A micro analysis of demand-side determinants of schooling in urban Pakistan *.

Hamid, Shahnaz

1. INTRODUCTION

Investment in human capital is as important for growth as physical capital. The objective of this paper is to establish the importance of some factors, that affect the decision of a household to send their children who are of school going age, between 5-14 years, to school in urban areas.

This paper is divided into five sections, Section 2 reviews the relevant schooling literature for Pakistan. Section 3 points some theoretical hypotheses describing the data and Section 4 presents the analysis, and finally Section 5 presents some conclusions.

2. A SUMMARY LITERATURE SURVEY OF THE DEMAND FOR SCHOOLING IN THE CONTEXT OF PAKISTAN

So far in Pakistan, limited work has been done in this field. Irfan using the PLM (1) survey data, concluded that the socio-economic status of a household is positively related to children's schooling. However, this may be a limited explanation.

Lack of schools and socio-cultural norms also play an important role to keep children away from school, especially in the case of gifts. Moreover, Irfan's analysis is limited to 10-14 years of age, whereas it is more important to examine the 5-9 years old age cohort.

Burney and Irfan again using the same PLM data set, carded out a more comprehensive analysis for both urban and rural Pakistan, and concluded that household income, parental education and tenurial status with respect to land has a positive impact on a household's decision to send a child to school.

We agree with the importance of parents' education, but consider that decisions taken by the head of the household also plays an important role, especially in case of girls.

3. DATA AND SOME HYPOTHESES

Data

The present study is based on survey data, collected for the project Food Security Management in 1986 by Pakistan Instiute of Development Economics and International Food Policy Research Institute. Our analysis is based on 792 households, who have children in the age cohort 5-14 years.

Hypotheses

Our hypothesised model is that, the demand for child schooling is constrained by household socio-economic characteristics. (2)

Keeping in view four important variables, income of the household, occupation, education and gender of household head, three different aspects of demand for child schooling will be analysed.

(a) Incidence of households sending children to school.

(b) Children attending school.

(c) (3) A comparison between male and female children's schooling.

Household Income

Our major hypothesis is that poverty often compels the household to keep their children away from school.

To see the impact of household's income, we are taking into consideration, the total income of the household earned by all adult members in the age cohort above 14 years.

We are also analysing the impact of per capita income because aggregate figures for household income do not give a reliable estimate, and household size modifies the effect of income.

Occupation

For more diverse estimates of socio-economic status, we disaggregated income into two variables, occupation of the household head, and the periodicity of their earning, monthly, weekly, daily. Occupation of the head of household in general reflects the level of income as well as social class. Our hypothesis is that occupational status of household head and demand for children schooling have a positive relationship.

Education

Occupation of the household head in general reflects their level of education. Education is important because of its impact on income and a bias towards children's education. Therefore, the educational level of the household head and children's schooling will also be positively related.

Gender of the Household Head

In a household, the head of the household has greater power in decision-making. In a household headed by a male, females will, therefore, have a lower decision-making power. Our hypothesis is that a lower proportion of children will go to school where household's head are males, especially in the case of female children because of prevalent social biases against women.

4. EMPIRICAL RESULTS

Incidence of Household Sending their Children to School

Household Income

Our major hypothesis is that household income is positively related to child schooling. The first explanatory variable is total income of the household per month.

Table 1 shows that in the lowest income group, below < Rs.1000, the percentage of households sending all their children to school is lowest, at 33.7 percent. Households with a total monthly income of Rs 1000 -< Rs 2000, 41.2 percent send all their children to school. In the income group, Rs 2000-< 3000, 43.2 percent of households send all their children to school. In the income group, Rs. 3000-< Rs 4000, 50.9 percent of the households send all their children to school. And for the highest income group, 50.8 percent of the households send all their children to school.

In keeping with this, another variable, households at least not sending one of their children to school, shows a negative relationship with household income. Our results support our hypothesis regarding income and child schooling.

Table 2 shows, that in the lowest per capita income group below Rs <100, only 32.4 percent of the households send all their children to school. Subsequently as the per capita income category increases the proportion of households sending all their children to school also rises from 32.3 percent to 54.3 percent. Table 2 makes it clear that as we move from the lowest per capita income group to a higher one, the percentage of households with at least one child not in school drops from 67.6 percent to 45.6 percent. Thus supporting our hypothesis regarding monthly per capita income.

Household Heads' Occupation and Periodicity of their Earning

To analyse our hypothesis regarding occupational status, we have a. value judgement, that occupation of the head of the household can be ranked in the following way. The highest status is for government employees, followed by private employees, the self-employed, the sick and labourers. We can not give any value judgement to undefined occupational status. The table show that the percentage of household with all children in school decreases from 58.6 percent to 43.2 percent, 38.5 percent, 36.0 percent, 22.8 percent with a change in the occupational status of the household head from government employee to labour and vice versa (Table 3).

The above evidence supports our hypothesis that the household head's occupational status is positively related to households with all children in school.

Periodicity of household heads' earning reflects stability in the job, income and socio-economic stares of a household. So our hypothesis is that, periodicity of household heads' earning will affect the household's demand for child schooling. Households in which the household head has a stable and reliable source of earning, the demand for children's schooling will be higher.

We have a value judgement that the periodicity of earnings can be ranked in the following way. The highest status is for monthly income earning, followed by profit, weekly, daily irregular, doing nothing and in kind.

As shown in Table 4, the percentage of households with all children in school drops from 50.50 (monthly earner) to 32.45 in case of irregular sources of earning. Our ranking of household heads clearly supports our hypothesis, that the periodicity of the household head's earning is related to income, and uncertainty in household head's earning is positively related to households sending all their children to school.

Education of Household Head

Our next hypothesis is that the household head's educational level is positively related to children's schooling (Table 5).

It becomes clear that the percentage of households with all children in school increases from 29. 4 percent in the case of an illiterate household head to 83.3 percent where the educational level of household head is the highest i.e. university education. Thus supporting our hypothesis of a positive relationship between the educational level of household head and households with all their children "in school. The table shows an inverse relationship between the household head's educational level and households with at least one child out of school.

Gender of the Household's Head

An hypothesis is constructed on the basis of our field observation, both from rural and urban Pakistan. Females realise the importance for education; but having no say in decision-making, can not impose their decision, especially in the case of girl.

The interesting results that emerges from Table 6 shows that among the households headed by males, 40 percent send all their children to school, whereas the percentage increases to 49.1 in ease of female-headed households. On the other hand, households headed by males contained a higher percentage of households with at least one child not in school, while for female-headed household, it drops to 50.9 percent. Evidence suggest that households headed by females are more inclined to educate their children.

Incidence of Child School Enrollment

In the following section, we have tried another estimate of the variable to explain child school enrollment. We will explain child school enrollment with all our previous determinants.

Household Income (Household Total and per Capita Income)

Table 7 clearly shows, that as the total household income per month increases, child school enrollment also increases. In the table, the households having total income < Rs 1000 per month, only 56.6 percent of the total children are in school, enrollment increases with each income category, from 56.64 percent to 59.42 percent, 61.83 percent, 71.9 percent and 70.53 percent.

From Table 8, it becomes obvious that as we move from the lowest per, capita income group to the highest per capita income group the percentage of total children attending school also improves from 59.76 to 69.07 percent.

Household Head's Occupation, and Periodicity of Earning

Table 9 presents the evidence for household head's occupational status, and child school enrollment. Households having occupational status i.e. government employee, gives the highest percentage i.e. 75.39 percent of child enrollment. Whereas the percentage of child enrollment drops down from 75.39 to 38.6 percent with a change in occupation status from government employee to labourer

Table 10 shows a positive relationship between the periodicity status of household head earning, and child school enrollment. This relationship is smooth, upto, where the periodicity status of the household head's earning is daily, for the irregular nature of periodicity status of household head earning the results are not very clear.

But on the basis of other evidence we may say that, there is a positive relationship between periodicity status of household head's earning and child schooling.

Educational Level of Household Head

Table 11 presents the relationship between the household bead's educational level and child school enrollment ratio. A look at Table 11 shows that there exists a positive relationship between the two. The increase in household head's educational level, the child school enrollment ratio also increases from 33.33 percent to 88.75 percent. Thus supporting our hypothesis concerning child school enrollment and household head's education.

So evidence clearly supports our hypothesis that child school enrollment is positively related to the household head's educational level.

Household Heads Gender

Table 12 shows the relationship of households headed by males and females to child schooling. Table 12 shows the positive impact of female-headed households, on child school enrollment. As can be seen, households headed by a male had a 59.97 percent enrollment ratio, while households headed by a female had a 71.25 percent enrollment ratio. So the evidence supports our hypothesis that households headed by a female will be more inclined to send their children to school.

CONCLUSION

The present study suggests that poverty is the key determinant of demand for child schooling. Various manifestations of poverty, such as the low income of households, low occupational status of the household head; periodicity of household head's earning, and low level of household head's education, together with household economy dominated by a male, play an important role in constraining the demand for children's schooling.

REFERENCES

Burney, Nadeem A., and M. Irfan (1991) Parental Characteristics, Supply of Schools, and Child School Enrolment in Pakistan. The Pakistan Development Review 30:1.

Irfan, M. (1980) Poverty and Household Demographic Behaviour in Pakistan: Insights from PLM Survey 1979. Islamabad: Pakistan Institute of Development Economics. (PLM Project Report No. 11.)

* Owing to unavoidable circumstances, the discussant's comments on this paper have not been received.

(1) Population, Labour Force and Migration.

(2) This present study is a part of a larger, on-going project on "Why children do not go to school". Our major hypothesis for that study is that the household, the educational system, and opportunities for child labour, all together determine children's schooling.

(3) Due to aspace constraint this part has not been included in the paper: any one interested can obtain it from the author.

Shahnaz Harold is Research Economist at the Pakistan Institute of Development Economics, Islamabad.

Table 1
Distribution of Households Sending their Children (5-14 Years of Age)
to School by Household Income

Total Income of Total Number Households with Household with
 the Household of Households All Children At-least One
 per Month Falling in in School Child not in
 (in Rs) Category School (in Rs)

 No. % No. % No. %

<1000 243 100.00 82 33.7 161 66.3
1000<2000 308 100.00 127 41.2 181 58.8
2000<3000 125 100.00 54 43.2 71 56.8
3000<4000 57 100.00 29 50.9 28 49.1
>4000 59 100.00 30 50.8 29 49.2
Total 792 100.00 322 40.7 470 59.3

Table 2
Distribution of Households Sending their Children (5-14 Years of Age)
to School by per Capita Income per Month

 Total Number of Households Household
 Households with All with At-least
Household per- Falling Children One Child not
Capita Income in Category in School in School
per Month
(in Rs) No. % No. % No. %

<100 148 100.00 48 32.4 100 67.6
100<200 332 100.00 124 37.3 208 62.7
200<300 148 100.00 64 43.2 84 56.8
300<400 85 100.00 43 50.6 42 49.4
>400 79 100.00 43 54.4 36 45.6

Total 792 100.00 322 40.66 470 59.34

Table 3
Distribution of Households Sending their Children
to School by Household Head's Occupation

 Total Number Households Households with
 Household Households with All At-least One
 Heads' Falling in Children in Child not
 Occupation Category School in School

 No. % No. % No. %

Govt Employee 111 100.00 65 58.6 46 41.4
Private Employee 125 100.00 54 43.2 71 56.8
Self Employed 268 100.00 102 38.05 166 61.94
Sick 25 100.00 9 36.0 16 64.0
Labour 101 100.00 23 22.8 78 77.2
Other 11 100.00 7 63.6 4 36.4
At Home 133 100.00 57 42.8 76 57.14
Students 1 100.00 0 0.0 1 100.0

Total 775 100.00 317 40.57 458 59.42

Table 4
Distribution of Households Sending their Children (Aged 5-14 Years)
to School by Periodicity of Households' Earning

 Number of Households Households
 Households with All with At-least
 Falling Children One Child not
Periodicity of in Category in School in School
 Household
Heads' Earning No. % No. % No. %

Monthly 206 100.00 104 50.50 102 49.51
Profit 141 100.00 53 37.58 88 62.41
Weekly 32 100.00 11 34.37 21 65.62
Daily 108 100.00 37 34.3 71 65.70
Irregular 114 100.00 37 32.45 77 67.54
Doing Nothing * 168 100.00 72 42.9 96 57.10
In kind 6 100.00 3 50.00 3 50.00

Total 775 100.00 317 40.9 458 59.1

* Retired, old, sick, unemployed etc.

Table 5
Distribution of Households Sending their School Going Age
Children (Aged 5-14 Years) to School by Educational Level
of Household Head

 Total Number Households Households
 Educational of Households with All with At-least
 Level of Falling in Children One Child not
Household Head Category in School in School

 No. % No. % No. %

Apprentices 2 100.00 0 0.0 2 100.00
Illiterate 337 100.00 99 29.4 238 70.6
Literate 123 100.00 45 36.6 78 63.4
Primary 131 100.00 53 40.5 78 59.5
Middle 80 100.00 36 45.0 44 55.0
Secondary 86 100.00 62 72.1 24 27.9
University 30 100.00 25 83.3 5 16.7
Other 3 100.00 2 66.7 1 33.3

Total 792 100.00 322 40.7 470 59.3

Table 6
Distribution of Households ,Sending their Children
(Aged 5-14 Years) to School by Gender of the Household Head

 Gender of Total Number Households Households
the Household of Households with All with At-least
 Head Falling in Children One Child not
 Category in School in School

 No. % No. % No. %

Male 737 100.00 295 40.0 442 60.00
Female 55 100.00 27 49.1 28 50.9

Total 792 100.00 332 40.7 470 59.3

Table 7
Distribution of School Going Children (Aged 5-14 Years)
by Total per Month Income of the Household

 Total No. of Children in
 Children whose Children in Income Category
 Household Falling Income Category not Going
Total House- in Category Going to School to School
hold Income
 Rs (pm) No. % No. % No. %

<1000 708 100.00 401 56.64 307 43.36
1000-<2000 892 100.00 530 59.42 362 40.58
2000-<3000 372 100.00 230 61.83 142 38.17
3000-<4000 171 100.00 123 71.9 48 28.07
>4000 190 100.00 134 70.53 56 29.47

Total 2333 100.00 1418 60.78 915 39.22

Table 8
Distribution of School Going Children (Aged 5-14 Years) by
per Capita Income of the Household per Month

 Children in
 Total No. of Children in Income Category
Household Children Falling Income Category not Going to
per Capita in Category Going to School School
Income per
Month (in Rs) No. % No. % No. %

<100 579 100.00 346 59.76 233 40.24
100-<200 970 100.00 560 57.73 410 42.27
200-<300 381 100.00 234 61.42 147 38.58
300-<400 209 100.00 144 68.90 65 31.10
>400 194 100.00 134 69.07 60 30.93

Total 2333 100.00 1418 60.78 915 39.2

Table 9
Distribution of School Going Children (Aged 5-14 Years)
by Occupational Status of Household Head

 Total Children Children in Children in
 whose Households Occupational Occupational
 Falling in Category Going Category not
 Category to School Going to School

Household
Heads
Occupation No. % No. % No. %

Govt. Employee 317 100.00 239 75.39 78 24.61
Private Employee 338 100.00 221 65.38 117 34.62
Self Employed 826 100.00 498 60.29 328 39.71
Sick 73 100.00 42 57.53 31 42.47
Labour 259 100.00 100 38.61 159 61.39
Other 31 100.00 21 67.74 10 32.26
At Home 487 100.00 297 60.98 190 39.01
Students 2 100.00 0 0.0 2 100.0

Total 2333 100.00 1418 60.78 915 39.22

Table 10
Distribution of Households Sending their Children
(Aged 5-14 Years) by Periodicity of Household Heads' Earning

 Children
 Total Children Children Falling in
 whose Households Falling in Category not
 Falling in Category Going Going to
 Periodicity of Category to School School
Household Head's
 Earning No. % No. % No. %

Monthly 591 100.00 412 69.71 179 30.29
Profit 419 100.00 263 62.76 156 37.23
Weekly 87 100.00 51 58.62 36 41.37
Daily 312 100.00 159 50.96 153 49.04
Irregular 323 100.00 172 53.25 151 46.75
Doing Nothing 582 100.00 353 60.65 229 39.35
In Kind 19 100.00 8 42.11 11 57.89

Total 2333 100.00 1418 60.78 915 39.22

Table 11
Distribution of School Going Children (Aged 5-14 Years)
by Educational Level of Household Head

 Total Children Children in
 Educational whose Households Children in Category not
Level of House- Falling in Category Going Going to
 hold Head Category to School School

 No. % No. % No. %

Apprentices 3 100.00 1 33.33 2 66.67
Illiterate 992 100.00 481 48.49 511 51.51
Literate 347 100.00 210 60.52 137 39.48
Primary 383 100.00 241 62.92 142 37.08
Middle 263 100.00 194 73.76 69 26.24
Secondary 255 100.00 211 82.75 44 17.25
University 80 100.00 71 88.75 9 11.25
Other 10 100.00 9 90.00 1 10.00

Total 2333 100.00 1418 60.78 915 39.22

Table 12
Distribution of School Going Children (Aged 5-14 Years)
by Household Head's Gender

 Total Children
Gender of the whose Households Children in Children in
 Household Falling in Category Category not
 Head Category Going to School Going to School

 No. % No. % No. %

Male 2166 100.00 1299 59.97 867 40.03

Female 167 100.00 119 71.25 48 28.74

Total 2333 100.00 1418 66.78 915 39.22