Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

23 

 Journal of Applied Economics and 
Business Studies (JAEBS) 

Journal homepage: https://pepri.edu.pk/jaebs 

ISSN (Print): 2523-2614 

   ISSN (Online) 2663-693X 

 

 

Female Unemployment Duration in Pakistan – A Survival 

Analysis  
 

GulRukh Zahid1*1 & Ambreen Fatima 
1 M.Phil Scholar, Applied Economics Research Centre (AERC), University of Karachi 

2 Associate Professor, Applied Economics Research Centre (AERC), University of Karachi 

 

 

ABSTRACT 

Female unemployment in Pakistan is very high in contrast to its 

neighboring countries. The finding of previously available literature 

concludes that staying out of the labour market for longer durations 

causes a lower possibility of getting a job again and for females; the 

probability is even lower compared to males. This study examines the 

unemployment situation and its duration, forcing females to drop out of 

Pakistan's labour force. Numerous factors are accountable for the 

higher inactivity rate of females in the labour market. Some examples 

are low education level, higher age, home responsibilities such as taking 

care of elders and children, spouse job status, marital status, and non-

availability of job-specific knowledge. Although in the case of Pakistan, 

literature shows many causes of higher inactivity of female 

unemployment, it does not focus on the time estimations and calculate it 

in days or months to figure out how long a female take to get back on a 

job once she becomes unemployed. More specifically, the aspects which 

determine unemployment are unavailable. Kaplan-Meier Survival 

analysis has been performed to assess the research by employing the 

data on the labour market and socio-economic variables from the 

Labour Force Survey (LFS) source of 2014-15. This study can help 

policymakers to figure out the duration of female unemployment, 

prevalent in the labour market for females in Pakistan. 

 Keywords  

Female 

unemployment, 

Survival 

Analysis, 

Kaplan-Meier 

technique   

JEL 
Classification 

J64, J13, J16, 

P34 

 
1 *gulrukhzahid@gmail.com 



GulRukh Zahid & Ambreen Fatima 

24 

 

1. Introduction 

One of the fascinating topics for academicians and policymakers is unemployment since 

it is one of the most distressing issues for developing and developed economies. Henceforth 

It is the favourite topic for labour economists, causing it the most debated matter among them 

as the strong spur on both micro-level as an influence of impacting the wellbeing of 

households and on the macro level, as it creates its impact on the aggregate level. The 

economic term differentiates the self-unemployed and unemployed by identifying that one 

who quit the job to stay at home to take care of young children or who does it for higher 

studies is not considered unemployed as they are not actively searching for a job. Numerous 

problems in society have the root cause of long-term unemployment. Survival for individuals 

becomes very difficult as many candidates contest for fewer positions. Social pressure and 

responsibilities make a person get into a severe depression when one fails to find any job, 

which ultimately pushes them to opt for any way possible to come out of that state. 

 Unemployment is equally stressful for both males and females as they spend their life's 

quality time acquiring skills and education for a prosperous future. Unfortunately, they 

comprehend a vast gap between the opportunities offered and the people willing to get them 

when they enter the job market. Undoubtedly, it is the state's responsibility to take policy 

measures that create job opportunities and ultimately stimulate growth in the long run. 

Nevertheless, if the state fails to make employment in the economy, it provides foundations 

for many social problems. Pakistan is considered the 6th largest economy globally by its size 

and its working-age population, and the number count is growing every minute. The statistics 

taken from the World Bank data on development indicators reported that 65.3% was declared 

as the population under the working-age category for 2015, highlighting that most of its 

population consists of youth. The average unemployment rate was 5.46% from 1985 to 2015 

and peaked in 2002 when unemployment reached 7.8%. 

 



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

25 

 

Figure1: Female unemployment 

Source: Data taken from World Bank 

Unemployment among females in Pakistan is alarming because the female population is 

more in number than males in Pakistan. Following the role model of developed economies, it 

is high time to encourage the country's females to become part of the labour force. However, 

female labour force participation has increased over the years. Still, it is not as desired 

compared to Pakistan's neighbouring countries. Being a developing country, Pakistan should 

increase female participation in the labour force; it is being perceived that unemployment 

affects females differently than males, especially educated females. 

Moreover, the persistent increase of females in the labour force is cumulating, causing the 

gap to grow faster, impacting the current employment opportunities. In terms of new arrivals 

and limited job opportunities, the constantly extending labour market creates a massive gap 

between jobs and the available labor force. The continuous increase in unemployment with 

fresh arrivals in the form of new graduates entering into the labour market creates havoc and 

will be problematic in the long run. 

0

1

2

3

4
1

9
9

1

1
9

9
2

1
9

9
3

1
9

9
4

1
9

9
4

1
9

9
6

1
9

9
7

1
9

9
8

1
9

9
9

2
0

0
0

2
0

0
1

2
0

0
2

2
0

0
3

2
0

0
4

2
0

0
5

2
0

0
6

2
0

0
7

2
0

0
8

2
0

0
9

2
0

1
0

2
0

1
1

2
0

1
2

2
0

1
3

2
0

1
4

2
0

1
5

2
0

1
6

2
0

1
7

2
0

1
8

Female Unemployment



GulRukh Zahid & Ambreen Fatima 

26 

 

 

Figure 2: Female labour force participation in Pakistan 

Source: World Bank 

It is witnessed that in Pakistan, women stay at home, for which many factors are involved. 

The primary reason for others is that household responsibilities (child care is included in it) 

did not allow her to go out to work. Other than this, lower-wage compared to males are also 

the reason for lower labour force participation. Spouse job status is also among those causes 

which do not permit her as it is suggested not to work if husbands earnings are reasonable; 

these all only a few among all those issues which are responsible for limited participation of 

females in the labour market(Tasnim Khan et al. 2009 Ahmed et al. 2013, Shabir et al. 2015). 

However, according to the World Bank statistics, the women residing in rural areas of 

Pakistan have more share in the labour force than urban areas due to the higher rate of poverty. 

Another cause can be less educated, or people with no education are more adaptable and ready 

to work than highly qualified or have any education as they prefer to work according to their 

education. Different studies concluded that those with higher education chose to stay 

unemployed if they failed to acquire a job that matched their qualifications (Kahn et al., 2009). 

The pioneering work of Mincer (1962) and Cain (1966) on the economic analysis of the labour 

force participation of females received significant interest. Extensive female labour force 

participation improvement is observed in developing countries. A developing economy like 

Pakistan needs more of its females to join the labour force for the country's rapid growth. If 

the issue is not appropriately addressed, it will discourage females from pursuing higher 

education in the future. It is essential to understand the importance of examining the duration 

of unemployment at the individual level especially long-term unemployment, because it is the 



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

27 

most significant factor which decides the life choices of individuals. Females living in a 

developing country like Pakistan let themselves deprives of many necessities due to the non-

availability of disposable income at their end. The meaningful life activities and better 

livelihood require better income resources Tansel, A., & Taşçı, H. M. (2010). 

1.1. Objective of the Study 

In order to estimate the unemployment duration among females Kaplan-Meier technique 

has been used. It is a nonparametric (actuarial) method for approximating time-related events 

(the survivorship function). Typically, A plot of the Kaplan-Meier estimate of the survival 

function is a sequence of horizontal steps of diminishing magnitude which, when a huge 

enough sample is taken, approaches the true survival function for that population. A 

significant advantage of the Kaplan-Meier curve is that the method can consider the 

"censored" data losses from the data before the final results are observed. 

Here, we are concerned to evaluate the duration of female unemployment, which is an 

enormous share of the dejected labour force. We can witness from the example of developed 

economies that they grew faster when they created equal opportunities for their female to play 

their part in society's wellbeing. The previously available literature has shed light on the fact 

that if good health amenities and equal chances of education are provided, it will increase its 

long-term growth. Specifically, the study will examine the role of age, education, gender, 

family size, children and other characteristics as determinants of unemployment duration. 

The paper is prearranged in the following manner: Section II shows the existing literature 

and highlights the critical factors related to the study—a review of the literature. The 

methodology part elaborates on the technique this study used for analysis purposes, the part 

after that debates the data sources; lastly, the study debated the study's findings by shedding 

light on the results and conclusion 

2. Literature Review 

Numerous social, economic, and psychological factors are involved in female 

unemployment. It is high time for females of our society to get encouraged and participate in 

the labour force to improve their economic stability and the country. It is being observed that 

females are the neglected part of our culture being stayed at home; however, if given a chance 

to them, an enormous change in society can be bought, as they can perform on both home and 

workplace if given few facilities. Females quit their job after having children. It is one of the 

primary reasons that they stay at home because of the non-availability of help at home who 

can take care of their children in their absence. It was concluded in many studies that children 

under the age of 5 increase females' unemployment probability. This is the reason that they 

stay out of the labour market. Moreover, it was recommended that to resolve this issue, more 



GulRukh Zahid & Ambreen Fatima 

28 

 

childcare facilities will help counter this problem (Michele A. et al. (2011), N Lázaroet et al, 

(2000), Shabbir et al, 2015). In another study performed by Samer Kharif (2015) concludes 

that higher wage reservations and discriminatory labour market conditions are one of the 

causes of female's longer duration of unemployment. 

In Spain, the analysis was performed on the Household Expenditure Survey Data (1990-

91). The study utilizes the socio-economic variables and personal characteristics such as age, 

education, and family background and concludes that there is a need to create a resolution for 

a better family and professional life. Enormous individual-level data of Romania and Hungary 

was examined by Daniela et al.(2012) and highlighted the impact of socio-economic variables 

such as marital status, education, health status, age, region and unemployment allowance on 

employment status and unemployment duration for women and indicated that for women 

unemployment duration and age is a significant variable. Khan et al. (2009) determined that 

married women become an active part of the informal sector's labour force due to poverty. 

Doing so increases productivity and working mothers' income, reducing poverty at the 

household level in the long run. Anderson (1993) took a sample of about sixty working 

women and forty jobless women to examine Glasgow women's unemployment experience 

and summarized the importance of paid employment for women's social identity. The study 

also concluded that unemployment is a crucial factor in the loss of women's economic and 

social identity and at the domestic level, there is no reward for doing household chores by 

losing paid work. 

It was examined by Ahmed et al., (2013) on the data of rural and urban areas of 

Bahawalpur, Pakistan that the socio-economic variable like education, age, number of family 

members employed, education of father, mother, and husband, job status of the mother and 

technical education are the reason for lower employment rate in females. The study also 

summarizes that large family size, number of children, and joint family positively impact 

female unemployment. The study performed on India's labour force by Stephan (2013) 

evaluated that India's economy is not utilizing its large working-age population. If the same 

scenario persists, the propensity to incorporate female labour force participation in coming 

years will reduce further, which will cause India to fall in the growth rate if they do not include 

their educated females in the labour force. Pieters (2015) identified by incorporating the five 

extensive cross-sectional microdata for the period of 1989-2009 that higher education level, 

high growth rates and decline in fertility rate. The labour force participation of females is 

unchanged which is caused by market forces. The supply side's situation causes the main 

reason for such an immobile participation rate: household income, low-level selection of 

highly educated females, and the cultural feeling of dishonour for working females. Nonyana, 



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

29 

Jeanette Zandile (2015) with the help of the survival technique, examine the semi-parametric 

and nonparametric estimates to understand the duration dependence and probabilities of 

exiting unemployment and the socio-demographic factors associated with it. 

Brigitte (2018) worked on Botosani County's labour market data. The information was 

taken from the Employment Agency of Botosani County between the time duration of 2012-

to 2015 on the data set of 200 unemployment spells by applying Kaplan- Meier estimator to 

predict the probability of remaining unemployed with characteristics like age, cohort and 

gender. The findings revealed that variable cohort influences the probability of staying 

unemployed if gender and age were considered as controlled. 

By working on the data of the United States for the period of 2008 to 2015 Sansale (2019) 

evaluated the importance of an individual's personality and its role in determining the 

unemployment duration among young adults. The methodology that was used for that purpose 

was Competing Risk Model. It was estimated in the paper that individuals with more concern 

to their job status got a job as soon as they lost the previous one compared to those who have 

neuroticism.  

The data for the time 2000-01 was utilized by Taşçi et al. (2010) to evaluate the duration 

of unemployment for both men and women on Turkey's data. The study examines household 

and personal characteristics and their effect on the labour market state and concluded 

similarities in both developed and developing countries. It was stipulated that with higher 

education, such as having a university degree for females has no impact on unemployment, 

while for males, it positively impacts it. Msigwa. (2013) employ Multinomial logistic 

regression model (MLM) on the data of Tanzania to evaluate youth unemployment. The study 

results suggest that skills, education, geographical location, gender, and marital status 

significantly explain the difference in youth employment status in Tanzania. A Semi-

parametric Cox regression approach was employed by Kavkler et al. (2009) to examine the 

duration of unemployment for five Central and Eastern European countries. The study utilizes 

the Cox proportional hazard model and Cox regression model with the inclusion of time-

dependent.  

3. Methodology 

This study focuses on unemployment duration using survival function; a nonparametric 

and parametric approaches. The nonparametric (actuarial) technique can handle the problem 

of censoring (both right and left). This is why, for duration analysis, the Kaplan-Meier 

approach, also known as product limit estimator, has been used. A breakthrough in medical 

science was witnessed, when Edward Kaplan and Paul Meier published their work in the 

Journal of American Statistical Association in June 1958. They evaluated that while 



GulRukh Zahid & Ambreen Fatima 

30 

 

performing a research trial, some patients may die while others survive at some stage. This 

motivates them to estimate the patient survival rate. They estimated the proportion of patients 

who survived at any point during the trial. Kaplan-Meier estimation is extensively used in 

recent years in other social science subjects as well. In economics, the technique is specifically 

used to research price and unemployment duration assessment by survival analysis.  

Numerous studies have estimated the duration analysis of unemployment by using a 

nonparametric technique such as Kaplan-Meier. Few of them are Flek al (2015), Vasilica, et 

al. (2011), Danacicaet al (2010), Gabriel et al. (2017), Ciucaet al (2010). The nonparametric 

method of estimating time-related events generally evaluates the survivorship function. It is 

commonly used to investigate the death as an outcome in biostatistics, but lately, this has 

become popular in industrial sciences and other social sciences.  An economist might get 

interested in estimating the duration of people who remained unemployed. The sequence of 

horizontal steps of reducing magnitude in the Kaplan-Meier estimation of survival function 

presents the population's survival function in a huge sample. It approaches its real survival 

function for that specific population. It is assumed that the value of the survival function in 

the distant sample is constant. Due to its significant advantage of taking into account the 

"censored" data, Kaplan-Meier curve considers the lost data before the final results observed. 

The survival function "S(t)" starts with preliminary steps that captivate the probability of 

survival past time "t". It initiates with the steps before approximating the survival function 

"S(t)", which can evaluate the probability of survival time. For this study, "survival" means 

the period when an individual remains unemployed – i.e. before he/she gets re-employed. The 

time "t" is measured in days; 

                           S ̂(t)=∏_(j/j≤t)  (u_j-a_j)/u_j                                              (1) 

Here, we have taken u_j as the spells of unemployment ongoing in numbers for j days, 

while a_j are those spells (in numbers), change into the employment promptly after j days. 

We have considered the concept of hazard rate λ(t). For this study, hazard rate λ(t) 

represents the short probability of leaving unemployment and moving into employment at 

time t, and it is entirely dependable on the condition of remaining unemployed till the time of 

unemployed until immediately before t. 

The universal classification of a continuous hazard rate is; 

              λ ̂(t)=lim┬(∆t→0)⁡〖(Pr⁡(t≤T<t+∆t\T≥t))/∆t〗                              (2) 

In the equation above, T represents the spell duration of unemployment (in our analysis, 

we have taken the days till any random individual stays unemployed). The proportional hazard 

model assumes the continuous hazard rate for the ith spell abides the following form   



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

31 

                            λ_i (t)= λ_o (t).e^(X_i^' (t).π)                                           (3) 

Equation (3) represents the λ(t) as the assumed baseline hazard, and X_i(t) is the vector 

of covariates (descriptive variables).  For example, the age of a person, gender, education 

level, marital status etc., and π presents the projected parameter vector. Finally, the term 

e^(X_i^' (t).π) represents a proportional shift – the observed explanatory variables change the 

complete hazard rate up or down. 

Semi-parametric analysis (cox regression) 

To estimate the relationship between survival rate and covariates selected. For the 

semiparametric analysis Cox regression is employed. Cox regression determines covariates' 

effect on the hazard model -its direction (increasing or decreasing) and importance. The 

common form of the Cox regression is   

         λ_(i )(t) =e^(X_i^' (β) ).λ_(o )(t) = c_i  .λ_(o )(t)                     (4)        i =  1,2 …n 

In the equation given above, x_(i )  ( x_(i 1) x_(i 2)….x_(i k)) is the vector of k covariates. 

The analysis includes age, marital status, spouse job status, education, head of household’s 

education, head of the household’s gender, head of household’s income, region, province and 

training taken if any. β (β_1 β_(2 ) 〖…  β〗_k) is the vector of regression coefficients and λ_(i 

)(t) is the hazard function of individual i and λ_(o )(t) is the baseline hazard. The baseline 

hazard starts with 0. The impact of the covariates on the hazard function in the Cox regression 

model does not depend on time due to the ratio (λ_i  (t))/(λ_o  (t))is equal to the constant ci. 

Therefore, the baseline hazard determines the shape of the hazard function. The ratio of the 

hazard functions of individual si and j, namely (λ_i  (t))/(λ_j  (t))is called the hazard ratio. 

(λ_i  (t))/(λ_j  (t))=           (〖e^(X_i^' (β) )  .  λ〗_0 (t))/(〖e^(X_i^' (β) )  .  λ〗_0  (t))=e^(〖(X〗_i^ 

- 〖x_j)〗_^(  ') (β) )       (5) 

The hazard ratio is the covariate ratio and is independent of time. This property is 

proportional, and its explanation is parallel to the odds ratio for logistic regression. A ratio 

higher than 1 shows increased risk while a ratio lower than 1 specifies decreased risk.  

                                    (λ_i  (t))/(λ_j  (t))= e 〖β_p〗^                       (6) 

The hazard ratio is said to be statistically significant at the given level when its confidence 

interval excludes 1. The null hypothesis that the variable is not correlated to survival can be 

rejected. This is the base for the explanation of the Cox regression results. By using Cox's 

partial likelihood estimator, it is possible to estimate the parameter vector - without 

postulating and estimating the baseline hazard (see Greene (2003) for details). 

The final model estimated is: 

H(t) = β_(1age )+ β_(〖age〗^2 )+ β_3edu+ β_(〖4edu〗^2 ) + β_5MS + β_(6spouse job 



GulRukh Zahid & Ambreen Fatima 

32 

 

status) +β_7traning + β_8region + β_9province + β_(10 head edu) + β_(11 house size) +β_(12 

head house) 

Where h(t) represents hazard rate, while the independent variable includes age, age 

square, education square, marital status, spouse job status, training, region, province, head 

education, house size and head of the house. 

As far as the effect of independent variables on hazard rate is concerned, the variable age 

as a determinant of hazard rate – unemployment survival function is taken in this study as age 

has a greater impact on employment decision and choice. Education too plays a significant 

role in finding a job. In order to evaluate the above statement, a non-linear term of education 

is also included in the model to see if after a certain education the job market condenses. It 

has been observed from the data that highly educated females, if once got unemployed their 

hazard of getting reemployed usually declines given that the offers did not match their 

qualifications, this is the reason, that the curve of education is decreasing after post-

graduation. Marital status is also assumed to have a negative impact on getting reemployed 

once a female got unemployed. The study has taken the spouse's job status for female 

employment as well indicating if a spouse has better job status, it will create a negative impact 

on a hazard of female employment. Head education is taken to see if the head of the house 

education influences female employment reemployment rate or not. The covariate of training 

is hypothesized as having a significant impact in the case of getting a job for females. Here 

training related to the job has a positive impact on job status.  

4. Data and Variables 

The study has taken into account the Labor Force Survey (LFS) data for the year 2014-

15. The analysis was made after extracting the required information from the data. 

Table 1: Duration and unemployment variables 

Variable Expressions Observations 

The span of the 

unemployment spell 

(duration of 

unemployment):  

The number of unemployment 

days. (the difference between 

the date of entry and the date of 

leaving unemployment).  

The minimum value is two days, and 

the maximum value is 3650 days 

(121.67 months and ten years). The 

mean value is 596.59 days, while the 

median duration is 730 days 

(months).  



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

33 

The status of the 

person (or the censored 

index)  

 

For analysis, all the diverse 

causes of the exiting are grouped 

into 2 major outgoing 

destinations:  

1 -economically active or 

employed (first employment or 

re-employment);  

0 - economically inactive 

Out of 1279 females in the labour 

force, there were 13 different reasons 

for leaving unemployment. The status 

variable allows for a distinction 

between the individuals who 

registered for the event (employment) 

and the individuals who did not 

register for the event.  

 

Table 2: Explanatory Variables  

Age We grouped the ages into 5 
categories  

Age_1 → 14-18 years  

Age_2 → 19-28 years  

Age_3 → 29-60 years  

Age_4 → 61-80 years  

Age_5 → 80 plus years 

Age_6  → 85 plus years 

The average age of a female 

registered in the database on LFS 

is 13 years and the maximum is 

78 years with the mean of 27.88. 

Age square  Age square is taken to see the 

impact of education for an 

individual 

Education The level of education is grouped 
into 5 categories:  

Educ_1 → Primary and secondary or 

vocational education  

Educ_2 → High school and post-

secondary education till graduation  

Educ_3 → post-graduation till 

Masters level 

Educ_4 → MPHIL 

Educ_5 → PhD 

The minimum education of the 

group is 1 and the maximum is 18 

years of education with the mean 

value of 12.73. 

Education square  We have taken the square of 

education to assess the impact of 

higher education on employment. 

Marital status (0) Unmarried ;(1) Married  

 

515 (40.27%) women were 

married in the data while 

764(59.73%) were unmarried. 

Spouse job status (1)Employed;(2) Unemployed; 

(3)Not in labour force  

 

The total number of employed 

husbands is 139 with 10.87%, 

unemployed is 98 with 7.66% 

and not in labour force is 1,042 



GulRukh Zahid & Ambreen Fatima 

34 

 

with 81.47%. 

Training (0) No; (1) yes Females with no training are 1, 

011 79.05% and with training are 

268 20.95%. 

Region (0) Rural; (1) urban People in rural areas are 742 with 

58.01% and in urban are 537 and 

41.99% 

Province (1) KPK  

(2) PUNJAB  

(3) SINDH  

(4) BALOCHISTAN  

(5) ISLAMABAD 

The percentage of population 

according to the province in 

KPK is 22.67%, 

 Punjab with 53.79%  

Sindh with 16.73% 

Balochistan with 6.49%  

Islamababad0.33% 

House hold size  House hold size is taken as how 

many members of a family living 

in any houes 

Head house (1) unemployed (2)employed  

Head education (0) NO; (1) YES About 89% of the head house is  

Results  

Specifically, in this study's analysis, we have studied and analyzed the time duration when 

any random individual leaves employment until he gets reemployed. The survival function 

generates a step for representing the time when an individual is exiting the state of 

unemployment in a graphical representation. In our sample data of the Labour Force Survey 

of 2014-15, the information of 1279 females was available, among those 135 were employed, 

289 were not employed, and 855 were out of the labour force (currently were not looking for 

job opportunities). The graph below shows the survival estimate. 

 

 

 

 

 

 

 

 

 



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

35 

Table 3: Kaplan-Meier survival function 

                                                                                
  3650        2      2      0             0.0000         .          .         .

  2920        5      3      0             0.0018    0.0012     0.0004    0.0061

  2555        6      1      0             0.0044    0.0020     0.0017    0.0099

  2190       11      5      0             0.0053    0.0021     0.0022    0.0111

  1825       23     12      0             0.0097    0.0029     0.0052    0.0168

  1460       47     24      0             0.0202    0.0042     0.0132    0.0297

  1265       48      1      0             0.0413    0.0059     0.0308    0.0540

  1095       91     43      0             0.0422    0.0060     0.0316    0.0550

  1030       92      1      0             0.0800    0.0080     0.0652    0.0967

   910       93      1      0             0.0809    0.0081     0.0660    0.0977

   730      814    721      0             0.0818    0.0081     0.0668    0.0986

   575      815      1      0             0.7158    0.0133     0.6888    0.7410

   545      816      1      0             0.7167    0.0133     0.6897    0.7418

   368      817      1      0             0.7176    0.0133     0.6906    0.7427

   366      818      0      1             0.7185    0.0133     0.6915    0.7435

   365     1027    209      0             0.7185    0.0133     0.6915    0.7435

   300     1028      1      0             0.9020    0.0087     0.8835    0.9177

   270     1030      2      0             0.9029    0.0087     0.8845    0.9185

   240     1040     10      0             0.9047    0.0086     0.8864    0.9201

   210     1042      2      0             0.9134    0.0082     0.8959    0.9282

   180     1078     35      1             0.9152    0.0081     0.8978    0.9298

   150     1086      8      0             0.9459    0.0065     0.9315    0.9573

   120     1097     10      1             0.9529    0.0061     0.9394    0.9635

    97     1098      0      1             0.9617    0.0055     0.9493    0.9711

    90     1156     12     46             0.9617    0.0055     0.9493    0.9711

    60     1188     12     20             0.9718    0.0047     0.9609    0.9797

    40     1189      0      1             0.9817    0.0038     0.9726    0.9878

    32     1190      1      0             0.9817    0.0038     0.9726    0.9878

    30     1228     10     28             0.9825    0.0037     0.9736    0.9885

    22     1229      0      1             0.9906    0.0027     0.9835    0.9946

    20     1237      1      7             0.9906    0.0027     0.9835    0.9946

    15     1263      2     24             0.9914    0.0026     0.9845    0.9952

    10     1269      1      5             0.9930    0.0023     0.9865    0.9963

     8     1270      1      0             0.9937    0.0022     0.9875    0.9969

     6     1273      3      0             0.9945    0.0021     0.9885    0.9974

     5     1275      0      2             0.9969    0.0016     0.9917    0.9988

     3     1276      1      0             0.9969    0.0016     0.9917    0.9988

     2     1279      3      0             0.9977    0.0014     0.9927    0.9992

                                                                               

  Time    Total   Fail   Lost           Function     Error     [95% Conf. Int.]

           Beg.          Net            Survivor      Std.



GulRukh Zahid & Ambreen Fatima 

36 

 

 

Figure 3: Kaplan-Meier survival Estimate 

Author's estimation using LFS (2014-15)                     

As per the outcomes of our study, a female's chance to get reemployed significantly 

reduces if she stays jobless for longer durations. The time projected in our analysis is ten 

years, i.e. 3650 days and in terms of months, it is 121.67 months. In graph 1 above, we can 

see that the starting steps are very close and comparatively small, predicting a higher 

probability of re-employment. In 2 days, the possibility of finding a job is 99.77%, and it 

continuously reduces, as the survival function increases. They are depicting a negative relation 

between re-employment and survival function. According to our findings, females' probability 

of finding a job reduces if they stay out of the labour market for more extended time durations, 

causing it as one of the main factors for females to quit job search voluntarily and come out 

of the labour force due to unavailability of the job permanently. 

The findings in figure 4 indicate that females with higher education such as a Master’s 

and higher than that have even fewer chances of entering the job market after becoming 

unemployed. This situation will make females feel overqualified and make them demoralized 

in persuading higher education. It is being witnessed already in Pakistan's case that females 

face numerous issues with respect to gender discrimination in the labour market. Fewer wages 

for the same work than male is already a problem; females have to face while working. Many 

studies show that females with children below five years of age and elders at home are the 

reason to generate impediments for them to proceed. 

 

 



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

37 

 

 

Figure 4: Kaplan-Meier survival estimate by education 
Author's estimation using LFS (2014-15 

Pakistan, being a developing country, belong to those societies, where females are 

associated with homemaking. It becomes unenviable to continue a career after marriage as 

uncountable household obligations do not allow them to do so. Employers are least concerned 

with hiring females with children as the unavailability of childcare pushes them to bequeath 

their profession. It will be beneficial to get back this discouraged labour force into the job 

market if females are provided with daycare facilities and similar to them in economical fees.  

 Figure 5 is presenting the distribution of females according to their age groups. According 

to this graph, females in the age group of 29 to 60 have scarcer likelihoods to get employed 

compared to the age group of 19 to 28. In figure 6, females are represented by marital status. 

The graph shows that married females face more problems than unmarried females in 

employment. It is high time for females to anticipate into the labour force, as almost half of 

Pakistan's population consists of females. At the macro level, this increase in the number of 

labour forces will spike up the opportunities of the higher growth rate in productivity and 

economic growth. Similarly, it will boost the affluence and additional income into the house 

on a micro level, which helps them invest more in their offspring's education and health.   

 

 

 



GulRukh Zahid & Ambreen Fatima 

38 

 

 

Figure 5: Kaplan-Meier survival estimate by age groups 
Author's estimation using LFS (2014-15) 

 

 

 

Figure 6: Kaplan-Meier survival estimate by Marital Status 
Author's estimation using LFS (2014-15) 

Spouse job status is another cause of intentional unemployment in females. The social and 

cultural norms of society force females to stay at home as husbands are financial providers of 

their wives. Is causes females due to family pressure or by their own free will leave their job 

after marriage and stay at home. The relationship of spouse’s job status and female 

employment is presented. In figure 7, the graph 1 indicates that spouse is employed, 2 as the 

spouse unemployed, and 3 as a spouse not in the labour force. Our findings depict females 

with spouse’s unemployed are putting more effort to get employed once being unemployed. 

 

 

 

 



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

39 

 

Figure 7: Kaplan-Meier survival estimate by spouse's employment status 

Author's estimation using LFS (2014-15) 

Cox Regression  

The results presented in the table using the cox regression show hthe azard rate of being 

unemployed. Marital status (being married) and age with the negatsignssign are the indicators 

showing that the hazard increases as age increases or if the female is married. On the other 

hand, years of education show a positive impact on females’ unemployment duration. 

Province-wise break up shows that other than Sindh and Balochistan, the sign for Punjab and 

KPK are positive. Job-related training is also showing a significant impact on unemployment 

duration. Head education and migration status also have a negative sign othe n hazard rate of 

being unemployed for females.   

Table 4: 

_t Coef. Z P>z 

Marital Status -.0809233 -1.11 0.266 

Age -.0387243 -2.12 0.034 

Square of age .0004067 1.71 0.086 

Years of education .0193234 1.84 0.065 

Percapita income 5.92e-07 1.13 0.259 

Province    

Punjab .0194688 0.26 0.798 

Sindh -.1493358 -1.54 0.123 

Balochistan -.1414082 -1.07 0.285 

Region .0483527 0.75 0.455 

Head education -.0119239 -1.32 0.186 



GulRukh Zahid & Ambreen Fatima 

40 

 

Traning .1253401 1.66 0.097 

Migration -.0632756 -0.41 0.681 

 

Table 5: 

Covariates  Coefficients Exp(coefficients) P value 

Age (-3.87E-02) 0.962 0.034 

Square of age (4.07E-04) 0.9999 0.086 

Years of education (1.93E-02) 0.019 0.065 

Training  (1.25E-01) 1.133 0.097 

Marital status (-8.09E-02) 0.922 0.0777 

 

5. Conclusion and Policy Recommendations 

The study related to the unemployment duration of females in Pakistan was evaluated by 

employing Kaplan Maier and Cox regression to examine survival analysis for the extended 

unemployment of females in Pakistan, which results in the discouraging labour force in the 

long run due to the unavailability of job opportunities. For that purpose, the data is taken from 

the Labor Force Survey (LFS) 2014-15. The research concluded that a longer duration of 

female unemployment causes them hard to find a job again. The longer the duration, the fewer 

will be the chances to get employed again. 

The study concluded that the total time duration of re-entering into employment once 

being unemployed is 3650 days which is equal to 10 years, and this time is 121.67 months. It 

is concluded that the chances of remain unemployment duration increase, as the survival 

function decrease. The study determines that providing childcare will positively impact 

female labour force participation, as it is one reason females quit their jobs. This increased 

female labour force participation will help them to contribute to society by utilizing their 

talent. 

Therefore, it is highly recommended that for the better economic development of any 

country, it is imperative to engage females' participation in the labour force. For doing so, the 

following are some of the policy recommendations. 

The most important measure that the government should take on an immediate basis is 

increasing female child education investment. By doing so, labour force participation will 

increase as it has a positive impact on employment. In this regard, both print and electronic 

media should play their part. More incentives should be offered to bring out children into 

school, especially female children, as studies suggest that an educated mother always puts 

more effort to send her child to school. This research also suggests that females with a Master's 



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

41 

degree or higher education faces a longer unemployment rate. To address this issue, the 

government should encourage private-sector employers to employ females in decision-

making posts, and for the government-sector job, they should increase the female quota in 

jobs. It is also observed that females are the least priority for employers due to their respective 

household responsibilities and childcare. The government should provide economic childcare 

centres and rules to provide this facility to increase the number of females to work after having 

children. In our society, the socio-economic culture has created an environment that 

discourages females from working after marriage. We, as a society, work on this norm to be 

changed. The government's responsibility is to generate such incentives and a safe and secure 

environment that facilitates females to pursue their careers and dreams without fear. We have 

the example of our armed forces. They employ highly talented and educated females into 

forces and facilitate them so that they can put their hearts and soul and work passionately for 

the better future of the country. 

References 

Anderson, Pauline. "Women and unemployment: a case study of women's experiences of 

unemployment in Glasgow." PhD diss., University of Warwick, 1993. 

Arslan, Muhammad, and Rashid Zaman. "Unemployment and Its Determinants: A Study of 

Pakistan Economy (1999-2010)." Journal of Economics and Sustainable 

Development, ISSN (2014): 2222-1700. 

Bell, D. N., & Blanchflower, D. G. (2010). Youth unemployment: déjà vu? 

Chapman, B., Weatherburn, D., Kapuscinski, C. A., Chilvers, M., & Roussel, S. (2002). 

Unemployment duration, schooling and property crime. 

Demidova, Olga, and Marcello Signorelli. "Determinants of youth unemployment in 

Russian regions." Post-Communist Economies 24, no. 2 (2012): 191-217. 

Eberwein, C., Ham, J. C., & LaLonde, R. J. (2002). Alternative methods of estimating 

program effects in event history models. Labor Economics, 9(2), 249-278. 

Flek, V., Hála, M., & Mysíková, M. (2015). Duration dependence and exits from youth 

unemployment in Spain and the Czech Republic. Economic Research-Ekonomska 

Istraživanja, 28(1), 1063-1078. 

Kavkler, Alenka, Daniela-Emanuela Danacica, Ana Gabriela Babucea, Ivo Bicanic, 

Bernhard Bohm, Dragan Tevdovski, Katerina Tosevska, and Darja Borsic. "Cox 

regression models for unemployment duration in Romania, Austria, Slovenia, 

Croatia and Macedonia." Romanian Journal of Economic Forecasting 10, no. 2 

(2009): 81-104. 



GulRukh Zahid & Ambreen Fatima 

42 

 

Kettunen, Juha. "Education and unemployment duration." Economics of education review 

16, no. 2 (1997): 163-170. 

Khan, T., & Yousaf, F. (2013). Unemployment duration of first time job seekers: A case 

study of Bahawalpur. Asian Journal of Economic Modelling, 1(1), 8-19. 

Kipesha, Erasmus Fabian, and Robert Msigwa. "Determinants of youth unemployment in 

developing countries: Evidences from Tanzania." (2013). 

Kisto, M. (2014). Discouraged Youth in Mauritius. International Journal of Recent Research 

in Interdisciplinary Sciences, 1(3), 1-8. 

Klasen, Stephan, and Janneke Pieters. What explains the stagnation of female labor force 

participation in urban India? The World Bank, 2015. 

Meghir, Costas, Yannis Ioannides, and Christopher Pissarides. "Female participation and 

male unemployment duration in Greece." European Economic Review 33, no. 2-3 

(1989): 395-406. 

Msigwa, R., & Kipesha, E. F. (2013). Determinants of youth unemployment in developing 

countries: Evidences from Tanzania. 

Moffitt, R., & Nicholson, W. (1982). The effect of unemployment insurance on 

unemployment: the case of federal supplemental benefits. The Review of Economics 

and Statistics, 1-11. 

Mortensen, D. T. (1986). Job search and labor market analysis. Handbook of labor 

economics, 2, 849-919. 

Novella, M. L., & Duvivier, M. (2015). The relationship between unemployment duration 

and education. 

Nonyana, J. Z. (2015). Statistical modeling of unemployment duration in South Africa 

(Doctoral dissertation). 

O'higgins, Niall. "The challenge of youth unemployment." International Social Security 

Review 50, no. 4 (1997): 63-93. 

Pekkala, Sari, and Hannu Tervo. "Unemployment and migration: does moving help?" 

Scandinavian Journal of Economics 104, no. 4 (2002): 621-63 

Scarpetta, S., Sonnet, A., & Manfredi, T. (2010). Rising youth unemployment during the 

crisis. 

Sansale, R., DeLoach, S. B., & Kurt, M. (2019). Unemployment duration and the 

personalities of young adults' workers. Journal of Behavioral and Experimental 



Journal of Applied Economics and Business Studies, Volume. 6, Issue 3 (2022) 23-44     https://doi.org/10.34260/jaebs.632 

43 

Economics, 79, 1-11. 

Tansel, A., & Taşçı, H. M. (2010). Hazard analysis of unemployment duration by gender in 

a developing country: The case of Turkey. Labor, 24(4), 501-530. 

Tansel, Aysit, and H. Mehmet Tasci. "Determinants of unemployment duration for men and 

women in Turkey." (2004). 

Verho, J. (2008). Determinants of unemployment duration over the business cycle in 

Finland. HECER Discussion paper 226. 

Verho, Jouko. "Determinants of unemployment duration over the business cycle in 

Finland." (2008) 

  



GulRukh Zahid & Ambreen Fatima 

44 

 

 

APPENDIX 

 

Figure:  Author's estimation using LFS (2014-15)                              Figure: Kaplan-Meier failure function 

 

Figure:  Survival functions showing the education related information