Uttarakhand Pre-Election Tracker Survey, Round-2, January 2017

About the Survey

This analysis is based on the second round of the pre-election tracker survey conducted in by Lokniti, Centre for the Study of Developing Societies (CSDS), Delhi, for ABP News. The survey was conducted from January 15 through January 20, 2017 among 1845 voters in 98 locations (polling stations) spread across 20 assembly constituencies. These are the same constituencies and polling stations where Lokniti had conducted the first round of the tracker survey in December 2016. The sampling design adopted was Multi-stage random sampling. The assembly constituencies where the survey was conducted were randomly selected using the probability proportional to size method. Thereafter, five polling stations within each of the sampled constituencies were selected using the systematic random sampling method. Finally, the respondents were also randomly selected from the electoral rolls of the sampled polling stations.

Before going to the field for the survey, field investigators were imparted training about the survey method and interviewing techniques at a day-long training workshop held in Srinagar, Garhwal. The field investigators conducted face-to-face interviews of the respondents in Hindi asking them a set of standardized questions. The duration of an interview was about 15 minutes. The survey could not be conducted at two locations in two assembly seats. At some locations the non-availability of sampled respondents or difficulty in finding households necessitated replacements or substitutions. Meanwhile, inaccessibility of some of the areas of the state due to snowfall resulted in a low completion rate in some constituencies.

The achieved sample has been weighted by age-groups, gender, locality, religion, and caste group, based on Census 2011. The poll has an overall margin of error of plus or minus 3.3 points. The error margin increases for cross-tabulations.

Profile of Achieved Sample in Uttarakhand Survey Census Sample 2011 (%) (%) Scheduled Caste (18+ years) 13.0 17.4 Muslim 9.9 13.9* Women (18+ years) 45.5 50.1 Urban (18+ years) 33.1 31.9 18-25 years 13.0 26.0 26-35 years 26.2 24.1 36-45 years 21.9 19.1 46-55 years 16.4 13.2 56+ years 22.4 17.6 *Includes all ages

The fieldwork of the survey was coordinated by Dr. Rakesh Negi, Assistant Professor (Ad hoc) at Hemvati Nandan Bahuguna Garhwal University, Srinagar. The survey was designed and analysed by a team of researchers at Lokniti. The team included Asmita Aasaavari, Arushi Gupta, Dhananjay Kumar Singh, Himanshu Bhattacharya, Jyoti Mishra, Souradeep Banerjee, Shashwat Dhar, Shreyas Sardesai and Vibha Attri. The survey was directed by Prof. Sanjay Kumar, Prof. Suhas Palshikar and Prof. Sandeep Shastri of Lokniti.

Survey details in a nutshell Conducted by Lokniti-CSDS, for ABP News Dates of fieldwork Jan 15-20, 2017 No. of Assembly Constituencies covered 20 No. of Polling Stations (Locations) covered 98 No. of interviews conducted (Sample size) 1845 Overall Margin of Error +/-3.3 Sampling method Multi-stage random sampling Fieldwork method Standardized face-to-face interview in Hindi

Summary of Findings

Lokniti team

The Bharatiya (BJP) continues to be ahead of the ruling (Congress) in Uttarakhand, however the vote difference between the two parties has narrowed considerably over the last one month. The Lokniti-CSDS pre-election Tracker survey conducted in the state between January 15 and 20, 2017 among 1845 voters spread across 98 locations in 20 assembly constituencies found the ruling Congress to be securing 36 percent of the votes, which is up by 3 percentage points since the first round of the Tracker survey conducted by Lokniti- CSDS in December 2016. Its main rival the BJP, which had been way ahead in the last survey, has seen its vote share drop by one point to 39 percent. In a state where such closely fought contests are the norm rather than the exception, these findings come as no surprise.

In the days ahead a lot may depend on indecisive voters. The survey found a little over one in every three (36 per cent) voters of the BSP and two in every five (41 per cent) voters of other smaller parties to be unsure about their voting choice. In other words, even though these voters had certain political preferences in mid-January when the survey was conducted, their voting decisions might change on or closer to the day of election. Interestingly, voters voting for these smaller parties were found to be slightly more inclined to believe that the Congress is winning the assembly election rather than the BJP. While this may just be a perception, it may benefit the Congress if there is a last minute shift of votes.

The survey which was conducted at a time when the Congress and the BJP had not announced all their candidates, found that nearly one-thirds (31 percent) of the respondents are planning to vote on February 15 keeping in mind the candidates standing from their respective constituencies. In the December survey, the proportion of those who had said they will vote on the basis of candidates had been much less at 19 percent or one in five. As one gets closer to the polls, voters seem to be more concerned about the candidates being fielded in their constituencies.

The latest survey found the vote difference between the BJP and Congress to have reduced considerably in the Maidaan and Garhwal regions of the state. In Maidaan, the BJP had been ahead of the Congress by a margin of ten percentage points in early December. This has now reduced to about six points. In Garhwal, the BJP continues to be comfortably placed but here too the Congress seems to be catching up. Meanwhile, the Congress maintains its overall vote lead over the BJP in the Kumaon region. Overall, the ruling party leads its challenger among rural voters but trails it by seven points among voters living in urban areas.

In terms of caste and community preferences, the Congress was found to be doing far better among Muslims, Dalits and OBCs than it was in the first round. Meanwhile, Brahmins seem to have further consolidated behind the BJP since December.

An increase in Chief Minister Harish Rawat’s popularity over the last one month seems to be one of the chief reasons for the rise in the Congress’s vote share. Whereas in December only 19 percent of the voters had expressed a spontaneous desire to see Rawat back as CM in response to an open-ended question on CM preference, in the latest survey the proportion of those favouring Rawat went up to 31 percent. Rawat’s performance as chief minister is also rated quite highly by the voters. About 80 percent of the voters were found to be satisfied with his performance, an eight point rise since the previous survey. Moreover, this overall rating of Rawat’s performance was found to be just five points less than the approval rating of Narendra Modi as Prime Minister.

However, despite Rawat’s appeal, factionalism in the ruling party might hurt its electoral prospects. The survey found 4 percent of the respondents favouring state Congress President Kishore Upadhyay for the post of chief minister and among them only only-fifth (21 percent) were found to be voting for the Congress. Similarly about two percent wanted to see other Congress leaders as chief minister and four in every five of them do not plan to vote for the Congress. The BJP seems on a much firmer footing in this respect, as almost all supporters of various BJP leaders are voting overwhelmingly for the party.

There is a bit of worrying news for the BJP, however. The groundswell of support which demonetization (notebandi) had evoked from the voters in December 2016, has dropped substantially (from 60 percent to 46 per cent). Four of every ten voters (40 percent) now think that the decision might have been well intentioned but was done without preparation. Meanwhile one in every ten (9 percent) voters said that demonetization should never have happened. Interestingly, the Congress seems to have gained a great deal from this shift in mood. The survey found a considerable rise in support for the party among those who were strongly critical of the move. The survey also found a stark reversal of opinions on whether the common man had suffered more than the rich owing to notebandi, with many more people (36 percent compared to 26 percent in December) now taking the view that the common man had suffered more.

Uttarakhand Pre-Election Tracker Survey, Round-2, January 2017 (Findings)

Overall Estimated Vote Share if Assembly Polls had been held in third week of January Vote estimate Change since Jan 2017 Dec 2016 (%) (% points) Indian National Congress (Cong) 36 +3 (BJP) 39 -1 Bahujan Samaj Party (BSP) 6 -2 Others 19 0 Note: Undecided voters who did not disclose their vote preference have been distributed proportionally after a careful analysis. The question asked was: If assembly elections in Uttarakhand are held tomorrow then what party will you vote for? We are giving you a slip which has the names and symbols of different parties. Please mark your vote on this slip and put it in this box.

About two in five voters of BSP and Other parties may change their vote on Election Day Will vote for same May vote for Gave no party on Election Day (%) someone else (%) response (%) All voters 73 17 10 Cong voters 83 8 9 BJP voters 84 9 8 BSP voters 58 36 6 Others’ voters 39 41 20 Note: Figures may not add up to 100 due to rounding The question asked was: On the day of voting will you vote for the same party which you voted for now or your decision may change?

Voters of smaller parties were slightly more likely to think that Cong is winning the election in Uttarakhand than BJP Cong will win BJP will win Some other No party Can’t say the election the election party will win will win who will win All voters 29 37 4 2 29 Those voting for Cong 68 6 <1 <1 25 Those voting for BJP 3 79 1 1 17 Those voting for Others 15 5 18 6 56 Note: Figures are percentages; they may not add up to 100 due to rounding The question asked was: In your opinion, which party will win or be the front runner in the upcoming election in Uttarakhand?

For nearly one-third voters, candidate will be a major voting consideration Will vote on Will vote on Will vote on Will vote in Will vote on No Party Local candidate CM candidate Modi’s name basis of Other response basis basis basis (unstated option) matters (stated option) (stated option) (stated option) (unstated option) All voters 47 (36) 31 (19) 9 (6) 9 (17) 1 (8) 4 (14) Cong voters 58 (44) 23 (23) 12 (8) 1 (3) 3 (10) 3 (12) BJP voters 39 (33) 29 (16) 8 (6) 21 (37) <1 (1) 3 (7) Note: Figures are percentages; they may not add up to 100 due to rounding. Figures in parentheses are from Dec 2016 survey. The question asked was: While deciding who to vote for in the upcoming assembly election, what will matter to you most - party, local candidate or the chief ministerial candidate of a party? The options of ‘Modi’ and ‘Other matters’ were unstated, and were not read out to the respondents.

Modi still assessed more positively than Rawat, however voters seemed more satisfied with Rawat’s work compared to December and a little less satisfied with Modi’s Fully Somewhat Somewhat Fully No Satisfied Satisfied Dissatisfied Dissatisfied response Harish Rawat’s performance as CM 38 (+2) 42 (+6) 9 (+3) 10 (-7) 1 (-4)

Narendra Modi’s performance as PM 52 (-5) 33 (+4) 5 (+3) 9 (+1) 1 (-3) Note: Figures outside parentheses are percentages; they may not add up to 100 due to rounding. Figures within parentheses represent percentage point change since December 2016 survey. The questions asked were: (1) What is your assessment of Harish Rawat as Chief Minister – are you satisfied or dissatisfied with his performance? (Probe further whether fully or somewhat) (2) What is your assessment of Narendra Modi as Prime Minister – are you satisfied or dissatisfied with his performance? (Probe further whether fully or somewhat)

Anti-incumbency sentiment has weakened in one month % Cong govt. should get another chance 36 (34) Cong govt. should not return 36 (44) Did not give a response 28 (22) Note: Figures may not add up to 100 due to rounding. Figures in parentheses are from the December 2016 survey. The question asked was: Should the present Congress government in Uttarakhand get another chance?

Harish Rawat spontaneous CM choice of one-third voters compared to one-fifth in Dec Dec 2016 Jan 2017 Harish Rawat 19 31 B. C. Khanduri 13 14 Kishore Upadhyay <1 4 Bhagat Singh Koshiyari 2 3 Other BJP leaders 5 11 Other Congress leaders 2 2 Leaders from other parties 5 1 Did not take any name 54 34 Note: Figures are percentages; they may not add up to 100 due to rounding. The question asked was: After the assembly election, who would you prefer as the next Chief Minister of Uttarakhand? The question was asked in an open ended way. No names were offered to the respondent.

Is factionalism hurting Cong? Only a fifth of Kishore Upadhyay’s supporters are voting for Cong; meanwhile almost all supporters of various BJP leaders are voting for BJP Vote for Vote for Vote for Cong (%) BJP (%) Others (%) Those who want Harish Rawat as next CM 87 4 9 Those who want Kishore Upadhyay as next CM 21 21 58 Those who want other Cong leaders as next CM 78 21 1

Those who want B C Khanduri as next CM 2 95 3 Those who want B S Koshiyari as next CM 2 98 0 Those who want other BJP leaders as next CM 1 99 0 Note: Figures may not add up to 100 due to rounding

BJP leads Cong in Garhwal and Maidaan but gap has narrowed; Cong ahead in Kumaon. Vote for Vote for Vote for Vote for Cong (%) BJP (%) BSP (%) Others (%) Garhwal - 25 seats 34 (+10) 47 (+3) <1 (-1) 18 (-12) Kumaon - 22 seats 43 (+1) 34 (-2) 7 (-4) 16 (+5) Maidaan - 23 seats 33 (+2) 39 (-2) 7 (-2) 21 (+2) Note: Figures outside parentheses may not add up to 100 due to rounding. Figures in parentheses represent percentage point change since December 2016 survey.

Cong takes narrow lead among rural voters, but loses urban support Vote for Vote for Cong (%) BJP (%) Rural 43 (+10) 40 (+1) Urban 25 (-8) 38 (-5) Note: The rest of the respondents indicated a vote preference for other parties. Figures within parentheses represent percentage point change since December 2016 survey.

Cong is now doing far better among Muslims, Dalits, OBCs; Brahmins consolidating behind BJP Vote for Vote for Vote for Vote for Cong (%) BJP (%) BSP (%) Others (%) Brahmin 33 (-5) 60 (+14) 3 (+3) 4 (-12) Rajput 30 (-4) 43 (-8) 5 (-1) 22 (+13) Other Upper castes 17 (-5) 46 (+5) 3 (+3) 35 (-2) OBC 38 (+11) 56 (0) 5 (+2) 1 (-13) Dalit 46 (+18) 34 (-7) 20 (-6) 1 (-4) Muslim 68 (+21) 10 (+6) 6 (+2) 16 (-29) Others 29 (+6) 29 (-26) <1 (-14) 42 (+35) Note: Figures outside parentheses may not add up to 100 due to rounding. Figures within parentheses represent percentage point change since December 2016 survey.

Cong has edge among youngest voters; BJP leads among middle-aged and elderly voters Vote for Vote for Cong (%) BJP (%) 18-30 years old 39 (+6) 38 (+2) 31-49 years old 34 (+1) 39 (-2) 46+ years old 36 (+4) 39 (-5) Note: The rest of the respondents indicated a vote preference for other parties. Figures within parentheses represent percentage point change since December 2016 survey.

Closer to the election, vikaas emerges as most important issue for one in three voters Important voting issues Dec Jan 2016 2017 Development 17 33 Unemployment 18 16 Inflation 11 9 Poverty 4 9 Corruption 6 6 Roads 6 4 Water, electricity supply 2 3 Demonetization 1 2 Other issues 6 9 No response 29 10 Note: Figures are percentages; they may not add up to 100 due to rounding. The question asked was: Suppose assembly elections are held in Uttarakhand tomorrow then while deciding whom to vote for, what will be the most important election issue for you? The question was asked in an open ended way. No options were given.

BJP leads among those for whom inflation, jobs & corruption are most important issues; Cong leads among those for whom development, poverty & roads are top concerns Vote for Vote for Cong (%) BJP (%) Development is most important voting issue (33%) 36 33 Unemployment is most important voting issue (16%) 37 54 Inflation is most important voting issue (9%) 42 51 Poverty is most important voting issue (9%) 32 24 Corruption is most important voting issue (6%) 22 58 Condition of roads is most important voting issue (4%) 58 22 Note: The rest of the respondents indicated a vote preference for other parties

Outright support for demonetization drops in Uttarakhand Demonetization was a… Dec Jan 2016 2017 Right move 60 46 Right move but done without preparation 28 40 Wrong move, should never have been taken 9 12 No response 3 2 Note: Figures are percentages; they may not add up to 100 due to rounding. The question asked was: People have different opinions regarding demonetization/cancellation of Rs 500 and Rs 1000 notes. Some people believe that this was a right move; some others believe that it was the right move but done without any preparation, while some others say it was a wrong decision that should never have been taken. What is your opinion?

Cong support among strong critics of demonetization sees huge increase; party continues to lead BJP among partial critics as well Vote for Vote for Vote for Cong (%) BJP (%) Others (%) Those who say demonetization was a right 20 (-3) 60 (+9) 20 (-6) move (46%) Those who say demonetization was a right 39 (-6) 26 (-2) 35 (+8) move but done without preparation (40%) Those who say demonetization was a 88 (+32) 4 (+1) 8 (-33) wrong move (12%) Note: Figures outside parentheses may not add up to 100 due to rounding. Figures within parentheses represent percentage point change since December 2016 survey.

Most now believe Common man suffered more than the Rich due to demonetization; in December it had been the other way round Dec 2016 Jan 2017 Common man has suffered more than the Rich 26 36 Rich have suffered more than common man 39 27 Both have suffered equally (unstated option) 29 36 No one has suffered (unstated option) 3 1 No response 3 1 Note: Figures are percentages; they may not add up to 100 due to rounding. The question asked was: People have different opinions regarding demonetization/cancellation of old Rs 1000 and Rs 500 notes. Some people say that due to this move, the common man has suffered more than the rich. On the other hand, some others say that the rich have suffered more than the common man. What is your opinion?