By Anjana K S

Introduction

India’s democracy is vast, heterogeneous and largely shaped by regional differences such as language, political cultures and histories. Political participation, experience, trust levels and general perceptions vary largely across regions, shaped by caste, economic structures and other factors such as media and the general cultural structures. We understand India’s system to be an example of democratic backsliding (how contemporary democracies often erode gradually rather than collapse abruptly). Researchers such as Levitsky and Ziblatt (2018) and Bermeo (2016) argue that modern democracies decline through subtle mechanisms such as weakening institutional autonomy, constraining media freedom, delegitimizing opposition, and normalizing dominance. An important outcome of this process is that democratic erosion reshapes citizen behavior and not just of the states. Citizens can continue voting while simultaneously losing trust, gaining fear or eventually withdrawing oneself from public expressions. Another major theme concerns the caste and social hierarchies. Marginalized groups are often the ones that experience biases and vulnerabilities when it comes to democracy and retaliation (Jodhka 2018; Still 2023). These inequalities/hierarchies influence not only participation but also willingness to be open about political preferences.

This report presents a three-way comparative analysis of political attitudes and experiences across Hindi-speaking regions, the Rest of India, and a third significant category of respondents who chose not to disclose their geographic location. We do this analysis using the frequency-percentage tables derived from the Questionnaire interviews carried out across India. Importantly, the latter group contributes the largest share of the sample, with 261 respondents, compared to 139 from the Hindi Heartland and 100 from the Rest of India. Hence, rather than treating this third group as a limitation, the report treats non-disclosure itself as an important analytical aspect.

Through the frequency tables, we examine how the respondents from two linguistic categories, along with the non-respondents, perceive/understand their political environment. We try to understand the percentages from different aspects such as political behavior and turnout, media presence, welfare, voter-government relationship and so on. This comparative analysis is structured around key themes of political experience:

• Political participation and Voting Behavior
• Motivations for voting
• Welfare Access and Governance
• Trust in Democratic Institutions
• Political outcomes and satisfaction
• Media exposure and Perceptions
• Civic Freedom and Safety

We start by understanding the differences in the socio-political context between the Hindi heartland and the rest of India. The Hindi belt comprises much of the North and Central India, which is mostly characterized by strong political mobilization, electoral participation, caste hierarchies and exposure to government welfare schemes. On the other hand, the rest of India is more linguistically and politically heterogeneous, characterized by region-specific parties, unique welfare schemes, diverse and varying media exposure.

An important theme we are trying to understand and address here is the significance and the implications of non-response in political surveys such as this. Non-response is often treated as some sort of nuisance that needs to be imputed/ignored in surveys. But in political surveys conducted on perceptions of democracy, political polarization, declining trust and such, non-response becomes a very important phenomenon, which reveals a lot about the existing systems. They reflect deeper issues such as fear, caution, disengagement or even strategic silence. As researchers of democratic erosion argue, declining democracy is often visible not only in institutions but also in how citizens choose to speak or remain silent (Kuran 1995; Bermeo 2016).

Data and Methodology

The analysis is done on post-poll survey data collected across India following the 2024 Lok Sabha elections. Field investigators conducted structured questionnaire interviews covering a wide range of political attitudes and experiences. After data cleaning, the final sample consists of 500 respondents. The composition of the respondents were as follows:

• Hindi heartland: N= 139
• Rest of India: N= 100
• No-response on location: N= 261

We calculate the frequency of each answer for each question, convert it into percentage for normalization and analyse the trends. The analysis is descriptive in nature, not inferential. Importantly, responses such as “cannot say” and “no response” are retained and analyzed rather than being dropped. This allows us to examine not only what the respondents say, but also where they hesitate or remain silent.

Political participation and Voting Behavior (Q1-Q4)

Reported Voting (Q1)

We see clear regional differences in reported electoral participation:

• Hindi Heartland: Approximately 84.35% report having voted.
• Rest of India: Around 52.57% report voting.
• No-response group: Only 49.40% report voting, with a substantial share no/do not want to answer categories

The high participation/reporting in the Hindi heartland reflects the characteristics that we had discussed previously. However, it could also be possible that the numbers are inflated due to social desirability (Karp & Brockington, 2005). Participation is lower among respondents from the rest of India and lowest among those who did not disclose their location.

Party disclosure

We also see a significant variation in reluctance to disclose party preference:

• Hindi Heartland: 36.70% select non-disclosure categories
• Rest of India: Non-disclosure stands at 26.01%
• No-response group: Non-disclosure rises to 43.41%

The non-response group’s high reluctance reinforces their strong preference for anonymity, showing deep political insecurity or distrust.

Motivations for voting

The survey asked for reasons for voting and we see that there isn’t a significant variation between Hindi Heartland and the rest of India. But, among the non-respondents, the responses vary a lot.

• Hindi Heartland:
  • Development-related motivations account for 13.67%.
  • Leadership and welfare motivations together account for roughly 15–18%.
• Rest of India:
  • Development motivates 14.46%.
  • Responses are more evenly distributed across categories.
• No-response group:
  • A striking 33.39% fall into ambiguous or non-committal categories.

This shows that while the Hindi Heartland reflects a more nationalized narrative of development and leadership, the non-response group avoids conveying clear political motivations.

Welfare Access and Governance

We asked a series of questions regarding the awareness of different welfare programs by the central government. The responses were as follows:

• Hindi Heartland: Awareness of welfare schemes reaches as high as 70.50%.
• Rest of India: Awareness is more moderate and evenly spread, generally between 52–60%.
• No-response group: Awareness drops sharply, with some schemes below 30%.

These differences align with the literature arguing that central government welfare schemes have been particularly effective in the Hindi belt, while non-Hindi states often rely on strong state-level welfare identities. Lower awareness among non-respondents may reflect marginalization, limited state contact, or political disengagement.

Trust in Democratic Institutions

When asked about free and fair elections,

• Hindi Heartland: 38.12% express conditional trust, saying there were still some discrepancies. Approximately 44% were sure that the elections were conducted in a free and fair manner.
• Rest of India: 42.26% expressed that they believed it was fair. The other responses were evenly distributed across categories “Not at all” and “Somewhat, there were still some discrepancies” and ~6% of the sample chose “Cannot say” and “No response”.
• No-response group: Only 17.79% express trust, with the majority choosing neutral or ambiguous categories. A striking 39% chose “cannot say” and ~25% chose not to respond altogether.

When asked about trust in EVMs (Electronic Voting Machines),

• Hindi Heartland: >50% express trust
• Rest of India: Trust is higher at ~70%.
• No-response group: Trust falls to ~20%, with over half choosing neutral options or no response

The consistent pattern is that institutional trust is lowest among non-respondents, moderate in the Hindi Heartland, and highest in the Rest of India. A large share of non-respondents chooses neutral options, saying they cannot judge, or choosing not to answer altogether. This suggests that institutional distrust may coexist with uncertainty and fear of expressing criticism.

Political outcomes and satisfaction

When asked if the respondents were satisfied with the election outcomes:

• Hindi Heartland: 33.52% satisfied, but ~44% were only somewhat satisfied
• Rest of India: 34% satisfied and ~60% were not satisfied or partially satisfied
• No-response group: Only ~12% were satisfied and >50% chose not to respond

When asked if they were happy that there is a proper opposition in the parliament after a long while,

• Hindi Heartland: ~23% gave a positive response and over 50% chose not to respond.
• Rest of India: ~20% gave a positive response and again, over 50% chose not to respond.
• No-response group: ~4% gave a positive response, ~6% negative responses and >80% chose not to respond.

Once again, the non-response group avoids giving out strong opinions, reinforcing the pattern of political ambivalence.

Media exposure and Perceptions

When asked if political parties connecting over whatsapp or over call would make a difference in reaching out to the voter base,

• Hindi Heartland: ~52% report WhatsApp/calls as influential, ~30% said they were not and the remaining chose not to respond
• Rest of India: ~60% said yes and ~20% said no and ~20 didn’t have an idea
• No-response group: ~20% said yes, ~7% said not, ~12% did not have an idea and around 60% chose not to respond.

When asked whether media was free and fair in its reporting,

• Hindi Heartland: ~18% perceive the media as fair, ~36% don’t and the rest are either not sure or chose not to respond.
• Rest of India: 10% perceive the media as fair, 50% don’t and the rest are not sure chose not to respond.
• No-response group: Only 3% perceive media as fair, 9% don’t and the rest significantly large population chose not to respond.

Perceptions of media fairness are low across all groups, but non-response dominates among the No-response (location-wise) group. The non-response group is not only less exposed to campaign messaging but also significantly more distrustful of media.

Civic Freedom and Safety

When asked about whether they feel that their civil liberties and rights are safe and that whether they are able to voice their opinions without any fear,

• Hindi Heartland: ~70% feel fully or at least partially safe, while 20% don’t feel safe at all.
• Rest of India: 57-60% feel fully or partially safe, while 34-38% don’t feel safe at all.
• No-response group: Only 13-15% feel safe, ~15% don’t feel safe and ~70% chose not to respond.

A majority of respondents from the Hindi-speaking heartland and the rest of India report feeling fully or partially safe expressing political opinions. Among respondents who did not disclose their location, only a small minority reported feeling safe. Most either say they do not feel safe or decline to answer the question. This stark difference suggests that non-disclosure is closely linked to fear and perceived risk in political expression.

Overview of Non-Response Patterns in the data

Questions with High Non-Response (≥30–70%)

Question	General Theme	Approx. No-Response %
Q2	“Who did you vote for?” (party choice)	~31%
Q3	Why did you vote?	~38%
Q13–Q16	Trust in institutions (ECI, EVMs, process fairness)	mostly >50%
Q28	Feeling safe to express political views	>40%
Q29	Fear of consequences when criticizing govt	~40–55%
Q30–Q32	Perceptions of media bias, civil liberties	>50%

These questions correspond to sensitive political content such as judgments about the government, perceived fairness of elections, and self-assessments of political safety.

Questions with Low Non-Response (≤5–10%)

These include factual, descriptive, or low-risk questions:

• Q1: Did you vote?
• Q7 series: Awareness of schemes
• Q9: Whether the election was peaceful in the area
• Q12: Whether campaigning reached them

These are socially safe questions with lesser political risk.

Conclusion

The comparative analysis done in this report gives us a very important insight. It is that democratic backsliding in India must be understood not only through institutional changes but also through how citizens experience, negotiate, and respond to the political environment in their everyday lives. The consistently high levels of non-response on politically sensitive questions, particularly those relating to party preference, institutional trust, media fairness, and civic safety suggest that silence itself has become a meaningful political response. For a lot of respondents, especially those unwilling to disclose their location, non-response appears to be a strategy of self-protection in situations of political polarization, declining trust, and perceived risks associated with open expression. This finding reinforces the literature on democratic erosion, which emphasizes that fear, self-censorship, and withdrawal are some of the important outcomes of weakening democratic systems.

Moreover, the fact that the non-response group constitutes the largest segment of the sample becomes an important issue to be addressed. Rather than treating non-response as a methodological inconvenience, this study shows that it can give us valuable insights into the health of a democracy. Overall, the findings suggest that while electoral processes and political participation continue in India, the quality of engagement is very much uneven. Trust in institutions, confidence in media, and the freedom to express dissent are fragile, particularly among those who feel most vulnerable. Recognising these gaps and addressing them are crucial for the country to improve its political systems and thereby the welfare of its people.

References

Bermeo, N. (2016). On Democratic Backsliding. Journal of Democracy, 27(1), 5–19. https://doi.org/10.1353/jod.2016.0012

Caste in contemporary India. (n.d.). Routledge & CRC Press. https://www.routledge.com/Caste-in-Contemporary-India/Jodhka/p/book/97811385729

Karp, J. A., & Brockington, D. (2005). Social desirability and response validity: A comparative analysis of overreporting voter turnout in five countries. The Journal of Politics, 67(3), 825–840. https://doi.org/10.1111/j.1468-2508.2005.00341.x

Kuran, T. (1995). Private truths, public lies: The social consequences of preference falsification.   Harvard University Press.

Levitsky, S., & Ziblatt, D. (2024). How democracies die. In Ideals and Ideologies (pp. 73-80). Routledge.

Still, C. (2023). Dalit Women: Honour and Patriarchy in South India. Routledge.

Anjana K. S. is a master’s student in Economics at Azim Premji University, Bangalore. Her research interests lie at the intersection of environmental and development economics and political economy, with a particular focus on low- and middle-income countries.