Kerala and Tamil Nadu Pre-Poll Observation

Today is Election Day in Kerala and Tamil Nadu. Originally, I was planning to write an election guide for each of the state elections going on right now. But time has not allowed for it. I will be heading back to the United States in less than two weeks and am starting to wrap up my research here. Where has the time gone?!

I spent the last week travelling through these two southern states to observe fieldwork for the ongoing CSDS pre-poll election studies. Unlike for Assam and West Bengal, these studies were being fielded before election day because of the quick turnaround on analysis. In both states, but particularly in Kerala, signs of the imminent elections were everywhere. Billboards and posters for the governing United Democratic Front (UDF) competed for space with the challenging Left Democratic Front (LDF). The BJP also had established a strong campaign presence in the capital of Thiruvananthapuram and the surrounding areas. In Tamil Nadu, I saw fewer campaign posters, particularly in the cities of Madurai and Chennai. In the villages surrounding Madurai, however, the two-leaf symbol of the governing Anna Dravida Munnetra Kazhagam (ADMK) and the rising sun symbol of the challenging Dravida Munnetra Kazhagam (DMK) decorated the walls of houses. Flags from the All India Forward Bloc (AIFB), which won the constituency I was visiting in 2011, were also in abundance. By the roadside, streamers with pictures of Jayalaalithaa, Tamil Nadu’s chief minister, or Karunanidhi, her rival from the DMK, hung from different campaign rally sites.

Like my previous post, a concise summary of the field visit is will be presented here rather than an extended narration.

On May 5, I flew down to Thiruvananthapuram by way of Chennai and checked into the guesthouse at the University of Kerala campus in Kazhakuttom. On May 6, I observed the full-day training for Kerala field investigators at the University of Kerala. Sandeep Shastri, National Coordinator of the Lokniti Network and Pro Vice Chancellor at Jain University in Bangalore, led the training. On May 7, I met up with two field investigators working in Nemom, a suburb that is 8 km south of Thiruvananthapuram city center (the assembly constitutency of Nemom was won by the Communist Party of India (Marxist), or CPM, in 2011). We met up late in the afternoon, around 4 PM, and so I only observed one interview since so many listed respondents were not home. On May 8, I met up with two other field investigators working in the village of Pallikkal, which is about 40 km north from Kazhakuttom in the assembly constitutency of Varkala (which was won by the Indian National Congress in 2011). Like the day before, we met late in the day — close to 4 PM — and so I could only observe two interviews. Like in Nemom, many of the selected respondents were not home at the time of our visit.

On May 9, I travelled by overnight train, the Ananthapuri Express, from Thiruvananthapuram to Madurai in Tamil Nadu. Madurai is the third largest city in Tamil Nadu, home to the legendary Meenakshi Amman temple complex. On May 10, I met up with four field investigators working in the area around Usilampatti, a town that is 40 km west of Madurai. I observed the work of FIs interviewing respondents in the countryside villages surrounding Usilampatti: Nallathathunayakkappatti, Sangakavundanppatti, Palaiyur, Karisalpatti, and Kattarripatti. On May 11, I squeezed in a morning visit to the Meenakshi Amman temple of Madurai before joining the FIs conducting in the city of Usilampatti. The assembly constituency of Usilampatti was won by the All India Forward Bloc (AIFB) in 2011. I took an overnight train, the Ananthapuri Express once again, from Madurai and arrived in Chennai on the morning of May 12. In the morning, I ate breakfast at Saravana Bhavan, visited the museum at Fort St. George, and stopped by Higgenbothams Booksellers and Stationers before heading to the airport and flying back to Delhi.

  1. Urban vs. rural areas: As with Assam and West Bengal, respondents in urban areas were more difficult to locate than respondents in rural areas. People on the street and in the neighborhoods of selected respondents were less likely to know the location of other selected respondents in urban areas than in rural areas. This meant that a lot of the time for fieldwork in urban areas (especially Menom) was spent asking people on the street if they know the location of selected respondents and walking around to identify households. In villages, particularly in village centres where older men gather, people were much more likely to identify the names of selected respondents and point us towards their houses.
  2. Low response rate in Kerala: Both researchers from the CSDS office in Delhi and the Lokniti coordinator for Kerala, Professor Sajad Ibrahim Ali, informed me that historically the response rate for Kerala is lower than that of other Indian states. This has been the case in state and national election studies, and as the data were being collected for this election study, Professor Ibrahim informed me that it would also be the case this time. Several team members from Kerala informed me that response rates are higher in north Kerala than in south Kerala, and particularly in Theruvananthapuram and the surrounding areas. No one had a clear answer about why this is. Some field investigators speculated that it had to do with the number of people working who were not home during the day. On my two admittedly short observations of fieldwork, field investigators experienced challenges locating houses and arranging interviews with the selected respondents once they had located the households.
  3. Caste in Tamil Nadu: Ranith, one of the field investigators whom I was observing, informed me that caste was a very sensitive issue in many of the villages around Usilampatti. The Lokniti state coordinator for Tamil Nadu, Professor P Ramajayam said, also said that Usilampatti is notorious for intercaste violence. Violent caste-related incidents in Tamil Nadu are most common between different backwards classes, such as Gounders and Vanniyars, and Scheduled Castes (SC) or dalits. In some of the villages we visited, it was dangerous for field investigators to even raise the issue or ask questions about it (every CSDS election study includes demographic questions about caste). For some interviews, the field investigators would therefore skip any such questions in the questionnaire. Field investigators in Tamil Nadu have to carefully navigate the complexities and sensitivities of caste while collecting data. That all four field investigators were from Usilampatti and the surrounding villages was critical, as they knew which villages it was safe to ask about caste.
  4. Religion: While Tamil Nadu is a state with a large (88%) Hindu majority, Kerala is a state with sizable Muslim and Christian minorities (54% Hindu, 27% Muslim, 18% Christian). The field investigators whom I observed in Kerala expressed skepticism that response rate was different across religion, though they said it was sometimes harder to speak with women in Muslim households (all the field investigators with whom I worked in Kerala were female). Though the neighborhood we were interviewing in Pallikkal was largely Muslim, I did not spend enough time in the field in Kerala to receive an informed impression about how religion shapes fieldwork in Kerala.
  5. Gender: The gendered dynamics of interviews which I observed in Assam and West Bengal were not as much in evidence in Kerala and Tamil Nadu. That is to say, while female respondents who were selected were nearly always interviewed in the presence of their father or husband in Assam and West Bengal, this did not seem to be the case in Kerala and Tamil Nadu. However, any difference may be due to the fact that the interviews I observed with female respondents in Kerala and Tamil Nadu were all conducted by female interviewers; in West Bengal, the interviewers whom I observed were male. In Tamil Nadu, the team of field investigators included two male FIs and two female FIs. This was for the safety of female FIs when traveling through villages in the countryside. It also allowed for male FIs to interview male respondents and female FIs to interview female respondents.
  6. Education level: For a number of respondents in Usilampatti and the surrounding areas, education level was so low that respondents could not understand the content of many of the questions. One respondent, for example, responded to a question about Prime Minister Narendra Modi’s different government initiative by saying that she did not know who Narendra Modi is. Field investigators thus had to make spur-of-the-moment judgments about which questions to ask and which could not be reasonably asked of the respondents. In the case of illiterate respondents who lacked even basic knowledge about political affairs in Tamil Nadu and India, the focus was on collecting correct demographic data and ensuring that the respondent answered the question about voting intention.
  7. Timing: Many of the respondents in Nemom, Pallikkal, Usilampatti, and the surrounding villages of Usilampatti were not home during the daytime because they were at their jobs. In the case of villages around Usilampatti, there was one village for which all or most of the villagers were out in a field working under the MGNREGA (Mohandas Gandhi National Rural Employment Guarantee Act). We went to the field to find the villagers assembled in the shade of the trees. When the FIs began conducting interviews with selected respondents, a village clerk from the panchayat (local administration) showed up and demanded that the interviews cease. The local official told the villagers not to answer any questions being asked by survey researchers. The FIs tried talking to the clerk and showed them their official CSDS letter, but the clerk insisted that we leave. This incident showed how field investigators sometimes must deal with obstacles presented by local government. It also illustrated how unemployed workers receiving benefits through MNREGA (which guarantees 100 days of agricultural work a year) may be less likely to be interviewed for surveys.
  8. Economic migration: Many of the selected respondents who were not available in the villages surrounding Usilampatti had left their village to work in either a larger city of Tamil Nadu or another state like Kerala. Such economic migration meant that many selected respondents in a given village could not be interviewed.

My pictures below. Unfortunately, my phone failed me on the date of May 6 so I lost pictures from the training workshop at the University of Kerala. What was perhaps most visually striking in Tamil Nadu was the placement of beautiful towering temples with elaborate carvings in village centres, even in the tiniest and poorest of villages.

Counting Day is only three days away. CSDS should have some analysis in the papers following the announcement of the election results.

May 7 (Nemom, Kerala)




Billboards for the United Democratic Front and National Democratic Alliance, over graffiti for the Communist Party of India (Marxist)








Picking jasmine



Graffiti for the Indian National Congress / United Democratic FrontIMG_20160507_174525

May 8 (Pallikkal, Kerala)


University of Kerala campus in Kazhakootam













May 9 (Kazhakootam to Madurai)


With Rajeena at the University of Kerala canteen



Aboard the Ananthapuri Express

May 10 (Villages around Usilampatti, Tamil Nadu)


Tamil dailies report on the election


In the Dalit section of a village





Anna Dravida Munnetra Kazhagam (ADMK) and Dravida Munnetra Kazhagam (DMK) posters










The two leaves are the symbol for the ADMK


The sun rising between the mountains symbol of the DMK










Ranith, Thanalakshmi, and Pechi

May 11 (Madurai and Usilampatti, Tamil Nadu)


At the Meenakshi Amman temple










Golden Lotus Tank








Lord Ganesha




A model of the temple complex


Dancing Shiva and Parvati


In the Thousand Pillar Hall








Usilampatti bus station


Interviews in Usilampatti






With Ram’s grandmother


Downtown Madurai

May 12 (Chennai)


Fort St. George museum



Higginbotham’s Booksellers, oldest bookstore in India

Jaipur and Jodhpur

Blogging has been backlogged — I am just now posting photos from Jaipur and Jodhpur in the days following the South and Central Asia Fulbright Conference at the end of February / start of March.

Tomorrow, I will be going out to Assam and West Bengal to observe exit polls being conducted by CVoter and fieldwork for the CSDS post-polls in both states. Reports to be posted on the blog once I return.

Jaipur (March 2-3)


Entrance to the Pink City


Hawa Mahal



Albert Hall Museum


Gateway to City Palace



Tikam Chand maintains a camera that dates back to the 1860s in front of Hawa Mahal


The photos look like they’re from the 1860s


Overlooking Jaipur from Suriya Mandir

Jodhpur (March 4-5)


Looking out across the rooftops of the Old City


Umaid Bhawan, the last palace to be built in India (completed in 1944)



View from Jaswant Thada


Mehrangarh Fort


Overlooking the Blue City from Mehrangarh Fort








Mandor Gardens, where the ancient capital of Marwar was located







The Clock Tower in the middle of the Old City



Around Delhi

Next week, I will be stepping away from my research for about a month. Before I do, I wanted to post some remaining photos of sightseeing around Delhi. These were taken over the past two months.


At the Hauz Khas Complex in South Delhi. The complex houses a madrasa which was one a center of Islamic learning during the Delhi sultanate era, thirteenth century (November 1).


From the other side of Hauz Khas Lake (November 1)


At the 35th India International Trade Fair held at Pragati Maidan. Vendors from all across India to the massive fair grounds to showcase their wares (November 21).


The Martyr’s Column at Gandhi Smirti, where Mahatma Gandhi spent the last 144 days of his life. This is the exact spot where he was assassinated by Nathuram Godse on January 30, 1948 (November 22).


A.R. Rahman, prolific musician and soundtrack composer for hundreds of Indian films, performs at the Weekender music festival held in Dwarka (November 28).


At a qawalli at Hazrat Nizamuddin Dargah. Every Thursday night, musicians perform qawwali, Sufi devotional music, at this dargah (shrine) for the Indian Sufi saint Hazrat Nizamuddin which was built in the sixteenth century (December 3).


Haus Kauz Connection, a band that fuses classical Indian instrumentation with jazz, performs a free concert at the Indian Habitat Centre (December 4).


The birds of Lodhi Garden (December 13)

Diwali and McLeod Ganj

A week and a half ago was Diwali, the largest festival of the year in India. To commemorate Ram’s lamp-lit return to Ayodhya twenty days after defeating the demon Ravan and rescuing his wife Sita in the Ramayana, lights are placed outside houses and shops. Our neighborhood of Lajpat Nagar was bedecked with twinkling lights. People also light diyas, tiny candle-lit lamps, to commemorate Ram’s return.

It should be said that for many people, Diwali is as much a commercial holiday as it is a religious one. The market was buzzing with activity and discounts in the weeks before. In this way, Diwali reminds me of Christmas in the United States.



Fireworks are a big part of Diwali. I have never seen so many little children setting off mortars. Brandon and I went over to Malviya Nagar to set off some fireworks with other Fulbright scholars, stopping at our neighborhood fireworks store first. The Delhi air was intensely polluted the next day.


To celebrate this year’s Diwali, Brandon and I decided to opt for a taste of home for dinner. Here’s to you, KFC Rock Box, you seamless combination of fried chicken sandwich, fried chicken pieces, fries, and Pepsi.


The weekend after was a visit to the Himalayan town of McLeod Ganj, where both the Dalai Lama’s and Tibetan government-in-exile are located. It was a weekend of fresh air, Western-style breakfasts in cozy cafes, bookstores crammed with tomes on Buddhism, and short treks from one mountain village to another (Dharamkot, Bhagsu, and Naddi). McLeod Ganj hosts an interesting mix of Western tourists, Indian tourists, the refugee Tibetan community, and the native Himacheli community. Highlights included the Tibet Museum, the Dalai Lama Temple Complex, an Israeli-owned bookstore with window-side table looking out on the Himalayas, and trying Tibetan tea, which strongly tastes of butter.










Election Day in Bihar: Field Notes from Patna

NOTE: The Electoral Commission of India prohibits the release of trends from exit polls before voting has been completed: 5 PM on November 5, in the case of the Bihar elections. Please note that the observations in this post are purely anecdotal and not indicative of any broader trends. To comply with ECI regulations, the names of information related to polling stations has been redacted until such information can be released at 5 PM on November 5.

UPDATE: The names of polling stations have now been added to this post (3:37 PM, November 12).

Last week, I traveled to Patna to observe CVoter’s exit polls in Patna during the third phase of the elections. I traveled with Ajit Shukla, the Qualitative Manager from CVoter. Below are my field notes from our visit to the polling stations on Wedneday, October 28 and the debriefing we held at the hotel with four field investigators on Thursday, October 29.

Wednesday, October 28

What is striking about Election Day in Patna is that the roads are nearly empty. The exhaust fumes and barrage of honking that filled the roads the day before are replaced with quiet, punctuated every so often by a police car driving by with election officials or the odd rickshaw. During voting hours — 7 AM to 5 PM — private vehicles are not allowed on the road. The city is effectively militarized by the police who are under the charge of the Electoral Commission of India.

Ajit and I started at one of the polling stations in Patliputra around 10 AM. Very few people had voted by this time. The field investigator with whom we had spoken had only talked to one voter thus far. It appeared that most people at the polling station — in an area that traditionally supported the BJP — would not be voting until the evening.

We were able to catch a ride to BN College, closer to downtown Patna. Along the way, we talked to the driver and his friend, who were both supporting the JD(U). The man in the passenger seat was supporting JD(U) even though he also liked Modi.

At the polling station at BN College, we encountered four women before walking near the station. One was a Gujarati woman married to a Bihari. When I asked her how she would vote, she told me, “We are reasonable people. We are voting for the BJP. Narendra Modi is my personal hero. I saw he handled the Gujarat riots in 2002.” Her three friends, from the Banya caste, were also supporting the BJP.

We approached the exit poll interviewer from CVoter. His coding sheet suggested the BJP was heavily favored at this polling station. We also spoke to a journalist from the Patna edition of Hindustan, who was curious about the research we were doing.

From BN College, we took a bicycle rickshaw to a polling station that was near Hathwa Market. The surrounding neighborhood appeared to have a sizable Muslim population, based on the mosque. We spoke to an old man with a beard who said that he was supporting Nitish because he had worked hard for the development. A group of men standing near the roadside were all Nitish supporters, except for one, who said he was supporting the BJP because “traditionally he supports the BJP.” We talked to a chaiwallah who said he was supporting Nitish because of the development he brought to Patna. His wife, standing close by, told us, “Whatever my husband does [with regard to voting], I am with my husband.”

Speaking to the exit poll interviewer, we learned that this polling station would be a close fight between the NDA and the grand alliance. There was a long queue of voters going into the polling station.

From this polling station, we walked towards the Patna Junction railway station. Along the way, we spoke to a couple who were supporting Nitish. The man said he was supporting Nitish because of “development,” while his wife — who was from Bengal — mentioned that she was supporting Nitish because of his programs to give out bicycles, school uniforms, books, and monthly scholarships. We spoke to one store owners who said he was voting for “Those who are winning, so we don’t destroy our votes.” When we asked him who he meant by this, he said he was voting for the BJP.

We took a bicycle rickshaw to get the rest of the way to Patna Junction. Our rickshaw puller was voting for Nitish because of his development record and his promises to do more development programs for the poor. We spoke to a chaiwallah who said he was a Kurmi voting for the BJP because he did not want Lalu to return to power and bring back “jungle raj.” We also spoke to a Muslim who was voting for Congress.

The fourth and last polling station we came to was in Phulwari Sharif. It is a heavily Muslim dominated area on the outskirts of Patna. We spoke to the exit poll interviewer there, who said that getting respondents to reveal their vote choice was challenging. This was reflected in his codesheet, which showed many “99”s of the ten or so respondents he had talked to. These “99”s symbolized “Don’t Know”s; most of these respondents did not want to reveal their votes to the interviewer, though they had told him the issues which they were considering when voting. We spoke to two men in the area who were JD(U) supporters; however, most people we talked to were very hesitant to share their vote intention. Ajit told me that this is often the case with Muslim respondents.

Thursday, October 29

We held a debriefing with four of the interviewers from CVoter’s exit polls — one was the interviewer from BN College (he had previously worked as a CATI interviewer for CVoter), one was the field coordinator for Hajipur (he had more than eight years of experience conducting field research), one was an interviewer in Digha (he had been working as a field investigator for CVoter for about six months), and another was the interviewer from the first polling station which we had visited in Patliputra (he had been working as a field investigator for CVoter for the last few months). I asked questions in English, which Ajit translated into Hindi. Ajit then translated the interviewers’ responses in Hindi back into English.

The interviewers said that the issues that mattered most to voters were “development,” “development and inflation,” “development and change of power,” and “development and change of government.” All four interviewers reported that respondents were most comfortable speaking with them about the issues that they considered while voting, and least comfortable speaking about whom they would be voting for. They all reported that Hindu and male voters were more willing to speak with them. Women, OBCs, SCs, STs, and Muslims were less likely to speak with respondents. Interviewers reported that roughly 20% of respondents did not want to respond to questions about how they voted.

I closed the debriefing by asking the interviewers who they thought would win in their polling stations, and who they thought would win the election. The interviewer for BN College thought the NDA would win, projecting a 60/40 margin. The coordinator for Hajipur thought it would be close in Hajipur, perhaps 50/50. The interviewers for Patliputra and Digha both thought the NDA would win their polling stations. All four interviewers predicted an NDA victory for all of Bihar.


Patliputra neighborhood


Gandhi maidan: These expansive field grounds in the heart of Patna are where the Quit India Movement was launched in 1942



The world’s largest statue of Gandhi is located in Gandhi maidan




Golghar: A granary built by the British in 1786, after the famine of 1770 killed 10 million people in Bihar and Bengal


The view from atop Golghar. Patna traffic on the day before the election


Facing west from atop Golghar


Facing north


Jay Prakash Nagar


Election Day: Voters emerge from the polling station


The roads are quiet on Election Day


Bicycle rickshaws are the easiest way to get around Patna on Election Day





An ink marking is left on voters’ fingers to prevent voter fraud

Research updates

Last week, I finally made contact with Impetus Research, a market research agency based in the southwest Delhi neighborhood of Dwarka. In contrast to my other research affiliates, most of Impetus’ surveys are for international business clients. Working with them should provide a different perspective on the challenges of conducting survey research in India.

I am traveling back to Patna this week to observe the exit polls that CVoter will be fielding on Wednesday, which is when the third phase of Bihar’s elections will be held. Blogging will be quiet until the end of the week.

Over the weekend, I visited Gurudwara Bangla Sahib, Delhi’s most magnificent gurudwara (Sikh house of worshop), and Agrasen ki Baoli, a step well that dates back to at least the 14th century.

Gurudwara Bangla Sahib

Agrasen ki Baoli

Uncle sign

Old Delhi and Indira Gandhi memorial

Below is the Jama Masjid, the Mughal-era mosque located in the heart of Old Delhi, which is also called Shahjahanabad. It was built in the middle of the 17th century when Shahjahan was constructing the Mughal capital in Delhi. Jama Masjid is the largest mosque in India.





Two snapshots of Old Delhi below. It is extremely congested. If you ever visit, make sure to stop by Karim’s, next to Jama Masjid. This restaurant was run for centuries during the Mughal era, before the family owners were expelled when the Mughal dynasty was ended by the British in 1857. They returned in 1911 and have maintained the restaurant ever since.


20151007_161703 (1)

Below are photos from Indira Gandhi’s house and memorial. Indira lived here during her prime ministership (1966-1977 and 1980-1984). The museum covers the events of her childhood and prime ministership, skating over the darker aspects of her legacy such as the suspension of India’s democracy during the Emergency from 1975-1977, and Operation Blue Star, a ham-handed raid on a Punjabi insurgency in the Harmandir Sahib of Amritsar in 1984. As you can see from the crowds below, she remains India’s most popular politician. Indira was assassinated outside her house by Sikh bodyguards several months after she ordered Operation Blue Star.

20151018_120833 (1)


Indira crowds


The sari Indira Gandhi was wearing when she was assassinated on October 31, 1984

20151018_125307 (1)

The spot where Indira Gandhi was assassinated

A section of the museum is dedicated to Rajiv Gandhi as well. Indira’s younger son Sanjay was her closest political adviser and expected to succeed her. When Sanjay died in a plane crash in 1980, Rajiv was pulled into politics.

He immediately succeeded his mother as Prime Minister in 1984, and led the country until 1989. When campaigning for elections in 1991, Rajiv Gandhi was assassinated by a suicide bomber from the Liberation Tigers of Tamil Eelam, several years after he had sent an Indian peace keeping force to Sri Lanka.

Rajiv Gandhi clothes

The clothes Rajiv Gandhi was wearing when he was assassinated on May 21, 1991

Rajiv’s wife Sonia Gandhi is now the President of the Indian National Congress. Her son Rahul is the Vice President, and led the Congress campaign during the 2014 Lok Sabha elections.