A Brief Introduction

On the whole, South Asia has seen considerable economic growth in recent years, in spite of ongoing labor-market issues (Yadav and Iqbal 22), many resulting from conflict in the region. In line with Okun’s law, a rising GDP has resulted in increased employment. However, general employment levels do not address who enters the workforce and who does not.

Given the prevalence of long-standing issues like low education levels, gender-based violence, and discriminatory societal norms, an investigation of the relationship between gender and employment is necessary. I aim do so with the following research question:

• Does gender impact the proportion employed in South Asia?
  • I chose gender as the independent variable because gendered issues serve as barriers to success for South Asian women in a variety of contexts (“South Asia Women”). As such, the unit of analysis is women in South Asia.
  • I hope to discover whether gender will once again serve as a predictor to success, specifically in regards to employment.

To help answer this question, I will employ R to generate line graphs, maps, and linear models based on the World Bank’s modeled estimates of data from the International Labour Organization (ILOSTAT). The ILO collects most of its data from South Asian countries through both data from national statistical offices and questionnaires sent to statistical offices and labor ministries. These data are available to R users through the wbstats package.

Does a Gender Gap in Employment Exist in South Asia?

• The line graph below allows for visualization of the relationship between gender and employment in each South Asian country.

• The countries are ordered by smallest to largest gaps in employment over ten years..

  • Nepal boasts the smallest gap, with the proportion of women and men employed being between 72% and 75% as of 2020.
  • On the other hand, Pakistan holds the largest gender gap, with only 19.6% of women employed and 74.5% of men employed.

• While it is less severe in Nepal and Bhutan, the data indicate that a gender gap in employment exists in a majority of South Asia.

  • Furthermore, plotting the data over time allows us to see that there has been little change in the percent of both women and men employed.
  • Such data suggest that one’s gender may influence their chances of gaining employment in the region, especially since women have struggled with this for the past decade.

NOTE: The employment to population ratio represents the proportion of a country’s population that is employed. By taking into account population density, the data avoid conflating differences in population of women and men with their resulting employment percentages.

A Broader View: Comparing South Asian Countries to the Globe

• To place the data in the bigger picture, it is necessary to connect the findings from South Asia to the rest of the world.
  • In doing so, we see how most of South Asia stands in contrast to its Central and East Asian counterparts: while their neighbors to the north and east generally see over 43% of women employed, South Asia sees fewer than 33% of women employed.
  • Alongside North Africa and West Asia, South Asians see some of the lowest rates of female employment in the world.

Data Source: ILOSTAT through the World Bank

• When mapping employment of men across the world, we see how employment trends for women and men are extremely different.
  • While South Asia (alongside West Asia and North Africa) held some of the lowest employment rates for women, it compares to others globally in its employment of men.
  • However, many countries also exhibit a gender gap in employment, as no countries have male employment below 33% while female employment starts at 5%.
    • These data emphasize how discrepancies based on sex are not localized to one portion of the globe. Still, South Asia exhibits one of the more severe cases of gender-based discrimination.

Data Source: ILOSTAT through the World Bank

What Can Statistics Tell Us?

• While the previous graphs are necessary to help visualize the possible relationship between gender and employment, analyzing a mixture of descriptive statistics and linear regression models is essential to answering the question of whether gender predicts employment in South Asia.

	Overall
	(N=2387)
Proportion Employed (Women)
Mean (SD)	46.8 (16.1)
Median [Min, Max]	47.2 [4.49, 83.4]
Missing	330 (13.8%)
Proportion Employed (Men)
Mean (SD)	66.1 (11.1)
Median [Min, Max]	65.5 [33.3, 96.1]
Missing	330 (13.8%)
Population, total
Mean (SD)	33900000 (134000000)
Median [Min, Max]	6320000 [10000, 1410000000]
Missing	9 (0.4%)

• The table above displays descriptive statistics of the proportion of women and men employed across the world.
  • As shown, the median proportion of men employed is 65.5%, which is 18.3% more than the median number of women employed.

	Overall
	(N=88)
Proportion Employed (South Asian Women)
Mean (SD)	36.0 (19.8)
Median [Min, Max]	31.2 [12.9, 80.1]
Proportion Employed (South Asian Men)
Mean (SD)	71.9 (6.60)
Median [Min, Max]	71.6 [58.7, 85.2]
Population, total
Mean (SD)	219000000 (421000000)
Median [Min, Max]	29200000 [366000, 1380000000]

• The table above displays descriptive statistics of the proportion of women and men employed in South Asia specifically.
  • The median proportion of men employed is 71.9%, which is not only 6.4% higher than the proportion of men employed globally but also a whopping 40.7% higher than the median proportion of women employed in South Asia.
  • These data points further highlight how female employment is much lower in South Asian countries than the rest of the world more broadly.

	(1)	(2)
(Intercept)	37.728 ***	46.611 ***
	(1.640)	(0.311)
emp_popMale	35.955 ***	19.323 ***
	(2.176)	(0.432)
pop	-0.000 **	0.000
	(0.000)	(0.000)
N	176	4096
R2	0.620	0.328
logLik	-718.076	-16571.162
AIC	1444.153	33150.325
*** p < 0.001; ** p < 0.01; * p < 0.05.

• In the table above, model (1) concerns South Asia only while model (2) concerns all countries included in the dataset.

• We see that model (1) has an R² value of .620 while model (2) has an R² value of 0.328. This stipulates that the given data explain about 62% of variation within South Asian employment and about 32.8% of variation worldwide.

  • These values tell us that, on the whole, gender may not serve as a reliable predictor of employment across the globe. However, since the R² in model (1) is almost twice as much as that of model (2), gender is more likely to be a predictor in South Asia.

• While the model does provide some evidence for the large gender gap in South Asia, it also lets us know that gender is not a fool-proof predictor of employment in the region or globally.

Conclusions

• As discussed thus far, gender alone may not be a reliable predictor of employment in South Asia

• However, we see comparatively that it is a more reliable predictor for South Asia than the entire globe.

• The data visualizations presented thus far–and the data they build on–allow us to see that gender gaps in the workforce need to be addressed so more women are employed.

  • The primary recommendation for increasing female employment is addressing issues like gender-based violence and girls’ education. Empowering women economically starts with broadening women’s agency through socialization and institutional reform that provides them with access to more opportunities. Specific recommendations include:
    • focusing on development: increasing women’s access to resources like schooling, electricity, and maternal healthcare place gender at the forefront of policymaking (“How to Boost”)
    • instituting quotas, particularly at the executive levels: as women see other women in positions of power, they will likely feel more empowered in their ability to perform jobs.

• What are the implications?

  • More working women will likely lead to a better economy. A study recently revealed that “the overall gain in GDP to South Asia from closing gender gaps in employment and entrepreneurship would be close to 25 percent” (“How to Boost”).

Involving women in the workforce and society as a whole is more important than ever. We must learn from the existing data to understand where gender gaps exist and how we can move forward to close them.