AN INVERSE U-SHAPE RELATIONSHIP BETWEEN RATE OF UNEMPLOYMENT AND LEVEL OF EDUCATION:SOME EVIDENCES FROM PAKISTAN

AYUB KHAN MEHAR

APPLIED ECONOMICS RESEARCH CENTRE

UNIVERSITY OF KARACHI

KARACHI (PAKISTAN)

1.  Introduction

Judgement about the importance of various influences on employment have a long history .since the early 1940’s when providing the job opportunities become an operative objective of economic development, knowledge on the subject has certainly improved. A number of analytical treaties have focused on the issues. Pareto[1972],Fishlow[1972],Pen[1974],Rado[1976],Solow[1980],Okun[1981] and more important data on unemployment are routinely reported for about a score of developing countries. From the reported data sets of developing countries, an ‘Inverse U shape” relationship between unemployment rate and level of education has commonly observed. In next section we tested the existence of this relationship .While in section: III, we discussed our opinions in respect of the causes of this relationship.

II. Existence of the Relationship

This study is mainly based on a sample of 12483 individuals covered in socio-economic survey of Karachi carried out by the Applied Economics research centre, University of Karachi[1988].However ,to confirm the validity and consistency of the conclusions we derived some tables of classification from Pakistan Labour force survey(Govt. of Pakistan1974;1988) and from urban Indian data(Krishnamurthy 1973).

For measuring the influences of various factors in determination of unemployment ,we applied a logit model approach. According to our hypothesis:

Ui = f (Xij)

Where

Ui = A dummy dependent variable; equal to one if ith individual is

Unemployed.

Xij = Jth explanatory variable which explain employment status of ith individual

While, in matrix Xt following explanatory variables are included.

AGE: Age have been included to proxy of time for searching or getting a job; SEX: it is envisaged that female unemployment would be greater than male. It has been included as a dummy variable with male serving as reference category. Education: Level of education is also supposed to influence on employment. We have included five level of education namely below metric (BELMAT), Intermediate (INTER),

VOLUME NO. 42.NO.4

Graduate(BACHLR),Post-graduate(MASTER) and professional qualified (PROFSNL) with metric being the reference category.

In our study the ethnic groups are include.2.ETHA represents to Muhajir, Gujrati and Kuchhi; ETHB represents to Sindhis and ETHC represents to all of the other ethnic groups.

The results of the logit regression presented in table 1; while table 2 shows the pattern of unemployment by educational levels and age group. Table 1 and Table 2 shows the results of Karachi based data set(AERC:1988).While table 3 to table 5 present the same conclusions from other data(Govt. of Pakistan: 1988; Krishnamurthy: 1973).

The principal conclusion of the study is that the unemployment rate at intermediate level of education is significantly higher than the unemployment rate in all the other level of education. This result is consistent in all the distributions and data sets. Through estimated logit equation, it is observed that probability of being unemploy of an intermediate is significantly higher than other level of education.

The results are also consistent with Pasha,Hyer and Arshad[1978] findings. According to them, in Pakistan unemployment is higher in matriculate, but in urban Sindh it is higher at intermediate level of education. However, according to Rado[1976],during 1974-75,unemployment of matriculate was higher than other level of education, in Karachi.

III. Causes of the Relationship

Unemployment is a two way problem:Supply side and demand side.

1). Supply side problems including:

1.  A higher drop out rate at a certain level of education because after that

competition become much higher or educational expenditures become much

higher.

2.  A higher drop out rate on a certain level of age which may be an average age of marriage in the community or may be an average age of maturity /earning responsibility.

2) Demand side Problems including:

1.  Demand patterns of educated labour force by the employer.

2.  Age limits for entry in labour market (18 years in Pakistan).

THE INDIAN ECONOMIC JOURNAL

TABLE 1

Logit Regression Results

Variable / Coefficient / T-Ratio
CONSTANT
AGE
SEX
ILLET
INTER
BACHLR
MASTR
PROFSNL
ETHB
ETHC / 0.636
-0.007
0.012
-0.025
0.072
0.052
0.043
0.041
-0.011
-0.022 / 59.47
-30.13
1.17
-2.95
1.18
4.65
2.22
2.78
-1.31
-2.98
-2 LOGLIKLIHOOD
NUMBER OF OBSERVATION / 10030
(12473)
12483
Table of classification
Predicted
0 1 / Percent Correct
Observed / 0 / 9663 / 892 / 91.55%
34.60%
1 / 1261 / 667
Overall / 82.75%

VOLUME NO. 42.NO.4

TABLE 2

Rate of Unemployment by Education and by Age

(Karachi based Survey:1988)

By Level of Education / By Age Group
Educational Level / Rate of Unemployment / Age Group / Rate of Unemployment
Overall
IIIeterate
Less than
Matric
Matric
Intermediate
Graduate
Master
Professional / 11.55
5.95
7.04
15.83
20.39
7.55
8.46
5.07 / Overall
15-19
20-24
25-29
30-34
35-39
40-44
45-49
50-54
55-59
60-+ / 11.55
23.39
19.73
9.49
4.47
3.27
2.35
1.80
1.38
1.44
1.59

TABLE 3

Rate of Unemployment by Education and by Age

(Urban India :1967)

By Level of Education / By Age Group
Educational Level / Rate of Unemployment / Age Group / Rate of Unemployment
Overall
IIIeterate
Less than Matric
Matric &
Intermediate
Graduate % + / 1.60
0.69
1.68
3.82
2.18 / Overall
15-19
20-24
25-29
30-39
40-49
50-59 / 1.60
5.64
4.14
1.48
0.61
0.51
0.52

THE INDIAN ECONOMIC JOURNAL

TABLE 4

Rate of Unemployment by Education Level and Territory(%)

(1974-75)

Education\Territory / Punjab / Sind / N.W.F.P / Bilochistan
Urban / Rural / Urban / Rural / Urban / Rural / Urban / Rural
Overall
IIIeterate
Less than Matric
Matric
Intermediate
Graduate
Master / 2.81
1.71
4.00
5.09
6.11
4.96
2.60 / 1.97
1.19
2.35
15.31
43.39
0.00
0.00 / 2.10
1.90
2.20
2.79
2.48
5.03
1.69 / 0.53
0.30
1.55
3.29
0.00
0.00
0.00 / 3.53
1.95
3.70
8.86
9.49
7.48
0.00 / 3.02
1.01
8.81
30.67
0.00
0.00
0.00 / 1.11
1.19
2.49
0.04
0.00
0.00
0.00 / 1.81
2.65
0.00
0.00
0.00
0.00
0.00

TABLE 5

Rate of Unemployment by Education Level and Territory(%)

(1988-89)

Education\Territory / Punjab / Sind / N.W.F.P / Bilochistan
Urban / Rural / Urban / Rural / Urban / Rural / Urban / Rural
Overall
IIIeterate
Less than Matric
Matric
Intermediate
Graduate
Master / 7.02
6.14
5.63
7.72
7.35
4.93
3.41 / 3.43
2.51
3.31
9.59
10.59
3.14
13.60 / 3.68
2.87
2.36
3.42
2.78
5.28
4.05 / 0.80
0.57
0.60
3.94
5.43
6.92
0.00 / 5.60
3.81
4.23
7.17
5.26
17.19
2.69 / 5.74
3.42
8.39
11.47
19.64
24.07
0.00 / 3.17
3.54
1.48
0.00
0.00
15.11
5.90 / 0.55
0.53
0.35
0.00
6.92
0.00
0.00

IV. Policy Implications

In the light of the above analysis, the following strategies for improving the relationship between education and employment are recommended:

Position of Unemployment Curve / Recommendations
1) Above than level of natural unemployment
/ 1. Minimize private cost of
Education at all levels.
2.  Increase in minimum age of entry in labor force
3.  Incremental salary of one year’s experience should be less than incremental salary of one year’s(further) education. So that, education share in life time increase and working share in life decrease.
4.  Motivate to employers and make rule to prefer higher education in recruitment.
2) Below than level of natural unemployment
/ 1. Maximize private cost of
Education at each level,so that
size of labor force
increase(shifting the students
into labor force)
2. Decrease minimum age of
entry in labor force
3.Incremental salary of one year’s experience should be higher than incremental salary of one year’s education.
3) Intersect to the line of natural unemployment
/ 1. Minimize private cost of
Education at kinked levels.
2.Reduce competition at above
kinked either through
increase in number of seats
or through decrease in fees.

THE INDIAN ECONOMIC JOURNAL

NOTES

1  By ‘Professional Qualification’ we mean only those qualification, where the period of completion the education is greater than required period of master degree.

2  In interpreting the result, it is notable that all the ethnic groups except Muhajir,Gujrati and Kuchhi belong to up country. In Pakistan the services in public sector are constraint to the population based provincial quota. In the province of Sindh, this quota is more divided into urban-rural areas(To some extend based on ethnic distribution)