Appendices

Selection of Sample Units and Variables

The study used vacant seats in PRI elections to measure the democratic deficit in electoral participation. The first indicator was the number of contestants for each electoral seat in the villages. This second indicator measured the degree of democratic contest for the Pradhan (village council chief) and village council membership positions. These indicators reflected the level of engagement in local polls, i.e. whether the PRI system achieved the constitutional mandate of filling all of the seats (Pradhan and other members) in each village council.

The proportion of vacant positions and contestants in elections at the village level was used as a proxy to measure the community representatives' participation in the villages' PRI elections. After adding vacant and uncontested positions, the Allahabad district falls near or above the upper quartile (Q3) in all three indicators: vacant, uncontested positions and both (vacant and uncontested combined). The high proportion of empty and uncontested seats indicates that PRIs cannot ensure participation at the village level's first electoral stage (i.e. low involvement). Due to this democratic deficit, the district of Allahabad was chosen for further examination at the level of blocks and villages.

Selection of Blocks, Villages, and Households

The 20 blocks of the Allahabad district were divided into three quartiles based on their performance in the employment guarantee scheme (Mahatma Gandhi National Employment Guarantee Act – MNREGA). The scheme has a mandate of bottom-up planning through village meetings. Hence, the ratio of households (HHs) that applied for a job vs people getting a job was used to proxy participation in open meetings across these blocks. From each quartile, one block was randomly selected. Three blocks were selected out of the three strata of 20 blocks using stratified random sampling. Two villages were then chosen at random from the lists of each block. Hence, a total of six villages were selected for the HH interviews. During the fieldwork, it was found that one village had no meeting, so another village was selected from within the same block, bringing the total count to seven villages. ¹

From these seven villages, I interviewed the head of the HHs from 135 HHs. A diverse pool of HHs was purposively selected from each village, containing HHs supporting the political party of the chief of the village (Pradhan), as well as HHs supporting the opposition leader. Since open meetings and participation in these meetings are the dependent variables in this study, when the information on the meeting was saturated from a village (e.g. 8–10 HHs reported no meeting or shared a similar experience in open meetings), another village was selected. ²

The study recorded gender, caste segregation and elite-brokerage relationships in interviews and village observations, as caste, education and gender are widely discussed in the relevant literature (Chattopadhyay & Duflo, 2004; Sanyal & Rao, 2018). The study reported the community narratives from the sample of seven villages, highlighting their satisfaction and concerns with PRIs. These narratives help explain and contextualize the data collected on vacant positions at the GP level in PRI elections. ³ Questions were posed regarding the frequency of open meetings, HH attendance and meeting participation. The interviews also contained two open-ended questions concerning the participatory local governance mechanisms of PRIs: (i) ‘What are the problems with the Panchayat system?’ and (ii) ‘In your opinion, what should be done to improve the system so that people can participate in the open meetings?’ Respondents were also given the ability to report inactive village councils, no/or low frequency of meetings. I conducted interviews with 135 HHs and translated them into English for this write-up.

Summary of Key Variables

To handle the data and conduct analysis, SPSS and R studio were used. CSS selectors, an add-on in the Chrome web browser and R studio, helped curate the data from the State Election Commission and census websites.

Readers can visit my website and refer to online tutorial for a better understanding of data entry and analysis (chi square test) that I used for my PhD data. ⁴

The field experiences that I have used here was part of my job in civil society organization and academic institutions.

During my Master's (2007–2009), I was extensively engaged in self-help group formation, survey and training of officials for Allahabad district (now Prayagraj), Uttar Pradesh, for the employment guarantee programme (Mahatma Gandhi National Rural Employment Guarantee Act) in India.

As a Research Assistant, I had conducted field visits in five minority concentrated districts of Uttar Pradesh in 2009.

I conducted interviews and focus group discussions with a range of community, viz. affected people, female sex workers and transgenders in Jaipur, Alwar, Jodhpur and Udaipur districts of Rajasthan (2011–2013). I conducted the similar field activity and imparted training to youth and women in the Himalayan region of Himachal Pradesh (2010–2011).

In Bihar (2009–2010), I stayed with the community for various days intermittently while conducting project-related activities during my tenure with Samta Gram Seva Sansthan (Posted through Ministry of Rural Development, Government of India). The field area in Bihar was Sitamarhi district to the extreme north of the state of Bihar bordering Nepal.

Currently, as a Postdoctoral Fellow (2021- till date), I have conducted interviews and coordinated field visits in Singrauli coal field area, Obra thermal power (Sonbhadra District) and that in Kanpur District of Uttar Pradesh as part of energy transition project at Indian Institute of Technology Kanpur.