Impact of educational attainment, caste dynamics and occupation on fertility behaviour among women in Bihar

INTRODUCTION

Fertility is a key component of population change, alongside mortality and migration.¹ Within population science research, the relationship between fertility and education has been extensively studied, with empirical evidence highlighting a strong influence of education on fertility patterns, particularly among tribal and non-tribal populations in India.² Achieving an optimum fertility rate is crucial for meeting sustainable development goals and factors such as caste, education and occupation have been identified as significant determinants of fertility reduction.³ In recent years, fertility rates have gradually declined across the country, including at the state level. Among these determinants, female’s education is a critical determinant in reducing the Total Fertility Rate (TFR). While fertility rates across caste groups are also decreasing, disparities persist. Fertility rates among Muslims, Scheduled Castes (SC) and Scheduled Tribes (ST) remain higher compared to other caste and religious groups. Another cultural determinant, son preference, has shown a slow decline. However, rural populations still hold traditional views that emphasize the necessity of having at least one son to continue the family lineage. Government initiatives and increased digital literacy and connectivity in rural areas have contributed to fertility rate reductions. These efforts have enhanced awareness among rural populations, including women, about modern family planning methods and government programs through television, the Internet, news- papers and awareness campaigns. This evolving scenario underscores the interplay between social, educational and cultural factors in shaping fertility trends in India.

Fertility trends and population density in Bihar: insights from census 2011 and NFHS data

According to the 2011 national population census, Bihar is India’s third-largest state by population. With a population density of 1,106 people per square kilometer, it is much greater than the national average of 382 people per unit area. This makes Bihar the state with the highest population density in the country. According to NFHS-4, Bihar’s total fertility rate (TFR) was 3.4 children per woman, the highest among all Indian states. Urban areas recorded a TFR of 2.4, while rural areas had a much higher TFR of 3.6.

Data from NFHS-5 reveals a slight decline in Bihar’s TFR to 3.0 children per woman, which remains considerably higher than the national average TFR of 2.0. The urban-rural fertility gap persists, with urban areas reporting a TFR of 2.4 and rural areas 3.1, compared to India’s average urban total fertility rate of 1.6 and rural total fertility rate of 2.1. In contrast, Sikkim reported the lowest TFR in India at 1.1 children per woman (urban: 0.7, rural: 1.3), significantly below the national average. Bihar’s fertility decline has been modest over the years, reducing by only 0.6 children between NFHS-3 and NFHS-4 and 0.4 children between NFHS-4 and NFHS-5.

REVIEW OF THE LITERATURE

A literature review is a key research methodology component, aiding in hypothesis formulation and objective framing. Fertility studies primarily assess fertility behavior, focusing on variables like caste, religion, socio-economic conditions, education, occupation and age at marriage. These studies span various states and communities, utilizing primary surveys or NFHS data. Methodologies include cross-sectional analyses and regression techniques to examine the impact of specific variables on fertility.

Several studies have examined the factors influencing fertility behavior in India.¹² Found that higher education levels in both spouses reduced fertility in Assam’s tribal societies, especially when wives were more educated.¹⁹ Highlighted higher fertility among Muslims and SC/ST populations compared to Hindus in Madhya Pradesh, using ANOVA and logistic regression.¹¹ Showed that Christians had the highest life expectancy, while SCs had the lowest, with significant mortality differentials across social groups.²⁰ Emphasized education’s roles in fertility reduction and noted higher fertility among marginal workers due to economic insecurity.⁴ Focused on how age, education and caste differences between couples influenced fertility, revealing higher fertility in inter-caste marriages.¹⁴ Found that women’s empowerment in tribal groups of Rajasthan, through education and decision-making, significantly impacted fertility.¹⁵ Identified higher fertility and lower contraceptive use among SC/ST populations, particularly in rural areas.⁶ Noted a slower fertility decline in Uttar Pradesh, with faster progress after 2000 but persistent gaps in contraceptive use.¹ Observed a limited influence of decision-making on fertility among SC/ST women in Odisha, while highlighted stagnation in fertility rates in Bihar, with maximum declines in urban, Hindu and affluent populations.^7,9 Linked increased female education to declining fertility rates. Together, these studies emphasize the significant roles of education, caste, religion, socioeconomic status and regional disparities in shaping fertility behavior.

Problem statement

While extensive literature exists on fertility behavior, there is limited research specifically focusing on women’s fertility in Bihar. Most studies have primarily examined the impact of factors like caste and education in other states. Despite the overall decline in fertility rates across India, Bihar’s Total Fertility Rate (TFR) remains higher than the national average. This is mainly due to widespread illiteracy and early marriages in rural areas, which hinder rapid fertility decline. A significant contributing factor to high fertility is the low awareness of contraceptive methods among women in Bihar, coupled with an unmet need for family planning services. Additionally, fertility rates are higher among SC and ST populations, whose lower socioeconomic status and reliance on children for economic support, as well as benefits from government programs like ration cards, contribute to their higher fertility rates.

Aims and objective

The present study examines the effect of female’s education and employment on fertility behavior in Bihar, focusing on the top five most populated districts and the bottom five districts with very low literacy rates. Women’s education in Bihar has significantly improved over recent decades. As per the 2011 Census, the female educational attainment rate in Bihar was 51.50%, showing a notable increase from 33.57% in 2001, reflecting a growth of approximately 21% in women’s literacy over the past decade. This study explores the potential linkage between female education, employment and fertility reduction in Bihar. These objectives have been derived from a comprehensive literature review and are categorized into three key areas. To find any association between Female’s education and Total Fertility Rate (TFR) among women of Bihar. To find any association between fertility and women’s employment status among women of Bihar. To find any association between women’s social status and fertility among women of Bihar.

Hypothesis of the study

The following null hypotheses have been formulated to address the research objectives. These hypotheses will be tested using a large secondary dataset, specifically the NFHS-5 data, which is highly relevant to the research:

H01

Is there any significant relationship between female education levels and fertility behavior in Bihar.

H02

Is there any significant effect of Female’s employment status on fertility reduction in Bihar.

H03

Social disparities in literacy rates and population densities do not significantly influence fertility behavior in Bihar.

STUDY AREA, DATA AND METHODOLOGY

Characteristics of study area

Bihar is geographically situated between 24°20’10” N to 27°31’15” N latitude and 83°19’50” E to 88°17’40” E longitude. Bihar is a landlocked state, bordered by Jharkhand to the south, Nepal to the north, Uttar Pradesh to the west and West Bengal to the east. The state is divided into North and South Bihar by the Ganges, with an average elevation of 52.73 meters over sea level.

The study focuses on specific districts in Bihar, targeting the top five districts with the highest population growth and the bottom five with the lowest literacy rates. Between 2001- 2011, Bihar experienced a population growth of 25.42%, compared to 28.62% during the previous decade (1991-2001). The districts with the highest growth rates are East Champaran (29.43%), Khagaria (30.19%), Araria (30.25%), Madhepura (31.12%) and Kishanganj (30.40%).

The five districts identified with the lowest literacy rates are Purnia, Sitamarhi, Katihar, Madhepura and Saharsa. These districts exhibit not only low overall literacy levels but also significant gender disparities in education. Purnia has the lowest average literacy rate at 51.08%, with male literacy at 59.06% and female literacy at just 42.34%. Sitamarhi follows with an average literacy rate of 52.05%, where male literacy is 60.64% and female literacy is 42.41%. Katihar records a slightly higher literacy rate of 52.24%, with male literacy at 59.36% and female literacy at 44.39%. Madhepura, with a literacy rate of 52.25%, shows a considerable gender gap male literacy is 61.77%, while female literacy drops to 41.74%. Saharsa has the highest literacy rate among the five at 53.20%, but the gender divide remains stark, with male literacy at 63.56% compared to 41.68% for female.

Data and methodology

This study utilizes data from the NFHS-5 of India, which provides comprehensive information on population and health measures. The survey includes details about both husbands’ and wives’ characteristics, lifestyles and topics like Health of mothers and children (covering breastfeeding), marital status, fertility trends and family planning practices. In 2019-20, the NFHS-5 covered approximately 6.1 lakh households across all 28 states and 8 Union Territories. While many indicators are consistent with those from NFHS-4, NFHS-5 introduces Emerging subjects such as early childhood education, disability, death registration, access to sanitation facilities, menstrual hygiene practices and factors influencing abortion.

Development and Research Services Pvt. Ltd. (DRS) conducted the NFHS-5 survey in Bihar from July 9, 2019, to February 2, 2020. Data were collected from 35,834 families, 42,483 females and 4,897 males. This study focuses on data from married women aged 15-49 in the selected districts of Bihar. The sample size for ever-married women in this age group included in the analysis is 9,389, derived from the NFHS-5 dataset.

The study draws on data from nine districts, with sample sizes determined based on the NFHS-5 dataset. Among these, East Champaran contributes the highest number of samples at 1,142, followed closely by Araria with 1,122 and Kishanganj with 1,116. Khagaria and Madhepura also have substantial representation, with sample sizes of 1,072 and 1,055, respectively. Sitamarhi contributes 1,004 samples, while Purnia and Saharsa provide 973 and 960 samples, respectively. Katihar has the smallest sample size among the districts included, with 945 individuals. Altogether, the total sample size for the study across these nine dis- tricts is 9,389. This distribution ensures a diverse and representative dataset for examining regional patterns and drawing meaningful insights relevant to the study’s objectives.

This study uses the Multinomial Logistic Regression (MLR) method to analyze the NFHS-5 data. The purpose of using MLR is to examine the relationship between dependent and independent variables. The Total Number of Children, representing the fertility rate, is considered the dependent variable. In contrast, independent variables include factors such as place of residence, female education level, economic status, spouse’s education level, employment status, religion and social status of women. MLR predicts the probability of category membership for the dependent variable based on multiple independent variables, which can be either dichotomous (binary) or continuous (interval or ratio scale). This method is an extension of binary logistic regression, designed to handle more than two dependent variable categories. MLR utilizes maximum likelihood estimation to evaluate membership probability in specific categories. This approach analyzes the relationships between a non-metric dependent variable and dichotomous independent variables. MLR compares multiple groups through a series of binary logistic regressions. It is beneficial when data do not meet the assumptions of normality, linearity or homogeneity of variance required for discriminant analysis, making MLR a preferable choice. The model’s overall fit is assessed by comparing the likelihood values of models with and without independent variables. The relative risk ratio (RRR) in a multinomial logistic model is obtained by exponentiating the coefficients, indicating the relative risk of outcomes compared to a reference group. A unit change in a predictor affects the RRR for a specific outcome, assuming other variables are constant. The 95% Confidence Interval (CI) and standard error help assess the estimate’s precision and significance through the z test and p value, which test whether a coefficient significantly differs from zero. Model significance is evaluated using the chi-square test, indicating the overall relationship between dependent and independent variables. In this study, based on NFHS-5 data, the model examines how factors like residence, education, wealth, employment, religion and ethnicity influence the number of children ever born (CEB), categorized as 0, 1, 2 or 3 children.

Table 1 presents the multinomial logistic regression results, highlighting the impact of independent variables on the dependent variable. Based on the p values, the only significant impact observed is women’s employment status in the case of one child. The p value for women’s employment status is 0.006, which is less than 0.05, indicating a significant relationship between employment status and fertility for one child. Table 2 presents the multinomial logistic regression analysis results, showing the impact of independent variables on the dependent variable. Based on the p values, a significant impact is observed for women with secondary education in the case of two children ever born. The p value for women with secondary education is 0.012, which is less than 0.05, indicating a significant relationship between secondary education and fertility in case of two children.

The multinomial logistic regression analysis shown in Table 3 of the study “impact of education, caste and occupation on fertility behaviour among the women of Bihar” provides important findings regarding the effects of education and employment on fertility behaviour. The results highlight a notable correlation between advanced education levels and current employment status with the likelihood of women having three or more children. Both p values for education (0.006) and employment (0.002) fall below the critical threshold of 0.05, indicating a significant statistical relationship.

DISCUSSION

Present study estimates RRR for number of child birth based on education, employment, religion and wealth index. The results highlighted in table 3-5 indicated notable a correlation between advanced education levels and current employment status with the number children’s women have. The finding of this study indicates that higher level of educational attainment lowers the number of children women have. Also, the result of this study correlate with wider studies exploring the connections between education, employment and fertility. For example, research conducted in a tribal community in Assam revealed that increased levels of female education led to lower fertility rates.

The findings of this study indicate that husbands’ education plays a vital role in the number of children women have. Furthermore, the study of indicates clear support for the findings of this research.⁵ Also, the study showed that as women’s educational attainment rises, the number of children they have tends to decline after reaching a certain level of schooling, highlighting the significant influence of education on fertility choices.^8,12,18 Also, higher education and fertility behaviour among women of Bihar are inversely related. Because of educational attainment, women have an awareness of family planning, which leads to making choices about fertility behaviour among women. Employment often brings about time constraints and challenges, particularly when it comes to balancing work and childcare responsibilities. This reality can encourage smaller family sizes. Additionally, educational attainment plays a significant role in shaping career aspirations, which increases the opportunity costs related to childbearing. As a result, these factors collectively influence fertility decision-making among women.¹³ The strengths of the present study include adopting a method appropriate to represent educational attainment and RRR of children born to a woman. This study helps identify the factors contributing to higher fertility rates. The findings reflect the situation of women in Bihar and can be generalized to similar socioeconomic conditions in other states of India.

CONCLUSION

This study reveals that women’s education and employment do not uniformly impact the fertility rate in the study area. For women with one child, the p-value for employment status is less than 0.05, indicating a significant relationship between women’s employment and fertility. Similarly, secondary education and employment substantially impact women with two children, as their p-values are less than 0.05.

In the case of three children, women’s education plays a crucial role, with a p-value below 0.005, leading to rejecting the null hypothesis. For women with three or more children, the analysis indicates significant effects of female education, husband’s education and female employment, all with p-values below 0.05, thus rejecting the null hypothesis.

Overall, the results suggest that women’s employment significantly affects fertility rates in cases of one child and three or more children. Women’s education also demonstrates a strong influence, mainly on two or more children.

This study focuses on districts in Bihar with low literacy rates and high population growth. A larger sample size would allow us to explore the impact of various determinants on fertility rates further. Expanding the sample could provide clearer insights into the effects of factors such as employment, education, social status and other variables on the number of children born.

Funding: No funding sources

Conflict of interest: None declared

Ethical approval: Not required

REFERENCES

1. 1. Pradhan MR, Mondal S, Saikia D, Mudi PK. Dynamics of caste and early childbearing in India: a perspective of three decades. BMC Women's Health. 2024;24(1):231.
2. 2. Bhagat RB. Recent Trends in Population Growth and Fertility Level Among Muslims in Major States of India, 1981-91. Muslims in India since independence. 1998:67-80.
3. 3. Das B, Tarai D. Decision-making and fertility behaviour: a comparative analysis of scheduled caste and scheduled tribe women in Odisha. Social Change. 2011;41(2):233-49.
4. 4. Das KC. “Age, education and caste differences among the couples and its impact on their fertility in India”. 2013.
5. 5. Dayanti AA, Nurrochmah S, Alma LR. The relationship between husband support and husband’s education level with fertility of women of childbearing age in East Nusa Tenggara Province, Indonesia. J Publ Heal Africa. 2022;13(2):2397.
6. 6. Halli SS, Ashwini D, Dehury B, Isac S, Joseph A, Anand P, et al. Fertility and family planning in Uttar Pradesh, India: major progress and persistent gaps. Reprod health. 2019;16:1-2.
7. 7. Jain AK, Nag M. Importance of female primary education for fertility reduction in India. Economic and Political Weekly. 1986;6:1602-8.
8. 8. Kumar A, Bordone V, Muttarak R. Influence of older generation's fertility behaviours on daughter's desired family size in Bihar, India. Vienna Institute of Demography Working Papers. 2014.
9. 9. Voranger J. Fertility trends and differentials in Arab countries. Population. 1974;29(3):657-9.
10. 10. Ghosh S, Sayyad AA. The Triadic Relationship Between Fertility, Education, and Religion Among Women in Bihar. In Pioneering Approaches in Data Management. IGI Global Scientific Publishing. 2025: 333-356.
11. 11. Kumari M, Mohanty SK. Caste, religion and regional differentials in life expectancy at birth in India: cross-sectional estimates from recent National Family Health Survey. BMJ open. 2020;10(8):35392.
12. 12. Mahanta A. Impact of education on fertility: Evidence from a tribal society in Assam, India. International J Pop Res. 2016;2(1):3153685.
13. 13. Manna S, Gupta V, Sanchaya S, Garg A. Inequities in maternal child health, education, and empowerment among tribal population in India. J Fam Med and Prim Care. 2022;11(11):7150-7.
14. 14. Hemant Patidar. “Women’s Empowerment and Fertility Behaviour among the Tribals of Rajasthan, India”. In: Space and Culture, India. 2018;4:129.
15. 15. Poluru R. “An Analysis of Fertility Differentials among Caste Groups in Andhra Pradesh”. In: esocialsciences.com, Working Papers. 2008.
16. 16. Zaveri A, Paul P, Saha J, Barman B, Chouhan P. Maternal determinants of low birth weight among Indian children: Evidence from the National Family Health Survey-4, 2015-16. PLoS One. 2020;15(12):244562.
17. 17. Sarkar S, Gupta P. Socio-demographic correlates of women’s infertility and treatment seeking behavior in India. J Reprod Infert. 2016;17(2):123.
18. 18. Ram U, Ram F. Demographic transition in India: Insights into population growth, composition, and its major drivers. InOxford Research Encyclopedia of Global Public Health. 2021.
19. 19. Som KS, Mishra RP. Assessing the impact of religions and castes on population fertility in an EAG district of India. Space Culture, India. 2019;7(1):206-22.
20. 20. Kalyan SO, MISHRA RP. Education, caste and occupational differentials in fertility in India. European J Geog. 2020;11(1):37-57.