INTRODUCTION
Problem statement
The "Doi Moi" renovation process in Vietnam, initiated in 1986, has significantly transformed the country from a centrally planned economy to a vibrant market economy over the past two and a half decades Key reforms included the legalization of private economic activities and the removal of price controls on most goods and services These changes primarily impacted the rural sector, granting farmers greater freedom in production choices and gradually reducing price distortions As a result, agriculture has directly benefited the majority of Vietnam's population, whose livelihoods rely on small-scale agricultural self-sufficiency in rural areas.
Vietnam has transformed from a rice importer in the mid-eighties to the second largest rice exporter globally, achieving higher yields in rice and other crops without expanding the cultivated area or decreasing domestic consumption.
Since 2000, Vietnamese farmers have embraced new agricultural opportunities, leading to the cultivation of diverse crops like pepper and rubber As a result, Vietnam has emerged as the world's second-largest coffee producer, while the production and export of fruits and vegetables have seen significant increases This diversification has played a crucial role in boosting the country's income growth.
-= activities such as aquaculture, livestock, and non-farm activities with substantial tructural changes towards more industry and services
The significance of various income sources in rural income growth influences policy and public investment strategies When technological advancements that enhance yields are the primary drivers of rural income growth, prioritizing investments in agricultural research and extension becomes essential Conversely, if income growth is largely attributed to crop diversification, efforts should concentrate on improving agricultural credit, transportation, and market information to support this transition Additionally, if rising income or poverty levels are mainly linked to diversification into non-farm activities, the focus should shift towards training, electrification, and commercial credit to foster non-farm employment growth.
Non-farm activities in rural areas can generate more jobs and higher incomes compared to traditional agricultural activities, while also positively impacting agriculture by addressing market failures related to credit and insurance It is essential to recognize the interconnectedness of non-farm and agricultural sectors, as both contribute to the livelihood strategies of rural households through shared investments, production, and consumption decisions To enhance household income through non-farm employment, existing constraints must be addressed The significant shift towards non-farm incomes has led to a division among rural households, distinguishing between those who remain solely in farming and those who engage in mixed agriculture and non-farm activities This raises the question of whether market signals, despite their imperfections in transitional economies, favor those with inherent qualities that make them better farmers compared to those who diversify away from agriculture.
The objectives of study
This thesis aimed at investigating the role of rural non-farm activities on household income by analyzing the result of a Vietnam household living standard survey in
2008 More specifically, the objectives of the study are:
( 1) First, reviewing current status of diversification level of household income
(2) The second objective is to examine the role of non-farm income in increasing household income and how to increase non-farm income for household in Vietnam.
Research question
In order to research the importance of income diversification and the role of non-farm income in total income of household during 2002 - 2008, the following research questions are raised:
What is current status of household's income diversification?
Does income diversification affect household's income? What factors do affect household income diversification?
Structure of thesis
The thesis will begin with Chapter 1, the introduction will present In the next, a review of existing empirical research related to diversification and non-farm economy
Chapter 3 outlines the methodology, while Chapters 4 and 5 present the data and explain the statistical results The thesis concludes with a summary of the findings and final remarks.
LITERATURE REVIEW
Definitions
Income diversification encompasses various concepts, primarily focusing on the increase in the number of income sources and their balance within a household A household with multiple income streams is considered more diversified than one relying on a single source, particularly when income contributions are more evenly distributed Non-farm activities, which exclude primary agriculture, forestry, and fisheries, include trade and processing of agricultural products, as well as various sectors such as manufacturing, construction, and services The distinction between 'non-farm' and 'off-farm' is crucial; 'non-farm' refers to any income-generating activity outside of agriculture, while 'off-farm' specifically denotes activities conducted away from the household's own farm, often related to agricultural labor on others' land Understanding these definitions is essential for analyzing the economic rationale behind household income diversification strategies.
Factors effect to income diversification
Income diversification should not be confused with livelihood diversification, which involves households creating a varied portfolio of activities and social support systems to enhance living standards and mitigate risks While income generation is a key aspect of livelihood strategies, livelihood diversification also includes social institutions, gender relations, property rights, and other non-income support mechanisms essential for sustaining a living The motivations for diversification and the opportunities available vary significantly among different settings and income groups This leads to a crucial distinction between diversification aimed at accumulation, primarily influenced by "pull factors," and diversification focused on risk management, coping with shocks, or moving away from stagnating or declining agricultural practices.
Barrett et al (2001) identify two main categories of motives influencing household decisions: "push factors" and "pull factors." Push factors include risk reduction, responses to diminishing returns, and challenges such as land constraints and population pressure, which compel households to self-provision due to high transaction costs In contrast, pull factors involve the realization of strategic complementarities between activities, such as integrating crop and livestock production, as well as specialization based on comparative advantages derived from superior technologies and skills.
Households often engage in income diversification as a strategy for pre-risk management and to mitigate the effects of economic shocks (Reardon, Delgado, and Malton 1992; Reardon et al 1998) In developing countries, few households rely primarily on a single income source, as the literature on livelihood sustainability highlights the tendency to minimize dependence on one or two income streams amid economic uncertainty (Reardon 1997; Bryceson 1999; Ellis 2000) Various factors contribute to this observed income diversification at the household level, as noted by Barrett et al (2001).
(1) Self-insurance against risk in the context of missing insurance markets (e.g., Kinsey, Burger, and Gunning 1998);
(2) An ex post coping strategy (e.g., Reardon, Delgado, and Malton 1992), with extra individuals and extra jobs taken on to stem the decline in income;
(3) An inability to specialize due to incomplete input markets;
( 4) A way of diversifying consumption in areas with incomplete output markets;
(5) Simple aggregation effects where the returns to assets vary by individual or across time and space
In developing countries, rural households are increasingly diversifying into non-farm income sources, which now represent a significant portion of their total income An analysis by Reardon et al (1998) reveals that non-farm income accounts for an average of 42% in Africa, 40% in Latin America, and 32% in Asia Research in rural Africa shows a positive correlation between non-farm diversification and improved household welfare Consequently, promoting non-farm employment in rural areas has emerged as a widely supported policy recommendation among development agencies, including the World Bank and various non-governmental organizations (Delgado and Siamwalla 1999).
Household non-farm activities
3.1 Declining a share of agriculture in GDP and labor
The World Development Report (2008) highlights that economic development involves a continuous redefinition of the roles of agriculture, manufacturing, and services It identifies two key patterns in structural transformation: initially, low levels of development see agriculture accounting for significant shares of GDP (up to 50%) and employment (up to 85%), but these figures decline as countries progress This trend has been historically observed in developed nations and is currently evident in developing countries experiencing growth Despite agriculture's essential role, its contribution diminishes as countries advance; for instance, in Sub-Saharan Africa, agriculture still represents an average of 34% of GDP and 64% of employment In countries with GDP per capita between $400 and $1,800, particularly in Asia, agriculture averages 20% of GDP and employs 43% of the labor force, with these ratios further decreasing to 8% and 22% in more advanced economies.
The GDP per capita in various regions, particularly Eastern Europe and Latin America, ranges from $1,800 to $8,100 Incorporating both forward and backward linkages to agriculture, known as extended agriculture, can significantly boost its contribution to the economy, often by 50% or more, especially in middle-income countries.
The decline in agriculture's GDP share is attributed to the significant growth in the manufacturing sector, particularly in Malaysia and Thailand, where manufacturing's share doubled during the structural transformation process While Malaysia's service sector has been substantial, it has primarily supported rather than led this transformation, unlike India, where the service sector's GDP share rose from 42% to 52%, largely due to the information technology industry This shift in growth rates among sectors also altered the export structures of these countries, with agricultural exports decreasing as a proportion of total exports, while manufacturing exports experienced considerable growth.
Figure 2-1: Share of labor and GDP in agriculture
Share of labor and GOP in agriculture
• Share of labor in agriculture !1990-2005, average)
+ Smre of GOP from agriculture (1990-2005, averaoel -Trajectories of the share ot labor in agriculture 1961-2003
GOP per capita, constant2000US$ flog scale I
Solllt'e:WOR 2D08tnm band on dat~ from World Sank 200ov
Nôo: The list o!l-!4nor coon and tho coontr!qs thty rtpruent can bolound on P"'l• xviiL ource: World Development Report (2008)
Non-farm activities play a crucial role in enhancing household economies, particularly during agricultural off-seasons Research indicates that 5-65% of farmers engage in secondary employment within the non-farm sector, contributing 15-40% of total family labor hours to these income-generating activities As countries develop, there is a noticeable increase in the commercialization of non-farm tasks and employment opportunities The rising demand for non-farm products and services is closely linked to advancements in agricultural technology and management, leading to increased incomes for landowning households These households often reinvest their earnings into labor-intensive goods and services produced by small-scale non-farm enterprises Consequently, households involved in non-farm activities experience higher incomes, prompting them to expand their operations or encourage neighbors to participate The decision to allocate additional labor to the non-farm sector reflects a behavioral model of factor supply, emphasizing the household's aim to maximize earnings while navigating resource constraints and risk minimization Ultimately, the choice to diversify into non-farm activities involves five interrelated decisions that are made simultaneously.
(1) Non-farm participation: choice of farm sector activity or non-farm activity
(2) Level of non-farm activity
(3) Sectored choice: manufacturing or services
(4) Location: whether to undertake it locally or elsewhere
(5) Form: whether to undertake self-employment or wage-employment
In developing rural economies with limited agricultural employment opportunities, income diversification through non-farm activities is becoming increasingly important The distribution of land plays a crucial role in this shift, particularly in land-scarce, labor-rich countries like China and India, where inadequate access to land can push poorer households towards non-farm sectors Research by Adams (1995) in Pakistan and Chinn (1979) in Taiwan demonstrates that non-farm income can alleviate rural income inequality, as it tends to benefit poorer households more significantly Specifically, Adams (1995) found that non-farm income is inversely related to both the size of land owned and total rural income, highlighting its potential to improve the economic situation of the rural poor.
Research in Africa reveals contrasting findings regarding non-farm income and rural income distribution Studies by Collier et al (1986) in Tanzania and Matlon (1979) in Nigeria indicate that non-farm income adversely affects rural income distribution, primarily benefiting large landowners This suggests that in land-rich, labor-scarce African nations, widespread access to land may encourage the majority to remain in agriculture, with only wealthier households transitioning to the non-farm sector.
Empirical Literature
Research by Piesse, Simister, and Thirtle (1998) indicates that non-farm income sources contribute to increased income inequality in remote areas, while in regions with better access to urban markets, they help reduce income inequality In less connected rural areas, existing agricultural power structures enable those with higher farm incomes to take greater advantage of non-farm income opportunities Conversely, improved access to urban markets provides more equitable non-farm employment opportunities, diminishing the influence of these power structures on income distribution.
A study by Kinsey, Burger, and Gunning (1998) on 400 resettled households in rural Zimbabwe over 13 years reveals that income diversification serves as a coping strategy during droughts, primarily involving low-return activities like day jobs and agricultural piecework However, existing research on income diversification in Zimbabwe has limitations, notably the lack of emphasis on urban settings where poor households face similar risks as rural ones These studies often rely on the share of non-farm income as a proxy for diversification, which is challenging to measure and less relevant for urban-rural comparisons Similar findings from studies in Ethiopia and Tanzania by Dercon and Krishnan (1996), and in India by Micevska and Rahut (2008), indicate that household composition significantly influences diversification strategies Specifically, larger household sizes positively impact on-farm diversification, while a greater male labor force increases the likelihood of local off-farm diversification and migration, attributed to the benefits of scale in household chores and labor availability.
Credit constraints significantly influence household decisions regarding diversification, as limited access to credit greatly diminishes the likelihood of pursuing both farm and non-farm diversification strategies This finding aligns with existing empirical literature, which indicates that activity diversification often faces entry barriers, necessitating financial resources or credit access to fund the initial investments required for new ventures (Barrett et al., 2001; Abdulai & CroleRees, 2001; Woldenhanna & Oskam, 2001).
In Latin America, non-farm wage earnings typically surpass self-employment earnings, with countries like Brazil, Chile, Colombia, Mexico, and Nicaragua showing a significantly higher proportion of income from wage employment Conversely, in nations such as Ecuador, Honduras, and Peru, self-employment plays a more crucial role, especially in poorer regions These disparities are also evident within individual countries; for instance, research by Berdegue et al (2001) indicates that the share of wage employment in rural non-farm enterprises is considerably greater in more favorable areas compared to less advantageous ones.
A study from 2004 indicates that in northern Honduras, areas with better infrastructure and higher rural town density experience significantly higher non-farm wage income compared to self-employment income Conversely, in the southern region, where infrastructure is less developed and town density is lower, self-employment plays a more crucial role in income generation.
The non-farm economy remains a crucial focus for researchers and policymakers in advancing rural development strategies, as it plays a significant role in poverty reduction by providing alternative income sources Additionally, it can enhance agricultural growth by increasing productivity through a reduction in agricultural labor, thereby indirectly boosting family incomes Policies that encourage the non-farm economy can also help mitigate rural-to-urban migration, a pressing issue in many transition economies Studies across seven African households reveal that in four cases, including Botswana, Kenya, Malawi, and Zimbabwe, non-farm wage income is nearly twice as significant as self-employment, while the remaining cases, such as Rwanda, Ethiopia, and Sudan, indicate the opposite trend Generally, the share of non-farm earnings from wages tends to increase near urban areas, whereas part-time self-employment is more prevalent in remote rural regions.
Alain de Janvry, Elisabeth Sadoulet, and Nong Zhu (2005) utilized household survey data from Hubei province to analyze the potential impacts on rural households' incomes, poverty, and inequality without access to non-farm income sources Their findings indicate that the absence of non-farm employment would significantly increase rural poverty and income inequality Key factors such as education, proximity to towns, and neighborhood and village effects play a vital role in enabling households to access these non-farm opportunities Additionally, those who remain solely as farmers possess unobservable characteristics that enhance their agricultural productivity, suggesting a positive selection bias Furthermore, engaging in non-farm activities positively influences household farm production, underscoring the growing significance of the rural non-farm sector in developing countries.
Peter Lanjouw, Abusaleh Shariff, and Dil Bahadur Rahut (2007) highlight the growing importance of the non-farm sector in rural India's economy since the 1970s, noting a strong correlation between non-farm employment and agricultural wage rates Their analysis reveals that non-farm incomes constituted a significant portion of household income in the 1980s, with 6 out of every 10 new jobs created in the non-farm sector between 1993/4 and 2004/5 The most substantial growth in non-farm employment occurred from 1999/0 to 2004/5, though access to non-farm occupations varies significantly across different quintiles and states, influenced by factors such as education, wealth, caste, and local agricultural conditions While the non-farm sector has potential for poverty reduction, its direct impact is limited for the poor due to a lack of assets Additionally, certain non-farm sub-sectors are linked to increased agricultural wage rates, yet the literature offers little insight into how trade reforms and policies affect rural households' decisions to engage in non-farm activities.
Mukesh Eswaran, Ashok Kotwal, Bharat Ramaswami, and Wilima Wadhwa (2005) examine the effects of liberalization in the 1980s and 1990s on earnings and gender disparity in India Their findings indicate that the non-farm sector significantly contributes to educational attainment by creating job opportunities for literate individuals and younger generations, allowing them to transition out of agriculture While non-farm sectors do not directly benefit female employees, men’s employment in these sectors often leads to women taking over agricultural tasks Additionally, advancements in agricultural productivity through technical changes have been crucial in increasing agricultural wages, thereby enhancing women's earnings as productivity rises.
S Ranjan (2007) agrees that there are trends in the level and nature of employment in the rural non-farm sector The rise in male workers was larger than the rise in female workers and the manufacturing units in the non-farm sector continued to absorb the highest number of workers The demand-pull factors at work are the expansion of employment in sub-sectors-construction, trade-hotels, restaurants, transport and communications sectors hold promise of employment opportunities The expansion in hese sectors could be due to both the push and pull factors The gender wise istribution gives a clear impression of distress-driven employment increase The urvey revealed that although linkages between the farm and non-farm sectors in rural ndia were multifarious and strong, yet there were examples of a vibrant non-farm ector that was emerging without the support of the agricultural sector The scenario as a whole make a believer of the role of both the demand and distress -pull as well as external factors in generation of non-farm employment That most of the non-farm activities took place in the unorganized sector
T.Q Trung and N.T Tung (2008) using data from Vietnam Household Living Standards Survey in 1993, 1998, 2002 to analyze multiple indirect effects of trade liberalization on performance and business behaviors of non- farm household enterprises in the context of economic environment change during the transition period in Vietnam As focus on trade liberalization, they found that Vietnamese economy has experienced high economic growth rate but the total non-farm household enterprises income in the selected industries affected by trade liberalization increased not much The reason is the entry and exit rates of non-farm household enterprises are quite high in comparison with other international findings Vietnamese non-farm household enterprises also faced with many constraints in terms of low competition, differentiation and value added chain of products; weak marketing; poor and obsolete technology; weak entrepreneurial skills and low qualifications of non-farm entrepreneurs; insufficient business and market information; and shortage of capital and of skilled laborers, limited access to credit
Remco H Ostendorp, T.Q Trung, and N.T Tung (2009) identified non-farm household enterprises as a significant pull factor for income generation, income inequality reduction, and income volatility mitigation in Vietnam Their research revealed that while these enterprises enhance income and decrease between-household inequality, their influence has waned from 1993 to 2002 Consequently, they argue that untargeted policies promoting non-farm household enterprises are increasingly unjustifiable, advocating instead for targeted, export-oriented policies that address market failures This aligns with Thai Hung Pham's (2007) findings, which utilized data from the Vietnam Household Living Standards Survey from 1993, 1998, and 2002, employing a Multinomial Logit Model regression The study highlights the diversification of the rural labor force into non-farm employment, with the non-farm sector emerging as a crucial employment source outside agriculture Key individual-level factors influencing non-farm diversification include gender, ethnicity, and education, while land ownership negatively impacts non-farm employment due to a tendency to concentrate on agriculture Additionally, both physical and institutional infrastructures significantly affect participation in the non-farm sector.
RESEARCH METHODOLOGY
Model specification-dependent variable
Efforts to quantify income diversification have primarily focused on rural areas, estimating the proportion of non-farm income within total household income Studies suggest that a higher percentage of non-farm income correlates with increased diversification and reduced vulnerability to weather-related shocks, a significant risk in agriculture-dependent regions However, using non-farm income share as a measure of diversification presents challenges, such as treating households with varying sources of non-farm income equally, regardless of the number of income streams Additionally, accurately measuring this share requires detailed accounting of all income sources, making it a complex indicator Furthermore, this measure is less applicable in urban settings, where the majority of income sources are typically non-farm.
The Shannon equitability index serves as a straightforward measure of household diversification levels, highlighting the importance of multiple income sources to mitigate income risk, particularly in response to macroeconomic policies that may lead to job losses, as observed in Vietnam during the 1990s In Vietnam, income sources are unevenly distributed among households, with significant differences in livelihood strategies between urban and rural areas Rural households exhibit greater income diversification, with only 3.55% relying on a single income source, compared to 6.21% of urban households In 2008, over 65.39% of rural households reported having three or four income sources, while 48.60% of urban households had at least three Despite experiencing a decline in income diversification following economic shocks, rural areas faced a more substantial reduction in the number of income sources Notably, around 93% of households in both urban and rural regions received money transfers, primarily from pensions and domestic remittances, underscoring their significance in household income stability.
The criticism of using the number of income sources as a measure of diversification highlights several key issues Firstly, households with more economically active adults tend to have more income sources, which may reflect labor supply choices rather than a genuine pursuit of diversification This critique is addressed by analyzing per capita income sources and considering household demographics, such as age, sex, and education Secondly, a discrepancy arises when comparing households with varying income distributions from similar activities; for example, a household earning 99% of its income from farming is classified the same as one earning 50% from farming and 50% from wage labor unless adjustments are made To accurately reflect income diversification, the Shannon equitability index can be employed, as it effectively weighs different income shares and is simpler to calculate than assessing non-farm income shares.
Households that earn 50 percent of their income from farming and 50 percent from wage labor exhibit greater income diversification than those that rely on more than 50 percent of their income from farming The income diversification index, which ranges from zero to 100, reflects the percentage of actual income diversity compared to the maximum potential diversity To measure overall income diversity, the Shannon equitability index is employed, a method adapted from the Shannon index traditionally used to evaluate species diversity (Magurran, 1988).
The Shannon index of income (Hincome) evaluates both the number of income sources (S) and the distribution of income shares (incsharei) from each activity within a household This index is calculated for each household and reflects greater diversity as it increases Additionally, the Shannon equitability index (E) is derived from this index (H), highlighting the balance in income distribution among various sources.
The income diversification index measures how concentrated or scattered household income is across various sources, indicating the level of income diversification Households with a greater variety of income sources will have a higher index value (E), while those relying on a single income source will have the lowest value of 0 The index ranges from 0, representing the least diversified households, to 100, indicating households with equal income from four different activities A higher E value reflects either an increased number of income sources or a more balanced distribution of income shares.
Figure 3-1: Distribution of the diversification index
The diversification index, calculated using the Shannon equitability index method, measures the diversification level of rural households in Vietnam Between 2004 and 2008, there was only a slight change in the index, with values predominantly ranging from 30 to 60 This trend indicates that households are less inclined to diversify their income sources equally and are shifting from quantitative to qualitative income diversification By specializing in two to three activities, these households enhance their product competitiveness and increase their profitability.
According to FAO (1998), household decisions to engage in economic activities are influenced by two main categories of factors: those affecting the return and risk of agricultural production and those determining the capacity for non-farm participation, such as education and access to credit De Janvry, Sadoulet, and Zhu (2005) suggest that these factors are shaped by the household's physical and human capital endowments and their environmental context Key determinants include human and social capital, household composition, assets, and local institutions Specifically, human and social capital encompasses the age and education level of the household head and members, while household composition pertains to size Household assets are represented by land holdings per capita, and local factors include village density and distance to the provincial center Furthermore, based on de Janvry and Sadoulet (2001), individual decisions within households are interconnected, highlighting the complexity of economic participation.
3 Econometric Model ased on the above research and the data of Vietnam Living Standards Survey 2008 hich conducted by World Bank (WB) and the General Statistic Office of Vietnam, ualitative and quantitative analysis are applied; in which, qualitative analysis is used o describe current status of household income and role of non-farm income in total ousehold income; quantitative analysis is used to find which factors are most effect o diversification trend by using Two Stage regression First, probit regression for ousehold with 1 income resource (Diversification Index equal to 0) and more than 1 ind of income recourse (remaining household), then Least Square regression will be
• se for household with more than 1 kind of income resource, model for both stage is uggested as follow:
Diversification Index = f (Gender ; Age ; Age square ; Education ; Education
; Training ; Landholding per capita ; Household size ; Dependency ratio illage Density; Non-farm percent; Distance from urban center) ariable descri tion:
1 Gender: Dummy for gender of household head Using dummies for gender differences instead of estimating separate equations by gender in order to directly compare differences by gender rather than differences among men and women When household header is women, she is tendency stable income and do not like risk when invest in new activities Men normally will accept the risk and using family resource into other activities Since an economy that is composed of households which interact as collective units, rather than one in which individuals interact as purely independent agents, the differences among households as defined by the gender of their head can reveal a lot about different economic experiences
2 Age, Age square: age of household head Age has a differential impact on participation in agricultural and non-agricultural, which might potentially be explained by different physical fitness requirements across sectors Manual agricultural labor is often harder than work in other sectors, so that older people are at a disadvantage
3 Education: Number of years of schooling of the household head It has positive impacts on income While schooling does not seem to be important for agricultural wage laborers, it significantly increases the probability of finding work in non-agricultural sectors
4 Education Level: The average number of years of schooling of household members 15 years old and above Households with higher education level engage more in non-farm activities, and that human capital has an important effect on the level of non-farm income achieved
5 Number Education: Number of people in household had pass Lower Secondary school degree Higher people number, the family will have more income from wage and non-farm activities
6 Training: Dummy variable if member of household trained m non-farm activities It had same effect as Education in households
7 Landholding per capita, is the total areas of cultivated land used for agriculture production divided by total member of household, measured by square meters per person For a rural household, land is the main form of physical capital Larger per capita landholdings also equip a household better to engage in agriculture Lower landholding per capita, income from agriculture is not enough for household's expenditures It makes pressure in household budget and they tend to be doing anything in non-farm activities to get more income
8 Household size: The size of the household: land ownership might proxy wealth and contacts, and thereby provides some indication of the extent to which individuals are better placed to take advantage of opportunities in the non-farm sector
9 Dependency ratio: The percentage of family members engaged in cultivation activities, proxy a latent demand to diversify out of agriculture (and thereby reduce exposure to agriculturally related risk)
Econometric Model
The research utilizes data from the Vietnam Living Standards Survey 2008, conducted by the World Bank and the General Statistics Office of Vietnam, employing both qualitative and quantitative analyses Qualitative analysis describes the current status of household income and the significance of non-farm income within total household earnings In contrast, quantitative analysis identifies the key factors influencing diversification trends through a Two Stage regression approach This includes a probit regression for households with a single income source (Diversification Index equal to 0) and a Least Squares regression for those with multiple income sources.
• se for household with more than 1 kind of income resource, model for both stage is uggested as follow:
Diversification Index = f (Gender ; Age ; Age square ; Education ; Education
; Training ; Landholding per capita ; Household size ; Dependency ratio illage Density; Non-farm percent; Distance from urban center) ariable descri tion:
1 Gender: Dummy for gender of household head Using dummies for gender differences instead of estimating separate equations by gender in order to directly compare differences by gender rather than differences among men and women When household header is women, she is tendency stable income and do not like risk when invest in new activities Men normally will accept the risk and using family resource into other activities Since an economy that is composed of households which interact as collective units, rather than one in which individuals interact as purely independent agents, the differences among households as defined by the gender of their head can reveal a lot about different economic experiences
2 Age, Age square: age of household head Age has a differential impact on participation in agricultural and non-agricultural, which might potentially be explained by different physical fitness requirements across sectors Manual agricultural labor is often harder than work in other sectors, so that older people are at a disadvantage
3 Education: Number of years of schooling of the household head It has positive impacts on income While schooling does not seem to be important for agricultural wage laborers, it significantly increases the probability of finding work in non-agricultural sectors
4 Education Level: The average number of years of schooling of household members 15 years old and above Households with higher education level engage more in non-farm activities, and that human capital has an important effect on the level of non-farm income achieved
5 Number Education: Number of people in household had pass Lower Secondary school degree Higher people number, the family will have more income from wage and non-farm activities
6 Training: Dummy variable if member of household trained m non-farm activities It had same effect as Education in households
7 Landholding per capita, is the total areas of cultivated land used for agriculture production divided by total member of household, measured by square meters per person For a rural household, land is the main form of physical capital Larger per capita landholdings also equip a household better to engage in agriculture Lower landholding per capita, income from agriculture is not enough for household's expenditures It makes pressure in household budget and they tend to be doing anything in non-farm activities to get more income
8 Household size: The size of the household: land ownership might proxy wealth and contacts, and thereby provides some indication of the extent to which individuals are better placed to take advantage of opportunities in the non-farm sector
9 Dependency ratio: The percentage of family members engaged in cultivation activities, proxy a latent demand to diversify out of agriculture (and thereby reduce exposure to agriculturally related risk)
10 Village Density: The population density in the village (total village landholdings divided by the village population) a high population density would be expected to push People out of agriculture and may stimulate non- farm activities (through lower transactions costs, economies of agglomeration, etc.)
11 Non-farm Percentage of the labor force employed in non-farm activities: capture the strength of clustering of non- farm activities, and access to the specific infrastructure necessary to promote non-farm activities
12 Distance from urban center: measured distance (km) from the village that households are living to the nearest urban center.
DATA ANALYSIS AND DISCUSSIONS
Data description
The analysis of household income structure utilizes data from the 2008 Vietnam Household Living Standard Survey (VHLSS 2008), which includes 9,189 rural and urban households across eight regions of Vietnam Among these, 5,967 households provide community data The survey is categorized into eight distinct sections, offering comprehensive insights into the relationships between various household income sources, household assets, characteristics, and public assets.
Table 4-1: Structure offamily income in the 2008 survey
Share of total family income % of income
Total family income, Thousand VND per year 500,413,635
Per household income, Thousand VND per year 27,583
Table 4.1 outlines the "Structure of Family Income in the 2008 Survey," highlighting eight distinct income sources categorized into four mutually exclusive groups: agriculture income, wage income, non-farm income, and money transfer income Agriculture income, which encompasses five sources—planting, livestock, agricultural services, aquaculture, and forest hunting—can vary significantly depending on the type of crops cultivated or livestock raised Non-farm income includes both earned and unearned sources, such as remittances from family members working in cities, welfare subsidies, pensions, interest from money transfers, and income from rural non-farm activities Adhering to the national accounting system, non-farm income encompasses all secondary sectors like manufacturing and construction, as well as tertiary sectors such as transport and services, along with certain primary sub-sectors like mining.
( 1) Government employment - includes wages from all government and public sector service;
(2) Private sector- includes wages from private sector companies;
(3) Unskilled labor- includes wages from any unskilled non-farm activity, such as construction, brick-making and ditch digging;
Self-employment encompasses profits and earnings derived from various activities, including production trade, agriculture, forestry, and aquaculture It also includes retail operations and artisan services, such as tailoring, home appliance repair, and shoe repair.
(5) Other- includes property benefit, gift, remittances, welfare, pensions, interest
I come from (1), (2) and (3) are consider as wage income, (4) still keep name as non- f:~rm income and (5) is namely money transfer category So there are 4 different i 1come sources considered
According to data from the VHLSS 2008, agriculture remains a primary source of income for rural households, contributing to 56.70% of their total family income This agricultural income is comprised of 27.71% derived from various forms of agricultural activities.
~ gricultural (activity that includes revenue from sales of farm products and value of
In rural households, a significant portion of income comes from various sources: 3.49% is generated from aquaculture activities and other agricultural endeavors, including livestock, forestry, and hunting Notably, 20.07% of family income is derived from non-agricultural sources, while 28.67% is earned through salaries Additionally, money transfers play a crucial role in supporting many rural families.
Approximately 3.13% of households receive various transfers, including pensions, unemployment benefits, gifts, and other social benefits, which constitute 14.23% of total family income This significant percentage of recipients is largely influenced by the high proportion of seniors within these households.
~oldier among the rural population Although planting and livestock income contribute relatively little to total family income, they are relative large number of households
~ngage in these activities For most households, family income is quite diversified
Over 3.04% of households are classified as "non-diversifiers," relying solely on a single source of income In contrast, more than 75.81% of rural households benefit from a combination of non-farm or salaried income alongside agricultural earnings.
Descriptive statistics
Variable Obs Mean Std Dev Min Max age 6504 49.46817 13.72242 16 97 age2 6504 2635.376 1485.211 256 9409 edu 6504 1.337946 1.345635 0 12 edulev 6504 6.986716 2.966165 0 12 numedu 6504 0.808579 1.085226 0 6 gender 6504 0.202183 0.401659 0 1 dependency 6504 0.294682 0.316826 0 1 distance 6504 39.25969 35.70457 0 446 diver index 6504 41.94158 23.47788 0 99.85677 house size 6504 4.198493 1.681861 1 15 land_p_c 6504 1897.226 4340.304 0 126675 training 6504 0.164668 0.370909 0 1 village_ dens 6504 661.8885 749.9889 0.3461 15661.4
According to the VHLSS 2008 data, the average household size in the surveyed sample is four, aligning with the national average Approximately 24.49% of households are headed by women Adults in these households have an average of 6.9 years of schooling, surpassing the national average of 5.5 years (UNDP, 2010) The average age of household heads is 49.4, but they have an unexpectedly low average of only 1.3 years of schooling, indicating that many individuals lacked educational opportunities during and immediately after the Vietnam War Additionally, the average landholding per capita is 1,512 m², and households are located at a considerable distance from the nearest urban center.
On average, individuals in rural areas travel 49.2 km, taking about one hour to reach the market by motorcycle The annual per capita income is approximately 11.85 million VND, equivalent to around 650 USD, reflecting a reasonable average income level in Vietnam.
Household's income diversification
Table 4-3: Structure of employed population by kind of economic activity
Wholesale and retail trade; repair of motor vehicles, motor cycles and personal and household goods
Public administration and defence; compulsory social security
Other activity and money transfer
Real estate, renting and business activities
(Source: General Statistics Office of Vietnam, 2009)
As of 2009, Vietnam, the 13th most populous country in the world with nearly 86 million residents, has seen a shift in its economy since the mid-1980s when petroleum emerged as a significant sector The contribution of agriculture, livestock, forestry, and fishing to the GDP decreased from 27.76% in 1996 to 22.10% in 2008, with agriculture alone dropping from 24.17% to 18.14% Despite this decline, agriculture remains the primary source of employment and livelihood for nearly half of the Vietnamese population The 2008 Vietnam Household Living Standards Survey (VHLSS) indicates that typical rural families earn income from 2 to 3 different sources, including money transfers, highlighting that income diversification is positively correlated with higher household earnings However, families must navigate potential entry barriers and constraints, as identified in Lapar et al (2003), which affect various household types.
Figure 4-1: Family income and the number of family income sources
Figure 4-2: Number of households and number of income sources
2008 ource: VHLSS 2004, 2006, 2008(at constant price year2000)
This section examines income diversification across different sectors and time periods at the household level, utilizing per capita income quintiles to highlight variations among households in different income categories While per capita income is a common metric, expenditure is often considered a more accurate reflection of household welfare Households with significant debt and substantial current income may be mistakenly perceived as enjoying a high standard of living.
Households may experience low consumption levels due to income being allocated towards debt repayment Additionally, savings can enable families to maintain high consumption levels even during temporary declines in their current income.
I owever, culturally people in rural Vietnam tend to rely on subsistence behavior and tpeir expenditure does not reflect exactly their economic condition
Table 4-4: Trends in income diversification by the number of income sources
Mean SD Mean SD Mean SD
~easuring diversity in the number of income sources, table 4-4 displays the average
~umber of income sources of rural households is conditional on household per capita ncome-based quintiles n 2004, each household had 2.68 income sources, on average while the most income
~iversified households had 4 sources of income Over time, there is limited variation n the number of sources and between the end-points of the period under consideration
2004 - 2008), the size of increase in overall income sources is small (0.0 1 ), indicating
The analysis reveals a consistent increase in income diversification throughout the period examined Notably, the level of income diversity remains relatively uniform across different income quintiles For instance, households in the "poorest quintile" report an average of 2.41 income sources, while those in the "highest quintile" have an average of 2.56 income sources, indicating only a slight variation in income diversity across the spectrum.
"ghest for those in the "middle quintile" (2.80) in 2004 This indicates that the rich d the poor are not much different in terms of the level of diversity in income sources
2004 This cross-sectional pattern of diversity remains unchanged in the period
According to Minot et al (2006), income diversity among households exhibits an inverted U-shape, with those in the middle quintiles having the most diverse income sources, while households at both ends of the distribution are less diversified Data from Table X illustrates this trend, revealing that the richest quintile consistently shows the greatest variation in income sources, supporting the "pull-distress diversification" strategy Notably, between 2004 and 2008, there was an overall increase in income sources across all quintiles, with the most significant rise observed among wealthier households (0.06), while the poorer quintile experienced a decline (-0.03) This suggests that income diversification in Vietnam is increasingly characterized by a blend of demand-pull and distress-push strategies.
3.2 Measurement of income share diversity
Between 2004 and 2008, households showed notable trends in income diversification, as outlined in Table 4-5 In 2004, agricultural income was the primary source, constituting 37.30% of total household income, followed by wage employment at 28.05% and money transfers at 18.07% By 2008, the share of agricultural income increased to 40.82%, likely influenced by a significant rise in agricultural commodity prices, particularly rice, following a food price shock This increase of approximately 3.5 percentage points from 2004 indicates that fluctuating prices may have impacted household income sources, although it's unclear if these changes stem from direct price effects or household decision-making influenced by relative price signals Meanwhile, contributions from non-farm businesses and money transfers decreased, while wage income remained relatively stable.
The growth of non-farm income sources is steadily increasing each year, reflecting a greater diversity in income streams As noted in section 3.1, the share of money transfer income decreased to 15.77 percent in 2008, a drop of 2.55 percent, while non-farm income remained relatively stable at 15.5 percent, with a smaller decline of only 1.21 percent.
Table 4-5: Trends of income diversification, by income shares
In 2004, income distribution among quintiles revealed distinct patterns, particularly in non-farm business and wage employment income Households in the poorest quintile received the least income from non-farm activities, while the wealthiest quintile enjoyed the highest share, with non-farm income reaching 24.5 percent This trend highlights the significant disparity in income sources across different economic groups, emphasizing that the richest households rely more heavily on non-farm business income compared to their poorer counterparts.
Between 2004 and 2006, households across all income quintiles showed a decline in reliance on agricultural income, with the poorest households depending on it for 50.21% of their income, compared to 21.61% for the richest The movement away from agriculture varied, with the poorest quintile reducing their reliance by 3.26 percentage points, while the second and fourth quintiles saw reductions of 0.3 and 0.25 percentage points, respectively Overall, the trend indicates a shift towards non-farm income sources, reflecting a broader diversification of income strategies among households This diversification, driven by different motivations for rich and poor households, ultimately leads to a balanced approach in income generation from various sources.
Roles of non-farm activities in Vietnam's rural household economy
In Vietnam, the rural labor force is expanding quickly, yet employment opportunities are lagging behind Each year, approximately 74,000 hectares of farmland are converted for housing, industrial parks, and infrastructure, leading to an annual farmland loss rate of about 1% due to urbanization and climate change Consequently, the per capita availability of farmland is declining, with the average agricultural land area per person in Vietnam at just 1,224 square meters, significantly lower than the global average.
The reduction of farmland in Vietnam is significantly impacting social development, leading to an increase in landless farmers and social differentiation From 1990 to 2008, the share of agriculture and forestry in total employment fell from over two-thirds to approximately 48.87%, resulting in more than 6 million farmers losing their jobs and a high unemployment rate in rural areas As agricultural land becomes increasingly scarce, the expansion of non-farm employment is crucial to prevent deepening rural poverty Developing non-farm sectors is essential for boosting rural employment, fostering economic growth, improving income distribution, and alleviating poverty By creating job opportunities outside of agriculture, we can also mitigate rural-to-urban migration, which exacerbates urban social issues Given the limited capacity of urban industries to absorb a rapidly growing labor force, it is vital for labor-intensive rural non-farm sectors to take on excess labor, drive economic growth, and diversify income sources.
In 2008, over 36% of households received non-farm income, with 32.16% in rural areas and 49.87% in urban regions, marking a 3% increase since 2006 Households with non-farm income reported a per capita income of 15,838 thousand VND, which is 1.24 times higher than those without such income, who earned 12,717 thousand VND A significant disparity exists in farm income; households with non-farm income earned only 16,275 thousand VND annually, compared to 22,417 thousand VND for those without This suggests that non-farm income is not merely supplementary but a crucial source of financial support for families, compensating for insufficient farm income.
Table 4-6: Income of household with and without non-farm income
Average income from Without non-farm With activities Farm
The analysis of income structures for non-farm households indicates that they rely significantly on non-farm activities to supplement their earnings These households earn an average salary that is 5,788 thousand VND lower than those without non-farm income Additionally, non-farm self-employment contributes an average of 29,765 thousand VND to their overall income, highlighting the crucial role of non-farm activities in enhancing financial stability.
Table 4-7 indicates that non-farm activities primarily focus on providing supply services for local communities, with only about 9% dedicated to producing commercial products or handicrafts for external markets, such as wood and textile products The largest portion of non-farm activities, accounting for 33.63%, comes from retail sales and the repair of household items The remaining income is evenly distributed among sectors including hotels and restaurants, transportation via road, rail, and pipelines, as well as food and beverage production.
Table 4-7: Detailed non-farm activities of household
Types ofNon-farm activities Whole country Rural areas Retail sale, repair family applicants 33.63% 33.27%
Road, railroad and pipeline transport 7.93% 6.91%
Whole sale and agent sale 5.73% 6.20%
Wood processing and production of wood,
Sale services to local rural residents 3.60% 2.84%
The potential for families to earn non-farm income is influenced by the structure and quality of their human capital Research indicates that higher education levels among family members correlate with increased likelihood of employment in non-farm sectors.
Table 4-8: Status ofTraining and Education of household
Highest training degree Percent Highest learning degree Percent
Short-term Training 3.71% Primary School 37.70%
Professional High Upper Secondary school 3.94% school 21.29%
With 89.82% of household members do not attain any professional training and over 73.57% just attain secondary school or lower levels So, the way to get more income is wage and non-farm income Young people could move to urban areas and join industries and services and get wage income but middle-aged and old people are not able do so since they have low education level and lack of training in professional work, cannot change their normal life and they want to stay in rural area and work in non- farm activities The reduction in the importance of agricultural activities and the importance of non-farm activities is a main feature of economic development.
Econometric evidence
Table 5-1 presents the probit regression estimation results for the diversification index, revealing that human and social capital variables have minimal influence on household income diversification decisions Notably, the age of the household head significantly affects diversification choices, while the education level of the household head shows no significant impact on the decision to pursue additional activities.
The education level of household members plays a crucial role in diversification decisions, as investing in education opens up new opportunities for individuals to engage in activities different from those of the household head, leading to additional income sources such as wages, non-farm work, and remittances from urban employment Furthermore, household size positively influences diversification strategies, with each additional member increasing the likelihood of income diversification by 0.09.
Table 5-1: Probit Regression results of diversification index
Robust [95% none diver Coef Std Err z P>lzl Con f Interval] age 0.065174 0.011991 5.44 0 0.041673 0.088675
~istance -0.00175 0.000936 -1.87 0.061 -0.00359 8.04E-05 edu -0.02551 0.03272 -0.78 0.436 -0.08964 0.038621 edulev 0.032327 0.015563 2.08 0.038 0.001825 0.062829 numedu -0.0489 0.044223 -1.11 0.269 -0.13557 0.037778 lg_ender -0.06102 0.072228 -0.84 0.398 -0.20258 0.080548 dependency -0.21442 0.094159 -2.28 0.023 -0.39897 -0.02987 house size 0.106196 0.024204 4.39 0 0.058756 0.153635 land _p_ c -6.74E-06 4.72E-06 -1.43 0.154 -1.6E-05 2.52E-06 training 0.283806 0.106385 2.67 0.008 0.075294 0.492317 vi !!age dens 4.79E-05 6.17E-05 0.78 0.438 -7.3E-05 0.000169 region
Second, household with members has training in non-farm work is associated with a much higher probability of diversification index with the factor of 0.28 These results
~an be explained by increasing returns to scale in household chores for households with a larger size and more labor availability that makes it easier for them to let some
Research by Dercon and Krishnan (1996) in Ethiopia, along with Micevska and Rahut (2008) in India, indicates that members' engagement in various activities is influenced by household dynamics Specifically, the presence of older members significantly decreases the likelihood of households participating in migration, as a higher dependency ratio limits labor availability for such activities Furthermore, households with more arable land per adult are more likely to diversify their agricultural production Generally, decisions regarding local non-farm activities are primarily driven by the household's asset position, rather than by human or social capital or household composition Detailed results from a linear regression analysis on the diversification index can be found in Table 5-2 (refer to the appendix for further information).
Table 5-2: Liner Regression results of diversification index
Robust [95% diver index Coef Std Err t P>lti Con f Interval] age 0.496044 0.132329 3.75 0 0.236635 0.755454 age2 -0.0039 0.001229 -3.18 0.002 -0.00631 -0.00149 distance -0.03639 0.00843 -4.32 0 -0.05292 -0.01986 edu -0.28528 0.306865 -0.93 0.353 -0.88684 0.316281 edulev 0.682239 0.155369 4.39 0 0.377663 0.986816 numedu 0.237208 0.349843 0.68 0.498 -0.4486 0.92302 gender 0.407125 0.720861 0.56 0.572 -1.00601 1.820259 dependency -2.10355 0.907021 -2.32 0.02 -3.88162 -0.32548 house size 0.393301 0.183138 2.15 0.032 0.034288 0.752315 land p c -0.00062 0.00012 -5.15 0 -0.00085 -0.00038 training 4.976478 0.791587 6.29 0 3.424696 6.52826 village dens 0.000159 0.000419 0.38 0.705 -0.00066 0.000981 regiOn
(1) Demographic factors: Household s1ze has a positive effect on the diversification index The larger household size, diversify trend of household
The gender of the household head does not significantly impact the diversification index, making it difficult to draw conclusions about its effects In contrast, the age of the household head plays a significant role; older heads tend to possess greater life experience, which can lead to increased income resources for the household.
Education significantly influences household diversification, with higher education levels correlating to increased diversification trends Studies by Corral and Reardon (2001), Yunez and Taylor (2001), and de Janvry and Sadoulet (2001) indicate that more educated households tend to earn higher overall incomes, although not necessarily more farm income The positive impact of education on total income is substantial, as individuals with higher education often relocate to urban areas for better job opportunities or engage in non-farm activities, thereby increasing their earnings Datt and Jolliffe (2005) highlight education as a crucial determinant of household living standards in both rural and urban settings, noting a substitutability between education and land ownership Their findings also reveal that adult education positively affects household welfare across various contexts.
(2003) find that education is the factor that mostly affects households' escape from poverty
Household assets and community variables significantly influence agricultural diversification A very small per capita landholding tends to reduce the diversification index, while increased village population density enhances it This correlation is logical, as higher density leads to limited agricultural land, prompting villagers to seek alternative income sources Additionally, although the distance from the nearest urban center does not show a significant impact, it suggests that households farther from urban areas tend to experience lower income levels.