Effect of women’s empowerment on lowering the number of children ever born in Bangladesh

Abstract

Introduction

Women’s empowerment (WE) has become a central focus for development at a national and global level. There is a need for a comprehensive and updated assessment of the existing evidence on WE and the total number of children ever born (CEB). This paper addresses the association between different dimensions of WE and the total number of CEB among women aged 15–49 years in Dhaka, Bangladesh.

Methods

This is a cross-sectional study. Using primary data, two binary and skewed regression models were applied to find the best-fitting model to investigate the association between different dimensions of women’s empowerment and the total number of CEB.

Results

Based on the Akaike Information Criterion (AIC) and the Bayesian Information Criterion (BIC) principles, the best-fit model was the binary logistic regression compared to other models. The study revealed that higher levels of women empowerment were associated with fewer CEB. Among the four dimensions—economic, psychological, household, and socio-cultural indices- psychological dimensions significantly influenced the number of CEB. Results also showed that age at first marriage, educational attainment of respondents, occupational level, wealth index, and use of contraceptives were found to be negatively associated with the number of CEB.

Conclusion

The study found that an increase in WE was linked to a reduction in CEB. Other predictors for CEB were age at first marriage, women’s educational attainment, working status, wealth index, and contraceptive use, which were all identified as factors associated with a lower number of children. Policymakers should focus on the dimension-specific and overall level of WE in reducing CEB.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12905-025-03624-3.

Introduction

Women’s empowerment (WE) has emerged as a critical factor in shaping the social, economic, and demographic dynamics of nations across the globe [1]. This multi-dimensional concept is defined and measured variously by researchers, making it difficult to assess through a single metric [2]. The study has considered this complicated notion through numerous lenses, utilizing distinct criteria and methods to measure WE [3]. A study identified three features of WE: agency, resources, and achievements [4]. In addition to agency and resources, other research introduced another model of WE that incorporates institutional structures as a dimension [5, 6]. A study comprehensively addressed seven categories of WE, encompassing economic, health, human development, leadership, psychological, security and justice, and socio-cultural elements [7]. Research on WE in Bangladesh examines attitudes toward domestic violence and their level of engagement in decision-making [8]. Participation in household decision-making, self-esteem, freedom of mobility, and control of material resources were also considered in WE studies [9]. WE encompass multiple dimensions, and researchers employ diverse dimensions and indicators to measure it [10, 11]. Hence, assessing the degree of WE is imperative to accurately reflect their position within the household and the broader community [12]. Research on WE presents its diverse collection of indicators and procedures, resulting in a thorough and somewhat fragmented discussion. Various approaches emphasize the absence of a universal measure for empowerment, highlighting the necessity for analyses that consider specific contexts. Empowerment improves the overall welfare of women and significantly influences the population dynamics of any country, particularly developing countries [13, 14]. As women gain more authority and autonomy in decision-making, they become more capable of influencing their reproductive decisions, affecting the population size, structure, and composition of all countries [15]. The correlation between women’s empowerment and fertility is significant for researchers [16].

Bangladesh’s total fertility rate (TFR) is 2.3 children per woman, which remained unchanged since 2011 [17]. This stagnation raises compelling questions about the causal factors influencing fertility rates and the efficacy of empowerment initiatives in the country. Social, economic, demographic, and cultural dimensions influenced fertility [18–21]. Women’s empowerment is a reliable indicator of fertility in both developed and developing countries [18, 20, 21]. Factors such as women’s educational attainment, employment status, household decision-making, involvement in economic decision-making, and freedom of movement influence fertility rates in Bangladesh [18].

Research shows maternal education increases women’s decision-making power regarding reproductive health [22]. Study findings indicate that providing women with education, economic opportunity, and decision-making authority substantially decreases fertility rates and enhances reproductive health [16, 23, 24]. A study conducted in 53 low and middle-income countries showed that WE is closely linked to fertility decision-making [25]. The study also showed the role of education, economic stability, and decision-making power on fertility choices. Educated women are more likely to have control over their reproductive decisions, resulting in smaller families [26]. Financial independence also influences family size, as economically empowered women generally prefer fewer children [27]. Additionally, women with a voice in household and community decisions are more likely to use family planning methods, leading to lower fertility rates [28, 29]. A review study included 35 South Asian research studies and found positive connections between WE, lower fertility, longer birth intervals, and reduced rates of unwanted pregnancy [23].

Existing studies have found a general link between WE and fertility. However, there is little research on how the economic, household, socio-cultural, and psychological dimensions of WE and socio-economic factors are associated with CEB. A recent paper on WE used 33 indicators under economic, household, socio-cultural, and psychological dimensions. A summary of the indicators by dimensions is presented in Table 1 [11]. Based on the indicators of the previous paper, the overall WE index and individual scores have been constructed. This paper examines how the particular dimension of WE, overall WE, and socio-economic characteristics are associated with children ever born.

Table 1.

Dimensions and their- indicators

Dimensions	Indicators
Economic (8 indicators)	Control over the family budget, Control over major HH purchases, Control over HH savings, Ability to buy something the respondent needs, Control over their income, Asset ownership, Contribution to household income, Involvement in income-generating activities The above indicators on economic dimension were used from the studies [7, [29–35]
Household (7 indicators)	Decision-making on—cooking food for meals, inviting guests to your home, the respondent’s health care, their own child’s healthcare, how many children to have, when to have a child, family planning, and contraceptive use The above indicators on household dimension were used from the studies [29, 32, 30, 31, 33, 30]
Socio-cultural (8 indicators)	Allowed to go out to the -local market/ bazaar/ bank, local health center/doctor’s clinic, home of family/relatives, other cities The above indicators on socio-cultural dimension were used from the studies [7 ,30, 33, 29, 36]
Psychological (10 indicators)	Ten indicators from the Rosenberg self-esteem scale (RSES) [7, 37, 38]

Source: [11]

As Bangladesh is undergoing rapid socio-economic changes, the fertility dynamics can change over time, and it is essential to monitor and understand these changes. Such transitions can have nuanced associations between WE and CEB. As the country strives to meet its Sustainable Development Goals, including those related to public health and gender equality, it becomes essential to have research tailored to its unique demographic landscape. This study aims to fill existing research gaps by examining how different WE dimensions and socio-economic aspects are associated with the total number of CEB in Bangladesh.

Materials and methods

Data and variables

Data were collected through a face-to-face interview using a self-administered questionnaire from 625 married women aged 15–49 with at least one child and currently living with a husband. The study was conducted in the eight Mohallas of Dhaka (Fig. 1). Mahalla, the smallest administrative unit under City Corporation, is the smallest urban geographic unit with identifiable boundaries. Probability proportional to size (PPS) sampling was used to determine the number of respondents from each mohalla. A list of eligible women was prepared with the help of the Ward Commissioner office of the respective Mohallas. The survey was administered to one eligible woman from each household, and a structured questionnaire was used to assess demographic characteristics, women empowerment, and child-related information. The dimensions of WE and dimension-specific indicators were taken from prior research and developed in the questionnaire. It took 40–45 min to complete the questionnaire. To avoid missing observations, we were careful about missed records of cases/ responses during data collection, field checks, and data entry.

Fig. 1.

Study area map (Map Source: Internet) [39]

Sampling technique and sample size

The sample size of the study was calculated by using the following formula:

$$n=\frac{z^2\times p\text{q}\times (deff)}{e^2\times rr}$$

Here, n = total sample size; z = the standard normal deviation, usually set at 1.96 at a 95% confidence level; p = predicted (anticipated) prevalence of women empowerment = 0.50; q = 1-p; deff (design effect) = 1.5; e (margin of error) = 0.05; rr (response rate) = 0.90. Using the formula, the study’s sample size was 640. The design effect 1.5 was used to account for any differences in the population, making the sample size estimate more reliable. This adjustment represents the likelihood of increased variability in the data. However, the study completed interviews among 625 women with a non-responselection rate of 2.34%. Before the start of the main survey, pretesting of the questionnaire was conducted by interviewing a sample of thirty women.

Dependent variable

The outcome variable of this study was the total number of children ever born (CEB) to a woman. Of 625 mothers, the researcher coded 0 for two or fewer children and 1 for three or more children.

Independent variables

Economic, familial, socio-cultural, psychological, and socio-economic dimensions of women’s empowerment were included as independent variables. Additionally, socio-economic aspects, age-related factors, women’s and husbands’ education, occupation, wealth index, religion, contraceptive use, and experience of familial violence were also included. The study used the same household items in BDHS to measure the wealth quintile. However, we classified the wealth index into three groups (“Poor,” “Middle,” and “Rich”) instead of four, as was used in the WE literature [40, 41]. This classification was created by using principal component analysis (PCA) based on various home assets (e.g., televisions, bicycles, drinking water sources, sanitation facilities, and construction materials) [42].

Women’s Empowerment Index (WEI)

This study employs a multi-dimensional framework that considers economic, household, social, socio-cultural, and psychological dimensions to assess women’s empowerment. These four dimensions were developed based on the review of more than thirty different studies. Each dimension is measured through several indicators that reflect various aspects of WE and decision-making power [11]. In total, 33 indicators were used in the framework to measure WE. To account for the multi-dimensional nature of WE, the study applied the Explanatory Factor Analysis (EFA), which categorized similar factors together and divided the four dimensions of empowerment into eight factors. Like positive self-esteem (SEP), negative self-esteem (SEN) [reverse coded], economic independence, control over economic decision-making, freedom of movement, women’s access to information and technology, household decision-making, and reproductive decision-making. The study performed confirmatory factor analysis (CFA) to validate the factor structure received from the EFA. All the items have sufficient loadings on their latent construct (> 0.5 or, ideally, > 0.7), which supports the idea that convergent validity is achieved in all the dimensions. The construct reliability value for all four constructs exceeded the threshold level of 0.7. All square correlation values were lower than the AVE values of their respective factors, indicating no issue in the model’s discriminant validity.

Factor scores obtained from the CFA, specifically the standardized regression weights, were utilized to construct a weighted sum of the indicator scores. The factor score for each construct was computed by summing the weighted values of each indicator, with weights determined by their respective factor score. The resulting factor scores were then standardized (z-scores) to make them comparable. The z-score is computed as follows: $z=\frac{x-\mu }{\sigma }$, where x is the factor score, µ is the mean of the factor scores, and σ is the standard deviation of the factor scores. Then, using the standardized factor scores, create individual scores for each respondent. We employed standard deviation (SD) based categorization techniques, as this method is particularly appropriate for z-scores, which inherently quantify deviations from the mean. The higher score indicated a higher level of empowerment and vice versa. The categorization aimed to assess the distribution of cases among women classified as low, moderate, and highly empowered [11]. Previous studies employed a similar technique [43, 44]. The individual indexes have been aggregated to form an overall index.

Statistical analysis

Cronbach’s alpha assessed the reliability and validity of the data. The socio-demographic characteristics of study variables were analyzed using descriptive statistics. To investigate the association between women’s empowerment and the number of CEB, we performed two regressions—binary logistic regression [45] and maximum likelihood skewed logistic regression [46] and [47]. Binary logistic regression is suitable for analyzing the relationship between women’s empowerment and the number of CEB as a binary outcome. In contrast, maximum likelihood skewed logistic regression accommodates potential skewness in the data, providing greater flexibility for non-symmetric distributions. The skewness value of the CEB was 1.163, indicating a significant positive skew in the data, thereby justifying the application of skewed logistic regression for improved model fit. The study then chooses the best-fitted model according to the criteria of AIC [46] and BIC [47].

In the current study, researchers can compare the results and validate their findings using these two methods. Utilizing both methods guarantees robustness, assists in identifying potential model biases, and validates that the observed relationships are not the result of a single modeling approach—variables selected for the study based on theoretical relevance, prior research, and statistical criteria. Bivariate analyses were conducted to observe the association between dependent and independent variables. This study employed fully adjusted logistic models. The selection of variables for the models was informed by theoretical reasoning, existing literature, and statistical criteria. The Variance Inflation Factor (VIF), which was employed to check for the presence of multicollinearity between the variables, revealed no multicollinearity issue, as the VIF ranges from 1.07 to 2.67) as our concern was to check the role of women empowerment indices in determining their total children ever born. The study’s Beta coefficients (effect sizes) were derived by maximum likelihood estimation (MLE) during the logistic regression modeling procedure. In binary logistic regression, the Beta coefficients signify the log odds of the dependent variable (fertility) for each unit change in the predictor variable. On the other hand, in skewed logistic regression, the Beta coefficients represent the log odds while accommodating the asymmetric distribution of the dependent variable, hence providing adjusted effect sizes. The data were analyzed using the Statistical Package for the Social Sciences (Version SPSS-22.0, AMOS 22, and STATA-16.0).

Results

Socio-demographic characteristics

A total of 625 women were included in the current study, and it was observed that exactly one child born was 153 (24.5%), two children 268 (42.9%), three children 146 (23.4%), and the rest, more than three was 9.4% as shown in Fig. 2. Table 2 shows that the majority of the women were aged 15 to 29 years (36.2%), had completed their secondary education 217 (32.0%), belonged to middle-class families 226 (36.0%), were not working 398 (63.7%) and belonged to Islam (81.6%). In addition, 244 (39.0%) of women got married before becoming 18 years old. Most of the women were moderately empowered in various dimensions, including economic 263 (42.0%), household 373 (60.0%), socio-cultural 351 (56.0%), and psychological 286 (46.0%) aspects. However, it is notable that a considerable proportion still experience low levels of empowerment in these dimensions. This study incorporated socio-economic, demographic, and crucial life-event-related variables besides empowerment indices.

Fig. 2.

Number of children born

Table 2.

Socio-economic and demographic variables of ever-married women along with frequency and percentage (%) distributions

Dependent variables	N (%)
Children Ever Born (CEB)
Two or less	421 (67.4)
More than two	204 (32.6)
Women empowerment-related variables
Economic Empowerment (ECO)
Low	259 (41.0)
Moderate	263 (42.0)
High	103 (16.0)
Household Empowerment (HH)
Low	107 (17.0)
Moderate	373 (60.0)
High	145 (23.0)
Psychological Empowerment (PSY)
Low	205 (33.0)
Moderate	286 (46.0)
High	134 (21.0)
Sociocultural Empowerment (SOC)
Low	150 (24.0)
Moderate	351 (56.0)
High	124 (20.0)
Overall empowerment index (WE)
Low	180 (29.0)
Moderate	345 (55.0)
High	100 (16%)
Age-related variable
Respondent Age
< 30	226 (36.2)
30–39	219 (35.0)
40 & more	180 (28.8)
Age at marriage
< 18	244 (39.0)
18 and above	381 (61.0)
Age at first birth
< 20	158 (25.3)
20–24	324 (51.8)
25 +	143 (22.9)
Age differences in spousal
< 5	168 (27.0)
5–9	251 (40.0)
> 10	206 (33.0)
Respondents related Variables
Respondent’s education
Illiterate	98 (15.7)
Primary	136 (21.8)
Secondary	200 (32.0)
Higher Secondary or more	191 (30.6)
Respondent’s occupation
Not working	398 (63.7)
Working	227 (36.3)
Husband related
Husband education
Illiterate	92 (15.0)
Primary	110 (18.0)
Secondary	205 (33.0)
Higher Secondary or more	218 (35.0)
Husband’s Influence
No	434 (69.4)
Yes	191 (30.6)
Household characteristics
Wealth index
Poor	204 (33.0)
Middle	226 (36.0)
Rich	195 (31.0)
Violence own case
No	412 (65.9)
Yes	213 (34.1)
Religion
Non-Muslim	115 (18.4)
Muslim	510 (81.6)
Family planning related
Contraceptive use
No (R)	240 (38.4)
Yes	385 (61.6)

Women with high economic (aOR = 0.261, 95% CI: 0.050–0.574), psychological (aOR = 0.224, 95% CI: 0.170–0.519), household (aOR = 0.052, 95% CI: 0.017–0.180), and socio-cultural (aOR = 0.232, 95% CI: 0.077–1.032) empowerment exhibit greater odds of CEB in comparison to the low empowered group (reference) (Table 3). Age positively influences the total number of CEB (aOR = 5.587, 95% CI: 2.438–12.803), age at marriage at 18 and above (aOR = 0.310, 95% CI: 0.152–0.633), higher educational level (aOR = 0.224, 95% CI: 0.056–0.898), working status (aOR = 0.259, 95% CI: 0.117–0.570), wealth index of middle income (aOR = 0.135, 95% CI: 0.057–0.318), high income (aOR = 0.173, 95% CI: 0.07–0.428), and contraceptive prevalence (aOR = 0.317, 95% CI: 0.166–0.608) these all showed comparatively lower odds but statistically significant than reference category in binary logistic regression.

Table 3.

Regression estimates for the effects of indices of women’s empowerment and other socioeconomic variables on the total number of CEBs

Characteristics of Sample	Binary logistic (95% CI) Model I	Skewed Logistic (95% CI) Model II
Economic index
Low (R)
Moderate	0.194*** (0.089, 0.423)	0.334*** (0.19, 0.588)
High	0.261** (0.05, 0.574)	0.328** (0.122, 0.879)
Psychological index
Low (R)
Moderate	0.079*** (0.034, 0.187)	0.154*** (0.082, 0.291)
High	0.224*** (0.17, 0.519)	0.316*** (0.164, 0.807)
Household index
Low (R)
Moderate	0.085*** (0.031, 0.236)	0.170*** (0.081, 0.358)
High	0.052** (0.017, 0.180)	0.274** (0.115, 0.650)
Sociocultural index
Low (R)
Moderate	0.249*** (0.110, 0.724)	0.363*** (0.204, 0.647)
High	0.232** (0.077, 1.032)	0.425** (0.179, 1.007)
Age group
< 30(R)
30–39	4.994*** (2.231, 11.176)	3.502*** (1.913, 6.41)
40 & more	5.587*** (2.438, 12.803)	3.498*** (1.901, 6.438)
Age at marriage
Below 18 (R)
18 and above	0.310*** (0.152, 0.633)	0.456*** (0.266, 0.781)
Age gap
< 5 (R)
5–9	0.813 (0.377, 1.751)	0.786 (0.432, 1.431)
> = 10	0.98 (0.459, 2.091)	0.834 (0.468, 1.487)
Respondent Education
Illiterate (R)
Primary	0.261** (0.076, 0.893)	0.408** (0.167, 0.995)
Secondary	0.422 (0.128, 1.393)	0.592 (0.247, 1.42)
Higher than secondary& +	0.224** (0.056, 0.898)	0.405 (0.137, 1.196)
Respondent Occupation
Not working (R)
Working	0.259*** (0.117, 0.57)	0.358*** (0.198, 0.646)
Spouse Education
Illiterate (R)
Primary	0.306* (0.082, 1.145)	0.393* (0.149, 1.04)
Secondary	2.128 (0.583, 7.769)	1.373 (0.545, 3.458)
Higher Secondary +	2.752 (0.6, 12.623)	1.473 (0.491, 4.417)
Wealth index
Poor (R)
Middle	0.135*** (0.057, 0.318)	0.252*** (0.137, 0.465)
Rich	0.173*** (0.07, 0.428)	0.284*** (0.14, 0.577)
Violence own case
No (R)
Yes	1.787 (0.895, 3.567)	1.5 (0.936, 2.405)
Religion
Non-Muslim (R)
Muslim	1.202 (0.535, 2.697)	1.083 (0.585, 2.006)
Husband’s desire for children
No (R)
Yes	2.67** (1.206, 5.909)	2.186** (1.198, 3.987)
Contraceptive use
No (R)
Yes	0.318** (0.166, 0.608)	0.41** (0.249, 0.675)
AIC	367.1	368.12
BIC	491.36	496.82

(R refers to the reference category; within the first parentheses, a 95% confidence interval is given, and asterisks define statistical significance, such as ***, **, and *, referring to 1%, 5%, and 10% significance levels, respectively.)

In Table 4, women with moderate (aOR = 0.039, 95% CI: 0.018–0.085) and high (aOR = 0.180, 95% CI: 0.061–0.529) empowerment levels had comparatively lower odds of having a child than low- empowered. Age at marriage at 18 and above (aOR = 0.249, 95% CI: 0.123–0.502), higher educational level (aOR = 0.329, 95% CI: 0.091–1.194), working status (aOR = 0.291, 95% CI: 0.145–0.587), women are in rich wealth index category (aOR = 0.188, 95% CI: 0.078–0.453), and contraceptive prevalence (aOR = 0.281, 95% CI: 0.15–0.527)-these regressors exhibited comparatively lower but statistically significant association with the number of CEB than reference category. Furthermore, the age group of 40 or more (aOR = 5.029, 95% CI: 2.209–11.451) and desire for more children from the husband showed a higher likelihood of having more children (aOR = 1.896, 95% CI: 0.94–3.824) than not desiring more. The minimum Akaike and Bayesian Information Criterion’s (AIC) model selection criteria were used to assess alternative models and pick the one that best fits the (AIC) and Bayesian Information Criterion’s (BIC) observed for binary logistic regression were the best performed, as both came with minimum values on binary logistic regression.

Table 4.

Regression estimates for the effects of overall women’s empowerment and other socioeconomic variables on fertility

Characteristics of Sample	Binary logistic (95% CI) Model III	Skewed Logistic (95% CI) Model IV
Overall Woman empowerment index
Low (R)
Moderate
High	0.039*** (0.018, 0.085)	0.09*** (0.051, 0.16)
Age group	0.180** (0.061, 0.529)	0.301*** (0.133, 0.68)
< 30(R)
30–39	3.901*** (1.82, 8.363)	2.497*** (1.414, 4.409)
40 & more	5.029*** (2.209, 11.451)	3.084*** (1.653, 5.751)
Age at marriage
Below 18 (R)
18 and above	0.249*** (0.123, 0.502)	0.395*** (0.232, 0.673)
Age gap
< 5 (R)
5–9	0.991 (0.473, 2.078)	1.029 (0.577, 1.836)
> = 10	1.218 (0.575, 2.581)	1.149 (0.641, 2.059)
Respondent Education
Illiterate (R)
Primary	0.173** (0.056, 0.539)	0.269*** (0.117, 0.618)
Secondary	0.616** (0.208, 1.827)	0.683 (0.312, 1.493)
Higher than secondary& +	0.329** (0.091, 1.194)	0.43 (0.16, 1.156)
Respondent Occupation
Not working (R)
Working	0.291*** (0.145, 0.587)	0.377*** (0.222, 0.64)
Spouse Education
Illiterate (R)
Primary	0.827*** (0.253, 2.695)	1.015 (0.433, 2.378)
Secondary	3.173 (0.925, 10.886)	2.635** (1.08, 6.434)
Higher Secondary +	3.547 (0.85, 14.793)	2.495 (0.857, 7.265)
Wealth index
Poor (R)
Middle	0.129** (0.057, 0.296)	0.262*** (0.146, 0.473)
Rich	0.188*** (0.078, 0.453)	0.345*** (0.178, 0.672)
Violence own case
No (R)
Yes	2.13** (1.103, 4.11)	1.961*** (1.218, 3.159)
Religion
Non-Muslim (R)
Muslim	0.902 (0.431, 1.887)	0.864 (0.483, 1.544)
Husband’s desire for children
No (R)
Yes	1.896** (0.940, 3.824)	1.618* (0.966, 2.713)
Contraceptive use
No (R)
Yes	0.281** (0.150, 0.527)	0.398** (0.245, 0.648)
AIC	376.04	379.76
BIC	468.11	477.38

(R referred refers to the reference category; within the first parentheses, a 95% confidence interval is given, and asterisks define statistical significance, such as ***, **, and *, referring to 1%, 5%, and 10% levels of significance, respectively.)

Discussion

Measuring WE using four dimensions and linked with CEB was a new initiative for Bangladesh, especially including the psychological dimension of WE and its association with the number of CEB. The study showed how, overall, WE contributes to reducing the total number of CEB. Similarly, an increase in WE in individual domains, such as economic, household, psychological, and socio-cultural, also reduces CEB. Economic empowerment allows them the financial resources to access contraceptives and healthcare services, while household empowerment helps women negotiate regarding healthcare, child-rearing, and reproductive choices. Socio-cultural empowerment ensures mobility, freedom to engage with their community, and access to information and technology. Enhanced mobility and access to resources often translate into better-informed fertility decisions and improved health outcomes. Psychological empowerment addresses women’s self-esteem and belief in their potential to change their lives. Positive self-esteem and reduced psychological barriers help women independently make reproductive health and fertility decisions. Studies in neighboring countries [48], including India [48], Pakistan [49], Srilanka [29], and Nepal [50], have also studied the relationship between WE and fertility, focusing on its indicators.

Bangladesh is a country with higher child marriage and teenage pregnancies and also a higher level of unmet needs [51]. The domain-specific focus on WE is essential and can be attributed to incremental adjustments in cultural views toward gender equality in the country [6]. Economic domain-specific focus may improve women’s financial inclusion and decision-making power [50, 51], affecting fertility decisions [52, 53]. A study in Bangladesh reveals a significant correlation between the number of live children and women’s status, precisely their education level, occupation, and conversations with their spouses regarding family planning matters [54]. Several other studies conducted in Bangladesh and other countries correspond with our findings [30, 55–58].

Teenage pregnancy in Bangladesh is very high [59, 60], and the study showed there is a positive connection between women’s age and the number of children. The findings showed that the number of CEB tends to rise as women’s age increases. This result aligns with the previous study conducted in Nepal and Bangladesh [19, 61]. A study revealed that age was a crucial factor in making the decision, as unlike women under 20, women 35 years of age or older had 2.7 times more freedom regarding selecting maternal health care [62]. In Bangladesh, child marriage is very high [63], and the mother’s age at marriage is also an essential factor, as teenage or child marriage has a higher chance of having more children as compared to a late or standard age of marital tie [64]. Moreover, this study highlights that in Bangladesh, working women have fewer CEBs than those who do not, which aligns with relevant studies in Nepal [50], Bangladesh [61], Tunisia [65], and Uganda [66]. In Bangladesh, 51.0% of women aged 20 to 24 married before age 18 [67]. So, age at marriage is an essential determinant of fertility pattern, as women who marry at a younger age tend to have more children. A study in Bangladesh revealed that an increase of 1 year in the age at first marriage for adolescents results in a postponement of 0.728 years at first birth [68].

The number of CEB is affected by women’s working status. This study highlights that in Bangladesh, working women have fewer CEB than those who do not. One probable explanation could be that working women may have less time to raise their children, leading to fewer CEB. Working women are crucial in decision-making and addressing fertility-related matters, leading to smaller family sizes and improved household lifestyles [30, 69–71]. The current study shows evidence of a negative relationship between fertility and education. This result agreed with studies conducted in India [48], Nepal [50], Pakistan [49], and Bangladesh [16]. In Bangladesh, women’s educational status, i.e., higher education, has significantly enhanced their awareness and utilization of family planning methods. As we are in the twenty-first century, there is plausibility and availability of better knowledge and accessibility of family planning-related medicine and options, derailing birth spacing among educated families [72]. Henceforth, empowering women is a striking agenda for meeting SDGs to maintain replacement fertility levels consistently for extended periods. Contraceptive prevalence was negatively associated with the total number of children ever born, though it also worked as an indicator of empowerment. These findings align with the previous research in Bangladesh [16, 57, 61]. A study conducted on ASEAN countries suggested a complex relationship between women’s empowerment and contraceptive use [73].

In this study, wealth index status is significantly related to the total number of CEB. The study found that the number of children decreases as the family’s wealth index status increases. Families with a higher wealth index status tend to have a better awareness of family planning, resulting in a reduction in the number of children they have. These findings are reinforced by a study conducted in Ghana, Nigeria, and Bangladesh [58, 74, 75]. In Bangladesh, the age of marriage and educational status are lower among those in the lower wealth index [76, 77]. In Bangladesh, a notable percentage of women from economically disadvantaged backgrounds enter into marriage before reaching the age of 18 [78, 79]. Due to the potential limitations in autonomy regarding reproductive choices and the reduced opportunities to postpone childbearing, these women may face significant challenges. In women whose husbands desire more children, the number of children born in that family tends to increase. This is a result of conventional gender roles in Bangladesh, where husbands often exert significant influence over reproductive decisions [61, 80, 81]. In Bangladesh, there remains a notable inclination towards favoring sons. This tendency increases fertility rates as families persist in having children until they reach their preferred number of sons [82].

The findings of this study carry important practical implications for women’s empowerment initiatives in Bangladesh. Firstly, policymakers should concentrate on dimension-specific empowerment strategies to enhance the empowerment process and lower fertility rates. Furthermore, it is essential to address socio-cultural barriers restricting women’s mobility and decision-making capabilities and hindering their participation in economic activities. This study presents several important considerations for future research. First, it primarily focused on married women from urban areas, so similar research needs to be done for rural areas. Expanding the research to include diverse geographic contexts would enhance our understanding of these dynamics of women’s empowerment and the number of CEBs. Second, future studies could benefit from exploring the bidirectional relationship between women’s empowerment and CEB to provide a more comprehensive view. Lastly, incorporating male perspectives could enrich the analysis and provide a more holistic understanding.

This study has some limitations. Firstly, we included only married women from urban areas, so we could not compare our findings with those of women living in rural settings. As this is a cross-sectional study and covers many respondents, the findings cannot be generalized to another part of the country or the whole country. Also, we collected data only from women currently living with their husbands. Still, women who are not currently living with their husbands may experience different outcomes compared to other participants, which were not captured in our analysis. Furthermore, the responses were gathered solely from women; we did not consider men’s perspectives, which may introduce response biases. Finally, CEB assesses the lifetime fertility of a female, with the data being restricted to births documented during the interview and omitting any future births.

Conclusion

Increased level of WE affects reducing CEB. Attention should be given to increasing domain-specific and over WE to decrease the total number of CEB. Factors like age at first marriage, women’s educational attainment, working status, wealth index, and contraceptive use should be in prime consideration as these factors are associated with lowering CEB. Measures should be taken to higher age, the marriage of women, interspousal age difference, and domestic violence as these factors contribute to a higher number of children. The husband’s desire for children increases the number of children. So, focus needs to be placed on husbands passionate about having more children. The current findings can help researchers, program planners, and policymakers achieve more effective measures to lower the number of CEBs. The findings can apply to countries with similar socioeconomic and demographic settings.

Authors’ contributions

S.C. has collected the data anay analyzed and wrote the manuscript; M.A.H has conceived the ideas, guided/supervised the research, reviewed the manuscript.

Funding

There was no funding for the research.

Data Availability

Data will be made availavle upon request to Md Aminul Haque at aminul.haque@du.ac.bd

Declarations

Ethics approval and consent to participate

This research is a part of the doctoral thesis and was reviewed and approved by the Research Evaluation Committee of the Department of Population Sciences, University of Dhaka. The study was conducted as per the principles outlined in the Declaration of Helsinki (https://www.wma.net/policies-post/wma-declaration-of-helsinki-ethical-principles-for-medical-research-involving-human-subjects/). Informed consent was also obtained verbally from the respondents before the interview’s commencement. There were no respondents below the age of 16. Respondents were assured about the anonymity of their names and identities and were informed that their data would be used only for academic purposes.

Competing interests

The authors declare no competing interests.