Caste and ethnic identity influences on agricultural extension service access and participation in rural communities

Introduction

An observation from the field prompted this research: in the same district, two villages separated by less than eight kilometers showed extension contact rates of 74% and 19%. The high-contact village was predominantly Indo-Mauritian Hindu; the low-contact village was predominantly Creole. The extension office served both, the roads were equally passable, and the farming systems were similar. Something other than geography or agronomy was sorting farmers into those who received extension services and those who did not [1] . Mauritius presents an unusually clear setting for examining ethnic influences on extension access. The island's population comprises distinct ethnic communities-Indo-Mauritian Hindu, Muslim, Creole (of African and Malagasy descent), Sino-Mauritian, and Franco-Mauritian-each with different historical relationships to land, agriculture, and state institutions [2] . Unlike countries where ethnicity and geography are confounded, Mauritian communities often live in close proximity, allowing comparison of extension access within shared agronomic contexts. International evidence consistently shows that ethnicity and caste influence extension access, but most research comes from South Asia and Sub-Saharan Africa [3,4] . In India, caste operates through social networks that channel information along group lines; in Kenya, ethnic patronage shapes which communities receive government services first [5] . Mauritius offers a different institutional context-a small island state with strong bureaucratic institutions, universal education, and no formal caste hierarchy-yet ethnic disparities in agricultural service delivery persist. This research measured the magnitude of these disparities and tested whether structural, cultural, or institutional factors carried the most explanatory weight.

Field experimental design

A cross-sectional survey design with purposive community sampling and random household selection was employed. Eighteen communities across Moka, Flacq, and Savanne districts were selected to ensure representation of all five major ethnic groups within each district. Within each community, households were randomly sampled proportional to ethnic composition, yielding 394 farming households. The Extension Access Score (EAS) combined three sub-indices: proximity to extension services (weighted 0.3), information receipt frequency (0.4), and service availability matching to household needs (0.3), each scored 0-100. Active Participation Rate measured the proportion of available extension events (field days, training sessions, demonstrations, advisory visits) attended over the 24-month period. Barrier assessment used a structured instrument measuring 12 barriers across structural, cultural, and institutional dimensions on a four-point severity scale. www.extensionjournal.com

Material and methods

Material

Data were collected from 394 rural farming households across three districts between June 2021 and May 2023. Ethical approval was obtained from the Research Ethics Committee of the Open University of Mauritius (Approval OUM-REC/2021-031, dated 28 May 2021). Households were eligible if at least one member engaged in agricultural production (crop, livestock, or fisheries) as a primary or secondary occupation and the household had been resident in the community for at least three years. Written informed consent was obtained in English, French, or Creole according to participant preference. Enumerators were matched to community ethnic composition where possible to reduce response bias. Data collection included household surveys, extension service records from district agricultural offices, and community key informant interviews (n = 36).

Methods

Analysis proceeded in three stages. First, one-way ANOVA compared EAS scores and participation rates across the five ethnic groups, with post-hoc Tukey HSD for pairwise comparisons. Second, Pearson correlation and linear regression examined the access-participation relationship.

Third, barrier analysis used chi-square tests and logistic regression to identify which barriers disproportionately affected which ethnic groups. Multiple regression modeled EAS as a function of ethnicity, controlling for household size, education, farm size, market distance, and land tenure status. All analyses used SPSS version 28 with significance at alpha = 0.05 [6] .

Seasonal variations

Extension contact patterns varied seasonally in ways that amplified ethnic disparities. The main sugarcane harvest season (July-November) saw peak extension activity concentrated in Hindu-majority communities with larger sugarcane holdings. Creole households, more often engaged in mixed vegetable production peaking in the cooler months (May-September), found fewer extension events scheduled around their agricultural calendar. Muslim households reported that Friday afternoon scheduling of extension events conflicted with religious obligations, reducing effective access by an estimated 15-20%. These scheduling mismatches reflected an extension planning process that defaulted to the majority community's agricultural calendar rather than accommodating diversity. Barrier patterns differed sharply across groups (Table 2). Creole households reported the highest severity across nearly all barriers, with land title requirements (52%) and transport cost (48%) most prominent. Muslim households were disproportionately affected by language mismatch (28%) and gender norms. Franco-Mauritians, despite high access, reported notable mistrust of state extension agents (18%), which partly explains their lower-than-expected participation. The scatter plot (Figure 1) reveals distinct clustering by ethnic group within a positive access-participation relationship. The quadrant analysis shows Creole households concentrated in the low-access/low-participation quadrant, while Indo-Mauritian Hindus occupy the high-access/high-participation space. Sino-Mauritians and Franco-Mauritians show an interesting pattern of high access but moderate-to-low participation, occupying the right side of the plot but below the diagonal.

Results

Cost analysis

Achieving equitable extension access carries measurable costs. Closing the access gap for Creole households to the sample median would require an estimated $28 per household annually-covering transport subsidies ($12), Creole-language materials ($8), and additional agent time ($8). Across the estimated 12,400 Creole farming households in three districts, the total annual investment would be $347,200. The projected productivity gain from equalizing extension access would add approximately $890 per household annually-a cost-benefit ratio of 1:31.8, making equity-focused extension investment substantially more cost-effective than marginal improvements in alreadyserved communities [7,8] .

Discussion

The 40-point access gap and 52.7% variance explained by ethnicity alone indicate that caste and ethnic identity operate as powerful sorting mechanisms in Mauritian extension delivery, despite the country's relatively strong institutional framework and absence of formal caste hierarchy [9] . Three patterns deserve particular attention. First, the Creole disadvantage is multi-layered. Creole households face the most severe barriers across all three dimensions-structural (remoteness, transport cost), cultural (language, mistrust), and institutional (land title requirements). The land title barrier is especially telling: 52% of Creole households reported that eligibility requirements excluding untitled farmers effectively excluded them, reflecting historical landlessness rooted in post-slavery settlement patterns [10,11] . Second, the Franco-Mauritian paradox-high access but low participation-suggests that access without cultural engagement is insufficient. Franco-Mauritian farmers, who tend to operate larger commercial holdings, reported that extension content was oriented toward smallholder concerns irrelevant to their operations. Their mistrust of state agents (18%) may also reflect historical tensions between large landholders and government agricultural policy [12,13] .

Conclusion

This research documents significant ethnic disparities in agricultural extension access and participation in Mauritius, with caste and ethnic identity explaining over half the variance in access scores. Creole households face the most severe disadvantages, scoring 40 points below the highestaccess group across structural, cultural, and institutional barrier dimensions. The strong access-participation correlation (r = 0.726) means access inequities translate directly into participation gaps. Achieving equitable extension delivery requires ethnically disaggregated monitoring, removal of exclusionary eligibility criteria, culturally adapted programming, and deliberate resource allocation toward underserved communities. The 1:31.8 cost-benefit ratio makes this not only an ethical priority but an economic one.

References

1. Food and Agricultural Research and Extension Institute. Annual report 2022: extension service performance review. Reduit: FAREI; 2023.
2. Eriksen TH. Common denominators: Ethnicity, nationbuilding and compromise in Mauritius. Oxford: Berg; 1998.
3. Ragasa C, Berhane G, Tadesse F, Taffesse AS. Gender differences in access to extension services and agricultural productivity. J Agric Educ Ext. 2013; 19(5):437-468.
4. Elias A, Nohmi M, Yasunobu K, Ishida A. Effect of agricultural extension program on smallholders' farm productivity: evidence from three peasant associations in the highlands of Ethiopia. J Agric Sci. 2013; 5(8):163-181.
5. IBM Corp. IBM SPSS Statistics for Windows, version 28.0. Armonk (NY): IBM Corp; 2021.
6. Ministry of Agro-Industry and Food Security. National agricultural investment plan 2022-2027. Port Louis: Government of Mauritius; 2022.
7. Doosarran SD. Cost-effectiveness of targeted extension delivery in multi-ethnic agricultural communities: a Mauritian perspective. J Agric Ext Rural Dev. 2023; 15(3):189-204.
8. Laville R, Mathur N. Social stratification without caste? Rethinking ethnic hierarchy in Mauritius. J South Afr Stud. 2019; 45(2):289-306.
9. Allen R. Slaves, freedmen, and indentured laborers in colonial Mauritius. Cambridge: Cambridge University Press; 1999.
10. Bunwaree S. The marginal in the miracle: human capital in Mauritius. Int J Educ Dev. 2001; 21(3):257271.
11. Selvon S. A comprehensive history of Mauritius: from the beginning to 2001. 2 nd ed. Port Louis: Mauritius Printing Specialists; 2012.
12. Minogue M. Mauritius: island of success? Political economy of multi-ethnic development. Commonw Comp Polit. 2018; 56(1):1-22.
13. World Bank. Mauritius: systematic country diagnostic. Washington (DC): World Bank Group; 2021.
14. Statistics Mauritius. Agricultural census 2020: detailed tables. Port Louis: Ministry of Finance; 2021.
15. Rao N, Singh CK. Social identity and agricultural extension in India: a critical review. Econ Polit Wkly. 2020; 55(12):47-54.
16. Diagne A, Demont M. Taking a new look at empirical models of adoption: Average treatment effect estimation of adoption rates and their determinants. Agric Econ. 2007; 37(2-3):201-210.
17. Food and Agricultural Research and Extension Institute. Guidelines for inclusive extension service delivery in multi-ethnic communities. FAREI technical series. Reduit: FAREI; 2022.
18. Gopal G, Jeetah P. Agricultural transformation in Mauritius: challenges and opportunities for smallholder farmers. J Rural Stud. 2022; 89:142-153.
19. Ramgulam S, Doosarran SD. Ethnic disaggregation of agricultural performance data: methodology and pilot results from Mauritius. Afr J Agric Res. 2023; 18(7):412-428.
20. United Nations Development Programme. Mauritius human development report 2022: Inequality and social cohesion. Port Louis: UNDP; 2023.