The impact of education household head on household income the case of rural an giang province luận văn thạc sĩ

Recognizing the benefits of education and knowing the important role of the household head in household, this research tries to explore the relationship between household head's educatio

VIETNAM- NETHERLANDS PROJECT FOR M.A IN

NGUYEN DUC THANG

ACADEMIC SUPERVISOR: MA NGUYEN TID SONG AN

HO CHI MINH CITY MAY, 2002

ABSTRACT

People can be died by poverty which suffers from extremely low income to cover the minimum needs in their lives, but no one died by lacking of education Unfortunately, an education level is a very important factor in comprising the income

People with low education level will earn low income, therefore they easily drop in vulnerable situation of poverty, hunger, or even crimes The benefit of education has been justified for times as a stimulating factor in income comprising through its effects on labour productivity, technical efficiency, allocative efficiency, managerial skills and even change one's perceptions or habits Education therefore benefits anyone who attains it, no matter the job they work and the sector they engage On the other hand, with higher education people are easier to work in many types of jobs, thus education also helps households diversify their income sources, and hence makes their income become more stable

An Giang province has a high rate of illiterate at 1 0%; and in the rural areas, 32% households are living in the poor situations caused by the low income level How to help these households overcome their current situations is the big question for everyone who concerns

Of course, there are many reasons push the households into the poor status One of main reasons is low education level This research aims to help people improve their life by stimulating to increase their income through upgrading education, especially for formal education

Recognizing the benefits of education and knowing the important role of the household head in household, this research tries to explore the relationship between household head's education level and household income, and it is found that this relationship is positive From the results of this research, some policies will be suggested to increase household income through the improvement of education factors

1.5 Organization of the research

Chapter 2 Literature Review

2.1 Concepts and some broader issues of concerned

2.1.1 Household

2.1.2 Household head

2.1.3 Household income

2.1.4 Education

2.2 The role of education

2.2.1 Education and household income

2.2.1.1 Schooling and farm income

2.2.1.2 Schooling and non-farm income

2.2.2 Other benefits of education

2.3 Determinants ofhousehold income

2.4 The relationship between education and income

2.5 Empirical evidences

Summary ofliterature review

Chapter 3 Model Specification and data collection

3.2.1 The study area

4.3 Education of household head

4.4 Relationship between household income and education

4.5 Results from econometric analysis

Chapter 5 Conclusions and suggestions

LISTS OF TABLES & FIGURES

Table No

Figure No.:

Figure 3.: Relationship between education level of household head on

household income 42

ABBREVIATION

ACR : An Giang Committee Report

AP AR : An Giang Poverty Alleviation Report

ASO : An Giang Statistical Office

CSO : Can Tho Statistical Office

HCEF : Human Capital Earnings Function

OLS :Ordinary Least Square Estimation

VND :Vietnam's Unit Currency

iv

I am also very appreciate with helpful opinions and supports of all my friends, especially Mr Tai,.Mr Dong, Mr Hong and Ms Phuong, who has given me a lot of supports in doing this thesis

I also send my thanks to people in An Giang province, local authorities and Woman Unions from three communes Thanh My Tay, Vinh Thanh Trung and Thoai Giang, who have helped

me a lot in the process of doing survey

Finally, I am indebted to my parents and relatives who have given me great encouragement and support

Carney (1998) argued that chief asset posed by the poor was their own labor Human capital refers to the labor available to the household: its education, skills and health On the other hand, Toulmin (1992) stated that human capital could be increased by investment in education and training or by getting the skills accumulated through the process of conducting their job with one or more occupations Modem theories of economic growth emphasized that the investment in human capital will get the sustainable growth and invest in public education, health is macro policies (Cornia and Jolly, 1987) The evidence about such expenditures fails to confirm the existence of a downward trend in real terms as a general proposition for low income adjusting economics (Sahn et al, 1996) This means that there have many factors affect income, however education is the most important In the broader view, Molin (1970) argued that next to the capital, the investment in education represented an important factor in economic development; a lack of knowledge and of skill will be a major obstacle to efficient production and productivity in either the agricultural or industrial sector

In the narrower view, Schultz (1964; 1975) proved that the education level affected farm productivity

by improving the quality of labor, by increasing the ability to manage and by implementing the tasks

on farm, as well as by the willingness to adopt innovations As a result, education will lead to increase productivity and hence income of households While many theories and evidences support for the benefit of education on income, many reverse results such as research ofRonnas (1992) reveal the negative relationship between them Therefore, empirical research about this relationship is necessary to be examined in order to have the exact conclusion in a specific area

Education contains all kinds of training and learning, in which the formal education is the most important one for people (Colear, 1990) The benefits of formal education have been known for a long time; Mincer (1974) stated the benefits of formal education on individual earnings; Basu, Narayan and Ravallion (2000) showed the knowledge shared within household; Lucas (1998) proved that education had benefits within village known as site effect; and as mentioned education can benefit entire the country at national level (Molin, 1970) This means that education can benefit people from individual, household, village to country level

Lockheed, et al (1980) stated that there was a consistent conclusion on the impact of education levels

on total household income in the areas where farms were modernizing In other words, education level has a relationship with household income in rural areas which are engaged in farm activities

As being stated in the chapter 2, in each household, the role of the household head is very important Household head is the decision-maker within the household about the production, the investment, children's education and many other household activities Household head is also the one who connects the household to the outside

In the South of Vietnam, An Giang is known as the largest paddy output and the highest productivity provinces through times Paddy productivity attains 5.3 tons/acre on average and output is 2.385 million

ton (ASO and CSO, 1998) However, it also is known as the highest population among Mekong Delta provinces (ACR, 1999}, and as the abundance of labor force 49.95% population (ASO, 1999) The

quality of this labor force is poor in terms of education and of working skills, unskilled labor are at 96%

(AS0,1999) Although there is high literacy at 92.8%, the general education level of population is still

at low rate with 51.04% household head under primary graduation (ACR, 1999) In addition, the number of poor households accounted for 38.841 households in 1996, the highest rate in Mekong Delta, and this number absolutely increased year by year and reached 39,210 poor households by 2000

The above general indices indicate that An Giang has high economic potential, but there have still many households which are living below the poverty lin/ How to raise their household income is the big question for all whom concern While the benefit of education has been generally demonstrated, the benefit of education of the household head- as the household decision-maker- to the household income

is not clearly explored, especially in rural areas That is the reason why I choose the topic: the impact of the education level of the household head on household income in rural An Giang

1

In rnral area, the household has income per capita per month less than 100,000 VND was defined as poor household In the urban area, the number is

I 50,000 VND (Steering Committee of Hunger Eradication and Poverty Alleviation of An Giang province, 200 1) This definition coincide with the definition of Vietnam

1.2 OBJECTIVE OF THE THESIS

This research justifies the importance of the education level by examining the impact of formal education of household head on household income in rural An Giang province by showing how the change in each year of schooling of household head corresponding with the change in household income is

Besides, throughout the research, the income sources of households in rural An Giang will also

be explored in details Based on these results, the research also tries to suggest some policies to increase household income by stimulating the education level

1.3 RESEARCH QUESTION AND RESEARCH HYPOTHESIS

To match the above objective, we will deal with the following research question:

• Is there a positive relationship between total household income and the household head's number of years of school attendance in rural An Giang Province?

And the hypothesis of the above research question is:

• There is a positive relationship between total household income and the household head's number of years of school attendance in rural An Giang Province

1.4 RESEARCH METHODOLOGY

This research studies the relationship between household income and the education level of household head, in which the household is the unit of this analysis For that purpose we need not

to explore this relationship in individual level

Both descriptive statistic and quantitative methods are used to answer the research questions The descriptive statistic method uses quartile analysis and the econometric analysis uses the OLS model to examine the relationship between household income and the education level of household head In which the education level of household head proxies by number of years of schooling attendance

In the OLS model, the dependent variable is the total household income in 2000 In which, farm and off-farm engaged in agricultural sector or farm sector Non-farm income was collected

on-by summing all income from outside agricultural sector All independent variables will be represented in Chapter 3 Primary data was collected from the survey by direct interviews with household heads under the help of local authorities Households were randomly selected from

3

available list also given by local authorities Interviews were based on the questionnaire in the appendix

With the OLS model, the results will be interpreted after the normality, hetterocedasticity, and multicollinearily tests Besides the primary data, secondary and tertiary data used in this study are collected from An Giang Statistical Office, and An Giang Report This methodology will be deeply presented in the Chapter 3

1.5 ORGANIZATION OF THE RESEARCH:

This thesis concludes five Chapters as follow:

Following the introduction Chapter, chapter two contains related concepts and some issues concerned The role of formal education level will be highlighted as the important element in household income All sources of household income will be discussed to clearly understand the structure of household income Following this, we examine the determinants of household income and then the relationship between education and household income The final part in this Chapter is about empirical evidences related to the research

In the Chapter three, the first section specifies the OLS model used for the analysis; all variables will be discussed to have expected signs The second section discusses data collection by reviewing the study area, calculating the sample size, constructing the sampling frame and preparing the questionnaire for the survey Finally, we will discuss limitations of the data

Chapter four opens with descriptive section, in which we describe the demography characteristics of the surveyed areas and of the households followed by household head The household income structure will be explored to understand their income diversification We also discuss the education of household head and finally we show the relationship between household income and the education level of household head by quartile analysis The second section is econometrics analysis which quantitatively supports to the first section in answering the research question

Chapter 5 summarizes all main and sub-findings from the research Based on these results, we suggest some policies to raise the household income through education elements The research will be ended with the further study suggestions

CHAPTER2

LITERATURE REVIEW

This chapter provides an introduction of the theoretical review for the study Section 2.1 first reviews and also discusses some relevant concepts Section 2.2 states the role and the benefit of education Section 2.3 explores the determinants of household income, in which education level

is the one of determinants From the Section 2.2 and 2.3, we can know that the education is one

of important factors that affects the household income, the Section 2.4 is about the relationship between education and household income in various aspects and theories Section 2.5 shows some empirical evidences that relevant to this research The final section is the summary of literature review

2.1 CONCEPTS AND SOME BROADER ISSUES OF CONCERN:

According to Ellis (1993), based on characteristics and income generating activitiei of

household, we can distinguish two kinds of household: farm household and non-farm household Farm household derives their livelihood predominantly from the agricultural production, and farm household can engage to other non-farm activities

VLSS (1998) defined that farm household was the household that had all or almost laborers working in agriculture and that got their income mainly from the agricultural production Therefore, farm household can be understood as a household derived its livelihood 3 mainly from agricultural production, or from their majorities of household labor force engaged in agricultural production Otherwise, it is the non-farm household

In Vietnam, farm household has major labor force engaging m agricultural activities or agricultural production determinant in their livelihood (VLSS 98) However, it is necessary to

5

explain why I choose household not individual for my study as being presented in the following part

In developed countries, most workers are wage earners; hence the return to education can be measured by examining the impacts of individual's school attainment on that individual's wage This means we can evaluate the effects of schooling at the individual level However, in the predominant feature of many developing countries, the largest share of household income is generated from self-employed activities, so the household income is therefore suitable to be measured at the household level (Jolliffe, 1998)

On the other hand, as stated many households and individuals are engaged in one or more than sources of income generating activity, hence the estimation of return to education by focusing

on either strictly farm income, off-farm income or non-farm income including wage income will provide an incomplete picture of the importance of education For this reason, this research will choose total household income to examine for highlight the role of education Of course, the effects of education on each type of household income are different, however this will not be covered in the scope of this research

2.1.2 Household Head

different across cultures and across surveys In Ghana, for instance, the household head classified does not base on educated attainment, management skills or other characteristics but

on age (Jolliffe, 1998: p5) However, Jolliffe (1998) supposed that the household head has been the person who acted as a household manager, making all of decisions in the household

In Vietnam, local authorities calls household head with a name "chu ho " 4 it means that household head is the person who presents for all household members in the resident book managed by local government, and in almost Vietnamese households they also are the household decision makers and represent for all members in the household when contact with the local government

"Most household surveys conducted in developing countries define one person as the head of the household in case of Vietnam, defining one person as the head of the household is not unreasonable Administrative procedures require the registration of one person as the head of the household, but definition of one person as the head goes deeper than current administrative requirements It can be traced to the influence of Confucian philosophy, which not only specifies a hierarchical structure for

4

'chu ho ' is the household head defined according to local government

different types of inter-personal relationships, but also clearly defines the position of individuals within each relationship "Desai (1995: pJO)

Household head is the concept and is not fixed for all studies; therefore, this is not covered in theories but in the empirical contexts This part tries to describe the role of household head by discussing whose education in the household has the most important effect to the household base on empirical studies When examining the education level of household, many indicators are possible to use for proxy For example, Jamison and Lau's (1982) surveyed in over 35 studies from Asia, Africa, and Latin America, they argued that only the education of household head affects to farm household income, while the education level of other household members had no effect Similarly, Fane (1975), Wu (1977) and Jamison (1984) also used education of household head to represent for household school attainment

By arguing that the spouse of household head is also a important person who influences indirectly on household making-decision, or sometimes this spouse is the decision-maker, Huffman (1974) used both education of household head and of the spouse to proxy for the entire

household; while Lin (1991) used both household head and average household education 5

Household head's education level and occupation can affect both farm and non-farm incomes D.T Nguyen (1999) asserted that the household head contributed to farm activities mainly by gathering information, planning, production decision making and these are more important than the time they spent on the farm Fafchamps (1998) justified that the household head had more decision-making power than other household members did

When analyzing the Viet Nam economy in transition stage namely 'Doi moi 6 ', Moock (1998) divided household head's education level into many subgroup~ and examined them; he found that 13,4% of the population in which household heads with university degree were poor in 1992-1993, but by the year 1997-1998 this rate downed to 4.5% The number of household with heads who attained technical school also downed from 47.7% to 19.2% In contrast, majorities

of households with the non-educated head were still living in poverty because of this rate was just downed from 69.9% to 57.3% Those evidences showed that education level of household head play an important role in poverty reduction in household

Basu and Foster (1998) argued that only one person expected to be educated in the household

for the entire household to benefit from the cognitive skill/ acquired in school, and he suggested

that it was better to use education level of the most important household member to proxy for the entire household education

From the above discussion, the household education level can be affected by all household members with the various impacts, however the household head education level is the most important and can not be separated from effecting elements

2.1.3 Household lncome 8

Income is the earnings come from labor work after the period of time or from the transfer These sources of income are the returns from investment, farm production, or from service supply Income can be in-cash or in-kind In-cash income is measured by money comprised from sold livestock and crop, from wage, rents, or from remittances In-kind income is the physical products of the return referring to the consumption of own-farm product, payment in-kind, and transfers or the exchanges of consumption items (Ellis, 2000)

Household income contains all income from household members There are many ways to divide the household income into each sub-categories of income source or activities that generates them (Ellis, 2000) However, the basic classification divided it into three kinds of income (Saith, 1992; Leones and Feldman, 1998; Ellis, 2000):

• On-farm income includes crops, livestock income and comprises both consumption kind of own-farm output as well as cash income earned from output sold which is a net amount that excluded all costs of production

in-• Off- farm income refers to wage or exchange labor on other farms

On-farm and off-farm income refers to the income generating activities from the agricultural sector On-farm and off-farm can be combined to have farm income

• Non-farm income relates to non-agricultural resources, these are:

(1) Non-farm rural wage or salary employment;

(2) Non-farm rural self-employment, sometimes called business income;

(3) Rental income obtain from land or property;

(4) Urban to rural remittances arising from within national boundaries

(5) Other urban transfers to rural households, for example pension payment to retirees;

(6) International remittances arising from cross-border and oversea migration;

8

also calls total household income

2.1.4 Education

Education, as the definition of Gillis et a! (1996) can broadly be defined as all forms of human

learning or more narrowly as the process that occurs in specialized institution called "school" Education is classified into three types: formal education takes place in schools and usually involves young people who have not yet started their working lives; informal education takes place outside any institutional framework or organizes program at home, on the job or in the community; non-formal education takes place outside schools whose attendants are often adults Its programs are usually shorter and more narrowly focused and more concerned with an applied knowledge than the programs of formal education are

Cotlear (1990) also describes education as three types: formal education is closely understand as the term schooling; Non-formal education included all knowledge getting from various extensions, apprenticeships and literacy training; Informal education refers to wide range of experience, 'learning by doing' or other activities which create new ideas and facilitate learning

"In VietNam, formal education begin with grade 1 and six-year olds should be in grade 1, seven-year olds in grade 2, etc Primmy schooling stretches from grade 1 to grade 5, so that a child who is on target should complete primary schooling by age 10 Similarly lower secondary schooling grade are 6 to 9 and if children are on target they should complete lower secondary schooling by age 14 After this point, individuals can switch to vocational training or go on to upper secondary schooling which consists of grades 10 through 12 If they choose the latter track they should complete general schooling by age 17 Those who move to the vocational track after lower secondary schooling spend three years obtaining vocational training Beyond upper secondary schooling there is technical training, university, etc " (Desai, 1995)

In rural An Giang, the definition of Cotlear is more appropriate in comparison with Gillis's definition because people can work after the time in schooling Therefore, throughout this research we follow the definition of Cotlear Moreover, in An Giang, with very weak non-formal educational system, there is almost no program such as agricultural extension which is opened for people to upgrade their knowledge All people who have to leave school seem to be finished their education Therefore, non-formal education is not needed to examine at this area

In addition, when examining the case of Vietnam, Desai (1995) stated "ordinarily schooling attainment should imply what individuals gain from formal schooling " This means that the

formal education is very important to reflect the schooling attainment This research will strictly focus onformal education 9

9

From here, the term 'education' can be understood as 'formal education'

9

Two types of education will be examined in this research are formal and informal education However, formal education is strictly focused, informal education will be examined as other determinants

As defined, formal education is reflected by schooling attainment or education level This education level can be measured by number of years of school attendance Informal education is accumulated from the time that people have done that type of work, and can be measured by years of experience

2.2 THE ROLE OF EDUCATION

2.2.1 Education and Household Income

To point out the benefits of education in term of earnings, several aspects of schooling benefits will be mentioned In which, education helps farmers conduct their own farm more efficiently and adapt to technical change, education thus affects productivity As the result, it leads to raise the farm household income included on-farm and off-farm (as stated) For non-farm workers, schooling benefits them directly through higher income

2.2.1.1 Schooling and farm income

For people who live in rural areas, one main source of their income is farm income (Ellis, 2000)

To justify that education has an impmiant influence on farm income, we should firstly focus on the relationship between farm production and education Moreover, efficiency and productivity are the keys to determine the output besides other factors in farm production

According to Weir (1999), the skills and the knowledge required to adapt to new technology are scarce in developing countries; a learning process is therefore necessary This learning process

is formed through farmer's experience and education, which includes three factors: the experience from the production, the knowledge about capital and equipment, and know-how

regarding to experiment action Weir (1999) concluded thatformal schooling might reduce the

Education is always necessary for farmer because education enhances farm productivity by improving quality of labor, by increasing ability to adjust disequillibria and by the willingness to

adopt innovation (Schultz, 1964; 1975) Croppenstedt et al (1998) conducted a fertilizer

marketing survey and found that literate farmers were more likely to adopt use of fertilizer than illiterate ones did

Psacharopoulos and Woodhall (1985) stated four stages of agricultural technology adoption

The first stage, traditional farming, little or no schooling is necessary because farmers can

obtain information from their parents The second stage, single input adoption, basic literacy

and numeracy are needed to understand the constructions The third stage, adoption of multiple inputs simultaneously, higher numeracy and literacy is required because it requires some basic

science knowledge The final stage, irrigation based farming, more education is required due to

some complex calculations happened

On the other hand, education help people increase the willingness to adopt technology, people with higher education have intention to need more information from agricultural extension to conduct trial test on their farms, for instance new inputs, crops or methods (Weir, 1999)

• Education and Productivity

As mentioned, in rural area, farm-income is a main source of household income To get high farm income, beside external factors such as: high output price and low input prices, the most important factor that strongly affects profit is productivity Higher productivity household will bring higher farm income To examine the effect of education on household income, we can examine the effect of education on farm income through agricultural productivity

Chraudhri (1979) stated four effects of schooling: worker 10 , allocative, innovative and external effects These effects help to raise productivity for farmers Similarly, Welch (1970) said that holding other inputs constant, the effects of schooling would increase the farm output The expected benefits of education come from the evidences that showed higher education leads to higher productivity in agricultural areas and higher wages in non-agricultural areas (Weir, 1999) Many economists believe that farmer education is more important for modem agricultural practices than for traditional ones "Increasing literacy and numeracy may help farmers to acquire and understand information and to calculate appropriate input quantities in a modernizing or rapidly changing environment" (Weir, 1999: p4) and schooling helps people

learn on-the-job more efficiently (Rosenzweig, 1995)

Mellor (1976) asserted that farmer education in rural areas should be a central ingredient of any strategy to improve agricultural productivity, and this productivity will again raise total household income through higher farm income Farm efficiency depends on the education level

of household head, which makes almost the farm decisions To make good decisions in using

10

'Worker effect' of schooling refers to the increase in farm output that is owing directly to education, holding other inputs constant (Chaudhri,

1979 and Welch, 1970)

11

new seeds in modernizing environment farmers are necessary to know how to read the instructions for the use of chemical pesticides, fertilizers, etc

It can be shown the benefits of education on farm income through farm production by the improvement of technical efficiency, allocative efficiency and productivity by the following figure

OutputY

Figure I: Effects of education on farm production

Source: Hussain and Byerlee (1995)

I

Hz

Y=f(Xz)

Input X

In Figure 1, suppose that farmer without knowledge of technical and allocative efficiency started

at point B The movement from B to T represents the achievement of technical efficiency The movement from T to A represents the allocative efficient and the shift of production frontier corresponding with the shift from point A to I represents the applying of new technology to get higher increase in output and profit

In brief, by this figure, we can see that when two farmers uses similar amounts of inputs, the output can be different base on their efficiency when using these inputs In other explanation, with the same inputs, depending on efficiency, one can get different output by enhancing their knowledge of effective production frontier (Jamison and Lau, 1982)

From above discussions, we can conclude that the education level affects strongly to technical efficiency and productivity The raises to technical efficiency and productivity surely help to

increasing the farm output and hence the farm income Finally, we can conclude that the higher formal education will lead to higher farm income

2.2.1.2 Schooling and non-farm income

Productivity of laborers is different from one another because they have different education, health and working skills Education affects labor power and hence wage On the other hand, education also helps labors find other opportunities to get additional income Laborers who have knowledge can use modem machines and learn new technology better than the less educated people, and they therefore get higher salaries or wages from employers (Schultz, 1993)

Education is an important determinant of earnings in the market economy The higher one's education is the higher starting salary and the steeper the rise in earnings during the working life (Moock et al, 1998)

Evans (1991) concluded that in self-employment, people who had higher education levels were better to be able to secure employment elsewhere, or succeed in business Knight and Sabot (1987) also showed that the number of years of schooling was an important element of cognitive skills and those cognitive skills were again an important determinant of wages

Moreover, educated farmers are able to interact more effectively with credit agencies because they can understand financial transactions and can keep records, and hence increase the likelihood of obtaining credit Consequently, the education level of household head brings more earnings in both farm and non-fmm income (Mellor,1976)

To understand clearly the benefit of education on income in the non-farm sector, we study the two following models

• The two-period models

Willis (1986) proposed the two-period models of the demand for formal education being presented as follows:

Assume that at the beginning of period 1, a person's stock of human capital is at level HC1, and his earning capacity is w1 per hour After investing ts hours in schooling, a person's stock of human capital increase to HC2 The increase in human capital is MIC(ts, A), with A is exogenous variable So, at the beginning of the second period the stock of human capital become:

(2.1)

13

Suppose there is no depreciation of human capital, but by time and with speed of technological changing, it will lead to reduce the values of old skills Calling the rate of deteriorates is 8, with (0<8<1) So the amount of human capital decreasing is (1-8) HC1

Therefore, the person's stock of human capital at beginning period 2 is:

HCz = (1-8) HC1 + LlliC(t5, A) (2.2)

Suppose the earnings in period 2 is w2 , from equation 2.2 we have:

Wz (HCz) = Wz (ts, HC1 8, A) (2.3)

IfHC1 8 and A are exogenous, w2 is only affected by t5, the time spent on education

So the relationship between w2 and schooling time t5 and ability A as follow:

According to Chiswick (1997) "Human Capital Earnings Function has become a fundamental

proposed the model as follow:

The model will focus only on formal education which is proxies by years of schooling, ignoring on-the-job training and other variables In one period of investment in schooling, earnings after completing the schooling is:

In which:

E0 : Earnings from no schooling attainment

E 1 : Earnings received each years after completing t years of schooling

C1 : Cost of investment in year t of schooling

R 1 : Rate of return on investments in year t of schooling

K 1 =C/T1_1 : Potential earnings after completing t years of schooling, in which T expressed the time one's consumed

to study and (t-1) is the previous point of time compare to point of timet

Suppose that there are two points of time, note by 1 and 2 We have the two periods of investment in schooling

And the two periods of investment in schooling is:

Base on this equation, we expand for t periods (attainted S years of schooling), (2 7) becomes:

2.2.2 Other benefits of education

Education helps people understand more about the world get higher utility from things they consume, as well as provide them the knowledge to absorb information relating to their production Therefore, "improved rural education would yield numerous important social, economic and human welfare benefits other than the productivity gain in agriculture" (Nguyen

and Cheng, 1999)

In addition, according to Knight and Sabot (1990), economic inequality is from an inequality among individuals, inequality in the lifetime income of individuals and inequality of

15

opportunities From these points of view, economic inequality relates to income distribution between individuals and households Therefore, to reduce this inequality gap, income must be improved for those individuals and households at the low-income level, and education is one of factors that make this gap shorter Educational expansion has a strongly effect on the distribution of the current income As mentioned, education plays a main role in the increase of income at both individual and the household level

Kooreman (1996: p199) stated the two concepts 'poverty' and 'income distribution' are closely

related, because of poverty refers to a particular state of well being which is wider than income However, to research the poverty state, it is very common to use income to measure state of well being And one more reason is that redistribution is the primary instrument for poverty relieving policies From this, we see that to reduce poverty, the compulsory solution is to increase income for people

Besides, it is easy to obtain that the higher education attainment helps reduce crimes, reduce population rate, that are also positive effects of education on society

• Formal education associated with informal education

As defined, experience is the element of informal education This section will show benefit of formal education while associate with experience The HCEF has been presented; however, to understand clearly the role of education when associate with experience years in earnings, we

refer to the revised HCEF 11 (Mincer 1974) This model shows the relationship between schooling, experience and individual earnings

where:

Y: annual earnings

ts: years of formal education

tw : years of experience

when w;=O, the starting point at level (f3 0 + f31 ts), only concluding formal education

The HCEF indicate that the tendency for earnings-experience profile to be steeper for more

educated people This also means education level consolidate with experience in creating the income (Knight and Sabot, 1990)

• External effects of education

An externality arises if those people without schooling copy the adoption behavior and productive practices of educated Hence, Nguyen (1999) argued that an increase in agricultural production requires a modernizing environment with more complex technology in which farmer education plays a crucial role and farmers with more education are more favorable than the less

educated ones Bmjas (1995) postulated and tested the external effect 12 of education In his research, human capital is passed from one generation to the next through the community This means in one household the human capital is passed not only from their parents but also through the average human capital of ethnic group in the parent's generation

To test the external effect of education, Lucas (1988) added average levels of primary and secondary schooling of other farmer at village into his equation; the results of external effect are found significant

According to Weir (1999), education is expected two effects: internal effect and external effect,

these two effects should be distinct, internal effect is return to schooling at household level, and

external effect is return to schooling at site level 13

Empirical studies of Lucas (1988) devoted that external effect was the main factor in success of raising the productivity for uneducated farmers who contacted with educated one Especially in rural areas, because the less educated can copy the innovative behavior of more educated households, and the effect of education at site level occurs when the less educated households live near higher educated households Many researchers had shown that education has external effects by adding average levels of primary and secondary schooling of other farmers in the village to their equation, and getting the positive result

However, not only the education causes the external effects, but also experience does Foster and Rosenzweig (1995) showed the external returns to experience in India case

2.3 DETERMINANTS OF HOUSEHOLD INCOME

To study the determinants of household income, Strauss (1986) supposed that household income includes farm and non-farm income This is only another way to comprise the household income In comparison with household income source classification of this research, it means on-farm, off-farm and non-farm income

13 "Site level" was considered as the village which the households are living (Weir, 1999: p37) l 1\tJCreG ~~~ li'UC I\ I ~Ill T{ I

1LI~~:!!.~iC'i' tiO 2Hi f ~!-~_1-1 I 17

t~.t.~L~-~~~~:~-~~!~ D '}~~-~-9.9

Strauss (1986) asserted that farm income is the return to household assets including land, labor and capital which are used in the agricultural activities This amount of farm income is simply different between gross farm output and expenditure on farm inputs Therefore, farm income can be understood as farm profit, and farm profit can be derived from the production function Saith (1992) in dealing with production function, defined it as "the technical relationship between inputs and output, indicating the maximum amount of output that can be produced with

The value of farm profit is thus assumed to be a function of land, labor, and other variable inputs; labor depends on household demographic (Jolliffe, 1998) Based on this argument he proposed a model of farm income as follow:

(2.12)

where:

Ar: acres of land cultivated and the value of farm equipment (fixed inputs);

Xr : household characteristics vector

Pr: vector of input and output price

In the equation (2.12), household characteristics are factors that affect farm income Education

is one of main elements of household characteristics, so we can conclude that education is one factor that affects to farm income

Besides income earned from farm income, farm household can earn mcome from off-farm income Lanjouw (1995) argued that off-farm 16 income is derived from all economic activities in rural areas other than crops, livestock, forest, hunting and aquaculture activities In other words, off-farm household income is derived from all activities of households or household members as independent producers in their household, the sub-contracting of work to farm families by urban-based firm, and off-farm wage employment in villages and rural enterprises

Similar to the explanation of farm income, Jolliffe (1998) showed the equation of off-farm income as follow:

X 0 : household characteristics vector

P 0 : off-farm prices including a vector of inputs and output in self-employment activities and wage rate in wage-paying activities

In the equation (2.13), household characteristics again include education factor

Combining the two equations farm income (2.12) and off-farm income (2.13) to be only one equation, the determinants of total household income are shown in the equation (2.14):

Labor vector includes household characteristics in which education is one of elements as above mentioned

Price vector includes vector of input and output price; off-farm prices include a vector of inputs and outputs in self-employment activities and wage rate in wage-paying activities

From the above discussion, we can summary the determinants of household income m the equation (2.15), in which capital, labor, price vectors are shown in details in table 2.5

Household income= f(Capital, Labor, Price) (2.15) Table 2.1 : Determinants of Household Income

Labor Household characteristics like: age, education, household size, religious

Price Price of inputs (seeds, fertilizers, chemical ), outputs; Price of labor hired-in (harvest,

transportation ), price oflabor hiring-out, wage & salary

Source: Own classified from discussed theories and evidences (will present in the section 2.4)

After showing that education level is a determinant of household income and indicating the positive relationship between education level and household income, it is necessary to make the short summary as presenting in the next section

16

As stated, there are many ways to divide household income Strauss (1986) divided it into two kinds:farm and off-farm However, if we combine them to get household income, this household income is now consistent with our definition that it is summed by on-farm, off-farm and non-farm income for rural households

19

2.4 THE RELATIONSHIP BETWEEN EDUCATION AND INCOME

This part will shortly summary the benefits of education level in terms of income From the previous discussions, we can conclude that education level and income has positive relationship And it can be showed as the figure 2

As stated in the two-period models, ability A is assumed exogenous variable, it depends on each

person (Willis, 1986) From the figure 2 proposed by Knight and Sabot (1990) we see this figure represents a simple structure model of the relationship between earnings, years of education, natural ability ( exogeneous element) and human capital The figure is consists of four elements

as mentioned above

Simplicity, we can see the figure contains six links between elements in which the link B is the most important for representing the influence of years of education on earnings, and this link cause directly effect

Years of education have directly effect on earnmgs through link F and link A, where F represents the human capital acquired in school The difference value of human capital leads to the result of different earnings So, we can conclude that years of education have actual positive effect on earnings The effect of ability on earnings can be directed by link Cor/and undirected

by link E when natural ability con-elated with years of education Link D shows the effect of A

on human capital through cognitive skill (Knight and Sabot, 1990)

Figure 2: Effects of Years of Schooling on Earnings

Earnings

Source: Knight and Sabot (1990)

Human capital acquired in school

Table 2.2: Summary of the Effects of Education on Household Income

Source: Own combined from theories and evidences as stated

21

As shown in the Table 2.2, the relationship between household income and formal education is positive

2.5 EMPIRICAL EVIDENCE

• Jolliffe (1997) conducted the research to examine whose education matters in the determinant

of household income in Ghana His research based on the two Type I Tobit 17 models The first model present the farm income is the function of three vectors (as stated in previous part, the three vectors are: household characteristics, capital and price) including land cultivated (is treated as fixed input), household farm labor hours, prices for farm products and farm inputs, household school attainment variables The second model present off-farm as dependant variable functioned with business assets and years of work experience, household off-farm labor hours, household school attainment variables Combine these two models he attain the specific model for testing factors on total household income 18 • He tested the effects of education of all members in household, and the main finding of this research is that: household head education level is a main factor affect household income

• Nguyen and Cheng (1999) conducted the research in China; they used data collected in 1995 with a sample of 978 households in five Chinese provinces (Guangdong, Jilin, Jiangxi, Sichuan and Shangdong) to invest the effects of education on farm efficiency and farm income In their analytical framework, farm income is affected by some following factors such as: sown area, chemical fertilizer cost, other costs (excluding labor chemical fertilizer and seeds), labors days, total value of capital stock owned by a household valued at current In their model, some factors were included to reflect household characteristics as: numbers of schooling years of household head, age of household head, average number of schooling years of household except household head, average age of household worker except household head, number of household workers Those chosen variables can be classified into three kinds of vector as mentioned With this framework, after the OLS 19 regression, they found that all variables had expected signs, and all coefficients are significant at least 5 percent level Specially the coefficient of variable numbers

of schooling years of household head is significant at 0.1 percent level and has expected positive sign

• Farchamps and Quisumbing (1998) argued that human capital includes nutrition status and education, then it may effect household income in such the ways: better nutrition influence to

17

"In the Type I Tobit model, the zeros are typically explained by an optimization problem which results in a negative value for the desired level

of the dependent variable" (Jolliffe, 1997)

physical capital and then to labor efficiency; and better education helps improve management, technological, allocative efficiency and labor supervision skills Based on this, they constructed

a model to evaluate the effect of human capital on of household income in rural Pakistan in

1998, the data used from the survey conducted by IFPRI (Food Policy Research Institution) In

his model, household income is functioned with vector of variable inputs, farm tools, equipment, and other semi-fixed factors and household human capital characteristics In which, the household human characteristics formed by education and nutrition The result indicate that

education level help to raise household income, one additional year of schooling of adult males

leads to increase household income by 4.5 percent, very strong effect

The above evidences from other country's rural area showed that the relationship between education level and household income is positive It is therefore needed to examine the case in Vietnam, and rural An Giang is the interested place for us (As identified from Chapter 1 )

Summary of Literature Review

Throughout this Chapter, concepts related to research was stated and discussed We knew why the household head plays the important role in determinant of household income How the household income sources were classified and then combine to have total household income, all types of education were also discussed

Education was highlighted in term of household income, and the education level was shown as the one of main determinants in household income Education can affect both farm and non-farm generating activities, therefore education level can affect all kinds of household income Determinants of household income is also explored in details, combined the three capital, labor and price vectors In which, education is the main element of labor vector

Empirical evidences justified that the effect of education level especially household head on household income is positive These evidences were consistent with the discussed theories From theoretical and empirical examination, we can expect the relationship between education level and income is the positive both individual and household level From this, it is necessary to construct the suitable model for testing the effect of household head's education level on the household income, then we should move to the next Chapter

23

CHAPTER3

MODEL SPECIFICATION AND DATA COLLECTION

This chapter contains two sections The first one is about the analytical framework, in which the model is identified to answer the research question In this section, beside the main explanatory variable which is the education level of household head, all variables which were chosen for the model will also be presented, and we will explain why those variables are important and should be focused The second part is about the general context of survey area, identify the sample size, how

to choose the sampling frame, and how to design a questionnaire for the research, this section also includes the limitations of the data we collected from rural An Giang province in 2000

3.1 MODEL SPECIFICATION

This research tries to explore the impacts of the education level ofhousehold head on household income To match the purpose of this research, we can apply both OLS and Tobit 20 models However, this research only aims to explore the impacts of education level of household head on household income, we thus do not want to examine the different effects of education level of household head on household income for the two poor and non-poor groups, and we do not intend to study effects of education level of household head on each source of the household income The OLS is therefore enough and suitable to match research objective, that is the reason why I decide to choose the OLS model for my study

After choosing the model, now I will explain selected variables which should be included in the model As stated in the Chapter 2, there are three groups of factors that affect household income

as capital, labor and price vectors In the two empirical studies in the Chapter 2, we can see that labor vector includes household schooling attainment and their age (for household head and the average of other members except the household head), household labor hours or labor days, years of working experience; Capital vector includes land cultivated, business assets, total value

of capital stock owned; Price vector includes the price of inputs (chemical, fertilizer, seeds) and that of the output

However, in order to meet the context of An Giang, in my study some variables should be dropped and other variables need be added, so the model is appropriate with the purpose of this research For example labor hours and labor days are not appropriate with the characteristics of

20 Also be called as 'limited dependent variable model' use for analysis two or more sub-groups (Gurazati, 1995)

this survey area where jobs are commonly temporary and self-employed, these variables should

be dropped Moreover, we should use productive assets instead of business assets, because the productive assets are more popular to the households that refer to all kinds of tools which can be used for earning money The price vector is not necessary to include, because of in An Giang province, price of output and inputs are stable 21 in 2000

Excluding the price vector, all of variables, which are included in my model, can be classified into labor vector and capital vector In which, capital vector only includes land-cultivated size and productive asset All of variables will be explained step by step in the following part

Table 3.1: Variable Presentation

coefficien tsign

('OOOVND)

characteristics

other household members

X7: Otherage Average age of other household Years ?

Source: classified from presented theories and evidences

21

Price stable according to local government and price indices of An Giang Committee's Report (2000)

25

Variable presentation

Dependent variable

• Log of total household income (ltotalinc)

The purpose of this research is to improve household income through education element

Therefore, the dependent variable is household income Chiswick (1997: p2) stated "Because of

positive skewness of earnings and the rise in earnings inequality as schooling level increases, by using the natural logarithm of earnings rather than earnings as the dependent variable the residual variance in the Human Capital Earning Function is less heteroskedastic and the

transform into logarithmic form 22 Moreover, by this transformation, we can examine the elasticity of education on income (one year of schooling affects how many percent of household income are)

Independent variable

• Number of years of schooling of household head (headedu)

As justified in the Chapter 2, numbers of year of schooling attendant has effects on total household income in general, especially for the household head case (Jolliffe, 1997; Nguyen,

1999 for instances)

• Age of household head (headage)

Vijverberg (1996) stated that one's age has a strong effect on household, both youngest and oldest members are less productive compared to members with the rank from 26 to 55

Grootaert (1995) found that in rural households, the household heads who are younger than 45 years old are more likely to increase the household income by their experience However, Emrys (1991) stated this experience can be off set by poorer health or reduced physical strength associated with age

Peter Matlon's (1981) showed that the household heads who are above 60 years old will begin

to disintegrate, and those who are under 25 years old may lack operating capital, adequate inherited land, and management experience

From the above discussions, the effect of household head age on household income is unclear

• Experience years of household head (headexp)

22 Analyzing your data (Haroon, 2001)

In our survey area, there is very weak institutional framework and hard to find any training programs for people, so the informal education has just been formed from the experience of doing work

For the farm household head cases, they influence on the agricultural productivity and efficiency

as mentioned in the previous section The older household heads may have more experience than the younger ones, and this experience will affect household income

For non-farm household head cases, as presented in the Chapter 2, according to Mincer (1974) years of experience affect individual income and then total income of the household

From the above discussions, we expect that household heads with more experience may get higher income

• Current working situation of household head (headwork)

The status of current working of the household head is very important to the household income because of his contribution This is the dummy variable defined that if household head is currently working, it is 1, otherwise it is 0

• Gender of household head (headgen)

Huou (1991) stated "Confucianism advocated that a family should have a head and the head

income from many types of heavy working jobs which the female household heads can not afford Male household heads are therefore expected to have stronger effect on household income in comparison with that of the female household head

• Average education level of other household members (otheredu)

As the benefits of education were mentioned, if all of members in household have higher education level they can earn more income compared to households with the lesser education level household Therefore, the expected sign is positive

• Average age of other household members (otherage)

While the expected sign of household head's age is unclear which was stated in the above section, the expected sign here can not be identified Therefore, it is also unclear relationship

• Household size (hhsize)

23 Cited from Desai (1995: pJO)

27

Nakajima (1986) stated that an increase in household size will lead to the increase in the family labor utilized, output and income This statement was supported by Desai (1995) that family labor source is important determinant in household income

Similarly, Moser (1998) argued that when there is a less developed labor market, the labor force

of the household depends on household size The large household size will reduce livelihood risks

Therefore, we expect that larger household size will help increase the household income

• Land cultivated size (landcult)

Norton (1993) said that the larger amounts of land cultivated would help to increase farm

income Moreover, "land is very clearly the most important natural asset of rural households"

(World Bank, 1999: p12) Obviously, the households with larger areas of cultivation will obtain more farm income from their output sold than the household with the limited areas of cultivation, and a larger scale of production helps them specialize in their production, and land size can help them overcome unemployment and under-employment in agricultural production Therefore, we expect that land cultivated size has a positive effect on household income

• Productive assets (Proasset)

Separating the productive assets from other assets is reasonable because it help the household in farm production or in conducting a business as well as service In our survey, the productive assets are, for instance, farm tools, water pump machine, boats, etc

It is accordance with likelihood approach of Ellis (2000) who advocated for income source

diversification of rural households Therefore, the higher productive asset value may lead to the

higher income the household earns

From the above discussions and the equation (2.4) we have:

Where:

LnY: Logarithm of total household income

And xi is variables shown in the table 3.2

ui: disturbance term

(3.1)

The OLS model can be interpreted after the heterocedasticity, multicollinearity, and normality tests

3.2 DATA COLLECTION

3.2.1 The study area

With the natural area of 3492 km2 equals to 1.03% of Vietnam, An Giang consists of 11 districts, and has many advantages in agriculture thanks to alluvium soils and aquaculture from rivers; however annual floods cause many difficulties for people life

The population of An Giang is 2,054,494 people (603 people/km2), one of the highest density province Average agricultural land per household is 0.67 acres or 1.615m2 per head 80,23% of the population live in rural areas and unevenly distributed (An Giang Report, 1999)

Agriculture took the largest share in provincial GDP, 43.61% and 47.59% in 1997 and 1998 respectively In which, crops are the most important source of agricultural outputs, accounting for over 70% total agricultural outputs from 1996 to1999 About crop structure, the major component

is paddy with 97 6% (Department of Planning and Investment of An Giang province, 2000)

An Giang has both average income per household and average income per capita higher than those of Mekong Delta 24 • The An Giang income per household and income per capita were VND 1,499,280 and VND 296,640 in comparison with those of Mekong Delta were VND 1,426,080 and VND 260,580, respectively (Own calculated from VLSS 1998)

An Giang has abundant young labor which accounted for 49.95% of the population, and people who are from 15 to 60 year olds accounted for 55% of the population (ASO, 1999)

The illiteracy rate was less than 10% and the average education level was 4.1 years of schooling attainment (VLSS, 1998) In which, 51.04% household heads under primary graduation, the rates for secondary and upper secondary school graduation are 5.21% and 6.25%, respectively But An Giang is facing with the decreasing of primary enrollment rate, from 1997 to 1998 this decreasing was 4.26% (ASO, 1999)

In short, An Giang province has many geographical and social advantages with abundant young labor force

24

Mekong Delta consists of 11 districts: An Giang Dong Thap, Long An, Tien Giang, Tra Vinh, Vinh Long, Can Tho, Bac Lieu, Kien Giang, Ca Mau, Ben Tre, Soc Trang

29