1.1.1. Graduate Unemployment and Mismatches

Unemployment among graduates has been widely investigated, especially in South Asian countries and the growing mismatch between higher education outputs and the labour market requirements. Existing literature points out that while tertiary education expands rapidly in developing economies, the structure of the labour market fails to recruit graduates [18]. Skills mismatch, particularly within academic knowledge and job-specific competencies, has been identified as a critical role in graduate unemployment [19]. Employing companies emphasise the importance of practical experience, problem-solving abilities and communication skills, but graduates’ lack of exposure to these components paves the way for an increase in unemployment rates [20].

Considering South Asian countries, this mismatch is further intensified by the limited integration between companies and universities. Weak collaborations between the two also reduce internships, apprenticeships and work-based learning opportunities, which will in turn help the graduates with the transition into employment. As a result, graduates often enter the labour market inadequately prepared for workplace expectations [21][22].

1.1.2. Graduate Employability Challenges in Bangladesh

Previous studies in Bangladesh consistently highlight structural deficiencies in the higher education and labor market. Researchers have found that university curricula often fall behind industry needs, offering limited practical training and not highlighting the importance of developing soft skills [23]. In the private sector and technology-driven industries, graduates face challenges in aligning their qualifications with employer expectations [24]. Despite the growing number of graduates, the creation of skilled employment opportunities has remained constrained, contributing to the rise in the unemployment rate [25]. Several studies also report that students show a strong preference for public-sector employment over the private sector, citing job security and social status [7][26]. Also, weak connections between the universities and employers reduce graduates’ exposure to the labour market [8]. Socioeconomic aspects like limited access to career guidance and unequal support within institutions further disadvantage graduates from low-income backgrounds [9].

1.1.3. Graduate Employability Challenges in Sri Lanka

Sri Lanka presents a contrasting yet equally complex graduate unemployment context. Despite the provision of free higher education through state universities, graduate unemployment remains a significant concern [27]. Previous literature suggests that unemployment rates tend to increase with higher levels of educational attainment, suggesting that education alone cannot guarantee employment [28]. Slow private sector expansion and limited formation of graduate-level jobs have been identified as a major contributor to this issue [29]. Additionally, field-of-study imbalances play a vital role in shaping graduate employment outcomes. Research also indicates that a concentration of graduates in non-technical disciplines, particularly arts and social sciences, experience weaker absorption in the employment market [30]. Although vocational and professional direction in careers exists within the island, political and economic instability over the last few years has increased the unemployment rate drastically [17].

1.1.4. Research Gap and Contribution

Although previous research has examined unemployment and employability challenges within Bangladesh and Sri Lanka, there is limited comparative evidence on how labour market outcomes have evolved across the two countries over time. Existing studies largely adopt country-specific perspectives and often focus on education and employability issues without considering broader labour market dynamics. Consequently, there remains a need for comparative analyses that investigate unemployment patterns and labour market characteristics within a wider economic context. Addressing this gap, this study undertakes a comparative statistical analysis of Bangladesh and Sri Lanka using annual labour market data for the period 2009-2024. Through an examination of overall unemployment, gender-specific unemployment, youth unemployment, and related labour market indicators, the study contributes to the comparative labour market literature and provides evidence-based insights for policymakers seeking to enhance employment outcomes in South Asian economies.

This study adopts a descriptive and comparative research design to examine labour market outcomes in Bangladesh and Sri Lanka over the period 2009-2024. The focus is on analysing unemployment trends, including overall unemployment, gender unemployment and youth unemployment, alongside selected labour market indicators to understand structural differences between the two economies.

Secondary data were obtained from internationally recognised sources including national labour force surveys, the World Bank and International Labour Organisation databases. The dataset includes annual observations for Bangladesh and Sri Lanka from 2009 to 2024 and covers labour force size, overall unemployment rates, unemployment by gender and youth unemployment (ages 15 - 24), labour force participation and self-employment indicators.

The analysis combines descriptive statistics with basic inferential statistical techniques. Trend analysis and graphical comparisons were used to identify unemployment patterns. Paired t-tests were conducted to compare gender unemployment gaps and youth versus overall unemployment differences. A two-way ANOVA without interaction terms was used to examine country and time effects. Pearson correlation analysis was also conducted to explore relationships between labour market indicators. Effect sizes were also calculated to analyse the results obtained [31][32].

Given the macroeconomic nature of the dataset, the statistical analysis is interpreted as identifying patterns rather than causal relationships.

To formally examine whether female unemployment consistently exceeds male unemployment in both countries, paired t-tests were conducted comparing annual male and female unemployment rates over the 16-year study period (2009-2024).

Figure 1. Female unemployment trends in Bangladesh and Sri Lanka, 2009-2024.

Figure 1 illustrates the trends in female unemployment in Bangladesh and Sri Lanka. The figure shows that female unemployment remained consistently higher in Sri Lanka throughout most of the study period, with noticeable increases during periods of economic stress such as the COVID-19 pandemic and the 2022-2023 economic crisis. In contrast, Bangladesh shows comparatively lower female unemployment, although fluctuations are still visible across the period. These patterns suggest persistent structural barriers affecting female labour market outcomes, particularly in Sri Lanka.

Figure 2 presents the corresponding trends in male unemployment. Compared to female unemployment, male unemployment remains substantially lower in both countries and exhibits less volatility over time. The relatively stable pattern of male unemployment suggests stronger labour market attachment among men, while the consistent gap between male and female unemployment visually confirms the gender disparities later tested through statistical analysis.

Figure 2. Male unemployment trends in Bangladesh and Sri Lanka, 2009-2024.

Taken together, these figures provide visual evidence of a persistent gender unemployment gap in both countries, with Sri Lanka showing a larger and more stable gap compared to Bangladesh. These descriptive patterns are formally confirmed by the paired t-test results presented in Table 1.

Table 1. Paired t-test results for gender unemployment gaps (2009-2024).

Note: pp = percentage points; SD in parentheses; * indicates significance at α = 5%.

In Sri Lanka, female unemployment (mean = 7.30%, SD = 0.95%) was significantly higher than male unemployment (mean = 3.36%, SD = 0.57%). The mean difference of 3.94 percentage points was statistically significant (paired t-test: t = 13.42, df = 15, p < 0.0001, 95% CI: 3.38 to 4.70). This confirms that the gender gap observed in Figure 1 and Figure 2 represents a persistent structural feature of the Sri Lankan labour market rather than random annual variation.

In Bangladesh, female unemployment (mean = 6.15%, SD = 1.01%) also significantly exceeded male unemployment (mean = 3.70%, SD = 0.74%) across the study period. The mean difference of 2.45 percentage points was statistically significant (paired t-test: t = 9.18, df = 15, p = 0.0014, 95% CI: 1.82 to 2.88). The overall pattern confirms significantly higher unemployment rates for female graduates in both countries across the 16-year period.

The magnitude of the gender gap differs between countries, with Sri Lanka showing a larger average gap (3.94 percentage points) compared to Bangladesh (2.45 percentage points). This indicates more pronounced gender disparities in Sri Lanka’s labour market. However, an independent samples t-test comparing these gaps found no statistically significant difference between countries (t = 1.21, df = 8, p = 0.26), suggesting that both South Asian economies face similarly severe gender disparities in graduate unemployment outcomes.

Youth unemployment rates in both countries were substantially higher than overall unemployment rates throughout the 2009-2024 study period. Paired t-tests confirm that this difference is statistically significant in both countries as shown in Table 2.

In Sri Lanka, youth unemployment averaged 20.30% compared to 4.73% overall unemployment, representing a difference of 15.57 percentage points. In Bangladesh, youth unemployment averaged 10.33% compared to 4.30% overall unemployment, a difference of 6.03 percentage points. These results confirm that young labour market entrants face disproportionately higher employment challenges, particularly in Sri Lanka.

Table 2. Paired t-test results: Youth vs overall unemployment (2009-2024).

Note: pp = percentage points; SD in parentheses; * indicates significance at α = 5%.

An independent samples t-test comparing the youth-overall unemployment gaps between countries revealed that Sri Lanka’s gap (15.57 pp) is significantly larger than Bangladesh’s gap (6.03 pp) (t = 7.84, df = 15, p < 0.0001). This statistically confirms that the school-to-work transition challenge is substantially more severe in Sri Lanka than in Bangladesh.

To examine whether unemployment differs significantly between countries and across time, a two-way analysis of variance (ANOVA) was conducted using unemployment rate as the dependent variable and country and year as fixed factors as shown in Table 3. The interaction term was excluded because only one observation exists per country-year combination.

Table 3. Two-way ANOVA: Effects of country and year on unemployment rates (2009-2024).

Note: *indicates statistical significance at α = 5%.

The ANOVA results show no statistically significant difference in average unemployment between Bangladesh and Sri Lanka. However, unemployment varies significantly across years, indicating that macroeconomic conditions play a stronger role than country differences in explaining unemployment variation. Figure 3illustrates unemployment trends in Bangladesh and Sri Lanka over the period 2009-2024. The figure shows that Sri Lanka experienced greater fluctuation in unemployment over time, with noticeable increases during 2020 and 2023, whereas Bangladesh displayed comparatively more stable unemployment patterns, despite a temporary rise during the COVID-19 period. These visual trends are consistent with the ANOVA results, which indicate significant variation across years but no statistically significant overall difference between the two countries.

Figure 3. Unemployment trends in Bangladesh and Sri Lanka, 2009-2024.

Figure 3 reinforces the descriptive finding that Sri Lanka experienced greater volatility in unemployment over the study period, while Bangladesh maintained a relatively more stable trend. Although average unemployment rates do not differ significantly between the two countries, the temporal patterns suggest that macroeconomic shocks affected each labour market differently.

Pearson correlation coefficients were calculated using the 16-year dataset (2009-2024) as depicted in Table 4. The larger sample improves statistical reliability compared to shorter time windows.

Strong positive correlations were found between overall unemployment and male unemployment in both countries. Sri Lanka also shows a very strong relationship between overall and female unemployment, suggesting that labour market shocks affect both genders simultaneously.

Youth unemployment shows strong positive correlations with overall unemployment in both countries, indicating that young workers are particularly sensitive to macroeconomic fluctuations.

Bangladesh shows a significant negative relationship between female labour force participation and female unemployment, suggesting improved participation may reduce unemployment risks. This relationship is weaker in Sri Lanka.

Self-employment shows moderate negative relationships with unemployment, suggesting a possible buffering effect, although these relationships are not statistically significant.

Table 4. Pearson correlation coefficients for labour market indicators (2009-2024).

Note: LFPR = Labour Force Participation Rate; * indicates at α = 5% significance level; n = 15 years.

The correlation analysis reveals several important relationships between labour market indicators in both countries. A strong positive relationship exists between overall unemployment and male unemployment in both Bangladesh (r = 0.83) and Sri Lanka (r = 0.96), indicating that general labour market conditions strongly influence male employment outcomes. However, the relationship between overall and female unemployment differs between countries. Sri Lanka shows a very strong correlation (r = 0.95), suggesting that female unemployment closely follows overall labour market trends. In contrast, Bangladesh shows only a weak relationship (r = 0.28), indicating that female unemployment may be influenced by additional structural or socio-cultural factors.

Youth unemployment also shows strong positive correlations with overall unemployment in both countries (Bangladesh: r = 0.81; Sri Lanka: r = 0.80), suggesting that young workers are particularly vulnerable to economic fluctuations. This finding supports existing literature highlighting the difficulty of school-to-work transitions in developing economies.

The relationship between female labour force participation and unemployment differs between the two countries. Bangladesh shows a significant negative correlation (r = −0.68), suggesting that increased female participation may improve employment outcomes. Sri Lanka shows a weaker negative relationship (r = −0.32), indicating that participation alone may not fully explain female unemployment patterns.

Finally, moderate negative relationships were observed between self-employment and overall unemployment (Bangladesh: r = −0.46; Sri Lanka: r = −0.38), suggesting that self-employment may act as a partial labour market adjustment mechanism. Although these relationships are not statistically significant, they suggest that informal or self-employment sectors may absorb workers during periods of labour market pressure.

Overall, these correlations indicate that while macroeconomic conditions influence unemployment patterns in both countries, structural labour market differences shape how these relationships operate. The longer 16-year dataset provides stronger descriptive evidence of these patterns compared to shorter time series studies.