Caste-Based Homophily and Value Delivery in a Service Context: Evidence from an Emerging Market

Abstract

Delivering value to consumers is a key challenge faced by social enterprises in emerging markets in which local social structures govern access to markets. We investigate how caste, a hierarchical social structure, impacts value delivered to consumers. Using data on 160,963 service provider-consumer transactions in rural India, we theorize that caste-based homophily occurs when the service provider and consumer are in the same caste category, and this results in 58% more value delivered compared to caste-mismatch interactions. However, if a consumer is disadvantaged due to physical disability or cognitive vulnerability, the value delivered by the service providers of the same caste category is reduced by 60%, while the value delivered by service providers in a lower caste category is increased substantially, by more than 6 times. Our findings uncover the nuanced role of caste and highlight pathways for leveraging local social structures to reduce inequalities and foster inclusive value delivery.

Keywords

caste-based homophily disadvantaged consumers rural India service providers value delivery

Introduction

Social enterprises bridge institutional voids to deliver value to consumers (Chatterjee et al., 2025; Hassan et al., 2023; Qureshi et al., 2023). One reason these social enterprises may struggle to achieve robust impact is the local social structures that shape social relations and hence value delivery in such communities. Caste is an illustrative example of this social order since it embeds social inequality into one’s daily life and creates an intentional, visible separation through unequal access to resources and opportunities (Ambedkar, 1918; Bapuji et al., 2024; Kumar et al., 2022). We study how caste category alignment between consumers and service providers in a social enterprise influences value delivery in a service context.

Given that consumers and service providers are embedded in local caste-based connections, caste category alignment (or misalignment) between them plays an important role in shaping the performance of social enterprises. We theorize that when a service provider and consumer are in the same caste category, it may foster homophily due to caste-based similarity, or a tendency of these two individuals to relate and connect better with one another (Byrne, 1971; McPherson et al., 2001). Such caste-based homophily (from hereon “homophily”) can influence exchanges between service providers and consumers, but we do not yet understand how. Hence, our research question: How does caste category homophily between the service provider and consumers influence value delivery?

The context of our study is the Digital Empowerment Foundation (DEF),¹ whose objective is to bring about a digital revolution that can eradicate information poverty and social backwardness in rural India. DEF recruits locals as service providers, equips them with digital technology, and trains them on government welfare programs. The service provider then distributes this information to consumers with the end goal of enrolling them in welfare programs. Value delivery in our context refers to the welfare amount associated with the government programs that is disbursed to the consumers through the service providers. We theorize that the caste category match between service providers and consumers is associated with homophily and increased value delivery. However, if the consumer has a physical or cognitive disadvantage, then caste match dilutes the effect of homophily, and value delivery declines. Instead, a higher caste consumer may derive more value from a lower caste service provider because they are less worried about losing their social status by revealing their physical or cognitive disadvantage to them.

Our study makes two key contributions. First, we unravel why social enterprises seeking to deliver value often struggle to generate robust impact. We demonstrate that prevailing social orders shape value delivery with a caste category match or mismatch between the service provider and consumer as one factor that drives these differences. Second, the existing literature has investigated how caste impacts business transactions in the context of urban entrepreneurs (Vissa, 2011), however, except for a few studies (e.g., A. V. Banerjee & Duflo, 2013; Mair et al., 2016; Patel et al., 2022), the impact of caste has been relatively understudied among rural service providers. Urban and rural contexts are distinct, with rural communities generally exhibiting greater knowledge of the identity of village residents (Albrecht, 2022; Cornwell & Behler, 2015; Wirth, 1938). We use rural India as an illustrative case to study local social structures because villagers tend to readily know the caste of one another.

Theory and Hypotheses

Homophily is the tendency of individuals to be attracted to, connect with, and form relationships with those individuals who share similar attitudes, values, or characteristics (Byrne, 1971; Heider, 1958; Sherif, 1958). Individuals tend to be attracted to individuals like themselves, which facilitates communication, trust, and predictability of behavior (Lincoln & Miller, 1979). Belonging to the same caste builds social capital and creates a trusted social community (Varshney et al., 2022), which lowers perceived risk in exchanges that take place in environments characterized by information asymmetry. Sales and services are intricately linked to local social structures (Anderson & Xue, 2022) through their effect on social relationships. Rural consumers often rely on word-of-mouth from these relationships when adopting products and services (A. Banerjee et al., 2013), reducing hesitation and encouraging continued usage.

Homophily occurs when two individuals have an affinity toward one another due to affiliation in the same caste category, and this has implications for service delivery (Meshram & Venkatraman, 2022). Consumers tend to rely on information from service providers of similar caste (Chen et al., 2015; Damaraju & Makhija, 2018; Vissa, 2011) because of ease of information-sharing, communication, and coordination (Afridi et al., 2020; Bapuji et al., 2024). For example, if both the service provider and consumer belong to the same caste category, the service provider may know cultural norms, dialect cues, and community symbols as well as understand the consumers’ day-to-day challenges. This familiarity can make consumers more open to receiving guidance on government program requirements. It can also facilitate sensitive document preparation and increase trust in service delivery. Hence, we derive the following hypothesis:

Hypothesis (H1): Caste category match between service providers and consumers will be positively associated with value delivered to the consumer.

When individuals belong to different caste categories, caste-based homophily and its advantages related to trust and lower cost are absent. Moreover, due to the stigma associated with being lower on the graded caste hierarchy (Bapuji et al., 2024; Mendonca et al., 2024), when service providers are in a lower caste category, they may feel a sense of apprehension in meeting and connecting with consumers in a higher caste category. At the same time, higher caste consumers may be reluctant to discuss their problems or financial conditions with lower caste service providers.

However, the same service provider may behave differently if they belong to a higher caste category than the consumer. Individuals in a higher caste category tend to be more educated and are assumed to be more competent than their lower caste peers. They tend to score higher on confidence, willingness to compete, and several personality traits such as conscientiousness and openness to experience (Dasgupta et al., 2023). To maintain their higher caste status in the eyes of the community (Jodhka & Newman, 2007), the service providers are likely to work harder to garner information and resources (Bodemann, 1988). Thus, we hypothesize:

Hypothesis (H2): Service providers in a higher caste category relative to their consumers will deliver greater value than service providers in a lower caste category.

Homophily typically facilitates trust and value delivery due to caste-based similarity (Byrne, 1971; Vissa, 2011), but such benefits may erode when the same caste consumers are in a visibly disadvantaged position, disrupting the symmetry associated with homophily. In general, lower caste category individuals are more likely to suppress caste markers, whereas higher caste category individuals tend to highlight visible markers that signal their relative advantage (Thomas, 2020). In particular, disadvantaged consumers may avoid highlighting their temporary or permanent disadvantage to preserve their social position with service providers of the same caste (Cohen & Silver, 1989), resulting in restrained communication or diminished help-seeking behavior (Deshpande & Spears, 2016). Thus, we predict:

Hypothesis (H3a): Disadvantaged position of the consumer moderates the effect of caste category match such that value delivered decreases when the service provider and disadvantaged consumer are from the same caste category.

Disadvantaged higher caste consumers may be more willing to work with and less apprehensive in sharing their challenges with lower caste service providers for several reasons. Although disadvantaged higher caste consumers belong to a higher caste category, they may feel uneasy working with the same or higher caste category service providers due to the loss of visible advantages. However, they may be less worried about losing their status with lower caste service providers, because their higher caste category provides them with a higher status relative to low caste service providers, despite other sources of disadvantage. This may aid more open communication and help-seeking between higher caste disadvantaged consumers and lower caste service providers, increasing the value delivered by the service providers in such situations. Hence, our final hypothesis proposes:

Hypothesis (H3b): Disadvantaged position of the consumer moderates the effect of caste category mismatch such that value delivered increases when service providers are of a lower caste category relative to their disadvantaged consumers.

Method

Data and Sample

Our empirical context is the DEF, a non-profit founded in 2002 to eliminate information poverty and social backwardness in rural India through digital inclusion.² Partnering with corporations such as Qualcomm and the European Union (under their corporate social responsibility and development funding initiatives, respectively), DEF launched the SoochnaPreneur (“Information Entrepreneur”) program to assist local villagers to become service providers, encourage women’s economic participation, and promote empowerment and financial self-sustenance.³ DEF recruits and trains local individuals as service providers, equips them with digital tools such as a laptop, mobile phone, and photocopier, and encourages them to operate as service providers. Their goal is to empower rural communities by helping them access financial support through government welfare programs. The SoochnaPreneur program provides access to information on 505 government welfare programs, including free health insurance, child nutrition, pension, financial assistance for marriage, and scholarships. These services enable rural residents to access financial aid and help DEF achieve its mission of digitally connecting remote villages and eliminating information poverty.

SoochnaPreneurs combine their knowledge of government programs with awareness of local needs to deliver value and reduce poverty. These service providers learn about various government programs, study demographics and needs of the consumers in the villages they cover, make people aware of government programs, enroll consumers into suitable welfare programs, and follow up and troubleshoot until the welfare amount is disbursed to the consumers, thereby delivering value through financial and convenience benefits (Chatterjee et al., 2025). For this, the service providers earn a nominal 200 INR fee per enrollment at the time of registration. DEF strictly prohibits any form of bribery and can terminate the service providers if they are found guilty. DEF’s service providers are known for their integrity, refusing bribes even under pressure, which makes consumers prefer them over government agents, whose informal costs are higher (Chatterjee et al., 2025). This integrity ensures that caste-based effects can be studied without confounding financial motives.

A proprietary mobile application supports data collection and monitoring and serves as our primary data source. Each transaction represents an interaction in which a service provider registers a consumer in an application with the intent of enrolling in welfare programs that provide financial benefits. The transactions include demographic details of both the service provider and consumer—age, caste, disability, district, education, gender, income, marital status, occupation, religion, and vulnerability status—and details of the welfare program. Financial benefits vary widely, from 300 INR for small scholarships to 1,000,000 INR for overseas scholarships. The status of these applications can also vary from approved, in-process, pending, and rejected, though only the ‘approved’ applications are considered for our analysis.⁴

The dataset contains 207,885 transactions between 385 service providers and their consumers in seven backward districts of rural India over a 36-month period (January 2017–December 2019). After removing 5,439 observations without caste data of the transacting parties, 202,446 interactions remain, of which 91,273 exhibit a caste category match, and 111,173 do not. Service providers belong to General, Other Backward Classes, Scheduled Caste, and Scheduled Tribe caste categories, while consumers span additional groups such as Backward Caste, Dalit Tribes, Denotified Tribes, Minority (Religious), Most Backward Class, and others; these additional groups are underrepresented among service providers. In terms of religion, service providers are mainly Hindu (85%), Muslim (13%), or Christian (1%), while consumers include these as well as Buddhist, Jain, and Sikh. Caste-based homophily occurs when both caste categories and religion match between the service provider and the consumer. Since 77% (160,963 of 207,885) of transactions involve Hindu service providers and consumers, and caste divisions are most entrenched within Hinduism (Srinivas, 1962), the analysis focuses on the Hindu subsample with the full sample used for robustness checks. The caste of service providers and consumers in count and percent at the transaction level can be seen in Supplemental Appendix Table A2.

Measures of Variables

Dependent Variable

Value delivered represents the welfare amount associated with each approved transaction under a government program. We define a transaction as an interaction in which a service provider has identified, informed, and ultimately enrolled a consumer in a government welfare program that delivers financial benefits. For programs without direct monetary transfers, such as obtaining a Permanent Account Number or caste certificates, we consider the approximate savings from reduced travel and time costs (e.g., not losing daily wages), which is estimated at INR 200 by DEF, for our analysis. Values are log-transformed (lnX+1) to account for outliers (Lall & Park, 2020).

Independent Variables

Same caste service provider is a binary variable indicating whether the service provider and consumer belong to the same caste category and Hindu religion, coded as 1 if the caste category matches and 0 otherwise.

Higher Caste Service Provider and Lower Caste Service Provider

To test the effect of caste category mismatch, we further classify service providers as lower caste or higher caste categories relative to consumers using binary variables to capture these relationships. To avoid multicollinearity (Gujarati, 2004), only lower caste service providers and higher caste service providers are included (with same caste service providers as the baseline) in Hypothesis 2. The caste hierarchy is derived from prior research and reports (Bapuji & Chrispal, 2020; Government of India, 1980; John et al., 2020; see Supplemental Appendix Tables A1 and A2).

Moderating Variable

Disadvantaged is a binary variable equal to 1 if the consumer has either a physical disability and/or a cognitive vulnerability, and 0 otherwise. This classification of consumers is based on their eligibility for relevant welfare programs. For more details, please refer to Supplemental Appendix A1.

Control Variables

Over and above caste category similarity or mismatch, the inherent caste of the service provider can influence their control on resources and opportunities (Bapuji & Chrispal, 2020). Service providers from different castes may have systematically different access to resources, which could be associated with delivering differential value to consumers regardless of caste category similarity. Therefore, we control for the service provider’s caste (Scheduled Tribe, Scheduled Caste, Other Backward Class, Minority, General; Scheduled Tribe is our baseline). In addition, we control for gender similarity between the service provider and consumer. Additional controls include consumer age, income, gender, marital status, employment, sickness (such as AIDS, cancer, leprosy), and service provider age and gender.

We also include district fixed effects for location-specific variation, and month and year fixed effects to account for time-based correlations (Certo & Semadeni, 2006). To examine the association between caste-based homophily and transaction outcomes, we estimated a pooled OLS regression (Conlon & Gortmaker, 2020; Correia, 2016) with two-way clustered standard errors at the service provider and consumer levels to account for potential within-group correlation in residuals due to potential repeated transactions and non-independence. The Variation Inflation Factors (VIFs) are below 3, indicating multicollinearity is not a concern.

Robustness Checks

As a robustness check, we run our analyses over the entire sample (results shown in Supplemental Appendix Table A3). We perform the Hausman-Durbin test (Nakamura & Nakamura, 1981), using two-stage least squares regression with instrumental variables of Education of the consumer and District location literacy rate. These variables are likely to impact value delivered (our dependent variable) through caste category (our independent variable) rather than directly, since education may reduce perceptions of caste-based divide. We did not find support for the alternative hypothesis that the OLS estimators are affected by endogeneity among regressors (p = .1790, χ²(1) = 1.8062; p = .1796, F(1, 28938) = 1.80113). Further, in the Supplemental Appendix Table A4 shows the influence of only higher caste service providers and only lower caste service providers (in comparison to others) on value delivered, and Supplemental Table A5 shows results after controlling for the type of welfare program, such as amenities, cash transfer, convenience, education, employment, health, women and children’s welfare, among others.

Results

Table 1 shows the descriptive statistics and correlations. The average value delivered per transaction for a consumer is 67,022 INR. Table 2 has the main regression results, with Model 1 showing the results for the control variables only. Model 2 shows the results for H1, which predicts that a caste category match increases the value delivered to consumers. It shows that the coefficient of Same caste service provider is positive and significant (p = .016, β = .456, e^β−1 = .577), indicating that the presence of homophily results in a 58% increase in value delivered (38,872 INR more value delivered). Model 3 examines the effect of caste category mismatch on value delivered (H2), using the same caste service provider as the baseline. Though both perform worse than the same caste service providers, higher caste service providers deliver more value than lower caste service providers. Results show that both higher caste service providers (p = .047, β = −.407, e^β−1 = −.334) and lower caste service providers (p = .014, β = −.579, e^β−1 = −.439) have negative and significant coefficients. A Wald test (F = 3.56, p = .0302) confirms these coefficients differ significantly (Greene, 2018). Lower caste service providers deliver 11% less value (−44% vs. −33%) than higher caste service providers, though both underperform relative to same caste service providers.

Table 1.

Descriptive Statistics and Correlation among Variables.

Variable	Mean	SD	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)	(12)	(13)	(14)
(1) Value delivered	6.932	3.691
(2) Same caste service provider	0.531	0.498	–.055
(3) Scheduled caste	0.167	0.373	–.186	.012
(4) Scheduled tribe	0.123	0.329	.0237	.085	–.119
(5) Other backward class	0.535	0.498	–.131	.296	–.441	–.318
(6) Disadvantaged	0.038	0.191	–.030	.007	.022	.003	.042
(7) Gender homophily	0.557	0.496	–.033	.017	.028	–.024	–.010	–.051
(8) Consumer age	32.778	50.234	–.015	.107	.008	.008	.118	–.001	.043
(9) Annual income	8.78e+07	1.31e+10	.007	–.009	–.003	–.002	.001	–.001	–.010	–.000
(10) Marital status	0.644	0.478	–.079	.097	.064	–.009	.132	.051	.032	.518	–.002
(11) Employment status	0.101	0.302	–.048	.134	–.073	–.059	.218	.005	.051	.117	.014	.134
(12) Female consumer	0.473	0.499	–.016	–.018	.033	.021	–.035	.046	–.036	–.065	.009	–.026	–.076
(13) Sickness	0.121	0.326	–.102	–.031	.062	–.033	–.136	.040	.071	.074	–.004	.041	–.110	–.070
(14) Service provider age	28.802	6.575	–.041	–.105	.167	–.152	–.302	–.002	.045	–.016	.005	–.012	–.067	–.023	.223
(15) Female service provider	0.533	0.498	–.060	–.051	.079	–.154	.019	–.076	–.082	–.072	–.008	–.054	–.069	.146	–.254	–.156

Note. Variables for relative caste categories are not included. n = 160,963 interactions. SD = standard deviation.

Table 2.

Effects of Caste-Based Homophily and Relative Caste Category on Value Delivered.

Variables	Model 1, only controls	Model 2, hypothesis 1	Model 3, hypothesis 2	Model 4, hypothesis 3a	Model 5, hypothesis 3b
Same caste service provider		0.456* (0.187)		0.489* (0.191)
Lower caste service provider			–0.579* (0.234)		–0.643** (0.240)
Higher caste service provider			–0.407* (0.203)		–0.428* (0.207)
Disadvantaged	–0.163 (0.416)	–0.113 (0.422)	–0.112 (0.423)	0.398 (0.451)	–0.534 (0.477)
Same caste service provider × dDisadvantaged				–0.932* (0.413)
Lower caste service provider × disadvantaged					1.871** (0.678)
Higher caste service provider × disadvantaged					0.641 (0.454)
Scheduled caste	–1.124 (0.842)	–1.060 (0.813)	–1.081 (0.812)	–1.059 (0.812)	–1.084 (0.810)
Other backward class	–0.735 (0.696)	–0.751 (0.676)	–0.812 (0.685)	–0.763 (0.675)	–0.827 (0.682)
General	1.961⁺ (1.024)	2.184* (1.042)	2.088* (1.048)	2.188* (1.041)	2.081* (1.046)
Gender homophily	–0.199 (0.159)	–0.204 (0.158)	–0.203 (0.158)	–0.203 (0.158)	–0.203 (0.158)
Consumer age	0.0124** (0.004)	0.012** (0.004)	0.012** (0.004)	0.012** (0.004)	0.012** (0.004)
Annual income	2.79e-12** (8.57e-13)	3.06e-12*** (7.74e-13)	3.03e-12*** (7.81e-13)	3.09e-12*** (7.71e-13)	3.04e-12*** (7.82e-13)
Marital status	–0.381 (0.253)	–0.370 (0.251)	–0.375 (0.251)	–0.370 (0.251)	–0.374 (0.251)
Employment status	–0.424 (0.386)	–0.469 (0.382)	–0.466 (0.383)	–0.474 (0.382)	–0.473 (0.382)
Female consumer	–0.026 (0.145)	–0.023 (0.143)	–0.025 (0.144)	–0.024 (0.143)	–0.025 (0.144)
Sickness	–1.512* (0.646)	–1.542* (0.638)	–1.535* (0.638)	–1.549* (0.638)	–1.543* (0.638)
Service provider age	–0.057 (0.041)	–0.059 (0.041)	–0.059 (0.040)	–0.059 (0.040)	–0.059 (0.040)
Female service provider	–0.675 (0.560)	–0.670 (0.554)	–0.671 (0.554)	–0.671 (0.554)	–0.676 (0.553)
District fixed effect	Yes	Yes	Yes	Yes	Yes
Year and month FE	Yes	Yes	Yes	Yes	Yes
Constant	9.521*** (1.329)	9.207*** (1.272)	9.699*** (1.323)	9.202*** (1.269)	9.733*** (1.322)
Sample size	67,561	67,561	67,561	67,561	67,561
Adjusted R²	.1715	.175	.175	.175	.175
Akaike information criterion	355509.6	355265.6	355258.0	355232.3	355206.2
Bayesian information criterion	355810.6	355575.7	355577.2	355551.5	355543.7
Log likelihood	–184094.4	–184094.4	–184094.4	–184094.4	–184094.4

Note. All tests are two tailed. SE values in parentheses.

p < .1. *p < .05. **p < .01. ***p < .001.

Models 4 and 5 test the moderating role of the disadvantaged consumer. Hypothesis 3a predicts a negative interaction between a same caste service provider and the disadvantaged position of the consumer. We find support for this hypothesis with a 60% reduction (40,615 INR reduction) in value delivered with reference to caste category match, with disadvantaged consumers (p < .025, β = −.932, e^β−1 = −.606). Hypothesis 3b predicts positive interaction effects for lower caste service providers and disadvantaged consumers, which we find support for (p < .006, β = 1.871, e^β−1 = 5.494), and the impact of higher caste service providers was not significant. Thus, disadvantaged consumers receive greater value from lower caste service providers (amounts to INR 368,219 more in comparison to service providers of the same caste category), supporting Hypothesis 3b. Robustness checks across religions yield similar results for our hypotheses (not shown), demonstrating how lower caste service providers deliver greater value to disadvantaged consumers in comparison to service providers of a higher or same caste category. Figure 1 shows the interaction effect of disadvantaged consumers on caste category match and mismatch.

Figure 1.

Interaction effect of disadvantaged consumer with caste category match and mismatch. (a) Moderation effect with the same caste service providers. (b) Moderation effect with relatively lower caste service providers.

Overall, our analysis finds that a caste category match between service providers and consumers is associated with 58% higher value delivery. Under the caste category mismatch, higher caste category service providers deliver 11% more value compared to lower caste category service providers. For disadvantaged consumers, however, caste category match reduces value delivery by 60% when compared with non-matched caste categories between service providers and consumers; but when service providers are of a lower caste category relative to disadvantaged consumers, value delivery increases by more than 500% in comparison to caste category match between service providers and disadvantaged consumers.

Discussion

Our study contributes to social enterprise performance research by examining how caste category matches and mismatches impact value delivery. We extend longstanding arguments that social similarity facilitates trust, communication, and transaction efficiency (Byrne, 1971). Furthermore, when service providers in a lower caste category assist disadvantaged consumers in a higher caste category, they deliver greater value.

Our study has important implications for poverty reduction efforts in India, where the caste system affects the entire population of 1.4 billion people (Chrispal et al., 2021). Governments and Non-Governmental Organizations (NGOs) should be cognizant of caste effects to ensure equality and deliver greater value to consumers. Since disadvantaged consumers work best with service providers from lower caste categories, programs that support these providers with training and resources can help deliver greater value. Sensitization training should also be provided to high caste category providers on serving disadvantaged individuals and those who are lower in the caste hierarchy (to improve it to the level of the same caste category).

While our study examines caste-based dynamics in India, our findings have broader implications for service delivery contexts characterized by social hierarchies and power asymmetries where providers and clients occupy unequal social positions. For example, in healthcare, patient–provider social distance or in financial services, loan officers’ social proximity or distance from borrowers can shape communication quality, trust formation, effort allocation, or perceived credibility. In both examples, mechanisms similar to caste-based interactions may exist, affecting value delivery. Thus, our findings can extend beyond caste to explain how social category alignment and status asymmetry can influence value delivery in frontline service delivery settings.

It is important to point out some limitations of this research. First, the sample is drawn from a single non-profit organization, which may limit the generalizability of our findings. Future studies could benefit from conducting experiments and examining data from multiple organizations and diverse welfare programs. Second, our study focuses on successful applications of caste-based value delivery in a service context. Investigating both successful and unsuccessful transactions would provide nuanced insights into the conditions under which homophily matters or does not matter. Third, while our analysis is primarily quantitative, future studies could leverage qualitative interviews in addition to quantitative analysis to gain deeper insights into caste in rural communities as well as examine potential district-level differences.

Governments of emerging countries such as India spend billions on welfare programs, yet implementation faces challenges, including information gaps and resource mobilization difficulties (A. V. Banerjee & Duflo, 2013). Mair and colleagues (2016) identify scaffolding as a process to transform behavior and tackle embedded societal inequality. However, services designed to promote inclusion frequently fall short due to underlying structural and social biases (Prakash, 2020), with caste-based societies particularly constrained by entrenched perceptions of power and hierarchy (Fisk et al., 2018; Meshram & Venkatraman, 2022). Our study demonstrates that not only does caste category match increase value delivery due to homophily, but lower caste category service providers also enhance value delivery to disadvantaged higher caste category consumers, which serves as a locally grounded lever for tackling caste-based social inequality. Understanding these dynamics can improve resource mobilization and enhance the effectiveness of welfare programs.

Supplemental Material

sj-docx-1-bas-10.1177_00076503261465114 – Supplemental material for Caste-Based Homophily and Value Delivery in a Service Context: Evidence from an Emerging Market

Supplemental material, sj-docx-1-bas-10.1177_00076503261465114 for Caste-Based Homophily and Value Delivery in a Service Context: Evidence from an Emerging Market by Aindrila Chatterjee, Amit J. Chauradia and Kiran Pedada in Business & Society

Footnotes

Acknowledgements

We thank the participants at the Third Global Conference on Caste, Business and Society held at the University of Oxford in 2025 for their valuable comments. We thank DEF for access to the data and data-related support from Amirullah Khan, Osama Manzar, and Arpana Sharma. We would like to thank the editors and reviewers for their detailed feedback and comments at each stage of the review process to improve the article.

ORCID iDs

Aindrila Chatterjee

Amit J. Chauradia

Kiran Pedada

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biographies

Aindrila Chatterjee (FPM, Indian School of Business; PhD, Jadavpur University) serves as a faculty member at the Institute of Management Technology, Hyderabad in India, where she teaches, mentors, and contributes to shaping future business leaders. Her research interests span strategy, organizational behavior, and contemporary social issues in management such as ethics and sustainability. With a focus on the intersection of strategy and social enterprises, she is motivated to study value creation through social impact in resource-constrained environments in emerging economies. She has published scholarly work in reputed academic journals like Acta Psychologica, Human Resource Development International, and Journal of the Academy of Marketing Science, and presented at international forums, earning recognition such as best paper awards at national and international business schools.

Amit J. Chauradia (PhD, University of Illinois) is an assistant professor of Instruction at the University of South Florida, Muma College of Business. His research studies the role of strategic human capital and social enterprises in creating value and improving performance. His articles have been published in journals such as Human Resource Development International, Journal of General Management, Journal of International Management, and Journal of the Academy of Marketing Science.

Kiran Pedada (PhD, Texas Tech University) is the associates fellow in Marketing and Inclusive Business and Associate Professor of Marketing at the University of Manitoba, Asper School of Business. His research focuses on marketing strategies for emerging markets and digital environments. His articles have appeared in such journals as the European Journal of Marketing, International Journal of Research in Marketing, Journal of Business Research, Journal of Marketing Research, and Journal of the Academy of Marketing Science.

References

Afridi

Dhillon

S. X.

Sharma

(2020). Using social connections and financial incentives to solve coordination failure: A quasi-field experiment in India’s manufacturing sector. Journal of Development Economics, 144, 1024–1045.

Albrecht

D. E.

(2022). Rural/urban differences: Persistence or decline. Rural Sociology, 87(4), 1137–1154.

Ambedkar

R. B.

(1918). Castes in India–Indian antiquary, May 1917. Journal of the Royal Asiatic Society, 50(3–4), 622–624.

Anderson

Xue

(2022), Transformative service research: Where we are and moving forward at the collective level. In Edvardsson

Tronvoll

(Eds.), The Palgrave handbook of service management (pp. 437–455). Springer International Publishing.

Banerjee

Chandrasekhar

A. G.

Duflo

Jackson

M. O.

(2013). The diffusion of microfinance. Science, 341(6144), 1236498.

Banerjee

A. V.

Duflo

(2013). Poor economics: Rethinking Poverty and the ways to end it. Random House.

Bapuji

Chrispal

(2020). Understanding economic inequality through the lens of caste. Journal of Business Ethics, 162, 533–551.

Bapuji

Chrispal

Attri

P. S.

Ertug

Soundararajan

(2024). An integrative review of management research on caste: Broadening our horizons. Academy of Management Annals, 18(2), 506–549.

Bodemann

Y. M.

(1988). Relations of production and class rule: The hidden basis of patron-clientage. In Wellman

Berkowitz

S. D.

(Eds.), Social structures: A network approach (pp. 198–220). Cambridge University Press.

10.

Byrne

(1971). The attraction paradigm (personality and psychopathology, 11) (1st ed.). Academic Press.

11.

Certo

S. T.

Semadeni

(2006). Strategy research and panel data: Evidence and implications. Journal of Management, 32(3), 449–471.

12.

Chatterjee

Chauradia

A. J.

Pedada

(2025). Rural women microentrepreneurs, consumer acquisition, and value delivery: Evidence from a quasi-experiment in rural India. Journal of the Academy of Marketing Science, 53, 213–231.

13.

Chen

Chittoor

Vissa

(2015). Modernizing without westernizing: Social structure and economic action in the Indian financial sector. Academy of Management Journal, 58(2), 511–537.

14.

Chrispal

Bapuji

Zietsma

(2021). Caste and organization studies: Our silence makes us complicit. Organization Studies, 42(9), 1501–1515.

15.

Cohen

B. P.

Silver

S. D.

(1989). Group structure and information exchange: Introduction to a theory. In Berger

Zelditch

Anderson

(Eds.), Sociological theories in progress: New formulations (pp. 160–181). Sage.

16.

Conlon

Gortmaker

(2020). Best practices for differentiated products demand estimation with pyblp. The RAND Journal of Economics, 51(4), 1108–1161.

17.

Cornwell

E. Y.

Behler

R. L.

(2015). Urbanism, neighborhood context, and social networks. City & Community, 14(3), 311–335.

18.

Correia

(2016). A feasible estimator for linear models with multi-way fixed effects. http://scorreia.com/research/hdfe.pdf

19.

Damaraju

N. L.

Makhija

A. K.

(2018). The role of social proximity in professional CEO appointments: Evidence from caste/religion-based hiring of CEOs in India. Strategic Management Journal, 39(7), 2051–2074.

20.

Dasgupta

Mani

Sharma

Singhal

. (2023). Social identity, behavior, and personality: Evidence from India. The Journal of Development Studies, 59(4), 472–489.

21.

Deshpande

Spears

(2016). Who Is the identifiable victim? Caste and charitable giving in modern India. Economic Development and Cultural Change, 64(2), 299–321.

22.

Fisk

R. P.

Dean

A. M.

Alkire

Joubert

Previte

Robertson

Rosenbaum

M. S.

(2018). Design for service inclusion: Creating inclusive service systems by 2050. Journal of Service Management, 29(5), 834–858.

23.

Government of India. (1980). Report of the Backward Classes Commission (Mandal Commission), Part I & II. Ministry of Home Affairs, Government of India.

24.

Greene

W. H.

(2018). Econometric analysis (8th ed.). Pearson Education.

25.

Gujarati

D. N.

(2004). Basic econometrics (4th ed.). Tata McGraw Hill.

26.

Hassan

Prabhu

Chandy

Narasimhan

(2023). When bulldozers loom: Informal property rights and marketing practice innovation among emerging market microentrepreneurs. Marketing Science, 42(1), 137–165.

27.

Heider

(1958). Psychology of interpersonal relations (99th ed.). John Wiley & Sons.

28.

Jodhka

S. S.

Newman

. (2007). In the name of globalisation: Meritocracy, productivity and the hidden language of caste. Economic and Political Weekly, 42(41), 4125–4132.

29.

John

R. K.

Xavier

Waldmeier

Meyer

A. H.

Gaab

(2020). The governmental ranking of class and the academic performance of Indian adolescents. PLoS One, 15(11), Article e0241483.

30.

Kumar

Bapuji

Mir

(2022). Educate, agitate, organize: Inequality and ethics in the writings of Dr. B. R. Ambedkar. Journal of Business Ethics, 178(1), 165–179.

31.

Lall

S. A.

Park

(2020). How social ventures grow: Understanding the role of philanthropic grants in scaling social entrepreneurship. Business & Society, 61(1), 3–44.

32.

Lincoln

J. R.

Miller

(1979). Work and friendship ties in organizations: A comparative analysis of relation networks. Administrative Science Quarterly, 24(2), 181–199.

33.

Mair

Wolf

Seelos

(2016). Scaffolding: A process of transforming patterns of inequality in small-scale societies. Academy of Management Journal, 59(6), 2021–2044.

34.

McPherson

Smith-Lovin

Cook

J. M.

(2001). Birds of a feather: Homophily in social networks. Annual Review of Sociology, 27(1), 415–444.

35.

Mendonca

D’Cruz

Noronha

(2024). Identity work at the intersection of dirty work, caste, and precarity: How Indian cleaners negotiate stigma. Organization, 31(1), 3–26.

36.

Meshram

Venkatraman

(2022). A transformative service research perspective on caste-based discrimination in microcredit lending in India. Journal of Services Marketing, 36(7), 964–976.

37.

Nakamura

(1981). On the relationships among several specification error tests presented by Durbin, Wu, and Hausman. Econometrica, 49(6), 1583–1588.

38.

Patel

P. C.

Lenka

Parida

(2022). Caste-based discrimination, microfinance credit scores, and microfinance loan approvals among females in India. Business & Society, 61(2), 372–388.

39.

Prakash

. (2020). Dalit capital: State, markets and civil society in urban India. Routledge India.

40.

Qureshi

Bhatt

Sutter

Shukla

D. M.

(2023). Social entrepreneurship and intersectionality: Mitigating extreme exclusion. Journal of Business Venturing, 38(2), 106283.

41.

Sherif

(1958). Superordinate goals in the reduction of intergroup conflict. American Journal of Sociology, 63(4), 349–356.

42.

Srinivas

M. N.

(1962). Caste in modern India and other essays. Oxford University Press.

43.

Thomas

(2020). Brahmins as scientists and science as Brahmins’ calling: Caste in an Indian scientific research institute. Public Understanding of Science, 29(3), 306–318.

44.

Varshney

Mishra

A. K.

Joshi

P. K.

Roy

(2022). Social networks, heterogeneity, and adoption of technologies: Evidence from India. Food Policy, 112, 102360.

45.

Vissa

(2011). A matching theory of entrepreneurs’ tie formation intentions and initiation of economic exchange. Academy of Management Journal, 54(1), 137–158.

46.

Wirth

(1938). Urbanism as a way of life. American Journal of Sociology, 44(1), 1–24.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.06 MB