1

Determinants of Urban Youth Unemployment; Evidence From East

Gojjam Zone of Amhara Region.

Aynalem Shita ( MSc)1

Mulugeta Dereje (Msc)2

Abstract

Recently, urban youths’ unemployment is a hot issue and becomes the main socio-economic problem in

Ethiopia. It creates many economic and social problems in the economy. Despite the severity of urban

youth unemployment worsens overtime in Ethiopia; researches done on the area are scanty. The main

objective of this study is to identify and examine the demographic and socio-economic determinants of

urban youth unemployment at East Gojjam Zone of Amhara Region. To achieve the specified objective,

both primary and secondary data sources were used. The primary data was collected from 397 sample

respondents through structured questionnaire from Debremarkos, Bichena and Motta towns

proportionally. To supplement the primary data, secondary data was also gathered from published and

unpublished sources. To come up with the results the researchers employed both descriptive and

inferential analysis. The study found that 48.1% of the respondents are unemployed while 51.9% of them

are employed. In the case of inferential analysis binary logit model was used. Hence, Variables such as

age, work experience, skill match, social network and family prosperity are identified as negative and

significant determinants of urban youth unemployment. Whereas education and migration status of urban

youths affected unemployment positively and significantly. Hence, efforts should be made to reduce the

level of unemployment by increasing job opportunity for educated and non- experienced youths, reducing

rural-urban migration and provision of relevant information for job seekers.

Keywords: unemployment, East Gojjam and Binary Logit

1 Lecturer, Department of Economics, Debre Markos University

2 Lecturer, Department of Economics, Debre Markos University

2

1. Background of the Study

Unemployment is one of the main problems in the world economy today. Many countries at different

levels of development are trying to cope with this problem. International Labor Organization (1992)

defines that unemployment is the situation of being out of work or need a job and continuously searching

for it in the last four week or unemployed ( age 16 or above) but available to join work in the next two

weeks. People who voluntarily do not want to work, full time students, retired people and children are no

included in unemployed category. In short, unemployment means the state when people who are willing

and able to do a job but fail to get the desired job.

Youth unemployment is a problem that affects most countries. The ability of youth to engage in

productive activities has both social and economic consequences for an economy. Youth unemployment is

often higher than the unemployment rate for adults highlighting the concerns that many countries face in

facilitating the transition from school to work. In developing countries, youth face not only the challenge

of obtaining productive employment, but also obtaining safe and acceptable works (Broussara and

Tsegay, 2012).

Government organizations, NGOs and civic association in different countries adopt and use various age

ranges for the concept “youth” from the stand point of the purpose which they stand for and the activities

they undertake. For Example, The United Nation (UN) defines the youth as persons between 15-24 years;

WHO, 10-24. In Ethiopia, according to the national youth policy, youth include part of the society who

are between 15-29 years (Ministry of youth, sports and culture, 2004).

According to ILO (2004) figures, the Sub-Saharan Africa region has the highest rate of youth

unemployment (18.4 per cent) after the Middle East and North Africa (21.3 per cent). If this trend

persists, it will have considerable effects on human capital in the region, as well as on the region’s

economic potential. Creating decent and productive work for young people in the Sub-Saharan region

could result in a potential GDP increase of 12 to 19 per cent (Berhanu etal., 2005).

Unemployment in Ethiopia is more of a problem of urban youth than that of rural. According to Ethiopian

labor force survey report, the unemployment rate of urban youth at country level were 22.9 while for rural

youth remained at 3.1% only (LFS, 2013).

3

Based on the above facts, this study was tried to analyze the socio-economic & demographic determinants

of urban youth unemployment in East Gojjam zone of Amhara National Regional State, Ethiopia.

2. Statement of the problem

Like many other developing countries unemployment has been one of the major problems in Ethiopia.

The excessive rate of unemployment negatively impacts on economy which causes unstable economic

conditions. This is troublesome because when workers are unemployed, there is an under-utilization of

resources. So the total production of a country is less than its potential level of output because resources

are not fully utilized in these countries (Maqbooletal.,2013)

A high level of un- and underemployment is one of the critical socio-economic problems facing Ethiopia.

While the labour force grows with an increasing proportion of youth, employment growth is inadequate to

absorb labour market entrants. As a result, youth are especially affected by unemployment. Moreover,

young people are more likely to be employed in jobs of low quality, underemployed, working long hours

for low wages, engaged in dangerous work or receive only short term and/or informal employment

arrangements (Berhanu et al., 2005).

In Ethiopia, the labour force is growing with an increasing proportion of youth and employment growth is

inadequate to absorb this high proportion of labor force specially the youth part in different sectors of the

economy (Guarcello and Rosati, 2007 cited at Alemnew, 2014). The country is also one of the highest

urban unemployment rates worldwide, at about 50% of the youth labour force (Berhanu et al, 2005).

A rapid increase in population raises many socio-economic problems in the economy. If people cannot

find jobs in their home country, they may be tempted to relocate to another country for getting jobs. This

can be dangerous for the future of a nation, particularly if other nations are attracting its brain drain.

Therefore, if this problem continuously persists in any economy, it could be a major factor in deteriorating

the economic growth. Additionally, persistent unemployment not only affects the status of a nation in

comparison to other nations, but it also leads to cruel home country problems. Long-term unemployment

always results in creating financial hardships, poverty, homelessness, crime, frustration and many other

4

problems like breakdown and family tension, social isolation, loss of confidence and self-esteem. All

these lead to the erosion of a healthy society (Maqbool et al., 2013).

Unemployment is of a special concern for Ethiopians and has a wider implication for the youth in addition

to leading their life as expected to help parents and extended families (Shumet, 2011). According to a

survey in 55 urban areas, unemployment was estimated at 41.3% and the incidence of youth

unemployment was 45.5% and 35.7% for females and males respectively (MOLSA 1992 cited in

Alemnew, 2014).

Even though few studies are conducted on youth unemployment in Ethiopia (Asalfew (2011), Tegegne

(2011), Amanuel (2016), Dejen et.al.(2016) and Alemnew (2014)), the results of these studies contradict

each other which needs further study based on the specific socio-economic situation of the study area.

Hence, this study is conducted on the determinants of urban youth unemployment at East Gojjam Zone of

Amhara Region, Ethiopia .

3. Objective of the study

The general objective of this study is to analyze the determinants of urban youth unemployment in the

case of East Gojjam zone of Amhara region, Ethiopia.

This study has the following specific objectives:

To assess the characteristics of unemployment at the study area.

To identify and examine the demographic and socio-economic determinants of urban youth

unemployment.

4. Literature Review

The ILO estimates that the number of unemployed youth is on the rise again since 2011, after declining

from the peak it reached at the height of the global financial crisis. It is expected to reach 74.2 million

young people by 2014 based on (ILO, 2010). The global youth unemployment rate has also been rising

since 2011. It is currently estimated at 12.6% and is projected to increase to 12.8% by 2018.

The study by Echebiri (2005), on the basis of youth in Umuahia city in Nigeria, finds that unemployment

is influenced by age, marital status, dependency ratio, education, current income and employment

5

preference (paid or self-employment). Alhawarin and Kreishan (2010) also indicate that age, gender,

marital status, region, work experience and educational level are the major determinants of unemployment

in Jordan.

Studies from Ethiopia indicate that the potential causes of unemployment in urban Ethiopia include

increasing number of youth labor force, the rising internal migration, literacy rate, poor to modest

macroeconomic performance, low level of job creation and low level of aggregate demand in the

economy (Getinet, 2003; WB, 2007). Youth unemployment is the outcome of different socio-economic

and demographic factors at macro and micro level. The micro level factors are directly associated to

individuals’ demographic and socioeconomic attributes while the macro level factors are related to the

national issues (Toit, 2003). This study emphasizes on assessing individuals’ demographic and

socioeconomic attributes that influence youth employment.

Asalfew (2011), the multivariate analysis showed that sex, migration, education, social network, job

preferences and access to business advisory services significantly determine youth unemployment in

Debre Birhan town. However, household income, father education, and marital status were found

insignificantly related to youth unemployment.

According to Tegegne (2011), examined the association between socio-demographic variables and

unemployment in Addis Ababa, the econometric analysis has confirmed that sex and age are statistically

significant and have negative relationship, signifying the inherent problem of unemployment among

women and the youth. Regarding migration status, in spite of the type of job, a migrant is more likely to

be employed than a non-migrant. This result can be an indication of the obvious fact that there is unmet

demand for domestic and casual labor in the city, a pull factor for the rural poor and marginalized youth,

particularly women. Thus, given the existing push and pull factors from rural areas and the unmet labor

demand in urban centers; the migrants' supply of labor would be mutually beneficial to both the urban as

well as the rural communities.

Dejene et al., (2016), conducted the binary logistic regression to assess the determinants of youth

unemployment at Ambo, Ethiopia. Their result showed that among the demographic variables, age of the

respondents and migration status were significantly related to youth unemployment whereas marital status

of the respondents was not significant. From the human capital variables included in the model, education

and health status of the respondents were significantly related to youth unemployment, whereas

participation in employment related trainings was not statistically significant. Among the economic

determinants, household income, access to credit and saving services and work experience were

6

significant. Access to job information and psycho-social factors were the two social capital variables that

were significantly related to youth unemployment. As youths are more vulnerable to unemployment,

efforts should be made by the government to provide credit and training so as to facilitate their entry into

business and entrepreneurship. Migrants are the victims of unemployment in town. Therefore, the pushing

factors of migrants should be identified to arrest the continuous drift of youth towards urban areas as this

may worsen the unemployment situation in urban areas.

5. Research Methodology

5. 1 Type and Sources of Data

To attain the stated objectives, data has been collected both from primary and secondary sources. A cross

sectional primary data was collected from selected respondents in the study area through structured

questionnaire. The questionnaire was designed in such a way that it could help the investigators to dig out

detailed information on respondents demographic, social and economic characteristics. Moreover,

secondary data was also gathered from published and unpublished documents obtained from necessary

institutions so as to generate additional information on the characteristics of unemployment in the study

area.

5. 2 Research population and techniques of sample size determination

This study is conducted at individual level, the sampling frame or the total population from which the

required number of sample drawn is the total number of active labour force of youth found at three towns

of East Gojjam Zone: Debre Markos, Motta and Bichena towns. These three towns are selected

purposefully due to greatest numbers of urban dwellers are stayed there.

According to Amhara region bureau of finance and economic development (2015) statistical report and

the Ethiopian labor force survey (2013), the total active urban youth in these towns is found 56,028.

In order to determine a representative sample size from the selected towns, the study used a sample size

determination formula given by Yamane (1967, 886) as cited in Glein D. (2013). The relation is given as

below:

-------------------------------------------------------------------- [1]

7

Where nis the representative sample size, N is the total population, and e is the desired level of precision.

For a 95% confidence level, the researchers have selected the representative sample of:

Moreover, the researchers applied the proportional probabilistic sampling technique and have selected

270 youths from Debre Markos town, 77 youths from Motta town and 50 youths from Bichena town.

5. 3 Method of Data Analysis

Descriptive and econometric analysis have been employed to meet the main objective of the study. In the

case of descriptive analysis graphs, tables, t-tests and chi-square tests have been employed. While the

econometric analysis has applied the binary logit model to identify the major determinants of urban youth

unemployment

Econometrics model Specification

Following Gujarati (2004) the logistic model could be written in terms of the odds ratio and log of odds

ratio, which enable one to understand the interpretation of the coefficients. In this study, the odds ratio is

the ratio of the probability that the youth will be unemployed (Pi) to the probability that he/she will be

employed (1-Pi).

Since, the above formula can be rewrite as shown below for easily understanding.

( --------------------------------------------------------------------------------- [2]

------------------------------------------------------------------------- [3]

Therefore,

------------------------------------------------------------- [4]

Taking the natural logarithm of equation (4)

8

Where: K=the number of explanatory variables; Xi= vector of independent demographic and socio-

economic variables of individuals, = the error term, = is the value of the log odd ratio when

X or explanatory variable is zero, and measures the change in L (logit) for a unit change in

explanatory variables (X).

The dependent variable is dichotomes or dummy variable: where it represents (1) when the urban youth is

unemployed and (0) when the urban youth is employed.

Based on the theoretical background and empirical results of different studies on urban youth

unemployment carried out in different countries including Ethiopia, the following variables are

hypothesized to influence youth unemployment status of urban dwellers in the study area.

Table-1: Summary of independent variables that may influence on the dependent variable

Variables Description Values/Categories References in the

model

Expected

sign

SEX Sex of the respondent 0 = female and 1= male Female -

AGE Age of the respondent Continuous variable ?

MARS Marital status of the respondent 0 = married and 1= otherwise Married ?

EDUC Educational category ranging from

illiterate to higher education.

0 = illiterate

1= primary education

2= secondary education

3= Higher education

Illiterate -

WOREXP Work experience of the respondents 0= no work experience

1= has work experience

has no experience -

SKIMAT

The relationship between the skill the

respondents have and the market need

0= skill mismatch

1= there is skill match

skill mismatch -

SOCIANET Availability of social network for the

respondent

0 = no any network

1 = 0 < net ≤3 persons networks

2 = net > 3 persons networks

no any network -

ACECRED Access of credit for the respondents 0 = no any access for credit

1 = has access for credit

no access -

MIGRN Migration status of the respondent 0= non-migrants

1= migrants

Non-migrants +

FAMPROS Prosperity level of their Family 0=poor

1=Medium

2=Rich

Poor -

RIGN Region of Residence 0=Debre Markos Town

1=Bichena Town

2=Motta Town

Debre Markos

Town

?

9

6. Results and Discussion

6. 1 Unemployment Status of Respondents

Based on the collected data individuals are categorized into employed and unemployed. Accordingly, it is

found that 48.11% out of the total respondents were unemployed where as 51.89% of them were

employed.

Table .1: Unemployment Status of Respondents

Category Frequency Percent

Unemployed 191 48.11

Employed 206 51.89

Total 397 100.00

Source: Field Survey, 2016

Unemployed respondents were asked to identify the major reasons for being unemployed and they

responded that shortage of vacancies and the existence of much completion are the major reasons which

hinder them to be employed in government offices, private organizations and NGOs. Moreover, they

replied that shortage of initial working capital and lack of profitable business ideas were the major

constraint that hinders them from doing their own business.

In order to examine the level of competition on vacancies at government offices, secondary data was

collected from Bichena and Motta towns for three consecutive years (2013/14, 2014/15 and 2015/16).

Figure 4.1: Average Number of registered candidates for a single Vacancy

Source: Own computation based on Bichena and Motta Town Civil Service Offices, 2016

10

As indicated in the figure 4.1 above the average number of registered individuals for a single vacancy has

increased over time. For instance, on average, there were 167, 204 and 447 registered job seekers for a

vacancy of cleaner and guard during 2013/14, 2014/5 and 2015/16 respectively. The number of registered

persons for a single vacancy decreases as educational level increases. But this does not mean that

unemployment is lower for higher educated individuals because most of lower educated individuals

participate in casual employment.

6. 2 Demographic Characteristics of Sample Respondents

From the sample respondents 54.7 % of them were males while the rest 45.3% of them were females.

From the total respondents 46.5% of females and 53.9% males were unemployed. The unemployment rate

for males and females respectively was 47% and 49.5% . However, there is no statistically significant

difference on unemployment status based on sex.

Table 4.2: Respondents Unemployment status by sex and Marital status

N

o.

variables Category Unemployed Employed Total X2

Number Percent Number Percent Number Percent

1 Sex Female 89 46.6 91 44.2 180 45.3 0.234

(0.628) Male 102 53.4 115 55.8 217 54.7

2 Marital

Status

Married 33 17.3 81 39.3 114 28.7 23.525***

(0.000) Non-

married

158 82.7 123 59.7 283 71.3

Source: STATA output based Field Survey, 2016

28.7% of the respondents were married but 71.% were non-married( single, widowed or divorced). Out

of the total unemployed respondents 82.7% of them are non-married while only 17.3% of them are

married. From the chi-square test it is found that married individuals are less unemployed than non-

married having unemployment rate of 28.9% and 55.8 respectively at 1% significance level.

The other important demographic variable is age. Since this study considers only youth individuals (15-29

years), the average age of the respondents were 24.04. The mean age of unemployed youths were 22.83

while for employed respondents it was 25.16. Table 4.3 shows that as age increases the probability of

getting job (being employed) increases significantly at 1% level of significance.

11

Table 4.3: Respondents Unemployment status by Age

Group Obs Mean Std.Err. T

Unemployed 191 22.83 0.2141 7.5928***

Employed 206 25.16 0.2132

Total 397 24.04 0.1616

Source: STATA output based Field Survey, 2016

6. 3 Socio-Economic Characteristics of Sample Respondents

As shown from table 4.4 below, 64.2 % of them attend higher education and 24% of the respondents

have completed their secondary education.

Table 4.4: Socio-economic characteristics of respondents

N

o

variables Category Unemployed Employed Total X2

No. % No. % No. %

1 Education

Level

Illiterate 4 3.1 7 6.8 11 2.8 24.01

(0.000)*** Primary 15 8.4 19 18.5 34 8.6

Secondary 44 22.5 53 32.5 97 24.4

Higher educ. 128 66.0 87 42.2 255 64.2

2 Work

Experience

No Experience 187 97.9 58 28.2 245 61.7 204.06***

(0.000) Has Experience 4 2.1 148 71.8 152 38.3

3 Skill Match Mismatch 98 51.3 53 25.7 151 38.0 27.527***

(0.000

Match 93 48.7 153 74.3 246 62.0

4 Social

Network

no any network 129 67.5 49 23.8 178 44.8 76.738***

(0.000) 1-5 persons networks 38 19.9 94 45.6 132 33.3

Above 5 persons networks 24 12.6 63 30.6 87 21.9

5 Access to

Credit

No access to credit 157 82.2 133 73 290 73.1 15.656***

(0.000) Has Access to credit 34 17.8 64.6 35.4 107 26.9

6 Migration Migrant 79 41.4 69 33.5 148 37.3 2.622

(0.105) Non-Migrant 112 58.6 137 66.5 249 62.7

7 Family

prosperity

Poor 106 55.5 83 40.3 189 47.6 9.506***

(0.009)

Medium 73 38.2 102 49.5 175 44.1

Rich 12 6.3 21 10.2 33 8.3

8 Region of

Residence

Debre markos 141 73.8 128 62.1 269 67.8 8.204**

(0.017)

Bichena 16 8.4 35 17 51 12.8

Motta 34 17.8 43 20.9 77 19.

Source: STATA output based Field Survey, 2016

12

The Chi-square test found that there is a statistically significant difference in terms of their

unemployment status among different categories of educational level. The more they educated, the more

they becomes unemployed. More specifically, unemployment rate is high for those whose education

level lies at higher education ladder. One basic reason behind this result may be the selection of job by

more educated individuals.

Another important variable is the match between the skill( education) they acquired and the demand by

the market. From the total respondents only 38% of them found that the skill they have is directly match

with what is demanded by the market, while 62 % of them believes the existence of skill mismatch which

make them unemployed. From the sample respondents the unemployment rate for those whose skills are

matched was 37.8% while for those who found it mismatch was that 64.9% and the difference is

statistically significant at 1%.

It is known that information plays a major role in any activity. Hence the respondents level of networking

with their employment status were examined. Based on the survey 44.8% of them replied that they have

no any network while 33.3% and 21.9% of have 1 to 5 and more than 5 number of networks which can

help for job searching respectively. It was found that there is a statistically significant difference among

categories networks; individuals who have less networks were more unemployed.

Family prosperity level is measured on relative base as rich, medium and poor based on the living

condition of the society at the study area. From the total respondents 47.6 % of them replied that their

families are poor. Whereas 44.1% and 8.3% of them were found from medium and rich families

respectively. Statistically there is a significant difference among the three groups individuals interms of

their unemployment status. the unemployment rate of individuals from poor families was 56.9% while

for individuals from medium families was 41.7%.

It is believed that those who have better access to credit may participate in small businesses. However, the

collected data revealed that only 26.9 % of them can have access to credit while 73.1 of them do not.

Moreover, the chi-square test also indicated the existence of statistically significant difference between

these two groups on their unemployment status. Respondents were also asked to identify the reasons for

the inaccessibility of credit and they responded that the major obstacle is lack of collateral to get credit

(71.2%) followed by the requirement of saving which accounts 26.1%.

13

4.4 Determinants of Unemployment

In this section attempts have been made in explaining the main demographic and socio-economic

determinants of urban youth unemployment. As mentioned earlier, Logit model was selected to identify

the determinants of Unemployment in the study area.

However, before fitting the Logit model, it was important to check whether serious problem of and

association exists between explanatory variables. Hence, pair wise correlation matrix was conducted so

as to test whether there exist or not the problem of multicollinearity. Accordingly, it is found that there is

no serious multicollinearity problem among the explanatory variables (see appendix-1). To avoid the effect

of heteroscedasticity robust logistic regression was employed for it compromises the effect of

heteroscedasticity even if it exists initially.

.

The estimated logit model is presented below in table 4.5, in which the dependent variable being

unemployed status regressed on different demographic and socioeconomic variables which are expected

to affect unemployment in the study area.

Various goodness-of-fit measures validate that the model fits the data well. The log likelihood ratio test

robustly rejects the hypothesis that all slope coefficients are simultaneously equal to zero and thus, the

model correctly predicted the observations (see table 4.5 below). Besides, the sensitivity, the number of

unemployed youths correctly predicted by the binary logit model is 91.1 percent and specificity, the

number of employed youths correctly predicted is 81.1 percent of the observations (see appendix-2).

Thus, the binary logit model under consideration fits the data very well and fairly.

In the model a total of 11 variables that may affect urban youth unemployment were considered.. Among

them 7 of the variables were found significant variables which affects urban youth unemployment.

However, it found that 4 of them (Sex, Marital status, Access to credit and region of residence) are

statistically insignificant. Hence, the relationship and the magnitude of influence of significant variables is

analyzed below.

14

Table 4.5: Logistic Regression Result

Number of obs = 397

Wald chi2(16) = 122.04

Prob > chi2 = 0.0000

Log pseudolikelihood =-122.88044 Pseudo R2 = 0.5530

UNEMPSTATUS Odds

Ratio

Robust

Std. Err.

Z P>z [95% Conf. Interval]

SEX

Male 0.823775 0.2720578 -0.59 0.557 0.4312142 1.573708

AGE 0.7946353 0.0461206 -3.96 0.000*** 0.7091925 0.8903722

MARS

non-married 1.412117 .5715961 0.85 0.394 .6387363 3.121905

EDUC

Primary 2.383696 2.667101 0.78 0.438 0.2659788 21.36262

Secondary 2.156389 2.346118 0.71 0.480 0.2556425 18.18952

Higher Edu. 8.078091 8.83274 1.91 0.056* 0.9475204 68.86982

WOREXP

Has Experience 0.0132952 0.0076473 -7.51 0.000*** 0.0043062 0.0410485

SKIMAT

Match 0.3687554 0.1220352 -3.01 0.003*** 0.1927717 0.7053966

SOCIANET

1 to 3 network 0.2705301 0.1145454 -3.09 0.002*** 0.1179798 0.6203314

more than 3 network 0.3290446 0.1434158 -2.55 0.011** 0.1400406 0.7731353

ACECRED

Access to credit 0.6327256 0.2620389 -1.11 0.269 0.2809935 1.424736

FAMPROS

Medium 0.5157164 0.1967133 -1.74 0.083* 0.2441919 1.089157

Rich 0.4770021 0.2916558 -1.21 0.226 0.1439017 1.581156

MIGRN

Migrant 2.239771 0.8101886 2.23 0.026** 1.1023 4.551005

RIGN

Bichena 0.9891485 0.5552689 -0.02 0.984 0.3291778 2.972298

Motta 1.710189 0.8132229 1.13 0.259 0.6734163 4.343146

_cons 475.0692 936.5522 3.13 0.002 9.970041 22636.9

Source: STATA output based Field Survey, 2016

15

Age

The result indicated that age negatively affects youth unemployment at 1 percent significance level.

Within the category (15-29), for a one year increase in age, odds ratio towards unemployment decreases

by 0.794. It shows that given the age boundary, as age increases the probability of being unemployed

reduces. According to Nganwa et.al (2015), an increase in age reduces the probability of being

unemployed because increase in age implies an increase in the years of schooling until the completion of

schooling into the job market. Youth in both these categories will seek employment or try to establish

businesses.

Education

In contrary to the researchers expectation, it is found that youths who attend higher education were more

likely to be unemployed. The odds ratio of being unemployed increases by 8.078 if the individual

attended higher education compared to those who are not educated. This may be due to job preference by

educated individuals, the existence of high completion in government sectors and slow growth of the

private sector as compared to the number of graduating individuals per year. This implies that in urban

areas, having an education certificate did not guarantee employment (Nganwa et.al., 2015). Another

justification for why unemployment rates tend to be higher among the more educated young is that there

is unavailability of resources to support full-time job search in Ethiopia like many other developing

countries unlike the situations in Latin American countries (Godfrey, 2003).

Work Experience

In line with the priori expectation of the researchers work experience affects unemployment negatively at

1 percent significance level. The result indicates that the odds ratio of being unemployed decreases by

0.013 if the individual have work experience. It means that lack of work experience increases the chance

of unemployment. This result is consistent to the results of ILO (2004) and Alemnew (2014) among

others.

Skill Match

It is believed that a match between the skill they acquired and what is demanded in the labor market have

a negative effect on unemployment. similarly, in this study it is found that skill match and unemployment

are negatively related at 1 % significance level. The odds ratio of being unemployed will decrease by

16

0.368 if the individuals skill and the demand by the market becomes match. This result is consistent with

the study of Alemnew (2014).

Social Network

Social network is important in order to get relevant information about different job opportunities. This

study found that social network affects individuals unemployment status negatively and significantly. The

odds ratio of being unemployed decreases by 0.271 ( at 1% significant level) and by 0.329 (at 5%

significant level), if individuals have 1 to 3 and more than 3 persons who helped them in finding a job

respectively compared to those who have no network. This result is similar to Asalfew (2011), Amanuel

(2016) and mesfin (2012)

Family prosperity Level (FAMPROS)

The result indicated that individuals from medium prosperous families are less unemployed that those

from poor families at 10% significant level. The odds ratio of being unemployed decreases by 0.516 if

their families are medium prosperous that those of poor. This is because families from medium income

families may have better situation for searching jobs and they can easily get initial capital to start their

own business. Similarly, youth’s occupational status in Ethiopia is significantly differs with respect to

family wealth index. With respect to this those whose family is poor are more likely unemployed

compared to those whose family are medium and above (Amanuel, 2016)

Migration Status (MIGR)

As it was hypothesized, migration status of individuals affects their unemployment status positively at 5%

significant level. The Logit model predicts that if individuals are migrants their unemployment status

increases by the odds ratio of 2.239 compared to non-migrants. This result is similar to Dejene et al.,

(2016) and Asalfew (2011)

7. CONCLUSION

The main objective of this study is to identify and examine the factors which determine urban youth

unemployment in East Gojjam zone of Amhara Region. To achieve its objective, the study has employed

Binary logit regression model. In the model unemployment status of urban youths were taken as

dependent variable and 11 explanatory variables were included. Based on the result of the logit model,

17

seven of the explanatory variables were found significant determinants of urban youth unemployment; of

which, age, work experience, skill match, social network, and family prosperity affects unemployment

negatively where as education and migration status of urban youths affects unemployment positively.

Within the age category of youth (15-29), as age increases by one year the odds ratio of being

unemployment decreases by 0.79 at 1% level of significance. The odds ratio of being unemployed

decreases by 0.013 for experienced urban youths at 1% significance level. Those urban youths who have

more social networks for the purpose of job searching are less unemployed. Moreover, it is found that

urban youths from medium prosperous families are less employed compared to poor family individuals.

The finding of the study indicated that urban youths who attend higher education are more unemployed

compared to illiterate at 10% level of significance. However primary and secondary education did not

affect unemployment significantly. In addition, it is found that migrant urban youths are more likely to be

unemployed compared to non-migrants.

8. RECOMMENDATIONS

Based on the finding of the study the following recommendation are forwarded. The study found that

comparatively those graduated from higher education institutions becomes more unemployed. Hence, the

government and concerned bodies should review job market laws and regulation in order to promote

educated youth to be employed in the formal sector which can help them to contribute their role for their

country. Moreover, emphasis should be given when departments are opened; a detailed study is required

in order to make a match between the demand and supply of education since mismatch between

individuals acquired skill and knowledge with the market demand is one factor for unemployment.

It is found that youths which have no work experience were more unemployed. Thus, intervention is

required to include more jobs to newly graduated youths by hiring institutions such as private

organizations , government offices and NGOs. Furthermore, the study recommends that the concerned

bodies should try to improve the living condition and employment opportunities to rural youths in order to

reduce rural-urban migration. Because migrants are more likely to be unemployed in the urban area and

in general can be a source for urban unemployment to rise. Finally, efforts should be made to increase the

availability of initial working capital, the identification of profitable business areas and provision of

practical training for urban youths to be engaged at their own business.

18

REFERENCES

Alemnew Getnet (2014), Socio-economic & Demographic Determinants of Graduate youths’

unemployment: In the case of Debre Markos town, Amhara Regional State, University of Gonder.

Alhawarin, I., & Kreishan, F. (2010), An analysis of long-term unemployment (ltu) in Jordan's labor

market. European Journal of Social Sciences, 15(1), 56-65.

Amanuel Disassa Abshoko (2016), Determinants Of Youth Unemployment; Evidence From Ethiopia,

Global Journal of Human-Social Science: A Arts & Humanities - Psychology Volume 16,ISSN: 2249-

460x & Print ISSN: 0975-587X

Asalfew A. (2011), Demographic and socio-economic determinants of youth unemployment in Debere

Birhan town, Master thesis, Addis Ababa University, Addis Ababa.

Berhanu, D., Abraham T. and Hannah D, (2005), Characteristics and Determinants of Youth

Unemployment, Underemployment and Inadequate Employment in Ethiopia.

BoFED (2015), Amhara Region Bureau of Finance and Economic Development statistical report, Bahir

Dar.

Dejene T.,J.PaulMansingh and Warkaw L., Determinants of Youth Unemployment: The Case of Ambo

Town, Ethiopia. International Journal of Economics and Business Management, 2016, 2(2), 162-169.

Echebiri, R.N. (2005). Characteristics and Determinants of Urban Youth Unemployment in Umuahia,

Nigeria: Implications for Rural Development and Alternative Labour Market Variables. OkparaUniversity

of Agriculture, Umuahia, Abia State

Getinet Haile. 2003. The Incidence of Youth Unemployment in Urban Ethiopia. Paper presented at the

2nd EAF International Symposium on Contemporary Development Issues in Ethiopia, 11-13 July, Addis

Ababa, Ethiopia.

Glenn D. Israel (1992, revised 2013), Determining sample size, (IFAS), University of Florida,

Gainesville, FL 32611.

Godfrey, M. (2003), Youth employment policy in developing and transition countries: Prevention as well

as cure: World Bank, Social Protection.

Gujarati N. Damondar (2004), Basic Econometrics, 4th

Edition, The McGraw-Hill Companies.

Guarcello, L. and Rosati, F. 2007. Child Labor and Youth Employment: Ethiopia Country Study. SP

Discussion Paper No. 0704, the World Bank.

ILO (1992). Surveys of Economically Active Population. Employment, Unemployment and

Underemployment: ILO manual on Concepts and Methodology. Geneva.

ILO (2004), Improving Prospects of Young Women and Men of Work.A guide To Youth Employment,

Geneva

19

ILO (2010). Global Employment Trends for Youth. Special Issue on the Impact of the Global Economic

Crisis on Youth, Geneva.

Labor Force Survey (2013), Statistical Report on the 2013 National Labour Force Survey, Central

Statistical Agency, Addis Ababa, Ethiopia.

Mesfin Mulu (2012), Determinants Of Women Unemployment In Ethiopia: A Multilevel Model

Approach; The Department of Statistics, Addis Ababa University.

Ministry of Youth, Sports and Culture (2004), National Youth Policy, Addis Ababa, Ethiopia.

Muhammad ShahidMaqbool, TahirMahmood Abdul Sattar And M. N. Bhalli (2013) ,Determinants Of

Unemployment Empirical Evidences From Pakistan, Pakistan Economic and Social Review, Volume 51,

No. 2 (Winter 2013), pp. 191-207

Nganwa, P., Assefa, D., and Mbaka, P.(2015). The Nature and Determinants of Urban Youth

Unemployment in Ethiopia. Public Policy and Administration Research. Vol.5, No.3, 2015. ISSN 2224-

5731(Paper) ISSN 2225-0972(Online).

Nzinga H. Broussara and TsegayGebrekidanTekleselassie (2012), Youth Unemployment: Ethiopia

Country Study, International Growth Center

Shumet, G. (2011). A Glimpse of Urban Youth Unemployment in Ethiopia. Ethiopian Journal of

Development Research, vol.33, No.2.

Tegegn Gebeyaw (2011), Socio-Demographic Determinants of urban Unemployment: The Case of Addis

Ababa Ethiopian Journal of Development Research Vol.33, No.2, October 2011

Toit, R. (2003).Unemployed Youth in South Africa: The Distressed Generation, Paper presented at the

MinnesotaInternational Counseling Institute, Minnesota.

WB. (2007), Urban labor markets in Ethiopia: Challenges and prospects: The World Bank.

20

Appendix 1: Correlation Matrix

RIGN 0.0480 0.3002 0.2115 1.0000

MIGR -0.0600 0.0397 1.0000

FAMPROS 0.2413 1.0000

ACECRED 1.0000

ACECRED FAMPROS MIGR RIGN

RIGN -0.0003 -0.0454 -0.2101 -0.1060 0.1071 0.0778 0.1689

MIGR -0.0534 -0.0195 -0.0173 -0.1754 -0.0036 -0.0031 0.0071

FAMPROS -0.0933 -0.0575 -0.0945 0.1043 0.1037 0.0462 0.2436

ACECRED -0.0398 0.0559 -0.0662 0.1508 0.2106 0.0666 0.2572

SOCIANET 0.0499 0.1525 -0.1287 0.1325 0.3888 0.1812 1.0000

SKIMAT -0.0465 -0.0386 0.0188 -0.0901 0.2435 1.0000

WOREXP 0.0095 0.3268 -0.2675 0.1001 1.0000

EDUC 0.0015 0.1267 -0.0280 1.0000

MARS 0.0594 -0.2482 1.0000

AGE 0.1074 1.0000

SEX 1.0000

SEX AGE MARS EDUC WOREXP SKIMAT SOCIANET

. pwcorr SEX AGE MARS EDUC WOREXP SKIMAT SOCIANET ACECRED FAMPROS MIGR RIGN

Appendix 2: Sensitivity and Specificity Test

Correctly classified 85.89%

False - rate for classified - Pr( D| -) 9.24%

False + rate for classified + Pr(~D| +) 18.31%

False - rate for true D Pr( -| D) 8.90%

False + rate for true ~D Pr( +|~D) 18.93%

Negative predictive value Pr(~D| -) 90.76%

Positive predictive value Pr( D| +) 81.69%

Specificity Pr( -|~D) 81.07%

Sensitivity Pr( +| D) 91.10%

True D defined as UNEMPSTATUS != 0

Classified + if predicted Pr(D) >= .5

Total 191 206 397

- 17 167 184

+ 174 39 213

Classified D ~D Total

True

Logistic model for UNEMPSTATUS

. estat classification, all