Survey Highlight

 32% of the respondents in Delhi and 45% of the respondents in Rajasthan have said that their opinionabout the BJP has improved by the manner in which the Central government had handled the Gujarat riots.

As mentioned earlier, a sample ofthe voting public were interviewed in 123 Assembly Constituencies in the four states of Delhi, Rajasthan, MPand Chhattisgarh.  

The respondents those who saidthat they were likely to vote in the coming elections, were asked if they had made up their minds about whichparty they would vote for. 

More than 90% across the fourstates stated that they definitely or most probably intended voting in the forthcoming assembly elections tobe held on December 1, 2003 and more than 90% in  Rajasthan and MPsaid that they had even decided on which party they would vote for.  Whereas in Delhi 14% and about a third (34%) in Chhattisgarh had not made up their minds about which party to votefor.

Voting intention and seatforecast for coming Vidhan Sabha elections  

BJP

34.0

33.4

--0.6

15

11

INC

47.8

49.7

+1.9

52

56

Others

18.2

16.9

--1.3

3

3

Total

100

100

0

70

70

In Delhi it seems that Congress islikely to improve upon its performance over 1998 when it had swept to power with a big majority. It is likelyto improve its tally to 56 seats from 52 seats at the expense of BJP, which will now be reduced further to 11seats from the 15 seats that it obtained in 1998.

 Votingintention and seat forecast for coming Vidhan Sabha elections  

BJP

33.9

36.4

+2.5

33

67

INC

45.2

40.7

--4.5

153

113

Others

20.9

22.9

+2.0

14

20

Total

100

100

0

200

200

In Rajasthan though there is a swing of 4.5% against the Congress and also a swing of 2.5% in favour of the BJP, it does not appear tobe enough to dislodge the Ashok Gehlot government which came to power in 1998 with a substantial lead of over11% in vote share over the BJP. 

BJP

38.8

43.1

+4.3

83

147

INC

41.2

34.2

--7.0

124

58

Others

19.9

22.6

+2.7

23

25

Total

100

100

0

230

230

In Madhya Pradesh BJP appears tobe a clear winner and is expected to sweep into power with a great majority of 147 seats out of a total of230.  The Congress being relegated to a distance second positionwith 58 seats.  The dissatisfaction with the Digvijay Singhgovernment in the State has led to a negative swing of 7% against the Congress as compared to 1998 with theBJP reaping in most of the advantage.

BJP

39.5

40.0

+0.5

36

43

INC

41.0

40.0

--1.0

48

43

Others

19.5

20.0

+0.5

6

4

Total

100

100

0

90

90

According to the poll, inChhattisgarh there has been a swing of 1.0% against the Congress and 0.5% in favour of the BJP and this is likely to result in an even share of votes between the two parties. Thus, both Congress and BJP arelikely to get 43 seats each out of a total of 90 seats in the state.  

This is on the assumption thatthere will be free and fair elections as some doubt has been expressed by people and even by the ElectionCommission about the impartiality of the officials in charge of the elections. 

Further, the fieldwork for thispoll was conducted before the exposure of Sh. Dilip Singh Judev accepting a bribe resulting in his resignationfrom the Central Government. Hence, the results do not reflect the likely impact of this episode on the finaloutcome as we go forward.

If the Judev scandal gathers steamin the next few days, it is likely that a reversal happens and Congress could win the elections especiallywhen the margin in vote share is so narrow.

The terms of State LegislativeAssemblies in four states viz., Delhi, Rajasthan, Madhya Pradesh and Chhattisgarh are coming to an end on thefollowing dates

Outlook commissioned ACNielsento conduct an Opinion Poll to gauge the current mood of the electorate and also to determine the likelyoutcome of the four State Assembly Elections. The Poll also sought todetermine the perceived strengths and weaknesses of the Political Parties in the fray and who was consideredto be the most capable of being the Chief Minister in each of the four states.

The main purpose of the research,as indicated earlier, was to determine the likely outcome of the State Assembly Elections and also to predictthe number of seats likely to be won by each of the main Political Parties (or alliances) contesting theelections.

More specifically we sought todetermine the following from the study;

(i)Past voting patterns both the last General Elections and State Assembly Elections

(ii)Voting intentions for the forthcoming Assembly Elections (Party of 1^st choice and Party of 2^ndchoice) and reasons for voting for a particular Party.

(iii)Image of the main Political Parties contesting elections in the State on a number of factors such aslaw and order, stability, efficiency, water scarcity; drought etc.

(iv)Level of satisfaction / dissatisfaction with the present Governments in the States.

(v)Evaluation of the performance of current State Chief Ministers – Shiela Dikshit in Delhi, AshokGehlot in Rajasthan, Digvijay Singh in Madhya Pradesh and Ajit Jogi in Chhattisgarh.

(vi)Major issues being faced in each of the four States and the Party that is perceived to be most capableof handling these issues.

(vii)Image and views on key leaders in the main Political Parties in each of the four States.

(viii)Who is considered to be the most capable of being the next Chief Minister in each of the four States.

(ix)Demographic profile of the voters including religion, caste, literacy level etc.

The poll was conducted using astructured questionnaire and personal interviews were conducted among a sample of the voting public chosen atrandom.

The proposed coverage ofAssembly Constituencies in each state and the actual coverage including the proposed and actual sample sizesachieved are given in the table below. In each Assembly Constituency approximately 180 men and women werepolled for the survey. In all a total of 21,936 interviews wereconducted in the four states as illustrated below:

The assembly constituenciesthat were chosen for the poll were selected on the basis of the voting pattern as observed over the last twostate assembly elections. Consideration was also given to:

·Swing in the selected assembly constituencies to closely match the overall swing in the state as awhole;

·Assembly constituencies that have shown shifts across parties over the two elections;

·Assembly constituencies that have elected representatives with small margins.

·VIP constituencies to be omitted.

While selecting the assemblyconstituencies care was taken to ensure that they were geographically spread to cover all the regions of theState. The list of assembly constituencies that were covered in thestudy in each state is appended to this report.

The sample in each assemblyconstituency was split between Urban and Rural areas with the actual split depending on whether the assemblyconstituency is primarily urban or rural. At the time of analysis thesefigures were weighted to reflect the actual Urban / Rural distribution in each assembly constituency / State.

In the urban areas interviewswere conducted with men and women in the ratio 60:40, and only males were interviewed in the rural areas, aswe normally do for such studies.

All interviews were conducted bytrained ACNielsen investigators working under supervision. They werepersonally briefed by a senior ACNielsen executive.

The fieldwork for the study wasconducted from 30^th October to 9^th November 2003.

We have been collectingpolitical opinion poll data since 1989 and have now developed a sophisticated and robust model for forecastingthe number of seats a party is likely to obtain.

In the forecasting model that wehave developed we lay a strong emphasis on attitudes people express towards a particular party, in terms ofhow they feel each party can tackle current political issues such as ‘riot handling’ / ‘communaldisharmony’ / ‘rising prices’ / ‘communalism’ / ‘law &order’ etc. These responses are weighted to arrive at a‘surrogate’ voting intention for each individual.

Also the model gives moreemphasis on the swing in voting intentions rather than actual voting intention.This is because it is difficult to get a totally representative sample covering all communities andbackward classes. The swing factor is subject to less error than theactual voting distribution. The swing data is usually superimposed onthe last election results to arrive at the new forecast.