Does More Education Always Mean Higher Earnings? Evidence from Bangladesh

Abstract

Essentially championed by the human capital theory, it is a widely accepted notion that the higher an individuals’ education level, the more likely they are to earn. However, is this true in the case of different groups of individuals across the type of education pursued, gender, and whether an individual grew up in a rural or urban area? This paper explores this question and tests whether there is a significant increase in earnings for every additional education level in the case of six particular groups of individuals: Individuals who studied in the general education stream, individuals who studied in the religious education stream, males, females, those who grew up in urban areas and those who grew up in rural areas. Using nationally representative data and ordinary least-squares estimation, this paper finds that individuals who studied in the religious stream and, in some cases, females, do not necessarily experience increased labor market earnings with an increase in their education level. Speculating on possible reasons behind these findings, I argue for criticalities and nuances to be better considered when assessing the change in earnings across education in a particular context.

Keywords

Bangladesh education employment human capital theory religious education I21 I24 I26 J01 J24

Introduction: Background and Aim

One of the most famous theories that explain the relationship between individuals’ education and their labor market earnings is the human capital theory, conceptualized and developed by Schultz and Becker in the United States during the 1960s (Becker, 1964; Schultz, 1961). Since its inception, the theory has been widely used as a guideline on the national level of different countries and has been a driving force behind the increased investment in education in recent years all around the globe. According to the simplest form of the theory, education increases individuals’ productivity and skills, which consequently leads to an increase in their labor market earnings (Becker, 1964; Chattopadhyay, 2012; Schultz, 1961).

Significantly, the human capital theory and subsequently this idea of “more education leads to more earnings” has informed the large “returns (economic returns) to education” literature all around the world (see for example (for an overview of global literature, see for example Psacharopoulos, 1972, 1981, 1985, 1994; for an overview of returns to education in developing countries, see Psacharopoulos & Patrinos, 2018; and for an overview of the returns to education in developed countries, see Gunderson & Oreopolous, 2020). Similarly, there is also a large return to education literature in Bangladesh (see for example, Ahmed & McGillivray, 2015; Asadullah, 2006; Hossain, 1990; Riboud et al., 2007; Shafiq, 2007). The question, however, remains: Does the “more education-more earnings” notion hold in the case of different groups of individuals within a context, e.g., within Bangladesh?

In investigating this important question, Brown et al. (2020) use evidence from more developed countries and find that education does not yield, as the orthodox human capital theory predicts, proportional economic returns for the majority in the labor market-women and racial minorities have lower returns to education compared to non-Hispanic white men. Evidence on the topic from the global south is scarce and inspired by Brown et al. (2020), this paper aims to fill this gap by exploring whether or not individuals’ labor market earnings, i.e., their returns to education, vary across different groups of individuals within Bangladesh—specifically across gender, type of education and whether individuals grew up in rural or urban areas. Thus, in testing whether the “more education-more earnings” notion functions as expected across different groups of individuals, I consider six different groups of individuals: Individuals educated in the general education stream, individuals educated in the religious education stream, males, females, individuals who grew up in a rural area and individuals who grew up in an urban area. In measuring “more education,” I account for additional educational levels beyond a particular education level e.g., primary, secondary, and higher secondary.

Therefore, the research question of this paper stands: Is there a significantly increasing association between additional education levels beyond primary, secondary, and higher secondary education and labor market earnings in the case of separately: Individuals who studied in the religious stream of education, individuals who studied in the general stream of education, males, females, individuals who grew up in a rural area and individuals who grew up in an urban area? The hypothesis of this paper is that more education would yield higher earnings in the case of different groups of individuals, but that the magnitude of the increase in earnings with an increase in education may vary. Therefore, there may be differences in the magnitude of the increase in earnings but there will always-regardless of the magnitude—at least be an increase in earnings with an increase in education level.

The Context of Bangladesh

Bangladesh is an apt and interesting setting for the analysis of this paper because there are different groups of individuals for whom it can be tested whether the “more education-more earnings” notion holds or not. Primarily, I choose to consider individuals across type/stream of education, gender, and whether they grew up in a rural or urban area. While the education stream categorization is important because the religious education stream, i.e., madrasahs, continues to have a substantial and increasing presence in the education system of Bangladesh (Roy et al., 2020), the gender categorization is important because gender equality is still an agenda for the country, especially, in terms of gender wage differentials (Asaduzzaman et al., 2015; Ferdaush & Rahman, 2011; Hossain & Tisdell, 2005), and the rural-urban categorization is important because there remains rural–urban disparities in opportunities which continues to lead to rural-urban migration across the nation (Afsar, 1999; Mortuza, 1992; Osmani & Sen, 2011). Therefore, it is important to test whether the “more education-more earnings” relationship holds across these different groups of individuals disaggregated by gender, type of education, and whether individuals grew up in rural or urban areas. Before proceeding to some more details on these different groups of individuals, let us first look at the education system of Bangladesh.

The education system of Bangladesh has, in general, four education levels: Primary, secondary, higher secondary or college, and tertiary. In Bangladesh, educational institutions can broadly be categorized in two ways: Firstly, according to ownership, e.g., public or private, and secondly, according to the relative emphasis on religious education, i.e., Islamic education, in the curriculum, lessons and mode of teaching. In terms of the latter, institutions are of two types: religious schools (madrasahs) and general schools. There are two types of religious schools in Bangladesh: The government-registered Alia madrasahs and the more independent Qawmi madrasahs that are not government registered.

Significantly, religious (Alia) and general schools are similar—to an extent and not completely—in terms of the subjects they teach, the syllabus within subjects and the textbooks of general subjects such as Mathematics and English. The main difference between the two education streams is that madrasahs place more emphasis on Islamic education, and in particular on Quranic teachings. In Bangladesh, the two different education streams—religious and general—run parallel to each other from the primary to the tertiary level of education. The two different education streams—religious (Alia) and general–are focal points in this paper. Due to the lack of data on qawmi madrasahs, I am only able to focus on Alia madrasahs in this paper and hereon refer to Alia madrasahs as madrasahs throughout the rest of the paper.

The general education system in Bangladesh is larger than the madrasah education system in Bangladesh in terms of both the number of institutions and the number of participation in education at different education levels (BANBEIS, 2023). However, there is little to no information comparing numbers and statistics across the two education streams, and there are even discrepancies in the percentage of madrasah-educated individuals in Bangladesh, especially in comparison to general-educated (Asadullah & Chaudhury, 2016; Roy et al., 2020). According to estimates from 2009, madrasas accounted for approximately 13.8% of primary education enrollment and about 21% of secondary education enrollment (World Faiths Development Dialogue and Georgetown University's Berkley Centre for Religion, Peace and World Affairs, 2015). Moreover, there is also little to no information on the labor force participation rate across madrasah-educated and general-educated individuals. Figure 1 shows the percentage of madrasah-educated and general-educated individuals at different levels of education, and it is evident that in both cases, the percentage of individuals who pursued formal education is higher at earlier levels of education and lower at higher levels of education.

Figure 1.

Percentage of madrasah-educated and general-educated populations at different levels of education in Bangladesh.

In terms of gender composition, around 50.4% of the population of Bangladesh is female, and interestingly, education statistics—in terms of both enrollment rates and completion rates—seem to be slightly better for females than males (males also have slightly higher dropout rates) at the primary and secondary levels-although the opposite is observed in the case of higher secondary and tertiary levels (BANBEIS, 2023; Bangladesh Bureau of Statistics et al., 2017; Women and Men in Bangladesh: Facts and Figures, 2022, 2022; World Bank, 2023b). On the other hand, in terms of the gender composition in the labor force, the labor force participation rate of the total population was 56% in 2016, of which the male and female labor participation rates were respectively 79.4% and 33% (World Bank, 2023a).

Figure 2, below, shows the percentage of males and females at different levels of education and two facts are evident—(1) the percentage of both males and females pursuing education decreases at higher levels of education, (2) the percentage of females in education is higher than males at earlier levels of education but lower at higher levels of education.

Figure 2.

Gender distribution (percentage of males and females) at different levels of education in Bangladesh.

Interestingly, in terms of rural-urban populations, it is worth keeping in mind that a comparatively lower percentage of the population pursue higher education, especially in rural areas. Consider, for example, the statistics in Figure 3. It is evident that only a small percentage of individuals attend higher education, while the percentage is higher in the case of earlier levels of education. In terms of labor force participation rates, the urban labor force participation rate was 55.7% in 2017, while this statistic was 59.3% in the case of rural populations (Bangladesh Bureau of Statistics, 2023a, 2023b).

Figure 3.

Percentage of rural-urban population at different levels of education in Bangladesh.

While there is information separately on the education participation (although unclear in many cases e.g., madrasah education) and labor force participation rates of the different groups of individuals under focus in this paper, there is little to no information on how employment statistics vary across education, and a particular gap, in this case, is whether or not individuals earnings’ indeed increases with an increase in education level in the case of these different groups of individuals. Intrinsically, filling this gap is the main focus of this paper.

Situating the Paper in Existing Literature

For the particular context of Bangladesh, this paper can be situated in the gradually expanding “returns to education” literature, and on a wider and more global scale—in the literature that focuses on testing the notions of the human capital theory, e.g., challenging the orthodox human capital theory as Brown et al. (2020) does. In essence, this paper situates itself in the “returns to education in Bangladesh” literature, but focuses on a different angle in addressing the topic: whether the association between higher education levels and increased earnings holds across different groups of individuals in Bangladesh. By and large, this paper focuses on a conceptual advancement in the topic of “returns to education” in Bangladesh.

Admittedly, although there is a scarcity of literature on whether the central notion of the human capital theory—i.e., that more education leads to more earnings—holds in the case of different groups of individuals, there is a substantive body of literature on the relationship between individuals’ education and their earnings, i.e., on the returns to education (see for example, Ahmed & McGillivray, 2015; Asadullah, 2006; Hossain, 1990; Riboud et al., 2007; Shafiq, 2007). Most of these studies focus on the relationship between earnings and education level controlling for different socio-demographic characteristics such as gender, but some studies—although limited in number—do explore the returns to education from a gendered and rural-urban perspective. For instance, Asadullah (2006) assesses returns to education disaggregated by gender and by rural-urban area, and finds that the returns to education are higher for individuals living in urban areas compared to those living in rural areas, and also higher for male individuals compared to females. Strikingly, there is no evidence as of yet on the returns to education across different streams of education—madrasah and general—in Bangladesh, and with an aim to fill this gap, this literature begins by exploring whether there is an increase in earnings with every additional education level in the case of separately madrasah-educated and general-educated individuals, males and females, and individuals from rural and urban areas. As such, this paper complements the existing literature on the returns to education in Bangladesh.

Interestingly, most of the more recent “returns to education” literature in Bangladesh have mostly been updating of preceding research in terms of data and methodology, but not in terms of perspective- and this is where this paper is different. This paper introduces an overlooked angle in viewing the topic of the relationship between education and earnings, and in doing so, brings to question whether the human capital theory—in its narrowest economic sense—holds in the case of different groups of individuals within the context of Bangladesh. However, unlike other more recent studies on the returns to education in Bangladesh, this paper does not offer a methodological advancement. The methodological advancement of later studies stems from the methodological weakness, identified in earlier studies, that there could be a problem of endogeneity and more specifically a problem of omitted variable bias when assessing whether education is correlated with earnings—for example, individuals’ labor market earnings could be influenced by unobservable characteristics such as an individual's ability which is often difficult to account for in simpler models of estimation. To mitigate this, later studies on the returns to education in Bangladesh use more sophisticated techniques such as the two-step-Heckman model and two-stage least squares instrumental variable approach (see for example, Asadullah, 2006; Kolstad et al., 2014; Mamun et al., 2021; Rahman & Al-Hasan, 2018; Taposh & Lin, 2009).

In this paper, I keep the analysis simple at the ordinary least squares (OLSs) estimation level and present the findings as groundwork for further research to be conducted through the building of more reliable and comprehensive datasets. Arguably, this does not take away from the fresh conceptual perspective that this paper provides by advocating for acknowledging overlooked criticalities, contextualization, and nuances when thinking about the relationship between education and earnings in a global south context such as Bangladesh. In conceptualization, there has been a lack of focus on whether or not returns to education varies, and if so, how it varies, across different groups of individuals within Bangladesh. This paper aims to fill this missing gap and complement the existing “returns to education” in Bangladesh literature by repositioning the focus to exploring whether each additional level of education yields, as promised by the core notion of the orthodox human capital theory, similar returns—i.e., labor market earnings—in the same way in the case of different groups of individuals.

Clarifying the Position of This Paper: Why is this Paper Important?

The central argument, put forth by Brown et al. (2020), that this paper aims to test is that the core notions and premises of the human capital theory do not hold in the same way across different groups of individuals, and in testing this—this paper focuses on the particular context of Bangladesh. Essentially, this paper asks whether the increase in earnings as individuals move from one level of the education system to a higher level varies from group to group. I draw from the notions of the human capital theory in the assumption that more education, usually, tends to lead to increased earnings. By focusing the analysis on the context of Bangladesh—a country from the global south—this paper contributes to the literature, which is limited in the global south, on how the impact of education on earnings may vary from case to case.

Fundamentally, the notion that more education leads to, generally, an increase in earnings is the underlying premise of this paper. To avoid being misunderstood, it is worth making an important clarification at this point. The “more education-more earnings” notion may be a simplified characterization of the human capital theory, but it is not to reject that the relationship between an individual's level of education and their labor market earnings is a complex one depending on a broad range of factors—individual characteristics, institutional and market factors. This paper recognizes that the relationship between individuals’ earnings and their education, i.e., the returns to education, may be expected to vary across individuals and settings, but its hypothesis stems from the expectation that there is usually an increase in earnings with an increase in education level, with “increase” being the key focus here. Thus, the expectation is that more education would lead to an increase, regardless of the magnitude, in earnings across all the groups of individuals under focus—thus, that there will always be a positive correlation between education levels and earnings, thereby implying that the more educated a person is, the higher their earnings are likely to be.

Importantly, this paper argues that it is important that we, in the study of the relationship between education and labor market earnings in Bangladesh, are aware of nuances relevant to different groups of individuals within Bangladesh, especially when it comes to the human capital theory. Particularly, although the application of the notions of the human capital theory has been accepted, implemented, and widespread without challenge in most countries like Bangladesh, the development of the theory reflected a very specific economic context—and thus failed to consider other particular contexts, contextualization or nuances—and so, assuming its generalizability is problematic (Marginson, 2019). Recognizing and acknowledging that the human capital theory continues to be used, in countries such as Bangladesh, in a prescriptive rather than a descriptive and contextualized way (Wheelahan et al., 2022), this paper sets out to conceptually complement the way we explore and understand the relationship between individuals’ education and their earnings in Bangladesh, mainly by considering different groups of individuals in the analysis.

The rest of the paper unfolds as follows. Section 2 discusses the data and highlights the empirical approach. Section 3 presents the findings. Section 4 discusses the implications of the findings, and then concludes.

Data and Methodology

Data

For the analysis in this paper, I use the Household Income and Expenditure Survey (HIES) data- a cross-sectional nationally representative dataset from 2016. The HIES 2016 data is the latest in the series of “Household Income and Expenditure Surveys” conducted by the Bangladesh Bureau of Statistics. The survey followed a stratified two-stage cluster sampling design and provided household-level data on household income, expenditures, assets, housing conditions, as well as individual-level data on education, employment, health, and disability (Bangladesh Bureau of Statistics, 2017). In the HIES 2016, a total of 46,076 households were randomly selected from eight divisions and sixty-four districts, with 32,096 households from rural areas and 13,980 households from urban areas. Each household may consist of multiple members including parents, children, and extended family. The total number of observations as individuals (members) is 1,86,076 according to HIES 2016, and the data collection was completed in around one year (1 April 2016 to 31 March 2017).

In the case of this paper, the final sample size for the analysis of the association between education and earnings for different groups of individuals reduces significantly to 55,211 observations because only a proportion of the total sample of individuals is employed and thus, in the labor market and obtaining labor market earnings. Those excluded from this sample of employed individuals include school-going children who are not yet of working age and also individuals who are of working age but are unemployed.

Variables of Interest

The main variables of interest are shown in Table 1. The dependent variable is “hourly labor market earnings” and the main covariates of interest are three types of “level of education” variables: (1) Level of education with primary education as the base category and secondary, higher secondary and tertiary levels as the additional higher levels of education beyond primary; (2) level of education with secondary education as the base category and higher secondary and tertiary levels as the additional higher levels of education beyond secondary; and (3) level of education with higher secondary education as the base category and tertiary education as the additional higher level of education beyond higher secondary.

Table 1.

Main Variables of Interest.

Variable	Type of variable	Categories in the case of a categorical dummy variable
Level of education¹ (Base category: Secondary education)	Covariate; Multinomial	0 if primary completed (up to grade five);
Level of education¹ (Base category: Secondary education)	Covariate; Multinomial	1 if secondary completed(grades six to ten) 2 if higher secondary completed (grades eleven and twelve) 3 if tertiary completed (university)
Level of education² (Base category: Higher secondary education)	Covariate; Multinomial	0 if secondary completed(grades six to ten)
	Covariate; Multinomial	1 if higher secondary completed(grades eleven and twelve) 2 if tertiary completed (university)
Level of education³ (Base category: Tertiary education)	Covariate; Binomial	0 if higher secondary completed (grades eleven and twelve)
Level of education³ (Base category: Tertiary education)	Covariate; Binomial	1 if tertiary completed (university)
Labor market earnings	Dependent variable/main outcome of interest; Continuous	Hourly labor market earnings in the Bangladeshi currency (Taka/BDT) as a continuous variable; No categories

Note. Household Income and Expenditure Survey (HIES) 2016 data was used to construct these variables.

Thus, the outcome of interest for the analysis at hand is individuals’ labor market earnings. As a result of several wage-related questions in the HIES questionnaire, the labor market earnings variable consists of the total take-home monthly renumerations after all deductions at the source and also includes in-kind payments and other benefits (for example, tips, bonuses, transportation cost) in the job. Using the information on the number of days an individual works per month and the average number of work hours per day, the labor market earnings variable I construct accounts for hourly earnings in the Bangladeshi currency-taka (BDT).

The main covariates of interest are each of the three education-level variables. Then, to constitute the different groups of individuals, the gender, education stream, and rural-urban variables are used, each of which is a dichotomous dummy variable: (1) Gender (0 if male and 1 if females); (2) stream of education (0 if general education and 1 if religious education); and (3) whether the individuals’ area of upbringing was rural or urban (0 if rural and 1 if urban).

Besides the main variables of interest related to individuals’ education and earnings, I assume that the association between education and earnings would plausibly vary across different types of jobs and different job sectors. And so, I also explore the association between education and earnings across different groups of individuals further disaggregated by type of job (day-laborer work, salaried work, or working as self-employed/an employer) and job sectors (agricultural or nonagricultural). Moreover, I control for a number of socio-demographic individual characteristics such as age, gender, stream of education, whether the individuals’ area of upbringing was rural or urban, religion, fathers’ education level, mothers’ education level and regional variables.

Importantly, the fathers’ education and mothers’ education variables are considered as key family background variables in this analysis, i.e., as a proxy for individuals’ socio-economic backgrounds. One issue with using parents’ education variables in the case of the data at hand is that there is a lot of missing data in the case of parents’ education information, and this may be because respondents’ do not know their parents’ level of education. Additionally, it could also simply be a matter of some respondents’ being reluctant to share their parents’ education information. To deal with this missing data, I recategorize each of the fathers’ and mothers’ education variables to include an “unknown” category, which constitutes the respondents whose parents’ education information is not available. This, however, does not mean that the information is not applicable to these respondents. To clarify, in this case, it is justified to construct an additional “unknown” category because respondents would have parents who either went to formal education or did not, and it is merely the case that this information is unknown in the case of these particular respondents.

Summary and Descriptive Statistics

Table 2, below, shows some selected summary statistics most relevant to the analysis of individuals’ labor market earnings across different education levels, and Table A1 in the appendix presents the full summary statistics for all the variables used in the analysis. Both tables present the summary statistics of different variables disaggregated across streams of education, gender, and whether the individuals’ upbringing was in a rural or urban area.

Table 2.

Selected Summary Statistics Across Stream of Education, Gender and Whether the Area of Upbringing was Rural or Urban.

	Full sample	Nonreligious education	Religious education	Male	Female	Rural	Urban
	(N = 55211)	(N = 34213)	(N = 1103)	(N = 47094)	(N = 8062)	(N = 37618)	(N = 17539)
Labor market returns (hourly)
Mean (SD)	50.29 (172.22)	56.23 (192.48)	57.03 (83.73)	50.10 (129.23)	51.35 (322.90)	44.65 (134.71)	60.64 (225.05)
Gender
Male	47094 (85.4%)	30012 (87.7%)	1015 (92.0%)	47094 (100.0%)	0 (0.0%)	32735 (87.0%)	14359 (81.9%)
Female	8062 (14.6%)	4201 (12.3%)	88 (8.0%)	0 (0.0%)	8062 (100.0%)	4883 (13.0%)	3179 (18.1%)
Stream of education
Nonreligious education	34213 (96.9%)	34213 (100.0%)	0 (0.0%)	30012 (96.7%)	4201 (97.9%)	21789 (96.2%)	12424 (98.0%)
Religious education	1103 (3.1%)	0 (0.0%)	1103 (100.0%)	1015 (3.3%)	88 (2.1%)	850 (3.8%)	253 (2.0%)
Area of upbringing
Rural	37618 (68.2%)	21789 (63.7%)	850 (77.1%)	32735 (69.5%)	4883 (60.6%)	37618 (100.0%)	0 (0.0%)
Urban	17539 (31.8%)	12424 (36.3%)	253 (22.9%)	14359 (30.5%)	3179 (39.4%)	0 (0.0%)	17539 (100.0%)
Education level
Less than primary or primary completed	26094 (73.9%)	25453 (74.4%)	638 (57.9%)	23019 (74.2%)	3075 (71.7%)	17990 (79.5%)	8104 (63.9%)
Secondary completed	3562 (10.1%)	3405 (10.0%)	157 (14.2%)	3204 (10.3%)	358 (8.3%)	2071 (9.1%)	1491 (11.8%)
Higher secondary completed	2819 (8.0%)	2701 (7.9%)	118 (10.7%)	2414 (7.8%)	405 (9.4%)	1497 (6.6%)	1322 (10.4%)
Graduate/Tertiary completed	2843 (8.0%)	2654 (7.8%)	189 (17.2%)	2392 (7.7%)	451 (10.5%)	1081 (4.8%)	1762 (13.9%)

Note. This table shows the summary statistics for the sample that includes only employed individuals. In the case of the continuous variables, we report the mean and standard deviation. In the case of the categorical variables, we report the number of observations for each category and the corresponding percentage in the parentheses. The first column presents the statistics in the case of the entire sample. Columns 2-7 present the statistics in the case of disaggregation across stream of schooling, gender and whether the area of upbringing was rural or urban. The variables in the case of each disaggregation are binary variables, whereby the variables stream of schooling, gender and rural-urban each take the value 1 if the individual attends the religious stream of schooling, if the individual is female and if the individual grew up in an urban area.

Interestingly, there is not much variation in the mean hourly labor market earnings across madrasah-educated and general-educated individuals, and the same low gap in earnings is observed in the case of comparing males’ earnings’ and females’ earnings. Surprisingly, the mean hourly earnings are slightly higher for individuals who studied in the religious stream than for individuals who studied in the general education stream and also slightly higher for females compared to males. On the other hand, there is a more substantial gap in the mean hourly earnings between individuals who grew up in rural areas and individuals who grew up in urban areas- individuals who grew up in urban areas have higher mean earnings (around 60 BDT per hour) than individuals who grew up in rural areas (around 44 BDT per hour), making the gap between the two groups to be almost sixteen taka, on average, per hour.

Since the aim of this paper is to assess whether individuals’ earnings significantly increase with the increase of an education level, Figure 4, below, plots the mean hourly earnings across education levels respectively across streams of education, gender, whether the area of upbringing was rural or urban and across both gender and stream of education.

Figure 4.

Mean hourly earnings at different education levels for different groups of individuals.

According to Figure 4, there is, generally, patterns suggesting that individuals’ earnings are higher in the case of higher education levels across all the different groups of individuals, but the gradient seems to be lower in the case of individuals who studied in the religious stream of education and, in some cases, females—especially females who studied in the religious stream of education. Nonetheless, there is a clear gradient suggesting that the “more education-more earnings” notion does hold across the different groups of individuals. The question that now arises is: Will this pattern hold in the case of a more rigorous regression analysis?

Empirical Approach: OLSs Estimation

To identify whether there is any significant association between each “level of education” variable and individuals’ labor market earnings, I use OLSs estimation with “if conditions” to specify the different groups of individuals, e.g., individuals who studied in the religious stream of education, individuals who studied in the general education stream, males, females, individuals who grew up in a rural area and individuals who grew up in an urban area. In each of the models, I control for several socio-demographic factors such as age, gender, stream of education, area of work, type of job, job sector, religion, and region. The functional form of the OLS models are as follows:

For the full sample of respondents:

L_{i} = β_{0} + β_{1} X_{i} + β_{2} Y_{i} + ε_{i}

(1)

For the different groups of individuals:

L_{i} = β_{0} + β_{1} X_{i} + β_{2} Y_{i} + ε_{i} if α_{i} = 0 | α_{i} = 1

(2)

For the different groups of individuals in the case of different types of jobs and job sectors:

L_{i} = β_{0} + β_{1} X_{i} + β_{2} Y_{i} + ε_{i} if α_{i} = 0 | 1 & α_{2} = 0 | 1 =

(3)

Notedly, equation 1 represents the OLS model in the case of the full sample, equation 2 represents the regression models in the case of the different groups of individuals using the “if” condition, and equation 3 represents the regression models in the case of different groups of individuals for different types of jobs and job sectors using multiple “if” conditions. In the case of all three equations,

L_{i}

is the hourly labor market earnings of the individual i;

α_{i}

represents either the stream of education (general or religious) or gender (males or females) or the area of upbringing (rural or urban) for individual i;

α_{2}

represents either the type of job (daily-laborer work or salaried work/work as employees or self-employed work/working as employers) or the job sector (agricultural sector or nonagricultural sector) of individual I; X_i is the level of education of individual i, Y_i is the vector of observable characteristics of individual i and ε _i is the error term of individual i.

Results: What Does the Evidence Tell us?

Tables 3–5 present the results of the analysis of the differences in earnings across education levels in the case of different groups of individuals and also disaggregated by type of job and job sector. Strikingly, across all the results in Tables 3–5, there is no significantly increasing association between education level and earnings at any additional level of education—either beyond primary, secondary, or higher secondary. The results in terms of day-laborers and agricultural sector workers makes some sense because these jobs usually require more manual labor skills than intellectual skills and so, higher levels of education perhaps may not be helpful to perform in such jobs. On the other hand, I find the results in terms of self-employed workers surprising and interesting, and I contend that there is a need to better define what is meant by “self-employed” by taking into account all possible nuances (e.g., a tea-stall owner and a multi-business owner billionaire CEO would both be categorized as self-employed but this would be an under-nuanced categorization).

Table 3.

Differences in Earnings to Higher Levels of Education Beyond Primary Education for Different Groups of Individuals.

		Labor market earnings (BDT per hour)
	[Base education level: Primary education]	All	Madrasah- educated	General- educated	Males	Females	Based in a rural area	Based in an urban area
Full sample	Secondary	11.701***	10.224	11.701***	11.507***	4.304	14.034***	8.587**
		(2.588)	(9.997)	(2.658)	(2.830)	(6.309)	(3.193)	(4.238)
	Higher secondary	30.985***	17.851	31.718***	24.997***	48.304**	27.248***	33.173***
		(4.231)	(11.477)	(4.378)	(3.348)	(20.257)	(3.911)	(7.770)
	Tertiary	62.004***	23.039*	64.418***	64.535***	42.420***	48.988***	68.101***
		(4.794)	(13.033)	(5.005)	(5.498)	(8.109)	(5.630)	(7.381)
	Observations (N)	21380	697	20683	18679	2701	12780	8600
Day-laborer work	Secondary	2.580	−0.615	2.822	3.308	−11.717	2.923	2.678
		(2.501)	(3.109)	(2.621)	(2.325)	(20.540)	(1.930)	(6.357)
	Higher secondary	4.522	29.849	3.436	6.784***	−51.808	5.747	1.353
		(3.032)	(18.769)	(2.795)	(2.315)	(53.544)	(3.872)	(5.590)
	Tertiary	14.437*	1.106	15.231*	12.875**	7.178	9.817	16.204
		(7.465)	(6.659)	(8.420)	(5.605)	(74.705)	(6.786)	(14.372)
	Observations (N)	9839	242	9597	9138	701	7263	2576
Salaried work or working as employees	Secondary	11.673***	15.010	11.384***	10.695***	11.379***	14.323***	9.799*
		(3.253)	(16.589)	(3.265)	(3.642)	(3.239)	(3.208)	(5.492)
	Higher secondary	33.211***	19.427	34.055***	25.542***	64.245***	31.146***	35.025***
		(5.161)	(13.921)	(5.328)	(4.031)	(20.403)	(4.633)	(9.181)
	Tertiary	60.147***	25.632	62.282***	61.582***	50.361***	49.926***	66.149***
		(5.284)	(16.732)	(5.490)	(6.199)	(4.603)	(6.276)	(8.216)
	Observations (N)	10700	431	10269	8789	1911	5000	5700
Self-employed work/ working as employers	Secondary	56.878*	−26.244	59.704*	60.551	15.005	103.425	6.149
		(33.625)	(18.784)	(36.014)	(36.815)	(16.313)	(64.099)	(12.579)
	Higher secondary	48.073**	88.784*	49.863**	52.575**	25.974	39.946*	51.942
		(21.179)	(43.040)	(21.557)	(24.504)	(17.329)	(22.612)	(32.850)
	Tertiary	92.481**	−9.730	97.413**	104.277**	31.559*	27.874***	132.834*
		(40.674)	(16.390)	(42.420)	(47.321)	(16.569)	(8.991)	(67.671)
	Observations (N)	841	24	817	752	89	517	324
Agricultural sector work	Secondary	9.154	2.722	9.953	10.788	−33.964	14.079	−20.749
		(10.950)	(6.175)	(11.558)	(10.995)	(129.309)	(12.419)	(25.423)
	Higher secondary	−0.672	12.277	−1.020	3.870	−98.434	−0.792	−12.161
		(6.300)	(11.479)	(6.479)	(2.568)	(235.236)	(4.430)	(32.027)
	Tertiary	46.988	1.636	51.313	44.759	−232.433	23.194*	84.901
		(31.314)	(5.264)	(35.387)	(32.816)	(342.041)	(13.978)	(89.667)
	Observations (N)	4166	106	4060	3870	296	3697	469
Non-agricultural sector work	Secondary	11.649***	11.229	11.570***	11.294***	11.211***	14.216***	9.644**
		(2.629)	(12.208)	(2.665)	(2.951)	(2.843)	(2.782)	(4.334)
	Higher secondary	33.045***	19.301	33.786***	26.439***	61.477***	30.808***	34.828***
		(4.537)	(12.570)	(4.684)	(3.643)	(18.871)	(4.209)	(7.927)
	Tertiary	62.136***	24.225*	64.462***	64.407***	49.340***	49.522***	68.806***
		(4.874)	(14.321)	(5.073)	(5.733)	(4.232)	(5.844)	(7.482)
	Observations (N)	17213	591	16622	14808	2405	9083	8130

Note. (1) Standard errors in parentheses, (2) * p < 0.10, ** p < 0.05, *** p < 0.01.

Table 4.

Differences in Earnings to Higher Levels of Education Beyond Secondary Education for Different Groups of Individuals.

		Labor market earnings (BDT per hour)
	[Base education level: Secondary education]	All	Madrasah- educated	General- educated	Males	Females	Based in a rural area	Based in an urban area
Full sample	Higher secondary	18.348***	10.046	19.344***	12.946***	51.278**	12.881***	24.276***
		(3.710)	(15.426)	(3.828)	(3.094)	(20.221)	(4.817)	(5.678)
	Tertiary	45.517***	14.117	47.836***	48.410***	39.682***	28.469***	58.065***
		(4.309)	(16.269)	(4.497)	(4.926)	(7.016)	(6.850)	(5.214)
	Observations (N)	5684	302	5382	4849	835	2542	3142
Day-laborer work	Higher secondary	3.884	13.162	2.265	4.098	1.928	5.802*	−0.963
		(3.222)	(19.738)	(3.133)	(3.349)	(11.180)	(3.182)	(7.104)
	Tertiary	11.233	3.176	12.414	8.690	36.758	4.240	16.967
		(7.402)	(8.571)	(8.224)	(6.209)	(60.191)	(6.796)	(13.328)
	Observations (N)	712	43	669	666	46	462	250
Salaried work or working as employees	Higher secondary	22.414***	7.641	23.827***	16.091***	58.642**	18.725***	26.219***
		(4.179)	(18.888)	(4.298)	(3.159)	(24.139)	(5.553)	(6.555)
	Tertiary	48.201***	16.834	50.704***	50.672***	44.594***	34.821***	57.530***
		(3.997)	(19.518)	(4.141)	(4.548)	(9.145)	(7.058)	(4.812)
	Observations (N)	4770	250	4520	4006	764	1996	2774
Self-employed work/ working as employers	Higher secondary	−1.721	−8.836	0.360	−6.409	17.023	−104.597	52.552
		(38.789)	(.)	(40.568)	(45.966)	(25.069)	(93.281)	(40.971)
	Tertiary	43.014	5.915	50.467	46.404	26.506	−80.298	143.465*
		(50.059)	(.)	(51.932)	(56.435)	(19.142)	(70.902)	(80.979)
	Observations (N)	202	9	193	177	25	84	118
Agricultural sector work	Higher secondary	−5.590	18.689	−7.196	−7.402	−0.775	−4.467	−13.191
		(8.126)	(23.572)	(9.391)	(9.831)	(13.555)	(8.649)	(24.723)
	Tertiary	23.519	−1.880	25.250	24.397	−38.980	−1.430	46.813
		(28.267)	(7.095)	(32.379)	(31.676)	(80.238)	(15.981)	(67.745)
	Observations (N)	327	22	305	308	19	260	67
Non-agricultural sector work	Higher secondary	20.609***	10.697	21.651***	14.900***	53.772**	16.147***	24.764***
		(3.801)	(16.016)	(3.896)	(3.034)	(21.060)	(5.070)	(5.821)
	Tertiary	47.064***	16.892	49.350***	49.766***	41.392***	31.584***	57.594***
		(4.132)	(17.088)	(4.281)	(4.746)	(7.367)	(6.643)	(5.128)
	Observations (N)	5357	280	5077	4541	816	2282	3075

Note. (1) Standard errors in parentheses, (2) * p < 0.10, ** p < 0.05, *** p < 0.01.

Table 5.

Differences in Earnings to Tertiary Education Beyond Higher Secondary Education for Different Groups of Individuals.

		Labor market earnings (BDT per hour)
	[Base education level: Secondary education]	All	Madrasah- educated	General- educated	Males	Females	Based in a rural area	Based in an urban area
Full sample	Tertiary	25.469***	5.662	26.526***	33.165***	−11.433	14.847**	33.275***
		(4.567)	(16.222)	(4.739)	(4.535)	(15.736)	(6.644)	(5.960)
	Observations (N)	3765	202	3563	3135	630	1541	2224
Day-laborer work	Tertiary	5.755	72.628	9.870	3.304	39.847	−4.367	20.057*
		(7.277)	(.)	(7.574)	(6.522)	(60.481)	(8.399)	(11.791)
	Observations (N)	224	15	209	204	20	137	87
Salaried work or working as employees	Tertiary	25.462***	11.133	26.504***	33.648***	−12.712	15.690**	32.092***
		(4.662)	(18.859)	(4.826)	(4.490)	(16.430)	(7.173)	(5.898)
	Observations (N)	3411	183	3228	2821	590	1358	2053
Self-employed work/ working as employers	Tertiary	42.129	−1.603	42.463	43.249	8.333	8.371	90.113
		(38.726)	(.)	(38.865)	(41.606)	(17.729)	(18.657)	(74.042)
	Observations (N)	130	4	126	110	20	46	84
Agricultural sector work	Tertiary	28.365	−37.776	34.531	32.755	116.648	0.607	41.039
		(28.461)	(.)	(30.021)	(32.640)	(158.366)	(4.306)	(71.398)
	Observations (N)	131	9	122	120	11	90	41
Non-agricultural sector work	Tertiary	25.063***	8.011	26.074***	32.753***	−11.616	14.349**	32.364***
		(4.658)	(17.694)	(4.819)	(4.599)	(15.971)	(6.962)	(6.001)
	Observations (N)	3634	193	3441	3015	619	1451	2183

Note. (1) Standard errors in parentheses, (2) * p < 0.10, ** p < 0.05, *** p < 0.01.

Table 3, above, shows the labor market earnings to higher levels of education beyond primary education for different groups of individuals. According to Table 3, in the case of the full sample, the results reveal that there is a significant and increasing association between the level of education and labor market earnings in the case of all the groups of individuals except in the case of females when considering the additional level of secondary education beyond primary education and in the case of individuals who studied in the religious stream of education considering all additional education levels beyond primary. For general-educated individuals, males, rural-background individuals, and urban-background individuals, the increase in earnings for secondary educated-individuals compared to primary-educated individuals is respectively, on average, 11.7, 11.5, 14, and 8.5 BDT per hour. Similarly, for general-educated individuals, males, females, rural-background individuals, and urban-background individuals, these statistics are, on average, respectively 31.7, 25, 48, 27, and 33 BDT per hour for higher secondary education beyond primary, and respectively 64.4, 64.5, 42.4, 49, and 68 BDT per hour for tertiary education beyond primary.

In the case of salaried work, there is a significant and increasing association between level of education and labor market earnings in the case of all groups of individuals except madrasah-educated individuals at all additional levels of education beyond primary and in the case of urban-background individuals at the additional secondary level of education beyond primary.

In the case of self-employed work, individuals’ earnings significantly increase with higher secondary and tertiary education beyond primary for general-educated individuals, males, and individuals from rural areas, whereas this was not the case for madrasah-educated individuals, females, and individuals from urban areas. In the case of non-agricultural work, there is a significantly increasing association between higher levels of education beyond primary for all groups of individuals except madrasah-educated individuals.

Similar to the results for additional levels of education beyond primary, Table 4, above, shows the labor market earnings to higher levels of education beyond secondary education for different groups of individuals and finds that there is a significant increase in earnings with every additional level of education in the case of the full sample, salaried work and nonagricultural work for all groups of individuals except madrasah-educated individuals. For instance, in the case of the full sample, for general-educated individuals, males, females, rural-background individuals, and urban-background individuals, the increase in earnings for higher secondary-educated individuals compared to secondary-educated individuals is respectively, on average, 19.3, 12.9, 51.3, 12.9, and 24.3 BDT per hour. Then, for general-educated individuals, males, females, rural-background individuals and urban-background individuals, these statistics are, on average, respectively 47.8, 48.4, 39.7, 28.5, and 58.1 BDT per hour for tertiary education beyond secondary.

Lastly, Table 5, above, shows the labor market earnings for tertiary education compared to higher secondary education for different groups of individuals. In each of the cases of the full sample, salaried work and nonagricultural work, there is a significant increase in earnings for tertiary education compared to higher secondary education for all groups of individuals except madrasah-educated individuals and females. For instance, in the case of the full sample, general-educated individuals, males, rural-background individuals, and urban-background individuals earn significantly 26.5, 33, 14.8, and 32.3 BDT per hour higher earnings at the tertiary education level beyond the higher secondary level.

In sum, across all the results presented in Tables 3–5, there is an absence of a significant and increasing association between additional higher education levels and earnings in the case of madrasah-educated individuals, whereas the significant and increasing association is observed in the case of general-educated individuals. Moreover, in some instances (at the secondary education level beyond primary for the full sample and for self-employed work and at the tertiary level beyond higher secondary for the full sample and for self-employed work as well as for nonagricultural work), there is an absence of a significant and increasing association between additional education levels and earnings in the case of females, whereas the significant and increasing association is observed in the case of males. In two instances (at the secondary level beyond primary for salaried work and at the tertiary level beyond primary for self-employed work), there is an absence of a significant and increasing association between additional education levels and earnings in the case of individuals who grew up in urban areas, whereas the significant and increasing association is observed in the case of individuals who grew up in rural areas.

Therefore, the supposed positive and significant association between the level of education and labor market earnings—a notion stemming from the orthodox human capital theory—is not evident in the case of all groups of individuals. Individuals who studied in the religious stream of education and, in some cases, females do not necessarily experience increased labor market earnings with an increase in their level of education. This is surprising and warrants further exploration. Contrary to the hypothesis of this paper, it is not simply a matter of differences in the magnitude of earnings but that an “increase” in earnings, as expected, is absent in some specific cases. The question is: Why is this the case?

Discussion and Implications

In essence, the aim of this paper was to test the functionality of the “more education-more earnings” notion of the human capital theory across different groups of individuals within Bangladesh—a country that has largely relied on the human capital theory in policymaking and in expanding education. This paper finds, contrary to its hypothesis, that it is the case that certain groups of individuals, e.g., madrasah-educated individuals, do not necessary earn more with each additional level of education. This finding is the key to beginning important discussions in Bangladesh about the differences between educational streams. Overall, by generating evidence that the more education-more earnings notion does not always necessarily hold for madrasah-educated individuals in Bangladesh, this paper draws attention to criticalities overlooked so far, and thus provides food for thought for future research on the topic of the relationship between education and earnings not only in Bangladesh but also in similar global south contexts across the globe.

Why is the Significant and Increasing Association Between an Extra Level of Education and Earnings Absent in the Case of Specific Groups of Individuals?

There could be several reasons behind the difference in the “more education-more earnings” relationship across different groups of individuals. For example, aspects such as structural inequality, information asymmetries, bias, and discrimination are likely to vary across contexts and nuances. It could also simply be a matter of differences in the perception of education and in the purpose of education. Overall, there could be three main sources behind the absence of a significant and increasing association between additional levels of education and earnings in the case of specific groups of individuals, especially in the case of madrasah-educated individuals:

Differences in skills, job network and availability, aspirations, and experiences of discrimination.

Differences at the root—in educational goals and in the perception of the purpose of higher levels of education.

The job market structure could be different for general-educated and madrasah-educated individuals. For example, there are some particular jobs that render madrasah education and training as essential—working as an imam in a mosque, an educator at a moktob (gathering for religious teachings) or madrasah.

It is worth elaborating on the four mechanisms, mentioned above in point 1, as sources of differences in the “more education-more earnings” relationship across difference groups of individuals: (1) Skills, (2) job availability and job search dynamics, (3) labor market discriminations and (4) aspirations.¹

Firstly, individuals’ labor market earnings may be associated with their skills. The assumption that skills may be a mechanism that explains how differences in educational attainment lead to different job market outcomes is supported by the tenets of the human capital theory (Becker, 1964) which posits that human capital differences matter in the attainment of labor market outcomes. Thus, it may be the case that different groups of individuals—e.g., individuals who studied in different education streams—have varying skillsets stemming from their education, which in turn translate into differences in labor market returns.

Secondly, there may be heterogeneity in the jobs and types of jobs available in different regions and to different groups of the population (Mourshed et al., 2012). Different groups of people may also have varying levels of access to job search tools (Huffman & Torres, 2001; Wei-Cheng & Kopischke, 2001). There may also be differences in networks (Falcón, 1995; Montgomery, 1991).

Thirdly, there may be labor market discrimination (Arrow, 2015; Becker, 1957). It has long been documented that certain groups of individuals often face more labor market discrimination than others, for example, colored people more than white people and females more than males (Arrow, 1998; Lissenburgh, 2000).

Finally, different groups of individuals may have, in general, different aspirations in terms of the type of career they want to pursue and this may be related to the careers they actually pursue in reality (Howard et al., 2011; Metz et al., 2009; Naafs & Skelton, 2020). So, it may be the case that individuals who study in the madrasah education stream tend to have certain career aspirations that differ from individuals belonging to the general educational stream.

What are the Most Noteworthy Implications of this Paper?

The most noteworthy implication of this paper is that it highlights the importance of testing notions of Western-driven theories such as the human capital theory across different groups of individuals in different contexts, especially in non-Western contexts such as Bangladesh.

By focusing on whether the more education-more earnings notion holds in the case of different groups of individuals, this paper argues for a more critical approach when applying a particular Western-originated theory such as the human capital theory to a different context, especially in the global south—a context such as Bangladesh. The norm has been to take the human capital theory as automatically applicable, which is true in many cases but without considering that there may be different groups of individuals within a particular context for whom the theory is perhaps not applying as expected. The argument is that these contextual criticalities and nuances be acknowledged, especially as the “returns to education” literature continues to expand in global south countries such as Bangladesh. This argument is where the merit of this paper, albeit modest, lies.

Moreover, another main implication of the findings of this paper, especially for future research, is that it raises the need to focus on the relationship between not only individuals’ education level and their labor market earnings but also between individuals’ type/stream of education—whether general or madrasah in the case of Bangladesh- and their labor market earnings. For instance, there is scope for future research on the topic to investigate whether there is a significant association, and causational relationship, between the stream of education that individuals studied in and their labor market earnings later in life-and this would be the next step to take following this paper.

Furthermore, if the more education-more earnings notion does not hold in the case of day-laborer work, self-employed work in the case of education beyond the secondary level and working in the agricultural sector, this raises the question of whether it is the case that individuals who pursue higher levels of education ultimately pursue these lines of work because they are not able to access other types of jobs and job sectors. That is, the question is whether these individuals who pursue higher levels of education have different job aspirations, and in such a case, the more important question becomes Why are these higher-educated individuals ultimately pursuing jobs that are not yielding higher earnings as expected? There is an important policy implication to consider here, and this is related to the question: If day-laborer work, self-employed work and working in the agricultural sector does not support the “more education-more earnings” notion, then what are the incentives that individuals in these jobs have to pursue higher levels of education?

Considering this question, then, leads us to contemplate whether these jobs require higher levels of education in the first place and whether the jobs can be properly executed with lower educational level qualifications. If it is the case that some jobs require lower levels of educational level qualifications (e.g., the job of a farmer) and other jobs need higher levels of educational level qualifications (e.g., the job of a teacher or doctor), then there may be a need to predict at earlier levels of education the types of jobs and job sectors that students are likely to be able to successfully compete for based on their educational trajectories (e.g., grades, learning curve, skills). This may perhaps allow for a more efficient system where individuals who do not need to pursue higher levels of education for their jobs do not pursue higher education, which would, in turn, allow for focusing more resources on a lower number of students at higher levels of education-students who would need higher educational level qualifications to pursue their jobs. This is, of course, assuming that the purpose of education, and especially higher levels of education, is primarily and dominantly labor-market-oriented, i.e., to get jobs and earn a living. Also, for such a prediction system to work, there are important factors to strongly institutionalize for the system to be bias-free-otherwise it may be the case that larger gaps arise in terms of intergenerational education mobility.

On the other hand, however, there may be a deeper equity issue at play here. As discussed above, it may be the case that the “more education-more earnings” notion is not holding in the case of madrasah-educated individuals, and in some cases females, because there could be differences in skills, job availability and network, experiences of discrimination and differences in aspirations. In the case of the “more education-more earnings” notion not holding for additional levels of education beyond the primary level for both salaried work and self-employed work in the case of individuals who grew up in urban areas, similarly, it could be a case of differences in skills, job availability and network, experiences of discrimination and differences in aspirations. Undoubtedly, this paper activates the need to dig into these issues deeper and consider these contextual criticalities in future research on the relationship between education and earnings in Bangladesh and other non-Western contexts.

Conclusion

In the case of Bangladesh, what has been missing so far is the introduction of criticalities and contextualization in thinking about the linkage between education and employment. Take the phenomenon of human capital investment for example. Human capital investment in Bangladesh has been largely driven by the human capital theory- by its notion that more education translates into higher earnings through the increase of skills and productivity (Maitra, 2018). Alarmingly, the awareness that this “one size fits all” approach may not be working as intended for everyone-either because of varied educational goals and the theory not being a right “fit,” or because there are differences in skills, or even discrimination for which the theory is dysfunctional- has been missing in the process. This paper introduces this missing criticality.

Overarchingly, the fact that the findings of this paper reveal that the “more education-more earnings” notion does not hold across all the groups of individuals as expected, for example in the case of madrasah-educated individuals, strengthens and further supports the argument, presented in this paper, that there is a need to acknowledge criticalities and nuances of particular contexts when thinking about the relationship between education and earnings. Notably, it is important to clarify that the evidence yielded in this paper is not enough to conclude that the “more education-more earnings” notion not holding in the case of madrasah education is a comment on the quality of the madrasah education steam, and it is not the aim of this paper to do so. Additionally, there remains a need for better data and more rigorous, causational, analysis to be able to strengthen the reliability of the findings of this paper. Until then, this paper advocates for the acknowledgment of overlooked criticalities in researching the relationship between education and earnings in a non-Western global south context such as Bangladesh and positions itself as a foundation for future research to build upon.

Footnotes

Declaration of Competing Interest

The author declares that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Declaration of Conflicting Interests

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

Funding

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

ORCID iD

Rubaiya Murshed

Notes

Author Biography

Rubaiya Murshed is an Education Economist and a PhD candidate at the Faculty of Education, University of Cambridge. By profession, she is a lecturer (currently on study leave) at the Department of Economics, University of Dhaka in Bangladesh. She is also a published writer- her first book, titled 'Nobody's Children', is a collection of short stories about the lives of homeless street children in Bangladesh.

References

Afsar

(1999). Rural-urban dichotomy and convergence: Emerging realities in Bangladesh. Environment and Urbanization, 11(1), 235–246. https://doi.org/10.1177/095624789901100106

Ahmed

McGillivray

(2015). Human capital, discrimination, and the gender wage gap in Bangladesh. World Development, 67, 506–524. https://doi.org/10.1016/j.worlddev.2014.10.017

Alvarado

Conde

Novella

Repetto

(2020). NEETs in Latin America and the Caribbean: Skills, aspirations, and information. Journal of International Development, 32(8), 1273–1307. https://doi.org/10.1002/jid.3503

Arrow

K. J.

(1998). What has economics to say about racial discrimination? Journal of Economic Perspectives, 12(2), 91–100. https://doi.org/10.1257/jep.12.2.91

Arrow

K. J.

(2015). The theory of discrimination. In The theory of discrimination (pp. 1–33). Princeton University Press. https://doi.org/10.1515/9781400867066-003

Asadullah

M. N.

(2006). Returns to education in Bangladesh. Education Economics, 14(4), 453–468. https://doi.org/10.1080/09645290600854144

Asadullah

M. N.

Chaudhury

(2016). To madrasahs or not to madrasahs: The question and correlates of enrolment in Islamic schools in Bangladesh. International Journal of Educational Development, 49, 55–69. https://doi.org/10.1016/j.ijedudev.2016.01.005

Asaduzzaman

Kabir

R. N. A.

Radović-Marković

(2015). Gender Inequality in Bangladesh. JWEE, 3–4, Article 3–4, 54–64.

BANBEIS (2023). Bangladesh Education Statistics 2022. Bangladesh Bureau of Educational Information and Statistics, Ministry of Education, Government of the People’s Republic of Bangladesh.

10.

Bangladesh Bureau of Statistics (2023a). Bangladesh LFS: Labour Force Participation Rate: Rural [dataset]. https://www.ceicdata.com/en/bangladesh/labour-force-survey-labour-force-participation-rate/lfs-labour-force-participation-rate-rural

11.

Bangladesh Bureau of Statistics (2023b). Bangladesh LFS: Labour Force Participation Rate: Urban [dataset]. https://www.ceicdata.com/en/bangladesh/labour-force-survey-labour-force-participation-rate/lfs-labour-force-participation-rate-urban. https://www.ceicdata.com/en/bangladesh/labour-force-survey-labour-force-participation-rate/lfs-labour-force-participation-rate-urban

12.

Bangladesh Bureau of Statistics, UCEP, & Diakonia (2017). Education Scenario in Bangladesh: Gender perspective .

13.

Becker

G. S.

(1957). The Economics of Discrimination. University of Chicago Press.

14.

Becker

G. S.

(1964). Human Capital: A Theoretical and Empirical Analysis, with Special Reference to Education. University of Chicago Press.

15.

Brown

Lauder

Cheung

S. Y.

(2020). The Death of Human Capital?: Its Failed Promise and How to Renew It in an Age of Disruption. Oxford University Press.

16.

Chattopadhyay

(2012). Education and economics: Disciplinary evolution and policy discourse. Oxford University Press.

17.

Falcón

(1995). Social networks and employment for Latinos, Blacks, and Whites. New England Journal of Public Policy, 11(1), Article 4. https://scholarworks.umb.edu/nejpp/vol11/iss1/4

18.

Ferdaush

Rahman

K. M. M.

(2011). Gender inequality in Bangladesh. http://archive.nyu.edu/handle/2451/33900

19.

Gunderson

Oreopolous

(2020). Chapter 3—returns to education in developed countries. In Bradley

Green

(Eds.), The economics of education (second edition) (pp. 39–51). Academic Press. https://doi.org/10.1016/B978-0-12-815391-8.00003-3

20.

Hossain

(1990). Returns from education in rural Bangladesh. In Khan

A. R.

Sobhan

(Eds.), Trade, planning and rural development (pp. 199–219). Palgrave Macmillan UK. https://doi.org/10.1007/978-1-349-11415-3_14

21.

Hossain

M. A.

Tisdell

C. A.

(2005). Closing the gender gap in Bangladesh: Inequality in education, employment and earnings. International Journal of Social Economics, 32(5), 439–453. https://doi.org/10.1108/03068290510591281

22.

Howard

K. A. S.

Carlstrom

A. H.

Katz

A. D.

Chew

A. Y.

Ray

G. C.

Laine

Caulum

(2011). Career aspirations of youth: Untangling race/ethnicity, SES, and gender. Journal of Vocational Behavior, 79(1), 98–109. https://doi.org/10.1016/j.jvb.2010.12.002

23.

Huffman

M. L.

Torres

(2001). Job search methods: Consequences for gender-based earnings inequality. Journal of Vocational Behavior, 58(1), 127–141. https://doi.org/10.1006/jvbe.2000.1770

24.

Kolstad

Wiig

Moazzem

K. G.

(2014). Returns to education among entrepreneurs in Bangladesh. Journal of Asian Economics, 34, 54–65. https://doi.org/10.1016/j.asieco.2014.07.001

25.

Lissenburgh

(2000). Gender Discrimination in the Labour Market. In PSI Research Discussion Series (3; PSI Research Discussion Series). Policy Studies Institute, UK. https://ideas.repec.org/p/psi/resdis/3.html

26.

Maitra

(2018). Investment in physical, human capital, economic growth and life expectancy in Bangladesh: An empirical investigation. South Asia Economic Journal, 19(2), 251–269. https://doi.org/10.1177/1391561418799110

27.

Mamun

S. A. K.

Taylor

B. R.

Nghiem

Rahman

M. M.

Khanam

(2021). The private returns to education in rural Bangladesh. International Journal of Educational Development, 84, 102424. https://doi.org/10.1016/j.ijedudev.2021.102424

28.

Marginson

(2019). Limitations of human capital theory. Studies in Higher Education, 44(2), 287–301. https://doi.org/10.1080/03075079.2017.1359823

29.

Metz

A. J.

Fouad

Ihle-Helledy

(2009). Career aspirations and expectations of college students: Demographic and labor market comparisons. Journal of Career Assessment, 17(2), 155–171. https://doi.org/10.1177/1069072708328862

30.

Montgomery

J. D.

(1991). Social networks and labor-market outcomes: Toward an economic analysis. The American Economic Review, 81(5), 1408–1418.

31.

Mortuza

S. A.

(1992). Rural-urban migration in Bangladesh. Herausgeberexemplar. https://doi.org/10.23689/fidgeo-3637

32.

Mourshed

Farrell

Barton

(2012). EDUCATION TO EMPLOYMENT: DESIGNING A SYSTEM THAT WORKS. Mckinsey center for government.

33.

Naafs

Skelton

(2020). Realities and Aspirations for Asian Youth: Education, Training, Employment. Routledge.

34.

Osmani

S. R.

Sen

(2011). Inequality in rural Bangladesh in the 2000s: Trends and causes. The Bangladesh Development Studies, XXXIV(4), 1–36.

35.

Psacharopoulos

(1972). Rates of return to investment in education around the world. Comparative Education Review, 16(1), 54–67. https://doi.org/10.1086/445569

36.

Psacharopoulos

(1981). Returns to education: An updated international comparison. Comparative Education, 17(3), 321–341. https://doi.org/10.1080/0305006810170308

37.

Psacharopoulos

(1985). Returns to education: A further international update and implications. The Journal of Human Resources, 20(4), 583–604. https://doi.org/10.2307/145686

38.

Psacharopoulos

(1994). Returns to investment in education: A global update. World Development, 22(9), 1325–1343. https://doi.org/10.1016/0305-750X(94)90007-8

39.

Psacharopoulos

Patrinos

H. A.

(2018). Returns to investment in education: A decennial review of the global literature. Education Economics, 26(5), 445–458. https://doi.org/10.1080/09645292.2018.1484426

40.

Rahman

Al-Hasan

(2018). Returns to schooling in Bangladesh revisited: An instrumental Variable quantile regression approach. Bangladesh Development Studies, XLI(2), 16. https://doi.org/10.57138/TRAO9374

41.

Riboud

Yevgeniya

Hong

(2007). The knowledge economy and education and training in South Asia. World Bank.

42.

Roy

Huq

Rob

A. B. A.

(2020). Faith and education in Bangladesh: A review of the contemporary landscape and challenges. International Journal of Educational Development, 79, 102290. https://doi.org/10.1016/j.ijedudev.2020.102290

43.

Schultz

T. W.

(1961). Investment in human capital. The American Economic Review, 51(1), 1–17.

44.

Shafiq

M. N.

(2007). Household rates of return to education in rural Bangladesh: Accounting for direct costs, child labour, and option value. Education Economics, 15(3), 343–358. https://doi.org/10.1080/09645290701306358

45.

Taposh

Lin

(2009). Returns to Education: Estimates from Two Countries (SSRN Scholarly Paper 1461165). https://doi.org/10.2139/ssrn.1461165

46.

Wei-Cheng

Kopischke

(2001). Job search methods, job search outcomes, and job satisfaction of college graduates: A comparison of race and sex. Journal of Employment Counseling, 38(3), 141–149. https://doi.org/10.1002/j.2161-1920.2001.tb00496.x

47.

Wheelahan

Moodie

Doughney

(2022). Challenging the skills fetish. British Journal of Sociology of Education, 43(3), 475–494. https://doi.org/10.1080/01425692.2022.2045186

48.

Women and men in Bangladesh: Facts and figures 2022 (2022). Government of the People’s Republic of Bangladesh.

49.

World Bank (2023a). Labor force participation rate-Bangladesh [dataset]. https://data.worldbank.org/indicator/SL.TLF.CACT.MA.NE.ZS?locations=BD.

50.

World Bank (2023b). Population, female (% of total population)—Bangladesh [dataset]. https://data.worldbank.org/indicator/SP.POP.TOTL.FE.ZS?locations=BD.

51.

World Faiths Development Dialogue and Georgetown University’s Berkley Center for Religion, Peace, and World Affairs (2015). Faith and Development in Focus: Bangladesh. World Faiths Development Dialogue and Georgetown University’s Berkley Center for Religion, Peace, and World Affairs. https://repository.library.georgetown.edu/bitstream/handle/10822/1052080/Faith%20and%20Development%20in%20Focus_Bangladesh.pdf?sequence = 1&isAllowed = y