4.07	32.52	21.14
Overall support	35.45	4.18	33.00	19.62

 Table 6
 Level of education of committed supporters of Political Leaders (Percentages)

Level of education	Indira Gandhi	Morarji Desai	M. G. Ramachandran	M. Karunanidhi
Illiterate	32.06	2.29	41.41	16.98
Up to VIII Std.	36.83	4.85	30.67	20.71
IX to XI Std.	36.61	5.95	29.17	20.54
College	37.66	3.90	23.38	22.08
Overall percentage	35.45	4.18	33.00	19.62

Table 7
Reading habits of committed supporters of Political Leaders (Percentages)

Reading of Newspapers / Magazines	Indira Gandhi	Morarji Desai	M.G.Ramachandran	M.Karunanidhi
In English only	53.85	15.38	30.77	00.00
In English and Tamil (and other languages)	40.74	5.35	24.69	18.52
In Tamil only	35.62	5.15	29.55	22.68
Listen to paper read aloud (Tamil)	34.38	2.73	37.89	17.97
Neither listen to nor read papers	31.32	2.30	42.53	14.94
Overall percentage	35.45	4.18	33.00	19.62

Table 8
Age-wise distribution of committed supporters of Political Leaders (Percentages)

Table 9
Sex-wise distribution of committed supporters of Political Leaders (Percentages)

Sex	Indira Gandhi	Morarji Desai	M. G. Ramachandran	M. Karunanidhi
Men	33.30	4.99	32.10	22.89
Women	39.40	2.69	34.65	13.61
Overall percentage	35.45	4.18	33.00	19.62

References:

1 "There is no substitute for polls. No inside information, no gut-feeling, no inspired journalism, no other soothsaying exists to give a better answer than that of the polls", D. Bubler, in "Discussion on public opinion polls", Journal of the Royal Statistics Society, Series A (1979), 142,454.
2 Fredderick E. Croxton and Dudley J. Cowden Applied General Statistics, Prentice-Hall, New Delhi, 1969.
3 Gift Siromoney and Rani Siromoney, Mathematics for the Social Sciences, National Book Trust, India, New Delhi, 1976.
4 S.Bala, S.Govindaraju, and Gift Siromoney, ' Surveys',  Madras Christian College Magazine (1978), 47, 21-26.
5 S.Bala, M. Bagavandas, and Gift Siromoney, ' Public opinion on Mrs. Gandhi's imprisonment, political leaders and parties' Madras Christian College Magazine (1979), 48,19-21.
6 The DMK candidate Mr. Ram Mohan defeated the CPI candidate Mrs. Parvati Krishnan by about 50,000 votes in the parliamentary election.
7 S.Bala, S.Govindaraju, and Gift Siromoney, 'Surveys', Madras Christian College Magazine (1978), 47, 21-26.
8 The percentage of popular votes received in the election by the two major fronts were very close to the figures obtained in the survey for the respective leaders. The Congress(I)-led alliance received 54.6% of popular votes. In our survey the total of strong supporters for Mrs. Gandhi and Mr. M. Karunanidhi was 55.1%