Intergenerational Social Mobility and Gender Equality Status in Bangladesh: A Cross-Sectional Study

1. Introduction

Gender equality is a primary focus of the sustainable development goals (SDGs) in Bangladesh. Despite its consistent GDP performance, Bangladesh remains afflicted by gender inequality (Asadullah and Chakravorty, 2019). The Global Gender Gap Index (GGGI) 2025 study indicates that Bangladesh has risen to 24th place among 146 nations, a significant improvement from its 99th position in 2024. This marks the largest jump in the global gender gap ranking in one year. Bangladesh's overall gender parity score increased from 68.9% in 2024 to 77.5% in 2025. In the sub-sections of Global Gender Gap Index (GGGI) 2025, economic participation score in Bangladesh improved due to labor-force data revisions, bringing its economic parity back to its 2023 level. The country made progress in bridging the gender gap in literacy, with women increasingly catching up to men in literacy rates. In political empowerment, Bangladesh ranks 3rd globally, with the proportion of women in ministerial positions rising from 9.1% to 22.2% between 2024 and 2025 (WEF, 2025). Ashraf and Ali (2018), on the other hand, while measuring the gender inequality index (GII), brought educational attainment and health, and survival under the social participation pillar. Thus, they have measured GII based on three pillars: economic participation, political participation and social participation (e.g., Ashraf and Ali, 2018). The present study, to explain Ashraf and Ali’s GII status in the Bangladeshi context, has considered intergenerational social mobility as the key determinant as found in an extensive body of empirical studies (Soharwardi and Ahmad, 2020; Baig et al., 2018).

Observational studies on gender inequality reveal determinants including social mobility, workplace facilities and environment for women, age, social norms, access to services, decision-making capacity, job security, institutional expansion, wage disparity, race, globalisation, geographic inequality, trade liberalisation, ethnicity, religion, and patriarchy (Soharwardi and Ahmad, 2020; Baig et al., 2018; Lin et al., 2019; Ragasa et al., 2019; Stavi et al., 2021; Montgomery and Dacin 2020). The findings in the Bangladeshi context suggest that patriarchy does not matter GII (Mishra, 2020). However, the SES index, developed in the American context by Blau and Duncan, derived from education, occupation, and income is the root mobility determinant for explaining women's empowerment or gender equality status (Blau et al., 1967; Ayella and Williamson, 1976; Klasen, 2019; Avvisati, 2020). Unfortunately, none has been conducted any study in the Bangladeshi context on examining the effect of intergenerational social mobility, derived from Blau and Duncan’s SES index, on the gender inequality index (GII) developed by Ashraf and Ali (2018) based on economic participation, political participation, and social participation. In addition, none have either developed or proposed to design a theory-based policy in the field of social mobility and gender equality in Bangladesh. Under this circumstance, the study has two central research questions: how should the theory-based policy for gender equality in Bangladesh be developed? What are the most influential factors of intergenerational social mobility while explaining gender equality in Bangladesh?

2. Literature Review

The research seeks to investigate intergenerational social mobility and its impact on gender equality in Bangladesh. This knowledge is crucial for understanding social mobility about an individual's movement between socio-economic strata in relation to their parents, significantly impacting societal frameworks. This literature debate is divided into two parts: theoretical frameworks and concepts, and actual evidence from Bangladesh or similar contexts.

2.1 Theoretical Literature

Three sociological classics such as Marx, Weber, and Durkheim did not directly utter the concept of social mobility but addressed modernity as a key to the processes of social change (Tiryakian, 1995). Though the assumptions, methods, and findings of the Big Three classics differed, all agreed that societal change is dependent on social and technological forces (Boamah and Rothfub, 2018; Oeij et al., 2019). For instance, Pollack (2015) argues that the continuing process of rationalization is an indispensable and irreversible change. Huber (2009) posits that the dialectical tension between production forces and social relations culminates in the emergence of a new social order. Salmi and Sonck-Rautio (2018) assert that the division of labour is fundamental to contemporary society.

The present study suggests that though the classical sociologists did not directly talk about social mobility, there are clear insights into social mobility theory in their theories of social change. To clarify, educational and occupational mobility are rooted in Durkheim’s notion of ‘division of labor’ which can well explain intergenerational social mobility and gender equality status. To argue, Durkheim’s idea of industrialized society metaphorized as organic solidarity is the product of the ‘division of labor’ due to educational expansion (Durkheim and Halls, 1997). Secondly,economic mobility and income inequality stem from Marx's dialectical battle between the forces of production and social relations (Marx and McLellan, 2000). Marx asserted that capitalism is the most dynamic economic model because of its inherent capacity for renewed profit, as illustrated by his definition of capitalism as M-C-M´ (where M represents money, C denotes commodity, and M´ signifies money + increment). Thus, Marx called capitalism the most dynamic economic model though Marx criticized capitalism for its product of social inequality (Swedberg, 2003). However, he predicted that communism was the panacea for resolving social inequality (Marx and McLellan, 2000). Thirdly, all of the above three mobility aspects are reflected in Weber’s idea of rational action practiced by individuals in rational capitalism to achieve their values or purposive goals. Thus, Weberian analysis of upward social class mobility is dependent on the life chances of the individuals to live in the modern rational society (Weber and Tribe, 2019; Swedberg, 2003).

Unfortunately, the classical sociologists as the root theorists of social mobility have very often been neglected. This, perhaps, happens due to their lack of clarification on the difference between social change and social mobility. Later, Sorokin’s work (1959) “Social and Cultural Mobility” clarified such a difference introducing the concept of ‘social mobility’ and stating that there is neither a completely open society (i.e., class system) nor a completely closed society (i.e., caste system). To him, neither those societies are identical nor their pace of mobility too. He also mentions that societal change can happen over time depending on social mobility in terms of individual movement from one position to another within the mobility ladder (Sorokin, 1959 & 1998). The present study, hence, argues that the classical sociologists are the real proprietors of the concept of social mobility which can explain the gender equality status. This argument will be more obvious from the analytical perspectives of the contemporary theories of social mobility.

2.2 Empirical Literature

There are many types of social mobility. They are horizontal-vertical, upward-downward and intergenerational-intragenerational. These types of mobility can further overlap with each other. The present study is related to intergenerational social mobility which is referred to as intergenerational (i.e., within past and present generations) movement between social positions such as the position of parents (social origins) and that of their children (social destinations) (Hertel, 2017). There are many aspects of intergenerational social mobility such as educational, occupational, income, lifestyle, status, prestige, and religion (Staff et al., 2017). Based on these aspects, contemporary social mobility theories can be explained from different perspectives.

Firstly, the functionalist perspective asserts that variations in social position should be seen as gradational inequality (Wright, 1979) and disparities in wealth and status or prestige (Duncan, 1961; Treiman, 1977). Blau and Duncan (1967), in "The American Occupational Structure," discovered that fathers' educational attainment and occupational standing significantly influence their sons' professional accomplishments. The influence of fathers' employment on their sons' careers was partially mitigated by the educational attainment of the sons. Blau and Duncan, consistent with functionalist theory, posited that the mobility observed between agricultural, manual, and non-manual occupations will diminish due to the rise of universalism, favouring achievement over ascription (Blau and Duncan, 1967; Treiman, 1970). Secondly, the occupational micro-class approach is directly theoretically connected to Durkheim’s idea of the 'Division of Labour' (Grusky and Galescu, 2005). This viewpoint posits that the large classes resulting from the distortions of early industrialisation would be supplanted by occupation-based micro-social classes due to the development of inequality within the labour market (Grusky and Sørensen, 1998). Thirdly, the neo-Weberian perspective referred to as the EGP (Erikson-Goldthorpe-Portocarero) scheme or CASMIN (Comparative Analysis of Social Mobility in Industrial Nations) categorises class distinctions solely in terms of life chances (Breen and Jonsson, 2005; Chan and Goldthorpe, 2007). Goldthorpe (2007) initially formalised the OED triangle, comprising class origins (O), educational attainment (E), and class destinations (D). Golthorpe's OED triangle elucidates the interaction between O, E, and D. Alterations may impact the relationships among O>E, E>D, or O>D. He elucidates the interaction effect of all three elements on the relationship between O and D, suggesting that the ED link may alter in relation to the OD link involving schooling (Hertel, 2017). Goldthorpe (2016), in his examination of advanced nations such as modern Britain, discovered that the extent of social mobility is diminishing due to the proliferation of educational qualifications. Fourthly, the neo-Marxist perspective developed by Wright which is one of the most creative Marxist class schemes based on exploitation (Wright, 1979). Wright’s class scheme is interesting while analyzing intergenerational social mobility (Western and Wright, 1994) in terms of expertise affecting the cultural skills and job preferences of their children. As a result, the exploitation associated with parental class origin matters in the intergenerational penetrability of class boundaries (Wright, 1997). Fifthly, Oesch's formulation of a novel class scheme resulted from the expansion of education, increased female participation in the labour force, and the emergence of service classes alongside routine non-manual vocations (Oesch, 2008). Oesch, disregarding Goldthorpe’s perspective on work relations, observes the direct impact of educational credentials on the "advantage associated with employment relations" (Oesch, 2006: 67). Finally, the Esping-Andersen class scheme analyses the relationship between the intragenerational socio-economic changes and stratification system in the post-industrial societies (Esping-Andersen, 1999).

Considering the sociological theories in the field of social mobility and gender stratification, the study has formulated five hypotheses.

Hypothesis 1: There is a significant effect of SES₁ on the SES mobility index.

Hypothesis 2: There is a significant effect of SES₂ on the SES mobility index.

Hypothesis 3: There is a significant effect of the SES mobility index on GII.

Hypothesis 4: There is a significant association between SES₁ and GII.

Hypothesis 5: There is a significant association between SES₂ and GII.

3. Research Gap

None has explained GII in the Bangladeshi context depending on Blau and Duncan’s SES mobility index derived from three indicators such as education, occupation, and income (Blau et al., 1967, 1978; Hopper et al., 1968) which is sociologically significant while studying gender equality status in Bangladesh. The first reason is that it incorporates all the classical aspects including educational mobility, occupational mobility, and income mobility. The second reason is that Bangladesh is a society where universalism is yet to be established but exists at the stage of increasing inequality (Paulus et al., 2020). The third reason is that Blau and Duncan’s model compares the effect of the social origin (the parental generation) on their future generations. The fourth reason is that the pre-test survey also supports Blau and Duncan’s model to explain GII. The fifth reason is that gender equality is one of the top priority areas of SDGs. So, the study follows Blau and Duncan’s SES mobility index as the root determinant of explaining gender equality status in Bangladesh. Based on this research gap, the study has an endeavour to assess the effect of intergenerational social mobility on gender equality status in Bangladesh.

4. Materials And Methods

4.1 Ethical Consideration

The research received approval in accordance with the ethics committee requirements of Bangladesh University of Professionals (BUP) and was supported by the same institution. The participants were informed that their involvement in the survey was voluntary and that the confidentiality of their responses would be preserved. They also stated that they could withdraw from the survey at any moment. The respondents' consent was obtained through a written form that outlined the research objectives, data collection methods, and data procedures. Verbal consent was also obtained from them. Participants were guaranteed that their identities would remain undisclosed in any subsequent presentations or publications stemming from the study. This study allocated a distinct identifying code to preserve the confidentiality of the data and the identity of the respondents. All datasets were secured with passwords and stored in several locations.

4.2 Research Design and Measurement of Variables

Quantitative study adheres to a rigorous experimental research approach. The study evaluated the importance of the offered hypotheses across four segments: Socioeconomic Status 1 (SES1), Socio-economic Status 2 (SES2), Mobility Index (MOBIN), and Gender Inequality Index (GII). The quantified values of exogenous variables, specifically SES1 and SES2, were employed to assess the GII mediated by MOBIN. The SES1 is defined as the socio-economic level of the parents or first generation, determined by the composite index score derived from their Parent’s Educational Score (PES), Parent’s Occupational Score (POS), and Parent’s Income Score (PIS). Likewise, child’s Educational Score (CES), Child’s Occupational Score (COS), and Child’s Income Score (CIS) pertain to the socio-economic status of children or the second generation in relation to SES2. The exogenous constructs SES1 (socio-economic status of the parents or first generation) and SES2 (socio-economic status of the children or second generation) were defined in relation to their PES and CES, POS and COS, and PIS and CIS. The PES and CES were defined as proportional scores reflecting the educational levels of parents and children in accordance with the general education system in Bangladesh. In this perspective, educational attainment was quantified as follows: PhD and above=7, MPhil=6, postgraduate=5, graduate=4, HSC=3, SSC=2, below JSC=1, and illiterate=0. Conversely, the study defined POS and COS as the proportional scores corresponding to the occupational levels of parents and children, respectively.

The occupational levels were derived from the Quarterly Labour Force Survey conducted by the Bangladesh Bureau of Statistics (2017), with modifications to the codes based on rank scores. In this perspective, occupational levels were redefined as follows: Managers=10, Professionals=9, Technicians and Associate Professionals=8, Clerical Support Workers=7, Service and Sales Workers=6, Skilled Agricultural, Forestry and Fisheries Workers=5, Craft and Related Trades Workers=4, Plant and Machine Operators and Assemblers=3, Elementary Occupations=2, and other occupations=1. The income levels of both parents and children were categorised as upper=3, middle=2, and low=1; ethnicity was classified as Hindu=1, Bengali=2, and others=3; and religion status was categorised as Hindus=1, Islam=2, and others=3. Furthermore, age categories were defined as follows: for dads and mothers, up to 55 years=1, 56-60 years=2, and above 60 years=3; for sons and daughters, up to 30 years=1, 31-35 years=2, and above 35 years=3.

The GIIs for each responder were determined using categorical dummies. Both sons and daughters were requested to indicate their level of agreement with each factor in the Gender Inequality Index between sons and daughters (GIIsd) on a scale from 1 to 5, where 1 represents "not at all" and 5 signifies "a great deal." Based on Eq. (1) formulated by Ashraf and Ali (2018), the GIIsd was initially computed using SPSS software (Windows version 25), followed by the calculation of GII using SmartPLS software (Windows version 4).

                               GIIsd = (EPsS/EPdS)*(PPsS/PPdS)*(SPsS/SPdS)* (1/3)                   (1)

 Here, EPsS, PPsS, and SPsS denote a son's scores in economic, political, and social engagement, respectively. Similarly, EPdS, PPdS, and SPdS indicate a daughter's score in economic, political, and social engagement, respectively. Within these frameworks, the endogenous construct GII was defined through six indicators: EPs and EPd encompass the assertion ‘I have a financial contribution to my family’; PPs and PPd incorporate the assertion ‘My opinion is valued during significant family decisions, and SPs and SPd include the assertion ‘I engage in societal activities’. The EPs, PPs, and SPs were allocated to sons, whereas EPd, PPd, and SPd were allocated to daughters. In comparison, the status of children's educational, occupational, and income levels (CES, COS, and CIS), combined with the status of parents' educational, occupational, and income levels (PES, POS, and PIS), comprise the mobility index (MOBIN). In accordance with Eq. (1), the MOBIcp was predominantly computed using SPSS software, whereas MOBIN was determined using SmartPLS software (Windows version 3.3), as seen in Eq. (2).

                                        MOBIN= (CES/PES)*(COS/POS)*(CIS/PIS)*(1/3)                       (2)

Thus, the study has specified the following conceptual framework (Fig. 1) where CES reveals the educational status of children; CIS represents the income status of children; COS denotes occupational status of children; PES expresses educational status of parents; PIS detects income status of parents; POS presents occupational status of parents; SES1 states socio-economic status of parents; SES2 presents socio-economic status for second generation or children; MOBIN indicates mobility index; MOBIcp shows mobility index for child/parent generation; GII indicates gender inequality index; and GIIsd expresses gender inequality index for sons/daughters.

Figure 1: Hypothesized Model

[Source: Modifiedly Adopted]

4.3 Sampling Technique

The study collected data following a multi-stage technique. In the 1^st stage, the study chose three clusters. Out of 64 districts of Bangladesh, the first cluster consisted of 27 districts of which Kushtia is the member district of this cluster. The second cluster covered 31 districts in which Panchagarh is the representative district of this cluster and the third cluster included Dhaka out of 4 districts. Findings from Hossain and Hossain (2019) about the district-wise clusters of socio-economic and demographic homogeneity in Bangladesh motivated us to split our sample size into three clusters. The study has, thus, chosen above mentioned 3 districts from three clusters for sample selection. In the second stage, three upazilas (Khoksa, Debiganj & Pallabi) were selected randomly. In the third stage, three unions were selected further randomly (Samaspur, Debiduba, and Ward No. 3). But in the final stage, the representative respondents were purposively selected considering the inclusion criteria of 4 respondents from each household (1 father, 1 mother, 1 son, and 1 daughter). The reason for selecting 4 respondents from each household is to develop the mobility index (MOBIN) based on SES1 and SES2 as well as the gender inequality index (GII).

4.4 Sample Size Determination

The minimal acceptable sample size was established using two iterations of Cochran’s method due to the finite study population. The initial phase was determining the number of respondents as n=384, accounting for a 5 percent margin of error. Subsequently, in the second stage, a sample of 384 was modified to account for a finite population size of N=30728 (i.e., N1+N2+N3 = 3254 + 5199 + 22275). The definitive sample size of households was established at 380. Utilising this sample strategy, data was gathered from 40 houses (n1) in the first cluster, 64 households (n2) in the second cluster, and 276 households (n3) in the third cluster.

4.5 Data Collection

 We conducted a pre-test survey for the appropriate empirical inquiry. Twenty-five individuals were solicited to partake in this pre-test survey. Upon concluding the pre-test, we proceeded to conduct household surveys in collaboration with other associates. Participants in the pre-test were excluded from the final home surveys. The household surveys were executed in three districts of our suggested clusters from July to December 2019. A suitable sample strategy was employed during data collection to mitigate sampling error. In this context, all survey interviews were executed by a team of skilled data collectors. The importance and consequences of the current investigation were elucidated to all participants. The interview of each responder was conducted over an extended duration. The data collectors refrained from engaging in personal and irrelevant gossip to prevent biassing the respondents' answers. The surveys adhered rigorously to data collecting processes to ensure compliance with human research standards.

5. Result And Discussion

The analysis revealed that the entire sample exhibited a bias towards a higher representation of individuals with varied education, occupation, income, and age among fathers, mothers, sons, and daughters. Table 1 illustrates the distribution of socioeconomic and demographic (SED) factors among respondents.

Table 1: Summary Statistics of Major SED Characteristics of Respondents (n=380)

PMO&A=Plant and Machine Operators & Assemblers; *CRTW=Craft and Related Trades Workers; ***SAFF=Skilled Agricultural, Forestry and Fisheries; ****SSW=Service and Sales Workers; *****CSW= Clerical Support Workers; ******TAP= Technicians and Associate Professionals.

[Source: Field Data]

Approximately 95 percent of the respondents, comprising fathers and mothers, belong to the oldest demographic category. Conversely, all children fall inside the category of individuals under 55 years of age. Regarding gender, 50 percent of responders are female, while the other 50 percent are male. Over 70 percent of responders have family sizes ranging from 6 to 8 members. Approximately 75 percent of respondents identify as Bengali, while around 92 percent identify as Muslim. A larger proportion of fathers and mothers failed to obtain the Junior School Certificate (JSC), whereas the educational attainment of their children appears to be significantly improving. The occupational status of fathers and mothers lags far behind that of their children. There has been an increase in social mobility for sons and daughters relative to their parental status or origins. The largest proportion of fathers and mothers resides in the lower class, followed by the middle class and upper class, with the majority of their offspring belonging to the middle class. To enhance empirical evaluation, we established the measurement model and structural model for assessing the endogenous and exogenous constructs.

5.1 Measurement Model

The study employs three steps to assess the proposed model. First of all, the internal consistency is initially assessed using Cronbach’s Alpha and composite reliability (CR). The analysis reveals that all values significantly exceed the criterion of 0.70 (Wong, 2013; Hair et al., 2010; Hair et al., 2011; Hair et al., 2014; Hair, Tomas et al., 2016; Nunnally and Bernstein, 1994; Urbach and Ahlemann, 2010) and the acceptable value of 0.50 (Hair et al., 1998). The internal consistency of the data has been verified. Secondly, convergent validity has been assessed via cross-loadings (refer to Figure 2 for further details) and average variance extracted (AVE), with both metrics required to exceed the acceptable threshold of 0.5 (Wong, 2013; Hair et al., 2010; Hair et al., 2011; Hair et al., 2014; Hair et al., 2016; Nunnally and Bernstein, 1994; Urbach and Ahlemann, 2010; Bagozzi and Yi, 1988). The cross-loadings of CES (0.966), CIS (0.900), and COS (0.983), together with those of PES (0.939), PIS (0.831), and POS (0.903) in our study indicate that SES2 and SES1 adequately reflect these two variables. Consequently, the cross-loadings and AVE values exceeding 0.50 in our study validate the convergent validity of our data. Finally, the discriminant validity has been evaluated through Heterotrait-Monotrait (HTMT) ratio which should be less than 0.90 (Gold et al., 2001). Benitez et al. (2020) argue that it should be less than 0.85. Our estimated HTMT ratios of SES1, SES2_, and GII2 are lower than 0.85 which makes conformity of discriminator validity of data. 

Figure 2: Measurement Model

[Source: Field Data]

5.2 Structural Model

Following confirmatory factor analyses and the assessment of the reliability and validity of exogenous and endogenous constructs, the study assesses the structural model. The structural model denotes the internal framework illustrating the interconnections among latent variables (Hair et al., 2011; Henseler et al., 2012). The study adheres to specific procedures in this context. Initially, multicollinearity has been assessed. The estimated variance inflation factor (VIF) values indicate that the model constructs are largely devoid of multicollinearity issues, as the majority of VIFs are less than 3, with two factors approximately at 5 (Hair et al., 2010, Hair et al., 2011; Hair et al., 2014; Hair et al., 2016; Kock and Lynn, 2012, Diamantopoulos and Siguaw, 2006).

Secondly, regarding the coefficient effect, the exogenous construct SES1 exerts a positive influence (0.414) on the endogenous construct GII. The exogenous construct SES2 exerts a negative influence (-0.737) on GII. The beneficial impact of SES1 indicates that the first generation undergoes downward mobility, whereas the detrimental impact of SES2 suggests that the second-generation experiences upward mobility. Current research indicates that the SES of the first generation may exacerbate gender inequality, while the SES of the second generation may mitigate it (Archambault et al., 2017; Tezcan, 2019). The research indicates that all structural path coefficients between SES1 and GII, SES1 and MOBIN, SES2 and GII, and SES2 and MOBIN are significant at the 1 percent level, with T-statistics above 2.58 and P values less than or equal to 0.01. Nonetheless, we cannot ascertain any correlation between MOBIN and GII due to its lower T-statistics and elevated P-value. (see Table 2 and Fig. 3 for more details).  

Table 2: Results of Estimated Path Coefficients

[Source: Field Data]

[Source: Field Data]

Thirdly, the coefficient of determination (R²) value of 0.23 signifies that 23 percent of the variation in the endogenous latent variable GII is elucidated by the corresponding exogenous variables SES1 and SES2, mediated by MOBIN. Iqbal (2020) posits that an R² value between 0.20 and 0.30 is deemed acceptable in the absence of variables in the model or index. Consequently, the proposed model is deemed acceptable. Fourthly, the effect size (f²) serves as a tool for evaluating the impact of exogenous constructions on endogenous constructs. Cohen (1988) posits that effect sizes of 0.02, 0.15, and 0.35 correspond to minor, medium, and large effects of exogenous variables on endogenous constructs, respectively. The current investigation indicates that the effect size of SES1 on MOBIN (2.444) is substantial.

However, the adverse impact of the structural coefficient suggests that the parental generation has undergone a decline in socioeconomic status mobility. The effect size of SES2 on MOBIN (1.209), coupled with the positive structural coefficient, suggests that the current generation of children has undergone upward socioeconomic mobility. The effect size of SES1 on GII is minimal at 0.024. However, the positive structural coefficient effect suggests that the parent generation has encountered gender discrimination. The effect magnitude of SES2 on GII (0.118), coupled with the negative structural coefficient, suggests that the generation has undergone gender equality. However, the effect magnitude of MOBIN on GII (0.000) has yet to be observed. Finally, according to the Blindfolding approach, the estimated predictive power or relevance of the endogenous construct (Q²) indicates that GII2 possesses substantial predictive power and relevance, as the Q² value is 0.229, exceeding the threshold value of 0. The calculated value aligns with the proposed Q² value of the Stone-Geisser model (Geisser, 1974; Stone, 1974). 

The aforementioned assessments of the social mobility model regarding GII suggest that it aligns with the Blau and Duncan (1967) model established in the American environment of the 1970s. The correlation between MOBIN and the GII indicates that the MOBIN remains inadequate and fails to mitigate gender inequality in Bangladesh. MOBIN, as a mediating variable, does not significantly contribute to the explanation of GII. The inverse link between SES1 and SES2 likely stems from the absence of universalism, as demonstrated by the examination of Blau and Duncan's (1961) predictions and EGP class mobility (i.e., OED triangle) and class framework. Blau and Duncan's mobility model is closely associated with traditional sociologists, like Durkheim, Marx, and Weber. The study findings suggest that gender inequality, as a component of social inequality, can be elucidated through Marx and Weber’s concepts of economic status mobility and Durkheim’s notions of educational and occupational status mobility. From a Durkheimian perspective, the socioeconomic level (SES) of parental generations parallels educational and occupational status in mechanical solidarity, as both are ascribed at birth. Conversely, SES2 in child generation exhibits parallels with the educational competencies and occupational division of labour found in organic solidarity, where both are attained rather than ascribed. According to Durkheim, while organic solidarity initially fosters educational and vocational stratification, the process of modernisation will promote universalism, ultimately diminishing inequality. From that perspective, the SES1 was significant, as mothers exhibited lower mobility compared to fathers. The average SES score of mothers (2.26) is lower than that of fathers (2.47). The average educational scores for mothers and fathers were 1.50 and 1.85, respectively. The average ratings for mothers and fathers for occupation were 4.19 and 4.20, respectively. The average income scores for mothers and fathers were determined to be 1.09 and 1.36, respectively. In mechanical solidarity, mothers had decreasing trends in educational, vocational, and financial mobility relative to fathers. Conversely, the SES2 of the child generation was determined to be organic as a result of the upward mobility of both sons and daughters. However, the proportion of daughters was determined to be lower than that of sons, as the mean SES ratings for daughters and sons were 3.78 and 4.08, respectively. The mean educational scores for daughters and sons were 3.47 and 3.86 respectively; the mean occupational scores were 5.99 for daughters and 6.47 for sons; and the mean income scores for both daughters and sons were 1.91. Consequently, while both daughters and sons exhibit upward mobility relative to their parental generation, gender inequality persists within Bangladeshi organic solidarity. In terms of the Gender Inequality Index (GII), the average score for daughters (3.45) is lower than that for sons (3.7). In a similar vein, SES2 is juxtaposed with Marx's capitalist modernity and Weber's rational modernity, whilst SES1 is contrasted with Marx's primitive communism and Weber's non-rational traditional society. Consequently, the study suggests that theories of social mobility are not novel. Instead, they are grounded in classical sociological theories, however frequently overlooked in mobility studies. This study advocates for the re-examination of classical theories of social mobility, as the topic of gender disparity can be effectively analysed through the three primary classical views, despite their lack of statistical testing on upward or downward mobility across various social classes.

6. Conclusion and Recommendations

The primary objective of this study was to assess the impact of intergenerational social mobility on gender equality in Bangladesh.

6.1 Summary of the Findings

The findings reveal a significant negative correlation between the socioeconomic status of the parental generation (SES1) and the mobility index (MOBIN), thereby supporting Hypothesis 1. This suggests that lower SES in the parental generation is associated with restricted upward mobility.

In contrast, the socioeconomic status of the child generation (SES2) shows a positive and significant relationship with the mobility index, confirming Hypothesis 2. This indicates improved mobility among the offspring compared to their parents, further supporting Hypothesis 3, which posits higher socioeconomic mobility in the child generation.

Moreover, SES1 demonstrates a strong positive effect on the gender inequality index, validating Hypothesis 4. This implies that lower parental SES contributes to the persistence of gender inequality. Conversely, SES2 has a significant negative effect on the gender inequality index, as stated in Hypothesis 5, indicating that improved socioeconomic conditions among the younger generation contribute to reducing gender inequality in Bangladesh.

6.2 Recommendations for the Policymakers

Based on the estimated results, this study has some implications for the policymakers of our government. First, the policymakers and government should take proper steps to boost mobility and increase universalism without compromising the past generations and ethnic minority groups in Bangladesh since the MOBIN is hitherto not satisfactory in level and cannot reduce gender inequality significantly. Secondly, administrators should prioritize the educational, occupational and income aspects of all generations, groups, and communities of Bangladesh. The reason is that educational mobility boosts occupational mobility which further boosts income mobility. These three kinds of mobility are the root mobility indicators.

6.3 Limitations and Future Directions

Despite following the scientific principles and steps of social research, the present study has some limitations. The first limitation is to formulate MOBIN as a function of the educational score of child generation (consisting of 50 percent sons and 50 percent daughters) divided by parent generation (consisting of 50 percent fathers and 50 percent mothers) multiplied by the occupational score of child generation divided by parent generation multiplied by income score of child generation divided by parent generation multiplied by 1 divided by 3 (i.e., (CES/PES) (COS/POS) (CIS/PIS) (1/3)). But, if the MOBIN was formulated as a function of each score of 100 percent fathers divided by 100 percent mothers (i.e., (FES/MES) (FOS/FOS) (FIS/FIS) (1/3)), the study might have produced another dimension of MOBIN from a gender perspective. So, future researchers are suggested to formulate MOBIN following the second alternative. Secondly, the study was based on the respondents from three districts of Bangladesh, adopting multi-stage sampling where the final step follows a purposive sampling technique to fulfil the inclusion criteria of the respondents within the limited study fund. However, if the number of districts had expanded, the final step could have followed random sampling which would be a more nationally representative study. In this regard, future researchers are recommended to expand the number of districts to follow the random sampling in each step of the multi-stage sampling technique. To do so, the researchers are recommended to conduct such a study having the necessary funds. Thirdly, gender equality status was explained only from a social mobility perspective though there are many other aspects of this kind of study. For example, there are scopes of explaining GII based on the perspectives of marriage, political pressure, displacement, climate change, and hazardous conditions. The final limitation is to include only SES mobility developed by Blau and Duncan though there are some other aspects of SES mobility (e.g., CAPSES aspect where SES is defined as a function of three capital factors such as Human Capital, Material Capital, and Social Capital). Furthermore, future researchers are recommended to conduct gender equality status depending on the CAPSES factors developed by Oakes and Rossi (2003). So, social scientists should conduct further studies in this field to move from the ‘intergenerational social mobility and gender equality status in Bangladesh’ to the ‘gender mobility and Bangladesh society’ to the ‘gender mobility and global society.

Competing Interests: The authors declare that they have no competing interests.

Funding: This research received a grant from the Centre for Higher Studies and Research (CHSR) of Bangladesh University of Professionals (BUP).

Acknowledgment: We are grateful to the respondents who participated in the study and formed our sample for collecting and analyzing the data.

Informed Consent Statement/Ethics Approval: Not applicable.

Declaration of Generative AI and AI-assisted Technologies: This study has not used any generative AI tools or technologies in the preparation of this manuscript.